v ■ o\ *n

HARVARD UNIVERSITY

LIBRARY

OF THE

Museum of Comparative Zoology

The Library -'
Museum of Comparative Zoology
Harvard University

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

MUS. CO MP. ZOOS
LIBRARY

FEB 25 1972

HARVARD
UNIVERSITY

REVISION OF THE BEE GENUS AGAPOSTEMON
(HYMENOPTERA: HALICTIDAE)

By

Radclyffe B. Roberts

Vol. XLIX Pages 437-590 Feb. 16, 1972 No. 9

S-N4- Lcatuv

ANNOUNCEMENT

The University of Kansas Science Bulletin (continuation of the Kansas Uni-
versity Quarterly) is an outlet for scholarly scientific investigations carried out
at the University of Kansas or by University faculty and students. Since its
inception, volumes of the Bulletin have been variously issued as single bound
volumes, as two or three multi-paper parts or as series of individual papers.
In each case, issuance is at irregular intervals, with each volume approximately
1000 pages in length.

The supply of all volumes of the Kansas University Quarterly is now ex-
hausted. However, most volumes of the University of Kansas Science Bulletin
are still available and are offered, in exchange for similar publications, to learned
societies, colleges and universities and other institutions, or may be purchased
at $15.00 per volume. Where some of these volumes were issued in parts, in-
dividual parts are priced at the rate of 1% cents per page. Current policy, ini-
tiated with volume 46, is to issue individual papers as published. Such separata
may likewise be purchased individually at the rate of 1H cents per page. Sub-
scriptions for forthcoming volumes may be entered at the rate of $15.00 per
volume. All communications regarding exchanges, sales and subscriptions should
be addressed to the Exchange Librarian, University of Kansas Libraries,
Lawrence, Kansas 66044.

Reprints of individual papers for personal use by investigators are available
gratis for most recent and many older issues of the Bulletin. Such requests should
be directed to the author.

Editor
Charles R. Wyttenbach

Editorial Board

Kenneth B. Armitage
Richard F. Johnston

Paul A. Kitos

Charles D. Michener

Delbert M. Shankel

George W. Byers, Chairman

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLIX Pages 437-590 Feb. 16, 1972 No. 9

TABLE OF CONTENTS

ABSTRACT 439

INTRODUCTION 439

ACKNOWLEDGMENTS 440

HISTORICAL REVIEW 441

EXCLUDED SPECIES 442

GENERIC DIAGNOSIS 444

INFRAGENERIC STRUCTURE 444

ZOOGEOGRAPHY 446

METHODS 449

KEY TO SPECIES

United States and Canada 451

Meso America 453

West Indies 456

South America 458

DESCRIPTIONS, SYNONYMIES AND DISTRIBUTIONS

aenigma 460

alayoi 461

angelicus 461

ascius 463

atrocaerulens 464

boliviensis 467

centratus 469

chapadensis 470

coc\erelli 471

coloradinus 473

columbi 476

cubensis 478

cyaneus 479

erebus 480

femoratus 482

heterurus 483

hispaniolicus 484

inca 484

insularis 485

intermedins 486

jamaicensis 490

kphliellus 491

lanosus 494

leunculus 495

melliventris 499

mexicanus 503

mourei 504

nasutus 508

obscuratus 513

ochromops 513

peninsularis 515

poeyi 517

radiatus 520

rhopalocera 523

sapphirinus 524

semimelleus 525

splendens 529

swainsonae 531

texanus 533

tyleri 542

viequesensis 544

virescens 547

viridulus 550

NOMINA DUBIA 554

APPENDIX A 555

LITERATURE CITED 556

FIGURES 559

INDEX TO INCLUDED TAXA 588

Revision of the Bee Genus Agapostemon
(Hymenoptera: Halictidae)

1.2

Radclyffe B. Roberts3

ABSTRACT

In this study 43 species of Agapostemon are recognized from the Americas.
Of these, the following 19 species are new: A. aenigma, A. alayoi, A. ascius,
A. boliviensis, A. columbi, A. cubensis, A. cyaneus, A. erebus, A. hispaniolicus,
A. inca, A. insularis, A. intermedins, A. jamaicensis, A. lanosus, A. mexicanus,
A. mourei, A. ochromops, A. peninsnlaris, A. sapphirmus. Forty-six names are
listed as synonyms, 26 for the first time. Twenty-two species previously placed
in Agapostemon but now placed in other genera are listed.

Separate keys are provided for species from America north of Mexico, Meso-
America, The West Indies and South America. All species are described and
most are illustrated. The geographic distribution and variation of the species
and species groups are discussed, with particular regard to speciation of North
American groups.

INTRODUCTION

The genus Agapostemon occurs only in the Western Hemisphere where
it ranges from southern Canada to Paraguay. It is the only member of a
group of allied genera to be found north of Mexico — South America being
the center of abundance of most of its relatives.

Agapostemon is polythetic and cannot be differentiated from related
genera on the basis of any one character or group of characters. Neverthe-
less, most species are characterized by their metallic green or blue head and
mesosoma and their contrasting black or black and yellow metasoma.

Until recently, the biology of this genus was poorly known. However, in
1969 Eickwort and Eickwort described in detail the nesting and foraging
behavior of A. nastitus in Central America. In the same year I reported on
the biology of the North American A. radiatus, A. splendens and A. texanus
and summarized available information on other species. These biological
investigations have complemented the present classification.

The primary objective of this revision is to describe and classify the
species of Agapostemon and to elucidate their evolutionary history. It is
possible to recognize species groups and in some cases provide reasonable

1 Contribution number 1452 from the Department of Entomology, University of Kansas,
Lawrence, Kansas.

"This study was supported by National Science Foundation Grant GB 91 to the University
of Kansas (C. D. Michener, principal investigator).

3 Present address: Department of Entomology, Oregon State University, Corvallis, Oregon
97330.

440 The University Science Bulletin

explanations of their origins, but the number of species with no apparent
affinities proscribes extensive speculation on the phylogeny of the genus.

A second objective of this work is to facilitate identification of species
of Agapostemon. To this end keys, descriptions and illustrations have been
prepared in as simple and uniform a style as possible. Species descriptions
are in alphabetical sequence, because I believe a phenetic or "phylogenetic"
sequence presumes too much knowledge on the part of the readers. The
keys are regional and artificial for the sake of brevity and clarity respectively.

ACKNOWLEDGMENTS

I would like to express my gratitude to Dr. C. D. Michener for pointing out
the need for this study; for examining specimens in Washington, London, Genoa,
and in Pretoria, South Africa; and for his interest, encouragement, and guidance
during the course of this study. Special thanks are also extended to Dr. G. W.
Byers for his very helpful advice and careful editing of the manuscript.

The taxonomic portion of this study was made possible by the efforts of the
curators, graduate students and others who selected and loaned nearly 50,000
specimens from the following collections: Academy of Natural Sciences of
Philadelphia; American Museum of Natural History; Brigham Young Univer-
sity; British Museum (Natural History); California Department of Agriculture
(Sacramento); Canadian National Collection; Carnegie Museum; Chicago
Natural History Museum; Colorado State University; Cornell University; Florida
State Collection of Arthropods, Gainesville; Illinois Natural History Survey;
Iowa State University; Instituto e Museo di Zoologia Universita di Torino;
Kansas State University; Los Angeles County Museum; Michigan State Univer-
sity; Milwaukee Public Museum; Montana State College; Museo Civico di Storia
Naturale (Genoa); Museum National d'Histoire Naturelle (Paris); Museum of
Comparative Zoology; Naturhistorisches Museum (Vienna); North Carolina
State of the University of North Carolina; Northwestern State College (Louisi-
ana); Ohio State University; Oklahoma State University; Oregon State Uni-
versity; Pennsylvania State University; Purdue University; Riksmuseum van
Natuurlijke Historie (Amsterdam); Rutgers, The State University; San Jose
State College; South Dakota State College; Stanford University; Transvaal Mu-
seum (Pretoria, South Africa); United States Department of Agriculture, Wild
Bee Pollination Investigations (Logan, Utah); United States National Museum;
University of Arizona; University of Arkansas; University of California at
Berkeley, Davis, and Riverside; University of Colorado; University of Georgia;
University of Idaho; University of Kansas; University of Louisville (Kentucky);
University of Michigan; University of Minnesota; University of Missouri; Uni-
versity of Nebraska; University of Nevada; University of North Dakota; Univer-
sity of Tennessee; University of Wisconsin; University of Wyoming; Utah State
University; Washington State University; Zoologisches Museum der Humboldt-
Universitat (Berlin).

The following individuals graciously loaned specimens from their personal
collections: Pastor Alayo D., Havana, Cuba; the late R. R. Dreisbach (collection

Revision of the Bee Genus Agapostemon 441

now at Michigan State University); R. A. Morse, Cornell University; D. W.
Ribble, University of Wyoming; and G. I. Stage, United States National Museum.
Through their cooperation and hospitality the following individuals did much
to make my visits to museums both profitable and pleasant: P. H. Arnaud, Jr.,
California Academy of Sciences; G. E. Bohart, U.S.D.A., Logan, Utah; K. V.
Krombein and G. I. Stage, U.S. National Museum; J. G. Rozen, Jr., American
Museum of Natural History; the late J. A. G. Rehn and the late H. J. Grant,
Academy of Natural Sciences of Philadelphia.

Finally, I am grateful to my wife, Guinnevere, whose multifarious talents as
collector, observer, secretary, editor and counselor have contributed immeasurably
to the completion of this study.

HISTORICAL REVIEW

The taxon Agapostemon was first proposed in 1844 by F. E. Guerin-
Meneville as a subgenus of Andrena in the following statements:

Nous connaissons plusieurs especies a cuisses ainsi renflees. Ce
sont des males. Peut-etre jugera-t-on a propos de les reunir en un
sous-genre, que nous proposerions de nommer Agapostemon. II
serait aux Andrenes ce qu'est le genre Nomia parmi les Halictes.

The only included species was Andrena (Agapostemon) femoralis
Guerin 1844. Dalla Torre (1896) correctly recognized this species to be a
junior synonym of Apis viridula Fabricius 1793. Therefore the correct name
for the type species of the genus Agapostemon is now Agapostemon viridulus
(Fabricius). Agapostemon was first described and accorded generic rank
in 1853 by Frederick Smith, who listed seven species, four of them new.
In 1896 Dalla Torre catalogued 16 species and placed four species in
synonymy. He also gave the Latin translation of Guerin's transliterated
Greek compound, Agapostemon, as "aya7ra<o amo, aT-q/jLon* stemen" In
English the translation is "lover of stamens."

In 1897 Robertson redescribed the species known from the United States
and listed their synonyms. The first key, published by Titus in 1901, included
only the species known to occur in Colorado, and in November of the same
year Crawford revised the North American species of Agapostemon. In
this revision Crawford re-described the genus, described seven new species,
and included a key to the 15 species then known from North America. In
1902 Robertson published keys to the genera and species of the North
American Halictinae in which he commented on the relationships of the
genera. Although he cited very few characters, the keys and generic con-
cepts were remarkably good.

In 1903 Vachal reduced Agapostemon to subgeneric rank in Halictus.
He included 28 species (nine of them new) in the subgenus Agapostemon
and proposed the new subgenus Paragapostemon for 25 additional species

442 The University Science Bulletin

with hairy eyes and without a complete propodeal carina. Although his
keys were good, Vachal's conservative generic concepts were not widely
accepted, and Cockerell (1905) accorded generic rank to both Agapostemon
and Paragapostemon. Vachal did not cite a type species for Paragapostemon,
but Cockerell (1905, in footnote) designated Halictus (Paragapostemon)
podager Vachal as the type species.

Schrottky (1909a) erected the genus Pseudagapostemon (type species
Agapostemon arenarius Schrottky), thereby removing most of the species
from Paragapostemon as well as additional species from Agapostemon.
Schrottky published another paper (1909b) in which he reduced seven
species of South American Agapostemon to synonymy and transferred four
more from Agapostemon to Pseudagapostemon. In 1918(a) Cockerell re-
duced Pseudagapostemon to subgeneric rank in Agapostemon, but never re-
ferred to it as such thereafter.

In 1936 Sandhouse revised the species of Agapostemon occurring in the
United States. Although she examined about 4,000 specimens, her work
does not make sufficient allowance for intraspecific variation, none of the
species are described, the only figures (of genitalia) are inadequate, and
finally, she did not recognize five of the species now known from the
United States; A. femoratus, A. tyleri, A. nasutus, A. peninsularis and A.
leunculus. Despite these shortcomings, Sandhouse's revision has remained
the best available reference on the genus for more than thirty years. More
recent systematic treatments of Agapostemon are the contribution by
Michener (1951) to the catalog of North American Hymenoptera; and the
key, descriptions and figures for the species in the eastern United States
(Mitchell, 1960).

EXCLUDED SPECIES

The following is a list of species which, although described in or sub-
sequently transferred to Agapostemon, are no longer considered to belong
to this genus. When possible the genus to which they belong has been
indicated. Some of the species are synonyms but listing of synonymies must
be postponed until Paragapostemon and Pseudagapostemon can be revised.

In 1918(a) Cockerell reduced Pseudagapostemon to a subgenus of
Agapostemon but this classification has never won acceptance. The only
species mentioned by Cockerell [Agapostemon (Pseudagapostemon)
xanthorhinus Cockerell, Halictus citricornis Vachal, Pseudagapostemon
paulista Schrottky, and Pseudagapostemon nasua Schrottky] have been
omitted from the following list.

Augochloropsis

Agapostemon caeruleus Ashmead 1890 was placed in Augochloropsis by
Titus (1901).

Revision of the Bee Genus Agapostemon 443

Paragapostemon

Notnia caelestina Westwood 1875, placed in Agapostemon by Cockerell
(1910a), was placed in Paragapostemon by Moure (1964).

Agapostemon bruneri Crawford 1901 was placed in Paragapostemon
by Moure (1964).

Halictus (Agapostemon) sicheli Vachal 1901 was placed in Paraga-
postemon by Moure (1964).

Nomia tacita Cameron 1902, placed in Agapostemon by Cockerell
(1910a), was placed in Paragapostemon by Moure (1964).

Nomia cillaba Cameron 1902, placed in Agapostemon by Cockerell
(1910a), was placed in Paragapostemon by Moure (1964).

Rttizantheda

Halictus emarginatus Spinola 1851, placed in Agapostemon by Cockerell
(1905), is a synonym of the type species (Halictus proximus Spinola
1851) o£ Rttizantheda Moure 1964.

Halictus mtttabilis Spinola 1851, placed in Agapostemon by Schrottky
(1903), became the type species of Rttizantheda (Ruizanthedella)
Moure 1964.

Halictus placid us Smith 1879 was placed in Agapostemon by Cockerell
(1905), but Moure (in litt.) places it in Rttizantheda.
Pseudagapostemon

Agapostemon arenaritts Schrottky 1902(b) became the type species of
Pseudagapostemon Schrottky 1909(a).

Agapostemon aeneus Schrottky 1902(a) was placed in Pseudagapostemon
by Schrottky (1909b).

Agapostemon arechavaletae Schrottky 1908 was placed in Pseudaga-
postemon by Schrottky (1909b).

Agapostemon bonaerensis Schrottky 1908 was placed in Pseudaga-
postemon by Schrottky (1909b).

Halictus (Agapostemon) pissisi Vachal 1903 was placed in Pseudaga-
postemon by Vachal (1904).

Halictus (Agapostemon) divaricattts Vachal 1903 was placed in Pseud-
agapostemon by Vachal (1904).

Agapostemon olivaceo-splendens Strand 1910 was placed in Pseudaga-
postemon by Moure (1947).

Agapostemon zosteronedys Moure 1940 belongs in Pseudagapostemon
divaricattts (Vachal) as indicated by Moure (footnote in Michener
and Lange, 1958).
Species Incertae Sedis

Halictus bruchianus Schrottky 1908 was placed in Agapostemon by
Schrottky (1913). Moure (in litt.) has not seen the type but believes
it should be placed in Corynttra or Rttizantheda (probably the latter).

444 The University Science Bulletin

GENERIC DIAGNOSIS

No one character or set of characters was found to be both necessary
and sufficient to distinguish all species of Agapostemon from all species in
other halictine genera. However, the genus may be recognized by a syn-
drome of characters. This set of characters is common to most species of
Agapostemon but not all of its characters are present in all of the species.

The character which is unique to Agapostemon and which best separates
this genus from other halictine genera is the carina which in most species
entirely surrounds the posterior surface of the propodeum of both sexes.
Also unique among halictines is the striking contrast between the non-
metallic coloration of the metasoma and the metallic coloration of the head
and mesosoma of most female and nearly all male Agapostemon.

Female Agapostemon may be distinguished from females of most other
halictine genera by the three or four (sometimes as many as seven) large
spatulate teeth on the posterior hind tibial spur and by the parallel con-
tiguous carinae extending postero-dorsally from the antero-ventral margin
of the gena. Male Agapostemon may be distinguished from males of the
augochlorine genera and the genera in the Halictus-Lasioglossum group by
the fusion of the first two tarsomeres of the hind tarsus. Unlike those of
almost all other halictine genera, the hind femora and, to a lesser extent,
the hind tibiae of many male Agapostemon are conspicuously swollen. The
only other genera with similarly modified legs have conspicuously hairy
eyes in contrast to the glabrous or nearly glabrous eyes of Agapostemon.

INFRAGENERIC STRUCTURE

While it seems premature to speculate at length on the phylogeny of
the species of Agapostemon, certain discrete groups of species can be
recognized on the basis of such features as genitalia, pronotum, legs,
metasomal sterna and color pattern. These species groups might have been
recognized as subgenera, but to be consistent I would have been forced to
recognize an unacceptable number of small or monotypic subgenera. Sub-
species are not recognized because there is little biological or utilitarian
justification for such in this genus. Of course, intra-specific variation is
described wherever encountered.

In the following list the name of the first member (chosen arbitrarily)
of each species group has not been indented and will serve to identify the
group in subsequent discussion (e.g., the species in the splendens group
are A. splendens, A. texanus and A. angelicas).

NORTH
AMERICAN

SOUTH
AMERICAN

(except atrocaenileus)

I{ohliellus

centratus

poeyi
insularis
jamaicensis
vieqitesensis
colit m bi
ochromops
sapphirinus
cyan ens
aenigma

virididus
obscuratus
hispaniolicits
c u ben sis
alayoi
swainsonae

splendens
texanus
angelicus

radiatus*
cockerelli
femoratus*

virescens*
tyleri
color -ad in us*

melliventris
peninsularis
mexicanus

rhopiilocera
erebus

as cms
leunculus

n as ut us

atrocaenileus
semimelleus*
chapadensis*

intermedins

heterurus
inca
mourei
boliviensis
lanosus

WEST
INDIAN

MESO-AMERICAN
(except those
with asterisk)

446 The University Science Bulletin

ZOOGEOGRAPHY

Although more than 50,000 specimens were examined in the course of
this study, I was hampered by inadequate collections from the West Indies
as well as Central and South America. While it seems unlikely that any new
North American species will be discovered, it is likely that new species
remain to be found in the West Indies, Central America and the Andean
region of South America.

It is difficult to comprehend what factors influence the distribution of
species of Agapostemon. Only in the correlation between the distribution
of A. splendens and that of moist sandy soils in the eastern United States
and eastern Mexico is there evidence of a causal relation between the
presence of an ecological parameter and the presence of the bees (Roberts,
1969).

Some species, such as A. coloradinus, are relatively restricted in range
while others are widespread, A. texanus occurring from Canada to Costa
Rica and Boston to San Francisco. In the United States alone, A. texanus
occurs in 70 of 116 plant communities (Appendix A). A. angelicas occurs
from below sea level in Death Valley, California to 12,000 ft. (3,658 m) on
Mt. Evans in Colorado (timberline 11,700 ft.). In view of the apparent
ecological plasticity among most species, it is not surprising that their dis-
tributions broadly overlap.

North American Species. There are four North American species
groups. Each of these groups is composed of two very similar species and
a third species which, although obviously closely related, stands somewhat
apart phenetically. In each group but one, the two most similar species are
found west of the 95th meridian and the third species is found north and
east of the others. There is some sympatry within each species group, but
the center of distribution is different for each species within a group (Fig.
228).

The simplest explanation for the occurrence of these similar geographic
and cladistic patterns in each of the North American species groups is to
assume: (1) North America was originally occupied by four species; (2)
each species was bisected into southeastern and southwestern populations
at the time of Pleistocene glaciation; and (3) the southwestern populations
were more recently subdivided (possibly on the Pacific and Gulf coasts of
Mexico) when forced to migrate farther south by still further cooling.
Presumably this pattern of evolution is not evident among the Meso-
American and West Indian species groups because they were too far south
to have been significantly displaced by the climatic changes accompanying
Pleistocene glaciation.

West Indian Species. The West Indian species are interesting in that
they are not found on the mainland (with the doubtful exception of A.

Revision of the Bee Genus Agapostemon 447

aenigma) and are not closely related to the mainland species. This is
somewhat surprising in view of their relative proximity to the Florida and
Yucatan peninsulas. Also surprising is their absence in the lesser Antilles;
yet their occurrence throughout the Bahama Islands and Greater Antilles
constitutes prima-facie evidence of high vagility.

Although it is extremely improbable that the bees fly of their own
accord between islands, it does seem likely that they are occasionally swept
up by hurricanes and deposited on other islands many miles distant. Mem-
bers of the poeyi and viridulus groups are widespread (cf. map, Fig. 1).
As there is very little discernible correlation of geographic factors with the
phenotype among members of each of these phenetically homogeneous
species groups, I am inclined to believe that, relative to their rates of evolu-
tion, their rates of dispersal have been rapid.

In studying the poeyi group I vacillated between considering all of the
island populations as conspecific and considering each population as a
separate species. A. poeyi and A. viequesensis, occurring in Cuba and Puerto
Rico respectively, have long been considered as distinct species which could
easily be differentiated morphologically. However, as morphologically in-
termediate forms exist on the Bahama Islands, it is tempting to classify the
entire group as a single highly variable species. However, phenetically
typical populations of A. poeyi and A. viequesensis are sympatric on New
Providence Island, and there is no evidence of hybridization. The phenetic
homogeneity within each of the phenetically different populations of this
complex on other islands leads me to believe that gene flow between these
allopatric populations is also inconsequential. Thus it seems that in the
poeyi group the rate of speciation, or creation of clades, is high relative to
the rate of evolution, or shift in gene frequency. Although not as large
and widespread, the viridulus group similarly appears to be speciating rela-
tively rapidly. My decision to regard the A. poeyi group as an Arten\reis
rather than a Rassen/^reis is based on scanty evidence. However, the de-
cision is a taxonomic necessity which in no way effects the biological at-
tributes of the organisms in question.

The species A. l{ohliellus and A. centratus do not seem closely related to
any other species of Agapostemon. Rare species, they may be restricted to
a particular ecological factor found only on the island of Hispaniola. Such
a severe ecological limitation could explain their absence on the other islands.

Meso-American Species. The distributions and relationships of the
species occurring in this region are difficult to interpret. Some of the species
such as A. mexicanus or A. tyleri clearly belong to North American groups
although A. mexicanus is found in northwestern Mexico and A. tyleri
ranges southward on the central plateau to the vicinity of Mexico City.

A. rhopalocera is a rare species known only from males, not obviously

448 The University Science Bulletin

related to other species and restricted to the vicinity of Mount Orizaba in
Vera Cruz. A. erebns is known from a single female collected in El Salva-
dor, but is obviously closely related to A. leunculus which ranges from the
southern tip of Texas to eastern Ecuador, and also to a third species, A.
ascitis, known from two females collected in Colombia.

Males of A. nasutus are unlike those of any other species inasmuch as
their genitalia and clypeal region are both highly modified. They also differ
from males of other species in less striking features such as the lateral tufts
of pubescence on the 6th metasomal sternum. Despite these obvious peculi-
arities of males, females of A. nasutus are somewhat similar to those of the
erebus group (differing principally in the shape of the pronotum). The
commonest species throughout most of its range, A. nasutus is found from
the southern tip of Texas to northern South America, as far east as Trinidad,
and west of the Andes as far south as lea, Peru.

A. atrocaeruleus has been found only in Costa Rica but is closely related
to a pair of species found primarily south of the Amazon Basin but reaching
as far north as Ecuador. A. intermedins is found from Costa Rica to Tingo
Maria, Peru and obviously links the atrocaeruleus group with the Andean
heterurus group.

In summary, most of the Meso-American species of Agapostemon belong
in North or South American species groups. Only the very widespread and
abundant A. nasutus and the very restricted and rare A. rhopalocera have no
apparent affinities with either North or South American species, or each
other.

South American Species. Some of the species occurring in South Amer-
ica such as A. nasutus, A. intermedins and A. ascitis are restricted to the
northwestern margin of the continent. Apparently these species, which are
recently evolved or recently arrived via Central America (e.g., A. nasutus),
have found further access to the continent blocked by the Andes to the east,
the Atacama Desert to the south, and the forests of the Orinoco Basin to the
north.

Agapostemon chapadensis and A. semtmelleus are the only species widely
distributed in South America. Both are found in the campos, or subtropical
grasslands, of southern Brasil, Paraguay and northern Argentina. In addition
isolated populations of A. semimelleus have been collected from the head-
waters of the Marafion, Huallaga and Cauca rivers. Presumably, these species
were able to ascend the valleys at some time in the past when the South
American grasslands were more extensive than today. Surprisingly, A.
semimelleus has been collected at the head of the Cauca Valley in Colombia.
I cannot explain this occurrence as the Cauca River drains into the Caribbean
Sea. Specimens of A. semimelleus have been collected from localities as near
as the headwaters of the Putumayo River, but the Cauca and Putumayo

Revision of the Bee Genus Agapostemon 449

rivers are separated by at least 200 kilometers of mountains (Cordillera Cen-
tral) rising as high as 5755 m. The closest relative of A. semimelleus and A.
chapadensis is A. atrocaerideas which has been collected only on the central
plateau of Costa Rica, nearly 1000 km from the headwaters of the Cauca
River.

The most aberrant species group in Agapostemon includes A. heterurus,
A. inca, A. boliviensis, A. lanosus and A. mourei. These species are very
closely related and are restricted to the eastern margins of the central Andes.
As many of the valleys in this region are both ecologically isolated and poorly
collected, it seems reasonable to predict that more species in this group will
be discovered.

A. intermedins is morphologically intermediate between the A. heterurus
group and the A. atrocaeruleus group. A. intermedins ranges from central
Costa Rica, where it is sympatric with A. atrocaeruleus, to northwestern
South America. It has also been collected from Tingo Maria at the head of
the Huallaga Valley where it is sympatric with A. lanosus. Possibly A.
intermedins is part of an Arten\reis running from Costa Rica down to the
grasslands of Brasil (the A. atrocaeruleus group) and down to the eastern
margins of the Andes (the A. heterurus group).

METHODS

Terminology. The terminology is essentially that utilized by Michener
(1944, 1965). However, the following terms have been proposed since no
appropriate ones existed for these features.

The propodeal shield is the flat or slightly concave area on the posterior
vertical surface of the propodeum usually delimited by a conspicuous propo-
deal carina (Fig. 25). Male genitalia often bear an apical stylus, medial plate
and basal stylus on the mesal surface of the gonostylus (Fig. 180). Although
often reduced and covered by long bristles, a ventral lobe (Fig. 223) is always
present on the ventral surface of the gonocoxite.

The specific importance of the maculations of male legs has necessitated
a more precise and morphologically consistent system for designating various
portions of the legs than that used by Michener and others. The legs of
bees, like those of most insects, may move anteriorly and posteriorly, and may
be flexed ventrally and extended laterally. The middle legs, when extended,
are nearly perpendicular to the longitudinal axis of the body (presumably
the "primitive" condition in the Arthropoda). The fore legs and hind legs,
however, are directed respectively forward and rearward to a greater or
lesser degree. Owing to the rearward orientation of the hind legs of bees
(especially in pinned specimens and in live bees in flight), that which is
designated by previous authors as the "inner" surface of the hind legs is
not homologous with what they term the "inner" surface of the fore legs.

450 The University Science Bulletin

Therefore, I propose a terminology consistent with the presumed serial
homology of the parts of the three pairs of legs rather than with the positions
in which they are borne. Although this system may at first seem strange
and cumbersome, it is economical in that one can refer, for example, to
serially homologous maculations as being on the posterior surfaces of all
three pairs of legs. With the system used by other authors one could not
refer to such maculations as being on the "inner" surfaces of all three
pairs of legs. The disadvantage of the proposed system is that the "dorsal"
quadrant of the tarsus may be visible only from below (due to the flexed
position of the leg), and the "anterior" surface of the hind leg only from
the side.

In the following descriptions, the legs are considered as extended at right
angles to a sagittal plane through the body of the bee. The hypothetical
leg segment is regarded as cylindrical, its cross-section divided by perpen-
dicular diagonals into dorsal, ventral, anterior and posterior quadrants.

Description. To conserve space, the reader is often referred to the de-
scription of another species or sometimes the opposite sex. All interspecific
comparisons are between members of the same sex unless otherwise specified.

Because sexual dimorphism is so pronounced in Agapostemon, characters
used to differentiate males (or females) of two closely related species are not
always present on members of the opposite sex. In some instances members
of only one sex can be differentiated morphologically. In other cases the
morphological differences between species are so subtle as to render identi-
fication difficult regardless of sex. For these reasons distributions have been
used (especially in the keys) wherever morphological differentiation was
difficult or impossible. Of course, no species is recognized solely on the basis
of its distribution. The reader must be cautious in relying on distributional
differentia because the ranges of species are subject to change without notice.

If two species are very similar then only one is described and the other
is described as identical but for the differentiating characters. Relative size
is presented as proportions and absolute size may be derived from scale
drawings. For each character, the drawings are to the same scale to
facilitate size comparisons between species.

Synonymies have been kept as brief as possible and no references to
synonyms have been made unless they involve nomenclatural changes of
specific epithets. An effort was made to locate and examine the primary
types of all nominal species and subspecies and, whenever possible, their
location has been noted in the synonymy. In quotations of label data on
the types of new species, a single slash indicates the break between lines
on a label and a double slash the break between labels on a pin.

Revision of the Bee Genus Agapostemon 451

KEY TO SPECIES OF THE UNITED STATES
AND CANADA

I. Female; 10 flagellomeres; scopa on hind leg 2

Male; 11 flagellomeres; scopa absent 14

2.( 1 ) Metasomal terga bright metallic green to blue, concolorous with

head and mesosoma 3

Metasomal terga black to pale amber, not concolorous with
metallic head and mesosoma 7

3.( 2 ) Wings almost hyaline, only slightly darkened on distal margins;

mandible yellow basally 4

Wings transparent brown, distal margins conspicuously dark-
ened; mandible usually amber basally (rarely yellow) splendens

4.( 3 ) Mesoscutum with numerous fine punctures interspersed with
fewer distinctly larger and deeper punctures; punctuation often

so fine as to leave mesoscutum shiny (Figs. 23, 24)

texanus and angelicus*

Mesoscutum coarsely punctate or rugose, lacking punctures of

two distinct sizes and never shiny 5

5.( 4 ) Mesoscutum coarsely rugose at least on inner borders of

parapsidal lines, if not rugose throughout jemoratus

Mesoscutum between parapsidal lines coarsely punctate, not
rugose 6

6.( 5 ) Common in eastern half of U.S., becoming rare in the Great

Plains (Fig. 18) radiatus

Common on Mexican Plateau and in arid southwestern U.S., rare

along eastern border of Rocky Mountains (Fig. 2) cockfrelli

7.( 2 ) Posterior lobe of pronotum with pale yellow or creamy spot at
apex (Fig. 134); clypeus with yellow transverse band (Fig. 42)

nastttus

Posterior lobe of pronotum metallic at apex, never creamy or
yellow; clypeus with or without yellow transverse band 8

8.( 7 ) Clypeus with transverse apical or subapical yellow band;

metasoma black or pale amber 9

Clypeus metallic with apical region dark brown to black;
metasoma always black 10

9.( 8 ) Pronotum with single conspicuous sharp carina extending
postero-ventrally from lateral angle (Fig. 135); metasomal terga
black with narrow basal bands of white tomentum; scape dark

brown to black, lacking yellow leunculus

Pronotum without conspicuous sharp carina extending postero-
ventrally from lateral angle (may have several small carinulae);
metasomal terga pale amber to black, with white tomentum on

basal halves; scape sometimes marked with yellow 12

10.(8) Genal ridges coarse (2-3 per 0.25 mm) (Fig. 28); mandible

usually yellow basally virescens

Genal ridges fine (5-6 per 0.25 mm); mandible amber to brown-
black basally 1 1

* The females of A. texanus and A. angelicus cannot be reliably separated on the basis of
their morphology; however, females occurring well outside the range of A. angelicus males are
probably A. texanus (see map, Fig. 20).

452 The University Science Bulletin

11.(10) Wings transparent brown; white tomentum lacking on anterior
part of first metasomal tergum; protuberance above clypeus

without large, central, shiny area coloradinus

Wings hyaline, not brown; white tomentum on anterior part of

first metasomal tergum 13

12.( 9 ) Dorsal area of propodeum moderately to coarsely rugose; scape
often marked with yellow; metasomal terga amber to black or

black with amber anteriorly on tergum 1 melliventris

Dorsal area of propodeum finely rugose to finely rugulose; scape
never with yellow markings; metasomal terga always entirely
black peninsularis

13.(11) Supraclypeal protuberance with smooth shiny central area with
scattered punctures; found only in Arizona, New Mexico and

on the Mexican Plateau (Fig. 27) tyleri

Supraclypeal protuberance with weakly and transversely rugose
central area; occurs around Gulf of California mexicanus**

14. ( 1 ) Clypeus conspicuously concave ventrally, more than three times

as broad as long (Figs. 86, 87) nasutus

Clypeus normal, little broader than long 15

15.(14) First metasomal tergum yellow to very pale amber basally, not
brown or black, and hind basitarsus always slender and simple
(Fig. 155); hind leg yellow with small brown or black area

apically on femur and basally on tibia melliventris

First metasomal tergum brown to black basally (if pale amber,
then hind basitarsus swollen and with basal ridge and apical
groove); hind leg usually with dark brown to black stripe on tibia 16

16.(15) Hind femur without tooth; slender 25

Hind femur with subapical tooth postero-ventrally; often con-
spicuously inflated 17

17.(16) Wings transparent brown, conspicuously darkened at distal
margins; hind basitarsus with crest of basal ridge grooved

(Fig. 166) splendens

Wings hyaline or nearly so, not conspicuously darkened at distal
margins; hind basitarsus with basal ridge (if present) not grooved 18

18.(17) Metasoma with last two visible sterna (5-6) dark brown to
brown-black, without yellow maculations; sterna 2-4 with yellow

(if present) restricted to basal margins 19

Metasoma with extensive yellow maculations on last two visible
sterna (5-6) as well as on sterna 2-4 21

19.(18) Hind femur lacking brown to black stripe on posterior surface

( Fig. 158) virescens

Hind femur with large brown to black stripe covering most or

all of posterior surface 20

20.(19) Brown to black streak less than % length of posterior surface of

hind tibia (Fig. 160) coloradinus

Brown to black streak extending entire length of posterior surface

of hind tibia (Fig. 159) tyleri

** Although not currently known to occur in the United States, this species may yet be
found in the vicinity of San Diego or Yuma, thus it is included in this key.

Revision of the Bee Genus Agapostemon 453

21.(18) Metasomal sternum 4 (antepenultimate visible) with low trans-
verse ridge not quite reaching posterior margin laterally;

metasomal tergum 4 usually without conspicuous metallic tints 22

Metasomal sternum 4 (antepenultimate visible) with low trans-
verse ridge reaching posterior margin laterally; metasomal
tergum 4 nearly always with conspicuous metallic tints 24

22.(21) Hind basitarsus lacking apical groove and basal ridge (Fig.

161) radiatus

Hind basitarsus with apical groove and basal ridge 23

23.(22) Hind basitarsus with inconspicuous basal ridge and slender apical

groove (Fig. 162) coc\erelh

Hind basitarsus with very large, conspicuous, slightly sinuate

basal ridge and broad, conspicuous apical groove (Fig. 163) .... jemoratus

24.(21) Base of apical stylus of gonostylus slightly inflated (Fig. 181);
brown to black stripe on posterior surface of hind tibia but

never on anterior surface (Fig. 164) angelicus

Base of apical stylus of gonostylus not inflated (Fig. 180); brown
to black stripe anteriorly on hind tibia (Fig. 165), or, if lacking,
then without black stripe on posterior surface texanus

25.(16) Posterior surface of hind tibia largely yellow (Fig. 143) leunculus

Posterior surface of hind tibia largely brown or black 26

26.(25) 2-4 large submarginal bristles on each side of metasomal

sternum 4 peninsularis

14-20 large, evenly spaced submarginal bristles on metasomal
sternum 4 mexicanus*

KEY TO MESO-AMERICAN** SPECIES

1. Males; 11 flagellomeres; scopa absent 2

Females; 10 flagellomeres; scopa on hind legs 15

2.( 1 ) Antenna normal 3

Antenna long, filamentous, and with apical flagellomere flattened

(Fig. 131 ) rhopalocera

3.( 2 ) Clypeus normal, little longer than broad 4

Clypeus conspicuously concave ventrally, more than three times

as broad as long (Figs. 86, 87) nasutus

4.( 3 ) Metasoma banded with yellow and dark brown or black, lacking

bands of white tomentum 5

Metasoma pale amber, with bands of white tomentum centrums

5.( 4 ) Sixth metasomal sternum flat or with inconspicuous medial ridge 6

Sixth metasomal sternum with medial "button" (flattened process
resembling human tongue and attached to sternum in much the
same way as human tongue is attached to floor of mouth
(Fig. 136) intermedins

6.( 5 ) Hind femur with conspicuous subapical tooth postero-ventrally 7

Hind femur lacking tooth 1 1

* Although not currently known to occur in the United States, this species may yet be
found in the vicinity of San Diego or Yuma, thus it is included in this key.
** Mexico, Panama, and intervening countries.

454 The University Science Bulletin

7.( 6 ) Wings hyaline or nearly so, not conspicuously darkened at distal
margins; hind basitarsus with basal ridge (if present) not

grooved 8

Wings transparent brown, conspicuously darkened at distal
margins; hind basitarsus with crest of basal ridge grooved

(Fig. 166) splendens

8.( 7 ) Metasoma with extensive yellow maculations on sterna; hind
tibia with brown or black posterior stripe (when present) not

extending from base to apex 9

Metasoma with sterna brown or black, lacking yellow macula-
tions; hind tibia with broad brown or black posterior stripe

extending from base to apex tyleri

9.( 8 ) Metasomal sternum 4 (antepenultimate visible) with low trans-
verse ridge reaching posterior margin laterally; metasomal

tergum 4 nearly always with conspicuous metallic tints 10

Metasomal sternum 4 (antepenultimate visible) with low trans-
verse ridge not quite reaching posterior margin laterally; meta-
somal tergum 4 without conspicuous metallic tints coc\erelh

10. ( 9) Base of apical stylus of gonostylus slightly inflated (Fig. 181);
brown to black stripe on posterior surface of hind tibia but

never on anterior surface (Fig. 164) angelicas

Base of apical stylus of gonostylus not inflated (Fig. 180); brown
to black stripe anteriorly on hind tibia (Fig. 165) or, if lacking,
then without black stripe on posterior surface texanus

ll.( 6 ) Base of metasomal tergum 1 dark brown or black 12

Base of metasomal tergum 1 yellow or pale amber melliventris

12.(11) Hind leg with broad brown or black stripe covering most of
posterior surface of tibia and extending from base to apex;

similar stripe on femur sometimes broken centrally 13

Hind leg largely yellow on posterior surface of tibia and femur 14

13.(12) Metasomal sternum 4 with 2-4 large submarginal bristles on

each side peninsularis

Metasomal sternum 4 with 14-16 large, evenly spaced sub-
marginal bristles mexicanus

14.(12) Pronotum with very acute lateral angle; mesoscutum with con-
spicuous flange laterally on anterior margin atrocaeruleus

Pronotum with lateral angle rounded; mesoscutum lacking flange

on anterior margin leunculus

15.( 1 ) Metasomal terga bright metallic green to blue-green 16

Metasomal terga pale amber to black, not metallic 19

16.(15) Mandibles amber with metallic green spot basally; dorsal area
of pronotum with very large, widely separated carinae extend-
ing from anterior margin to propodeal carina aenigma*

Mandibles amber or yellow, lacking metallic spot basally; pro-
notum rugose dorsally 17

17.(16) Wings almost hyaline, only slightly darkened on distal margins;

mandibles yellow basally 18

Wings transparent brown, distal margins conspicuously dark-
ened; mandibles usually amber (sometimes yellow) basally .... splendens

* Although labeled "Costa Rica" the two specimens of A. aenigma are closely related to
West Indian species and may be mislabeled.

Revision of the Bee Genus Agapostemon 455

18.(17) Mesoscutum coarsely punctate to finely rugose, lacking punctures

of two distinct sizes and never shiny coc\erelli

Mesoscutum with numerous fine punctures interspersed with
fewer distinctly larger and deeper punctures, punctation often

so fine as to leave mesoscutum shiny (Figs. 23, 24)

texanus and angelicas**

19.(15) Lateral angle of pronotum acutely pointed; mesoscutum with

conspicuous flange laterally on anterior margin 20

Lateral angle rounded, not acutely pointed; mesoscutum lacking
conspicuous flange on anterior margin 21

20.(19) Head and mesosoma black (usually with inconspicuous dark
blue tints and with yellow band on clypeus); metasoma black

with bands of white tomentum atrocaeruleus

Head and mesosoma metallic green or coppery; metasoma pale
amber to black, with yellow integumental bands replacing usual
bands of white tomentum intermedins

21.(19) Posterior lobe of pronotum metallic at apex, never creamy or

yellow; clypeus with or without yellow transverse band 22

Posterior lobe of pronotum with creamy or yellow spot at apex;
clypeus with yellow transverse band nasntus

22.(21) Clypeus with transverse, subapical. yellow band; metasoma

amber to black 23

Clypeus black at apical margin, lacking yellow band; metasoma
always black 25

23.(22) Pronotum without single conspicuous sharp carina extending
postero-ventrally from lateral angle ( may have several small
carinulae); metasomal terga pale amber to black, with white
tomentum on basal halves; scape dark brown to black, sometimes

marked with yellow 24

Pronotum with single conspicuous sharp carina extending
postero-ventrally from lateral angle; metasomal terga with nar-
row basal bands of white tomentum; scape dark brown to black,
lacking yellow leunculus

24.(23) Dorsal area of propodeum moderately to coarsely rugose; scape
often marked with yellow; metasomal terga amber to black, or

black with amber anteriorly on tergum 1 mellwentris

Dorsal area of propodeum finely rugose to finely rugulose; scape
never with yellow markings; metasomal terga always entirely
black peninsularis

25.(22) Head and mesosoma bright metallic green 26

Head and mesosoma black with metallic purple tints erebus

26.(25) Supraclypeal protuberance with smooth shiny central area and
with scattered punctures; occurs in New Mexico, Arizona, and

on the Mexican Plateau (Fig. 27) tyleri

Supraclypeal protuberance with weakly and transversely rugose
central area; occurs around the Gulf of California mexicanus

s

** The females of A. texanus and A. angelicits cannot be separated reliably on the basis
of their morphology; however, females occurring well outside the range of A. angelicus males
are probably A. texanus (see map, Fig. 20).

456 The University Science Bulletin

KEY TO WEST INDIAN SPECIES

1. Males; 11 flagellomeres; scopa absent 2

Females; 10 flagellomeres; scopa on hind leg 16

2.( 1 ) Hind femur without tooth 3

Hind femur with tooth on postero-ventral margin 4

3.( 2 ) Lower portion of clypeus abruptly flattened, glabrous and im-
punctate (Figs. 109-110); metasomal terga yellow with black

bands; from Hispaniola, Cuba and Jamaica h^ohliellus

Lower portion of clypeus normal, not abruptly flattened,
glabrous (Figs. 121-122) or impunctate; metasomal terga honey-
colored; from Haiti centratus

4.( 2 ) Metasomal terga black or pale amber, never with metallic tints;

clypeal region elongate 5

Metasomal terga with yellow and brown to black bands, often

with metallic tints; clypeal region normal 9

5.( 4 ) Head and mesosoma bright metallic green to blue 6

Head and mesosoma shiny brown-black, not metallic; from

Cuba obscuratus

6.( 5 ) Metasomal terga dark brown to black, pale bands of tomentum

conspicuous 7

Metasomal terga pale amber with brown bands, pale bands of
tomentum inconspicuous or absent 8

7.( 6 ) Sculpturing of mesosoma slightly finer than that of vindulus,
contiguous punctures giving the mesoscutum a dark blue ap-
pearance when viewed from above; from Hispaniola hispaniolicus

Sculpturing of mesosoma slightly coarser than that of hispanioli-
cus, subcontiguous punctures giving the mesoscutum slightly
shiny green to blue reflections; from Cuba vindulus

8.( 6 ) Basal ridge of hind basitarsus very broadly and deeply grooved
(Fig. 147); mesoscutellum extremely shiny, only weakly sculp-
tured; from Jamaica swainsonae

Basal ridge of basitarsus very narrow and without broad, deep
groove (Fig. 149); mesoscutellum weakly shiny with contiguous
punctures; from Cuba cubensis

9.( 4 ) Eye normal brown; pterostigma translucent amber to dark brown 10

Eye and (or) underside of pterostigma (except for dark mar-
gins) pale cream colored to opaque yellow 14

10.( 9 ) Metasomal tergum 3 weakly to strongly metallic green to blue

medially 11

Metasomal tergum 3 without metallic green to blue medially 13

11.(10) Metasomal tergum 1 with metallic green to blue tints postero-

laterally 12

Metasomal tergum 1 without metallic green to blue tints postero-
laterally; from Hispaniola insidaris

12.(11) Hind femur 60% as wide as long (Fig. 146); from Jamaica .. jamaicensis
Hind femur less than 55%, as wide as long (Fig. 144); from
Cuba, Hispaniola, New Providence I. and probably from Andros
I. and Cat I poeyi

Revision of the Bee Genus Agapostemon

457

13.(10) Metasomal terga 4 and 5 with conspicuous metallic green tints

medially; from San Salvador I. (= Wading I.) columbi

Metasomal terga 4 and 5 without conspicuous metallic green
tints medially (may have faint metallic tints laterally); from
Puerto Rico (including Mona I. and Vieques I.) and New

Providence I viequesensis

Pedicel, unlike scape or flagellum, pale amber to yellow below 15

Pedicel, unlike scape or flagellum, dark brown below; from

Crooked I cyaneus

From Long I sapphirinus

From Cat I., Rum Cay, Conception I., New Providence I. and

Mariguana (=Mayaguana?) I ochromops

Metasomal terga black to pale amber, not metallic 17

Metasomal terga largely metallic green to blue to purple 22

Lower portion of clypeus brown or black, metasomal terga pale

amber to black, lacking yellow bands 18

Lower portion of clypeus with broad yellow band; metasomal

terga with yellow bands; from Hispaniola, Cuba and Jamaica .. \ohliellus

Metasomal terga uniformly brown-black with narrow, white

bands of tomentum 19

Metasomal terga 1 and 2 honey-colored becoming brown-black

on posterior terga; from Hispaniola alayoi

Head and mesosoma metallic green to blue 20

Head and mesosoma shiny brown-black, not metallic; from

Cuba obscuratus

More than 50% of interocular area above antennal sockets and

below median ocellus rugose to rugulose 21

More than 50% of interocular area above antennal sockets and
below median ocellus with fine, deep, contiguous punctures;

from Hispaniola hispamolicus

Mandible with metallic green tints basally, about as dark

ferruginous-brown basally as apically; from Cuba viridulus

Mandible without metallic green tints basally, much darker

apically than basally; from Jamaica swainsonae

Metallic coloration dark blue or dark purple 23

Metallic coloration green 24

Gena with fewer than 10 extremely coarse carinae; tegula and

legs with some pale amber areas; from Long I sapphirinus

Gena with many more than 10 fine carinae; tegula and legs

dark brown; from Crooked I cyaneus

Eye normal brown; pterostigma translucent amber to dark brown 25

Eye or underside of pterostigma (usually both) pale cream-
colored to opaque yellow; from Cat I., Rum Cay, Conception I.,

New Providence I. and Mariguana (=rMayaguana?) I ochromops

Mandible with metallic green spot basally 26

Mandible lacking metallic green spot basally; from Puerto Rico
(including Mona I. and Vieques I.) and New Providence I. .. viequesensis

Metasomal sterna 3 and 4 with conspicuous medial metallic

green tints 27

Metasomal sterna 3 and 4 brown, without medial metallic green

tints 28

458 The University Science Bulletin

27.(26) Mesoscutum rugose laterally along parapsidal line, becoming
coarsely rugose anteriorly, punctate centrally and posteriorly;

from Cuba, Hispaniola, New Providence I., Cat I. and Andros I poeyi

Mesoscutum finely punctate, with extreme antero-lateral margin
rugulose; from Hispaniola insularis

28.(26) Scopal hairs on femur and hairs on postero-ventral surface of

hind tibia golden; from San Salvador I. ( = Watling I.) columbi

Scopal hairs on femur and hairs on postero-ventral surface of

tibia white; from Costa Rica aenigma*

KEY TO SOUTH AMERICAN SPECIES

1. Female; 10 flagellomeres; scopa on hind leg 2

Male; 11 flagellomeres; scopa absent 10

2.( 1 ) Mesoscutum rounded on anterior margin; lateral angle of

pronotum inconspicuous and not projecting antero-laterally 3

Mesoscutum with acute upturned flange on anterior margin;
lateral angle of pronotum conspicuous and projecting antero-
laterally 5

3.( 2 ) Pronotum with creamy or yellow spot at apex of posterior lobe
and without carina extending postero-ventrally from lateral angle

(Fig. 134) nasutus

Pronotum without creamy or yellow spot at apex of posterior lobe
and with conspicuous sharp carina extending postero-ventrally
from lateral angle (Fig. 135) 4

4.( 3 ) Metasoma black; lateral angle of pronotum rounded; 3 spatulate

teeth on posterior hind tibial spur leunculus

Metasoma amber; lateral angle of pronotum angular; 5 (rarely

4?) spatulate teeth on posterior hind tibial spur ascius

5.( 2 ) Metasomal terga 2 and 3 amber or black with yellow basally

(best seen in posterior view) 6

Metasomal terga 2 and 3 amber or black but lacking yellow
semimelleus or chapadensis*

6.( 5 ) Head and mesosoma dull coppery, metallic greenish-black or
bluish-black; mesoscutal pubescence dense and woolly; restricted

to western South America 7

Head and mesosoma usually bright metallic green (may be
coppery in Central America); mesoscutal pubescence not espe-
cially dense and finely branched; Central and South America
intermedins

7.( 6 ) Propodeum metallic blue-black or green-black dorsally; clypeus

without yellow maculations heternms

Propodeum metallic coppery dorsally; clypeus usually with
yellow maculations 8

* Agapostcmon aenigma is included in this key because its similarity to West Indian species
casts doubt on the correctness of its "Costa Rica" label. A specimen with identical label data
belongs to the Haitian species A. centratus.

* The females of these species cannot be distinguished reliably (cf. A. semimelleus,
Variation).

Revision of the Bee Genus Agapostemon 459

8.( 7 ) Propodeum with coarse parallel striae dorso-laterally; clypeus

without yellow maculations lanosus

Propodeum coarsely rugose dorso-laterally; clypeus with yellow
maculations 9

9.( 8 ) Posterior half of each metasomal tergum pale amber, prepygidial

fimbria brown to black mourei

Posterior half of metasomal tergum 1 and 2 dark amber, tergum
3 brown, terga 4 and 5 and prepygidial fimbria brown-black .. bolwiensis
10. ( 1 ) Last visible (6th) metasomal sternum with medial, button-like

protrusion basally (Figs. 136-140) 14

Last visible (6th) metasomal sternum without button-like
protrusion 11

11.(10) Clypeus normal, little longer than broad 12

Clypeus conspicuously concave ventrally, more than three times

as broad as long (Figs. 85-87) nasutus

12.(11) Lateral angle of pronotum prominent, acute and projecting
antero-laterally; first metasomal tergum dull, with large, deep,

subcontiguous punctures 13

Lateral angle of pronotum rounded; first metasomal tergum

shiny, with small, shallow, scattered punctures leunculus

13.(12) Hind femur with brown on posterior extending from base to
apex (Fig. 172); gonostylus with distal stylus about twice as

long as width of medial plate (Fig. 173) chapadensis

Hind femur with brown on posterior limited to base and apex
(Fig. 171); gonostylus with distal stylus little longer than width
of medial plate (Fig. 174) semimelleus

14.(10) Hind femur less than twice as long as broad; hind basitarsus ex-
tensively marked with brown or black; basal ridge and apical
groove on hind basitarsus large and conspicuous (Figs. 167-

170) 15

Hind femur more than twice as long as broad; hind basitarsus
yellow, only rarely marked with brown; basal ridge and apical
groove on hind basitarsus reduced and inconspicuous (Fig.
151) intermedins

15.(14) Antepenultimate visible sternite (4th) without conspicuously

long, contiguous, hooked setae laterally on posterior margin 16

Antepenultimate visible sternite (4th) with two or three con-
tiguous, long (about twice length of adjacent setae), thickened,
hooked setae laterally on posterior margin lanosus

16.(15) Last visible sternite (6th) with "button" about lA as wide as
sternite (Figs. 137, 139); basitarsus relatively broad (Figs.

167, 170) 17

Last visible sternite (6th) with "button" about l/2 as wide as
sternite (Fig. 140); basitarsus relatively slender (Fig. 168) inca

17.(16) Tegula pale amber with yellow lunule anteriorly; medial plate

of gonostylus large (Fig. 176) mourei

Tegula dark ferruginous, without yellow lunule anteriorly;
medial plate of gonostylus small and rounded (Fig. 177A) bolwiensis

460 The University Science Bulletin

DESCRIPTIONS, SYNONYMIES AND DISTRIBUTIONS
Agapostemon aenigma n. sp.

The name of this species refers to its mysterious affinity with certain
West Indian species.

Agapostemon aenigma is known from two female specimens labeled
"Costa Rica" in the same handwriting and type of label as two males of
the Haitian A. centratus. This species appears closely related to some West
Indian species and may be mis-labeled as are the similarly labeled specimens
of A. centratus. The holotype and paratype are in the Academy of Natural
Sciences, Philadelphia.

Diagnosis. The female may be distinguished from other green Central
American Agapostemon by the coarse parallel carinae on the dorsal surface
of its propodeum. Among West Indian females with bright metallic green
to blue metasomas, A. aenigma may be distinguished from females of
A. ochromops by its lack of milky to yellowish eyes and pterostigma; from
females of A. viequesensis by the basal metallic spot on its mandible; from
females of A. cohtmbi by the white (not golden) pubescence on the under-
side of its hind femur and tibia; and from A. poeyi and A. insularis by its
lack of strong metallic tints medially on sterna 3 and 4. This species is very
close to A. insularis and the two may eventually prove to be conspecific.

Description

female

General coloration of head, mesosoma and metasoma bright metallic
greenish blue. Head: pubescence as in A. viequesensis. (1) La brum as
in A. hispaniolicus. (2-4) Clypeus, interocular area and vertex as in A.
viequesensis. (5) Gena much more coarsely carinate than in A. viequesensis
but not quite as coarse as in A. poeyi. (6) Malar area absent. (7-8)
Mandible and antenna as in A. poeyi. Mesosoma: pubescence as in
A. viequesensis but slightly more fuliginous on mesonotum and metanotum.
(9) Pronotum as in A. viequesensis but with lateral angle and posterior
lobe more angular, and with fewer and coarser (not as coarse as in A. poeyi)
horizontal carinae postero-laterally. (10-12) Mesoscutum, mesoscutellum
and metanotum as in A. viequesensis. (13-14) Mesepisternum and metepi-
sternum as in A. viequesensis but with sculpturing very slightly coarser.
(15) Propodeum as in A. poeyi but with finer sculpturing (not quite as
fine as in A. viequesensis). (16-17) Wing and tegula as in A. viequesensis
but with tegula darker with metallic tints more extensive. (18-20) Fore,
middle and hind legs dark brown, coxae with strong metallic tints.
Pubescence dark amber to brown, scopal hairs white. Metasoma: as
in A. viequesensis but with faint metallic tints laterally on sterna 3 and
4 of paratype (cf. A. insularis 5 ) .

Revision of the Bee Genus Agapostemon 461

Agapostemon alayoi n. sp.

This species is named after Dr. Pastor Alayo D. of La Habana, Cuba,
who has contributed significantly to our knowledge of Antillean bees.

I have seen a single specimen labeled "Ace. 1007-25/Port au Prince,/
Haiti., 1925//G. N. Wolcott/Coll." The type is in the United States Na-
tional Museum, Washington, D.C.

Diagnosis. The female may be distinguished from females of A.
Itphliellus by the lack of yellow on its clypeus. No other West Indian
species has females with metasomal terga 1-2 pale amber. This species is
phenetically closest to A. viridiilns and A. hispaniolicus. It differs from
both in the coloration of its metasoma, in having a more densely punctate
mesoscutellum and in being larger than females of A. hispaniolicus.

Description

female

General coloration of head and mesosoma metallic greenish blue,
metasoma honey-colored anteriorly, brown-black posteriorly. Head: as
in A. vmduliis but metallic blue, with very dark fuliginous hair on
interocular area and vertex. Mesosoma: as in A. viridiilns but metallic
blue, with slightly finer, deeper and more numerous punctures on meso-
scutum and mesoscutellum. Mesoscutellum not shiny as in A. viridiilns.
Sculpturing of mesepisternum, metepisternum and propodeum slightly
deeper than in A. viridiilns. Pubescence of mesoscutum, mesoscutellum and
metanotum much darker fuliginous than as in A. viridiilns. Metasoma:
as in A. viridiilns but terga 1-2 pale amber instead of brown-black.

Agapostemon angelicus Cockerell

Agapostemon angelicus Cockerell 1924. Type 9, California Academy of Sciences.

I have seen the type in San Francisco and it is from Pond Island Bay,
Angel de la Guarda Island in the Gulf of California, Mexico. As the males
of A. texanns are rare relative to those of A. angelicus in this region, it
seems reasonable and expedient to consider this type to be conspecific with
those males which differ in genitalia and color pattern from those of A.
texanns.

Distribution. All males of A. angelicus have been found well within
the range of the males of A. texanns. Owing to this sympatry and my in-
ability to distinguish between females of A. texanns and A. angelicus, I have
here considered the distribution of males and disregarded females.

Although common only in the arid regions of the southwestern United
States and northern Mexico, males of A. angelicus have been found as far
north as west-central North Dakota, as far south as northern Durango
(Mexico), as far east as central Iowa, and as far west as southwestern

462 The University Science Bulletin

California. This species ranges from below sea level in Death Valley,
California, to 12,000 ft. (3,658 m) on Mount Evans (timberline is at 11,700
ft.) in Colorado. The altitudinal range of A. angelicas is as astounding as
the latitudinal range of A. texanus. In Arizona males of A. angelicas have
been collected from April through November; in Kansas from June through
October; and in Mexico from June through October. (Map, Fig. 20.)

Albeit uncommon, A. angelicas does occur in the Great Plains, where
it is found together with A. texanus in the gallery forests of the tributaries
of the Missouri and Mississippi Rivers. This habitat is the "Northern
Floodplain Forest" of Popalas, Salix, and Ulmas (no. 98 in Appendix A).
While not as diverse as those of A. texanas, its habitats are numerous (39
of the 116 given by A. W. Kiichler 1964— cf. Appendix A).

Diagnosis. The male may be distinguished from many other species by
its toothed hind femora, the apical stylus on its gonostylus, and the lack of
a low medial ridge on the apical half of its last visible sternum; and from
A. texanus by the shorter apical stylus with swollen base and enlarged apex
on its gonostylus, and by the presence of a posterior stripe and absence of
an anterior stripe on its hind tibia. The female may be distinguished from
most other species by its bright metallic green to blue metasomal terga, its
almost hyaline wings, and by the presence of two distinct sizes of
mesoscutal punctures.

I have labeled females of A. angelicas as "Agapostemon texanas or A.
angelicas" because I am unable to separate them from A. texanas. Sand-
house (1936) claimed that these species differ in the punctation of the
mesoscutum and in the color of the pubescence. After having examined
many thousands of females of these species I am forced to conclude that
the variation within A. texanas nearly encompasses the range of variation
within A. angelicas, thus invalidating the characters used by Sandhouse.
It is probably true that the modes of A. angelicas characters differ from those
of A. texanas but this cannot be demonstrated without positively identified
females of both species. If one were able to positively identify A. angelicas
females, perhaps by means of rearing, it might be possible to utilize a
discriminant function to distinguish between females of the two species.
Owing to the occurrence of A. texanas males within the range of A.
angelicas males, I was unable to obtain females which were indisputably
A. angelicas. Even if one were to find an area where only A. angelicas
occurred, one could not be sure that differences between these females and
females of A. texanas were not simply geographic variations.

Variation. Basing the probable identity of females on the relative
abundance of males, it would seem that most females of A. angelicas have
a slightly shinier mesoscutum and whiter pubescence than most of those of
A. texanas. Like the males, the females of A. angelicas probably average

Revision of the Bee Genus Agapostemon 463

slightly smaller than those of A. texanus, but the variation within each
species is far greater than any difference between them. In both males and
females of A. angelicas the metallic coloration is the same green as that of
the sympatric A. texanus. The most noticeably variable character of A.
angelicits is the amount of melanic pigmentation on the fore coxae of
males, which range in color from yellow to brown-black, but intra-locality
variation is far greater than inter-locality variation.

Description

male (Figs. 73-74, 164, 181)

Males as in A. texanus but only about 85% as large with whiter
pubescence on the mesonotum and metanotum; with dark stripe present
posteriorly and absent anteriorly on its hind tibia (Fig. 164) ; with apical
stylus on its gonostylus shorter, swollen basally, and larger apically; with
medial plate on its gonostylus smaller; and with basal stylus on its
gonostylus broader and blunt apically (Fig. 181).

female (Figs. 67-68)

As in A. texanus but always green and probably smaller with shinier
mesoscutum and whiter pubescence.

Agapostemon ascius n. sp.

The name of this species comes from the Greek sl{ias, or shadow. As
used in Pliny, ascius literally means without shadow and refers to in-
habitants of those regions (i.e., the tropics) where the sun is directly over-
head once or twice each year.

Agapostemon ascius is known from two females labeled, "5 mi. W.
Cajamarca/ Tol, COLOMBIA/ 11-20-1965// Figleaf/ gourd/ A. E. &
M. M./ Michelbacher/ Collectors// 210.1/ 0855-/ 0905." The holotype
and paratype are at the University of California, Berkeley.

Diagnosis. A. ascius may be distinguished from most other South Amer-
ican species by its lack of an acute upturned flange on the anterior margin
of its mesoscutum, from A. nasutus by the lack of yellow on its posterior
pronotal lobe, and from the closely related A. leunculus and A. erebus by
its amber metasoma and 5 (rarely 4?) spatulate teeth on its posterior hind
tibial spur.

Description
female
General coloration of head and mesosoma bright metallic green, meta-
soma amber with black at base. Head: (1-4) Labrum, clypeus, interocular
area and vertex as in A. nasutus. (5) Gena with coarse, parallel and

464 The University Science Bulletin

anastomosing carinae. (6) Malar area dark amber to brown, very short.

(7) Mandible yellowish amber basally, becoming ferruginous on apical %.

(8) Antenna black above, dark brown below, undersides of first two
flagellomeres darker than others. Mesosoma: metallic green, never
with yellow spot on posterior lobe of pronotum as in A. nasutits. (9)
Pronotum with prominent carina running postero-laterally from lateral
angle; lateral angle projecting slightly but not markedly as in A. inter-
medins. (10-12) Mesoscutum, mesoscutellum and metanotum as in A.
nasutus but punctures slightly coarser and pubescence slightly more dense.
(13) Mesepisternum very coarsely rugose anteriorly to moderately rugose
posteriorly; pubescence creamy. (14) Metepisternum as in A. leunculus.
(15) Propodeum as in A. nasutus but more coarsely sculptured. (16) Wing
as in A. nasutus. (17) Tegula dark brown with transparent spot postero-
laterally. (18-20) Fare, middle and hind legs as in A. leunculus. Metasoma:
(21) Terga amber with anterior half of tergum 1, posterior half of tergum 5
and lateral margins of 2-4 dark brown to black; pubescence short, sparse and
inconspicuous. (22) Sterna brown to brown-black with long pale hairs on
exposed areas.

Agapostemon atrocaeruleus Friese

Agapostemon atrocaeruleus Friese 1916.

I am unable to locate the type of this species but have seen males and
females labeled "Typus" (which only means they were identified by Friese)
and they agree with his descriptions.

Distribution. I have seen specimens from the following localities in
Costa Rica: Cartago (2 $), San Jose (9 8, 7 2 ), San Mateo (1 8),
Santiago de Puriscal (12 8,7 5), and Turrialba (7 8 , 12 2). Males
have been collected from July through November and females from July
through September. Members of this species have been collected from
1,950 ft. (594 m) in Turrialba up to 3,963 ft. (1,200 m) in San Jose.

Diagnosis. The male may be distinguished from males of all Central
American Agapostemon except A. intermedins by the very acute lateral
angle and posterior lobe on its pronotum and the conspicuous flange laterally
on the anterior margin of its mesoscutum; and from males of A. inter-
medins by the lack of a medial tongue-shaped process, or "button," at the
base of its 6th metasomal sternum (small bump may be present, but not a
distinct "button"). The female may be distinguished from females of all
other species of Agapostemon by the combination of acutely pointed lateral
pronotal angle and its almost wholly black color (usually with yellow band
on clypeus and dark metallic blue tints on head and mesosoma).

Revision of the Bee Genus Agapostemon 465

Description

male (Figs. 79-80, 142, 196, 213)

General coloration of head and mesosoma bright metallic green,

metasoma banded with black and yellow. Head (Figs. 79-80): pubescence

golden to amber, becoming creamy on gena. (1) Labrnm as in A. texanus

but with apical portion shorter and much more acute at apex. (2) Clypeus

with large scattered punctures on lower (yellow) half, becoming very weakly

punctate and with weak horizontal carinulae on upper (green) half. (3)

Interocular area with moderately small and deep contiguous punctures

below ocelli, becoming shallower and larger (almost rugulose) at level of

antennal sockets; supraclypeal protuberance with large punctures separated

by about their own diameters and inclined medially. (4) Vertex with

moderately small, deep, contiguous punctures anteriorly becoming shallower

laterally, very weakly rugose between ocelli, and becoming transversely

carinate posteriorly. (5) Gena with parallel subcontiguous weak carinae

extending dorso-laterally from antero-ventral margin. (6) Malar area yellow,

very short. (7) Mandible yellow with apical !/3 ferruginous to amber. (8)

Antenna brown to brown-black above; scape yellow below, pedicel and

flagellum yellowish to amber below. Mesosoma: pubescence creamy,

becoming golden to amber on mesonotum and metanotum. (9) Pronotum

with small, inconspicuous, yellow spot on posterior lobe; lateral angle and

posterior lobe acutely pointed, lateral angle projecting antero-laterally,

sharp carina extending postero-ventrally from apex of lateral angle, and

weak horizontal carinulae postero-ventrally. (10) Mesoscutum with large

conspicuous flange laterally on anterior margin extending antero-dorsally;

punctures small (slightly larger than those of A. intermedins), very deep

and contiguous, becoming rugose antero-laterally. (11) Mesoscutellum with

medial punctures like those of mesoscutum but separated by about their

own diameters and becoming larger and much less dense laterally; large

shiny area with few punctures on each side of mid-line. (12) Metanotum

coarsely and moderately shallowly rugose. (13) Mesepisternum moderately

coarsely rugose anteriorly, becoming finely rugose posteriorly with largest

rugae extending postero-ventrally in arc from anterior margin of metepis-

ternum. (14) Metepisternttm with coarse, subcontiguous punctures on

lower Yi-Yi becoming horizontally and irregularly rugose on upper l/i-x/i.

(15) Propodeum with propodeal carina strong and conspicuous; propodeal

shield very coarsely and transversely to irregularly rugose (sometimes with

scattered punctures); dorsal area very coarsely and deeply rugose; lateral

area with large, deep, subcontiguous punctures and faint rugae (punctures

and rugae slightly finer anteriorly than posteriorly). (16) Wing hyaline or

slightly fuliginous, veins dark brown. (17) Tegula transparent amber with

yellow on proximal margin and submarginally on anterior portion (some-

466 The University Science Bulletin

times small amount on posterior margin). (18) Fore leg yellow with dark
brown basally and posteriorly on coxa; femur with brown apical spot
dorsally and often with brown streak posteriorly; brown streak posteriorly
on tibia. (19) Middle leg as fore leg but with coxa dark brown, with brown
spot at apex of femur larger, and with dark brown streak on posterior of
tibia extending to dorsal surface. (20) Hind leg (Fig. 142) with coxa bright
metallic green dorsally and posteriorly, yellow ventrally; trochanter dark
brown, becoming yellow ventrally; femur yellow with dark brown on
proximal margin and with dark brown patch at apex; tibia with dark
brown to brown-black streak antero-dorsally on apical /4-% and with large
dark brown spot anteriorly on distal margin. Metasoma: (21) Terga
2-6 dark brown to brown-black on posterior %-% and yellow on anterior
Vi'Vi (often partially concealed by overlap); tergum 1 dark brown to brown-
black with yellow band antero-dorsally (often interrupted laterally); very
conspicuously punctate with subcontiguous punctures largest on tergum 1,
becoming progressively smaller on posterior terga. Pubescence golden to
amber anteriorly on tergum 1, laterally on terga 1-7 and dorsally on terga
5-7; very short dorsally on terga 1-4, golden to fulvus on yellow bands and
dark brown to brown-black on dark bands. (22) Sterna yellow, brown
tinged with green on sternum 1 and with transverse bands on posterior
margins of sterna 2-5 narrowest on sternum 2, becoming progressively wider
(especially medially) posteriorly; sternum 6 with slight proximo-medial
bulge but without distinct "button"; pubescence limited to scattered, moder-
ately long, fulvus to golden hairs on exposed areas. (23) Genitalia (Figs.
196, 213) with ventral lobes very short and bearing brush of hairs apically.

female (Figs. 43-44)

General coloration of head and mesosoma black or blue-black, pubescence
black or fuscous. Head (Figs. 43-44): pubescence black or nearly black,
becoming fuliginous on gena. (1) Labrum as in A. texanus. (2) Clypcus
with transverse submarginal yellow band distally (sometimes interrupted
medially or entirely absent); large scattered punctures on lower l/2, becoming
smaller and interspersed with short, weak, horizontal rugulae. (3) Inter-
ocular area with deep, contiguous, moderately small punctures below ocelli,
becoming coarsely rugose above antennal sockets, becoming moderately
coarsely rugose below antennal sockets; supraclypeal protuberance with
deep, moderately small, scattered punctures and weak horizontal rugae.
(4) Vertex with moderately small, deep, contiguous punctures anteriorly and
laterally, becoming smaller and weaker between ocelli; transversely carinate
posteriorly. (5) Gena with fine, contiguous, parallel carinae extending
postero-dorsally from antero-ventral margin. (6) Malar area amber; very
short, almost absent. (7) Mandible yellow, becoming amber to ferruginous

Revision of the Bee Genus Agapostemon 467

on basal half. (8) Antenna with scape and pedicel black to brown-black;
flagellum brown to brown-black above, becoming amber below. Meso-
soma: pubescence fuscous (white tomentum between posterior lobe of
pronotum and mesepisternum) becoming black on mesonotum and metano-
tum. (9) Pronotum with lateral angle and posterior lobe acutely pointed,
lateral angle projecting antero-laterally; with conspicuous sharp carina ex-
tending postero-ventrally from apex of lateral angle; with weak irregular
horizontal rugae postero-laterally. Pubescence above lateral angle and pos-
terior lobe black. (10) Mesoscutum with conspicuous flange laterally on
anterior margin extending antero-dorsally; punctures small (slightly larger
than those of A. intermedins), very deep and contiguous, becoming rugose
antero-laterally. (11) Mesoscutellum with punctures small, deep and con-
tiguous medially, becoming widely scattered on shiny lateral area. (12)
Metanotitm coarsely and shallowly rugose with very deep scattered punc-
tures. (13) Mesepisternum moderately coarsely rugose anteriorly, becoming
more finely rugose laterally and with irregular rugae extending antero-
ventrally from anterior margin of metepisternum. (14) Metepisternum
horizontally rugulose with few interconnecting rugae. (15) Propodeum
with propodeal carina very strong and conspicuous; propodeal shield with
very strong carinae extending dorso-laterally from medial groove; dorsal
area with very strong, deep rugae extending postero-laterally from anterior
margin (irregular medially); lateral area with numerous weak horizontal
carinae anteriorly becoming fewer but larger posteriorly. (16) Wing as
in $ . (17) Tegula brown to black. (18-20) Fore, middle and hind legs
brown-black to black; pubescence black dorsally, becoming dark amber
below. Metasoma: (21) Terga black; punctures separated by about
their diameters; pubescence very short and appressed dorsally, becoming
moderately long and erect anteriorly, laterally and posteriorly; pubescence
black with narrow basal band of sparse white tomentum on terga 2-4. (22)
Sterna brown-black to black with long black hairs scattered on exposed areas.

Agapostemon boliviensis n. sp.

This species is named for the country where it is found.

The male holotype and paratype as well as the allotype are labeled,
"BOLIVIA-Beni, Rur-/ renabaque, 175 mts./ 5 October 1956/ (L. Pefia)."
These types are in the Snow Entomological Museum at the University of
Kansas.

Distribution. This species seems to be sympatric with A. mourei. In
addition to the types, I have seen specimens in the Snow Entomological
Museum with the following labels: "Cristal Mayu./ Chapare, Cochab/
amba Bol. 200 m./ XII-5-49 L E Pena" (2 6 ); same as preceding but col-
lected the next day (2 8); "BOLIVIA-Dpto. La Paz/ Alta Marani, N. of/

468 The University Science Bulletin

Rurrenbaque, 10 Nov/ 1956 (L. Pena)" (1 S). Two males in the U.S.
National Museum have the following labels: "Covendo/ Bolivia/ Wm M
Mann// Aug// Mulford/ BioEpl/ 1921-22" and "nr. mouth/ Rio Mapiri/
Sept// Mulford/ Bio Expl.,,

Diagnosis. The male may be distinguished from those of most other
species by the presence of a medial "button" at the base of its last visible
(6th) sternum (Fig. 137). Agapostemon boliviensis is distinguished from
the closely related A. inca by its lack of large hooked setae on the postero-
lateral margin of the antepenultimate visible (4th) sternum, from A. mourei
and A. inca by the smaller and more rounded medial plate of its gonostylus,
and from A. intermedins by its enlarged hind femur, tibia and basitarsus
(Fig. 170).

The female may be distinguished from most other species by the acute
lateral angle and posterior lobe of its pronotum and by the dense woolly
pubescence of its mesoscutum; from those of A. lanosus and A. heterurus by
the yellow maculations on its clypeus and by its more finely branched meso-
scutal pubescence; from those of A. mourei by its slightly coarser genal
striae and its darker pubescence.

Description

male (Figs. 137, 170, 177A, 215)

As in A. mourei but head and mesosoma bright metallic green usually
with strong, dark copper tints; antenna dark brown to black with scape
yellow below; tegula dark brown to brown-black; middle leg brown with
narrow yellow streak anteriorly on trochanter and sometimes on tibia, and
with yellow anteriorly (and usually posteriorly) on femur; hind leg (Fig.
170) with ventral ridge of basitarsus more convex; sternum 6 (Fig. 137)
with proximo-medial projection, or "button," smaller and with each lateral
flange approximately x/2 as wide as base (as measured on transverse section
through widest part of "button"); genitalia (Figs. 177A, 215)."

female (Figs. 127-128)

As in A. mourei but with sculpturing slightly coarser and pubescence
darker; metasoma brown anteriorly, becoming black posteriorly and with
yellow basally on tergum 3.

Variation. The color of the head and mesosoma of males ranges from
a bright metallic blue-green to a dark coppery luster. Furthermore, the
black markings of the appendages and metasoma are often replaced by pale
fawn markings. There is no obvious correlation between the color of the
head and mesosoma and the color of the markings on the appendages and
metasoma.

Revision of the Bee Genus Agapostemon 469

Agapostemon centratus (Vachal)

Halicttts (Agapostemon) centratus Vachal 1903. Type o* , Museum National LTHistoire
Naturelle, Paris.

A metasoma, all that remains of the holotype, is dark brown to black
with broad bands of white tomentum. I have seen two males labeled "Costa
Rica" and both have amber metasomas. The genitalia of the type are in-
distinguishable from those of the males labeled "Costa Rica." The type is
from Haiti and this species resembles the West Indian A. viridulus group in
sculpturing and genitalia. Like the similarly labeled specimens of A.
aenigma, the two males labeled "Costa Rica" are probably from the West
Indies.

Diagnosis. This species is unique in the genus inasmuch as the metasoma
resembles that of females of A. melliventris in being amber to black with
broad bands of white tomentum basally on the terga. The resemblance is
so great that Sandhouse misdetermined the two males with amber meta-
somas as females of A. melliventris.

Description

male (Figs. 121-122, 152, 182, 206)

General coloration of head and mesosoma bright metallic green, meta-
soma amber to black with conspicuous bands of white tomentum.
Pubescence of head and mesosoma white to creamy. Head (Figs. 121-
122): (1) Labrum yellow with transparent margin, evenly rounded apically
(not acute) ; basal area with conspicuous median depression. (2) Clypeus
with shallow punctures separated by nearly twice their diameters; pubescence
sparse, short. (3) Interoadar area shallowly rugose at level of antennal
sockets, becoming finely rugulose above, sparsely punctate on supraclypeal
protuberance becoming shallowly rugose laterally; pubescence nearly as
long as scape medially, becoming shorter laterally and appressed ventro-
laterally. (4) Vertex with contiguous, fine punctures, becoming rugose be-
hind ocelli; pubescence like that of interocular area. (5) Genu with weak,
discontinuous, parallel carinulae running postero-dorsally from antero-
ventral margin; pubescence dense, nearly 1.5X as long as scape postero-
medially, becoming shorter anteriorly and laterally. (6) Malar area yellow
to amber, short, with pubescence white and tomentose. (7) Mandible yellow,
apical quarter transparent and ferruginous. (8) Antenna. Flagellum dark
brown above, amber below and on shiny dorsal spot on apex of distal
fiagellomere; pedicel dark brown above, becoming light brown below;
scape yellow below, dark brown above. Mesosoma: white pubescence
moderately dense and evenly distributed (except pronotum). (9) Pronotum
with lateral angle and posterior lobe rounded, 8-12 parallel carinulae running
postero-ventrally from below lateral angle; pubescence above lateral angle

470 The University Science Bulletin

and posterior lobe like that of mesoscutum, extremely short and fine on
postero-lateral area. (10) Mesoscutum with fine, contiguous punctures be-
coming finely rugulose antero-laterally. (11) Mesoscutellum shiny with
small scattered punctures becoming larger and more numerous on lateral
margins. (12) Metanotum with very small, sparse punctures medially and
with weak, short, lateral carinulae directed postero-laterally. (13) Mesepis-
ternum coarsely but shallowly rugose anteriorly, becoming finely rugose
posteriorly. (14) Metepisternum finely, regularly and horizontally carinulate.
(15) Propodeum rounded posteriorly; propodeal carina absent or very weak
and discontinuous; weak, closely parallel carinulae running ventro-laterally
from antero-dorsal margin; propodeal shield with few, weak, irregular
carinulae running laterally and antero-laterally from mid-line. (16) Wing
transparent and colorless with radius obviously darker than other veins.
(17) Tegula transparent with metallic green on proximal margin and faint
yellow transverse band. (18-20) Fore, middle and hind legs (Fig. 152) amber
to fawn, slightly darker on dorsal surfaces of tibiae of middle and hind
legs; all coxae metallic green. Pubescence short, creamy to amber. Meta-
soma: (21) Terga transparent amber with subapical brown band on
tergum 1; brown band larger, darker and broader on terga 2 and 3 and
covering exposed areas of terga 4-7. Broad, basal band of dense white
tomentum on terga 2-5; sparse, moderately long, creamy pubescence on
anterior surface of tergum 1, ventro-lateral surfaces of terga 1-6, and
postero-dorsal surfaces of terga 5-7. (22) Sterna transparent, amber, with
short creamy pubescence on exposed areas directed posteriorly. (23)
Genitalia (Figs. 182, 206) with weak parallel carinae dorso-laterally on
gonocoxa. Ventral lobes extremely large, cupped and densely fringed with
short hairs.

Agapostemon chapadensis Cockerell

Agapostemon chapadensis Cockerell 1900. Lectotype o , Carnegie Museum.

? Agapostemon castanens Schrottky 1902b. Type $ ? Departamento de Zoologia, Sao Paulo?

Schrottky (1909b) synonymized the male of this species with A. chapadensis. Not having

seen a type I cannot dispute his judgment.
? Agapostemon azarae Holmberg 1903. Type $ .
? Agapostemon argentimts Holmberg 1903. Type $.
? Agapostemon experiendtis Holmberg 1903. Type $.

I have seen the syntypes of A. chapadensis and have labeled one male
lectoholotype, the other male lectoparatype and the female lectoallotype.
All specimens are from Chapada. I selected a male as the lectoholotype be-
cause I am unable to distinguish the females of A. semimelleus and A.
chapadensis, although the female lectoallotypes are easily distinguished.

Distribution. This species is found in the campos, or tropical grass-
lands, of southern Brasil, of Paraguay and of northeastern Bolivia. Unlike

Revision of the Bee Genus Agapostemon 471

A. semimelleus, it has not been found in the river valleys of the eastern
slopes of the Andes.

Diagnosis. The male may be distinguished from those of most other
South American species by the acute lateral angle on its pronotum and
lack of a "button" on its last visible sternum (6th). It may be distinguished
from males of A. semimelleus by the longer distal stylus of its genitalia
(Figs. 173-174) and by the greater amount of black on the posterior of its
hind femur (Figs. 171-172).

The female cannot be distinguished from that of A. semimelleus (cf.
A. semimelleus, Variation) but it may be distinguished from those
of other South American Agapostemon by the acute upturned flange on
its anterior mesoscutal margin and lack of basal yellow bands on its meta-
somal terga.

Description
male (Figs. 172-173, 219)

General color of head and mesosoma bright metallic green, metasoma
banded with black and yellow. Like male of A. semimelleus but for the
following features. Mesosoma: (18) Fore leg yellow with brown (some-
times with slight metallic tints) posteriorly on trochanter, femur and tibia.
(19) Middle leg yellow with brown on coxa, basitarsus, posterior of
trochanter and tibia, and posterior of apical x/z of femur. (20) Hind leg
(Fig. 172) yellow with upper x/2 of coxa metallic, with brown on trochanter,
on posterior and apex of femur, on posterior and antero-dorsal regions of
tibia, and on all but antero-ventral region of basitarsus. Metasoma:
(22) Sterna 2-5 brown with amber posterior margins and sometimes with
yellow on anterior margins, sternum 1 amber with metallic tints basally,
and sternum 6 amber becoming brown on anterior margin. (23) Genitalia
(Figs. 173, 219) ventral lobes very short with apical brush of long stiff
hairs. Distal stylus about twice as long as width of medial plate.

FEMALE

Like that of A. semimelleus (cf. A. semimelleus, Variation).
Agapostemon cockerelli Crawford

Agapostemon cockerelli Crawford 1901. Type $ , U.S. National Museum.
Agapostemon martini Cockerell 1927. Males only (misidentifieil).

I have examined the types of Agapostemon cockerelli Crawford and
Agapostemon jemoratus Crawford (1901) in Washington, D.C. and, con-
trary to the opinions of Sandhouse (1936) and Michener (1951), find them
specifically distinct. The males (paratypes) of Agapostemon martini Cock-
erell (1927) were misdetermined and belong in A. cockerelli. Nomia cillaba
Cameron (1902), listed by Michener {In Muesebeck, et al., 1951) as a

472 The University Science Bulletin

possible synonym of Agapostemon coc\erelli, does not belong in the genus
Agapostemon (cf. Excluded Species).

Distribution (Fig. 2). Agapostemon cockerelli occurs as far north as
Pat Creek, Park Co., Wyoming; as far south as Petlalcingo, Puebla, Mexico;
as far east as Tehuacan, Mexico; and as far west as Wickenburg, Maricopa
Co., Arizona. Most abundant at elevations of 4,500-7,500 ft. (1,372-2.286 m),
this species has been collected as high as 8,000 ft. (2,438 m) on Pine Top
Mountain, Culberson Co., Texas, and as low as 2,100 ft. (640 m) in Big
Bend National Park, Brewster Co., Texas. In Mexico it is apparently re-
stricted to the central plateau, where it has been collected as high as 7,300
ft. (2,225 m) 20 miles north of Zacatecas, Zacatecas, and as low as 1,850 ft.
(564 m) in Boquillas del Carmen, Coahuila. In the United States females
of A. cockerelli have been collected from March through October and males
from April through November. In Mexico females have been collected from
June through October and males from April through December.

Diagnosis. The male may be distinguished from males of many species
by its toothed and conspicuously inflated hind femora, the lack of an
apical stylus on its gonostylus, or by its nearly hyaline wings; from males
of A. radiatus by the basal ridge on its basitarsus; and from males of A.
femoratus by the much smaller basal ridge on its hind basitarsus (Figs.
162-163). The female may be distinguished from females of many species
by its metallic green metasomal terga, its coarsely punctate or rugose meso-
scutum without punctures of two distinct sizes, and its nearly hyaline
wings; and from females of A. femoratus by its largely punctate (rugose
only anteriorly and laterally) mesoscutum. The female cannot be mor-
phologically differentiated from females of A. radiatus' but may be dis-
tinguished by its southwestern distribution.

Description

male (Figs. 107-108, 162)

General color of head and mesosoma bright green to blue-green, meta-
soma with black and yellow bands. Head (Figs. 107-108): pubescence
white, becoming pale yellowish on vertex and sometimes on interocular
area; more dense than in A. radiatus. (1) Labrum as in A. texanus. (2-5)
Clypeus, interocular area, vertex and gena as in A. radiatus but with slightly
coarser and shallower sculpturing. (6-7) Malar area and mandible as in
A. radiatus. (8) Antenna as in A. radiatus but with underside of flagellum
pale amber to yellowish and with upper side of apical half of apical flagello-
mere pale amber to yellow. Mesosoma: pubescence white, commonly
becoming yellowish on mesonotum and metanotum. (9) Pronotum with
lateral angle and posterior lobe slightly more angular than in A. radiatus
and with sculpturing weaker than in A. radiatus. (10-15) Mesoscutum,

Revision of the Bee Genus Agapostemon 473

mesoscutelhim, metanotum , mesepisternum, metept sternum and propodeum
as in A. radiatus but with sculpturing shallower and slightly finer. (16-17)
Wing and tegida as in A. radiatus but slightly paler. (18-19) Fore and
middle legs as in A. radiatus but with brown maculations reduced (may be
absent on fore leg). (20) Hind leg (Fig. 162) as in A. radiatus but with
trochanter yellow with femur and tibia more swollen and with small basal
ridge and apical groove on basitarsus. Metasoma: (21-22) Terga and
sterna as in A. radiatus but with dark brown or black areas reduced, and
with stronger metallic tints postero-laterally on terga 3-4. (23) Genitalia
indistinguishable from those of A. jemoratus.

female (Figs. 59-60)

General coloration of head, mesosoma and metasoma bright metallic
green to blue-green. Head (Figs. 59-60) : pubescence white, commonly
with faint yellow tints on vertex. (1) Labrum as in A. texanus. (2-4)
Clypeus, interocular area and vertex as in A. radiatus but with sculpturing
slightly shallower and slightly finer. (5) Gena as in A. radiatus but with
genal carinae usually slightly coarser. (6-7) Malar area and mandible as in
A. radiatus. (8) Antenna as in A. radiatus but often paler on underside of
flagellum and with yellowish amber at base and apex of scape. Meso-
soma: pubescence white, becoming slightly yellowish to fuliginous on
mesonotum and metanotum. (9) Pronotum as in A. radiatus but with
sculpturing slightly weaker. (10-13) Mesoscutum, mesoscutelhim, metano-
tum, mesepisternum as in A. radiatus but with sculpturing slightly coarser.
(14) Metepisternum as in A. radiatus but irregularly anastomosing carinae
with greater tendency to rugose condition. (15) Propodeum as in A. radiatus
but with slightly finer rugae on propodeal shield; dorsal surface with much
shallower sculpturing and more commonly rugose or with more frequently
anastomosing carinae. (16) Wing as in A. radiatus but slightly paler. (17)
Tegula as in A. radiatus £ but slightly paler. (18-20) Fore, middle and
hind legs as in A. radiatus but with pubescence paler; with yellow markings
on fore leg and middle leg; and usually with more extensive yellow mark-
ings anteriorly on fore tibia. Metasoma: (21) Terga as in A. radiatus
but bands of white tomentum on terga 2-4 slightly broader and more dense.
(22) Sterna as in A. radiatus but rarely with weak metallic tints on
sternum 4.

Agapostemon coloradinus (Vachal)

Agapostemon coloradensis Crawford 1901. Lectotype* 9 , U.S. National Museum.
Halictus (Agapostemon) coloradinus Vachal 1903.

When Vachal placed Agapostemon in the genus Halictus, Agapostemon

* The specimen is labeled "Lectotype." Although I have not been able to find a lectotype
designation in the literature, I accept this specimen as a lectotype and formalize it by so
designating it here.

474 The University Science Bulletin

coloradensis Crawford became a junior secondary homonym of Augochlora
coloradensis Titus. Vachal renamed the former, Halictus (Agapostemon)
coloradinus. Few workers agreed with Vachal's placement of Agapostemon
as a subgenus of Halictus, but the International Code of Zoological Nomen-
clature (anonymous, 1964) states (Article 59c) that secondary homonyms
rejected before 1961 cannot be revived, contrary opinions (Cockerell, 1937a
and Michener, In Muesebeck, et al., 1951) notwithstanding.

Sandhouse (1936) included Agapostemon tyleri Cockerell and Aga-
postemon martini Cockerell in Agapostemon coloradinus (Vachal), but hav-
ing examined the types I find her decisions unjustified (cf. A. tyleri).

Distribution (Fig. 3). This is by far the rarest of the North American
species and its range is more restricted than that of any other North Ameri-
can species. I have seen fewer than 300 specimens, from the high, western
plains of South Dakota, Nebraska, Kansas, Oklahoma and Texas and from
both the plains and mountains of New Mexico, Colorado, Wyoming and
central Utah. Females have been collected from May through October and
males from August through October. Specimens have been collected as low
as 2,060 ft. (628 m) in Rush Co., Kansas and as high as 10,600 ft. (3,231 m)
at Echo Lake, Mt. Evans, Colorado.

Diagnosis. The male may be distinguished from all other North Ameri-
can species except A. tyleri by the lack of yellow maculations on its dark
brown to black metasomal sterna; from A. tyleri by its larger size, by the
brown streak on the central % of the length of the posterior surface of its
hind tibia (Figs. 159-160), and by its more slender penis valve which has a
conspicuous row of hairs on its dorsal crest (Figs. 226-227). The female
may be distinguished from many other species occurring north of Mexico
by its black metasoma and the lack of yellow on its clypeus; from A.
virescens by its finer and more numerous genal carinulae (2-3 per 0.25 mm
in A. virescens and 5-6 per 0.25 mm in A. coloradinus) ; and from A. tyleri
by its brown wings, paler tegulae and less shiny supraclypeal protuberance.
This species is very close phenetically to A. tyleri.

Description

male (Figs. 99-100, 160, 185, 199, 226)

General coloration of head and mesosoma bright metallic blue to yreenish
blue, metasoma with black and pale yellow to creamy bands. Head (Figs.
99-100): pubescence snowy white. (1) Labrum as in A. texanus but more
rounded at apex. (2) Clypeus as in A. virescens but punctures shallower
and pubescence more dense. (3) Interocular area with coarse contiguous
punctures (some specimens finely rugose around antennal sockets) ; supra-
clypeal protuberance as in A. virescens. (4) Vertex as in A. virescens. (5)
Gena as in A. virescens but with carinae much finer. (6) Malar area yellow,

Revision of the Bee Genus Agapostemon 475

never amber; short. (7) Mandible as in A. virescens. (8) Antenna dark
brown to brown-black above with apex of apical flagellomere amber;
flagellum amber below and underside of scape and pedicel yellow (scape
of some specimens from Boulder, Colorado, area black with yellow apically
on underside). Mesosoma: pubescence snowy white. (9) Pronotum as
in A. virescens but with weaker sculpturing. (10) Mesoscutum as in A.
virescens but with finer punctures and with anterior margin punctate.
(11-13) Mesoscutellum , metanotum and mesepistemum as in A. virescens
but with finer sculpturing. (14) Metepisternum rugose or with irregularly
anastomosing carinae. (15) Propodentn with propodeal carina weak, often
broadly interrupted laterally; propodeal shield finely and weakly rugulose;
finely rugose dorsally and postero-laterally becoming punctate antero-
laterally. (16) Wing as in A. virescens but paler and with darker radial
vein. (17) Tegnla as in A. virescens but slightly paler and transparent.
(18-19) Fore and middle legs as in A. virescens but with brown-black mark-
ings of tibiae and femora more extensive and sometimes with yellow apically
on ventral surface of trochanter. (20) Hind leg (Fig. 160) as in A. virescens
but tibia with brown streak on apical Y2 of posterior surface, femur with
brown streak entire length of posterior surface, trochanter usually with
yellow spot apically on antero-ventral surface, and tooth on femur directed
distally. Metasoma: (21) Terga brown-black with creamy bands on
basal halves of terga 2-5 and centrally on tergum 1 (commonly interrupted
medially on tergum 1); pubescence short dorsally on terga 1-4, black on
dark bands and white on creamy bands; moderately long and white else-
where. (22) Sterna brown to brown-black, usually with faint carina medially
on apical l/2 of sternum 6; moderately long white hairs scattered on exposed
areas. (23) Genitalia (Figs. 185, 199, 226) as in A. virescens but with
smaller fold on medial plate of gonostylus and with conspicuous hairs on
prominent dorsal crest of penis valve.

female (Figs. 49-50)

General coloration of head and mesosoma bright metallic blue to blue-
green, metasoma black with white hair bands. Head (Figs. 49-50): (1)
Labrum as in A. texanus but with distal keel slightly broader. (2-4) Clypeus,
interocular area and vertex as in A. virescens but with sculpturing slightly
finer. (5) Gena as in A. virescens but with much finer and more numerous
(5-6 per 0.25 mm) carinulae. (6) Malar area dark ferruginous to brown-
black; very short, almost absent. (7) Mandible very dark amber with apical
l/3 darker ferruginous and base almost black (with very faint metallic tints
on some specimens). (8) Antenna brown-black, lower half of flagellum
usually slightly paler brown than upper l/2. Mesosoma: snowy white
pubescence sometimes very slightly fuliginous on mesonotum and meta-

476 The University Science Bulletin

notum. (9-15) Pronotum, mesoscutum, mesoscutellum , metanotum , mese-
pisternum, metepisternum and propodeum as in A. virescens but with
sculpturing slightly shallower and finer. (16) Wing brown in contrast to
that of male; unlike that of A. splendent, not markedly darker on distal
borders; radius brown-black, other veins and pterostigma dark brown.
(17) Tegida as in A. virescens. (18-20) Fore, middle and hind legs as in
A. virescens but with pubescence basally on antero-dorsal surface of hind
leg slightly paler (usually amber but may be dark brown). Metasoma:
(21-22) Terga and sterna as in A. virescens.

Agapostemon columbi n. sp.

This species is named after Christopher Columbus who first landed in
the New World on San Salvador Island (=Watling I.).

I have seen one male and one female of A. columbi. The male is labeled,
"Watlings [sic] Is./Bahama Islands/W. W. Worthington/III. 13.1909/7
Cam. Mus./Acc. 3960." The female is labeled, "Wading Is./Bahama
Islands/W. W. Worthington/III.20.1909//Carn. Mus./Acc. 3960." The
male holotype and the allotype are both at the Carnegie Museum, Pittsburgh,
Pennsylvania.

Diagnosis. The female may be distinguished from A. viridulus, A.
obscurants, A. hispaniolicus, A. swainsonae , A. alayoi and A. hohliellus by
its metallic metasomal terga; from A. sapphirinus, A. cyaneus and A.
ochromops by its lack of milky yellowish eyes and pterostigma; from A.
viequesensis by the metallic green tints on the base of its mandible; from
A. poeyi and A. insularis by its lack of metallic green tints on metasomal
sterna 3 and 4; and from A. aenigma by the golden (rather than white)
pubescence on the undersides of its hind femur and tibia.

The male may be distinguished from A. viridulus, A. obscurants, A.
hispaniolicus, A. swainsonae and A. cubensis by its lack of an elongate
clypeal area; from A. \ohliellus and A. centratus by the tooth on its hind
femur; from A. sapphirinus, A. cyaneus and A. ochromops by its lack of
milky to yellowish eyes and pterostigma; from A. insularis, A. jamaicensis
and A. poeyi by its lack of metallic green or blue medially on metasomal
tergum 3; and from A. viequesensis by its conspicuous metallic green tints
medially on metasomal terga 4-5. Both males and females of A. columbi
are slightly larger than those of other members of the A. viequesensis
species group.

Description

MALE

General coloration of head and mesosoma bright metallic blue-green,
metasoma dark brown with narrow yellow bands. Head: as in A. vie-
quesensis but with pubescence amber on interocular area and vertex.

Revision of the Bee Genus Agapostemon 477

Mesosoma: pubescence on mesonotum and metanotum pale amber, else-
where pale yellowish to creamy. (9) Pronotum as in A. vie que sen sis but
with fewer and coarser horizontal carinae on postero-lateral area, and with
lateral angle and posterior lobe slightly more angular. (10-14) Mesoscutum,
mesoscutelhim , metanotum , mesepisternum and metepisternum as in A. vie-
quesensis but with coarser sculpturing as in A. poeyi (but not as coarse). (15)
Propodeum with sculpturing much coarser than in A. viequesensis and with
prominent carinae extending postero-laterally from posterior margin of
metanotum. (16) Wing as in A. viequesensis but slightly darker and with
darker veins and pterostigma. (17) Tegula as in A. viequesensis but with
transparent amber portion slightly darker. (18-20) Fore, middle and hind
legs as in A. viequesensis. Metasoma: pubescence as in A. viequesensis.
(21) Terga. Tergum 1 dark amber on anterior l/2, dark brown on posterior
dorsal area, and with yellow band centrally, broadly interrupted medially.
Terga 2-3 dark brown with yellow band on anterior l/3. Terga 4-7 like
2-3 but with conspicuous metallic tints on dark brown areas. (22) Sterna
as in A. viequesensis. (23) Genitalia as in A. viequesensis but with apex of
distal stylus on gonostylus slightly enlarged, but not so much as in A. poeyi.

FEMALE

General coloration of head, mesosoma and metasoma bright metallic
blue-green to blue (metasoma with purple-brown tints). Head: pubes-
cence white, only very faintly fuliginous on vertex. (1) Labrum as in A.
hispaniolicus (cf. A. viequesensis). (2-5) Clypeus, interocular area, vertex
and gena as in A. viequesensis but with sculpturing much coarser, although
not as coarse as in A. poeyi. (6-8) Malar area, mandible and antenna as in
A. viequesensis but with metallic spot at base of mandible. Mesosoma:
as in A. viequesensis but pubescence paler on mesonotum and metanotum.
(9) Pronotum with lateral angle and posterior lobe slightly more angular
than in A. viequesensis, although not so much as in A. poeyi; and with
postero-lateral horizontal carinae fewer in number and coarser than in A.
viequesensis although not as few and as coarse as in A. poeyi. (10) Meso-
scutum with punctures slightly coarser than in A. viequesensis, but not
nearly as coarsely sculptured as in A. poeyi. (11) Mesoscutelhim as in A.
viequesensis but less shiny and with punctures deeper and denser. (12-15)
Metanotum, mesepisternum, metepisternum and propodeum as in A. poeyi
but with finer sculpturing, although not as fine as in A. viequesensis. (16)
Wing as in $ . (17) Tegula as in A. viequesensis. (18-20) Fore, middle and
hind legs as in A. viequesensis but with pubescence amber, becoming brown-
black antero-dorsally on tibia and basitarsus (not as dark on latter) of hind
leg. Metasoma: as in A. viequesensis but with metallic areas bluer and
nonmetallic areas darker.

478 The University Science Bulletin

Agapostemon cubensis n. sp.

This species is named for the island of Cuba where it is found.

I have seen a single male labeled, "Cuba 5.K./S. of Pinar Rio,/ Sept.
12-23 '13." The type is in the American Museum of Natural History, New
York.

Diagnosis. The male of this species may be distinguished from A.
kohliellus and A. centratus by its toothed hind femora; from A. viridulus,
A. obscuratus and A. hispaniolicus by its pale amber and brown banded
basal metasomal terga; from A. swainsonae by its small and narrowly
grooved basitarsal ridge; and from males of A. poeyi, A. vie que sen sis, A.
cyaneus, A. sapphirinus, A. ochromops, A. columbi, A. insularis and A.
jamaicensis by its elongate clypeal area.

Description

male (Figs. 119-120, 149)

General coloration of head and mesosoma shiny, metallic blue and
green, metasoma pale amber and brown. Head (Figs. 119-120): pubes-
cence white. (1) Labrum as in A. viridulus but white and darkened only
centrally. (2) Clypeus metallic blue above, white below; elongate as in A.
viridulus; shiny and sparingly punctate medially, becoming more densely
punctate laterally; pubescence sparse. (3) Interocular area metallic blue;
shallow contiguous punctures below median ocellus, shallowly and finely
rugose just above and at level of antennal sockets, becoming shiny with
scattered punctures below antennal sockets, supraclypeal protuberance shiny
and sparsely punctate and with weak transverse carinulae; pubescence be-
tween antennal sockets about % as long as scape, becoming shorter laterally,
dorsally and ventrally; tomentose ventro-laterally, near malar area. (4)
Vertex metallic blue, changing to blue-green posteriorly; punctures very fine
(finer than in A. viridulus), shallow anteriorly and laterally, becoming
rugose posteriorly; area between ocelli sparsely and shallowly punctate;
pubescence white, not fuliginous as in A. viridulus. (5) Gena metallic blue-
green in contrast to metallic blue of clypeus, interocular area and vertex;
rugae as in A. viridulus but weaker; pubescence as long as scape postero-
medially, becoming shorter laterally, anteriorly and dorsally. (6-8) Malar
area, mandible and antenna as in A. viridulus but with white replacing
yellow. Mesosoma: pubescence as in A. viridulus but white on meso-
notum and metanotum. (9) Pronotum as in A. viridulus but blue dorso-
medially, becoming blue-green ventro-laterally. (10) Mesoscutum metallic
blue with very fine punctures, slightly shallower and less dense than in
A. viridulus and becoming weakly rugulose anteriorly. (11) Mesoscutellum
metallic blue-green, punctures as in A. viridulus but slightly shallower.
(12) Metanotum metallic blue-green, with sculpturing as in A. viridulus.

Revision of the Bee Genus Agapostemon 479

(13) Mesepisternum metallic blue-green antero-dorsally becoming green
posteriorly and ventrally, with sculpturing as in A. viridulus but slightly
coarser. (14) Metepisternum metallic blue-green dorsally, becoming green
ventrally, with sculpturing as in A. viridulus but slightly coarser. (15)
Propodeum metallic green, with sculpturing as in A. viridulus. (16) Wing
as in A. viridulus but with veins not as dark. (17) Tegula as in A. viridulus
but transparent portion amber and metallic tints blue. (18-19) Fore and
middle legs as in A. viridulus but with pale amber replacing dark brown.
(20) Hind leg (Fig. 149) with coxa and trochanter as in A. viridulus; femur
amber with white antero-ventrally, on basal % antero-dorsally, and on
postero-ventral margin between tooth and distal margin; tibia pale amber
ventrally becoming pale brown dorsally; tarsus amber. Femur less swollen
than in A. viridulus, basitarsus with narrow basal ridge much smaller than
in A. viridulus, apical groove much narrower and shallower than in A.
viridulus. Pubescence as in A. viridulus. Metasoma: (21) Terga. Ter-
gum 1 pale amber with narrow band of pale brown anteriorly on dorsal
surface; tergum 2 pale amber with central brown band darker and broader
than on preceding tergum 1; tergum 3 with broad dark brown band cen-
trally, pale amber basally, and brown apically; terga 4-7 brown-black basally,
becoming brown apically. Moderate to long white hairs scattered on ventro-
lateral margins, anteriorly on tergum 1, and posteriorly on terga 5-7; short,
dense, simple, prostrate hairs on terga 1-5 brown and inconspicuous; thin,
white inconspicuous tomentum basally on terga 1-6 (usually hidden by
preceding terga). (22) Sterna as in A. viridulus but amber with subapical,
pale brown band narrow on sternum 3, becoming broader and darker on
sterna 4-6. (23) Genitalia (examined dry) as in A. viridulus but with about
l/2 as many grooves on gonocoxa.

Agapostemon cyaneus n. sp.

The name of this species refers to the dark blue color of the female.

I have seen ten specimens (7 3,3 2 ) of A. cyaneus. The female holo-
type, the allotype and seven paratypes are labeled, "Bahamas/Crooked Isl./
1-2 Mar//Utowana/Exp. 1934," and one male paratype is labeled, "Ba-
hamas/Crooked Isl./20-II//Greenway/coll." All of the types are in the
Museum of Comparative Zoology, Harvard University, Cambridge, Mas-
sachusetts.

Diagnosis. Both males and females may be distinguished from all other
West Indian species, except A. sapphirinus and A. ochromops, by their
milky to yellowish eyes and pterostigma. The male of A. cyaneus may be
distinguished from A. sapphirinus and A. ochromops by its pedicel which
is yellow to amber below (not dark brown). The female of A. cyaneus may
be distinguished from A. ochromops by its very dark blue color, and from
A. sapphirinus by its finely carinate gena.

480 The University Science Bulletin

Description

male

General coloration of head and mesosoma dark metallic green to blue,
metasoma black with pale yellow bands. Head: darker than in A. vie-
quesensis and with paler eyes; pubescence as in A. viequesensis. (1-7) La-
brum, clypeus, interocular area, vertex, gena, malar area and mandible as in
A. viequesensis. (8) Antenna as in A. viequesensis but with black on upper
surface replaced by yellow on distal l/2 of apical flagellomere and basal Vr1/*
of scape, with underside of flagellum pale amber to yellow, and with un-
derside of scape dark brown to black. Mesosoma: pubescence as in A.
viequesensis. (9-15) Pronotum, mesoscutum, mesoscutellum , metanotum,
mesepisternum, metepisternum and propodeum as in A. viequesensis but
with very slightly coarser sculpturing. (16) Wing as in A. viequesensis but
with pale yellow pigment on pterostigma and underside of larger veins.
(17) Tegula as in A. viequesensis but with amber replaced by dark brown.
(18-20) Fore, middle and hind legs as in A. viequesensis but with brown-
black replacing amber, with yellow areas paler, with brown-black postero-
dorsally on apex and base of middle femur (and sometimes fore femur),
and with average size of basal ridge of basitarsus slightly greater. Meta-
soma: pubescence as in A. viequesensis. (21) Terga brown-black with
narrow, pale yellow bands basally on terga 2-7 (those on posterior terga
usually hidden by overlap), and centrally on tergum 1 (often interrupted
medially); apex of pygidium nearly as acute as in A. ochromops (Fig. 225).

(22) Sterna as in A. viequesensis but brown-black to, amber, and sometimes
with metallic spot medially on transverse submarginal ridge of sternum 4.

(23) Genitalia as in A. poeyi but distal stylus on gonostylus not as swollen
at apex (cf. A. viequesensis).

FEMALE

General coloration of head, mesosoma and metasoma very dark metallic
bluish purple. Head: as in A. viequesensis but with eyes whitish, pubes-
cence wholly white, mandibles brown-black with metallic tints at base and
ferruginous apically. Mesosoma: as in A. viequesensis but with pubes-
cence of mesonotum and metanotum white and fuliginous dorsally on mid-
dle tibia, hind tibia and apex of hind femur; with wings as in S ; and with
tegula brown-black with metallic tints antero-basally. Metasoma: as in
A. viequesensis but with basal bands of white tomentum on terga broader
and denser, and with sterna very dark brown with faint metallic tints.

Agapostemon erebus n. sp.

This species is named after the Greek god Erebus, personification of
darkness and brother of Night.

Revision of the Bee Genus Agapostemon 481

The female holotype is in the California Academy of Sciences, San
Francisco. It is labeled, "El Salvador/Mt. San Salvador/ VII-8-1963//M. E.
Irwin &/D. Q. Cavagnaro/Collectors." A second female is in the University
of Kansas collection labeled, "Guatemala 6.3 mi./NE. Escuintla/30 July
1966/U. Kansas Mex. Exped."

Diagnosis. A. erebus may be distinguished from the similarly colored
females of A. atrocaeruleus by its rounded lateral pronotal angle and its
coarse genal carinae. It may be distinguished from those females of A.
nasutus with dark blue head and thorax by the prominent carina running
postero-ventrally from the lateral angle of its pronotum, by the lack of
yellow on the posterior lobe of its pronotum, and by its coarse genal carinae.
It seems most closely related to A. leunculus but differs from females of this
species by its blue-black head and thorax and shiny lateral area on its
mesoscutellum.

Description

FEMALE

General coloration of head and mesosoma very dark blue to black and
only slightly metallic, metasoma black with narrow bands of white pubes-
cence dorsally. Head: (1-4) Labrum, clypeus, interocular area and vertex
as in A. nasutus but with pubescence predominantly black. (5) Gena with
coarse, parallel and anastomosing carinae and whitish pubescence as in A.
leunculus. (6) Malar area black, short. (7) Mandible yellow basally, be-
coming ferruginous on apical %• (8) Antenna as in A. leunculus. Meso-
soma: (9) Pronotum as in A. nasutus but dark blue and with prominent
carina running postero-ventrally from lateral angle. (10) Mesoscutum dark
purple-blue marginally, dull green-black centrally; sculpturing as in A.
nasutus but slightly finer; pubescence black. (11) Mesoscutellum dark
purple-black; punctation as in A. nasutus but with lateral shiny area more
shiny and conspicuous; pubescence black. (12) Metanotum dark purple-
black, rugulose; pubescence black. (13) Mesepisternum purple-black; sculp-
turing as in A. leunculus; pubescence dark fuliginous on dorsal l/}, becoming
white below. (14-15) Metepisternum and propodeum dark blue, sculpturing
as in A. nasutus but slightly coarser; pubescence white, sparse except around
and posterior to propodeal spiracle. (16) Wing transparent, not darkened
apically, radius scarcely darker than other veins. (17) Tegula black, dark
transparent brown postero-laterally. (18) Fore leg brown-black with small
yellow spot at base of tibia; pubescence of coxa and trochanter white, re-
mainder dark brown to black. (19-20) Middle and hind legs brown-black;
pubescence similar in color to that of fore leg, but scopal hairs creamy.
Metasoma: (21-22) Terga and sterna as in A. leunculus (on the holotype
the normally hidden, pale brown basal areas of terga are exposed).

482 The University Science Bulletin

Agapostemon femoratus Crawford

Agapostemon femoratus Crawford 1901, Type $ , U.S. National Museum.
Agapostemon calijornicus Crawford 1901. Females only (misidentified).

Sandhouse (1936) and Michener (1951) both considered A. femoratus
synonymous with A. cocl^erelli Crawford (1901), but after examining the
types of both species in Washington, D.C., I concluded that they are spe-
cifically distinct. The females (paratypes) of Agapostemon calijornicus
Crawford (1901) were misdetermined and belong in A. femoratus.

Distribution. Agapostemon femoratus occurs as far north as Chilcotin,
British Columbia, Canada; as far south as San Vicente, Baja California,
Mexico; as far west as the coast of California; and as far east as Williston,
North Dakota. In the northern part of the range (Washington), females
have been collected from April through October and males from June
through October. In California south of Los Angeles, females have been
collected from February through November, and males in March and May
through November. This species is equally abundant at high and low eleva-
tions. A. femoratus has been collected as high as 10,150 ft. (3,094 m) at
Blanco's Corral, Mono Co., California; 10,000 ft. (3,048 m) at Blue Ridge
Fire Lookout, Fremont Co., Wyoming; 9,200 ft. (2,804 m) at Warner
Ranger Station, Grand Co., Utah; and 8,600 ft. (2,541 m) on Hart Mountain,
Lake Co., Oregon. (See map. Fig. 4.)

Diagnosis. The male may be distinguished from many other species by
its conspicuously inflated and toothed hind femora, the large and conspicuous
apical groove on its hind basitarsus, or its nearly hyaline wings; from A.
radiatus by the large basal ridge on its basitarsus; and from A. cocherelli
by its much larger basal ridge on its hind basitarsus and much more in-
flated hind leg (Figs. 162-163). The female may be distinguished from
other species by its metallic green metasomal terga or its coarsely rugose
mesoscutum.

Description

male (Figs. 101-104, 163, 194, 202)

General color of head and mesosoma bright metallic green, metasoma
with black and yellow bands. Head (Figs. 101-104): pubescence white,
usually yellowish on vertex and often pale yellow on interocular area,
denser than in A. radiatus. (1) Labrum as in A. texanus. (2-7) Clypeus,
interocular area, vertex, gena, malar area and mandible as in A. radiatus.
(8) Antenna as in A. radiatus but with underside of flagellum pale amber
to yellowish and with upper side of distal |/2 of apical flagellomere pale
amber to yellow. Mesosoma: pubescence white, commonly becoming
yellowish on mesonotum and metanotum. (9) Pronotum as in A. radiatus
but with lateral angle and posterior lobe slightly more angular, and sculp-

Revision of the Bee Genus Agapostemon 483

turing weaker. (10-15) Mesoscutum, mesoscutellum, metanotum, mesepi-
sternum, metepistemum and propodeum as in A. radiatus, but with sculp-
turing slightly shallower. (16-17) Wing and legula as in A. radiatus but
paler. (18) Fore leg as in A. radiatus but lacking brown stripe on tibia.

(19) Middle leg as in A. radiatus but with brown tibial stripe very reduced.

(20) Hind leg (Fig. 163) as in A. radiatus but with trochanter yellow; brown
stripe on tibia reduced; femur, tibia and basitarsus swollen (femur sub-
globose) ; and basitarsus with very large, slightly sinuate basal ridge and
large broad apical groove. Metasoma: (21-22) Terga and sterna as in
A. radiatus but with dark brown or black areas reduced, and with stronger
metallic tints postero-laterally on terga 3-4. (23) Genitalia (Figs. 194, 202)
with gonocoxite partially fused with gonobase; gonostylus with large medial
plate but lacking apical and basal styli ; ventral lobe of gonocoxite moderately
large and with distal fringe of hairs.

female (Figs. 57-58)
General coloration of head, mesosoma and metasoma bright metallic
blue-green to blue. Head (Figs. 57-58): pubescence white, commonly
very pale yellow on vertex. (1) Labrum as in A. texanus. (2-4) Clypeus,
interocular area and vertex as in A. radiatus but with sculpturing slightly
deeper and more coarse. (5) Gena as in A. radiatus. (6) Malar area dark
brown to brown-black; short. (7) Mandible as in A. radiatus but with
pigment creamy and less opaque. (8) Antenna as in A. radiatus but with
flagellum slightly paler on underside. Mesosoma: pubescence white, be-
coming very faintly yellowish on some specimens. (9) Pronotum as in
A. radiatus but with slightly finer sculpturing. (10) Mesoscutum very
coarsely and deeply rugose, becoming coarsely and contiguously punctate
postero-medially. (11-15) Mesoscutellum , metanotum, mesepisternum, me-
tepistemum and propodeum as in A. radiatus but with sculpturing slightly
coarser; dorsal area of propodeum with carinae more commonly anastomos-
ing or rugose. (16) Wing as in A. radiatus. (17) Tegula as in A. radiatus
6 but much paler. (18-20) Fore, middle and hind legs as in A. radiatus.
Metasoma: (21) Terga as in A. radiatus. (22) Sterna as in A. radiatus
but darker and never with metallic tints on sternum 4.

Agapostemon heterurus Cockerell

Agapostemon heterurus Cockerell 1917 (a). Type 9, U.S. National Museum.

In addition to the type and paratype, I have seen a third specimen of A.
heterurus. All three of these specimens are females; all are in the U.S.
National Museum; all are labeled, "Palcazul Peru//Collection/ Rosenberg."

Diagnosis. The female may be distinguished from most other species by
the acute lateral angle and posterior lobe of its pronotum; from A. mourei
and A. boliviensis by the lack of yellow on its clypeus and by its less finely

484 The University Science Bulletin

branched mesoscutal pubescence; from A. lanosus by its finer genal striae
and its dark metallic blue-green head and mesosoma.

Description. Female as in A. mourei but head and mesosoma dark
metallic blue to blue-green; clypeus lacking yellow; pubescence on meso-
and metanotum less dense and woolly; tegulae dark brown; metasomal terga
often with dark brown bands centrally.

Agapostemon hispaniolicus n. sp.

This species is named for the island of Hispaniola where it occurs.

The male holotype is labeled, "Port-au-Prince/ & vie, Haiti/3-X 1934/
Darlington." The allotype is labeled, "Manneville/Haiti/Nov 16-17, '34/
Darlington." The types in the Museum of Comparative Zoology, Harvard
University, Cambridge, Massachusetts.

Distribution. In addition to the holotype and allotype, I have seen
specimens from Santo Domingo (2 9) and Samana (1 8 , June) in the
Dominican Republic; and from La Moriniere (1 2 , March, 125 ft.) and
Port au Prince (12, April, 300 ft.; 1 $ ) in Haiti.

Diagnosis. The male may be distinguished from other West Indian spe-
cies, except A. viriduhts and A. obscuratus, by its black metasomal terga and
toothed hind femora; from A. obscuratus by its bright metallic blue-green
head and mesosoma; and from A. viridulus by the contiguous punctures on
its mesoscutum. The female may be distinguished from other West Indian
species, except A. viridulus and A. swain sonae, by its bright green head and
mesosoma and black metasomal terga; from A. viridulus and A. swainsonae
by the finer sculpturing of its interocular area.

Description
male (Figs. 115-116, 148)
As in A. viridulus but with slightly finer and denser sculpturing; with
slightly narrower apical groove on basal ridge of basitarsus; and with shal-
lower and more numerous longitudinal grooves laterally on gonocoxa.

female (Figs. 37-38)

As in A. viridulus but bluer, smaller, and more finely sculptured, clypeal
area not as elongate, and with narrower distal keel on labrum.

l6

Agapostemon inca n. sp.

This species is named for the Indians of the Quechuan tribe inhabiting
Peru.

The male holotype and one male paratype are labeled, "Hda San Juan/
Col Perene Peru/ 16 June 1920/ Cornell Univ. Exp.// Cornell Univ. Ex-/
pedition Lot 607/ Sub 95." The above two specimens are in the collection
of Cornell Univ., Ithaca, N.Y. The only other specimen I have seen is a

Revision of the Bee Genus Agapostemon 485

male (paratype) in the American Museum of Natural History labeled,
"PERU: Montenegro/ Bagua, Amazonas/ Sept. 29-Oct. 2, 1963/ 350m.
Wygodzinsky."

Diagnosis. The male may be distinguished from those of most other
species by the presence of a medial "button" at the base of its last visible
(6th) sternum. This sternal "button" is much longer and broader than that
of the closely related A. mourei (Figs. 139-140), and the hind femur, tibia
and basitarsus are more slender than those of A. mourei (Figs. 167-168).
Agapostemon inca may be distinguished from the related A. intermedins,
A. boliviensis and A. lanosus by the larger and differently shaped medial
plate of its gonostylus (Figs. 175, 177 A-B, 192).

Variation. The general appearance of the males differs strikingly be-
cause the black markings of the appendages and metasoma may be replaced
by pale amber markings.

Description

male (Figs. 140, 168, 175, 218)

As in A. mourei but with head and mesosorna dark coppery green; an-
tenna dark brown to black with yellow spot at base of lower side of scape;
mesosorna with slightly finer punctures and less coarse rugae; tegula brown
to brown-black, nearly obscuring anterior, submarginal, yellow crescent;
fore leg yellow with coxa brown, trochanter brown but for anterior yellow
spot, and brown on posterior of femur and tibia; middle leg brown with
yellow spot on anterior of trochanter, yellow streak on anterior of femur,
and small yellow spot anteriorly at either end of tibia; hind leg (Fig. 168)
more slender, and dark brown with antero-dorsal yellow patch on femur;
metasomal sterna brown with yellow on sterna 2-3; proximo-medial projec-
tion, or "button," on sternum 6 (Fig. 140) very broad, base with proximal
portion wider than distal portion; genitalia (Figs. 175, 218).

Agapostemon insularis n. sp.

The name of this species refers to its island distribution.

I have seen four specimens (3 $ , 1 9 ) from the island of Hispaniola.
The female holotype and the allotype are labeled "Aux Cayes, Haiti/Mch.
15-20, '22/F. 4644." The two male paratypes are from Carrefour, Haiti
(April) and the San Francisco Mts., Dominican Republic (Sept.). The
types are in the American Museum of Natural History, New York.

Diagnosis. The male may be distinguished from other West Indian
species, except A. jamaicensis and A. poeyi, by its medial metallic tints on
metasomal tergum 3; from A. jamaicensis and A. poeyi by its lack of postero-
lateral metallic tints on metasomal tergum 1. The female may be distin-
guished from other West Indian species, except A. poeyi, by its conspicuous

486 The University Science Bulletin

medial metallic green tints on metasomal sterna 3-4; and from A. poeyi
by its finely punctate mesoscutellum.

Description

MALE

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma brown with yellow bands and tinted with green posteriorly.

Head: as in A. vieqitesensis. Mesosoma: as in A. viequesensis but with
sculpturing slightly finer, pronotum with lateral angle and posterior
lobe both slightly more angular and with fewer postero-lateral carinulae,
and tegula with metallic green tints slightly more extensive. Metasoma:
(21) Terga as in A. columbi but with conspicuous metallic tints on tergum
3; pygidium rounded apically as in A. viequesensis. (22) Sterna as in A.
viequesensis. (23) Genitalia as in A. poeyi but with apex of apical stylus
on gonostylus slightly smaller (much closer to that of A. poeyi than that
of A. viequesensis).

FEMALE

General coloration of head, mesosoma and metasoma bright metallic
greenish blue. Head: pubescence as in A. viequesensis. (1-4) Labrum,
clypeus, interocular area and vertex as in A. viequesensis. (5) Gena much
more coarsely carinate than that of A. viequesensis but not quite as coarse
as in A. poeyi. (6) Malar area absent. (7-8) Mandible and antenna as in
A. poeyi. Mesosoma: pubescence as in A. viequesensis but slightly more
fuliginous on mesonotum and metanotum. (9) Pronotum as in A. vieque-
sensis but with lateral angle and posterior lobe both more angular and with
fewer and coarser (not as coarse as in A. poeyi) horizontal carinae postero-
laterally. (10-12) Mesoscutum, mesoscutellum and metanotum as in A.
viequesensis. (13-14) Mesepisternum and metepi sternum as in A. vieque-
sensis but with sculpturing very slightly coarser. (15) Propodeum as in
A. poeyi but with finer sculpturing (not quite as fine as in A. viequesensis).
(16-17) Wing and tegula as in A. viequesensis but tegula darker with metal-
lic tints more extensive. (18-20) Fore, middle and hind legs dark brown,
coxae with strong metallic tints. Pubescence dark amber to brown, scopal
hairs white. Metasoma: as in A. viequesensis but with terga wholly
metallic, and sterna 3-4 with strong metallic tints.

Agapostemon intermedins n. sp.

The name refers to the phenotype of this species which is intermediate
between that of A. semimelleus and that of A. heterurus.

The male holotype, the allotype, and the 28 paratypes (4 $ , 24 $ ) are
all from Costa Rica: Gromaco, 34 km southeast of Potrero Grande, on the

Revision of the Bee Genus Agapostemon 487

Rio Coto Brus, southern Puntarenas Province, 21 July 1963, 1,000 ft. (C. D.
Michener and W. Kerfoot) and are in the Snow Entomological Museum
at the University of Kansas.

Distribution. In addition to the types I have seen specimens from the
following localities: Turrialba (2 <$ , 8 July; 1 $ , 31 July; 15,1 Aug.;
15,3 Aug.; 1 S , 10 Aug.; 1 S , 16 Aug.; 2 $ , 28 Aug.; 1 9 , 12 Sept.),
4.5 mi. SW Cartago (19,3 July) and Pozo Azul at the confluence of Rios
Parrita and Candelaria (19,9 Dec), Costa Rica; 10 miles east of Cumana,
Venezuela (2 9 , 13 March); Macas, Rio Upano (1 $, 25 Jan.) and
Pichilingue, Los Rios (59,2 Feb.), Ecuador; Monson Valley, Tingo Maria,
Peru (1 5, 2 Nov.; 2 6, 21 Nov.; 3 S, 29 Nov.).

Diagnosis. The male is easily distinguished from all other North and
Central American Agapostemon by its medial "button" (flattened process
resembling human tongue and attached to sternum in much the same way as
human tongue is attached to floor of mouth) at the base of the last visible
(6th) sternum (Fig. 136), and from the South American Agapostemon with
similar "buttons" on sternum 6 by the low basal ridge and inconspicuous
apical groove ventrally on its slender basitarsus (Fig. 151). The female may
be distinguished from the other North and Central American species of
Agapostemon by the basal yellowish bands on its metasomal terga 2-5 (may
be obscured by overlapping terga posteriorly), and from South American
species with yellow metasomal bands by its bright metallic green or coppery
head and mesosoma and the unbroken yellow band on its clypeus (the
yellow-banded South American females have only faint metallic tints and
nearly always have the yellow clypeal band broken medially).

Variation. The coloration of females of this species varies considerably.
While the female head and mesosoma are usually bright metallic green with
amber and fuliginous pubescence, I have collected a female from Turrialba,
Costa Rica, with bright coppery head and mesosoma and with fulvus pubes-
cence. At first I was inclined to regard it as a separate species but the lack
of morphological differences, and the slightly fulvus pubescence of the other-
wise typical males from Turrialba, together with the similar pattern of varia-
tion in the sympatric A. leunculus, lead me to regard these differences as
geographic variation.

The color of the yellow-banded metasomal terga of females varies from
pale amber (Gromaco and Pozo Azul, Costa Rica; Cumana, Venezuela; and
Pichilingue, Ecuador) to brown-black (Turrialba, Costa Rica; Cumana,
Venezuela; and Pichilingue, Ecuador). Although it is difficult to assess
such variation on the basis of so few specimens it appears that the color of
the yellow-banded metasomal terga varies both within and among popula-
tions as does that of the North American A. melliventris.

488 The University Science Bulletin

Description
male (Figs. 81-82, 136, 151, 192, 214)

General coloration of head and mesosoma bright metallic green, meta-
soma banded with black and yellow. Head (Figs. 81-82) : with pubescence
golden to fulvus on interocular area and vertex, becoming creamy on gena.
(1) Labrum as in A. texanus but with apical portion much shorter and more
acute at apex. (2) Clypens with large, scattered, shallow punctures on
lower (yellow) Y-1/^ becoming contiguous on upper (green) Y}-l/i. (3)
Interocular area with small to moderately small, deep, contiguous punctures
below ocelli, becoming shallower and medium-sized at level of antennal
sockets and still shallower below antennae; supraclypeal protuberance with
large contiguous moderately deep punctures. Pubescence between antennal
sockets Yi-7/} as long as scape, becoming shorter laterally, dorsally and ven-
trally and subtomentose ventro-laterally. (4) Vertex with small, deep, con-
tiguous punctures anteriorly and laterally, becoming shallower and slightly
larger between ocelli and transversely carinate posteriorly. (5) Gena with
parallel, subcontiguous, short (often very weak) carinae extending dorso-
lateral^ from antero-ventral margin. Pubescence about as long as scape
postero-medially, becoming shorter anteriorly, laterally and dorsally. (6)
Malar area yellow, very short, and with minute, appressed hairs. (7)
Mandible yellow with apical !/3 ferruginous to amber. (8) Antenna brown
to brown-black above, scape yellow below, pedicel and flagellum yellowish
to amber below. Mesosoma: pubescence creamy or yellowish becoming
golden to fulvus on mesonotum and metanotum. (9) Pronotum with lateral
angle and posterior lobe acutely pointed, lateral angle projecting antero-
lateral^; with conspicuous sharp carina extending postero-ventrally from
apex of lateral angle; with weak irregular rugae postero-laterally. Pubescence
above lateral angle and posterior lobe golden to fulvus, yellowish to creamy
below. (10) Mesoscutum with large conspicuous flange laterally on anterior
margin extending antero-dorsally; punctures small, very deep and contiguous,
becoming rugose antero-laterally. (11) Mesoscutellum with medial punc-
tures like those of mesoscutum, becoming slightly larger and less dense
laterally; relatively shiny area with widely separated punctures on each side
of midline. (12) Metanotum coarsely punctate to rugose medially, becoming
moderately coarsely rugose laterally. (13) Mesepistemum moderately
coarsely rugose anteriorly, becoming finely rugose posteriorly. (14) Metepi-
sternum moderately finely rugose with horizontal rugae longer and slightly
more prominent than other rugae, or with irregular horizontal carinae. (15)
Propodeum with propodeal carina usually strong and conspicuous; propodeal
shield very coarsely and deeply punctate to very coarsely rugose; dorsal area
coarsely and very deeply rugose anteriorly and medially, becoming deeply
and moderately rugose to coarsely punctate laterally; lateral area with small

Revision of the Bee Genus Agapostemon 489

and very deep subcontiguous punctures anteriorly and centrally, becoming
coarsely rugose adjacent to propodeal carina. (16) Wing hyaline or nearly
hyaline, veins dark brown. (17) Tegula pale to dark transparent amber with
broad submarginal yellow band anteriorly (partially obscured by pubes-
cence) and amber to dark brown on proximal margin. (18) Fore leg yellow
with brown streaks posteriorly on femur and tibia. (19) Middle leg yellow
with coxa brown and with brown postero-dorsally on tibia and postero-
dorsally at apex of femur (also present posteriorly on trochanter of some
specimens). (20) Hind leg (Fig. 151) yellow with upper x/2 of coxa brown
and tinted metallic green; with trochanter brown basally, becoming amber
apically; with brown apically on femur and antero-dorsally on tibia.

Metasoma: (21) Terga 2-6 dark brown to brown-black on posterior l/2
and yellow on anterior l/2 (often partially concealed by overlap), tergum
1 dark brown to brown-black with broad yellow band antero-dorsally (cen-
trally) ; very conspicuously punctate with subcontiguous punctures largest
on tergum 1, becoming progressively smaller on posterior terga. Pubescence
golden to fulvus anteriorly on tergum 1, laterally on terga 1-7 and dorsally
on terga 5-7; very short dorsally on terga 1-4, golden to fulvus on yellow
bands and dark brown to brown-black on dark bands. (22) Sterna yellow
with brown on sternum 1 and with transverse bands on posterior margins
of sterna 2-5 narrowest on sternum 2, becoming progressively wider (espe-
cially medially) posteriorly; sternum 6 with proximo-medial projection, or
"button," in form of human tongue (Fig. 136) ; pubescence limited to scat-
tered, moderately long, fulvus to golden hairs on exposed areas. (23)
Genitalia (Figs. 192, 214) with ventral lobes very short and with a brush
of large stiff hairs.

female (Figs. 47-48)

General coloration of head and mesosoma bright metallic green or cop-
pery, metasoma pale amber to black and with yellow bands. Head (Figs.
47-48) : pubescence golden to fulvus, becoming creamy to golden on gena.
(1) Labrum as in A. texanus. (2) Clypeus with large scattered punctures
on lower l/2, becoming subcontiguous on upper l/2. (3) Interocular area
with small, moderately deep, contiguous punctures below ocelli, becoming
rugose from just above antennal sockets to lower margin; supraclypeal pro-
tuberance with medium sized punctures separated by nearly twice their
diameters and with short, inconspicuous horizontal rugae. (4) Vertex with
small, moderately deep, contiguous punctures anteriorly and laterally, be-
coming subcontiguous between ocelli and transversely carinate posteriorly.
(5) Gena with fine, contiguous, parallel carinae extending postero-dorsally
from antero-ventral margin. (6) Malar area amber, very short to absent.
(7) Mandible yellow, becoming dark amber or ferruginous on apical !/3-

490 The University Science Bulletin

(8) Antenna dark brown to brown-black with underside of flagellum
amber. Mesosoma: pubescence creamy to golden, becoming golden (with
large scattered hairs dark) to fulvus (with large scattered hairs pale)
on mesonotum and metanotum. (9) Pronotum as in $ but with weak
carina extending from lateral angle to posterior lobe. (10) Mesoscutum as
in $ but with most punctures slightly smaller and shallower (large scattered
hairs in large punctures). (11) Mesoscittellum with sculpturing like that of
mesoscutum but with small punctures slightly shallower. (12) Metanotum
moderately finely rugose. (13) Mesepisternum as in $ but with sculpturing
slightly more coarse. (14) Mete pi stern um horizontally carinulate with inter-
connecting rugae. (15) Propodettm with strong propodeal carina; propodeal
shield with conspicuous carinae extending dorso-laterally from medial groove
and with fine rugae between conspicuous carinae; dorsal area coarsely and
extremely deeply rugose anteriorly and medially, becoming moderately
deeply and less coarsely rugose laterally; lateral area horizontally carinulate
with small scattered punctures anteriorly becoming very coarsely and
horizontally carinate posteriorly. (16) Wing as in S . (17) Tegula as in
S but slightly darker. (18) Fore leg with dark brown on coxa and tro-
chanter; femur dark brown, becoming yellow anteriorly and dorsally on
apical l/3; tibia and tarsus yellowish to pale amber anteriorly, becoming
brown posteriorly. Pubescence creamy on coxa, trochanter and femur;
golden on tibia and tarsus. (19) Middle leg dark brown, becoming pale
amber to yellowish anteriorly and dorsally on apical % of femur and at base
of tibia; pubescence creamy on coxa and trochanter, becoming golden on
femur, tibia and tarsus, and dark brown dorsally on tibia and tarsus. (20)
Hind leg dark brown to pale brown with metallic green dorsally on coxa.
Pubescence creamy on coxa, trochanter and femur; golden ventrally and
dark brown to brown-black dorsally on tibia and tarsus. Metasoma:
(21) Terga 2-5 pale amber to black, with yellow band on basal l/3 of each
(may be partially obscured by overlap) ; tergum 1 pale amber to black with
inconspicuous pale spot dorso-laterally. Punctures small, separated by about
Y2 their diameters. Pubescence moderately short, amber to brown dorsally
on terga 1-4 but with sparsely scattered longer and darker hairs increasing
in size and abundance posteriorly from tergum 1; medium length, creamy
to amber anteriorly on tergum 1 and laterally on terga 1-4; moderately long,
dense and dark brown to black on terga 5-6. (22) Sterna dark brown on
specimens with brown to brown-black on tergum 1, but sterna 1-3 and
base of sternum 4 amber on specimens with tergum 1 amber; metallic tints
basally on sternum 1. Long amber pubescence scattered on exposed areas.

Agapostemon jamaicensis n. sp.

This species is named for the island where it occurs.

I have seen a single male specimen labeled, "Moneage, Jamaica./Sept. 14,

Revision of the Bee Genus Agapostemon 491

1917 (497)/Harold Morrison." The holotype is in the United States Na-
tional Museum, Washington, D.C.

Diagnosis. The male may be distinguished from other West Indian
species, except A. poeyi and A. insularis, by its posterior metasomal terga
strongly tinted with metallic green medially and by its brown eyes and
translucent amber to brown pterostigma; from A. insularis by its metallic
green tints postero-laterally on metasomal tergum 1; and from A. poeyi
by its more inflated hind femur and tibia.

Description

male (Fig. 146)

General coloration of head and mesosoma bright metallic blue-green,
metasoma dark brown banded with yellow anteriorly, becoming bright
metallic green posteriorly. Head: as in A. viequesensis but with sculp-
turing very slightly coarser and with pubescence pale amber. Mesosoma:
pubescence as in A. viequesensis. (9) Pronotum as in A. viequesensis but
with lateral angle and posterior lobe slightly more angular. (10-12) Meso-
scutum, mesoscutellum and metanotum as in A. viequesensis. (13-14)
Mesepisternum and metepistemum with sculpturing coarser than in A.
viequesensis but not as coarse as in A. poeyi. (15) Propodeum as in A. poeyi
females but with sculpturing slightly finer and antero-lateral area with
large punctures replacing carinae. (16-19) Wing, tegula, fore and middle
legs as in A. viequesensis. (20) Hind leg as in A. viequesensis but with
femur and tibia more inflated, apical groove on basitarsus broader and
convex centrally, and basal ridge more prominent with broad flat area on
apical portion (Fig. 146). Metasoma: (21) Terga as in A. viequesensis
but slightly darker and with bright metallic green replacing brown band
completely on tergum 4 and partially on terga 1-3 and 5-6. (22) Sterna as in
A. viequesensis but dark amber. (23) Genitalia as in A. poeyi but with
apex of apical stylus of gonostylus more slender than that of A. poeyi (not as
slender as A. viequesensis).

Agapostemon kohliellus (Vachal)

Halictus (Agapostemon) kohliellus Vachal 1903. Type 9, Naturhistorisches Museum, Vienna.

I have examined the type of H. kohliellus and the locality label is illegible.
This species is atypical of the genus inasmuch as the labrum of the female
is trilobed (Fig. 31). For the present study it is included in Agapostemon
because it is phenetically closer to this genus than to any other. Ultimately
it may be desirable to erect a separate genus for this species.

Distribution. Other than the type, I have seen specimens from Port au
Prince (1 9, March) and Cul-de-Sac Plain (3 S, Dec.) Haiti and from
Rio Toa Baracoa, Cuba (1 S , April).

492 The University Science Bulletin

Diagnosis. The male may be distinguished from all other West Indian
species by its truncate clypeus, hind femur lacking tooth, and unique hind
basitarsus. The females may be distinguished from all other West Indian
species by the yellow band on the clypeus and by the yellow bands on the
honey-colored metasoma.

Description

male (Figs. 109-110, 141, 191, 212)

General coloration of head and mesosoma metallic green, metasoma
yellow with brown bands. Head (Figs. 109-110): pubescence white to
yellowish on clypeus, interocular area and vertex; white on gena. (1)
Labrum yellow, with apical margin transparent; apex truncate to very
slightly bilobed, basal x/i without medial depression on transverse basal ridge;
labral surface shiny and impunctate with row of small, simple submarginal
setae. (2) Clypeus yellow with metallic green on upper margin, truncate,
but not nearly as much as in A. nasutus (Figs. 85-87); flattened lower por-
tion with several small shallow punctures and nearly glabrous, narrow upper
portion with shallow scattered punctures and with short sparse pubescence.
(3) Interocular area with deep, coarse, contiguous punctures above; deeply
rugulose around antennal sockets, becoming less deeply rugulose below.
Supraclypeal protuberance small but relatively acute, shallowly and coarsely
rugose below, weakly rugulose above. Pubescence between antennal sockets
about % as long as scape, becoming slightly shorter dorsally, laterally and
ventrally; becoming short appressed tomentum ventro-laterally. (4) Vertex
with deep, contiguous, medium-sized punctures anteriorly, laterally and
between ocelli; becoming rugose posteriorly. Pubescence like that of dorsal
portion of interocular area. (5) Gena with weak rugae extending postero-
dorsally from antero-ventral margin; postero-medial pubescence slightly
longer than that between antennal sockets, becoming shorter anteriorly,
laterally and dorsally. (6) Malar area absent. (7) Mandible yellow, ferru-
ginous on apical l/?,. Row of hairs on postero-ventral margin only l/^ as long
distally as proximally. (8) Antenna brown above, but scape with yellow on
basal x/i of upper surface; scape and pedicel yellow below, flagellum amber
below. Mesosoma: pubescence white, yellowish on mesonotum and
metanotum. (9) Propodeum with lateral angle and posterior lobe rounded;
sculpturing weak and inconspicuous. (10) Mesoscutum with deep, small,
contiguous punctures becoming rugose antero-laterally. (11) Mesoscutellum
with small, deep punctures separated by at least their diameters, becoming
rugose laterally. (12) Metanotum shallowly rugose with small, deep punc-
tures more abundant than those of mesoscutellum. (13) Mesepisternum
finely rugulose anteriorly, becoming very finely rugulose posteriorly. (14)
Metepisternum horizontally carinulate. (15) Propodeum rounded posteriorly

Revision of the Bee Genus Agapostemon 493

with propodeal carina absent; small, deep punctures laterally separated by
at least their own diameters, becoming rugulose or contiguously punctate
postero-dorsally; propodeal shield shallowly rugulose. (16) Wing trans-
parent; veins and pterostigma dark brown, radius almost black. (17) Tegitla
pale transparent honey-colored with yellow on basal margin and with yellow
band almost reaching distal margin. (18-19) Fore and middle legs yellow
with coxae metallic green. (20) Hind leg (Fig. 141) yellow with trochanter
brownish, brown apically on dorsal surface of femur and amber on all but
ventral surface of tibia, and with coxa metallic green; femur without tooth;
femur and tibia moderately inflated; basitarsus broad but flat with basal
ridge narrow and gently curved apically; basitarsus lacking apical groove.
Metasoma: (21) Terga. Tergum 1 yellow, amber anteriorly and brown
apically; terga 2-6 with basal yellow band and apical brown band of equal
width; tergum 7 amber. Pubescence minute and inconspicuous dorsally,
amber on yellow bands, and black on brown bands; long brown hairs
dorsally on terga 5-7; long pale hairs ventro-laterally on terga 4-6 and
anteriorly on tergum 1. (22) Sterna amber with white hairs on exposed
areas. (23) Genitalia (Figs. 191, 212), ventral lobe short with long brush
of hairs on short inner portion.

female (Figs. 31-32)

General coloration of head and mesosoma bright metallic green, meta-
soma honey-colored with yellow bands. Head (Figs. 31-32): with short
white pubescence. (1) Labrum (Figs. 31A-B) with acute lateral lobes on
apical l/2. (2) Clypeus yellow below, green above; yellow portion shiny,
sparsely punctate, green portion with medium-sized deep, contiguous punc-
tures. (3) Interocular area rugulose, becoming finely and deeply punctate
above; supraclypeal protuberance low and gently rounded with medium-
sized punctures separated by their own diameters and with numerous
horizontal rugae; pubescence between antennal sockets only about V3 as
long as scape. (4) Vertex with fine, deep, contiguous punctures anteriorly,
becoming smaller, shallower and subcontiguous laterally and faintly rugulose
posteriorly; area between ocelli shiny with small scattered punctures. (5)
Gena with fine, contiguous rugulae extending postero-dorsally from antero-
ventral margin; pubescence postero-medially about 1.5 X as long as pubes-
cence between antennal sockets and becoming shorter anteriorly, laterally
and dorsally. (6) Malar area absent. (7) Mandible yellow with apical J/3
ferruginous; single row of hairs on postero-ventral margin long basally,
becoming Y3 as long distally. (<S) Antenna brown, flagellum yellowish to
amber below. Mesosoma: short white to yellowish pubescence. (9)
Pronotum with lateral angle and posterior lobe rounded; sculpturing very
fine, weak and irregularly rugulose ventro-laterally, becoming smooth dorso-

494 The University Science Bulletin

medially. (10) Mesoscutum with small contiguous punctures, becoming
faintly rugulose antero-laterally. (11) Mesoscutellum shiny, with scattered
large punctures and more numerous very small punctures. (12) Metanotum
sparsely punctate, irregularly and transversely rugulose. (13) Mesepistemum
with shallow, fine rugae anteriorly, becoming much finer and transversely
rugulose to carinulate posteriorly. (14) Metepi sternum with shallow, regular
transverse carinulae. (15) Propodeum with distinct but weak propodeal
carina; very fine, regular, horizontal carinulae laterally and on propodeal
shield; dorsal area with very fine carinulae extending postero-laterally from
anterior margin. (16-17) Wing and tegula as in $ . (18-20) Fore, middle
and hind legs yellowish to amber with white to yellowish pubescence be-
coming fuliginous dorsally on middle and hind tibiae. Metasoma: (21)
Terga amber to pale brown with yellow bands basally on terga 2-5; tergum
1 may have pale yellow band anteriorly on dorsal portion interrupted
medially; pubescence as in 6 . (22) Sterna amber to brown with long white
hairs scattered on exposed areas.

Agapostemon lanosus n. sp.

The name of the species refers to the woolly appearance of the mesoscutal
pubescence.

The male holotype and the allotype are labeled, "PERU:/ Monson
Valley/ Tingo Maria/ XI-21-1954// E. I. Schlinger/ &E. S. Ross collectors."
The two male paratypes were collected by Ross and Schlinger from the
same locality on "29-XI-1954" and "XII-23-1954." At present, these types
are the only known specimens and they are in the California Academy of
Sciences, San Francisco.

Diagnosis. The male may be distinguished from most other species by
the presence of a medial "button" at the base of its last visible (6th) sternum
(Fig. 138). It may be distinguished from other species with sternal "buttons"
by the presence of 2-3 long (about twice the length of adjacent setae), thick-
ened, apically hooked setae laterally on the posterior margin of the antepenul-
timate visible (4th) sternum.

The female may be distinguished from most other species by the acute
lateral angle and posterior lobe of its pronotum and by the dense woolly pu-
bescence of its mesoscutum; from A. moitrei and A. boliviensis by the lack of
yellow on its clypeus, by its less finely branched mesoscutal pubescence and
by its coarser genal sculpturing; from A. heterurus by the parallel striae
dorsally on its pronotum and by the coppery luster of its head and mesosoma.

Variation. The general appearance of the males varies strikingly be-
cause the black markings of the appendages and metasoma may be replaced
by pale amber markings.

Revision of the Bee Genus Agapostemon 495

Description
male (Figs. 138, 169, 177B, 216)

As in A. mourei but with head and mesosoma bright coppery green;
antenna dark brown to black with scape yellow below; mesoscutellum with
punctures widely scattered in central region; tegula darker amber with an-
terior, submarginal, yellow crescent partially obscured proximally; fore leg
yellow with brown on posterior of coxa, trochanter, femur and tibia and
anteriorly at base of coxa; hind basitarsus (Fig. 169) yellow with dark
brown to black on ventral carina; metasomal sterna dark brown with yellow
on sterna 2-3; sternum 6 (Fig. 138) with proximo-medial projection, or
"button," with each lateral flange slightly less than l/2 width of base (as
measured on transverse section through widest part of "button") ; genitalia
(Figs. 177B, 216).

FEMALE

As in A. mourei but clypeus without yellow band; gena with deeper
striae; meso- and metanotum with pubescence darker and slightly less
woolly; propodeum dorsally with coarse, deep striae extending postero-
laterally from anterior margin; metasoma dark brown with yellowish amber
on tergum 1 and proximo-laterally on terga 2-5.

Agapostemon leunculus Vachal

Agapostemon leunculus Vachal 1903. Lectotype 2, Museum National D'Histoire Naturelle,

Paris.
Agapostemon vulpicolor Crawford 1906 (new synonymy). Type 2, U.S. National Museum.
Agapostemon nasutus var. ater Friese 1916 (new synonymy). Type 5, whereabouts unknown.

From the 30 syntypes of Agapostemon leunculus in the Paris Museum,
I have designated a female as lectoholotype. I have seen the type of A.
vulpicolor but not that of A. nasutus ater. As noted by Michener (1954),
A. nasutus ater is not conspecific with A. nasutus. I have seen a female
"Typus" (not a type, but identified by Friese) of A. nasutus ater from the
U.S. National Museum labeled, "Costa Rica,/San Jose/1903." Friese did
not describe the male, probably because he was expecting to find something
with the "hog-nosed" aspect of A. nasutus. The only difference between
A. leunculus and A. nasutus ater is the coloration of the pubescence and
integument of the head and the dorsal surface of the thorax. I have not
found any differences among the males that can be correlated with the varia-
tion among females. Because I have seen females intermediate in coloration
(see discussion of variation) I am considering the difference between A.
leunculus and A. nasutus ater as local variation.

Distribution. Agapostemon leunculus is found in Mexico exclusive of
the central plateau, as far north along the eastern coast as ten miles south-
west of Pharr, Texas (1 2 collected by C. D. Michener) and as far north

496 The University Science Bulletin

along the western coast as Rosario, Sinaloa. It has been collected in the
mountainous regions of southern Mexico but has not been taken on the
Yucatan Peninsula. It has also been collected in the mountains of El
Salvador, Honduras, Costa Rica, Panama, Colombia and Ecuador (Fig. 5).
Although information on altitudinal distribution is scant, this species has
not been found at low altitudes (less than 1,000 m) in the hot tropics. It has
been collected from 32 m (105 ft.) at Rosario, Sinaloa, Mexico, up to 7,100
ft. (2,164 m) near Nochixtlan, Oaxaca, Mexico; from 3,300 ft. (1,006 m) at
Santiago de Puriscal, Costa Rica, up to 9,000 ft. (2,743 m) on Volcan Irazu,
Costa Rica; 7,800 ft. (2,377 m) at Popayan, Colombia, and 2,800 m (9,184
ft.) at Calacali, Ecuador. A. leunculus has been collected in February and
October in Ecuador; February in Colombia; every month but January,
March, June and November in Central America; every month but January
and November in Mexico; and on March 31 near Pharr, Texas. Undoubtedly
this bee is more widely distributed altitudinally and seasonally than is in-
dicated by these records. I have seen fewer than 300 specimens, and this
sample is probably biased because bee collectors seldom work as high as
3,000 m and because most of the specimens were collected by North Ameri-
cans, who seldom have an opportunity to collect from October through May.

Diagnosis. The male may be differentiated from males of many other
species by its slender, untoothed hind femora and rounded lateral pronotal
angle; from A. nasutus by its normal clypeal region; from A. rhopalocera
by its normal antennae; from A. peninsidaris and A. mexicanns by its lack
of enlarged subapical hairs on metasomal sternum 4; and from A. melli-
ventris by the black on the basal portion of its metasomal tergum 1. The
female may be distinguished from other species with black metasoma and
green or coppery head and mesosoma by the prominent carina extending
postero-ventrally from the rounded lateral angle of its pronotum.

Variation. The coloration of the head and dorsal mesosomal region, as
well as the pubescence thereon, is quite variable in females of this species.
Most of the females are metallic green on the head and thorax. Some of the
females from Costa Rica (Volcan Irazu, San Jose, Cartago and San Mateo)
have reddish or coppery integument on the head and the dorsal mesosomal
area. The pubescence on these areas is orange. Although nowhere else so
extreme, I have seen females from Mexico (e.g., Orizaba, Vera Cruz;
Ajijic, Jalisco; Tehuacan, Puebla; and near Oaxaca, Oaxaca) with orange
pubescence and with coppery tints on the head and the dorsal mesosomal
area. In still other areas (e.g., near Ciudad del Maiz and San Luis Potosi)
the integument is green but the pubescence orange. In those regions where
the orange of the pubescence and the coppery color is most pronounced, the
yellow band on the clypeus of females is often reduced or absent. As these
regions are relatively cool and wet, it is not surprising that the females are

Revision of the Bee Genus Agapostemon 497

more darkly pigmented. The pigmentation of the males is correlated with
that of the females. Those males with enlarged black or brown markings
are found with the darker females. As in the females, intergrades may be
found between the lightest and darkest males. It is the presnce of color
intergrades and the lack of structural differences that lead me to synonymize
A. leunciilus and A. nasutus ater. Some of the females from Costa Rica,
Colombia and Ecuador have fuliginous pubescence on the head and dorsal
mesosomal regions, but sympatric intermediates exist for this condition also.
The tegulae of those females with orange or fuliginous pubescence are
often dark brown to black, unlike the more common yellow and trans-
parent tegulae of females from warmer, drier areas.

Description

male (Figs. 83-84, 143, 193, 211)

General coloration of head and mesosoma metallic green, metasoma with
black and yellow bands. Head (Figs. 83-84): (1) Labrum as in A. texamis
but more acute apically. (2) Clypeus with small scattered punctures on
yellow portion, larger and deeper scattered punctures on green portion;
scattered yellowish hairs directed ventrally. (3) Interocular area with punc-
tures above antennal sockets contiguous, deep and nearly rugose; coarsely
and shallowly punctate at level of antennal sockets becoming coarsely,
shallowly and irregularly rugose below; supraclypeal protuberance with
coarse, contiguous punctures; pubescence pale amber. (4) Vertex with
fine, contiguous punctures becoming coarse posteriorly; pubescence pale
amber. (5) Gena with sparse, shallow carinae extending postero-dorsally
from antero-ventral margin; pubescence dense, white, longest posteriorly.
(6-7) Malar area and mandible as in A. texantts. (8) Antenna dark brown
to black above except basal y2 of scape; yellow to amber below, basal !/2 of
scape usually yellow. Mesosoma: pubescence moderately long, evenly
distributed (except pronotum), and whitish to amber. (9) Pronotutn with
weak carina running postero-ventrally from lateral angle; condition of lateral
angle and posterior lobe intermediate between A. nasutus and A. texanus.
Pubescence above and between lateral angle and posterior lobe similar to
that of mesoscutum; short, white, appressed, inconspicuous pubescence on
postero-ventral surface. (10) Mesoscutum with punctation as in A. nasutus
but slightly finer and with rugosity restricted to antero-lateral portions.
(11) Mesoscutellum with punctures slightly deeper and larger than those
of mesoscutum. (12) Metanotum coarsely punctate medially to shallowly
rugose laterally. (13) Mesepisternum with punctation as in A. nasutus.
(14) Metepisternum with predominately horizontal, irregularly anas-
tomosing rugae not as regular as in A. nasutus. (15) Propodeum coarsely
punctate antero-laterally becoming moderately rugose postero-laterally and

498 The University Science Bulletin

very coarsely rugose dorsally; propodeal carina weak, evenly rounded dor-
sally, sometimes obscured by coarse rugae; propodeal shield very coarsely
and irregularly rugose. (16) Wing transparent, not darkened apically, radius
scarcely darker than other veins. (17) Tegula as in A. nasutus but trans-
parent brown proximally and with less yellow. (18) Fore leg yellow, often
with postero-dorsal black streak on any or all segments but tarsus; pubes-
cence as in A. nasutus but yellow to amber on tibia and tarsus. (19) Middle
leg yellow with coxa brown to black, postero-dorsal brown streak on
proximal l/2 of tibia, and commonly with brown streak on part or entire
length of postero-dorsal surface; pubescence like that of fore leg but shorter
on coxa, trochanter and femur. (20) Hind leg (Fig. 143) coxa metallic green
and usually with some yellow on ventral surface; trochanter predominately
yellow, brownish dorsally to wholly brown or black; femur with brown
dorsally and posteriorly at apex; proximal l/2 of tibia with brown streak on
dorsal x/2 sometimes extending to apex; tarsus yellow to brown; pubescence
as in A. nasutus but slightly shorter and darker. Metasoma: (21) Terga
shiny; 2-6 with yellow on basal l/2, apical 2/3 black; tergum 1 black with
yellow band on antero-dorsal surface; tergum 7 yellow basally, pygidial
region brown to black; pubescence sparse, simple, moderately long, and
white on vertical portion of tergum 1; terga 2-4 with short simple hairs,
white on yellow bands and black on dark bands; terga 5-7 with moderately
long, stiff, simple, black hairs. (22) Sterna 2-5 yellow with brown bands
subapically; sternum 1 green basally, apical l/2 yellow with amber medial
stripe; pubescence white, short and simple. (23) Genitalia (Figs. 193, 211)
very simple; penis valves broad and without conspicuous prominences;
gonostylus elongate with short curved apical stylus and long medial plate of
moderate width; ventral lobes notched as in A. nasutus but smaller and with
smaller notches.

female (Figs. 45-46, 135)

General coloration of head and mesosoma dark metallic green to blue-
green, metasoma shiny black with narrow, inconspicuous basal bands of
white tomentum. Head (Figs. 45-46): (1-4) Labrum, clypeus, interocular
area and vertex as in A. nasutus but with slightly coarser sculpturing and
with pubescence white to yellow or fuliginous. (5) Gena with deep, coarse,
parallel carinae running postero-dorsally from antero-ventral margin ir-
regularly anastomosing on some specimens; pubescence white (never yellow)
and shorter than in A. nasutus. (6) Malar area ferruginous, very short.
(7) Mandible as in A. nasutus. (8) Antenna black above, dark brown be-
low, undersides of first two flagellomeres often lighter than others. Meso-
soma: metallic green, never with yellow spot on posterior lobe of prono-
tum, as in A. nasutus. (9) Pronotum (Fig. 135) as in A. nasutus but with

Revision of the Bee Genus Agapostemon 499

prominent carina running postero-ventrally from lateral angle. (10-12)
Me so scut um, mesoscuteUum and metanotum with punctation as in A.
nasutus but slightly finer and without shiny lateral area on mesoscuteUum;
pubescence more dense than in A. nasutus and white to orange or fuliginous.
(13) Mesepisternum very coarsely rugose anteriorly to moderately rugose
posteriorly; white to yellowish pubescence. (14) Metepisternum with
moderately coarse horizontal carinae anteriorly, becoming rugose pos-
teriorly; pubescence creamy. (15) Propodeum coarsely rugose dorsally and
postero-laterally, becoming finely rugose antero-laterally; propodeal carina
prominent, strongly curved dorsally and laterally; propodeal shield with
irregularly anastomosing carinulae extending laterally from mid-line. (16)
Wing as in B . (17) Tegula as in £ but may also be brown to black. (18)
Fore leg dark brown, usually with some yellow on antero-apical portion of
femur and antero-basal portion of tibia; pubescence yellowish brown ven-
trally, brown to black dorsally. (19-20) Middle and hind legs dark brown,
with pubescence brown ventrally and black dorsally. Metasoma: (21)
Terga black, shinier than in A. nasutus. Narrow basal bands of white
tomentum on terga 2-5; long, pale hairs scattered on vertical portion of
tergum 1 and lateral margins of terga 2-5; long black hairs on apical half
of terga 4-5. (22) Sterna black with long pale hairs on exposed areas.

Agapostemon melliventris Cresson

Agapostemon melliventris Cresson 1874 (not 1875). Type $, Academy of Natural Sciences,

Philadelphia.
Agapostemon fasciatus Crawford 1901 (new synonymy). Type $, U.S. National Museum.
Halictus (Agapostemon) plurifasciatus Vachal 1903 (new synonymy).
Agapostemon digueti Cockerell 1924. Type $, California Academy of Sciences, San Francisco.

Cresson described this species as new in two different publications (1874,
1875). Obviously the earlier publication is valid and the latter superfluous.

I have seen the types of Agapostemon melliventris, A. fasciatus and A.
digueti and believe they are conspecific. Agapostemon fasciatus Crawford
has been considered by most authors to be of subspecific rank because it
differs from A. melliventris sensu stricto in having females with black meta-
somal terga. I have accorded it neither subspecific status, because in many
areas (cf. Arizona, Fig. 6) it is sympatric with A. melliventris s.s., nor
specific status because intermediate forms are too common (cf. discussion
of variation).

When Vachal placed Agapostemon in the genus Halictus, Agapostemon
fasciatus became a junior secondary homonym of Halictus fasciatus Ny-
lander. Thus Vachal renamed the former Halictus (Agapostemon) pluri-
fasciatus. Few workers agreed with Vachal's placement of Agapostemon
as a subgenus of Halictus. If other workers wish to apply a name to this
dark color form they should use the epithet proposed by Vachal. Cockerell

500 The University Science Bulletin

(1937a) and Michener {In Muesebeck, et al., 1951) believed Crawford's
epithet should be used, but according to the International Code of Zoological
Nomenclature (anonymous, 1964) secondary homonyms rejected before
1961 cannot be revived (cf. Agapostemon coloradinus).

Diagnosis. The male may be distinguished from other species of
Agapostemon by its yellow metasomal terga with brown to brown-black
restricted to narrow subapical bands. The female may be distinguished
from other North American species (except A. nasutus, A. leuncitlus and
A. peninsitlaris) by the yellow band on the lower margin of its clypeus;
from A. nasutus by the lack of a yellow spot on the posterior lobe of its
pronotum; from A. leuncitlus by its lack of a prominent carina extending
postero-ventrally from the lateral angle of the pronotum; and from A.
peninsitlaris by its coarser propodeal sculpturing and less dense pubescence.
In most females of A. melliventris the metasomal terga are pale amber,
unlike the black metasomal terga of A. peninsitlaris (and some specimens
of A. melliventris).

Variation. The most striking variation is the color of the metasoma
of females (see map, Fig. 6). In females from the central plains of Okla-
homa, Kansas, Nebraska and South Dakota, the metasoma is almost always
black. From the opposite side of the range, in females from Baja California
and Sonora, it is more commonly black than amber. In the remaining por-
tion of the range the female metasoma is usually amber, but a few females
with the metasoma black or dark brown may be found in these regions.
I have seen specimens intermediate with respect to color of abdomen in all
areas, but they are rare in the central plains. The amount of yellow on the
clypeus and scape is positively correlated with the paleness of the metasoma,
those specimens from the central plains having a narrow band of yellow on
the clypeus and little or no yellow on the scape.

There is no variation in the color or color pattern of males corresponding
to that of females. The size and darkness of the brown regions on the meta-
soma, legs and antennae varies slightly within, but not among, populations.

One of the most obvious variables in the phenotype is size. As shown
in Figures 93-94, the linear dimensions of larger specimens may be nearly
twice those of smaller specimens collected at the same time and place. This
size variation is somewhat greater among males than among females and
does not appear to be influenced by geographic factors.

Description

male (Figs. 92-94, 155, 188, 203)

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma yellow with narrow brown bands; all pubescence white.

Revision of the Bee Genus Agapostemon 501

Head (Figs. 92-94): (1) Labrum as in A. texanus but with transparent
distal margin slightly wider. (2) Clypeus with medium-sized punctures
separated by about their diameters below, nearly contiguous above. (3)
Interocular area with medium-sized, moderately deep, contiguous punctures
becoming minute, shallow and sub-contiguous ventro-laterally. (4) Vertex
with medium-sized, moderately deep, contiguous punctures anteriorly, be-
coming slightly smaller and shallower laterally, and rugulose posteriorly
and between ocelli. (5) Gena with very weak, short, parallel carinulae ex-
tending postero-dorsally from antero-ventral margin. (6) Malar area yellow;
short. (7) Mandible yellow with distal third ferruginous to amber. (8)
Antenna with scape yellow (small brown spot apically on upper side on
some specimens); pedicel dark brown above, yellow below; flagellum dark
brown above (yellow at apex of last flagellomere) and yellow to pale amber
below. Mesosoma: (9) Pronotum with lateral angle and posterior lobe
rounded; low rounded vertical ridge below lateral angle and faint, parallel,
horizontal carinae below posterior lobe. (10) Mesoscutum with moderately
fine, moderately deep, contiguous punctures becoming rugose antero-laterally.
(11) Mesoscutellum with punctures like those of mesoscutum but slightly
less dense and sometimes subcontiguous laterally. (12-13) Metanotum and
mesepisternum moderately finely rugose. (14) Metepistemum with sculp-
turing as in mesepisternum but slightly coarser and horizontal. (15) Propo-
deum with propodeal carina moderately weak; propodeal shield coarsely
but shallowly rugose; dorsal area coarsely and moderately deeply rugose;
lateral area with moderately coarse rugae postero-dorsally, becoming moder-
ately finely rugose antero-ventrally. (16) Wing hyaline; radial vein brown,
other veins and pterostigma pale amber. (17) Tegnla colorless and trans-
parent with opaque yellow band submarginal anteriorly, marginal basally
and posteriorly. (18-19) Fore and middle legs yellow with coxae bright
metallic green. (20) Hind leg (Fig. 155) like fore and middle legs but with
brown spot distally on dorsal surface of femur and basally on dorsal surface
of tibia; femur without tooth and basitarsus without apical groove or basal
ridge. Metasoma: (21) Terga yellow with narrow, subapical, brown
bands and with apical !4_1/3 transparent, pale amber; tergum 1 with anterior
surface very pale amber to yellow; tergum 7 with pygidium pale amber.
Short appressed pubescence dorsally becoming moderately long anteriorly,
posteriorly and laterally. (22) Sterna yellow with faint metallic tints basally
on sternum 1; narrow brown bands subapically on sterna 2-4; broad brown
band subapically on sternum 5; and sternum 6 largely brown, with pale
amber to yellow disto-laterally; moderately short to moderately long scat-
tered hairs on exposed areas. (23) Genitalia (Figs. 188, 203) with medium-
sized and slightly elongate ventral lobe on gonocoxite fringed with hairs
and with short hairs rather dense on posterior surface.

502 The University Science Bulletin

female (Figs. 61-62)

General coloration of head and mesosoma bright metallic green to blue-
green; metasoma pale amber to black and with broad bands of white
tomentum. Head (Figs. 61-62): pubescence white, rarely tinged faintly
with pale yellow on vertex. (1) Labrum as in A. texanus. (2) Clypeus with
moderately small, scattered punctures below, becoming subcontiguous to
contiguous above. (3) Interoctilar area rugose below, becoming rugulose to
coarsely punctate above; supraclypeal area shallowly punctate to weakly
rugose. (4) Vertex contiguously and coarsely punctate to rugulose anteriorly,
becoming contiguously and less coarsely punctate laterally; area behind and
between ocelli weakly rugulose. (5) Gena with numerous moderately fine,
contiguous carinae extending postero-dorsally from antero-ventral margin.
(6) Malar area amber; very short. (7) Mandible yellow with apical %
ferruginous. (8) Antenna with scape sometimes black, but usually yellow
with brown or black on apical l/z of upper surface; pedicel brown below,
brown-black above; flagellum brown to brown-black above, pale amber to
brown below. Mesosoma: pubescence white, rarely tinged faintly with
very pale yellow on mesoscutum. (9) Pronotum as in 6 . (10) Mesoscutum
with medium-sized, moderately deep, contiguous punctures, becoming finely
rugose anteriorly and moderately rugose antero-laterally. (11) Mesoscutel-
lum with punctures like those of mesoscutum but sometimes less dense and
subcontiguous laterally. (12) Metanotum moderately finely and shallowly
rugose. (13) Mesepisternum moderately coarsely and deeply rugose. (14)
Metepisternum rugose (like mesepisternum) to horizontally carinate with
carinae anastomosing most often near posterior margin. (15) Propodeum
with strong propodeal carina; propodeal shield very faintly rugose; dorsal
area moderately to coarsely rugose; lateral area with shallow, horizontal,
anastomosing rugae, very fine to moderately fine anteriorly, becoming much
coarser and slightly deeper posteriorly. (16) Wing as in $ . (17) Tegula
as in 6 but with metallic green tints on proximal margin. (18) Fore leg
with coxa brown to brown-black, tinged with metallic green; trochanter
sometimes black, but usually pale amber with yellow on distal l/2 of anterior
surface; femur entirely yellow to brown-black with small yellow spot dorsally
at apex; tibia yellow to brown with yellow antero-dorsally on basal %; and
tarsus yellow to brown. Pubescence white to amber. (19) Middle leg like
fore leg but dark specimens with faint metallic tints on trochanter and with
yellow area on tibia reduced to small spot dorsally at base; pubescence like
that of fore leg but faintly fuliginous on darker specimens. (20) Hind leg
with coxa bright metallic green dorsally and brown, strongly tinged with
metallic green elsewhere; trochanter brown to brown-black; femur, tibia
and tarsus pale amber to brown-black. Pubescence white to pale amber, be-
coming fuliginous basally on dorsal area of tibia of darker specimens.

Revision of the Bee Genus Agapostemon 503

Metasoma: (21) Terga pale amber with small black spots laterally on
terga 3-4 to entirely black with distal margins transparent (see preceding
discussion of variation). Pubescence white, becoming pale amber around
pygidium; bands of dense tomentum on basal halves of terga 2-5, and cen-
trally on tergum 1 (often interrupted medially on tergum 1); short, ap-
pressed, inconspicuous hairs distal to bands of tomentum; moderately long
hairs anteriorly on tergum 1, ventro-laterally on terga 1-5, and dorsally on
terga 5-6. (22) Sterna pale amber to black with scattered, long, white hairs
on exposed areas.

Agapostemon mexicanus n. sp.

The name of this species refers to the country where it is found.

The male holotype and the three male paratypes are labeled, "14 mi.
S./El Arco Mine/L. Cal. VI-23-38//Michelbacher & Ross Collectors," The
allotype is labeled, "Mesquital/L. Cal. VII-28-38//Michelbacher & Ross
Collectors." All of these types are in the California Academy of Sciences,
San Francisco.

Distribution. In addition to the localities listed above I have seen
specimens from: Santa Maria Valley (3 6 , 11 Aug.), Magdalena Bay
(1 $ , 30 May), and 28 miles south of El Arco, Baja California, Mexico
(1 9,3 July); 25 miles south of Navajoa, Sonora, Mexico (1 9 , 13 Sept.);
17 miles south of Navajoa, Sonora, Mexico (1 $ , 18 Aug.).

Diagnosis. The male may be distinguished from all other North Amer-
ican species of Agapostemon by the comb of 14-20 large, evenly spaced sub-
marginal bristles on metasomal sternum 4. The female may be distinguished
from A. s plena7 ens, A. texanus, A. angelic us, A. jemoratus, A. cocker elli and
A. aenigma by its black metasoma; from A. melliventris, A. peninsularis,
A. nasutus, A. leunculus and A. intermedins by the absence of yellow on its
clypeus; from A. erebus and A. atrocaer ideas by the bright metallic green
to blue-green of its head and mesosoma; and from A. tyleri by the lack of
a smooth shiny area (with scattered punctures) centrally on its supraclypeal
protuberance.

Description

male (Figs. 90-91, 156, 189, 204)

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma banded with black and yellow. Head (Figs. 90-91):
pubescence white, becoming yellowish on interocular area and vertex. (1)
Labrum as in A. texanus but slightly more obtuse apically. (2) Clypeus
as in A. melliventris but with medial enlargement of yellow band not quite
as prominent. (3-7) Interocular area, vertex, gena, malar area and mandible
as in A. melliventris. (8) Antenna brown-black with scape and pedicel

504 The University Science Bulletin

yellow below, flagellum pale amber below. Mesosoma: pubescence
white, becoming yellowish on mesonotum and metanotum. (9-12) Prono-
tum, mesoscutum, mesoscutellum and metanotum as in A. melliventris.
(13) Mesepisternum more coarsely rugose than metanotum (especially
anteriorly). (14-15) Metepisternum and propodenm as in A. melliventris
but with finer sculpturing. (16) Wing as in A. melliventris but veins darker
brown. (17) Tegula as in A. melliventris but with metallic green tints
basally and with transparent areas very pale amber. (18-20) Fore, middle
and hind legs as in A. peninsularis but with brown-black streak on tibiae
posterior (Fig. 156). Metasoma: (21) Terga as in A. peninsularis but
with pubescence shorter and less dense. (22) Sterna 2-4 yellow with brown
submarginal band posteriorly, sterna 5-6 brown, sternum 1 with metallic
green tints basally; pubescence long and pale on brown areas; sternum 4
with low submarginal ridge posteriorly and row of 14-20 large, evenly spaced
bristles just distal to submarginal ridge. (23) Genitalia (Figs. 189, 204)
with medium-sized ventral lobe on gonocoxite fringed with hairs.

female (Figs. 63-64)

General coloration of head and mesosoma bright metallic green, meta-
soma black. Head (Figs. 63-64): as in A. melliventris but lower margin
of clypeus black instead of yellow and with pubescence becoming slightly
fuliginous at vertex. Mesosoma: pubescence white, becoming fuliginous
on mesonotum and metanotum. (9-15) Pronotum, mesoscutum, mesoscutel-
lum, metanotum, mesepisternum, metepisternum and propodenm as in A.
melliventris but with finer sculpturing. (16) Wing hyaline with veins
brown, radius much darker than other veins. (17) Tegula as in A. melli-
ventris. (18) Fore leg brown-black with faint metallic green on coxa and
with yellow to pale amber anteriorly and dorsally on tibia and apically on
femur. (19-20) Middle and hind legs brown to brown-black with hind
coxa metallic green dorsally and tinged with metallic green anteriorly.
Pubescence dark brown dorsally on tibiae. Metasoma: (21) Terga
brown-black to black. Pubescence white becoming brown-black around
pygidium; moderately long, becoming short and appressed dorsally; dense
tomentum on basal third of terga 2-5, sparse tomentum on anterior % of
tergum 1. (22) Sterna brown to brown-black with metallic green tints basally
on sternum 1; pubescence white and moderately long.

Agapostemon mourei n. sp.

This species is named after Padre J. S. Moure who has contributed so
much to knowledge of the South American bee fauna.

The male holotype, the allotype and the paratypes (16 $ , 2 2 ) are all
labeled, "Region Chapare/ Bolivia 400 M./ VIII 1950 Zischka." All of

Revision of the Bee Genus Agapostemon 505

these types are in the Snow Entomological Museum at the University of
Kansas.

Distribution. This species seems to be sympatric with A. boliviensis.
In the Snow Entomological Museum I have seen additional specimens (14

3,3 9) from the type locality and specimens labeled as follows: "BO-
LIVIA-Beni, Rur-/ renabaque, 175 mts./ 5 October 1956/ (L. Pena)" (6

(5); same as preceding but collected Oct. 17 (3 $); "BOLIVIA-Beni,
Rey-/es. 10 December/ 1956 (L. Pena)" (2 $); "BOLIVIA- Chapare,/
Tropical Reg. 400/ mts. August 1951 (Zischka)" (3 $); same as preceding
but 25 August 1950 (1 $ ) ; same as preceding but Oct. 1953 (5 6 ) ; "Cristal
Mayu./ Chapare, Cochabamba Bol. 200 m./ VIII-4-49 L E Pena" (1 8 );
"Prov. del Sara/ Bolivia, 450 m./ J. Steinbech// Ace. 4548" (16); same
as preceding but Ace. 5072 (1 6 ). In the U.S. National Museum I have
seen specimens labeled: "nr. mouth/ Rio Mapiri/ Sept// Mulford/ Bio-
Expl/ 1921-22" (35); "Covendo/ Boliv./ W M Mann// September// Mul-
ford/ BioExpl/ 1921-22" (19); "Ixiamas/ Bolivia/ W M Mann// Dec//
Mulford/ BioExpl/ 1921-22" (2$); "Rosario, Lake/ Rogagua, Boliv/ W M
Mann, Oct / 28-Nov. 9, 1921// Mulford/ Biol. Expl.,/ 1921-1922" (2 $).

Diagnosis. The male may be distinguished from most other species by
the presence of a medial "button" at the base of its last visible (6th) sternum.
This sternal "button" is narrowed basally unlike that of the closely related
A. inca (Figs. 139-140), and the hind femur, tibia and basitarsus are less
slender than those of A. inca (Figs. 167-168). Agapostemon mourei may be
distinguished from the related A. lanosits, A. boliviensis and A. intermedins
by the larger and differently shaped medial plate of its gonostylus (Figs.
176,177A-B, 192).

The female may be distinguished from most other species by the acute
lateral angle and posterior lobe of its pronotum (Figs. 132-133) and by the
dense woolly pubescence of its mesoscutum. It may be distinguished from A.
lanosits by its bright yellow clypeal band and shallower genal striae; from
A. boliviensis by its finer sculpturing and paler pubescence; from A.
heteriirus by the yellow on its clypeus and by the coppery luster of its
head and mesosoma.

Variation. The general appearance of the males varies strikingly be-
cause the black markings of the appendages and metasoma are frequently
replaced by pale amber markings.

Description

male (Figs. 127-128, 132-133, 139, 167, 176, 217)

General coloration of head and mesosoma bright metallic green to blue,
metasoma with black and yellow or amber and yellow bands. Head (Figs.
127-128) : with short creamy to yellow pubescence on interocular area and

506 The University Science Bulletin

vertex, becoming long and creamy to white on gena. (1) Labrum opaque
lemon yellow with transparent pale amber margins, % as long as wide;
anterolateral margins forming obtuse angle at apex; large transverse ridge
on basal x/3 without median depression. Punctures on crown of basal ridge
shallow and separated by several times their diameters; submarginal punc-
tures large and shallow at apex becoming smaller laterally and disappearing
just anterior to lateral margins of basal ridge. Pubescence restricted to
minute scattered hairs on crown of basal ridge and single submarginal row
of simple hairs deflected distally from anterolateral margins. (2) Clypeus
with large, shallow, sparsely scattered shallow punctures on lower (yellow)
Vi-Vi) becoming more abundant, slightly smaller and contiguous on upper
(green) x/3-x/i. (3) Interocular area with small, deep, contiguous punctures
below ocelli, becoming medium-sized at level of antennal sockets and
shallower, subcontiguous and smaller below antennae; supraclypeal pro-
tuberance with medium contiguous moderately deep punctures. (4) Vertex
with small, deep, contiguous punctures anteriorly and laterally, becoming
shallower and slightly larger between ocelli; transversely carinate posteriorly.
(5) Gena with parallel, subcontiguous, short, weak carinae extending dorso-
lateral^ from antero-ventral margin. (6) Malar area yellow, short, and
with minute hairs. (7) Mandible yellow with apical x/$-x/3 amber. (8)
Antenna dark brown to black, scape and pedicel yellow below. Mesosoma:
pubescence short, pale fulvus to yellow on mesoscutum and mesoscutel-
lum, becoming shorter and creamy- to white posteriorly and ventro-
laterally. (9) Pronotttm (Figs. 132-133) with lateral angle and posterior
lobe acutely pointed, lateral angle enlarged and projecting antero-laterally;
with conspicuous sharp carina extending postero-ventrally from apex of
lateral angle; with weak irregular rugae postero-laterally. Pubescence
above lateral angle and posterior lobe pale fulvus to yellow, creamy to white
below. (10) Mesoscutum with very large, conspicuous flange laterally on
anterior margin extending antero-dorsally; punctures very fine, deep and
contiguous. (11) Mesoscutellum with punctures slightly larger and shallower
than those of mesoscutum and subcontiguous. (12) Metanotum deeply and
moderately coarsely rugose. (13) Mesepisternum coarsely rugose anteriorly,
becoming moderately rugose posteriorly. (14) Metepisternum moderately
rugose with horizontal rugae slightly more prominent than other rugae.
(15) Propodeum with propodeal carina nearly obliterated by extremely
coarse and deep rugae posteriorly and dorsally; posterior and dorsal rugae
becoming moderately fine, deep, contiguous punctures antero-laterally. (16)
Wing sub-hyaline, tinted with amber on distal margins, veins and stigma
brown. (17) Tegula pale transparent amber with anterior, submarginal,
yellow crescent. (IS) Fore leg yellow with brown streak posteriorly on coxa,
trochanter, femur and tibia. (19) Middle leg elongate, yellow with brown

Revision of the Bee Genus Agapostemon 507

markings on posterior of trochanter, tibia and sometimes femur. (20) Hind
leg (Fig. 167) : femur sub-globose, tibia and basitarsus inflated. Coxa yellow
below, dark brown with strong metallic tints above, trochanter dark brown,
femur dark brown with yellow antero-dorsally, tibia dark brown sometimes
streaked with yellow dorsally, and basitarsus varying from dirty yellow to
dark brown. Metasoma: (21) Terga 2-6 very pale amber to dark brown-
black on posterior 2/3 and yellow on anterior l/3 (often partially concealed
by overlap), tergum 1 with amber to brown-black band on anterior margin
and brown to amber on anterior face; large, conspicuous, subcontiguous
punctures anteriorly on tergum 1 becoming smaller posteriorly on tergum
1 and progressively smaller on posterior terga. Pubescence sparse and yel-
lowish. (22) Sterna dark brown and with yellow or amber markings on
sterna 1-3; sternum 6 (Fig. 139) with proximo-medial projection, or "button,"
with each lateral flange slightly more than l/2 width of base (as measured on
transverse section through widest part of "button") ; pubescence moderately
long, scattered, and pale fulvus to yellowish on exposed areas. (23) Genitalia
(Figs. 176, 217).

female (Figs. 125-126)

General coloration dark copper with tints of metallic green. Head
(Figs. 125-126) : with pubescence fulvus, becoming creamy on genal region.
(1) Labrum dark amber to ferruginous. Basal ridge prominent, proximal
face sloping gently toward proximal margin, distal face sharply declivous
and delimited from basal area by carina; additional carina present between
basal area and distal process; distal region narrow, rounded apically and
with large central keel. Pubescence restricted to single submarginal row of
coarse, stiff, amber bristles curving anteriorly at their apices and forming
fimbria on antero-lateral margins of labrum. (2) Clypeus with shallow
medial concavity; submarginal, yellow band broadly interrupted medially;
punctures widely scattered on lower l/2, becoming sub-contiguous on upper
1/2- (3) Interocular area with fine, moderately deep, contiguous punctures
below ocelli, becoming finely rugose laterally and below; supraclypeal pro-
tuberance with moderate-size, subcontiguous punctures. (4) Vertex with
fine, moderately deep, contiguous punctures anteriorly and laterally, and
with fine, transverse striae posteriorly. (5) Gena with regular fine carinae
extending postero-dorsally from antero-ventral margin. (6) Malar area
amber to brown-black, very short, pubescence minute, white, appressed
hairs. (7) Mandible yellow with apical !/3 ferruginous. (8) Antenna dark
brown to brown-black with nagellum pale yellowish amber below. Meso-
soma: pubescence short, dense, woolly and pale ferruginous on meso-
and metanotum, becoming sparse, moderately long and slightly paler below.
(9) Pronotum like that of $ (Figs. 132-133) but with dense fulvus hair
above lateral angle and posterior lobe. (10) Mesoscutum with small, deep,

508 The University Science Bulletin

contiguous punctures. (11) Mesoscntelhtm with punctures slightly smaller
and shallower than those of mesoscutum. (12) Metanotum finely puncto-
rugulose. (13) Mesepisternum with moderately coarse, shallow rugae an-
teriorly, becoming slightly finer posteriorly. (14) Metepisternum shallowly
rugose with horizontal rugae predominant. (15) Propodeum with prom-
inent propodeal carina; propodeal shield coarsely rugose, with horizontal
rugae predominant; dorsal area coarsely and deeply rugose; lateral area
with moderately coarse, deep striae extending from propodeal carina to
anterior margin. (16) Wing as in $ . (17) Tegula as in male, but slightly
darker amber. (18) Fore leg brown with amber anteriorly on femur and
tibia as well as dorsally and posteriorly; pubescence amber, becoming brown
dorsally on tibia and tarsus. (19) Middle leg brown with amber anteriorly
on tibia and femur and posteriorly on femur; pubescence pale amber, be-
coming dark brown to brown-black dorsally on tibia and tarsus. (20) Hind
leg brown to brown-black, pubescence pale amber, becoming dark brown to
black antero-dorsally on tibia and basitarsus. Metasoma: (21) Terga
finely and contiguously punctate anteriorly with punctures becoming smaller
and shallower posteriorly; amber with yellow on basal !/3 of terga 2-5 and
brown apically on tergum 5; pubescence short and sparse, pale amber
anteriorly and dark brown posteriorly. (22) Sterna pale amber with brown
basally on sterna 1, 4-6; long, sparse, yellowish pubescence on exposed
areas.

Agapostemon nasutus Smith

Agapostemon nasutus Smith 1853. Type $, British Museum (Natural History).
Agapostemon peruvianus Cameron 1903 (new synonymy). Type $, British Museum (Natural

History).
Agapostemon nasutus gualanicus Cockerell 1912a (new synonymy). Type $, U.S. National

Museum.
Agapostemon purpureopictus Cockerell 1924 (new synonymy). Type 9, California Academy

of Sciences, San Francisco.
Agapostemon melanurus Cockerell 1949 (new synonymy). Type 9. U.S. National Museum.

I have seen the types of all the above forms except A. nasutus and A.
peruvianus, which were examined for me by Michener. There is no doubt
that they are all the same species. A. nasutus gualanicus Cockerell is an
ill-defined geographic variant which does not seem to warrant formal
recognition as a subspecies (see discussion of variation).

Distribution (Fig. 7). Agapostemon nasutus is the most common and
widespread of the tropical Agapostemon. Absent from the central plateau of
Mexico, it occurs along the eastern coast as far north as Brownsville, Texas,
and along the western coast as far north as Hermosillo, Sonora, and El Arco,
Baja California. It is widespread in southern Mexico and Central America
but has not been found in the wet Atlantic lowlands of Nicaragua, Costa
Rica and Panama. It has been collected along the coast of Venezuela as

Revision of the Bee Genus Agapostemon 509

far east as Trinidad, in the Cauca and Magdalena valleys of Colombia, and
west of the Andes as far south as lea, Peru.

It has been collected at or near sea level throughout its range and at
elevations as high as 8,500 ft. (2,591 m) at Calderon, Ecuador; 1,200 m
(3,963 ft.) at San Jose, Costa Rica; 5,000 ft. (1,524 m) at Guatemala City,
Guatemala; and 6,400 ft. (1,951 m) at Tecamachalco, Puebla, Mexico. It
has been collected in South America in every month but October, in Central
America in every month but December and in Mexico north of the Isthmus
of Tehuantepec in every month but January. Thus it seems that this species
flies most, if not all, of the year wherever it occurs. My field observations
in Costa Rica indicate that it is hard to find toward the end of the dry season,
when flowers are also scarce.

Diagnosis. Males are easily distinguished from all other species by the
unique "hog-nosed" appearance of the broad, concave, ventral portion of the
short clypeus (Figs. 85-87) and by sternum 6 which is broad with laterally
directed disto-lateral tufts of branched hairs usually visible from above.
Females, as well as males, may be distinguished from all other species with-
out acutely pointed lateral pronotal angles by a yellow or creamy spot at the
apex of the rounded posterior lobe of the pronotum (Fig. 134).

Variation. The females normally are metallic green on the head and
thorax but in the central highlands of Central America (e.g., Gualan,
Guatemala and San Jose, Costa Rica) the green is replaced by blue. This
is the form named A. nasutus gitalanicus by Cockerell. The same blue
color is also found in many females from the southern limits of the range
(e.g., Lima, Peru) and the northern limits of the range (e.g., Brownsville,
Texas and Hermosillo, Sonora) where it may be very dark purple-blue.

Normally black, the abdomen is pale amber in the majority of females
from coastal Venezuela. I have seen females with black or intermediate
abdomens from this same region although these constitute less than 10%
of the total sample (Fig. 7).

The amount of yellow on the males and females is quite variable and
is slightly correlated with the coloration of the head and mesosoma of females.
Blue females fly with males usually more extensively marked with black
or brown than are males from regions where the metallic color of the
females is green.

Description
male (Figs. 85-87, 153, 195, 221-223)

General coloration of head and mesosoma metallic green to blue-green
or bronze, metasoma yellow with black bands. Head (Figs. 85-87): (1)
Labrum yellow, 40% as long as broad, flattened basal area slightly concave,
20% as long as broad, distal area triangular, 85 % as long as basal width and
separated from it by sharp step. Distal area shiny, impunctate; basal area

510 The University Science Bulletin

with scattered punctures most abundant distally. Distal margin with row
of about 32 widely separated bristles, stout medially, finer laterally, re-
mainder of distal area glabrous; basal area with scattered, prostrate hairs
directed distally, longest and most abundant distally and forming fringe
on distal margin of basal area. (2) Clypeus with ventral area smooth and
shiny, dorsal area shiny and coarsely punctate; flattened ventral area glabrous,
at right angle to upper area which is densely covered with white to yellowish
hairs. (3) Interocular area with dense, fine punctures below vertex, rugose
laterally, coarsely punctate or rugose above antennal sockets; pubescence
dense, white to yellowish hairs nearly as long as scape and much shorter
along inner margins of eyes. (4) Vertex with dense fine punctures below
and in ocellar triangle; coarsely and shallowly rugose behind ocellar triangle.
(5) Gena with sparse, weak carinulae extending postero-dorsally from antero-
ventral margin; pubescence dense, white (never yellowish), medially about
1.5 times as long as on interocular area. (6) Malar area yellow; pubescence
sparse, white, tomentose. (7) Mandible yellow with apical third transparent
ferruginous; nearly straight, not strongly curved as in other species. Pubes-
cence on flat outer surface and posterior margin white to yellowish and
directed apically; single longitudinal row of short simple bristles on inner
surface. (8) Antenna (except scape and last two flagellomeres) dark brown
above and light brown to yellow below; scape yellow with brown on apical
V4-1/? above; last flagellomere brown with pale spot on upper side at apex;
penultimate flagellomere brown, becoming yellow or light brown on basal
V2 of underside. Mesosoma: metallic green, blue-green or bronze, with
yellow or creamy spot on apex of posterior lobe of pronotum; moderately
dense, white or creamy pubescence evenly distributed on all but anterior
and lateral surfaces of pronotum. (9) Pronotum with lateral angle and
posterior lobe rounded; 2-4 inconspicuous parallel carinae slanting rearwards
and downwards near postero-lateral margin. Tomentum on lower lateral
area white. (10) Mesoscutum densely covered with moderately deep, con-
fluent punctures, fine posteriorly, becoming shallowly rugose on anterior
lA-lA- (H) Mesoscntellum with fine, moderately deep punctures separated
by about their own diameters. (12) Metanotum moderately and shallowly
rugose medially, rugulose laterally. (13) Mesepistermtm coarsely and shal-
lowly rugose anteriorly, becoming moderately rugose posteriorly. (14)
Metepistemum with weak, parallel, interconnected, horizontal carinulae.
(15) Propodeutn coarsely and shallowly rugose laterally, rugae slightly
coarser and deeper dorso-medially; propodeal carina weak, evenly rounded
dorsally; propodeal shield weakly and irregularly carinulate. (16) Wing
pale, transparent brown, not darkened apically, radius much darker than
other veins. (17) Tegula yellow with large transparent area in center of
posterior half and transparent on posterior, distal and anterior margins.

Revision of the Bee Genus Agapostemon 511

(18) Fore leg yellow, sometimes with brown on posterior surface of coxa,
trochanter and base of femur. Pubescence on coxa, trochanter and basal l/2
of femur sparse, long, white and branched; on tarsus, tibia and apex of
femur sparse, short, simple, white to yellowish. (19) Middle leg yellow, coxa
brown; trochanter and posterior area of femur sometimes brown; pubescence
like that of fore leg. (20) Hind leg (Fig. 153) with coxa brown, becoming
green dorsally; trochanter brown; brown streak on basal % of posterior
surface of femur; tibia and tarsus yellow; on some specimens with brown
on basal % of tibia and all but anterior of femur; pubescence similar to
that of preceding legs except posterior surfaces of basitarsus and femur
which are densely covered with short, stiff, simple, yellow hairs directed
distally. Metasoma: (21) Terga 1-6 transparent apically, each with
subapical brown or black band slightly broader than transparent area (may
seem broader if abdomen is flexed and brown basal band of underlying
tergum shows through overlapping transparent margin of preceding
tergum), central yellow band almost twice as wide as subapical black band,
and narrow brown or black basal band; tergum 1 with additional yellow
basal region; tergum 7 brown or black; pygidial plate fawn to yellow. Pu-
bescence fine and short dorsally, longer laterally, at base of tergum 1, and on
posterior margins of terga 5-7; white except on dark portions of terga 5-7
where yellow or brown; terga shiny with small, shallow punctures separated
by about their own diameters. (22) Sterna 2-5 yellow, each with subapical
brown band, apical margin transparent, narrow brown basal band often
concealed by overlap of preceding sternum but sometimes so broad as to
be confluent with subapical brown band; sternum 1 brown, tinged with
green basally, paler subapically, distal and lateral margins transparent;
sternum 6 yellow with basal brown band and transparent distal margin;
yellow areas on abdominal sterna sometimes partially or completely replaced
by brown; sternum 6 very broad with disto-lateral bulge and longitudinal
medial ridge; sternum 1 with few scattered long white hairs and thin
fringe of short hairs on distal margin; sterna 2-5 with scattered long yellow
hairs subapically, distal margins fringed with short, prostrate hairs, basal %
of each with minute, simple, white hairs; sternum 6 with numerous long
yellowish hairs conspicuously branched, short medially but with long disto-
lateral tufts directed postero-laterally and usually visible from above. (23)
Genitalia (Figs. 195, 221-223) far more complex than those of other species,
and relatively large.

female (Figs. 41-42, 134)

General coloration of head and thorax metallic green to blue. Metasoma

black or pale amber, with white bands of fine pubescence. Head (Figs.

41-42): pubescence short (l/2JA as long as shortest distance between antennal

sockets) and white or yellowish on face, twice as long on gena and pure

512 The University Science Bulletin

white. (1) Labrum dark amber to ferruginous. Basal ridge prominent,
proximal face sloping gently toward proximal margin, distal face sharply
declivous, and apex of ridge with deep punctures; distal portion narrow,
abruptly rounded at apex; conspicuous median keel on antero-distal portion,
continued basally almost to basal ridge; anterior margin of keel flattened
and slightly flared laterally; single row of dark amber, wide, flat, stiff
bristles curving anteriorly at their apices forming fimbria on lateral margins
of distal portion of labrum. (2) Clypeus green above with yellow band on
lower Yi-Yi bordered with black; yellow almost entirely replaced with dark
brown or black on darker specimens; median subapical area slightly concave,
yellow band with large scattered punctures, upper green area with numerous
short horizontal carinulae or horizontally elongate punctures. (3) Inter-
ocular area coarsely and shallowly rugose, supraclypeal protuberance with
dense, long, horizontal carinulae. (4) Vertex densely and finely punctate
between, anterior to, and lateral to ocelli; coarsely and shallowly rugose be-
hind ocelli. (5) Gena with numerous weak, contiguous rugulae running
postero-dorsally from antero-ventral margin. (6) Malar area brown and
very short. (7) Mandible yellow, apical Y3 ferruginous. (8) Antenna
dark brown to black, slightly paler brown on underside of flagellum.
Mesosoma: pubescence (except on pronotum) nearly twice as long as
on face, moderately dense, white laterally and below, but may be yellowish
or fuliginous on mesoscutum and mesoscutellum of blue specimens. (9)
Pronotum (Fig. 134) with yellow to creamy spot at apex of posterior lobe;
lateral angle and posterior lobe rounded as in $ \ faint punctures on dorsal
portion and faint parallel striae running downward and rearward on lower
lateral area; pubescence white, very fine, short, and appressed on lower
lateral area. (10) Mesoscutum as in $ . (11) Mesoscutellum finely and
densely punctate anteriorly, becoming coarsely punctate posteriorly; shiny,
less densely punctate area lateral to center. (12) Metanotum rugulose. (13)
Mesepisternum coarsely rugose anteriorly, somewhat less coarsely so pos-
teriorly. (14) Metepisternum as in $ . (15) Propodeum coarsely rugose
dorsally and postero-laterally, becoming finely striate antero-ventrally;
propodeal carina prominent, strongly curved dorso-laterally, weakly convex
dorsally; propodeal shield with irregularly anastomosing carinulae running
laterally and slightly upward from mid-line. (16-17) Wing and tegula
as in $ . (18) Fore leg brown to black with yellow on anterior apical %
of femur and basal % of tibia, may be more extensive on pale specimens;
pubescence of coxa, trochanter and femur white, becoming yellow on tibia
and tarsus. (19) Middle leg similar in color and pubescence to fore leg but
pubescence on dorsal surface of tibia and tarsus fuliginous. (20) Hind leg
brown to black, pubescence similar to that of middle leg but even darker on
dorsal surface of tibia and tarsus and with pale penicillus at apex of

Revision of the Bee Genus Agapostemon 513

basitarsus. Metasoma: (21) Terga usually black but often pale amber
with black at apices of terga 4-6 in Colombian and Venezuelan specimens;
anterior vertical portion of tergum 1 with sparse white hairs, upper x/z wim
thin white tomentum; basal halves of terga 2-4 with dense white tomentum;
prepygidial fimbria and hairs adjacent to pygidium and on apical portion
of tergum 4 brown to black. (22) Sterna brown to black, often pale amber in
Colombian and Venezuelan specimens; long, scattered, whitish pubescence
on exposed portions.

Agapostemon obscuratus Cresson

Agapostemon obscurata Cresson 1869. Type $ (not 9), Academy of Natural Sciences,

Philadelphia.
Agapostemon obscurants. Dalla Torre, 1896.
Agapostemon obscuratus abjectus Cockerell 1917b (new synonymy). Type $ (not 9), U.S.

National Museum.

By peculiar coincidence the descriptions by both Cresson and Cockerell
are headed by the symbol " 9 ", but both descriptions are of males. I have
seen the types of A. obscuratus and A. obscuratus abjectus and both speci-
mens are males. The color differences described by Cockerell were not
visible to me.

The specific rank of A. obscuratus is questionable, as I am unable to find
any differences between specimens of A. viridulus and A. obscuratus other
than the replacement of metallic coloration on the former by dark green-
black on the latter. The sympatric distribution of A. obscuratus and A.
viridulus and the absence of intermediate forms seems to refute any con-
tention that they are merely subspecies. It is possible that they are a single
species with two color phases. In the absence of biological data, I choose
the simplest solution and regard A. obscuratus as specifically distinct from
A. viridulus.

Distribution. I have seen specimens from Habana, Cuba (Jan., 4 8 ;
Feb., 2 $ ; April, 1 6 ; Sept., 1 8 ; Dec, 3 i ; 15 £ and 2 9 , no date).

Diagnosis. Both sexes may be distinguished from all other West Indian
species of Agapostemon by their total lack of metallic coloration.

Description. As in A. viridulus but for the color differences mentioned
above.

Agapostemon ochromops n. sp.

The name of this species refers to the pale yellowish eyes of both sexes.

The male holotype is labeled, "Bahamas/Mariguana [Mayaguana?] /22-
II//Greenway/coll." The female allotype and paratype are labeled, "Ba-
hamas/Mariguana [Mayaguana?] /25-II // fls. Ctf/^7<?//Fairchild/coll." All
of these types are in the Museum of Comparative Zoology, Harvard Univer-
sity, Cambridge, Massachusetts.

Distribution. Aside from the type series listed above, I have seen

514 The University Science Bulletin

specimens collected on the following Bahamian islands: Rum Cay (Feb.; 23
$ , 3 9 ), Cat I. (March; 1 9 ), Concepsion [Conception?] I. (Feb.; 9 S ,
1 9 ) and New Providence I. (1 S ).

Diagnosis. Both sexes may be distinguished from A. centratus, A.
/{ohlielhis, A. obscurants, A. hispaniolicus, A. viridulus, A. swainsonae,
A. cubensis, A. alayoi, A. insularis, A. jamaicensis, A. poeyi, A. viequesensis,
A. columbi and A. aenigma by their pale milky eyes and creamy to yellow
pigment on the underside of the pterostigma. The female is metallic green
in contrast to the metallic dark blue of A. sapphirinus and the metallic dark
purple of A. cyaneus. The male may be distinguished from A. cyaneus by
the pale yellow to amber of the underside of its pedicel but is indistinguish-
able from males of A. sapphirinus.

Variation. The anterior surface of the first metasomal tergum of the
allotype is honey-colored with an interrupted yellow band. On the other
males the honey-color is replaced by light to dark brown. The brown on the
legs is lightest on the allotype and darker on the other males. Metallic
color varies in intensity and varies from blue-green to green both between
and within island populations. The genal carinulae of females are finest
and most numerous on the specimen from* Conception I., most coarse and
fewest on specimens from Rum Cay, and intermediate on the holotype and
paratype from Mariguana (Mayaguana?) I.

This is the most heterogeneous of the Bahamian species and it may be a
composite species. Only more collections from the Bahamas can resolve
the problem.

Description
male (Fig. 225)

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma brown to amber banded with yellow. Head: as in A.
viequesensis but with eyes pale milky to yellowish, and ferruginous apical
portion of mandible slightly reduced. Mesosoma: as in A. viequesensis
but with slightly coarser sculpturing, with creamy to yellow pigment on
pterostigma and on underside of larger wing veins. Metasoma: as in
A. viequesensis but pygidium with apex more acute (Figs. 224-225), meta-
somal terga 3-5 of the holotype and a specimen from Rum Cay with faint
metallic tints and with pale bands slightly paler and dark bands slightly
darker than in A. viequesensis, and genitalia with apex of distal stylus of
gonostylus more spoon-shaped than in A. viequesensis (but not so much
as in A. poeyi).

FEMALE

Like A. viequesensis but with eyes milky to yellowish; with creamy to
yellow pigment on pterostigma and underside of larger wing veins; and with

Revision of the Bee Genus Agapostemon 515

sculpturing less coarse than in A. poeyi, but coarser than in A. vie que sen sis
(closer to the latter), and somewhat more variable than in other West
Indian species (even within island populations).

Agapostemon peninsularis n. sp.

The name of this species refers to the Baja California peninsula where
it is found.

The male holotype, the allotype and nine male paratypes are all labeled,
"Mgdlena Bay/Lower Cal./May 30, 1925//H. H. Keifer/Collector" and
are all in the California Academy of Sciences, San Francisco.

Distribution. In addition to the type locality above I have seen specimens
from the following localities: Ensenada (1 9 , 31 May; 1 $ , 24 June; 3 $ ,
30 Aug.), Camalu (1 2 , 23 June), 3 miles south of San Quintin (3 9 , 8-12
March), 20 miles west of San Augustine (1 S , 24 Sept.), and 20 miles north
of Mesquital (1 6 & 1 9, 27 Sept.), Baja California, Mexico; San Diego
(IS) and La Jolla (1 S ,16 July), California.

Diagnosis. The male may be distinguished from most species by its
lack of a prominent hind femoral tooth and by the black streak extending
from the base of the posterior surface of its hind femur to the apex of the
posterior surface of its hind tibia; and from A. mexicanus by the lack of
enlarged submarginal bristles medially on its 4th metasomal sternum (only
two to four submarginal bristles are present on each side). The female
may be distinguished from many species by its black metasoma, bright
green head and thorax, transverse yellow clypeal band, and rounded lateral
pronotal angle; from A. tyleri by the lack of a smooth shiny (with scattered
punctures) area centrally on its supraclypeal protuberance; from A. nasutus
by the absence of a creamy spot on the apex of its pronotal lobe; from A.
leunculus by the white to creamy pubescence (brown to black in A.
leunculiis) around the basitibial plate on its hind leg; from A. melliventris
by the finely rugulose dorso-medial region of its propodeum (rugose in
A. melliventris').

Description

male (Figs. 88-89, 157, 190, 205)

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma banded with black and yellow. Head (Figs. 88-89):
pubescence white. (1) Labrum as in A. texanus but more obtuse apically.
(2) Clypeus as in A. melliventris but with transverse yellow band slightly
narrower and with medial enlargement only about l/2 as great. (3-5) In-
terocular area, vertex and gen a as in A. melliventris but with sculpturing
slightly coarser. (6-7) Malar area and mandible as in A. melliventris. (8)

516 The University Science Bulletin

Antenna with scape and pedicel brown-black above and yellow below to
entirely brown-black; flagellum brown-black above, becoming pale amber to
yellow below. Mesosoma: pubescence white. (9-12) Pronotum, meso-
scutum, mesoscutellum and metanotum as in A. melliventris. (13) Mesepi-
sternum more coarsely rugose than metanotum (especially anteriorly). (14)
Metepisternum with interconnected horizontal carinulae, becoming rugose
posteriorly in some specimens. (15) Propodeum with propodeal carina
usually absent but may be visible as very weak carina dorsally; propodeal
shield very coarsely and shallowly rugose; dorsal area moderately rugose
laterally, becoming very finely and shallowly rugose medially; lateral area
coarsely and shallowly rugose posteriorly, becoming moderately finely rugose
anteriorly. (16) Wing as in A. melliventris but with veins darker brown.
(17) Tegula pale transparent amber with submarginal yellow band anteriorly
and posteriorly, and with strong metallic tints basally. (18) Fore leg with
coxa metallic green; trochanter brown-black with metallic green tints basally
and apical yellow spot anteriorly; femur yellow with broad brown-black
streak postero-dorsally extending from base almost to apex; tibia yellow
with brown-black streak ventrally; tarsus yellow. (19) Middle leg as fore
leg but with brown-black ventral streak on tibia extending to posterior sur-
face. (20) Hind leg (Fig. 157) as fore leg but with yellow spot on trochanter
ventral; with brown to brown-black streak on femur posterior, extending
to apex, and extending dorsally at apex; tibia with brown to brown-black
streak posterior and extending dorsally at base and apex. Femur without
distinct tooth and basitarsus without apical groove or basal ridge. Meta-
soma: (21) Terga 2-6 black with yellow band on basal l/z (usually
partially obscured by overlapping terga posteriorly) ; tergum 1 black with
yellow band centrally; terga 1-6 with posterior and ventro-lateral margins
transparent; tergum 7 with pygidium dark amber to brown-black. Pubes-
cence white, dark brown on postero-dorsal Y2 of terga 1-4; moderately long,
becoming short and appressed dorsally on terga 1-4. (22) Sterna 2-5 yellow
with brown-black bands basally and anteriorly (these bands very narrow
on sternum 2, becoming progressively broader on posterior sterna until
yellow nearly absent on sternum 5), sternum 1 with strong metallic tints,
sternum 6 brown-black with small medial ridge in shallow depression on
apical Y2. Pubescence long and pale on exposed areas, sternum 4 with 2-4
large hooked bristles at each postero-lateral margin. (21) Genitalia (Figs.
190, 205) with medium-sized ventral lobe on gonocoxite fringed with hairs.

female (Figs. 65-66)

General coloration of head and mesosoma bright metallic green, meta-
soma black. Head (Figs. 65-66): as in A. melliventris but with transverse
band on clypeus narrower and bordered with brown-black, and with scape

Revision of the Bee Genus Agapostemon 517

entirely brown-black. Mesosoma: pubescence white, becoming creamy
on mesonotum and pronotum. (9-14) Pronotum, mesoscutum, mesoscutel-
lum, metanotum, mesepisternum and metepisternum as in A. melliventris.
(15) Propodeum with weak propodeal carina; propodeal shield with large
but very shallow rugae; dorsal area finely rugose laterally, becoming very
finely rugulose medially; lateral area with shallow, anastomosing horizontal
carinulae anteriorly, becoming shallowly rugose posteriorly. (16) Wing
as in 6 . (17) Tegula as in A. melliventris. (18-19) Pore and middle legs
with coxa dark brown tinged with metallic green; trochanter dark brown;
femur dark brown with yellow apically; tibia yellow with dark brown
streak ventrally; tarsus yellow. (20) Hind leg brown, coxa metallic green
dorsally. Metasoma: pubescence white. (21) Terga brown-black to
black. Pubescence moderately long, becoming short and appressed dorsally
on terga 1-4; dense tomentum on basal !/2 of terga 2-5 and on all but posterior
!/3 of tergum 1. (22) Sterna brown-black to black; moderately long pubes-
cence on exposed areas.

Agapostemon poeyi (Lucas)

Andrena poeyi Lucas 1856. Type 2, Gribodo collection, Museo Civico di Storia Naturale,

Genoa.
Agapostemon jestivus Cresson 1865 (new synonymy). Type $, Academy of Natural Sciences,

Philadelphia.
Halictus (Agapostemon) chalets Vachal 1903 (new synonymy). Type $, Naturhistorisches

Museum, Vienna.

The "type" of Andrena poeyi has been examined by C. D. Michener.
He states that it is labeled "Andrena poeyi Guerin Cuba Type" but notes
that the word "type" is not in Guerin's hand. While this specimen may
not be the type, I am regarding it as such until such time as a more likely
candidate is found. The description coupled with the type locality leave
little doubt as to the identity of this species. I have seen the type of A.
jestivus and it is obviously the male of A. poeyi. I have seen the type of
H. chalets and it is unquestionably conspecific with A. poeyi.

Distribution. I have seen about 70 specimens from Cuba; males have
been collected in all months but January, June, October and December and
females have been collected in January, April, June, September and Novem-
ber. I have also seen specimens from Guaiameti, Santo Domingo, Hispaniola
(IS, July); Arthurs Town, Cat I., Bahamas (19, July-Aug.); Mangrove
Cay [part of Andros I.], Bahamas (19,1 Aug.); Nassau, New Providence
I, Bahamas (5 9 , 12 Aug.; 1 9 , Nov.; 1 6 , 16 Dec; 7 6 & 3 9 ).

I have also seen three males of this species labeled "Mexico-Veracruz,/
Nogales/April 1956/ (N. L. H. Krauss)." In his revision of the banded sub-
genera of Nomia, D. W. Ribble (1965) noted specimens of the Cuban Nomia
robinsoni with these label data and questioned the correctness of the labels.

518 The University Science Bulletin

In the face of this additional evidence it seems almost certain that the
labels are incorrect.

Diagnosis. Both sexes may be distinguished from those of A. cyanens,
A. sapphirinus and A. ochromops by their lack of milky to yellowish eyes
or pterostigma. The male may be distinguished from A. columbi and A.
viequesensis by its strong metallic green to blue tints medially on metasomal
tergum 3; from A. insularis by its metallic blue to green tints postero-
laterally on tergum 1; and from A. jamaicensis by its narrower hind femora.
The female may be distinguished from A. centratus, A. alayoi, A. obscuratus,
A. hispamohcus, A. viridulus and A. swainsonae by its bright metallic green
metasomal terga; from A. viequesensis by the basal green spot on its man-
dible; from A. columbi and A. aenigma by its conspicuous metallic green
tints medially on metasomal sterna 3-4; from A. insularis by its rugose area
adjacent to the parapsidal line.

Description

male (Figs. 144, 186, 210)

General coloration of head and mesosoma bright metallic green to blue-
green, metasoma dark brown with yellow bands. Head: color and shape
as in A. viequesensis; sculpturing very slightly coarser than in A. vieque-
sensis; pubescence of clypeus, interocular area and vertex yellowish to pale
amber, that of gena white. Mesosoma: pubescence as in A. viequesensis.
(9) Pronotum with lateral angle and posterior lobe more angular than in
A. viequesensis; carinulae on postero-lateral area larger and fewer in number
than in A. viequesensis. (10-12) Mesoscutum, mesoscutellum and metano-
tum as in A. viequesensis but with very slightly coarser sculpturing. (13-15)
Mesepisternum, metepi sternum and propodeum as in A. viequesensis but
with much coarser sculpturing. (16) Wing as in A. viequesensis but with
veins darker. (17) Tegula with dark amber to brown almost obscuring yellow
band, metallic green tints antero-basally slightly more extensive than in
A. viequesensis. (18-19) Fore and middle legs with coxae metallic green
(only weakly tinted on middle leg), trochanters brown with amber to
yellow on anterior surface, femora yellow with apical, brown postero-dorsal
spot, tibiae yellow with brown postero-dorsally, and tarsi yellow to pale
amber (some specimens from New Providence I. with brown areas on fore
leg reduced) ; pubescence slightly darker than that of A. viequesensis. (20)
Hind leg (Fig. 144) as in A. viequesensis but with brown areas much
darker, with femur, tibia and basitarsus less swollen, and with ridge and
apical groove on basitarsus narrower and less prominent. Metasoma:
(21) Terga as in A. viequesensis but with brown areas darker, with apex of
pygidium nearly as acute as that of A. ochromops (Figs. 224-225), and with
metallic green replacing brown completely on terga 4-6 and partially on 2-3.

Revision of the Bee Genus Agapostemon 519

(22) Sterna as in A. vie que sen sis but slightly darker and with small metallic
medial green spot on transverse submarginal ridge of sternum 4. (23)
Genitalia (Figs. 186, 210) with apex of apical stylus on gonostylus enlarged
like bowl of spoon. Ventral lobe large, with fringe of hairs on lower margin.

female (Figs. 8, 10, 12, 14, 16)

General coloration of head, mesosoma and metasoma bright green to
blue-green. Head: pubescence as in A. viequesensis. (1) Labrum as in
A. hispaniolicus. (2) Clypeus with broad, deep punctures on brown-black
portion becoming deeply rugose on upper l/2. (3) Interocular area deeply
rugose, with rugae largest just above antennal sockets; supraclypeal pro-
tuberance rugose. (4) Vertex finely rugose anteriorly, becoming coarsely
rugose laterally and posteriorly. (5) Gena with less than ten extremely
large rugae extending postero-dorsally from antero-ventral margin (Fig. 16).
(6) Malar area as in A. viequesensis. (7) Mandible as in A. viequesensis
but with basal area darkened and with metallic basal spot. (8) Antenna
brown-black, slightly paler on underside of flagellum, apex of apical flagel-
lomere sometimes conspicuously paler than other flagellomeres. Meso-
soma (Figs. 8, 10, 12, 14): with pubescence as in A. viequesensis. (9)
Pronotum with lateral angle and posterior lobe slightly angular; dorsal por-
tions of lateral angle and posterior lobe rugose and separated from lower
portion by distinct carina extending from lateral angle to posterior lobe;
postero-lateral area with 3-6 strong horizontal carinae. (10) Mesoscutum
very coarsely rugose laterally and anteriorly, becoming coarsely and con-
tiguously punctate medially and posteriorly. (11-12) Mesoscutellum and
metanotum as in A. viequesensis. (13) Mesepistemum evenly, extremely
deeply and coarsely rugose. (14) Metepisternum with extremely coarse
horizontal carinae. (15) Propodeum with very strong propodeal carina;
propodeal shield with strong carinae extending dorso-laterally from medial
groove; dorsal area with extremely coarse carinae extending postero-laterally
from antero-dorsal margin; lateral area with coarse, anastomosing carinae
extending from metepisternum to lateral portion of propodeal carina. (16)
Wing as in A. viequesensis but with veins darker. (17) Tegula as in A.
viequesensis but with amber replaced by dark brown, metallic tints slightly
more extensive, and yellow band largely or wholly obscured. (18-20) Fore,
middle and hind legs dark brown with coxae wholly to partially covered
with bright metallic green (least on middle coxa); pubescence as in A.
viequesensis but darker (almost black) dorsally on middle tibia, hind tibia
and apex of hind femur. Metasoma: (21) Terga as in A. viequesensis
but with central brown bands absent or very faint. (22) Sterna dark brown
with strong metallic tints on all or most of exposed portions; pubescence as
in A. viequesensis.

520 The University Science Bulletin

Agapostemon radiatus (Say)

Halictus radiatus Say 1837. Type probably destroyed.

Agapostemon pulchra Smith 1853. Type $ , British Museum (Natural History).

Agapostemon sitlcatulits Cockerell 1909. Type $ , U.S. National Museum.

The type of Halictus radiatus has probably been destroyed, but the
description leaves no doubt of its identity. Michener has examined the
type of Agapostemon pulchra (which is not labeled "California" as reported
by Sandhouse, 1936) and I have examined the type of Agapostemon sul-
catulus.

Distribution (Fig. 18). Specimens have been collected as far north as
Treesbank, Manitoba; as far south as Orange Co. in central Florida; as
far east as Cape Cod, Massachusetts; and as far west as Sioux Co., Nebraska.
In the northern part of the range (Wisconsin) females have been collected
from April through October and males from July through October. In the
central part of its range (Kansas)- females have been collected from April
through November and males in April and July through November. In the
southern part of its range females have been collected from April through
September and males from June through August and in October. Most
abundant at elevations of less than 2,000 ft. (610 m), A. radiatus has been
collected as high as 4,100 ft. (1,250 m) in the Smoky Mountains of North
Carolina.

Diagnosis. The male may be distinguished from many species by the
yellow maculations on its 6th metasomal sternum and by the tooth on its
hind femur; from A. mellwentris by the brown-black on the anterior surface
of its first metasomal tergum; from A. texanus and A. angelicus by its lack
of conspicuous metallic tints on metasomal tergum 4, and by its lack of a
broad, subapical, transverse ridge on metasomal sternum 4; and from males
of A. splendens, A. cocherelli and A. femoratus by its unmodified hind
basitarsus. The female may be distinguished from many species by its
metallic green metasomal terga; from A. texanus and A. angelicus by its
coarsely punctate or rugose mesoscutum without punctures of two distinct
sizes; from A. splendens by its smaller size, coarse parallel carinae dorso-
medially on its propodeum, its bright yellow mandibles (ferruginous
apically), and its very pale amber to clear wings; from A. femoratus by its
mesoscutum (coarsely punctate with rugae anteriorly and laterally in A.
radiatus, almost entirely coarsely rugose in A. femoratus). The female can-
not be morphologically differentiated from females of A. cocherelli but may
be distinguished by its eastern distribution.

Description

male (Figs. 105-106, 161)

General coloration of head and mesosoma bright metallic green to
blue-green, metasoma with black and yellow bands. Head (Figs. 105-106) :

Revision of the Bee Genus Agapostemon 521

pubescence white, pale yellowish on interocular area, becoming pale amber
on vertex. (1) Labrum as in A. texanus. (2) Clypeits with large scattered
punctures below, becoming weakly rugose above. (3) Interocular area with
very deep, fine, contiguous punctures below vertex, becoming weakly and
finely rugose at level of and below antennal sockets; supraclypeal pro-
tuberance rugose. (4) Vertex finely and deeply punctate below ocelli and
laterally, becoming deeply rugulose behind and between ocelli. (5) Gena
with short, weak, parallel, well separated carinae extending postero-dorsally
from antero-ventral margin. (6) Malar area bright yellow; very short. (7)
Mandible bright yellow with apical x/^-x/i dark ferruginous. (8) Antenna
brown-black above with apical l/2 of apical flagellomere amber; underside
of flagellum amber to pale amber; underside of scape and pedicel bright
yellow (underside of pedicel sometimes brown). Mesosoma: pubescence
white, becoming yellowish to fuliginous on mesonotum and metanotum.

(9) Pronotum with distinctly, but not acutely, angular lateral angle and
posterior lobe; distinct carina extending postero-ventrally from lateral angle;
distinct vertical carina laterally; and numerous weak horizontal carinulae.

(10) Mesoscutitm with fine, deep, contiguous punctures becoming rugose
antero-laterally. (11) Mesoscutelhtm with fine, deep, contiguous punctures
medially, becoming scattered in shiny lateral area. (12) Metanotum finely
but deeply rugulose. (13) Mesepi sternum moderately deeply and moderately
coarsely rugose anteriorly, becoming slightly less deeply and coarsely rugose
posteriorly. (14) Metepisternum moderately deeply and moderately coarsely
rugose. (15) Propodeum with strong propodeal carina; propodeal shield
moderately deeply and moderately coarsely rugose; dorsal area with short,
coarse, parallel carinae extending postero-laterally from anterior margin,
becoming moderately deeply and moderately coarsely rugose posteriorly
(entire dorsal area rugose on some specimens) ; coarsely rugose postero-
laterally, becoming punctate or finely rugulose antero-laterally. (16) Wing
very pale brown, slightly darker on distal margin; veins and pterostigma
pale brown, radial vein very dark brown. (17) Tegula amber with metallic
tints antero-basally, with yellow on posterior margin, and with yellow sub-
marginal band anteriorly. (18) Fore leg yellow; coxa metallic green and
with small brown stripe posteriorly on tibia; pubescence yellowish to amber.
(19) Middle leg with coloration and pubescence like fore leg with small,
brown, subapical spot dorsally on femur and with brown posterior stripe
on tibia larger. (20) Hind leg (Fig. 161) with coxa metallic green; tro-
chanter brown to amber with yellow on apical l/2 of lower side (extending
to base on some specimens) ; femur yellow with brown spot dorsally at
apex and sometimes posteriorly at base (much smaller than apical spot);
tibia yellow with small brown to amber spot apically on anterior surface,
with brown stripe antero-dorsally often connected with brown basal spot;

522 The University Science Bulletin

tarsus yellow, basitarsus lacking basal ridge and apical groove. Pubescence
yellowish to amber. Metasoma: (21) Terga black with yellow bands
on basal halves of terga 2-6 and centrally on tergum 1 ; ventro-lateral margins
of terga 1-6 transparent amber; and sometimes with faint metallic tints
anteriorly on tergum 1 and postero-laterally on tergum 3. Pubescence dor-
sally on terga 1-4 very short and inconspicuous, pale on yellow bands, dark
on brown bands; moderately long, yellowish pubescence anteriorly on tergum
1 and laterally on terga 1-4; moderately long and dark brown dorsally and
laterally on terga 5-7. (22) Sterna yellow with narrow brown band on distal
margin of sternum 2 pale and slightly broadened medially, becoming pro-
gressively darker and larger on sterna 3-5; sternum 6 with brown medial
stripe varying from very narrow to l/2 width of sternum; sternum 1 metallic
green basally. Broad, low, transverse, subapical (not apical as in A. texantis
and A. angelicas) ridge on sternum 4, most prominent laterally, and often
with faint metallic tints medially; moderately long scattered hairs on ex-
posed areas, 2-4 large stout bristles disto-laterally on sternum 4. (23)
Genitalia indistinguishable from those of A. femoratus.

female (Figs. 55-56)

General coloration of head, mesosoma and metasoma bright metallic
green to blue-green. Head (Figs. 55-56): pubescence white, becoming
pale yellowish amber on vertex and, commonly, on interocular area. (1)
Labrum as in A. texanns. (2) Clypeus with punctures deep, scattered below,
becoming subcontiguous and interspersed with weak horizontal rugae above.
(3) Interocular area with coarse, deep, contiguous punctures dorso-medially,
becoming moderately rugose laterally and ventrally; supraclypeal pro-
tuberance sparsely punctate and with weak horizontal carinae. (4) Vertex
with coarse, deep, contiguous punctures anterior to ocelli, becoming more
shallowly punctate or finely rugulose laterally; shallowly and finely rugose
posteriorly and between ocelli. (5) Gena with moderate to coarse, contiguous,
parallel carinae extending postero-dorsally from antero-ventral margin. (6)
Malar area dark amber to dark brown; very short, nearly absent. (7) Man-
dible yellow with apical Yi-Vi dark ferruginous. (8) Antenna brown-black
with underside of flagellum very slightly paler than upperside. Meso-
soma: pubescence white, becoming slightly yellowish to fuliginous on
mesonotum and metanotum. (9) Pronotum as in $ . (10) Mesoscutum with
deep, contiguous, moderately large punctures becoming rugose anteriorly
and antero-laterally. (11) Mesoscutellum with small subcontiguous punc-
tures medially, becoming larger and scattered on shiny lateral area. (12)
Metanotum finely and shallowly rugulose. (13) Mesepisternum deeply and
coarsely rugose anteriorly, becoming slightly less coarsely and less deeply
rugose posteriorly. (14) Metepi sternum with even, parallel, horizontal

Revision of the Bee Genus Agapostemon 523

carinae anastomosing posteriorly (on some specimens becoming rugose
posteriorly). (15) Propodeum with very strong and prominent propodeal
carina; propodeal shield weakly rugose (rugae often absent on upper
margin) ; dorsal surface with very coarse, irregular and anastomosing carinae
extending posteriorly from anterior margin; postero-lateral margin with
moderately coarse horizontal carinae, becoming much finer and weaker
antero-laterally. (16) Wing as in S . (17) Tegirfa as in $ but slightly
darker. (18-19) Fore and middle legs dark brown to brown-black with
coxae metallic green and with antero-dorsal pale amber to yellowish areas
apically on femora and basally on tibiae; pubescence white to pale fuliginous
or amber. (20) Hind leg dark brown to brown-black with metallic tints
dorsally on coxa; pubescence creamy, becoming amber posteriorly on tarsus
and dark brown to brown-black dorsally on tibia. Metasoma: (21)
Terga with very fine subcontiguous punctures on metallic regions; pubes-
cence white, becoming brown-black on terga 5-6; terga 2-4 with narrow
band of tomentum basally, otherwise largely devoid of hairs; moderately
long hairs anteriorly on tergum 1, laterally on terga 1-5, and dorsally on
terga 5-6. (22) Sterna brown to brown-black with metallic tints basally on
sternum 1 and some specimens with very weak metallic tints on sternum 4;
exposed areas with long, white, scattered hairs.

Agapostemon rhopalocera Smith

Agapostemon rhopalocera Smith 1853. Type $, British Museum (Natural History).

The type of this rare and peculiar species was examined at the British
Museum by C. D. Michener. I know of only two additional specimens
(males), both loaned to me by the British Museum.

Distribution. The holotype is from "Mexico." The other two males
were collected in May, 1913-14 at Atoyac, Vera Cruz, Mexico. This town
lies just 3.5 km ESE of Pico de Orizaba at an elevation of about 600 m
(1,968 ft.). As well as this area has been collected, it is surprising that no
more specimens have been collected in the past 50 years. Perhaps the habitat
of this species has been destroyed by human activities.

Diagnosis. The male may easily be distinguished from all other members
of Agapostemon by its long filamentous flagellum with the apical flagello-
mere flattened (Fig. 131).

Description

male (Figs. 77-78, 131, 154, 207)

General coloration of head and thorax metallic green, abdomen with
black and yellow bands. Head (Figs. 77-78): (1) Labrmn as in A. texanus
but slightly more acute apically. (2) Clypeus with punctation as in A.
leunculus, pubescence white to yellowish. (3) Interocular area above an-

524 The University Science Bulletin

tennal sockets very coarsely rugose or with dense contiguous punctures,
coarsely rugose at level of antennal sockets, becoming moderately rugose
below antennal sockets; pubescence white to yellowish. (4) Vertex with
sculpturing as in A. leunculus; pubescence white to yellowish. (5) Genu
with sculpturing as in A. leunculus; pubescence white to yellowish as in
A. leunculus but denser and medially more coarsely branched and bent
medially so as to cover the proboscidial fossa. (6-7) Malar area and mandible
as in A. texanus. (8) Antenna (Fig. 131) with apical flagellomere and distal
% of subapical flagellomere dark brown to black, remainder of flagellum
and pedicel brown above and yellow-amber below, scape amber with dark
streak on distal % of upper surface. Mesosoma: pubescence moderately
long, evenly distributed (except pronotum), and whitish to amber as in A.
leunculus. (9) Pronotum as in A. leunculus. (10) Mesoscutum finely and
contiguously punctate, sometimes finely rugose anteriorly and laterally.

(11) Mesoscutellum with punctures separated by about l/2 their diameters,
slightly coarser than those of mesoscutum and becoming rugulose laterally.

(12) Metanotum finely rugulose medially becoming shallowly rugose
laterally. (13) Mesepisternum as in A. nasutus. (14) Metepisternum rugose
with few weak horizontal carinulae. (15) Propodeum as in A. leunculus
but with more prominent propodeal carina. (16) Wing as in A. leunculus.
(17) Tegula as in A. nasutus but with reduced yellow area. (18) Fore leg
as in A. leunculus but more slender; pubescence white to amber. (19)
Middle leg very slender; coxa dark brown to black, remainder of leg yellow
with brown-black on all but anterior surface of trochanter and with brown-
black stripe on basal Vz-% of posterior surface of tibia; pubescence white to
amber. (20) Hind leg (Fig. 154) very slender, coxa metallic green, trochanter
brown-black, remaining segments yellow with brown streak on posterior of
femur and brown at base of tibia; pubescence white to amber. Meta-
soma: (21-22) Terga and sterna as in A. leunculus. (23) Genitalia (Fig.
207) very similar to A. leunculus but penis valves broader and flatter, medial
plate of gonostylus longer and narrower; and ventral lobe smaller and not
notched.

Agapostemon sapphirinus n. sp.

The name of this species refers to the sapphire blue color of the female.

I have seen 15 specimens (14 $ , 1 $ ) of this species. The female holo-
type, the allotype and the 13 male paratypes are all labeled, "Bahamas/
Clarencetown/Long Isl. Febr//Utowana/Exp. 1934." All of the types are
at the Museum of Comparative Zoology, Harvard University, Cambridge,
Massachusetts.

Diagnosis. Both males and females may be distinguished from all other
West Indian species, except A. cyaneus and A. oc/iromops, by their milky to
yellowish eyes and pterostigma. The male may be distinguished from

Revision of the Bee Genus Agapostemon 525

A. cyaneus by the pale amber to yellow on the underside of its pedicel, but
cannot be reliably distinguished from A. ochromops. The female may be
distinguished from A. ochromops by its dark purple metallic color and
from A. cyaneus by its very coarsely carinate gena.

Description
male

General coloration of head and mesosoma bright metallic green to dark
metallic green to blue, metasoma with black and yellow bands. Head:
as in A. viequesensis but with eyes cream-colored, underside of flagellum
yellower, pubescence of interocular area slightly yellow, and sculpturing
very slightly coarser. Mesosoma: pubescence as in A. viequesensis but
slightly yellowish laterally and slightly darker on mesonotum and metano-
tum. (9) Pronotum with lateral angle and posterior lobe more angular than
in A. viequesensis. Carinulae on postero-lateral area larger and fewer in
number than in A. viequesensis. (10-12) Mesoscutum , mesoscuteUum and
metanotum as in A. viequesensis but with very slightly coarser sculpturing.
(13-15) Mesepisternum , metepisternum and propodeum as in A. viequesensis
but with much coarser sculpturing. (16) Wing as in A. viequesensis but
with yellow pigment on pterostigma and underside of larger veins. (17)
Tegula as in A. viequesensis but with metallic tints less extensive. (18-20)
Fore, middle and hind legs as in A. viequesensis. Metasoma: (21)
Terga as in A. viequesensis but with metallic tints on some specimens (never
as extensive as in A. poeyi), with dark bands almost black, and with yellow
bands paler. Apex of pygidium nearly as acute as in A. ochromops. (22)
Sterna as in A. viequesensis but often yellower basally and darker along
apical margin. (23) Genitalia as in A. poeyi (cf. Figs. 186, 210).

FEMALE

Female as in A. poeyi but deep metallic blue, with metallic tints on sterna
less intense and less extensive, with indistinct brown to ferruginous bands
on terga 1-4, with creamy pigment on pterostigma and underside of larger
wing veins, and with eyes pale.

Agapostemon semimelleus Cockerell

Agapostemon semimelleus Cockerell 1900. Lectotype $ , Carnegie Museum, Pittsburgh.
Halictus (Agapostemon) andensis Vachal 1903 (new synonymy). Type 9, Museum National

D'Histoire Naturelle, Paris.
? Agapostemon castanetis Schrottky l(M)2b. Type $ ? Departamento de Zoologia, Sao Paulo?

Schrottky (1909b) tentatively synonymized the female of this species with A. semimelleus.

Not having seen a type I cannot dispute his judgement.
? Agapostemon coryliventris Holmberg 1903. Type $.
? Agapostemon multicolor Holmberg 1903. Type $. According to J. S. Moure (pers. comm.)

the types of most Holmberg species have been destroyed by dermestids. Therefore I am

following Schrottky (1909b) in his synonymy of A. coryliventris and A. multicolor with

526 The University Science Bulletin

A. semimelleus. Although these nominal species may be synonyms of A. chapadensis they
would still be junior synonyms.

I have seen the syntypes of A. semimelleus and have labeled the male as
lectoholotype and the female as lectoallotype. Both specimens are from
Chapada. I selected the male as lectoholotype because I am unable to dis-
tinguish the females of A. semimelleus and A. chapadensis, although the
female lectoallotypes are easily distinguished.

Distribution. This species is widespread in the campos, or tropical grass-
lands, of southern Brasil, of Paraguay and of northeastern Bolivia. It has
also been collected at the heads of the Maraiion and Ucayali Valleys in Peru,
the heads of the Putamayo and Cauca Valleys in Colombia, and along the
eastern slopes of the Andes in Bolivia and northern Argentina. It ranges
from sea level in southeastern Brasil to 1,700 m in Bolivia. It has been
collected during every month but May, but in the southern part of its
range is found only during the summer months.

Diagnosis. The male may be distinguished from most other South Amer-
ican species of Agapostemon by the acute lateral angle on its pronotum and
lack of a "button" on its last visible sternum (6th). It may be distinguished
from A. chapadensis by the shorter distal stylus of its genitalia (Figs. 173-
174) and by the lesser amount of black on the posterior of its hind femur
(Figs. 171-172).

The female cannot be distinguished from A. chapadensis (cf. Variation
below) but it may be distinguished from other South American Aga-
postemon by the acute upturned flange on its anterior mesoscutal margin
and lack of basal yellow bands on its metasomal terga.

Variation. When A. semimelleus and A. chapadensis were first de-
scribed, the females were separated chiefly on the basis of metasomal color.
Supposedly the female metasoma of A. semimelleus is honey-colored while
that of A. chapadensis is black. However, I feel that this character is un-
reliable because I have seen specimens where the anterior of the metasoma
was honey-colored while the posterior was black. I have also seen specimens
with a dark brown metasoma.

Independent evidence leads me to believe that intra-specinc variation
masks interspecific differences between the females of these two species. The
males of A. semimelleus are far more abundant and widespread than those
of A. chapadensis. Unless one assumes that there are great disparities be-
tween the distributions and relative abundance of females and males, one
must conclude that the metasoma of A. semimelleus females is frequently
black. It is also possible that some females of A. chapadensis have brown or
amber metasomas. At present there seems no solution to this problem. A
similar problem exists with respect to the sculpturing of females of A.
texanus and A. angelicus.

Revision of the Bee Genus Agapostemon 527

Description
male (Figs. 123-124, 171, 174, 220)

General color of head and mesosoma bright metallic green, metasoma
banded with black and yellow. Head (Figs. 123-124): with pale amber
pubescence on interocular area and vertex, becoming creamy on gena.
(1) habrum opaque yellow with apical and lateral margins transparent
amber. Two-thirds as long as wide, apex narrowly rounded, basal ridge
broad. Small, shallow, scattered punctures on dorsal ridge. Fringed with
stiff submarginal bristles apically. (2) Clypeus with large (sometimes very
large) scattered punctures on yellow region, becoming contiguously punctate
to rugose on upper margin. (3) Interocular area with small to moderately
small, deep, contiguous punctures below ocelli, becoming shallower and
medium-sized at level of antennal sockets and still shallower below antennae;
supraclypeal protuberance with large, contiguous, moderately deep punc-
tures. Pubescence between antennal sockets Yi-7/} as long as scape, becoming
shorter laterally, dorsally and ventrally and sub-tomentose ventro-laterally.
(4) Vertex with deep contiguous punctures anteriorly and laterally, becom-
ing shallower and slightly larger between ocelli and transversely carinate
posteriorly. (5) Gena with parallel, subcontiguous, short carinae extending
dorso-laterally from just behind smooth shiny area on antero-ventral margin.
Pubescence postero-medially about as long as scape, becoming shorter an-
teriorly, laterally and dorsally. (6) Malar area yellow, very short and with
minute appressed hairs. (7) Mandible yellow with apical x/i ferruginous.
(8) Antenna brown to brown-black above, scape yellow below and some-
times with yellow extending dorsally at base, pedicel and flagellum amber to
yellow below. Mesosoma: pubescence creamy to yellowish, becoming
golden to fulvus on mesonotum and metanotum. (9) Pronotum with lateral
angle and posterior lobe acutely pointed, lateral angle projecting antero-
laterally; with conspicuous sharp carina extending postero-ventrally from
apex of lateral angle; with weak irregular rugae postero-laterally. Pubescence
above lateral angle and posterior lobe golden to fulvus, yellowish to creamy
below. (10) Mesoscutum with large conspicuous flange laterally on anterior
margin extending antero-dorsally; punctures small, very deep and con-
tiguous, becoming rugose antero-laterally. (11) Mesoscutellum with lateral
and medial punctures like that of mesoscutum, becoming larger and less
dense laterally; area on each side of mid-line shiny with scattered punctures.
(12) Metanotum coarsely punctate to rugose medially, becoming moderately
coarsely rugose laterally. (13) Mesepisternum coarsely rugose anteriorly, be-
coming moderately coarsely rugose posteriorly. (14) Metepisternum moder-
ately finely rugose with horizontal rugae longer and slightly more prominent
than other rugae. (15) Propodeum with propodeal carina strong and con-
spicuous; propodeal shield very coarsely and deeply punctate to very

528 The University Science Bulletin

coarsely rugose; dorsal area very deeply rugose anteriorly and medially, be-
coming moderately rugose to coarsely punctate laterally; lateral area with
small, deep, subcontiguous punctures anteriorly and centrally, becoming
coarsely rugose adjacent to propodeal carina. (16) Wing hyaline or nearly
hyaline, veins dark brown. (17) Tegula pale to dark transparent amber with
broad submarginal yellow band anteriorly (partially obscured by pubescence)
and yellow to dark brown on proximal margin. (18) Fore leg entirely
yellow, or yellow with brown postero-proximally on tibia and postero-dorsally
on distal y3 of femur. (19) Middle leg yellow with brown postero-dorsally
on tibia, usually postero-dorsally at apex of femur, and occasionally on coxa.
(20) Hind leg (Fig. 171) yellow with upper l/2 of coxa metallic, with
trochanter brown basally and sometimes amber apically, with brown apically
on femur and antero-dorsally on tibia. Metasoma: (21) Terga 2-6
amber to brown-black on posterior l/2 and yellow on anterior l/2 (often
partially concealed by overlap) ; tergum 1 varying from pale amber with
brown band on posterior margin to brown-black with lateral yellow spot
antero-dorsally, but most commonly intermediate in condition with antero-
dorsal yellow band separating dark brown anterior and posterior regions.
Very conspicuously punctate with subcontiguous punctures largest on
tergum 1, becoming progressively smaller on posterior terga. Pubescence
golden to fulvus anteriorly on tergum 1, laterally on terga 1-7 and dorsally
on terga 5-7; very short dorsally on terga 1-4, golden to fulvus on yellow
bands and dark brown to brown-black on dark bands. (22) Sterna 2-6
yellow but with amber to brown band at posterior margin of sterna 2-5,
narrow on sternum 2, becoming progressively wider on successive sterna
3-5; sternum 1 amber with metallic tints basally. Sternum 6 with low
proximo-medial ridge. Pubescence limited to scattered, moderately long,
fulvus to golden hairs on exposed areas. (23) Genitalia (Figs. 174, 220)
with very short ventral lobes bearing apical brush of large stiff hairs. Distal
stylus little longer than width of medial plate.

female (Figs. 129-130)
General coloration of head and mesosoma bright metallic green, meta-
soma honey-colored to black. Head (Figs. 129-130): bright metallic green;
pubescence white below becoming pale yellow above; sculpturing as in A.
atrocaeruleus. Mesosoma: bright metallic green, but with tegula, legs
and wing venation amber to brown; pubescence yellow or amber above,
becoming creamy below; sculpturing as in A. atrocaeruleus but dorsal and
lateral propodeal rugae more irregular. Metasoma: honey-colored to
black; pubescence creamy, very short and appressed dorsally, becoming
moderately long and erect anteriorly, laterally and posteriorly, and with
narrow band of white tomentum basally on terga 2-4; sculpturing as in
A . atrocaeruleus.

Revision of the Bee Genus Agapostemon 529

Agapostemon splendens (Lepeletier)

Halictus splendens Lepeletier 1841. Type $, Museum National D'Histoire Naturelle, Paris.
Agapostemon aeruginosas Smith 1853. Type 2, British Museum (Natural History).
Agapostemon splendens. Robertson, 1897.

I have seen the type of Halictus splendens, which was loaned to me by
Dr. S. Kelner-Pillault. The type of A. aeruginosas was examined by Cock-
erell, and more recently by Michener, and both believe it is conspecific with
A. splendens.

Distribution (Fig. 19). Agapostemon splendens is the most common
species of Agapostemon in the southeastern United States and along the
Gulf Coast. It is also common in the sandy areas of the Central States and
Michigan. It is found as far south as Veracruz, Mexico, as far north as
southern Saskatchewan, Canada, and as far west as south-central Utah. In
Canada females have been collected in May, June, July and August; males in
August and September. In Kansas females have been collected from April
through October and males from July through October. In Florida both
sexes have been collected every month of the year. While more common at
elevations of less than 2,000 ft. (610 m), A. splendens has been collected as
high as 6,600 ft. (2,012 m) in the Great Smoky Mountains National Park
in Tennessee and 3,700 ft. (1,127 m) near Wray, Colorado.

Diagnosis. The male may be distinguished from other species by its
toothed hind femora, the grooved basal ridge on its basitarsus, the apical
stylus on its gonostylus, and its brown wings. The female may be dis-
tinguished from other species by its metallic green metasomal terga, rounded
lateral pronotal angle, punctate mesoscutum and brown wings.

Description

male (Figs. 71-72, 166, 179, 200)

General coloration of head and mesosoma bright metallic green, meta-
soma with black and yellow bands. Head (Figs. 71-72): pubescence white,
yellowish on vertex. (1) Labrum as in A. texanus. (2) Clypeus with large,
shallow, scattered punctures on lower l/2, becoming contiguous above. (3)
Interocular area with deep, contiguous, medium punctures below ocelli,
becoming rugose laterally and below; sculpturing of supraclypeal protuber-
ance like that of adjacent clypeal area. (4) Vertex with deep, contiguous,
medium punctures anteriorly, becoming rugose laterally, posteriorly and
between ocelli. (5) Genu with short coarse carinae extending dorso-laterally
from antero-ventral margin. (6) Malar area bright yellow, short. (7) Man-
dible bright yellow, pale ferruginous on apical V4-V3- (8) Antenna black
above with yellow sometimes present on basal xA-x/i of scape and present on
apical flagellomere; scape yellow below, pedicel with yellow spot on under-
side, and underside of flagellum pale amber. Mesosoma: pubescence

530 The University Science Bulletin

yellow to amber on mesonotum and metanotum, white elsewhere. (9)
Pronotum with lateral angle rounded and rugose above; postero-lateral area
with large vertical carina just anterior to several irregular horizontal carinae.
(10) Mesoscutum with coarse contiguous punctures becoming rugose antero-
laterally. (11) Mesoscutellum with coarse contiguous punctures anteriorly,
becoming rugose posteriorly. (12) Metanotum rugose. (13) Mesepisternum
coarsely rugose anteriorly, only slightly less coarse posteriorly. (14) Metepi-
sternum with very coarse irregular horizontal carinae or coarsely rugose.
(15) Propodeum with propodeal carinae distinct; propodeal shield coarsely
rugose; very coarsely rugose postero-laterally, becoming rugulose antero-
laterally; coarsely rugose dorsally with triangular medial area depressed,
rugulose. (16) Wing yellowish brown, conspicuously darker on distal xA-x/i\
veins and pterostigma amber to brown with radius conspicuously darker
than other veins. (17) Tegula pale transparent amber with green tints
antero-basally and short transverse yellow band on anterior area curving
rearward and almost reaching distal margin. (18) Fore leg yellow with
coxa metallic green and trochanter with amber tints dorsally. (19) Middle
leg yellow with metallic tints on coxa and dark brown streak centrally on
postero-dorsal surface of tibia. (20) Hind leg (Fig. 166) yellow; coxa and
trochanter metallic green; femur with brown on apical xA-x/i of anterior,
dorsal and posterior surfaces; tibia with brown posteriorly on distal l/2,
antero-dorsally on proximal l/2, basally, and apically on anterior and dorsal
surfaces; basitarsus with amber on grooved portion of basal ridge and on
apical groove. Metasoma: (21) Terga brown to brown-black with yellow
bands on basal halves of terga 2-6, and with central yellow band on
tergum 1; anterior face of tergum 1 pale amber to brown. Pubescence short,
minute and dense on dorsal areas, brown-black on brown surfaces and white
on yellow or amber surfaces; pubescence longer on anterior surface of
tergum 1 and ventro-laterally on terga 5-7. (22) Sterna brown apically with
varying amounts of yellow basally (always some yellow present); low,
transverse, subapical ridge on sternum 4 (and sometimes 3) largest laterally;
long, white, scattered pubescence on exposed areas. (23) Genitalia (Figs.
179, 200) with ventral lobes large and fringed with hairs distally.

female (Figs. 69-70)

General coloration of head, mesosoma and metasoma bright metallic
green to blue-green. Head (Figs. 69-70): pubescence white to yellowish
on clypeus and intcrocular area, becoming yellow to amber on vertex; white
on gena. (1) Labium as in A. texanus. (2) Clypeus with large scattered
punctures below, becoming subcontiguous on upper %. (3) Interocular area
rugose with coarse contiguous punctures below ocelli; supraclypeal pro-
tuberance with coarse punctures and faint horizontal carinae. (4) Vertex

Revision of the Bee Genus Agapostemon 531

with coarse contiguous punctures anteriorly, becoming rugulose laterally
and between ocelli to rugose posteriorly. (5) Gena with coarse carinae ex-
tending postero-dorsally from antero-ventral margin. (6) Malar area very
short, dark brown. (7) Mandible yellowish to amber with apical YrYz
ferruginous. (8) Antenna dark brown to brown-black with underside of
flagellum slightly paler. Mesosoma: pubescence white, becoming yel-
lowish on mesonotum and metanotum. (9) Pronotum with lateral angle
rounded; dorsal aspect of lateral angle and posterior lobe shallowly rugose;
postero-lateral area with several carinae extending postero-ventrally and
antero-dorsally. (10) Mesoscutum with coarse, deep, contiguous punctures
becoming rugose antero-laterally (punctures sometimes appearing to be of
two sizes as in A. texanns but surface never as shiny). (11) Mesoscutdhtm
with large scattered punctures and with much smaller contiguous punctures;
central region just lateral to mid-line often shiny and devoid, or nearly
devoid, of small punctures. (12) Metanotum rugose to rugulose. (13)
Mesepisternum coarsely rugose. (14) Metepistemum rugose or with very
irregular horizontal carinae. (15) Propodeum with strong propodeal carina;
propodeal shield coarsely rugose or with irregular carinae extending dorso-
lateral^ from medial groove. (16) Wing as in $ . (17) Tegula as in 6
but slightly darker amber. (18) Fore leg dark brown to brown-black; coxa
metallic green; femur with pale yellow anteriorly, dorsally and posteriorly
at apex; tibia with basal yellow spot dorsally and pale brown on anterior
surface; tarsus pale brown; pubescence white to amber. (19) Middle leg
like fore leg but with coxa largely brown and with pubescence brown-black
dorsally on tibia and basitarsus. (20) Hind leg brown to brown-black, coxa
with metallic tints dorsally; pubescence white to amber, becoming brown-
black dorsally on tibia and basitarsus. Metasoma: (21) Terga bright me-
tallic green to blue dorsally, becoming brown ventro-laterally and brown
basally on tergum 1. Pubescence white, becoming black on terga 5-6; long
ventro-laterally on anterior surface of tergum 1 and dorsally on terga 5-6;
terga 2-5 with basal bands of white tomentum. (22) Sterna brown to brown-
black with long white hairs scattered on exposed areas.

Agapostemon swainsonae Cockerell

Agapostemon swainsonae Cockerell 1910b. Type $, British Museum (Natural History).

The type has been examined by C. D. Michener. Its shiny mesoscutum
leaves no doubt as to its identity.

The female is recognized here for the first time.

Distribution. This species is apparently limited to the island of Jamaica.
I have seen specimens from the following localities: Yardley Chase, Santa
Cruz Mtns., St. Elizabeth, 1,500 ft., (2 S , Oct.) ; Mandeville (15); Port-
land (2 9); Christina, 3,000 ft. (1 5 , July); Hanover, Great River District
(19, Jan.) ; Baron Hill, Trelawny (19, June) .

532 The University Science Bulletin

Diagnosis. The male is easily distinguished from all other West Indian
species by its very shiny, weakly sculptured mesoscutum. The female has
not been recognized previously owing to its similarity to A. viridulus. It may
be distinguished from all but A. viridulus and A. hispaniolicus by its bright
metallic green head and mesosoma and black metasoma; from A. hispanioli-
cus by its more rugose interocular area; and from A. viridulus by the lack of
metallic green tints at the base of its mandibles and by the lighter color of
its mandibles.

Description
male (Figs. 113-114, 147,209)

General coloration of head .and mesosoma bright metallic green and
shiny, metasoma amber to yellowish brown. Head (Figs. 113-114): pu-
bescence yellowish on clypeus, interocular area and vertex; white on gena.
(1) Labrum as in A. viridulus but yellow to white on basal ]/2 and pale
transparent amber on apical l/2. (2) Clypeus yellow to white below and
nearly as elongate as in A. viridulus; punctures scattered and with numerous
transverse rugulae on metallic area; pubescence short and sparse. (3)
Interocular area shallowly rugulose around antenna! sockets, becoming finely
and shallowly punctate dorsally and ventro-laterally; supraclypeal protuber-
ance with punctures and transverse rugulae similar to those on clypeus;
pubescence as in A. viridulus. (4) Vertex with very shallow, small, scattered
punctures anteriorly and laterally, becoming finely rugose posteriorly; area
between ocelli almost impunctate; pubescence as in A. viridulus. (5) Gena
as in A. viridulus but with rugae weaker and more numerous. (6) Malar
area as in A. viridulus but slightly shorter. (7) Mandible as in A. viridulus.
(8) Antenna medium brown above, scape yellow to white below, flagellum
amber to creamy below, underside of pedicel slightly darker than underside
of flagellum. Mesosoma: pubescence creamy to yellowish on mesonotum
and mctanotum, white laterally, ventrally, and on propodeum. (9) Pro-
notum as in A. viridulus. (10) Mesoscutum very shiny (more so than in any
other species of Agapostemon); very shallow, small punctures separated by
their diameters. (11) Mesoscutellum very shiny with very small, shallow,
widely scattered punctures. (12) Metanotum shiny, with minute, shallow,
scattered punctures medially, becoming finely and irregularly rugulose later-
ally. (13) Mesepisternum as in A. viridulus but sculpturing finer and shal-
lower. (14-15) Metepisternum and propodeum as in A. viridulus but with
slightly shallower sculpturing. (16) Wing as in A. viridulus (1 S with white
pigment on underside of pterostigma and veins). (17) Tegula pale trans-
parent, metallic tints absent or very faint, and yellow to creamy transverse
lunule truncate basally and almost reaching distal margin. (18-19) Fore and
middle legs yellowish amber with strong metallic tints on coxae; pubescence
white on coxae, amber on other segments. (20) Hind leg (Fig. 147) as in

Revision of the Bee Genus Agapostemon 533

A. viridulus but with brown areas much paler brown (or amber). Meta-
soma: (21) Terga amber with indistinct, slightly darker bands centrally.
Pubescence moderately long, scattered, white ventro-laterally and on anterior
surface of tergum 1; very short, inconspicuous white tomentum basally on
terga 2-6 (usually hidden by overlapping terga) ; inconspicuous, dense, short,
prostrate hairs dorsally on terga 1-4; long fuliginous hairs dorsally on terga
4-7. (22) Sterna amber to yellowish with long, white, scattered hairs on ex-
posed areas. (23) Genitalia as in A. viridulus but with only about l/2 as
many lateral grooves on gonocoxa (Fig. 209).

female (Figs. 35-36)

As described for A. viridulus but slightly bluer on head and metasoma;
with finer sculpturing; with lower portion of clypeus, basal % of mandible,
and labrum amber; and without metallic tints basally on mandible.

Agapostemon texanus Cresson

Agapostemon texanus Cresson 1872. Type 9, Academy of Natural Sciences, Philadelphia.

Agapostemon texanus subtilior Cockered 1898. Type 9 (?), whereabouts unknown; neotype
9 , University of Colorado Museum, Boulder.

Agapostemon borealis Crawford 1901. Type 9, Academy of Natural Sciences, Philadelphia.

Agapostemon calijornicus Crawford 1901. Lectotype 9, U.S. National Museum.

Halictus (Agapostemon) brachycerus Vachal 1903 (new synonymy). Type $, Museum Na-
tional D'Histoire Naturelle, Paris.

Agapostemon texanus iowensis Cockerell 1910c (new synonymy). Type 9, U.S. National
Museum.

Agapostemon proscriptus Cockerell 1912b (new synonymy). Type 9> U.S. National Museum.

Agapostemon joseanus Friese 1916 (new synonymy). Type 9, Zoologisches Museum, Hum-
boldt Universitat, Berlin.

Agapostemon suljuripes Friese 1916 (new synonymy). Type $, whereabouts unknown.

Agapostemon cyanozonus Cockerell 1924 (new synonymy). Type 6 , California Academy of
Sciences, San Francisco.

Agapostemon proscriptellus Cockerell 1924 (new synonymy). Type 9, California Academy of
Sciences, San Francisco.

Agapostemon texanus vandy\ei Cockerell 1925. Type 9, California Academy of Sciences,
San Francisco.

Agapostemon calijornicus psammobius Cockerell 1937b (new synonymy). Type $ , California
Academy of Sciences, San Francisco.

Agapostemon angelicas idahoensis Michener 1937 (new synonymy). Type 9, Snow En-
tomological Museum, University of Kansas, Lawrence.

Agapostemon calijornicus clementinus Cockerell 1939 (new synonymy). Type 9, University
of Colorado Museum, Boulder.

I have examined the types of all the nominal species and subspecies
listed above except A. texanus subtilior and A. suljuripes. As I find the
females of A. texanus are indistinguishable from those of A. angehcus, female
synonymies are based upon the distributions of the males (see map, Fig. 20).
The type localities are as follows: A. texanus, Texas; A. borealis, Vancouver
I., Canada; A. calijornicus, Pacific Grove, California; H. (A.) brachycerus,
Guatemala; A. texanus iowensis, Ames, Iowa; A. proscriptus, Guatemala
City, Guatemala; A. joseanus, San Jose, Costa Rica; A. cyanozonus,

534 The University Science Bulletin

Guaymas, Mexico; A. proscriptellits, Guaymas, Mexico; A. texanus vandy\ei,
Yosemite Valley, California; A. californicus psammobius, San Miguel I.,
California; A. angelicas idahoensis, Challis, Idaho; A. californicus demen-
ti mis, San Clemente I., California.

The type of Agapostemon texanus subtilior Cockerell has not been
identified by subsequent revisors and it is possible that Cockerell failed to
designate the type series with appropriate labels. I have designated a female
in the University of Colorado Museum as the neotype. This specimen is
probably from the series described by Cockerell as it bears a label identical
to that of some of the females mentioned by Cockerell, "Pasco, Wash/5
25 96."

I have not been able to examine a type of Agapostemon sulfuripes
Friese, but the description is good and its identity with A. texanus is obvious.

Distribution (Fig. 20). The range is greater than that of any other
species in the genus. It may be found from southern Canada (52° N) to
central Costa Rica (10° N) and from Cape Cod, Massachusetts (70° W)
to Vancouver Island, Canada (125° W). Agapostemon texanus is most
abundant on the western coast of the United States; fairly abundant in the
United States west of the Mississippi River, except for the arid Southwest;
uncommon in the arid Southwest, the United States east of the Mississippi
River, Mexico, and Central America. This species occurs from sea level on
the Adantic, Gulf and Pacific coasts to about 8,000 ft. (2,438 m) in the
Sierra Nevada of California. At the northern limits of its range (British
Columbia) females of A. texanus have been collected from May through
September and males from July through October; in Kansas females (in-
cluding those of A. angelicus) have been collected from April through
November and males in April (very rare) and from June through October;
in Arizona females (including those of A. angelicus) have been collected
from February through November and males in July and August; and in
Mexico females (including those of A. angelicus) have been collected from
February through October and in December, and males from April and
June through September.

From its geographic distribution it appears that A. texanus is able to live
under a wide variety of climatic conditions and in 70 of the 116 plant com-
munities (see Appendix A) occurring in the United States (Kiichler, 1964).
Considering the lack of apparent climatic, biotic or edaphic factors correlated
with its distribution, I find the distribution of A. texanus inexplicable. In a
widespread and heterogeneous as this (cf. discussion of variation),
> which limit its distribution in one area may be relatively unimportant
in another.

is. The male may be distinguished from males of many other
5 toothed hind femora, the apical stylus on its gonostylus, and

Revision of the Bee Genus Agapostemon 535

the lack of a low medial ridge on the apical l/2 of its last visible sternum;
from A. splendens by the lack of a grooved basal ridge on its hind basitarsus;
and from A. angelicas by the apical stylus of its gonostylus (Figs. 180-181)
and usually by the presence of a black stripe anteriorly on its hind tibia (if
absent, then also lacking black stripe on posterior surface of hind tibia).
The female may be distinguished from most other species by its bright
metallic green to blue metasomal terga, its almost hyaline wings, and its
two distinct sizes of mesoscutal punctures. I have not distinguished the
females of A. texanus from those of A. angelicits except where they occur
well outside the distribution of A. angelicas males (Fig. 20).

Variation. The amount of variation is reflected by the length of the
synonymy, which consists of 15 names. Although specimens from different
parts of the range differ in color, pattern and sculpturing, the male genitalia
are all alike and there are intergrades among the other characters. There-
fore, I conclude that the described variants all belong to one species. A.
angelicas is superficially more similar to the typical form of A. texanus than
is A. texanus vandy\ei. However, I regard A. angelicits as a separate species
because: the male genitalia — which provide excellent specific characters in
many other species of Agapostemon — are different from those of A. texanus
(Figs. 180-181); there are no intergrades between the two species with
regard to certain color patterns; the two are sympatric.

The variation within A. texanus falls into three major categories. These
are the area and density of melanic pigmentation, the amount and hue of
structural (metallic) coloration, and the sculpturing.

Melanic pigmentation is best seen in males, as it is obscured or replaced
by the structural colors predominant in females. The melanic pigmentation
varies in density from pale amber to brown-black, and varies both within
and among localities. It was expected that there might be concordance of
variation in the size of melanic maculations on the appendages. However,
specimens from San Clemente Island, California have no yellow (0%) on
either the fore coxa or lower l/2 of the scape; whereas specimens from Jalisco,
Mexico, have little (10%) yellow on the fore coxa but much (95%) on the
lower Yz of the scape. Nor could precipitation, temperature or other ecolog-
ical factors account for the similarity between, for example, populations
from New Hampshire and San Clemente Island, California (underside of
scape 100% black) or between Saskatchewan, Canada, and Missouri (under-
side of scape 100% yellow).

There are considerable differences of color pattern among various popula-
tions (Table 1). However, they can all be explained in terms of an increase
or decrease in the amount (area) of melanic pigmentation. For example,
although the black stripe on the posterior surface of the hind femur is
absent in some populations and present in others, it can be shown that this

536

The University Science Bulletin

03

to
to

re

re
O

c
re

<u

a
re

c

Ui

a,
1-1

3 2

8

re

^ 3

.

rt

^

(-i

6

<-w

c

o

i-

M-l

to

CV5

a

1-1

o

o

• ~*

4-»

-

re

U

3

a,

"re

o

3

Dh

t~*

u

-a

V

3
u

4— »

4— >

u

co

JJ

u

to

tJD

a

o

B

re

c

o

■ 1—1

4-»

re

1h

re

>

u

us

3
O

^ co

£ a

£ E

n

21

t/5

re

rt

rt

£

"O

U

rt

-C

C

o

c3

rt

(A

rt

CO

a

o

So

_y

o

►J

ft

o

rt o

•5 O

.^ c

rt "C

n

i — i

'a

u

o

c

a

'-t-H

rt

QJ

00

E

"rt

OJ

u

u

C

o

u

u

X

vs

flj

•^

rt

^.

1

00

o

IA

ON

I

o

I

CM

I

in

CM

I
O

a\

o
I

"-N

I
O

-2«o

•^

<u rt

o
■q-

I

"?

o

on
I

o

7

o

o

I
cs

"-N

I

o

7

o

ON

C5

On

I
o

o

rt "«

X rt

O tu

o ™

O cN

to, w

l/N

I

<=>

°?

o

ON

o
l/N

I
<=>

°P

in

on

o

o

I
"A

7

o

o

7

>n

7

O

o

I

°p

in

ON

-5
o *o ~

"< O i-,
'S u °

■o 3 n

•C Ji V4-,

" "« „

o

H-N

IX

I

o

o

o

I

T

o
o

1)

c

rt

u

"3
u

JO

96

s?

C

i>

4-J

.2
-a

IE

00

o

I

"N.

I

o
in

T

o

5^

3

a

3

Q

o

o

1

1

3

Q

J3

o

w 3

.2 S

U. M-4

O .3

CH-

"rt o+

E 5 o

O C

3
O

jo aj

u

S^

rt

4J C

CO

u

3
o

u

c

o+ y

(O

E g,

5 «

3 O
O) u,

o .2

to O

s

u

CQ §

o

—

V5

a

>>

3
O

QJ

c

9 o

IE

rt

CQ —

00

t/3

CO ^

u

-S

H

w

CO

C

3

u

o

u

OJ

tJ

Jh o

rt

4->
CO

O-B

PQ CO

o

z

_o

?, "a

"3

o+
o+

co r^i

£ Z

o —i

cj

— >N

rt rt

C (50

J3 (5
00 ^-^

Revision of the Bee Genus Agapostemon 537

stripe is absent only in those populations in which the anterior stripe is
reduced or absent, and that the reverse is never true. Every specimen in
which the posterior hind femoral stripe is present and the anterior stripe
absent proves, upon examination of the genitalia, to be A. angelicas rather
than A. texanas. Therefore, although differences of color pattern do exist
among individuals and populations, they are a result of the amount of
melanic pigmentation.

One of the most obvious variables is the hue of the structural (metallic)
coloration. It is usually pure green, but in some populations is yellowish or
blue. The extremes of the yellowish and blue forms are so distinctive that
they may be distinguished at a distance of several feet. The yellowish form
(A. vandy\ei) occurs in the Sierra Nevada of California and in the Coast
Ranges north of San Francisco. Limited to montane areas, it is not found
along the coast or in the Central Valley. This form has strong brassy yellow
overtones which are most striking in the Sierra Nevada south of Yosemite
National Park and in the southern portion of the Coast Ranges north of
San Francisco. The yellow cast becomes less pronounced further north to
the point where it is replaced by the pure green of A. texanus texanus along
the California-Oregon border.

The blue form of A. texanus (A. californicus) is found on the coast of
California south of Cape Mendocino, in the Central Valley, and on the
islands off the coast of southern California. It occurs only at low elevations
and is not sympatric with the yellow form. It is most easily recognized by
the bluish (rather than green) cast of the metallic coloration, which is
particularly pronounced in the mesoscutum of males. This hue is most
intense in the southern portion of the range, most notably on San Clemente
Island (A. calijornicus clementinus), where it is a very dark violaceous blue.
However, it becomes greener further north to the point where, along the
California-Oregon border, it is replaced by the pure green of A. texanus
texanus.

It is interesting that there do not seem to be intermediates between the
yellow and blue forms in the southern portions of their ranges. It would be
logical to regard them as separate taxa were it not for the green intermediate
forms to the north. Although the origins of this geographic distribution of
color forms are doubtless environmental, the populations must be relatively
homogeneous genetically for this character because annual or local environ-
mental fluctuations have no apparent effect upon the colors of the popula-
tions.

Variation in sculpturing is best seen on the mesosoma of females. The
punctures on the mesoscutum range from deep and contiguous (Fig. 24) to
relatively shallow and far apart (Fig. 23). The former condition produces
a rough and dull appearance, the latter a relatively smooth and shiny surface,

538 The University Science Bulletin

There are slight geographical variations in this character (Table 1), but
the variation within a given locality is considerable. Although the rugae on
the propodeum all follow the same general pattern, they vary considerably
in number, coarseness, angle of divergence and number of interconnections
(Figs. 25-26). This variability is so great within a locality that it is im-
possible to discern a significant difference among localities. In fact it is
so great as to invalidate the contention of Sandhouse (1936) and her pre-
decessors that A. texanus females may be distinguished from those of A.
angelicits by propodeal sculpturing.

Description
male (Figs. 75-76, 165, 180, 201)

General coloration of head and mesosoma bright metallic green to blue,
metasoma with black and yellow bands. Head (Figs. 75-76): (1) Labritm
opaque lemon yellow with transparent pale amber margins (distal % amber
on specimens from San Clemente I., California). Nearly as wide as long;
distal margin rounded; and large transverse ridge on basal l/3 with very
slight medial depression. Punctures on crown of basal ridge shallow and
separated by 2-3 times their diameters; submarginal punctures large and
shallow medially, becoming smaller laterally and disappearing just anterior
to basal ridge. Pubescence on crown of basal ridge minute, simple and erect;
simple hairs in submarginal punctures about three times as long as puncture
is wide and deflected distally. (2) Clypeus with coarse punctures contiguous
at upper margin, becoming subcontiguous at lower margin and becoming
smaller and shallower laterally; pubescence white, dense and about 2/5
length of clypeus medially, decreasing to 1/5 length of clypeus and loosely
appressed at lateral margin. (3) Interocular area with small, deep, contiguous
punctures below ocelli, becoming coarser laterally and moderately deeply
rugulose above antennal sockets and shallowly rugulose ventro-laterally;
supraclypeal protuberance coarsely punctate to shallowly rugose; pubescence
white, erect, l/2 as long as scape medially, becoming l/4 as long laterally.
(4) Vertex with small, deep, contiguous punctures anteriorly, becoming
shallower laterally; weakly rugose posteriorly, becoming weakly rugulose to
weakly punctate between ocelli; pubescence white, % as long as distance
between lateral ocelli medially, decreasing to l/2 as long as distance between
lateral ocelli laterally. (5) Gena with hypostomal carina about 2.5 times as
long as width of proboscidial fossa; short parallel carinulae extending
postero-dorsally from antero-ventral margin; pubescence white, ranging in
length from /5 as long as hypostomal carina centrally to 1/5 as long
laterally and dorsally. (6) Malar area yellow to amber; very short; pubes-
cence short, white, appressed. (7) Mandible opaque lemon yellow with
distal y3 ferruginous; punctures few, shallow and scattered thinly over basal

Revision of the Bee Genus Agapostemon 539

Y2; pubescence scant with small patch of short white hairs on base of man-
dible. Row of approximately eight evenly spaced amber hairs along lower
margin of basal 2/3 of mandible, with proximal hairs nearly twice as long
as distal hairs. (8) Antenna brown to brown-black above with pale amber
or yellow on apical l/2 of apical flagellomere; flagellum pale amber below;
scape and pedicel yellow to black below (see remarks on Variation).
Mesosoma: pubescence white, becoming creamy to golden on mesono-
tum and metanotum; metallic coloration brassy green to blue. (9) Pronotum
with lateral angle and posterior lobe slightly angular. Weak carina extend-
ing postero-ventrally from lateral angle; low carina extending postero-
ventrally from center of lateral portion; weak horizontal carinulae near
postero-lateral margin; long pubescence above lateral angle and posterior
lobe and very fine white tomentum on postero-lateral area. (10) Meso-
scutum with numerous small, dense, deep, contiguous punctures becoming
rugose anterolaterally. (11) MesoscuteUum with punctation like that of
mesoscutum but shallower, less dense, and often with shiny lateral area;
pubescence like that of mesoscutum. (12) Metanotum coarsely punctate to
rugulose; pubescence like that of mesoscutum. (13) Mesepistermtm coarsely
rugose anteriorly and only slightly less coarsely rugose posteriorly; pubes-
cence long. (14) Metepisterniim coarsely rugose, sometimes with horizontal
rugae more prominent than others; pubescence long. (15) Propodeum with
propodeal carina moderately weak to weak; propodeal shield moderately
coarsely to moderately finely rugose; dorsal area moderately coarsely rugose
medially and anteriorly, becoming finely rugose to finely rugulose postero-
laterally; lateral area moderately finely rugose posteriorly, becoming punctate
anteriorly. Pubescence very short dorsally, moderately long laterally and
posteriorly. (16) Wing nearly hyaline, with pterostigma and all veins but
radius amber; radius dark brown. (17) Tegula dark to very pale trans-
parent amber, with large pale yellow lunulate maculation parallel to — but
separated from — anterior margin, with metallic tints on basal margin, and
with short pale yellow band on posterior margin. (18) Fore leg with coxa
metallic green; trochanter yellow to brown-black; femur entirely yellow
to yellow with dark brown on all but distal portion of posterior surface;
tibia entirely yellow to yellow with dark brown on all but distal l/3 of
ventral surface and short brown stripe sub-basally on dorsal surface; tarsus
yellow. Pubescence white, becoming pale amber on tarsus. (19) Middle leg
with coxa brown-black tinted with metallic green; trochanter yellow with
brown spot basally on posterior surface to entirely brown-black; femur en-
tirely yellow to yellow with brown-black stripe broadest basally and narrowest
distally extending from base to apex on posterior surface; tibia yellow with
dark brown sub-basal stripe on basal l/z of dorsal surface to yellow with
brown-black sub-basal stripe on basal % of dorsal surface and brown-black

540 The University Science Bulletin

stripe entire length of ventral surface; tarsus yellow. Pubescence similar to
that on fore leg. (20) Hind leg (Fig. 165) with coxa bright metallic green;
trochanter brown-black with metallic tints to yellow with dark brown
basally; femur yellow with large brown apical spot postero-dorsally and
sometimes with basal brown spot posteriorly; tibia yellow with small brown-
black area at base and usually with apical streak of brown to brown-black
on posterior surface and sub-basal streak of brown to brown-black antero-
dorsally (posterior streak or both posterior and antero-dorsal streak may be
absent) ; tarsus yellow. Basal ridge on basitarsus low and inconspicuous.
Pubescence white, usually pale amber posteriorly on tarsus. Metasoma:
(21) Terga black with yellow bands on basal halves of terga 2-5 and centrally
on tergum 1; ventro-lateral margins of terga 1-6 transparent amber; metallic
tints apically on terga 4-5, pygidium amber to yellow. Pubescence dorsally
on terga 1-4 very short and inconspicuous, pale on yellow band, dark on
brown-black bands; moderately long and white anteriorly on tergum 1,
laterally on terga 3-5, and dorsally and laterally on terga 5-7. (22) Sterna
yellow with brown band apically on sterna 2-5 (narrow on sternum 2 but
occupying most of sternum 5) ; sternum 6 brown to yellow with brown band
basally; sternum 1 yellow to amber, with metallic tints basally; sternum 4
with low, transverse ridge reaching distal margin laterally and usually with
faint metallic spot medially. Moderately long, scattered, white hairs on ex-
posed areas and with 2-4 large stout bristles disto-laterally on sternum 4.
(23) Genitalia (Figs. 180, 201) with apical stylus, medial plate and basal
stylus present; ventral lobe moderately large, fringed with short hairs
distally.

female (Figs. 21-26, 67-68)

General coloration bright metallic green to blue or brassy. Head (Figs.
21-22, 67-68). (1) Labrum dark amber to ferruginous. Basal ridge prom-
inent, proximal face sloping gently toward proximal margin, distal face
sharply declivous, and apex of ridge with deep punctures; distal portion
narrow, abruptly rounded at apex; conspicuous median keel on antero-distal
portion, continued basally almost to basal ridge; anterior margin of keel
flattened and slightly flared laterally; single row of dark amber, wide, flat,
stiff bristles curving anteriorly at their apices and forming fimbria on antero-
lateral margins of labrum. (2) Clypeus brown-black below, becoming green
above; large, scattered, deep punctures below, becoming slightly smaller and
subcontiguous above; pubescence creamy to amber with single row of amber
bristles curving downwards from just above lower margin of clypeus. (3)
Interocular area with deep, moderate sized, contiguous punctures below
ocelli becoming rugose laterally and below; supraclypeal protuberance with
moderate sized, scattered punctures between short transverse rugae. (4)

Revision of the Bee Genus Agapostemon 541

Vertex with deep, small to moderate sized, contiguous punctures anteriorly,
becoming smaller and shallower laterally; rugulose posteriorly and between
ocelli; pubescence creamy to amber. (5) Genu (Fig. 22) with regular,
moderately coarse carinae extending postero-dorsally from antero-ventral
margin; pubescence white and long. (6) Malar area amber to brown-black;
very short; pubescence very short, sparse, white and tomentose. (7) Mandible
yellow to pale amber with apical V^-Yz ferruginous; moderately long, white,
scattered pubescence at base and single row of moderately long amber pu-
bescence on lower margin. (S) Antenna dark brown to brown-black with
underside of flagellum amber. Mesosoma: pubescence white, becoming
creamy to amber on mesonotum and metanotum. (9) Pronotum as in $
but with lateral angle and posterior lobe slightly more angular. (10)
Mesoscutum (Figs. 23-24) with medium sized, moderately deep punctures
separated by 2-3 times their diameters and interspersed with more numerous
smaller and shallower contiguous to subcontiguous punctures; rugose antero-
laterally (see discussion of Variation); pubescence moderately long. (11)
Mesoscutellitm with punctation like that of mesoscutum but small punctures
smaller, more numerous and shallower; often shinier than mesoscutum;
pubescence moderately long. (12) Metanotum finely and shallowly rugulose;
pubescence moderately long. (13) Me sepi sternum with coarse rugae an-
teriorly, becoming slightly less coarse posteriorly; pubescence moderately
long. (14) Metepistemum with coarse, interconnected, horizontal carinae;
pubescence moderately long. (15) Propodeum (Figs. 25-26) with prominent
propodeal carina; propodeal shield weakly rugose with stronger rugae ex-
tending dorso-laterally from medial groove; dorsal area with weak, irregular,
interconnected rugae extending posteriorly from anterior margin, becoming
moderately finely rugose postero-laterally; lateral area with contiguous,
horizontal carinulae and scattered small punctures; pubescence short dorsally,
moderately long laterally and posteriorly. (16) Wing as in $ . (17) Tegula
as in i but slightly darker amber. (18) Fore leg with coxa dark brown to
brown-black and tinted with metallic green; trochanter dark brown to
brown-black, often with narrow, inconspicuous, yellow band apically on
posterior surface; femur brown to brown-black, with apical yellow band;
tibia brown-black, with amber to yellow anteriorly; tarsus brown-black to
pale amber; pubescence creamy on coxa, becoming amber on tarsus. (19)
Middle leg similar in color to fore leg but with pubescence fuliginous dor-
sally on distal portion of tibia and amber ventrally on femur and tibia.
(20) Hind leg with coxa dark brown to brown-black with strong metallic
green tints dorsally; trochanter brown to brown-black; femur, tibia and
tarsus brown-black to pale amber; pubescence creamy on coxa, trochanter
and femur, becoming amber on tibia and tarsus and dark fuliginous to
black on dorsal surface of tibia (darkest basally). Metasoma: (21) Terga

542 The University Science Bulletin

metallic green to blue or brassy; white bands of pruinose pubescence cover-
ing basal l/$ of terga 1-4 and white, very short, simple hairs covering dorsal
% of terga 1-4; terga 5-6 with long, dark brown to black pubescence. (22)
Sterna brown to brown-black with metallic tints basally on sternum 1;
long, creamy, scattered hairs on exposed areas.

Agapostemon tyleri Cockerell

Agapostemon tyleri Cockerell 1917a. Type 9 , U.S. National Museum.

Agapostemon martini Cockerell 1927, 9 only, (new synonymy). Type 9, California Academy
of Sciences.

I have examined the holotypes of Agapostemon tyleri and A. martini
and find them conspecific. Contrary to the opinion of Sandhouse (1936)
I do not find the type of A. tyleri conspecific with that of A. coloradensis
Crawford (cf. Diagnosis). Cockerell correctly associated the sexes of A. tyleri
but incorrectly associated the male of A. cocker elli with the female of A.
martini. The male paratypes of A. martini are misidentified and belong to
Agapostemon coc\erelli Crawford.

Distribution (Fig. 27). Agapostemon tyleri occurs in the desert South-
west and in arid parts of Mexico. I have seen specimens from as far north
as Jerome, Arizona; as far south as Tehuacan, Puebla, Mexico; as far east
as the Gulf Coast in Tamaulipas, Mexico; and as far west as Prescott,
Arizona. It ranges from sea level in Tamaulipas, Mexico, to 8,200 ft. (2,499
m) at Pachuca, Hidalgo, Mexico. Throughout most of its range it occurs
at altitudes of over 4,000 ft. (1,219 m) on the central plateau. In the United
States males have been collected from February through April, June through
September, and November; and females in April and May, July through
September, and November. In Mexico males have been collected in Feb-
ruary, May through September, and November; and females from June
through September, and December.

Diagnosis. The male may be distinguished from all North American
species except A. coloradinus by the lack of yellow maculations on its dark
brown to black metasomal sterna; from A. coloradinus by its smaller size,
by the black streak extending the entire length of the posterior surface of
its hind tibia (Figs. 159-160), and by its broader penis valve, which lacks
conspicuous hairs on its dorsal crest (Figs. 226-227). The female may be
distinguished from many other species occurring north of Mexico by its
black metasoma and by the lack of yellow on its clypeus; from A. virescens
by its finer genal carinulae (2-3 per 0.25 mm in A. virescens and 5-6 per 0.25
mm in A. tyleri, cf. Figs. 28-29); and from A. coloradinus by its colorless
wings, darker tegulae, and shinier supraclypeal protuberance. This species
is very close to A. coloradinus.

Revision of the Bee Genus Agapostemon 543

Description
male (Figs. 95-96, 159, 184, 198, 227)

General coloration of head and mesosoma bright metallic blue-green to
blue, metasoma with black and pale yellow to creamy bands. Head
(Figs. 95-96): pubescence snowy white. (1) Labrum as in A. texanus but
more rounded at apex. (2) Clypeus as in A. virescens but punctures shal-
lower and pubescence more dense. (3) Interocular area with coarse con-
tiguous punctures (some specimens finely rugose around antennal sockets) ;
supraclypeal protuberance as in A. virescens. (4) Vertex as in A. virescens.
(5) Gena as in A. virescens but with carinae much finer. (6) Malar area
yellow, never amber; short. (7) Mandible as in A. virescens. (8) Antenna
as in A. coloradinus. Mesosoma: pubescence snowy white. (9) Prono-
tum as in A. virescens but with weaker sculpturing. (10) Mesoscutum
as in A. virescens but with finer punctures and with anterior margin punc-
tate. (11-13) Mesoscutelhtm, metanotum and mete pi sternum as in A.
virescens but with finer sculpturing. (14-17) Metepisternum, propodeum ,
wing and tegula as in A. coloradinus. (18-20) Fore, middle and hind legs
as in A. virescens but posterior surfaces of tibiae and femora nearly com-
pletely covered with dark brown to brown-black, and usually with small
yellow rim apically on ventral surface of trochanters (Fig. 159). Meta-
soma: (21-22) Terga and sterna as in A. coloradinus but with creamy
band on tergum 1 almost always broadly interrupted medially. (23) Geni-
talia (Figs. 184, 198, 227) as in A. virescens but with much smaller fold on
medial plate of gonostylus and with prominent dorsal crest on penis valve.

female (Figs. 29, 51-52)

General coloration of head and mesosoma bright metallic blue-green,
metasoma black with white hair bands. Head (Figs. 51-52): (1) Labrum
as in A. texanus. (2-4) Clypeus, interocular area and vertex as in A.
virescens but with center of supraclypeal protuberance shiny and nearly im-
punctate. (5) Gena as in A. virescens but with much finer and more nu-
merous (5-6 per 0.25 mm) carinulae. (6) Malar area very dark ferruginous
to brown-black; very narrow, almost absent. (7-8) Mandible and antenna
as in A. coloradinus. Mesosoma: snowy white pubescence sometimes
very slightly fuliginous on mesonotum and metanotum. (9-15) Pronotum,
mesoscutum, mesoscutellum , metanotum, mesepisternum, metepisternum
and propodeum as in A. virescens but with sculpturing slightly shallower
and finer. (16) Wing as in A. virescens but always clear, never slightly
amber. (17) Tegula as in A. virescens but dark amber (also darker than in
A. coloradinus). (18-20) Fore, middle and hind legs as in A. virescens but
with pubescence slightly paler. Metasoma: (21) Terga as in A. virescens

544 The University Science Bulletin

but with pubescence on 5-6 paler brown. (22) Sterna as in A. virescens but
darker.

Agapostemon viequesensis Cockerell

Agapostcmon tiequesensis Cockerell 1918b. Type 9 . U.S. National Museum.

Agapostemon radiants portoricensis Cockerell 1919 (new synonymy). Type 8, U.S. National

Museum.
Agapostemon portoricensis. Wolcott, 1948.

I have seen the types of both A. viequesensis and A. radiatus portoricensis
and have no doubt that they are conspecific.

Distribution. In addition to the types I have seen specimens from the
following Puerto Rican localities: Aguada (1 8,9 Nov.), Arecibo (4 9,
24-26 June; 1 8 & 7 9 , 30 July-1 Aug.), Caguas (4 8 , 28-29 May), Coamo
Springs (2 6* , 28 Dec; 1 9 , 29 Dec), Dorado (1 8 , 15 July; 1 9 , 12 Apr.),
Guayanilla (1 $ , 22 July), Manati (2 8 , 5 Mar.; 15,7 June), Mayaguez
(1 8, Mar.; 2 9, 21-23 June; 1 9 , 16 Dec), Ponce (1 8, Jan.), Puerto
Cangrejos (1 8 , 28 Jan.), Rio Piedras (1 8 , 23 Mar.), San Juan (3 8 , 11-14
Feb.; 1 9 , 14 Apr.; 1 8 , 9-12 July), Vega Alta (1 9 , 27 Oct.) and Mona
Island (1 8 & 1 9 , 21-26 Feb.; 2 8 , Aug.). I have also seen specimens from
Nassau, New Providence I. (1 8 & 2 9 , 16 Dec.) and Arthurs Town, Cat
I. (1 9,July-Aug.).

Diagnosis. The male may be distinguished from A. centratus and A.
kohhellus by its toothed hind femora; from A. viriditlus, A. obscuratus, A.
hispaniohcus, A. swainsonae and A. cubensis by its lack of an elongate
clypeal region; from A. cyaneus, A. sapphirinus and A. ochromops by its
lack of milky to yellowish eyes and pterostigma; from A. insularis, A.
jamaicensis and A. poeyi by its lack of metallic green or blue medially on
metasomal tergum 3; and from A. columbi by its lack of metallic green
tints medially on metasomal tergum 3. The female may be distinguished
from other West Indian species with metallic metasomas by its lack of milky
to yellowish eyes and pterostigma and by the tints at the base of its mandible.

Description
male (Figs. 117-118, 145, 187, 224)

General coloration of head and mesosoma metallic blue-green, metasoma
with brown-black and yellow bands. Head (Figs. 117-118): (1) Labrum
yellow with apical margin transparent; basal ridge with very slight or no
medial depression, apical l/2 triangular; glabrous but for submarginal row
"t simple bristles. (2) Clypeus green above, yellow below; punctures scat-
tered and small on yellow area, becoming larger and more numerous on
green area; short white scattered hairs on green area almost absent from
yellow area. (3) Interocular area blue-green above, becoming yellow-green
below; rugose, becoming coarsely, deeply and contiguously punctate above

Revision of the Bee Genus Agapostemon 545

antennal sockets; pubescence yellowish to creamy, % as long as scape be-
tween antennal sockets, becoming shorter laterally and ventrally, and be-
coming white and tomentose ventro-laterally. (4) Vertex with deep, con-
tiguous, medium punctures anteriorly, becoming smaller and shallower
laterally; rugose posteriorly, becoming irregularly rugulose between ocelli;
pubescence as on interocular area but somewhat darker in color. (5) Gena
with weak, fine, short rugulae extending postero-dorsally from antero-ventral
margin; pubescence white, 1.5-2.0 times as long postero-medially as between
antennal sockets but becoming shorter anteriorly, laterally and dorsally.
(6) Malar area absent or very short. (7) Mandible yellow, apical /3 ferru-
ginous; single row of bristles postero-ventrally, longest basally, becoming
l/3 as long distally. (8) Antenna with scape yellow, becoming dark brown
to brown-black on distal J4-/4 °^ uPPer surface; pedicel and flagellum dark
brown to brown-black above, becoming pale amber to yellow below; apical
flagellar segment yellow above and pedicel often slightly darker than flagel-
lum. Mesosoma: pubescence wholly white or tinged with yellow on
mesonotum and metanotum. ( 9) Pronotum with lateral angle and posterior
lobe rounded; several to many fine, horizontal carinulae on postero-lateral
area; pubescence on dorsal portions of lateral angle and posterior lobe as on
mesoscutum, becoming very short, fine and inconspicuous on postero-lateral
area. (10) Mesoscutum blue-green from above, yellow-green in oblique
view; very finely and contiguously punctate, becoming rugose antero-
laterally. (11) Mesoscutellum shinier than mesoscutum, with scattered
large punctures interspersed with more numerous smaller and shallower
punctures (faintly rugulose in some specimens). (12) Metanotum rugose.
(13) Mesepisternum rugose anteriorly, becoming rugulose posteriorly and
ventrally. (14) Metepisternum horizontally rugose. (15) Propodeum with
distinct propodeal carina; coarsely rugose on propodeal shield, and posteriorly
on dorsal portion, becoming moderately rugose antero-laterally; antero-
dorsal margin with short rugae extending postero-laterally. (16) Wing
transparent, slightly tinged with brown; radius brown-black, other veins
and pterostigma brown. (17) Tegula pale transparent amber with metallic
green tints antero-basally and yellow band almost reaching distal margin.
(18-19) Fore and middle legs with coxae metallic green (only weakly tinted
on middle leg), trochanters amber and remaining segments yellow with
posterior of tibiae amber; pubescence pale amber. (20) Hind leg with coxa
green, trochanter brown, femur yellow with brown dorsally on apical l/3,
tibia dark amber to brown dorsally, becoming yellow ventrally, tarsus
yellow; pubescence amber (Fig. 145). Metasoma: (21) Terga with
yellow band on basal l/2 of dorsal areas; anterior portion of tergum 1 amber;
anterior l/2 of dorsal areas brown, darkest anteriorly and becoming trans-
parent on posterior margin; terga 4-5 often with very faint metallic green

546 The University Science Bulletin

tints laterally. Apex of pygidium rounded (Fig. 224). Pubescence very mi-
nute, inconspicuous; white on yellow bands becoming brown on brown
bands. Longer hairs white on anterior surface of tergum 1 and ventro-later-
ally on terga 2-5, and becoming brown dorsally on terga 5-7. (22) Sterna yel-
lowish to amber or pale brown with metallic green tints basally on sternum
1. Sternum 4 with posterior margin slightly concave, and with low sub-
marginal transverse ridge most prominent laterally; scattered white hairs on
exposed areas. (23) Genitalia (Fig. 187) as in A. poeyi but apical stylus on
gonostylus not as swollen at apex.

female (Figs. 9, 11, 13, 15, 17, 39-40)

General coloration of head,' mesosoma and metasoma bright metallic
green to blue-green. Head (Figs. 15, 17, 39-40): pubescence white, be-
coming fuliginous on vertex. (1) Labrum as in A. hispaniolicus. (2)
Clypeus green above and brown to brown-black below; lower x/2 with large
punctures separated by at least their diameters, punctures on upper l/2
slightly obscured by faint transverse rugae; pubescence sparse, white, longest
dorsally, shortest and least abundant ventrally, row of long amber bristles on
lower margin. (3) Interocular area with large, deep, contiguous punctures
centrally on upper % becoming moderately rugose laterally and on lower
%; rugae nearest inner margins of compound eyes irregularly vertical in
orientation. Pubescence Y3-y2 as long between antennal sockets as scape,
becoming shorter ventrally and laterally. (4) Vertex with large, deep, con-
tiguous punctures anteriorly and laterally, becoming rugose posteriorly and
rugulose between ocelli. (5) Genu with fine carinulae extending dorso-
lateral^ from antero-ventral margin; pubescence postero-medially twice as
long as pubescence between antennal sockets, becoming shorter anteriorly,
laterally and dorsally. (6) Malar area brown, very short to absent. (7)
Mandible amber with apical '/3 ferruginous; single row of hairs on postero-
ventral margin long basally, becoming l/3 as long distally. (8) Antenna
brown-black above, slightly paler below; pedicel, proximal and distal
flagellomeres often paler below than scape and remainder of flagellum.
Mesosoma: (Figs. 9, 11, 13): pubescence white, becoming slightly fulig-
inous on mesonotum and metanotum. (9) Pronotum with lateral angle
and posterior lobe rounded; postero-lateral area with 3-6 weak, horizontal
carinulae; pubescence above lateral angle and posterior lobe like that on meso-
scutum; lateral surface with very fine, inconspicuous, short, white hairs.
(10) Mesoscutum with very fine, contiguous punctures, becoming shallowly
rugulose antero-laterally. (11) Mesoscutellum with scattered large, deep
punctures interspersed with far more numerous very small, shallow punc-
tures. (12) Metanotum rugose. (13) Me se pi sternum rugose with rugae
finest posteriorly (specimens from Nassau, Bahamas with slightly coarser

Revision of the Bee Genus Agapostemon 547

sculpturing). (14) Metepisternum with regular horizontal carinulae. (15)
Propodeum with strong propodeal carina; propodeal shield rugose or with
irregular anastomosing rugae extending dorso-laterally from medial groove;
dorsal area finely rugulose medially, laterally with anastomosing short carinae
extending postero-laterally from anterior margin; lateral area with weak
carinulae extending postero-dorsally from antero-lateral margin; pubescence
very short and fine dorsally (absent medially), with long hairs posteriorly
and laterally. (16) Wing as in £ . (17) Tegula dark amber to brown with
green tints antero-basally with faint (often obscure) yellow band almost
reaching distal margin. (18-20) Fore, middle and hind legs with coxae dark
brown; with metallic tints on anterior surface of fore coxa and dorsal surface
of hind coxa; trochanters brown, remaining segments amber; pubescence
white to pale amber, becoming dark brown dorsally on middle tibia, hind
tibia and apex of hind femur. Metasoma: (21) Terga shiny metallic
green with central brown band (varies greatly in size) devoid or nearly
devoid of metallic tints; very fine, inconspicuous basal band of tomentum;
medium length white to yellowish hairs ventro-laterally on terga 2-4 and
anteriorly on tergum 1; long black hairs dorsally on terga 5-6. (22) Sterna
amber to brown (faint metallic tints on specimens from New Providence I.) ;
long white to yellowish, scattered hairs on exposed areas.

Agapostemon virescens (Fabricius)

Andrena virescens Fabricius 1775. Type $ , British Museum (Natural History).

Andrena nigricornis Fabricius 1793. Types £ , Zoologiske Museum, Copenhagen.

Halictus dimidiatus Lepeletier 1841. Type $ , whereabouts unknown.

Halictus tricolor Lepeletier 1841 (new synonymy). Type £, Institute e Museo di Zoologia,

Universita di Torino, Italy.
Agapostemon bicolor Robertson 1893. Type 2 , Illinois Natural History Survey.

The synonymy of Andrena nigricornis Fabricius and Halictus dimidiatus
Lepeletier with Andrena virescens Fabricius is based on their descriptions.
C. D. Michener examined the type of A. virescens in the British Museum.
G. C. Eickwort examined the type of Agapostemon bicolor in Urbana,
Illinois, and confirmed Robertson's synonymy (1895) of this species with
A. virescens. 1 have examined the type of hi. tricolor and it is a synonym
of A. virescens, not A. radiatus as stated by Sandhouse (1936).

Van der Vecht (1959) suggests that Apis sericea Forster (1771) may be
a senior synonym of A. virescens, but the description given by Forster does
not warrant such a conclusion (cf . Nomina Dubia) .

Distribution (Fig. 30). Specimens have been collected as far east as
Penobscot Co. in central Maine; as far west as Coos Co. on the Oregon
Coast; as far south as Hernando Co. on the central western coast of Florida;
and as far north as Vernon, British Columbia. This distribution is unusual
inasmuch as it is wide, ranging from coast to coast, yet this species is com-

548 The University Science Bulletin

mon only from about 37° to 47° North latitude. In the northwestern part
of the range (British Columbia and Alberta) males have been collected from
August through October, and females from April through August and in
October. In the central part of the range (Nebraska) males have been
collected from July through October, and females from May through October.
In the southeastern part of the range (Virginia, Tennessee, North Carolina
and Georgia) males have been collected from June through September and
in November, and females from May through August and in October.
Although A. virescens occurs primarily at elevations of less than 2,000 ft.
(610 m), I have seen specimens from as high as 4,000 ft. (1,219 m) on
Middle Mountain, Virginia; 6,900 ft. (2,103 m) near Boulder, Colorado; and
7,000 ft. (2,134 m) near Salt Lake City, Utah.

Diagnosis. The male may be distinguished from many other North
American species by the lack of yellow maculations on its brown to brown-
black sterna 5-6, by its medial ridge on the apical !/2 of sternum 6 (last
visible sternum), and by its gonostylus (Fig. 183); from A. coloradinus and
A. tyleri by the restriction of brown to brown-black areas on the posterior
surface of its hind leg to basal and apical regions of femur and tibia, and
by its lack of a prominent dorsal crest on its penis valve. The females may
be distinguished from other species occurring north of Mexico by its black
metasomal terga and the lack of yellow on its clypeus; from females of A.
coloradinus and A. tyleri by the coarser carinae on its gena (2-3 per 0.25
mm in A. virescens, 5-6 per 0.25 mm in A. coloradinus and A. tyleri), and
its largely yellow mandibles (amber to brown in A. coloradinus and dark
brown in A. tyleri).

Description
male (Figs. 97-98, 158, 183, 197)
General coloration of head and mesosoma bright metallic green, meta-
soma with black and yellow bands. Head (Figs. 97-98): pubescence white
becoming yellowish to pale amber on vertex. (1) Labrum as in A. texanus
but slightly blunter at apex. (2) Clypeus with medium size, shallow,
scattered punctures on yellow portion, becoming more dense and slightly
larger above. (3) Interocular area with large deep contiguous punctures
below ocelli, becoming rugose and more shallowly sculptured below; supra-
clypeal protuberance shallowly rugose to punctate with few faint horizontal
rugae. (4) Vertex with deep contiguous punctures anteriorly, becoming
more finely punctate laterally, and coarsely but shallowly rugose posteriorly
and between ocelli. (5) Gena with numerous short, parallel, subcontiguous
carinae extending dorso-laterally from antero-ventral margin. (6) Malar
area bright yellow to pale amber; short. (7) Mandible bright yellow with
apical V3 ferruginous. (8) Antenna brown to brown-black above with shiny
pale area on apical l/2 of apical flagellomere; flagellum pale amber to brown

Revision of the Bee Genus Agapostemon 549

below; and underside of scape and pedicel yellow to largely black with yellow
apically. Mesosoma: pubescence white, becoming pale amber to yel-
lowish on mesonotum and metanotum. (9) Pronotum with lateral angle
and posterior lobe neither very rounded nor acute; postero-lateral area with
large, broad, low, vertical ridge just anterior to numerous very fine and
faint carinulae. (10) Mesoscuhim with fine contiguous punctures becoming
finely rugose anteriorly and antero-laterally. (11) Mesoscutellum with punc-
tures like those of mesoscutum but somewhat larger and less dense (especially
laterally). (12) Metanotum finely and deeply rugose. (13) Mesepisternum
coarsely rugose anteriorly, becoming finely rugose posteriorly. (14) Metepi-
stcrnum with irregular horizontal carinae, anastomosing so as to appear
largely rugose on some specimens. (15) Propodeum with distinct propodeal
carina; propodeal shield coarsely and shallowly rugose; dorsal area coarsely
rugose, postero-lateral area moderately rugose, becoming finely rugose to
coarsely punctate antero-laterally. (16) Wing clear to transparent amber
with apical regions slightly darkened. Radial vein dark brown, stigma and
remaining veins pale brown. (17) Tegula pale transparent amber with
green tints antero-basally, short transverse yellow band on anterior area
curving rearward and almost reaching distal margin; posterior margin
yellow. (18) Fore leg yellow with coxa bright metallic green; trochanter
dark brown; femur with brown basally and postero-dorsally on basal x/2 of
tibia. Pubescence white to pale yellow. (19) Middle leg like fore leg but
with only faint green tints on coxa and with brown streak postero-dorsally
on basal %-% of tibia. Pubescence white to pale yellow. (20) Hind leg
(Fig. 158) with coxa bright metallic green; trochanter brown (sometimes
with faint metallic tints); femur yellow with brown at base and apex;
tibia yellow with brown at base (and sometimes at apex) and brown streak
centrally on antero-dorsal surface. Pubescence white to pale yellow. Meta-
soma: (21) Terga brown to brown-black with yellow bands on basal
halves of terga 2-5 and narrow yellow band centrally on tergum 1; pubes-
cence white, short and appressed on dorsal portions of terga 1-4, moderate
length elsewhere. (22) Sterna brown to brown-black with yellow baso-
laterally on sterna 2-4 (often hidden by overlapping sterna), and sternum 1
with metallic tints basally; apical !/2 of sternum 6 with median ridge in
shallow depression. Pubescence white and moderately long on exposed
areas. (23) Genitalia (Figs. 183, 197) with dorsal crest of penis valve
rounded and broad, not prominent as in most other species; ventral lobes
of moderate size and distally fringed with hairs.

female (Figs. 28, 53-54)

General coloration of head and mesosoma bright metallic green to
blue-green, metasoma black with white hair bands. Head (Figs.

550 The University Science Bulletin

53-54) : pubescence white, becoming faintly yellowish at vertex (and on
interocular area of some specimens). (1) Labium as in A. texanus. (2)
Clypeus with large scattered punctures below, becoming contiguous above.
(3) Interocular area with deep, medium sized punctures above, becoming
deeply but finely rugose below. (4) Vertex with deep, small, contiguous
punctures anteriorly, finer and shallower laterally and becoming finely rugose
posteriorly; area between ocelli finely rugose to punctate. (5) Gena with
moderately fine (2-3 per 0.25 mm) parallel carinae extending postero-dorsally
from antero-ventral margin. (6) Malar area amber to ferruginous and very
short, almost absent. (7) Mandible yellow or rarely very pale amber; with
apical Yi-1/^ ferruginous. (8) Antenna dark brown to brown-black with
underside of flagellum slightly paler brown. Mesosoma: pubescence white,
becoming pale amber to slightly fuliginous on mesonotum and meta-
notum. (9) Pronotum with lateral angle and posterior lobe neither very
rounded nor acute; numerous very faint parallel carinulae extending antero-
dorsally from postero-ventral area. (10) Mesoscutum with deep, medium
sized, contiguous punctures becoming rugose anteriorly and antero-laterally.
(11) Mesoscutellum with fine contiguous punctures medially, becoming
large and separated by as much as 2-3 times their diameters laterally. (12)
Metanotum rugulose. (13) Mesepisternum coarsely rugose, only very slightly
finer rugae posteriorly. (14) Metepisternum with moderately coarse carinae
regular and parallel anteriorly, irregular and anastomosing posteriorly. (15)
Propodeum with prominent propodeal carina; propodeal shield with nu-
merous weak carinae extending laterally and slightly dorsally from median
groove; dorsal area with irregularly anastomosing carinae extending postero-
laterally from anterior margin; lateral area with anastomosing horizontal
rugae coarse posteriorly, becoming fine anteriorly. (16) Wing as in $ .
(17) Tegula as in $ but slightly darker. (18-20) Fore, middle and hind legs
brown to brown-black with yellow dorsally at apex of femur and base of tibia
of fore and middle legs; dorsal surface of hind coxa with strong metallic
tints; pubescence white to amber, becoming fuliginous apically on dorsal
surface of middle leg and brown-black basally on antero-dorsal surface of
hind leg. Metasoma: (21) Terga black; white tomentose hair bands
basally on terga 2-4 and centrally on tergum 1 (often broadly interrupted
medially); pubescence anteriorly on tergum 1 and laterally on terga 1-5
white and moderately long; pubescence dorsally on terga 5-6 moderately long
and brown to black. (22) Sterna brown to brown-black with long, scattered
white hairs on exposed areas.

Agapostemon viridulus (Fabricius)

Apis viridula Fabricius 1793. Type $, Lund collection, Zoologiske Museum, Copenhagen.
Indrena (Agapostemon) femoralis Guerin-Meneville 1844. Type $, Museo Civico di Storia
Naturale di Genova, Genoa.

Revision of the Bee Genus Agapostemon 551

Agapostemon semiviridis Cresson 1865 (new synonymy). Type 9 , Academy of Natural Sciences,

Philadelphia.
Agapostemon stridulus. Dalla Torre, 1896.

J. S. Moure has selected a lectotype of Apis viridula and redescribed it
(Moure, 1960). C. D. Michener has examined the type of Andrena (Aga-
postemon) jemoralis, and I have examined the lectotype of Agapostemon
semiviridis. I have no doubt that all three are conspecific. It should be noted
that owing to the inadequacy of the original description Agapostemon
viridulus was believed by many authors to be a synonym of Agapostemon
virescens.

Except for the coloration of the head and thorax this species is indis-
tinguishable from Agapostemon obscitratus Cresson. Sympatry and the
absence of intermediate forms lead me to conclude that A. obscitratus is not
conspecific with A. viridulus.

Distribution. I have seen specimens from the following localities in
Cuba: Baracoa (3 $ , Aug.), Castillo de Juaga, Cienfuegos (8 2 , 4-5 Sept.),
Central Jaronu (4 8 , 31 Mar.), Cristo (4 8 , 3 Oct.), Guane (2 8 & 1 2 ,
24-26 Sept.), Guantanamo (5 8 & 1 2 ), Havana (8 $), Puerto Baniato,
Santiago de Cuba (1 $ , Nov.), Rio Toa, Baracoa (2 8 , 26 Apr.), San Bias
(1 8 , 12 Aug.), Santiago de las Vegas (1 9 , 10 Feb.), Siboney (12, Feb.),
Soledad (1 8,27 Feb.; 1 8 , 4 Mar.; 1 8 , 5 May), Trinidad Mts. (1 2 ,
28 Aug.), Vinales (2 2 , 16-22 Sept.).

Diagnosis. The male may be distinguished from West Indian species,
except A. obscuratus and A. hispanioticus, by its black metasomal terga;
from A. obscuratus by its bright metallic blue-green head and mesosoma;
from A. hispaniolicus by the subcontiguous punctures of its mesoscutum.
The female may be distinguished from West Indian species, except A.
obscuratus, A. hispaniolicus and A. swainsonae, by its black metasomal
terga; from A. obscuratus by its bright metallic blue-green head and meso-
soma; from A. swainsonae by the metallic green tints on the base of its
mandible; and from A. hispaniolicus by the coarser sculpturing of its in-
terocular area.

Description
male (Figs. 111-112, 150, 178, 208)
General coloration of head and mesosoma metallic green to blue-green,
metasoma dark brown to black. Head (Figs. 111-112): (1) Labrum
yellow to white on basal l/2, amber to brown on apical l/2; apex smoothly
rounded (not acute), basal l/2 with shallow depression medially on trans-
verse ridge; shiny and impunctate; pubescence limited to about 20 subapical
bristles. (2) Clypeus yellow to white below and extraordinarily elongate;
punctures coarse and scattered, weak transverse rugae medially on upper Y2;
pubescence short, sparse and white. (3) lnterocular area with contiguous

552 The University Science Bulletin

punctures below median ocellus, becoming rugose midway between median
ocellus and antennal sockets, rugosity at level of antennal sockets changing
to scattered punctures ventro-laterally; supraclypeal protuberance with coarse
scattered punctures and transverse medial rugae; pubescence white, nearly
as long between antennal sockets as scape, becoming shorter laterally,
dorsally and ventrally; ventrolateral area near malar space with short,
sparse, white tomentum. (4) Vertex with fine, contiguous punctures an-
teriorly and laterally, becoming rugose posteriorly; area between ocelli
sparsely punctate; pubescence white, fuliginous between ocelli. (5) Genu
with fine parallel rugae running postero-dorsally from antero-ventral margin;
pubescence white and about 1.5 times as long postero-medially as pubescence
between antennal sockets, becoming shorter laterally, anteriorly and dorsally.
(6) Malar area yellowish, more than l/z as broad as long; shiny and im-
punctate; pubescence sparse, appressed, short and white. (7) Mandible
yellow to creamy, apical y$-xA transparent ferruginous; simple hairs on
postero-ventral margin of mandible longest basally, becoming y5 as long
apically. (8) Antenna with scape brown-black above, yellow below; pedicel
dark brown; flagellum brown-black above, paler brown below. Meso-
soma: pubescence fuliginous on mesonotum and metanotum, white
laterally, ventrally and on propodeum. (9) Pronotum with lateral angle and
posterior lobe rounded; sculpturing inconspicuous. Pubescence postero-
laterally very fine, short and inconspicuous; dorsally on lateral angle and
posterior lobe similar in length and color to that of mesonotum. (10)
Mesoscutum with very fine, deep, contiguous punctures, becoming rugulose
antero-laterally. (11) MesoscntelUim shiny with large, scattered, central
punctures, becoming finer and more numerous anteriorly. (12) Metanotum
finely and irregularly rugulose with large, scattered punctures. (13) Mesepi-
sternum coarsely to moderately rugose anteriorly, becoming finely rugose
posteriorly. (14) Metepisternum finely and horizontally carinulate. (15)
Propodeum rounded posteriorly; propodeal carina absent; fine, even,
horizontal carinulae laterally and posteriorly (uppermost of these becoming
stronger along upper margin of propodeal shield) ; dorsal area of propodeum
with fine, even carinulae running posten>laterally from posterior margin
of metanotum. (16) Wing transparent, faintly fuliginous; veins and ptero-
stigma dark brown, radius almost black. (17) Tegula dark transparent
brown with metallic green tints on basal margin and submarginally on
anterior !/3 and with very fine scattered punctures. (18) Fore leg dark
brown with yellow to white on antero-ventral surface of femur, on all but
posterior surface of tibia, and on tarsus; coxa metallic green; pubescence
white proximally, becoming amber distally. (19) Middle leg similar in
color and pubescence to fore leg but with white to yellow areas on femur
and tibia reduced, and with tarsus amber to brown, coxa slightly metallic

Revision of the Bee Genus Agapostemon 553

green. (20) Hind leg (Fig. 150) with coxa largely metallic green to blue;
trochanter brown with metallic tints; femur brown with yellow to white
antero-ventrally, on basal % antero-dorsally, and posteriorly around large
tooth; tibia brown with yellow to white ventrally and antero-ventrally
(yellow sometimes very reduced or absent); tarsus brown to amber. Hind
leg very swollen, especially femur; basal ridge on basitarsus very prominent,
broad and concave; apical groove broad and delimited by strong carina;
pubescence amber on tarsus, white on other segments. Metasoma: (21)
Terga brown-black (pygidium somewhat paler) and shiny; with long,
sparse, white pubescence becoming fuliginous around pygidium and with
very short, moderately dense pubescence brown-black and inconspicuous,
and with narrow, basal bands of white tomentum on terga 2-6 (often hidden
by preceding terga). (22) Sterna not quite as dark as terga, sternum 1 with
metallic green tints basally; with sparse white pubescence on exposed por-
tions. (23) Genitalia (Figs. 178, 208) with penis valves smooth basally on
dorsal surface, with large dorsal crest grooved at apex. Gonocoxa with
numerous deep longitudinal grooves laterally; gonostylus with simple apical
stylus; medial plate with small fold on antero-medial margin; ventral lobes
large, curled posteriorly and with dense, short, simple pubescence marginally
on posterior surface.

female (Figs. 33-34)

General coloration of head and mesosoma metallic green to blue-green,
metasoma black with narrow bands of white tomentum. Head (Figs. 33-
34) : pubescence white, fuliginous on vertex, and amber on lower margin of
clypeus and on mandibles. (1) Labrum (Figs. 34 A-C) brown, with large,
broad, distal keel; basal area with rounded medial protuberance and sharply
delimited from distal process by marginal carina; additional transverse
carina present between base of distal process and outer margin of basal area.
(2) Clypeus elongate and with broad, very shallow, subapical depression or
flattened area; punctures very coarse, widely separated distally, becoming
more numerous above; irregular transverse rugae often scattered on metallic
area. (3) Interocular area rugose, becoming punctate at vertex; supraclypeal
protuberance with scattered punctures and regular transverse rugae. (4)
Vertex finely and densely punctate anterior to and lateral to ocellar triangle,
becoming shallowly and weakly rugose posterior to ocelli; area between
ocelli with fine, scattered punctures. (5) Gena with numerous, regular, con-
tiguous rugulae running postero-dorsally from antero-ventral margin. (6)
Malar area very short, almost non-existent. (7) Mandible dark ferruginous
brown with metallic green tints at base. (8) Antenna brown-black, slightly
paler on underside of flagellum. Mesosoma: pubescence as in
Pronotum with lateral angle and posterior lobe rounded; weak carina
running postero-laterally downward from lateral angle. (10) Mesoscutum

554 The University Science Bulletin

with dense, fine punctures contiguous centrally, becoming separated by
about l/2 tbeir diameters posteriorly and becoming finely rugulose anteriorly.
(11) Mesoscutellum with scattered medium punctures interspersed with
more numerous fine punctures anteriorly, and finely rugulose posteriorly and
laterally. (12) Metanotum with very faint rugulae running postero-laterally
from anterior margin. (13) Mesepisternum moderately rugose; rugae
moderately prominent anteriorly, becoming finer posteriorly. (14) Metepi-
sternum finely, horizontally carinulate, with carinulae more numerous and
less prominent than on S . (15) Propodeum not rounded posteriorly and
with definite propodeal carina; carinulae on lateral and dorsal areas similar
to $ ; numerous parallel carinulae on propodeal shield running dorso-
lateral^ from median groove. (16-17) Wing and tegula as on $ . (18-20)
Fore, middle and hind legs brown-black to amber; pubescence creamy to
amber, fuliginous on dorsal surfaces of tibiae, dark brown to black adjacent to
basi-tibial plate on hind leg; posterior tibial spur with 3-5 (usually 4)
spatulate teeth. Metasoma: (21) Terga brown-black to black, anterior
area of tergum 1 often brown to amber; dorsal surfaces finely granulose,
not shiny; pubescence on ventrolateral areas of terga 1-4 as well as on
anterior surface of tergum 1 scattered, of moderate length and white; dorsal
portions of terga with very numerous, minute, simple, brown-black hairs
directed posteriorly; basal bands of white tomentum on terga 2-5 usually
hidden by preceding terga but may be visible laterally; abundant long
brown-black pubescence on terga 5-6. (22) Sterna brown to black with
scattered, simple, long white hairs on exposed portions.

NOMINA DUBIA

In 1771, J. R. Forster described Apis sericea from North America. Van
der Vecht (1959) has suggested that it is a senior synonym of Agapostemon
virescens Fabricius 1775. Forster 's very brief description is sufficient only
to indicate that he had a male Agapostemon. Forster made no mention of
characters of specific significance and, in the absence of the type, I can
accord only the status of nomen dubium to Apis sericea Forster.

In 1903, J. Vachal described Ha/ictus (Agapostemon) chiriquiensis from
a single female. It would seem from the description to be either A. nasutus
or A. leuncidus. I requested the type but Dr. S. Kelner-Pillault was unable
to find it in the Museum National D'Histoire Naturelle in Paris. Therefore,
I consider Halictus (Agapostemon) chiriquiensis a nomen dubium.

Revision of the Bee Genus Agapostemon 555

APPENDIX A

The following list of plant communities in which Agapostemon texanus and A. angelicus
may occur was derived by superimposing distributions of these species on the vegetation map
of Kiichler (1964). Only those communities present in the United States are listed, as com-
parable vegetation maps do not exist for Mexico and Central America. The letter "T" preceding
a community indicates records of A. texanus, "A" indicates A. angelicus, and "TA" indicates
both species. The numbers are those assigned to the communities by Kiichler.

As Kiichler's map is based on potential vegetation, this list may be unrealistically large,
with the bees actually living in disturbed areas with little, if any, semblance of the potential
climax or subclimax vegetation.

T 1. Spruce — Cedar — Hemlock Forest (Picea — Thuja — Tsuga)

T 2. Cedar — Hemlock. — Douglas Fir Forest (Thuja — Tsuga — Pseudotsuga)

T 3. Silver Fir — Douglas Fir Forest (Abies — Pseudotsuga)

T 5. Mixed Conifer Forest (Abies — Pinus — Pseudotsuga)

T 6. Redwood Forest (Sequoia — Pseudotsuga)

T 7. Red Fir Forest (Abies)

T 8. Lodgepole Pine — Subalpine Forest (Pinus — Tsuga)

T 9. Pine — Cypress Forest (Pinus — Cypressus)

T 10. Ponderosa Shrub Forest (Pinus)

T 11. Western Ponderosa Forest (Pinus)

T 12. Douglas Fir Forest (Pseudotsuga)

T 13. Cedar — Hemlock — Pine Forest (Thuja — Tsuga — Pinus)

T 14. Grand Fir — Douglas Fir Forest (Abies — Pseudotsuga)

A 15. Western Spruce — Fir Forest (Picea — Abies)

TA 17. Black Hills Pine Forest (Pinus)

TA 18. Pine — Douglas Fir Forest ( Pinus — Pseudotsuga)

A 19. Arizona Pine Forest (Pinus)

A 20. Spruce — Fir — Douglas Fir Forest (Picea — Abies — Pseudotsuga)

A 21. Southwestern Spruce — Fir Forest (Picea — Abies)

TA 23. Juniper — Pinyon Woodland (Juniperus — Pinus)

T 29. California Mixed Evergreen Forest (Quercus — Arbutus — Pseudotsuga)

T 30. California Oakwoods (Ouei-cus)

A 31. Oak — Juniper Woodland (Quercus — Juniperus)

TA 32. Transition between 31 and 37

TA 33. Chaparral (Adenostoma — Arctostaphylos — Ceanothus)

T 35. Coastal Sagebrush (Salvia — Eriogonum)

T 36. Mosaic of numbers 30 and 35

TA 37. Mountain Mahogany — -Oak Scrub (Cercocarpus — Quercus)

TA 38. Great Basin Sagebrush (Artemesia)

A 39. Blackbrush (Coleogyne)

A 40. Saltbush — Greasewood (Atriplex — Sarcobatus)

TA 41. Creosote Bush (Larrea)

TA 42. Creosote Bush — Bur Sage (Larrea — Franseria)

TA 43. Palo Verde — Cactus Shrub (Cercidium — Opuntia)

TA 44. Creosote Bush — Tarbush (Larrea — Flourensia)

TA 45. Ceniza Shrub (Leucophyllum — Larrea — Prosopis)

T 47. Fescue — Oatgrass (Festuca — Danthonia)

T 48. California Steppe (Stipa)

T 50. Fescue — Wheatgrass (Festuca — Agropyron)

T 51. Wheatgrass — Blue Grass (Agropyron — Poa)

TA 53. Grama — Galleta Steppe (Bouteloua — Hilaria)

A 54. Grama — Tobosa Prairie (Bouteloua — Hilaria)

TA 55. Sagebrush Steppe (Artemesia — Agropyron)

A 58. Grama— Tobosa Shrub Steppe (Bouteloua— Hilaria — Larrea)

TA 59. Trans-Pecos Shrub Savanna (Flourensia — Larrea)

TA 60. Mesquite Savanna (Prosopis — Hilaria)

556 The University Science Bulletin

T 61. Mesquite — Acacia Savanna (Prosopis — Acacia — Andropogon — Setaria)

T 62. Mesquite — Live Oak Savanna (Prosopis — Quercus — Andropogon)

T 63. Foothills Prairie (Agropyron — Festuca — Stipa)

T 64. Grama — Needlegrass — Wheatgrass (Boitteloua — Stipa — Agropyron)

TA 65. Grama — Buffalo Grass (Bonteloua — Buchlo'e)

TA 66. Wheatgrass — Needlegrass (Agropyron — Stipa)

TA 67. Wheatgrass — Bluestem — Needlegrass (Agropyron — Andropogon — Stipa)

TA 68. Wheatgrass — Grama — Buffalo Grass (Agropyron — Botitelotta — Buchlo'e)

TA 69. Blue-stem— Grama Prairie (Andropogon — Boutcloua)

A 70. Sandsage — Bluestem Prairie (Artemesia — Andropogon)

TA 71. Shinnery (Quercus — Andropogon)

T 74. Bluestem Prairie (Andropogon — Panicum — Sorghastrum)

TA 75. Nebraska Sandhills Prairie (Andropogon — Calamovilja)

TA 76. Blackland Prairie (Andropogon — Stipa)

T 77. Bluestem — Sacahuista Prairie (Andropogon — Spartina)

T 81. Oak Savanna (Quercus — Andro'pogon)

T 82. Mosaic of numbers 74 and 100

T 84. Cross Timbers (Quercus — Andropogon)

TA 85. Mesquite — Buffalo Grass (Prosopis — Buchlo'e)

TA 86. Juniper — Oak Savanna (Junipcrus — Quercus — Andropogon )

TA 87. Mesquite — Oak Savanna (Prosopis — Quercus — Andropogon)

T 88. Fayette Prairie (Andropogon — Buchlo'e)

T 95. Great Lakes Pine Forest (Pinus)

TA 98. Northern Floodplain Forest (Populus — Salix — Ulnius)

T 99. Maple — Basswood Forest (Acer — Tilia)

T 100. Oak — Hickory Forest (Quercus — Carya)

T 102. Beech — Maple Forest (Fagus — Acer)

T 104. Appalachian Oak Forest (Quercus)

T 106. Northern Hardwoods (Acer — Betula — Fagus — Tsuga)

T 109. Transition between numbers 105 and 106

T 110. Northeastern Oak — Pine Forest (Quercus — Pinus)

T 111. Oak — Hickory — Pine Forest (Quercus — Carya — Pinus)

T 113. Southern Floodplain Forest (Quercus — Nyssa — Taxodium)

LITERATURE CITED

(Anonymous). 1964. International code of zoological nomenclature. London: International

Trust for Zoological Nomenclature. 176 pp.
\siimead, W. H. 1890. On the Hymenoptera of Colorado; descriptions of new species, notes,

and a list of the species found in the state. Colorado Biol. Assoc. Bull. 1:408-409.
Cameron, P. 1902. Descriptions of new genera and species of American Hymenoptera. Trans.
Amer. Entomol. Soc. 28:369-377.
— . 1903. Descriptions of new species of Hymenoptera taken by Mr. Edward Whymper
on the "Higher Andes of the Equator." Trans. Amer. Entomol. Soc. 29:225-238.
Cockereli., T. D. A. 1898. Note on Agapostemon texanus. Entomol. News 9:27.

— . 1900. Descriptions of new bees collected by Mr. H. H. Smith in Brazil — I. Proc. Acad.

Natur. Sci. Philadelphia, pp. 356-377.
— . 1905. Notes on some bees in the British Museum. Trans. Amer. Entomol. Soc.

31:309-364.
— . 1909. Descriptions and records of bees-XXI. Ann. Mag. Natur. Hist. (8)4:25-31.
— . 1910a. The north American bees of the genus Nomia. Proc. U.S. Nat. Mus.
38:289-298.

. 1910b. Some neotropical bees. Psyche 17:142-143.

. 1910c. Descriptions and records of bees-XXVII. Ann. Mag. Natur. Hist. (8)5:361-369.
— . I'M 2a. Descriptions and records of bees-XLIV. Ann. Mag. Natur. Hist. (8)9:554-568.
— . 1912b. Descriptions and records of bees-XLV. Ann. Mag. Natur. Hist. (8)10:21-31.
— . 1917a. Descriptions and records of bees-LXXVI. Ann. Mag. Natur. Hist. (8)20:235-241.
— . 1917b. Descriptions and records of bees-LXXVIII. Ann. Mag. Natur. Hist. (8)20:436-
441.

Revision of the Bee Genus Agapostemon 557

. 1918a. New halictine bees from Chile. Canadian Entomol. 50:343-345.

. 1918b. Descriptions and records of bees-LXXXI. Ann. Mag. Natur. Hist. (9)2:418-425.

. 1919. Bees in the collection of the United States National Museum-3. Proc. U.S. Nat.

Mus. 55:167-221.
. 1924. Expedition of the California Academy of Sciences to the Gulf of California in

1921. Proc. California Acad. Sci. 12:529-560.'
. 1925. Bees in the collection of the California Academy of Sciences. Proc. California

Acad. Sci. 14:185-215.
. 1927. Bees of the genera Agapostemon and Augochlora in the collection of the

California Academy of Sciences. Pan-Pacific Entomol. 3:153-162.

. 1937a. New and little-known American bees. Amer. Mus. Novitates No. 889, pp. 1-5.

. 1937b. Bees from San Miguel Island, California. Pan-Pacific Entomol. 13:148-158.

■ . 1939. The bees of southern California Islands. Proc. California Acad. Sci. 23:427-436.

. 1949. Bees from Central America, principally Honduras. Proc. U.S. Nat. Mus.

98(3233) :429-490.
Crawford, J. C, Jr. 1901. North American bees of the genus Agapostemon Guerin. Proc.

Nebraska Acad. Sci. 7:156-165.

■. 1906. Some Costa Rican bees. Trans. Amer. Entomol. Soc. 32:157-163.

Cresson, E. T. 1865. On the Hymenoptera of Cuba. Proc. Entomol. Soc. Philadelphia 1865:

1-200.
. 1869. Notes on Cuban Hymenoptera with descriptions of new species. Trans. Amer.

Entomol. Soc. 2:293-298.

. 1872. Hymenoptera Texana. Trans. Amer. Entomol. Soc. 4:153-292.

. 1874. Descriptions of new Hymenoptera. Trans. Amer. Entomol. Soc. 5:99-102.

. 1875. Report upon the collections of Hymenoptera made in portions of Nevada, Utah,

Colorado, New Mexico, and Arizona, during the years 1872, 1873, and 1874. Report

U.S. Geogr. Survey 5:705-736.
Dalla Torre, C. G. 1896. Catalogus Hymenopterorum. Leipzig. 10:643 pp.
Eickwort, G. C, and K. R. Eickwort. 1969. Aspects of the biology of Costa Rican halictine

bees, I. Agapostemon nasutus (Hymenoptera: Halictidae). J. Kansas Entomol. Soc.

42(4):421-452.
Fabricius, J. C. 1775. Systema Entomologiae. Flensburgi et Lipsiae. 832 pp.
. 1793. Entomologia Systematica Emendata et Aucta. Copenhagen. Christ. Gottl. Proft.

2:519 pp.
Forster, J. R. 1771. Novae Species Insectorum, Centuria I. London. 100 pp.
Friese, H. 1916. Zur bienenfauna von Costa Rica (Hym.). Stett. Entomol. Zeit. 77:287-348.
Guerin-Meneville, F. E. 1844. Iconographie du Regne Animal de G. Cuvier, Insectes.

3:447-448.
Holmberg, E. L. 1903. Delectus Hvmenopterologicus Argentinus. Anales del Museo Nacional

de Buenos Aires. (3)2:377-517.
Kuchler, A. W. 1964. Manual to Accompany the Map Potential Natural Vegetation of the

Conterminous United States. Amer. Geogr. Soc. Special Publ. 36. 116 pp.
Lepeletier, L. M. 1841. Histoire Naturelle des Insectes; Hymenopteres. Paris. 2:680 pp.
Lucas, P. H. 1856. Ordre des hymenopteres. In Ramon de la Sagra, Histoire physique,

politique et naturelle de l'ile de Cuba. (Insects) Paris. 7:775-776.
Michener, C. D. 1937. Records and descriptions of North American bees. Ann. Mag. Natur.

Hist. (10)19:313-410.
. 1944. Comparative external morphology, phylogeny, and a classification of the bees

(Hymenoptera). Bull. Amer. Mus. Natur. Hist. 82:151-326.

. 1954. Bees of Panama. Bull. Amer. Mus. Natur. Hist. 104:1-175.

. 1965. A classification of the bees of the Australian and South Pacific regions. Bull.

Amer. Mus. Natur. Hist. 130:5-362.
Michener, C. D., and R. B. Lange. 1958. Observations on the behavior of Brazilian halictid

bees (Hymenoptera, Apoidea) I. Ann. Entomol. Soc. Amer. 51:155-164.
Mitchell, T. B. 1960. Bees of the Eastern United States. Vol. 1. North Carolina Agr. Exp.

Station Tech. Bull. No. 141, 538 pp.
Moure, J. S. 1940. Apoidea Neotropica. Rev. Mus. Paulista (25)2:39-64.
. 1947. Sobre algunas abejas de la provincia Salta. Rev. Soc. Entomol. Argentina

13:218-253.
. 1960. Notes on the types of the neotropical bees described by Fabricius. Studia Entomol.

[Rio] 3:103-104.
. 1964. Two new genera of halictine bees from the Araucanian subregion of South

America (Hymenoptera; Apoidea). J. Kansas Entomol. Soc. 37:265-275.

558 The University Science Bulletin

Muesebeck, C. F. W., K. V. Krombein, H. K. Townes, et al. 1951. Hymenoptera of America

North of Mexico. Svnoptic Catalog. U.S.D.A. Agr. Monogr. No. 2, Washington, D.C.

1420 pp.
Ribble, D. W. 1965. A revision of the banded subgenera of Nomia in America (Hvmenoptera:

Hahctidae). Univ. Kansas Sci. Bull. 45(3) :277-359.
Ridgway, R. 1912. Color Standards and Color Nomenclature. A. Hoen Co., Washington, D.C.

47 pp.
Roberts, R. B. 1969. Biologv of the bee genus Agapostemon (Hymenoptera: Halictidae).

Univ. Kansas Sci. Bull. 48(16) :689-719.
Robertson, C. 1893. Notes on bees, with descriptions of new species. Trans. Amer. Entomol.

Soc. 20:145-149.
. 1895. Notes on bees, with descriptions of new species. Trans. Amer. Entomol. Soc.

22:115-128.
. 1897. North American bees-descriptions and synonyms. Trans. Acad. Sci. St. Louis

7:315-356.

. 1902. Synopsis of Halictinae. Canadian Entomol. 34:243-250.

Say, T. 1837. Descriptions of new species of North American Hymenoptera and observations

on some already described. Boston J. Natur. Hist. 1 (4) :361-416.
Sandhouse, G. A. 1936. The bees of the genus Agapostemon (Hymenoptera: Apoidea) oc-
curring in the United States. J. Washington Acad. Sci. 26:70-83.
Schrottky, C. 1902a. Hymenopteres nouveaux de l'Amerique meridionale. Anales Mus.

Nac. Buenos Aires 7:309-315.

. 1902b. Ensaio sobre as abelhas solitarias do Brazil. Rev. Mus. Paulista 5:330-613.

. 1903. Enumeration des Hymenopteres connus jusqu'ici de la Republique Argentine, de

l'Uruguay et du Paraguay. Anales Soc. Cient. Argentina 55:176-186.

. 1908. Nuevos Himenopteros. Anales Soc. Cient. Argentina 65:225-239.

.1909a. Nuevos Himenopteros Sudamericanos. Rev. Mus. La Plata (2)16:137-149.

. 1909b. Synonymische bemerkungen iiber einige sud-amerikanische Halictinae (Hym.).

Deutsche Entomol. Zeitschr. 4:479-485.
. 1913. Distribucion Geographica de los Himenopteros Argentinos. Anales de la

Sociedad Cientifica Argentina. 75(5) :240.
Smith, F. 1853. Catalogue of Hymenopterous Insects in the Collection of the British Museum.

Part 1. Cat. Hym. British Mus. 1:197 pp.
. 1879. Descriptions of New Species of Hymenoptera in the Collection of the British

Museum. London, 240 pp.
Spinola, M. 1851. Himenopteros. I. Apisiteos. II. Andrenitas. In Claudio Gay, Hist. Fis.

Pol. Chile, Zoologia 6:158-232.
Strand, E. 1910. Beitrage zur kenntnis der hymenopteren fauna von Paraguay. Zool.

Jahrbucher, Abteilung Syst. Geog. Biol. Tiere 29:455-562.
Tins, E. S. G. 1901. Notes on Colorado bees. Canadian Entomol. 32:303-305.
Vachal, J. 1901. Contributions hymenopteriques. Hymenoptera mellifera americana nova.

Ann. Soc. Entomol. France 68:77-81.
— . 1903. Etude sur les Halictus d'Amerique (Hym.). Misc. Entomol. 11:89-104, 121-136.
—.1904. Etude sur les Halictus d'Amerique (Hym.). Misc. Entomol. 12:9-24, 113-128,

137-144.
Vecht, J. van der. 1959. Notes on aculeate Hymenoptera described in the period 1758-1810.

Entomol. Berichten. 19:65-70.
\\ estwood, J. O. 1875. Descriptions of some new species of short-tongued bees belonging to

the genus Nomia of Latreille. Trans. Entomol. Soc. London (3)1875:207-222.
Wolcott, G. N. 1948. Insects of Puerto Rico: Apoidea. J. Agr. Univ. Puerto Rico 32(4):

865-873.

Revision of the Bee Genus Agapostemon

559

Fig. 1. Distributions of West Indian Agapostemon (stippled regions less than 100 fathoms
in depth).

-a
c
<

c
o

rt
U

a
o

s ™

0 ^

3
OS

bo
c
o
l-J

u

-^

o
o

IH

u

rt

c
rt

bo
rt

rt
3

u

rt

u

rt

O

PS
O

u

C

rt
u

u

Pi

rt

rt

3

rt

C
O

Li
3
P-.

a"

>

O

U

poeyi

I'iequesensis

ochromops

columbi

sapphirinus

cyaneus

insularis

jamaicensis

aenigma

X

X
X
X

X
X
X

X

X

X

X

X

X

X X

XXX

X

X

viridulus

obscaratas
cubensis
alayoi
hispaniolicus

swainsonae

X
X
X

X
X

X

\ohliellns

XXX

centratus

X

560

The University Science Bulletin

3=

z(f

€=

-9

•

=o"&?

A. cockerelli

Fig. 2. Distribution of Agapostemon cockerelli.

Revision of the Bee Genus Agapostemon

561

Fig. 3. Distribution of Agapostemon coloradimts.

562

The University Science Bulletin

Fig. 4. Distribution of Agapostemon jemoratus.

Revision of the Bee Genus Agapostemon

563

jf^

\ 5

'••€.!

1/

.-'

.'

J 3

3=d"

C=9
•=d"&9

s/VS

■^-A,--'

A. leuncul

us

)$

Fig. 5. Distribution of Agapostemon leuncidus.

564

The University Science Bulletin

0 = 9 pole al

tf^ = Qdark abdomen

C) = 9 Pale & dark abd-
^fc = 0T g, O pale abdomen

^ —Qf1 & y dark abdomen

^fc=QTT&YPale & dark abdomen

A. melliventris

©

3

Fig. 6. Distribution of Agapostemon mellivetitns. Color of the metasomal terga of females
is indicated by "dark abdomen" (= brown-black to black) and "pale abdomen" (= amber).

Revision of the Bee Genus Agapostemon

565

Fig. 7. Distribution of Agapostemon nasutus with arrows indicating localities where females
with pale amber metasomal terga have been found.

566

The University Science Bulletin

Figs. 8-13. Coarse sculpturing of A. poeyt (left) and finer sculpturing of A. viequesensis
(right): 8-9, mesonotum; 10-11, propodeal shield; 12-13, dorsal region of propodeum.

Revision of the Bee Genus Agapostemon

567

Figs. 14-17. Coarse sculpturing of A. poeyi (left) and finer sculpturing of A. vieqtiesensis
(right): 14-15, lateral view of mesosoma; 16-17, ventral view of genal region.

568

The University Science Bulletin

Fig. 18. Distribution of Agapostemon radiattts.

Revision of the Bee Genus Agapostemon

569

Fig. 19. Distribution of Agapostemon splendens.

570

The University Science Bulletin

Fig. 20. Map showing distribution of Agapostemon texanus males and females (vertically
divided circles); Agapostemon angelicas males only (horizontally divided circles); A. texanus
and A. angelicas males (black circles); A. texanus and/or A. angelicas females (open circles).

Revision of the Bee Genus Agapostemon

571

Figs.
cence re
turing.

21-26. Agapostemon texanus females: 21-22, head with pubescence, and with F
moved; 23-24, variation in mesoscutal punctation; 25-26, variation in propodeal sculp-
Note propodeal carina (Fig. 25, arrow) characteristic of female Agaposten,

572

The University Science Bulletin

Fig. 27. Distribution of Agapostemon tyleri.

Revision of the Bee Genus Agapostemon

573

Figs. 28-29. Ventral view of genal region of Agapostemon virescens (left) and A. tyleri (right).

Fig. 30. Distribution of Agapostemon virescens,

574

The University Science Bulletin

Figs. 31-46. Heads of female Agapostemon in anterior and lateral views: 31-32, A.
\ohliellus (stippling yellow) with lateral (A) and anterior (B) views of labrum; 33-34, A.
viridulus (stippling brown) with anterior (A), antero-distal (B) and lateral (C) views of
labrum; 35-36, A. swainsonae (stippling amber); 37-38, A. hispaniolicus (stippling brown) with
anterior view (A) of apex of labrum; 39-40, A. vieqaesensis (stippling brown); 41-42, A.
nasutus (stippling yellow); 43-44, A. atrocaeruleus; 45-46, A. leunculus (stippling yellow).

Revision of the Bee Genus Agapostemon

575

Figs. 47-62. Heads of female Agapostemon in anterior and lateral views: 47-48, A. inter-
medins (stippling yellow); 49-50, A. coloradinns (stippling brown-black); 51-52, A. tyleri
(stippling brown-black); 53-54, A. virescens (stippling brown-black, hatching yellow); 55-56,
A. radiatus (stippling black, hatching yellow); 57-58, A. femoratus (stippling black, hatching
orange); 59-60, A. coc\erelli (stippling black, hatching yellow); 61-62, A. melliventris (stippl-
ing yellow).

576

The University Science Bulletin

Figs. 63-70. Heads of female Agapostemon in anterior and lateral views: 63-64, A,
mexicanus (stippling amber, hatching yellow); 65-66, A. peninsularis (stippling yellow); 67-68,
A. texanus or A. angelicus (stippling black, hatching yellow); 69-70, A. splendens (stippling
brown-black, hatching amber).

Figs. 71-78. Heads of male Agapostemon in anterior and lateral views (stippling yellow):
71-72, ./. splendens; 73-74, A. angelicus; 75-76, A. texanus; 77-78, A. rhopalocera.

Revision of the Bee Genus Agapostemon

577

Figs. 79-94. Heads of male Agapostemon in anterior and lateral views (Fig. 86, ventral
view): 79-80, A. atrocaeruleus ; 81-82, A. intermedins; 83-84, A. leunculus; 85-87, A. nasutus;
88-89, A. peninsnlaris; 90-91, A. mexicanus; 92-94, A. mellirentris (note size variation of two
specimens, 93-94, collected at the same time and place). Yellow or creamy maculations are
indicated with stippling.

578

The University Science Bulletin

Figs. 95-108. Heads of male Agapostemon in anterior and lateral views: 95-96, A. tyleri;
97-98, A. virescens; 99-100, A. coloradimts; 101-104, .-/. jemoratus (note size variation, 101-
102); 105-106, A. radiatus; 107-108, A. coc/^erelli. Yellow or creamy maculations
indicated with stippling.

are

Revision of the Bee Genus Agapostemon

579

Figs. 109-124. Heads of male Agapostemon in anterior and lateral views: 109-110, A.
\ohliellus; 111-112, A. viridulus, 113-114, A. swainsonae; 115-116, A. hispaniolicus; 117-118,
A. viequesensis; 119-120, A. cubensis; 121-122, A. centratus; 123-124, A. semimelleus. Yellow
or creamy maculations are indicated with stippling.

580

The University Science Bulletin

133

136

137

138

135

140

125-135

136-140

Figs. 125-140. 125-130, heads of Agapostemon in anterior and lateral views, with yellow
maculations stippled: 125-126, A. mourei female; 127-128, A. mourei male; 129-130, A. semi-
melleus female; 131, dorsal view of A. rhopaloccra male right antenna; 132-133, dorsal and lat-
eral views of male A. mourei pronotum showing overlapping mesoscutum (A), acute lateral angle
(B), and angular posterior lobe (C); 134, lateral view of female A. nasutus pronotum showing
yellow maculation (stippling); 135, lateral view of female A. leunculus pronotum showing
carina (arrow) not found on A. nasutus; 136-140, last visible (6th) male sternites: 136, A.
intermedins; 137, A. boliviensis; 138, A. lanosus; 139, A. mourei; 140, A. inca.

Revision of the Bee Genus Agapostemon

581

Figs. 141-154. Anterior views of male Agapostemon hind femur, tibia and basitarsus. Color
yellow or creamy with brown or black maculations. Anterior maculations with left oblique or
vertical hatching; posterior maculations with right oblique or horizontal hatching; anterior and
posterior maculations with cross hatching. 141, A. \ohliellus; 142, A. atrocaeruleus ; 143, A.
lennculus; 144, A. poeyi; 145, .-/. viequesensis; 146, A. jamaicensis ; 147, A. swainsonae; 148,
A. hispaniolicus; 149, A. cubensis; 150, A. viridulus; 151, A. intermedins; 152, A. centratus;
153, A. nasutus; 154, A. rhopdocera. Arrows indicate ventral views of basal ridge and apical
groove on basitarsus.

582

The University Science Bulletin

Figs. 155-166. Anterior views of male Agapostemon hind femur, tibia, basitarsus. Color
yellow or creamy with brown or black maculations. Anterior maculations with vertical hatch-
ing; posterior maculations with horizontal hatching; anterior and posterior maculations with cross
hatching. 155, A. melliventris; 156, A. mexicanus; 157, A. peninsularis; 158, A. virescens; 159,
A. tyleri; 160, A. coloradinus; 161, A. radialus; 162, A. cockerelli; 163, A. femoratus; 164, A.
angelicus; 165, A. texanus; 166, A. splendens. Arrows indicate ventral views of basal ridge and
apical groove on basitarsus.

Revision of the Bee Genus Agapostemon

583

167-172

173-177

Figs. 167-172. Anterior views of male Agapostemon hind femur, tibia, basitarsus. Color
yellow or creamy with brown or black maculations. Anterior maculations with left oblique
hatching; posterior maculations with right oblique hatching; anterior and posterior maculations
with cross hatching. 167, A. mowei; 168, A. inca; 169, A. lanosus; 170, A. boliviensis; 171,
A. semimelleus; 172, A. chapadensis. Arrows indicate ventral views of basal ridge and apical
groove on basitarsus.

Figs. 173-177. Apical view of male gonostyli (right side): 173, A. chapadensis; 174, A,
semimelleus; 175, A. inca; 176, A. mowei; 177A, A. boliviensis; 177B, A, lanosus,

5S4

The University Science Bulletin

Figs. 178-196. Apical view of male gonostyli (right side): 178, Agapostemon viridulus;
179, A. splendens; 180, A. texanus (A, apical stylus; B, medial plate; C, basal stylus); 181,
./. angelicus; 182, A. centratus; 183, A. virescens; 184, A. tyleri; 185, A. coloradinus; 186, A.
poeyi; 187, A. viequesensis ; 188, A. melliventris; 189, A. mexicanus; 190, A. peninsularis;
191, A. \ohliellus; 192, A. intermedins; 193, A. leunculus; 194, A. jemoratus; 195, A. nasutus;
196, A. atrocaeruleus.

Revision of the Bee Genus Agapostemon

585

Figs. 197-214. Dorsal view of male genitalia (showing striae and pubescence on right side
only): 197, Agapostemon virescens; 198, A. tyleri; 199, A. coloradinus; 200, A. splendent:
201, A. texanus; 202, .-/. femoratus; 203, A. melliventris ; 204, A. mexicanus; 205, A.
peninsidai is ; 206, A. centratus; 207, A. rhopedocera; 208, A. viridulus; 209, .-/. swainsonae;
210, A. poeyi; 211, A. leunculus; 212, A. kphliellus; 213, A. atrocaeruleus ; 214, A. intermedins.

586

The University Science Bulletin

215

216

217

218

219

220

215-223

224-227

Figs. 215-227. 215-221, dorsal view of male genitalia: 215, A. boliriensis; 216, A. lanosus;
217, A. mourei; 218, A. inca; 219, A. chapadensis; 220, A. semimelleus; 221, A. nasutus; 222,
lateral view of right penis valve of A. nasutus; 223, ventral view of male genitalia of A.
nasutus (ventral flap removed on right); 224, posterior view of A. viequesensis male
pygidium; 225, posterior view of A. ochromops male pygidium; 226, distal view of A. colo-
radinus right penis valve; 227, distal view of A. tyleri right penis valve.

Revision of the Bee Genus Agapostemon

587

Fig. 228. Distribution maps of the four North American species groups. The dendrograms
indicate phenetic and presumed cladistic relationships within each species group.

588

The University Science Bulletin

INDEX TO INCLUDED TAXA

Valid Agapostemon species are in boldface. When different from current
generic placement, the original genus of description is preceded by a semicolon.
Page numbers are in boldface for descriptions or synonymies and in italics for
illustrations.

abjectus Cockercll, Agapostemon obscuratus,
513.

aeneus (Schrottky), Pseadagapostemon; Aga-
postemon, 443.

aenigma Roberts, Agapostemon, 445, 454, 458,
460, 469, 476, 503, 514, 518, 559.

aeruginosas Smith, Agapostemon, 529.

alayoi Roberts, Agapostemon, 445, 457, 461,
476, 514, 518,559.

andensis (Vachal), Agapostemon: Halictus,
525.

Andrena, 441, 547, 550, 551.

angelicus Cockerell, Agapostemon, 444-446,
451, 453-455, 461-463, 503, 520, 522,
534, 535, 537, 538, 555, 570, 571, 576,

582, 584, 587.

Apis, 441, 547, 550, 551, 554.

arechavaletae (Schrottky), Pseadagapostemon;

Agapostemon, 443.
arenarius (Schrottky), Pseadagapostemon;

Agapostemon, 442, 443.
argentinus Holmberg, Agapostemon, 470.
ascius Roberts, Agapostemon, 445, 448, 458,

463-464.
atcr Friese, Agapostemon nasutas, 495, 497.
atrocaeruleus Friese, Agapostemon, 445, 448,

449, 454, 455, 464-467, 481, 503, 528, 574,

577, 581, 584, 585.
Augochlora, 474.
Aagochloropsis, 442.
azarae Holmberg, Agapostemon, 470.
bicolor Robertson, Agapostemon , 547.
boliviensis Roberts, Agapostemon, 445, 449,

459, 467-468, 483, 485, 494, 505, 580,

583, 586.

bonaerensis (Schrottky), Pseadagapostemon;
Agapostemon, 443.

borealis Crawford, Agapostemon, 533.

brachycerus (Vachal), Agapostemon; Halictus,
533.

brachianas (Schrottky), Ruizantheda ?; Halic-
tus, 443.

bruneri (Crawford), Paragapostemon; Aga-
postemon, 443.

caelestina (Westwood), Paragapostemon;
Nomia, 443.

caeruleus (Ashmead), Aagochloropsis; Aga-
postemon, 442.

californicus Crawford, Agapostemon, 482, 533,
537.

castaneus Schrottky, Agapostemon, 470, 525.

centratus (Vachal), Agapostemon; Halictus,
445, 447, 453, 456, 460, 469-470, 476, 478,
514, 518, 544, 559, 579, 581, 584, 585.

chalets (Vachal), Agapostemon; Halictus, 517.

chapadensis Cockerell, Agapostemon, 445, 448,
449, 458, 459, 470-471, 526, 583, 586.

chiriqaiensis (Vachal), Agapostemon; Halictas,
554.

cillaba (Cameron), Paragapostemon; Nomia,
443, 471.

citricornis (Vachal), Pseadagapostemon ; Halic-
tus, 442.

Clementinas Cockerell, Agapostemon californi-
cus, 533, 534, 537.

cockerelli Crawford, Agapostemon, 445, 451,
453-455, 471-473, 482, 503, 520, 542, 560,
575, 578, 582, 587.

coloradensis Crawford Agapostemon 473, 542.

coloradensis Titus, Augochlora, 474.

coloradinus (Vachal), Agapostemon; Halictus,
445, 446, 452, 473-476, 500, 542, 543, 548,

561, 575, 578, 582, 584-587.

columbi Roberts, Agapostemon, 445, 457, 458,
460, 476-477, 478, 486, 514, 518, 544, 559.

coryliventris Holmberg, Agapostemon, 525.

Corynura, 443.

cubensis Roberts, Agapostemon, 445, 456, 476,
478-479, 514, 544, 559, 579, 581.

cyaneus Roberts, Agapostemon, 445, 457, 476,
478, 479-480, 514, 518, 524, 525, 544, 559.

cyanozonus Cockerell, Agapostemon, 533.

digaeti Cockerell, Agapostemon, 499.

dimidiatus (Lepeletier), Agapostemon; Halic-
tus, 547.

divaricatus (Vachal), Pseadagapostemon;
Halictus. 443.

emarginatus (Spinola), Ruizantheda; Halictus,
443.

erebus Roberts, Agapostemon, 445, 448, 455,
463, 480-481, 503.

experiendus Holmberg, Agapostemon, 470.

fasciatus Crawford, Agapostemon, 499.

fasciatus Nylander, Halictus, 499.

femoralis (Guerin), Agapostemon; Andrena,
441, 550, 551.

femoratus Crawford, Agapostemon, 442, 445,
451, 453, 471-473, 482-483, 503, 520, 522,

562, 575, 578, 584, 585, 587.
festivus Cresson, Agapostemon , 517.
gualanicus Cockerell, Agapostemon nasutus,

508.
Halutus. 441-444, 473, 474, 499, 517, 525,

529, 533, 547, 554.
heterurus Cockerell, Agapostemon, 445, 448,

449, 458, 468, 483-484, 486, 494, 505.
hispaniolicus Roberts, Agapostemon, 445, 456,

457, 460, 461, 476-478. 484, 514, 518, 519,

532, 544, 546, 551, 559, 574, 579, 581.

Revision of the Bee Genus Agapostemon

589

idahoensis Michener, Agapostemon angelicas,

533, 534.
inca Roberts, Agapostemon, 445, 449, 459,

468, 484-485, 505, 580, 583, 586.
insularis Roberts, Agapostemon, 445, 456, 458,

460, 476, 478, 485-486, 491, 514, 518, 544,

559.
intermedius Roberts, Agapostemon, 445, 448,

449, 453, 455, 458, 459, 464, 465, 467, 468,

485, 486-490, 503, 505, 575, 577, 580, 581,

584, 585.
iowensis Cockerell, Agapostemon texanus, 533.
jamaicensis Roberts, Agapostemon, 445, 456,

476, 478, 485, 490-491, 514, 518, 544, 559,

581.
joseantis Friese, Agapostemon, 533.
kohliellus (Vachal), Agapostemon; Halicttis,

445, 447, 456, 457, 461, 476, 478, 491-494,

514, 544, 559, 574, 579, 581, 584, 585.
lanosus Roberts, Agapostemon, 445, 449, 459,

468, 484, 485, 494-495, 505, 580, 583, 586.
Lasioglossum, 444.
leunculus Vachal, Agapostemon, 442, 445,

448, 451, 453-455, 458, 459, 463, 464, 481,

487, 495-499, 500, 503, 515, 523, 524, 554,

563, 574, 577, 580, 581, 584, 585.
martini Cockerell, Agapostemon, 471, 474,

542.
melanurus Cockerell, Agapostemon, 508.
melliventris Cresson, Agapostemon, 445, 452,

454, 455, 469, 487, 496, 499-503, 504, 515-

517, 520, 564, 575, 577, 582, 584, 585, 587.
mexicanus Roberts, Agapostemon, 445, 447,

452-455, 496, 503-504, 515, 576, 577, 582,

584, 585, 587.
mourei Roberts, Agapostemon, 445, 449, 459,

467, 468, 483-485, 494, 495, 504-508, 580,

583, 586.
multicolor Holmberg, Agapostemon, 525.
mtttabilis (Spinola), Ruizantheda (Ruizan-

thedella); Halicttis, 443.
nasita Schrottky, Pseudagapostemon , 442.
nasutus Smith, Agapostemon, 439, 442, 445,

448, 451-453, 455, 458, 459, 463, 464, 481,

492, 496-500, 503, 508-513, 515, 524, 554,

565, 574, 577, 580, 581, 584, 586.
nigricomis (Fabricius), Agapostemon; An-

drena, 547.
Nomia, 441, 443, 471, 517.
obscuratus Cresson, Agapostemon, 445, 456,

457, 476, 478, 484, 513, 514, 518, 544, 551,

559.
ochromops Roberts, Agapostemon, 445, 457,

460, 476, 478-480, 513-515, 518, 524, 525,

544, 559, 586.
olivaceo-splendtns (Strand), Pseudagaposte-
mon; Agapostemon, 443.
Paragapostemon, 441-443.
paidista Schrottky, Pseudagapostemon, 442.
peninsularis Roberts, Agapostemon, 442, 445,

452-455, 496, 500, 503, 504, 515-517, 576,

577, 582, 584, 585, 587.
peruvianus Cameron, Agapostemon, 508.

pissisi (Vachal), Pseudagapostemon; Halicttis,
443.

placidus (Smith), Ruizantheda; Halicttis, 443.

plunfasciatus (Vachal), Agapostemon; Halic-
ttis, 499.

podager (Vachal), Paragapostemon; Halicttis,
442.

poeyi (Lucas), Agapostemon; Andrena, 445,

447, 456, 458, 460, 476-478, 485, 486, 491,
514, 515, 517-519, 525, 544, 546, 559, 566,
567, 581, 584, 585.

portoricensis Cockerell, Agapostemon radi-

atus, 544.
proscriptelltts Cockerell, Agapostemon, 533,

534.
proscriptus Cockerell, Agapostemon, 533.
proximus (Spinola) Ruizantheda; Halicttis,

443.
psammobius Cockerell, Agapostemon calijorni-

cus, 533, 534.
Pseudagapostemon , 442, 443.
pulchra Smith, Agapostemon, 520.
purpweopictus Cockerell, Agapostemon , 508.
radiatus (Say), Agapostemon; Halicttis, 439,

445, 451, 453, 472, 473, 482, 483, 520-523,

547, 568, 575, 578, 582, 587.
rhopalocera Smith, Agapostemon, 445, 447,

448, 453, 496, 523-524, 576, 580, 581, 585.
rohinsoni Cresson, Nomia, 517.
Ruizantheda, 443.

(Ruizanthedella) , Ruizantheda, 443.
sapphirinus Roberts, Agapostemon, 445, 457,
476, 478, 479, 514, 518, 524-525, 544, 559.
semimelleus Cockerell, Agapostemon, 445, 448,

449, 458, 459, 470, 471, 486, 525-528, 579,
580, 583, 586.

semiriridis Cresson, Agapostemon , 551.

sericea Forster, Apis, 547, 554.

sicheli (Vachal), Paragapostemon; Halicttis,
443.

splendens (Lepeletier), Agapostemon; Halic-
ttis, 439, 444-446, 451, 452, 454, 476, 503,
520, 529-531, 535, 569, 576, 582, 584, 585,
587.

subtilior Cockerell, Agapostemon texanus, 533,
534.

sulcatulus Cockerell, Agapostemon, 520.

suljuripes Friese, Agapostemon, 533, 534.

swainsonae Cockerell, Agapostemon, 445, 456,
457, 476, 478, 484, 514, 518, 531-533, 544,
551,559, 574, 579, 581, 585.

tacita (Cameron), Paragapostemon; Nomia,
443.

texanus Cresson, Agapostemon, 439, 444-446,

451, 453-455, 461-463, 465, 466, 472, 473,
475, 482, 483, 488, 489, 497, 501-503, 515,
520-524, 529, 530, 533-542, 543, 548, 550,
555, 570, 571, 576, 582, 584, 585, 587.

tricolor (Lepeletier), Agapostemon; Halictus,

547.
tyleri Cockerell, Agapostemon. 442, 445, 447,

452, 454, 455, 474, 503, 515, 542-544, 548,
572, 573, 575, 578, 582, 584-587.

590

The University Science Bulletin

vandykpi Cockerell, Agapostemon texamis,
533-535, 537.

viequesensis Cockerell, Agapostemon, 445, 447,
457, 460, 476-478, 480, 486, 491, 514, 515,
518, 519, 525, 544-547, 559, 573, 574, 579,
581, 584, 586.

virescens (Fabricius), Agapostemon; Andrena,
445, 451, 452, 474-476, 542, 543, 547-550,
551, 554, 575, 578, 582, 584, 585, 587.

viridulus (Fabricius), Agapostemon; Apis,
441, 445, 447, 456, 457, 461, 469, 476, 478,
479, 484, 513, 514, 518, 532, 544, 550-554,
559, 574, 579, 581, 584, 585.

I'ltlpicolor Crawford, Agapostemon, 495.

xanthorhinus (Cockerell), Psendagapostemon;
Agapostemon, 442.

zosteronedys (Moure), Psendagapostemon;
Agapostemon, 443.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLIX Pages 591-729 March 10, 1972 No. 10

TABLE OF CONTENTS
ABSTRACT 593

INTRODUCTION 593

ACKNOWLEDGMENTS 597

MATERIALS AND METHODS 598

ACCOUNT OF SPECIES 600

Family Emballonuridae 602

Family Noctilionidae 608

Family Phyllostomatidae 615

Chilonycterinae 615

Phyllostomatinae 622

Glossophaginae 629

Carollinae 636

Sturnirinae 643

Stenoderminae 649

Desmodontinae 657

Family Natalidae 664

Family Vespertilionidae 669

Family Molossidae 682

HISTOCHEMISTRY OF THE GASTRIC MUCOSA AND
DUODENAL JUNCTION 690

GASTRIC MORPHOLOGY IN RELATION TO FEEDING . .... 702

SYSTEMATIC RELATIONSHIPS AS REVEALED BY
GASTROINTESTINAL STUDIES

LITERATURE CITED 727

Comparative Morphological and Histochemical
Studies of Stomachs of Selected American Bats

G. Lawrence Forman*

Museum of Natural History
The University of Kansas

ABSTRACT

Stomachs of 13 species of North American hats representing genera of six
families and several different types of feeding habits were examined and com-
pared morphologically, histologically, and histochemically. All species studied
were found to have morphologically distinct stomachs; relationships hetween
gastric structure and physiology, and between gastric structure and consumption
of particulate food materials were investigated. Four histochemical procedures
were employed to examine qualitative and quantitative features of mucins in the
gastric mucosa and the glands of Brunner. Stomachs of insectivorous and carniv-
orous bats studied generally were of simpler structure than the more specialized
stomachs of frugivorous and sangivorous kinds, with various degrees of inter-
mediacy in structure noted in two more or less omnivorous species. Results of
histochemical studies indicate greatest overall production of mucus in insecti-
vorous and carnivorous bats.

Some systematic relationships among families, subfamilies, and genera are
discussed in light of results of gastric studies. Apparent coevolution of gastric-
morphology with a progressive change from albuminous to cellular feeding
within the family Phyllostomatidae also is discussed.

INTRODUCTION

Definitive phylogenetic studies of bats have been hampered by lack of
an adequate fossil record for this group. The earlier-supposed relationships
(Dobson, 1875; Miller, 1907) of major taxa in the Chiroptera have been
questioned in recent years, leading to a search for, and subsequent use of,
a whole suite of approaches and characteristics not heretofor employed
in the classification of these mammals. The purpose of this study was to
examine the structure of the gastric portion of the intestinal tract for possible
use as a criterion in studying relationships among bats. Additionally, be-
cause mammalogists and mammalian paleontologists have long used selected
structures of the teeth, palate, and masticatory musculature in examining
coevolution of morphology and food habits in various groups of mammals,
this study also attempts to correlate gastric morphology with food habits.

The objectives of the present study were fourfold. The first aim was to
increase our knowledge of the gastric structure of bats along with detailed
comparisons of the gastric mucosae. Second, some histochemical properties

♦Present address: Department of Biology, Rockford College, Rocktord, Illinois 61101.

594 The University Science Bulletin

of the gastric and duodenal mucins were examined and described. Third, the
results of the above investigations were evaluated with respect to diverse and
usually obligate feeding habits of various bat species, in an attempt to
demonstrate the presence or absence of coevolution in feeding preferences
and gastric structure. Lastly, gastric morphology and distribution of various
mucosubstances (as revealed by selected histochemical procedures) were
considered in the light of the presently recognized phylogenetic arrange-
ment of the species under examination.

Published comparative studies of mammalian soft anatomy are, for the
most part, inadequate, because the variability among species is rather poorly
documented. Deficiencies are especially apparent for the post-lingual por-
tions of the gastrointestinal tract of mammals. With the exception of a
few accounts dealing with domesticated animals and various groups of
rodents and insectivores, most comparative anatomical studies of stomachs
make reference to only a few species, or concern specific physiological prob-
lems in specific species. Such studies have left gaps in our knowledge of
how much stomachs of relatively closely related taxa differ structurally and
physiologically. Also, because there are few comparative studies of the
gastric mucosal membrane, relatively little has been suggested regarding the
relationship of gastric structure to feeding habits, other than a few well-
entrenched generalizations about certain domesticated mammals. This paper
presents new information regarding the coevolution of food habits with
gastric structure in bats.

According to Cuvier (1805), there were three distinguishable forms that
could be recognized; the transversely elongated tubular stomach possessed
by bats that eat fruit, the globular stomach with closely approximated cardiac
and pyloric orifices exhibited by insectivorous bats, and the elongated
"conical" stomach, with a specialized pyloric caecum, of blood feeders. The
last of these categories is immediately questionable because of the fact that
the species upon which the category is based were probably not blood-
feeding species, but nectar- or fruit-eaters. Huxley (1865), subsequent to
reviewing Cuvier's classification, described the stomach of the truly blood-
feeding vampire bat (Desmodns), in which he found the cardiac portion
to be a tremendously elongated caecum, the cardiac and pyloric orifices
close together, and the region of the pyloric mucosa notably abbreviated.
Owen (1868) gave his own interpretation of various types of stomachs of
bats in relation to diet, including one poorly denned category concerning
"times of taking the food and to the quantity taken."

The important work of Robin (1881) represents the first detailed pub-
lished comparative account of gastrointestinal structure in bats. Robin's
paper includes an excellent review of earlier work, along with new infor-
mation that led him to discard the classifications of Cuvier and Owen as

Studies of Stomachs of Selected American Bats 595

inconsistent with his own observations. Robin noted three shapes of stom-
achs as demonstrated by the 1) Megachiroptera (Pteropodidae), 2) Micro-
chiroptera (Nyctimene) and 3) Desmodus (distinctive type of vampire
bats). He also discussed variation between species in orientation of mucosal
folds and variation in distribution of certain cell types. He also noted the
lack of a "portio cardiaque" in the insect-eating Nyctimene cephalotes, a
structure characteristically present in fruit-eating megachiropterans. Grasse
(1955) reiterated the basic classification of Robin, and additionally noted
differences between the vampire genera Desmodus and Diphylla in the
lengths of the caecal pouches, and that the distribution and structure of
gastric glands were notably different in Desmodus from those of other bats.

Several early authors, notably Rollett (1871) and Langley (1880-82), de-
scribed the structure of the gastric mucosa of bats, but, unfortunately, ne-
glected to name the kinds examined. Langley noted that the fundic glands
were generally short, that those of the greater curvature were especially
long, that parietal cells were relatively few in number, and, although perhaps
unwittingly, was probably the first to describe a pylo-fundic transition zone.
His observations seem to agree well with those included here for Plecotus
townsendii. Rollett's study was similar to Langley's, although Rollett
claimed to have observed reduced numbers of "adelomorphous" or border
cells in summer-taken as opposed to winter-taken specimens. Present evi-
dence questions the reliability of these findings.

Until relatively recently, the bulk of investigation concerning gastroin-
testinal structure in the Chiroptera was almost entirely macroscopic in
nature. Several noteworthy earlier papers deserve consideration. Mathis
(1928a), in an extensive treatise restricted to a discussion of the intestinal
portion of the digestive tract, described the patterns of intestinal folds and
selected glandular areas in the intestines of nine species of bats, and noted
differences in intestinal length and relationships between patterns of in-
testinal folds and types of nutrition. Eisentraut (1950) noted that among
the species he examined the stomachs of insect-eaters tended to be simpler
and less saccular than those of the megachiropteran fruit-eaters, and that
the tendency toward the development of an extensive blind-ended, saccular
fundic caecum could be found only in the Phyllostomatidae of the Micro-
chiroptera. He also noted that two glossophagine bats, Glossophaga and
Choeronycteris, had stomachs that closely resembled the condition in insect-
eating vespertilionid and emballonurid species. However, Eisentraut's illus-
trations of stomachs of some species strongly suggest that they were dis-
tended and therefore not typical of the "empty" condition; his drawings of
stomachs of Carollia and Artibeus are, however, nearly identical to those
included for these genera herein. Fischer (1909) established that the mucous
membrane was better developed in the cranial than the caudal portion of

596 The University Science Bulletin

the stomach. The same was stated for the musculature, but, unfortunately,
no reference was made as to the species examined. Fischer held that a
condition of localized hypertrophied mucous membrane and musculature
was traceable to an adaptation to eating hard insects. Kolb (1954) examined
stomachs of Rhinolophus, Nyctalus, and Old World Eptesicus, and consid-
ered reductions and specializations of certain portions of the gastric mucosa
in these bats in relation to diets predominantly composed of insects. Kolb
concluded that specializations such as folding of the stratified epithelium
at the gastroesophageal junction, greater development of the musculature
and gastric mucosa in the lesser as opposed to the greater curvature (as
noted by Fischer, 1909), and secondary reduction of gastric mucosa in the
blind end of the caecum were in some way related to insectivorous feeding
habits, or to protection of the mucosal membrane from abrasion by chi-
tinous fragments.

Park and Hall (1951) examined the tongues and stomachs of eight
species of New World bats, commenting on the apparent specializations in
both tongue and stomach to incorporate a heavily liquid diet in Desmodus.

Studies that deal specifically with the histological structure of the gastro-
intestinal tract of bats are few, and even fewer are comparative in approach.
Fischer (1909) examined microscopic sections of the gastrointestinal wall,
showing for Barbastella and Vespertilio that the musculature was thicker
within the lesser curvature. Fischer was the first worker to examine in detail
the gastroesophageal junction (Barbastella). He noted that the junction
between the esophageal epithelium and gastric mucosa was abrupt and
described a local thickening of the circular muscle layer at the junction
similar to that described herein for Rhynchonycteris naso. Mathis (1928b)
examined the cellular structure of the glands of Brunner of three vesper-
tilionid species and a rhinolophid. He observed in all cases a clear-cut dis-
tinction in cellular morphology between the glands of Brunner and the
pyloric glands. Mathis noted slight differences in Brunner's glands in dif-
ferent species; his figure of the pyloric junction in Vesperugo (=Pipistrel-
lus) pipistrellus is notably similar to that of Plecotus included here. Kolb
(1954) examined the cardiac glands of two bats (probably Nyctalus and
Rhinolophus), concluding that they probably functioned as a source of lubri-
cant tor materials entering the stomach, and also observed in Rhinolophus
hipposideros that the cardiac glands penetrated the lower esophageal epi-
thelium within a narrow belt at the gastroesophageal junction. An especially
important recent contribution is that of Schultz, 1970. This paper proposes
familial status for a megachiropteran, Harpionycteris whiteheadi, based on
studies of gastrointestinal structure.

A recent paper by Rouk and Glass (1970) presents comparative accounts
<>l gastric structure in five species of North American bats, and reviews
additional pertinent literature not referred to here.

Studies of Stomachs of Selected American Bats 597

Two significant contributions concerning comparative gastrointestinal
morphology, although not specifically considering chiropteran anatomy, arc
those of Allison (1948) and Myrcha (1967) on the Insectivora. These papers
include comprehensive accounts of gross gastric morphology of both Old
and New World insectivores, and I have referred to them occasionally for
comparative purposes, because the Insectivora and Chiroptera are basically
similar in several aspects of gross gastric structure.

Histochemical examination of intestinal sections was undertaken in an
attempt to examine the qualitative and quantitative nature of mucopolysac-
charides in the gastric mucosa and the submucosal glands of Brunner.
Owing to the limited amount of material available for these studies, and
to the nature of the fixatives employed, only a few techniques were used
and the study must be regarded as preliminary in nature. Longitudinal
sections of stomachs examined histologically were included in all mucin
reactions when possible.

Prior to the relatively recent development of procedures for differen-
tially staining and identifying mucous substances, it long had been recog-
nized that mucins of various cells of the gastrointestinal epithelium reacted
differently, and that differences in staining properties of the same cell type
could be found in different species. Clara (1940) noted staining differences
between types of mucous cells specifically with reference to stomachs of
different species, and Liebermann (1SS7-SS, as cited by Jennings and Florey,
1956) had previously suggested, on the basis of chemical investigations on
extracts of mucosa, that different kinds of mucin were present in mammalian
stomachs. Liebermann categorized mucins into those that contained sulphur,
and those that did not. It is of some interest to note that present-day classi-
fications of mucopolysaccharides (see Spicer, 1963) employ the condition
of the presence or absence of sulfate radicals. A few workers, including
Clara (1940) and Gomori (1952), have suggested that gastrointestinal mucins
may be distinguished as either true mucin or mucoid on both histological
and histochemical grounds, although different techniques of fixation and
staining apparently yield inconclusive results (Jarvi and Lauren, 1951). A
comprehensive review and summary of the best known and most widely
used procedures for characterizing mammalian mucopolysaccharides is that
of Spicer (1963).

ACKNOWLEDGMENTS

I am especially indebted to Dr. J. Knox Jones, Jr., for his constant assistance
during the course of this study, and to Dr. William Coil for helpful suggestions
regarding technique and histochemical procedure. I am grateful also to Drs.
Jones, Coil, and William M. Balfour for critical review of the manuscript; to Dr.
William E. Duellman for the loan of microscopic equipment; to Ronald W.
Turner, Timothy E. Lawlor, Hugh H. Genoways, Duane A. Schlittler, and Dr.

598 The University Science Bulletin

Roderick Suthers for providing gastrointestinal tracts of bats; and to Carleton J.
Phillips and James D. Smith for providing stimulating discussion and criticism
during preparation of the manuscript.

Support for this study was provided through an award to me from the
Department of Zoology at The University of Kansas, Lawrence, and by way of a
grant (no. 4193-5706-2) to Dr. Jones from the Bio-Medical Research Fund at
the university. Many Neotropical specimens were collected under the aegis of a
contract (DA-49-193-MD-2215) from the U.S. Army Medical Research and De-
velopment Command.

MATERIALS AND METHODS

Whole gastrointestinal tracts of bats were taken from specimens collected in nets or shot
during crepuscular feeding, and, in one case (Plecotus townsendii) , from torpid individuals
taken in a winter hibernaculum. Most tropical species were netted, and usually were fasted
for a minimum of 10 hours. A total of 28 stomachs, chosen because they were adequately
preserved and devoid of gastric contents, were used for histological and histochemical exami-
nation. An additional 12 specimens, including one of Noctilio leporinus (a species of which I
had no suitable specimens) were examined only macroscopically. Specimens listed as examined
were those sectioned and stained. Descriptions of topography include observations made on
non-sectioned material, whereas the recorded measurements are only from stomachs prepared
for histological examination.

The exact physiological state of specimens at the time of preservation is unknown, al-
though, if used for histological examination, specimens had stomachs devoid of gastric contents.

All material was fixed in either Bouin's or 8% formalin in the ratio of at least 10 parts
fixative to one part tissue by volume, and subsequently preserved in 70% ethyl alcohol.
Several stomachs from specimens preserved in spirits as museum specimens were sectioned
and stained with hematoxylin and eosin, but judged unsuitable for microscopic study.

Material was imbedded in 52.0° Tissuemat and sectioned frontally at 5-7 microns. The
following histological techniques were employed:

1) Harris' hematoxylin and eosin (Harris' formula after Davenport, 1960).

2) Mallory's triple connective tissue stain (Pantin, 1948).
The following histochemical procedures were used:

1) Toluidine blue O (Lillie, 1965).- — A 1:1000 aqueous solution was employed with ex-
posure for 15 seconds with immediate transfer to acetone.

2) Steedman's Alcian blue 8GX for mucins (Steedman, in Lillie, 1965). — Alcian blue
8GX 300 (Allied Chemical Co.) was employed for 60 seconds with a distilled water rinse.
This procedure employs an aqueous solution of the dye and was used at a pH of 7.1.

3) Maxwell's periodic acid Alcian blue for gastric mucus (Maxwell, in Lillie, 1965). This
procedure, which employs exposure of sections to periodic acid oxidation prior to staining was
used as a supplement to Steedman's method, and as an additional check for glycogen.

4) Periodic Acid SchifT sulfite leucofuchsin reaction (Lillie, 1965). Control slides (no ex-
posure to periodic acid) also were employed.

In addition to successful histochemical procedures described above, several other types were
attempted, but did not provide conclusive results. One of these was digestion with testicular
hyaluronidase, followed by toluidine blue O (Lillie, 1065). Sections exposed to 18 hours
digestion in 1:10,000 aqueous hyaluronidase at 37°C revealed no distinguishable alteration
ot toluidine blue metachromasia in comparison to control sections, therefore yielding negative
results in a test for chondromucins and hyaluronic acid.

The two Alcian blue procedures were run as single series, one with, and one without,
periodic acid oxidation prior to staining. The toluidine blue O procedure was carried out
twice, once as a single series, then as a control series for the hyaluronidase lysis.

Scoring of the intensity of Alcian blue and PAS reactions was as follows: exceptionally
strong (-}--)- _)--(-), moderately strong ( + + + ), moderate ( + + ), weak (+), barely precepti-
ble (trace), not observed ( — ). Toluidine blue O was scored simply as plus ( + ), the presence
of demonstrable metachromasia, and minus ( — ), metachromasia not observed.

Studies of Stomachs of Selected American Bats

599

VENTRAL
VIEW

pyloric constriction

sulcus intermedius

RIGHT

sulcus angulans

cardiac vestibule

LEFT

fundic caecum

POSTERIOR

CRANIAL
VI EW

DORSAL

RIGHT

LEFT

VENTRAL

Fig. 1 (upper). Diagrammatic representation of the stomach of Carollia perspicillata (ven-
tral view). Note the location of various sulci.

Fig. 2. (lower). Stomach of C. perspicillata (diagrammatic), viewed cranially.

Anatomical terminology and descriptions of location and direction of parts used with
reference to the gastrointestinal tract frequently are undefined or unclear, sometimes inap-
propriate, and otherwise variously misused in the literature. For these reasons, a list of defined
terms is included here in addition to the brief discussion of orientational terminology to follow.

All descriptions and discussions of gastric morphology are based on location of the stomach
in situ in the abdominal cavity. Anterior (craniad) and posterior (caudad) are here con-
sidered those directions of the lesser and greater curvatures, respectively (Fig. 1), usually con-

600 The University Science Bulletin

sidered dorsal and ventral, respectively, in the literature. Dorsal is here considered that surface
of the stomach nearest the vertebral column, whereas ventral is that surface farthest from (or
opposite to) the vertebral column (Fig. 2). All gastric structures are distal to the apex of the
fundic caecum, and likewise proximal to the pyloric sphincter.

All measurements of gross structures are in millimeters, and are included to give an
indication of stomach size and orientation of various features with respect to others. Measure-
ments of depths of the various zones of gastric mucosa were obtained using a dial micrometer.
The ranges are recorded in microns.

All photographs are of structures stained with Harris' haematoxylin and eosin unless other-
wise indicated.

Mid-longitudinal outline drawings represent composites of specimens of each species, and
were produced by projecting representative sections onto paper using a photographic enlarger,
tracing the outlines for each specimen, and producing a composite outline by means of a light
table. Within some mid-longitudinal drawings, structures not observed in medial sections,
although important enough to be included in the illustrations, are depicted by broken lines.

Uncommon morphological terms used in the accounts that follow are listed and defined
below.

Cardia. — Limited area of stomach surrounding gastroesophageal junction.

Cardiac vestibule (same as cardia). — Part of stomach between gastroesophageal junction and

general plane of lesser curvature.
Elastic stomach. — That region of the stomach lateral to gastroesophageal junction terminating

in a blind end, usually containing some elastic fibers.
Fundic caecum. — Generally, area between gastroesophageal junction and left lateral terminus

of stomach.
Fundus. — That part of the stomach that secretes acid and pepsin, lying between cardiac and

pyloric parts.
Incisura cardiaca. — Notch formed between cardiac vestibule and fundic caecum at entrance of

esophagus into stomach.
Pyloric constriction. — External furrow at pyloric sphincter.
Pyloric bend. — Cranial recurvature or arching of terminal stomach.
Pylorus. — Glandular region of stomach composed entirely of cells resembling mucoid neck

cells, located proximal to pyloric sphincter.
Saccular stomach. — That portion between gastroesophageal junction and apex of fundic

caecum.
Sulcus angularis. — Furrow (external) formed generally at, or slightly proximal to, pyloric bend.
Sulcus intermedius. — Furrow (external) approximately midway between pyloric bend and

pyloric sphincter.
Terminal stomach. — That portion between gastroesophageal junction and pyloric sphincter.
Transition zone. — Zone of transitional gastric glands between fundic mucosa and other glandu-
lar types.

ACCOUNT OF SPECIES

The gastrointestinal tracts of all Chiroptera examined consisted of the
same typical segments that have been described for other kinds of mammals:
a thin-walled, extremely narrow esophagus lacking submucosal or mucosal
glands and lined entirely with stratified squamous epithelium, within which
the usual four strata of squamous cells could be observed; a distinctive,
saclike, muscular stomach (with the exception of the blind tube of Des-
modus rotundus) completely lined with well-differentiated, usually simple,
branched, tubular mucosal glands which are differentiated by the occurrence
of particular cell types and identifiable within specific zones; a relatively
long small intestine (especially long in Desmodits) with thin, muscular

Studies of Stomachs of Selected American Bats 601

walls, little loose submucosal tissue, and long, narrow villi; and a well-
developed colon including numerous goblet cells and somewhat more ex-
tensive non-muscular submucosa than the small intestine.

The wall of the chiropteran stomach is composed of the tunica mucosa,
tunica submucosa, tunica muscularis, and tunica serosa. Tunica serosa lines
the entire external surface of the stomach; no squamous tissue is present
except in the uppermost portion of the cardiac vestibule of Natalus stra-
mineus. The amount of non-muscular submucosa in the stomachs of most
bats examined was found to be extremely small in comparison to many
other mammals. Infrequent to numerous accumulations of lymphatic tissue
within the tunica propria were observed in some, but not all, specimens.
Vascularization of the tunica propria was similar in all species examined
with the exception of Desmodus rotundas, in which the complements of
arterioles and venules are extensive.

In all species the tunica mucosa, submucosa, and thin sheet of smooth
muscularis mucosa just beneath the gastric glands are thrown into longi-
tudinal or transverse (or both) and anastomosing rugae. These rugae, few
to many depending on the species, extend throughout the stomach, termi-
nating at the gastroesophageal and pylo-duodenal junctions. Rugae may or
may not be reduced at the apex of the elastic stomach (fundic caecum),
and are often poorly defined in a fully distended stomach. The tunica
muscularis proper consists of an inner circular and outer longitudinal layer,
with extensive variation in depths of layers in various species. A third,
extremely thin layer of smooth muscle fibers lies to the inside of, and
oblique to, the two outer layers, usually separated from them by a thin
zone of loose submucosal tissue.

In all species examined, I noted a partial or pronounced cranial recurva-
ture of the aboral end of the stomach, often beginning substantially proximal
to the pyloric sphincter (Natalus perhaps an exception). This condition has
been observed in other species of bats, including members of the Mega-
chiroptera, and immediately distinguishes stomachs of Chiroptera from
many members of the Insectivora, notably soricids examined by Myrcha
(1967). Curvature of the stomach is known in some Insectivora (Erina-
ceidae, Talpidae) and its significance as a diagnostic character in the
Chiroptera will be discussed later.

All species examined (except Desmodus) have four distinctively recog-
nizable zones of gastric mucosa — cardiac, fundic, pylo-fundic transition, and
pyloric. Although Desmodus may in fact have all four zones represented,
mucous glands occupying the area between oral and aboral orifices of the
stomach are identical and are here considered to be pyloric glands. Cardiac
glands are mucus-producing and appear to contain a single cell type. Fundic
mucosa usually occupies the majority of the inner surface area and is com-

602 The University Science Bulletin

posted of tubular glands containing mostly zymogenic cells that are the
proposed sites of pepsin secretion, parietal cells that secrete hydrochloric
acid, and various mucus-producing cells. The transitional zone is a highly
variable area, both dimensionally and structurally, representing the change
from fundic to pyloric mucosa within the terminal stomach. The pyloric
zone, always proximal to the pyloric sphincter, is a second mucus-producing
area composed of relatively straight tubular glands with a single cell type.

The stomach of the vespertilionid Plecotus townsendii was considered to
be relatively generalized among bats examined. Because of its notable sym-
metry and seeming lack of unusual specialization, the stomach is often
used in the accounts in comparison or contrast to structures in other species
where clarity by means of comparison appeared necessary. Species are
arranged in the following accounts in phylogenetic order following Hall and
Kelson (1959).

Due to the small sample sizes of species employed in this study, the
conclusions drawn here must be regarded as preliminary, and the study of
a pilot nature. Further study utilizing many more specimens is encouraged,
in order to document intraspecific variation.

Family EMBALLONURIDAE

Rhynchonycteris naso (Wied-Neuwied, 1820)

Specimens examined (2). — Two specimens from Nicaragua (KU 105911,
105915).

Overall dimensions. — Greatest length 7.4-8.1; greatest breadth 3.1-3.8;
gastroesophageal junction to pyloric sphincter 2.4-2.9; gastroesophageal junc-
tion to apex of fundic caecum 2.5-2.8; breadth pylorus at sphincter 1.3-1.4.

General morphology. — Stomach tubular, reniform, generally symmetrical
although tapering gradually throughout from fundic caecum to pylorus;
pronounced curvature in frontal plane throughout body with apex of fundic
caecum directed antero-laterad; tubular stomach gradually curved anteriad
two-thirds distance from apex of fundic caecum to pyloric sphincter, con-
tinuing craniad as narrow, relatively long, tubular endpiece directed cranially
and parallel to cardiac vestibule; lesser and greater curvatures nearly parallel
throughout; terminal endpiece lateral to pyloric curvature largest among
insectivorous species examined; stomach rounded on sides, convex on both
dorsal and ventral surfaces. Cardiac vestibule present although short and
narrow, musculature of wall well developed as through much of lesser
curvature; vestibule at marked angle to lesser curvature and relatively
uniform in breadth with little expansion posteriorly; incisura cardiaca rela-
tively deep and narrow, created by marked angle of cardiac vestibule to
lesser curvature; distinctive local thickening of both longitudinal and circu-
lar muscle layers within incisura, but no apparent separation of circular

Studies of Stomachs of Selected American Bats

603

2mm

Fig. 3. Mid-longitudinal representation of the stomach of Rhynchonycteris naso. Explanation
of symbols: CG, zone of cardiac glands; BG, zone of Brunner's glands; PG, zone of pyloric
glands; TZ, transitional zone; FG, zone of fundic glands.

fibers into laminated sheets at incisura; additional local thickening of circu-
lar muscle layer within cardiac vestibule opposite incisura cardiaca, possibly
forming anatomical cardiac sphincter (Fig. 3) (see discussion of muscula-
ture). Fundic caecum well developed, dilated, expanded superiorly and
with surface rounded throughout; structurally continuous with midstomach
(not anatomically distinctive chamber), oriented longitudinally on dorsal
surface (no dorsal hook). Terminal stomach distal to cardiac vestibule
long, decreasing gradually in breadth to just beyond midpoint where promi-
nent sulcus intermedius distinguishes proximal funnel-like portion from
narrow, relatively long terminal endpiece (2.5 mm in length); terminus
lateral to sulcus recurved slightly, nearly parallel to cardiac vestibule, uni-
form in breadth to pyloric sphincter. Pyloric sphincter long and narrow,
valves of equal length as measured on lesser and greater curvatures; ex-
panded through medial half to appear narrowly spoon-shaped in cross-

604 The University Science Bulletin

section; differing from that of other bats examined in being equivalent in
depth to circular muscle of pyloric tube that gives rise to sphincter within
both lesser and greater curvatures; valve aperture narrow. Pyloric junction
slightly superior to gastroesophageal junction in situ. Musculature generally
thin throughout much of stomach wall, thickened locally; unusual arrange-
ment of circular layer in lesser curvature. Rugal folds numerous, reduction
of mucosal lining throughout relatively large portion of fundic caecum
and greater curvatures. Lamina propria mucosae extremely thin except
within fundic caecum and portions of terminal endpiece of terminal stomach.
(Hands of Brunner moderately abundant. Stomach conforming in most
features to general configuration common to other obligate insectivorous
bats.

Musculature. — Typical three layers observed in both specimens. Longi-
tudinal layer a simple thin sheet throughout most of stomach, locally thick-
ened 1) in incisura cardiaca where there is substantial increase in depth
within the depression, 2) throughout longitudinal zone of fundic caecum
where it is somewhat thicker than overlying circular layer through much of
caecal region, and 3) in the mid-region of terminal stomach on greater
curvature. Circular muscle relatively narrow throughout pyloric tube in
comparison to other bats examined, uniform in depth throughout length
of pyloric tube as well as equal in depth in lesser and greater curvatures;
circular layer locally thickened within broad area in lesser curvature im-
mediately opposite incisura cardiaca, equivalent in depth to that of incisura
and extending from midpoint between gastroesophageal junction and sulcus
intermedius to just beneath gastroesophageal junction, resulting in "ring"
of thickened musculature at lower end of cardiac vestibule, an anatomically
plausible cardiac sphincter inferior to gastroesophageal junction although
clearly not a valvular closure mechanism; circular layer extremely thin
through fundic caecum and majority of greater curvature, shallower than
longitudinal layer; bundling of circular muscle occurring only in zone of
greater curvature proximal to pyloric bend. Muscularis mucosae typical,
thickest at incisura cardiaca and in fundic caecum. Lamina propria mucosae
(beneath muscularis mucosae) abundant only in fundic caecum. Depth
of longitudinal layer 15 to 100 (incisura cardiaca), depth of circular layer
15 to 140 (incisura cardiaca and opposing wall), depth of mucularis mucosae
15 to 25.

Organization of gastric mucosa. —Rugal folds deepest and most abundant
within tubular portion, originating at pyloric sphincter and radiating into
mid-stomach, becoming progressively more widely spaced as tube expands;
frequency of folds constant through tubular portion; least number of folds
originating along lesser curvature from pyloric bend to cardiac vestibule,
consistently directed toward lower apex of fundic caecum; upper wall of

Studies of Stomachs of Selected American Bats

caecum (lesser curvature) with relatively frequent, short, transversely-
directed rugae extending downward to mid-region of dorsal and ventral
surfaces of caecum; folds well developed and longitudinal within cardiac
vestibule; major portion of greater curvature and all of lateral wall of
fundic caecum lacking prominent rugal folds (gastric mucosa also distinc-
tively reduced in depth), which originate in pyloric tube and along lesser
curvature and terminate in lower middle area of ventral and dorsal sur-
faces of body of stomach; resulting pattern that of transverse to progressively
more longitudinally directed rugae from caecum to pyloric tube through
approximately upper three-fourths of body of stomach and all of tubular
stomach. All mucosal gland types represented; cardiac and pyloric zones
relatively short, fundic mucosa occupying remaining surface area except
for unusually broad transitional zone occurring through nearly all of
pyloric tube; no Brunner's glands within gastric mucosa. Progressive de-
crease in mucosal depth from middle fundus through transitional zone
and pyloric glands to pyloric sphincter with slight secondary increase within
terminal pyloric mucosa, cardiac glands comparable in depth to transitional
zone; deepest mucosa within lesser curvature of fundus; glands within
apex of fundic caecum and in the basal, non-tubular portion of greater
curvature reduced in length.

Cardiac glands. — Zone of true cardiac glands extremely narrow; transi-
tional zone to fundic mucosa slightly longer than cardiac zone; cardiac
glands short and broad, composed of single type of cell although some local
variation in morphology of nuclei occurring within gland; cells within
bulbous base small, elongate, with giant spherical nuclei; upper cells smaller,
more cuboidal with nuclei decreasing progressively in size although remain-
ing spherical throughout to epithelium; lumen broad, increasing only
slightly in breadth toward surface; epithelium in small clumps between
adjacent glands because luminal openings of glands broad and close together,
cells elongate, nuclei compressed laterally. Transitional zone to fundic mu-
cosa with numerous (but scattered) parietal cells throughout length of
glands. Interglandular and basal lamina propria no more abundant than
within fundic zone. Depth of cardiac mucosa 160 to 170.

Fundus.— Most variable in structure of all bats examined; true simple
tubular glands of branched type only within upper three-fourths of stomach,
reduced (dwarfed) to thin layers of oxnyctic, chief and mucous neck cells
within lower fourth (including much of fundic caecum); glands deepest
between fundo-pyloric bend and medial limit of fundic caecum (inclusive
of cardiac vestibule), moderate in depth within most of tubular stomach,
decreasing slightly into transitional zone; reduction of glands within lower
fundus resulting from general reduction, sometimes complete loss, of one
or more cell types characteristic of fundic mucosa. Depth of fundic gland:

606 The University Science Bulletin

Fig. 4. Argentaffin cell (AC) at base of fundic gland of Rhynchonycteris naso. ZC, zymo-
genic cell.

180 to 220 within lesser curvature, cardiac vestibule, part of fundic caecum,
and zone beneath transitional area within greater curvature; 40 to 100 within
zone of dwarfed glands.

Zymogenic cells occupying only lower fifth of gland, most abundantly
concentrated in stomach within lesser curvature and tubular portion of
greater curvature, and only scattered infrequent basal elements through most
of remaining area; zymogenic cells small, pyramidal, largest within ex-
panded basal arc, decreasing in size and becoming more elongate within
upper area of intraglandular distribution; nuclei of zymogenic cells small,
spherical, and juxtaposed to basal membrane.

Argentaffin cells (Fig. 4) observed infrequently, usually among parietal
instead of chief cells, approximately same size as mucous neck cells, being
extremely narrow, triangular, and with nucleus at inner border; point of
triangle peripheral; nucleus extremely small, spherical, dense argentaffin
granules lateral (Fig. 4).

Parietal cells abundant within upper 75% of glands throughout lesser
curvature from upper midpoint of fundic caecum to pylo-fundic transition;
also common through small area of lesser curvature just beneath transitional
zone, extremely sparse to absent through most of greater curvature; parietal
cells highly variable in size, distinctive intraglandular size gradient with

Studies of Stomachs of Selected American Bats 607

increase from bottom to top of distribution, uppermost cells extremely
large; cells most densely distributed just beneath necks of gastric pits where
unbroken rows of five to seven often observed, rarely found among chief
cells; sparsely distributed in greater curvature and generally smaller there
than in areas of abundance; usually ovoid.

Mucous neck cells extremely small as compared to adjacent parietal cells;
rarely occurring singly, usually observed in clumps of four or five, most
often situated along inner margin of tubule; nucleus half size of parietal
cell nuclei, oblong, irregular in shape, and with little cytoplasm; mucous
neck cells composing most of cellular complement within stunted glands
of greater curvature and fundic caecum in absence of many chief and
parietal cells.

Gastric pits shallow through lesser curvature (15-20% length of glands),
extending to base within dwarfed glands of greater curvature in absence
of non-mucoid cellular elements; cells lining pits similar to lower mucous
neck cells, although lumenal mucous border moderate in depth; no promi-
nent lateral flattening of nuclei within lesser curvature although neck cell
nuclei compressed within dwarfed glands of lower stomach; surface epi-
thelium thick, nuclei typically basal, club-shaped, and twice length and
breadth of gastric pit nuclei; interglandular and basal lamina propria ex-
tremely sparse.

Pylo-jitndic transition zone. — Extensive area; parietal cells persisting
through terminal endpiece nearly to pyloric sphincter, resulting in extremely
narrow pyloric zone; zymogenic cells gradually reduced in number from
initial margin of zone, large numbers of mucous cells occurring basally
with parietal cells throughout zone; loss of parietal cells along terminal
border of zone also gradual, with persisting cells at aboral margin usually
basal within glands; increase in depth of gastric pits progressive along
with loss of parietal elements, beginning in midregion of transition; slight
reduction in depth from fundic mucosa; depth of transitional mucosa ap-
proximately 150.

Pyloric glands. — Pyloric mucosa confined to extremely narrow area at
pyloric sphincter owing to extensive aboral advancement of parietal cells;
glands short and broad, equivalent in breadth to fundic glands; cells within
basal third elongate and narrow, with extremely small, spherical, basal
nuclei; cells within middle third more globular, nuclei twice size as those
of basal cells and ovoid, upper cells becoming progressively broader and
shallower toward surface epithelium with nuclei increasing correspondingly
in degree of lateral compression; glands relatively uniform in depth and
breadth throughout distribution, basal fourth of tubule expanded into bulb,
remaining portion uniform in diameter to surface epithelium; lumen nai
rowest within bulb, increasing gradually in breadth to surface; surface epi

608 The University Science Bulletin

thelium is in small clumps as orifices are broad and close together, cells
narrow, elongate with similarly shaped nuclei; loose lamina propria more
abundant beneath pyloric mucosa than fundic, typically sparse between
glands; depth of pyloric mucosa 150 to 170.

Glands of Brunner.—Con&ned to short, although broad, massive ring
occupying only proximal 1.0 mm of duodenum; deepest along distal surface
of pyloric sphincter, decreasing sharply distally; tubules moderate in breadth,
cross-sectional morphology identical to that of most other bats examined
except nuclei extremely small, juxtaposed to basement membrane, and
perfectly spherical; glands restricted to duodenum, always submucosal.

Remarks. — The antero-posteriorly expanded tubular stomach of Rhyncho-
nycteris is well representative of the "insectivorous" condition, with a few
noteworthy specializations.

The musculature is substantially better developed on the lesser, as com-
pared to the greater, curvature, a characteristic unique to this species. This
condition is greatest at the gastroesophageal junction where local thickening
occurs, suggestive (although not conclusively indicative) of the presence
of a non-valvular cardiac closure mechanism. Continuation and expansion
of the work of Botha (1958, a-c) on gastroesophageal closure mechanisms
in mammals would be enlightening in this regard.

The pattern of gathering of the mucosal folds is unique in two respects.
Many rugae originate along the lesser curvature and are directed somewhat
downward, toward the lower apex of the fundic caecum, rather than strictly
longitudinally as in other obligate carnivores. Also, although not unique
among carnivorous bats (observed to some extent in Plecotus townsendn in
this study), transversely directed rugae are much in evidence within the
upper fundic caecum. Secondly, unparalleled reduction of mucosal folds
and depth of glands occurs throughout much of the greater curvature.

The gastric mucosa is unique in its impoverished complement of zymo-
genic cells within the fundic zone (excepting the overall reduction of the
gastric mucosa in Desmodus). The transitional zone is extensive in terms
of lengdi and parallels that of Pteronotas in being not only a distinctive
portion of the mucosa, but by limiting the pyloric mucosa to an extremely
narrow band at the pyloric sphincter.

Family NOCTILIONIDAE

Noctilio labidis (Kerr, 1792)

Specimens examined (2). — Two specimens from Nicaragua of Noctilio
labidis minor (KU 105941, 116690).

Overall dimensions. —Greatest length 12.9-13.5; greatest breadth 7.0-8.0;
gastroesophageal junction to pyloric sphincter 5.8-6.1; gastroesophageal junc-
tion to Eundic caecum 4.4-4.7; breadth pylorus at sphincter 2.4-2.6.

Studies of Stomachs of Selected American Bats

609

2mm

Fig. 5. Mid-longitudinal representation of the stomach of Noctilio labialis; SSE, stratified
squamous epithelium; CSPH, cardiac sphincter. For explanation of other symbols, see Figure 3.

General morphology. — Stomach tubular, relatively short and narrow,
only slightly curved in frontal plane; lesser curvature arched anteriorly;
anterior and posterior surfaces of stomach convex resulting in egg-shaped
stomach. Fundic caecum short and narrow, tapering to narrow apex, di-
rected laterally, not chambered. Esophagus entering stomach at slight angle;
cardiac vestibule lacking gastric mucosa, stratified squamous epithelium
extending posteriorly beyond arc of lesser curvature forming lip at entrance
of esophagus into stomach, and incorporated as lining of well-developed
cardiac sphincter of mixed type as defined by Botha (1958 a,b); sphincter
composed of both striated and smooth muscle elements.

Cardiac sphincter occurring as distinctive fold of stomach wall (Figs.
5-6) at entrance of esophagus into stomach, asymmetrical, best developed
on pyloric side; composed of circular muscle layer, muscularis mucosae and
surrounding lamina propria, and stratified squamous epithelium, but longi-
tudinal muscle layer not incorporated into fold; circular muscle projecting
downward to apex of lip and recurving sharply craniad to lesser curvature,
subdivided into extensive numbers of laminated sheets, nearly uniform in
breadth throughout junction; muscularis mucosae extensive throughout
sphincter, fibers separating into radiating bundles at apex of lip, continuing
as narrow double layer at onset of gastric mucosa; most of sphincter lip
lined with thin, rugose layer of esophageal epithelium, lateral edge of lip

610 The University Science Bulletin

partially lined with gastric glands; loose connective tissue abundant above
and below muscularis mucosae within that portion of lip lined with squa-
mous epithelium. Prominent external constriction at junction of abdominal
esophagus and gastric sac, incisura cardiaca absent, cardiac orifice broad;
squamous epithelium at cardiac junction and adjacent cardiac mucosa
similar in depth.

Terminal stomach distal to cardiac junction moderate in length, curved
relatively slightly craniad, tapering gradually to sphincter; pyloric bend slight;
orifice at sphincter narrow because flaps of sphincter long; external pyloric
constriction prominent, partly owing to extensive expansion laterally within
duodenum of submucosal glands of Brunner; musculature of pyloric sphinc-
ter narrow, flaps relatively long, longer and broader on greater than lesser
curvature; circular muscle layer of duodenum sole muscular contribution to
sphincter as circular layer discontinuous from pylorus to duodenum with
pyloric complement ending abruptly at sphincter; sulcus intermedius present
and well developed approximately 2 mm proximal to pyloric sphincter;
sulcus angularis occurring midway within terminal stomach. Pyloric and
cardiac orifices at approximately equivalent levels but not unidirectional.
Lamina propria mucosae abundant and dense within rugae, less abundant,
although prominent beneath muscularis mucosae of outer wall. Rugal folds
extensive in number and depth throughout stomach, unusually prominent
through portions of greater curvature; gastric mucosa moderate in depth,
showing familiar pattern of reduction in depth between folds.

Musculature — Musculature generally deep throughout stomach, best de-
veloped within most of greater curvature and terminal 3 mm of lesser
curvature. Longitudinal layer a thin sheet throughout all of lesser curva-
ture and cardiac junction, abruptly thickened and moderate in depth
throughout fundic caecum; locally thickened to prominent layer through
medial 4 mm of greater curvature, again narrowing and remaining narrow
through terminal portion of greater curvature to pyloric sphincter. Circular
layer relatively narrow through fundic caecum, locally thickened within
medial portion of greater curvature (along with underlying layer), again
narrowing just proximal to sulcus intermedius where local thickening of
circular layer only is present, narrowing again before increasing sharply
in breadth to pyloric sphincter; correspondingly thickened zones between
cardiac junction and pyloric sphincter within lesser curvature, although
locally thickened areas slightly deeper; gathering of circular fibers into
bundles extreme in, and restricted to, fundic caecum and greater curvature
laterally through thickened midregion. Muscularis mucosae typically thin
layers within stomach wall, slightly thicker within rugae, most prominent
at cardiac junction. Ganglia of Auerbach's plexus more abundant within
greater curvature than in lesser curvature. Depth of longitudinal layer

Studies of Stomachs of Selected American Bats 611

30 to 60, circular layer 30 to 90 (caecum) to 300 (pyloric tube), muscularis
mucosae less than 10 up to 20.

Organization of gastric mucosa. — Mucosa gathered into longitudinal
parallel ridges from apex of fundic caecum to pyloric sphincter; rugal folds
best developed of any bat examined; shallow within caecum, increasing
gradually in depth from origin in fundic caecum through midstomach,
decreasing again into pylorus; folds gradually decreasing in length and in-
creasing in depth from lesser curvature posteriad to greater arc. Folds within
medial half of greater curvature longitudinally directed, relatively short
finger-like projections of tremendous depth (up to 3.5 mm) directed an-
teriorly into gastric lumen; few folds originating from lesser curvature (only
within upper wall of fundic caecum). No transversely oriented rugae;
moderately developed "zig-zag" configuration in folds within posterior
three-fourths (absent within region of lesser curvature) from apex of fundic
caecum to sulcus intermedius, never truly interdigitating; folds shallow,
infrequent at gastroesophageal junction. Mucosa shallowest in depth at
gastroesophageal junction, uniform in depth throughout remaining stomach
excepting slight reduction within interrugal furrows in comparison to sides
and apexes of rugae; no appreciable reduction within (or at apex of) fundic
caecum; deepest mucosal glands on sides of rugae within terminal portion
of transitional zone.

Zones of gastric mucosa atypical among carnivorous types examined,
apparently accounting for extremely large transition zone, cardiac glands
occupying typically narrow band at gastroesophageal junction, equivalent
in breadth throughout; true fundic mucosa restricted to approximately left
half, inclusive of region of cardiac junction; transitional zone to pyloric
mucosa occupying most of terminal half, thus limiting true pyloric glands
to narrow belt at sphincter; all zones relatively symmetrical regarding length
of distribution on lesser and greater curvatures (Fig. 5).

Cardiac glands— Zone of cardiac glands in complete absence of parietal
cells, extremely narrow; glands short and broad relative to remaining mu-
cosal elements; single cell type apparently present but extensive morpho-
logical variation within tubules; cells within basal fourth small and many
with extremely large, centrally located nuclei; cells in upper three-fourths
of tubule becoming progressively larger, more elongate, with increasingly
small, more ovoid nuclei toward upper limit of tubule; tubules not highly
convoluted, bases usually although not always bulbous and uniform in
breadth through upper portion; lumen moderate in breadth, slightly broader
at base than neck; smooth muscle fibers and areolar connective tissue more
abundant beneath and between tubules than in other mucosal areas; surface
epithelium thin, cells narrow with large ovoid nuclei located in lower
middle portion; depth 120 to 140.

612 The University Science Bulletin

Fundus. — Relatively restricted fundic zone characterized by marked uni-
formity in depth and morphology of glands; deepest glands within mid-
stomach, decreasing only slightly toward apex of fundic caecum, slightly
reduced in depth within furrows. Consistent zonation of cellular elements
within fundic glands with little variation in length of zones throughout;
lower 30% restricted to zymogenic cells with occasional rare argentaffin cell;
middle 40% occupied by numerous parietal cells with less abundant inter-
mixed mucous neck cells; upper 30% forming gastric pit; all cellular types
comparatively small as compared to those of other bats examined. Depth
of fundic mucosa 150 to 180.

Zymogenic cells typically pyramidal in form; nuclei spherical, small, and
usually positioned in middle or lower half of cells, occasionally truly basal;
basal portion of gland with chief cell complex rarely expanded into bulbous
structure. Argentaffin cells not observed.

Parietal cells small with small, slightly ovoid, centrally located nuclei;
slight variation in cell size apparent but variation mixed, not clinal; parietal
cells occurring extremely rarely among zymogenic cells, equally uncommon
within gastric pits of upper portion of tubule, little or no variation in distri-
bution throughout fundic zone, relatively abundant in view of only mod-
erate spatial distribution resulting from small size.

Mucous neck cells among parietal cells compressed horizontally with
large similarly-shaped nuclei; cells extremely small, infrequent within pa-
rietal cell zone, never locally abundant within zone. Gastric pits moderate
in depth, equal in breadth to remainder of gland, hence narrow, and with
little or no apparent relative increase in depth within lateral arc of fundic
caecum; pit cells markedly compressed laterally, large with extremely large
disc-shaped nuclei heavily compressed laterally; cells increasing progressively
in diameter from neck of pit to surface epithelium. Surface epithelium
relatively shallow; nuclei large, rounded, and tapered toward base; mucous
border narrow. Lamina propria mucosae typically scant basally and between
glands.

Pylo-fundic transition. — Extensive breadth of transitional area attributable
not only to pronounced invasion laterally of parietal cells into pylorus, but
also to relatively abrupt diminution and loss of chief cells considerably
proximal to pyloric sphincter (zone appearing expanded bidirectionally) ;
glands essentially like adjacent fundic mucosa, differing in absence of chief
cells and in that parietal cells distinctively more abundant along with
numerous mucous neck cells in basal arc; more variation in depth of glands
and slight increase in depth over that of fundus. Gradual increase in depth
ot j^istric pits beginning just proximal to sulcus intermedius, zone of in-
crease occupying approximately terminal third of transitional area, progres-

Studies of Stomachs of Selected American Bats

1 «* *-»

613

*> J**d$

0**$*

1mm

6

»•-» . f

- 2 . v.

7

Fig. 6. Cardiac sphincter at gastroesophageal junction of Noctilio labialis. The fundic
caecum is toward the left.

Fig. 7. Pyloric glands of Noctilio labialis.

614 The University Science Bulletin

sive increase corresponding to gradual loss of parietal cells within upper
tubule; lamina propria equivalent to that of fundus. Depth 160 to 250.

Pyloric glands. — Terminal mucous-type glands (completely lack parietal
cells) restricted to exceptionally narrow, symmetrical zone extending distally
slightly beyond lip of sphincter into duodenal lumen; glands moderate and
slightly variable in depth, slightly broader than fundic glands, remarkably
uniform in cellular structure (usual single cell type present) ; cells consistent
in morphology and diameter throughout tubule (Fig. 7), as well as pyloric
zone, narrow with relatively large, circular, flattened nuclei (twice diameter
of parietal nuclei) ; cells elongate and with moderately extensive mucous
border; glands uniform in breadth excepting slight basal dilation, only
slightly broader than fundic or transition glands. Lamina propria con-
siderably more abundant basally, slightly more abundant between glands
than within oxyntic type mucosa. Surface epithelial cells elongate and
narrow compared to those of fundic glands. Depth of pyloric mucosa
160 to 200.

Glands of Brunner. — Restricted to single mass in immediate region of
pyloric sphincter, relatively abundant; occupying entire distal surface of
pyloric sphincter and proximal 1.0-1.5 mm of duodenum; thick layer of
uniform depth, tapering little toward termination of distribution, ending
aborally at lip of sphincter, not observed within pylorus of stomach; highly
convoluted tubules of moderate and uniform breadth; cells long and narrow
with extremely small, spherical, basally located nuclei (Fig. 8). Interposed
smooth muscle fibers between tubules infrequent compared to other bats
examined, except Pteronotus parnellii; lumen narrow throughout.

Remarks. — Noctilio labtalis is unique among bats examined in the occur-
rence of an apparent anatomical cardiac sphincter at the gastroesophageal
junction. Botha (1958a) examined the cardiac junctions of 13 species of
distantly related mammals, describing that of "the bat" as a large, unbroken,
conical papilla at the cardiac orifice, lined throughout with gastric mucosa,
the edges of which form an extremely watertight "cork." This description
is in marked contrast to the situation at the junction in N. labialis, although
aptly describes cardiac junctions of most other species included in the present
study. The squamous-lined flaps found in N. labialis also were observed
macroscopically in one specimen of Noctilio leporinus, and must be con-
sidered unique to the family Noctilionidae among bats thus far described;
additional investigations will be required to demonstrate that this structure
is a physiologically efficient closure mechanism. The specialization at the
cardiac junction of N. labialis may represent a distinct and independently
evolved structure, unique among bats to the Noctilionidae and perhaps
unique within the insectivore-primate-chiropteran line of mammalian evo-
lution.

Studies of Stomachs of Selected American Bats 615

U. i-ki, I

■v W ff-%*. \

Ik. ^

> ^W-fc*t> Os. - 1

Fig. 8. Glands of Brunner of Noctilio labialis. Note the small nuclei.

Several unique features of the stomach of N. labialis are shared only
with Pteronotus pamellii. The general outline of the stomach and great
breadth of the pylorus at the sphincter are similar in these two species, as
is the extensive pylo-fundic transition zone, with subsequent limiting of
the pyloric glands to an extremely narrow region at the sphincter. Addi-
tional features common to N. labialis and P. pamellii include narrow
oxynctic glands with comparatively small cellular elements, limited distri-
bution of zymogenic cells within fundic tubules, and lack of distinctive
reduction of the fundic mucosa within the caecum.

The stomach of N. labialis clearly follows the basic morphological pat-
tern exhibited by other carnivorous bats examined, with notable specializa-
tion of the cardiac junction, perhaps a mechanism to aid in prevention of
reversed passage of relatively large amounts of food into the abdominal
esophagus.

Family PHYLLOSTOMATIDAE

Subfamily Chilonycterinae
Pteronotus pamellii (Gray, 1843)

Specimens examined (1). — One specimen from Nicaragua of Pteronotus
pamellii juscus (KU 116992).

616

The University Science Bulletin

2mm

Fig. 9. Mid-longitudinal representation of the stomach of Pteronotus parnellii. For ex-
planation of symbols, see Figure 3.

Overall dimensions. — Greatest length S.6; greatest breadth 4.6; gastro-
esophageal junction to pyloric valve 2.4; gastroesophageal junction to apex
of fundic caecum 4.0; breadth pylorus at sphincter 1.8.

General morphology. — Stomach tubular, short relative to breadth, highly
symmetrical with gastroesophageal junction located approximately midway
along lesser curvature (Fig. 9) ; upper and lower curvatures nearly parallel,
but dorsolateral curvature pronounced (dorsal surface convex); fundic
caecum short, directed slightly anteriorly with no evidence of dilation within
dorsal surface, apex narrowly rounded; terminal portion of stomach ex-
tremely short and broad. Gastroesophageal junction and cardiac vestibule
oriented nearly directly craniad, junction and cardiac orifice extremely
broad; inferior border of stratified squamous epithelium only half depth
of adjacent cardiac glands, gastroesophageal junction thus similar morpho-
logically to that of Natalus; muscularis externa and muscularis mucosae
continuous between esophagus and cardiac vestibule with no breaks or
unusual orientation of fibers and no evidence of an anatomical cardiac
sphincter; cardiac vestibule moderate in length, pattern of gradual coalition
with lesser curvature similar to that in Plecotus (but vestibule shorter);
incisura cardiaca absent (but slight thickening of circular muscle layer
within wall that corresponds to zone of incisura cardiaca in other bats).

Studies of Stomachs of Selected American Bats

617

Fig. 10. Diagrammatic mid-longitudinal illustration of gastroduodenal junction of Pterono-
tits parnellii. BG, Brunner's glands; a, extension of Brunner's glands beyond lip of pyloric
sphincter; b, pronounced reduction of pyloric sphincter (lesser curvature); c, small, narrow
musculature of sphincter within greater curvature; anterior is toward the top in this figure.

Fundic caecum shallow; mucosa within caecum deep, muscular comple-
ment of caecal wall thick owing in large part to longitudinal muscle layer
rather than circular layer. Stomach distal to cardiac vestibule directed
antero-laterad, tapering gradually to extremely broad pyloric orifice. Muscu-
lature of pyloric sphincter poorly developed on greater curvature as an ex-
tremely narrow mass of circular muscle and muscularis mucosae extends
outward and upward toward duodenal lumen (mass less in length than
depth of glands of Brunner resting on its distal surface) (Fig. 10); sphincter
of lesser curvature reduced to simple bulge in circular muscle layer; pyloric
mucosa lining sphincter equivalent in depth to that of remainder of pylorus;
pyloric orifice broad and displaced to left within tube; pyloric orifice inferior
to esophageal junction in situ, although not pronounced because stomach
nearly horizontal within body cavity. Lamina propria mucosae never
abundant within stomach wall proper, moderately represented within rugae;

618 The University Science Bulletin

gastric mucosa comparatively deep among bats examined, occupying entire
inner surface and never locally reduced (except extremely slight reduction
within furrows of pylorus).

Musculature. — Muscular wall comparatively thick among bats examined
and best developed among phyllostomatids studied; cardiac vestibule thick
walled, musculature as deep as that of major portion of lesser curvature
with circular layer principle component and longitudinal layer simple
thin sheet. Wall of lesser curvature shallower than corresponding area of
greater curvature only within terminal stomach, remaining portions (upper
and lower walls of fundic caecum) approximately equal in depth, being
slightly thicker than terminal half of lesser curvature; wall of greater curva-
ture increasing progressively in depth from midpoint to pyloric sphincter,
becoming two to three times breadth of comparable area within lesser arc
through terminal 2 mm. Gathering of circular fibers into bundles prominent
in lower wall of caecum, bundles becoming fewer in number and larger
through upper wall to base of cardiac vestibule; pattern of depth of longi-
tudinal layer similar to that observed in Phyllostomus, being exceptionally
deep through lower and apical wall of caecum, decreasing through lower
wall of terminal stomach, and extremely thin throughout lesser curvature;
muscularis mucosae thin double sheet throughout wall, slightly deeper
within rugae. Ganglia of Auerbacffs plexus numerous except within cardiac
vestibule and between vestibule and pyloric sphincter along lesser arc;
lamina propria uniformly scant between muscularis mucosae and tunica
muscularis throughout entire stomach wall, slightly more abundant beneath
muscularis mucosae within rugae. Depth longitudinal layer 15 to 45, circular
layer 30 (fundic caecum) to ISO (pylorus), muscularis mucosae less than
10 to 15 (within rugae).

Organization of gastric mucosa. — Rugal folds fewest in absolute number
as well as least frequent per given area of gastric mucosa among bats
examined; strictly directed longitudinally throughout entirety of dorsal and
ventral surfaces of body of stomach from apex of fundic caecum to pyloric
sphincter; isolated, transversely-oriented folds not observed anywhere within
stomach lining, and folds not locally reducted in frequency; folds of cardiac
vestibule continuous upward to gastroesophageal junction, although con-
forming to general longitudinal orientation (also the case in Plecotus,
Natalus, Noctilio, and Rhynchonycteris). Depth of mucosa relatively uni-
form throughout except for cardiac glands; fundic mucosa slightly deeper
than pyloric mucosa and slightly deeper in lesser curvature than in greater
curvature, little reduced in fundic caecum; cardiac glands exceptionally
deep in contrast to sharp decrease in depth of adjacent fundic mucosa;
cardiac zone similar in depth, breadth, and morphology to that of Phyllo-
stomus. Combined fundic glands and transitional zone occupying nearly

Studies of Stomachs of Selected American Bats 619

all of surface area as pyloric zone extremely short; fundic glands uniform
in structure within zone and with no secondary regions of noteworthy
variation from basic plan; transitional zone to pyloric mucosa broad in
comparison to that of other phyllostomatids and occupies entire lesser curva-
ture between gastroesophageal junction and pyloric mucosa (also extensive
along greater arc) ; area of true pyloric mucosa relatively slight and asym-
metrically distributed on upper and lower surfaces owing to extensive
transitional zone and asymmetry of pyloric sphincter.

Cardiac glands. — Zone relatively narrow, equal in length throughout
distribution; glands similar in morphology to those of Phyllostomus (long
and narrow) although cellular components not identical; glands nearly
uniform in breadth throughout, not prominently expanded at base; lumen
extremely broad and uniform in breadth throughout; cells in lower third
of tubule larger than those in upper portion, pyramidal, with large elongate
nuclei compressed against basement membrane; cells in upper two-thirds of
tubule extremely long and shallow with elongate, prominently compressed
nuclei; mucous layer of lumen border shallow (in contrast to broad border
of comparable cells in Phyllostomus). Lamina propria mucosae between
glands better developed than within remainder of mucosa. Transitional area
from cardiac to fundic mucosa twice breadth of cardiac zone; gastric pits
in transitional zone decrease abruptly in depth, glands identical to cardiac
mucosa except for presence of substantial number of scattered parietal cells;
surface epithelium equivalent to that of remaining stomach (excepting
slightly smaller nuclei). Depth of cardiac glands 350 to 390.

Fundus. — Fundic glands consistently narrow and highly convoluted,
relatively long; fundic mucosa generally best developed in midportion of
distribution, decreasing only slightly but progressively in depth in either
direction laterally; all cellular types comparatively small in relation to other
bats examined (except Noctilio labialis) by virtue of apparent uniform
reduction in cytoplasmic mass of cells (because nuclei are universally large
regardless of cell type); only slight variation in gland morphology within
zone, variation entirely in number of zymogenic cells and depths of gastric
pits. Cellular components more or less restricted to specific zones with slight
alteration in lengths of zones within apical arc of fundic caecum; lower
15% restricted to zymogenic cells, middle 65% to parietal and mucous
neck cells and upper 20% to gastric pits. Depth of mucosa 150 to 270
(cardiac vestibule).

Zymogenic cells never occupying more than basal 30% of gland; uni-
form in size, small and globular, not numerous, nuclei extremely large
relative to cell size and always spherical with nuclei usually displaced
laterally; zymogenic cells most abundant within glands of cardiac vestibule
and through area directly posterior to greater curvature; argentaffin cells

620 The University Science Bulletin

extremely rare in bases of glands and less than one-half size of chief cells,
rounded medially, tapering gradually toward periphery (outer limit of
tubule), nucleus small and spherical.

Parietal cells, although within extensive spacial distribution, only moder-
ately abundant owing to large numbers of interposed mucous neck cells;
parietal cells normally elongate, ovoid, with extremely large, centrally lo-
cated nuclei; clear cytoplasmic intercellular canals especially numerous;
pronounced increase in cell size with corresponding decrease in nuclear
size from bottom to top of distribution, uppermost cells extremely large;
parietal cells never observed among basal zymogenic cells and reduced
slightly in frequency within apex of fundic caecum as gastric pits are
comparatively deeper.

Mucous neck cells uncommonly abundant among parietal cells as ob-
served in Phyllostomus, increasing in numbers within gland from origin
at chief-parietal cell boundary to surface epithelium and often observed
in large clumps of 5 to 10 cells; mucous neck cells small, irregular in shape
with large ovoid nuclei usually larger than those of adjacent parietal cells,
nuclei remain large through gastric pits. Gastric pits shallow throughout
most of fundic zone, increasing from usual 15% of thickness of tubule to
20-25% within furrows along apical arc of fundic caecum; neck cells of
gastric pits with typically large nuclei previously described, nuclei decreasing
in size only slightly to surface epithelium; mucous border deep throughout,
lumen both shallow and narrow. Surface epithelium moderate in depth;
nuclei extremely large, ovoid, not appreciably compressed laterally; lamina
propria typically sparse basally, poorly developed between rather closely
packed fundic glands.

Pylo-fimdic transition zone. — Distribution of transitional glands to pyloric
mucosa so extensive that true pyloric glands absent in portions of terminal
pyloric tube; along lesser curvature transitional mucosa occupies all of area
between inferior border of cardiac mucosa and initial increase in depth
of musculature near sphincter (see Fig. 9), resulting in extremely broad
zone; within greater curvature transitional glands extend through approxi-
mately one-third total length of stomach, essentially all of pyloric tube;
glands identical in morphology along both curvatures up to terminal margin
with pyloric glands where junction abrupt on lesser curvature (non-existent
on greater curvature as true pyloric glands absent); glands essentially identi-
cal in structure to fundic glands (excepting absence of chief cells) with
parietal and mucous cells extending to base; deepening of gastric pits abrupt
at terminal margin of zone on lesser arc and gradual through greater curva-
ture continuing to duodenal junction (parietal cells persisting within basal
half of gland to duodenal junction); loss of zymogenic cells abrupt at initial
margin; surface epithelium identical to that of fundus; depth equivalent
to that of fundus; depth 210 to 250.

Studies of Stomachs of Selected American Bats 621

Pyloric glands. — Extremely limited distribution on lesser curvature;
glands with parietal cells absent in greater arc and 50% broader than fundic
or transitional glands; basal third of tubules broadly expanded with large
cells containing ovoid nuclei moderate in size; cells of upper two-thirds of
tubule small and more closely resembling mucous neck cells previously
discussed, and with large, spherical nuclei; glands increasing gradually
in breadth in upper half of tubule; basal two-thirds of lumen narrow,
slightly broader in upper third; terminal glands along greater curvature
generally fitting description given above although much narrower and con-
taining mixed mucous and parietal cells (considered transitional glands in
this account); intertubular and basal lamina propria as in fundic and
transitional zones; depth ISO to 220.

Glands of Brunner. — Comparatively, best developed among phyllostom-
atids examined; always submucosal and occur throughout proximal 2.5
mm of duodenum; in single mass with no isolated lateral clumps; deepest
along distal surface of pyloric valve, gradually decreasing in depth distally;
not observed in stomach; typically highly convoluted and branched with
tubules moderate in breadth; differing from other bats examined (except-
ing Noctilio) in having extremely small nuclei (always spherical) juxta-
posed to basement membrane; cells numerous within cross-section of tubule,
elongate and narrow; glands reduced slightly in breadth toward surface
in comparison to basal region; pattern of exit into duodenal lumen as in
other species.

Remarks. — Simplification of the stomach to a short, symmetrical, tubular
crescent is characteristic of many obligate carnivorous bats as it is of Ptero-
notus parnellii. The presence of features observed in insectivorous bats in
general is more apparent than in Phyllostomus discolor, another insect-
feeding phyllostomatid. The tunica muscularis is best developed relative
to stomach size, and the fundic caecum least distinctive, among the seven
phyllostomatids examined. The terminal portion of the stomach is short
and broad, totally unlike any other member of the family studied, although
at least one side of the sphincter is long and narrow as in several other
leaf -nosed bats.

General outline, topography, and orientation of orifices are somewhat
similar to that observed in Noctilio labialis and Plecotus townsendii. The
most unusual feature is the near loss of true pyloric mucosa at the terminal
end, with extension of transitional glands laterally. This feature is in contrast
to the relatively narrow transitional zone in most phyllostomatids. Orienta-
tion of mucosal folds is strictly longitudinal, a feature also unique among
phyllostomatids studied.

Cellular elements of glandular types throughout the gastric mucos
lend distinctiveness to the stomach by consistently possessing uncommonly

622 The University Science Bulletin

large nuclei, in relation to the relatively slight cytoplasmic content of the
cells.

Subfamily Phyllostomatinae

Phyllostomus discolor Wagner, 1843

Specimens examined (3). — Three specimens from Nicaragua of Phyllo-
stomus discolor verrucosus (KU 105952-53, 10595S).

Overall dimensions. — Greatest length 12.1-12.4; greatest breadth 5.0-5.2;
gastroesophageal junction to pyloric valve 3.5-3.6; gastroesophageal junction
to apex of fundic caecum 4.1-4.4; breadth of pylorus at sphincter 1.9-2.0.

General morphology. — Stomach reniform, long and generally symmetri-
cal; greater and lesser curvatures nearly parallel; marked curvature in
frontal plane with less marked dorsolateral curvature (dorsal surface con-
vex) from gastroesophageal junction to apex of fundic caecum; fundic
caecum long and slightly dilated at apex on dorsal surface; terminal stomach
long but pyloric endpiece abruptly tapered and relatively short. Cardiac
vestibule extremely short and tubular, junction with lesser curvature abrupt
(extends only about 1 mm superior to lesser curvature) ; uniform in diam-
eter, not tapered, breadth equivalent to that of esophagus; enters body of
stomach perpendicularly beneath sharply-angled esophagus; thick-walled
as most of remaining stomach in that musculature well developed. Gastro-
esophageal junction abrupt; circular and longitudinal layers continuous at
junction, muscularis mucosae permeated by strands of collagenous tissue
although most muscle fibers continuous. Incisura cardiaca present, although
extremely shallow and arises as a result of anterior expansion of the fundic
caecum, with slight depression resulting at junction of lesser curvature and
cardiac vestibule; circular muscle layer markedly thickened at incisura
with compartmentalization of fibers into radiating sheets but no conclusive
evidence of anatomical cardiac sphincter. Fundic caecum relatively long
and tubular, directed antero-laterally; dilated at apex and hooking slightly
dorsally; apex narrowly rounded; caecum not appearing as a distinctive
chamber but rather as a tubular extension as observed in Plecotus and
Molossus. Tubular stomach distal to gastroesophageal junction extensive,
although broader and less extensively recurved than other phyllostomatids
examined (excepting Pteronotus) ; decreases gradually in breadth to pro-
nounced constriction (sulcus angularis) approximately 2.5 mm below pyloric
sphincter (Fig. 11); terminal endpiece to sphincter narrow; gastroesopha-
geal junction superior to pyloric sphincter in stomach in situ. Pyloric
sphincter greater in length and breadth on greater curvature and expanded
along lumenal rim to form narrow lip at orifice; circular layer and muscu-
laris muscosae form contributing muscle layers; external constriction at
sphincter slight. Musculature thick with all muscle components well repre-

Studies of Stomachs of Selected American Bats

623

2mm

Fig. 11. Mid-longitudinal representation of the stomach of Phyllostomns discolor. For ex-
planation of symbols, see Figure 3.

sented (including muscularis mucosae) and little reduction anywhere
except in terminal arc of fundic caecum.

Rugal folds relatively few in number; gastric mucosa thick throughout,
reduced in depth only within furrows of fundic caecum, mucosa deepest
within fundus of lesser curvature in pylorus. Lamina propria best devel-
oped of all bats examined, particularly within pyloric region and immedi-
ately adjacent fundic mucosa between muscularis mucosae and circular
layer, although typically narrow sheets within wall of fundic caecum.
Stomach generally similar in gross morphology to those of other carnivorous
species examined.

Musculature. — Both external layers exceptionally deep throughout (ex-
cepting major portion of fundic caecum); circular layer substantially deeper
than longitudinal except in terminal two-thirds of fundic caecum; muscu-
lature between gastroesophageal junction and lesser curvature moderate in
depth in comparison to most of remaining stomach, notably deeper than
musculature of adjacent esophagus and somewhat thickened at junction
with lesser curvature.

524 The University Science Bulletin

Longitudinal layer deep in comparison to other phyllostomatids ex-
amined; moderate in depth through greater arc of pyloric tube, becoming
deeper through greater curvature, but decreasing slightly within apex of
caecum; reduced to extremely thin sheet in upper surface of caecum and
moderate in depth throughout remainder of lesser curvature.

Circular layer deep throughout stomach wall (except for apex and upper
wall of caecum where equivalent in depth to longitudinal layer) ; generally
deeper throughout greater than lesser curvature; especially pronounced
local thickening midway along greater curvature of tubular stomach, slight
local thickening at bases of rugae; circular layer extensively gathered into
large bundles surrounded by speta and permeated centrally with moderate
amounts of loose lamina propria originating from normally thin interlayer
connective tissue sheet; distribution of bundles from midpoint of greater
curvature to most lateral extremity of fundic caecum.

Muscularis mucosae uncommonly thick throughout and double-layered;
substantially thickened in rugal folds as compared to stomach wall proper;
continuous with muscularis mucosae of esophagus at junction in absence of
anatomical cardiac sphincter. Connective tissue between circular muscle
and muscularis mucosae best developed as a continuous layer among bats
examined; especially pronounced within tubular stomach, but occurring
as relatively deep region throughout stomach (excluding apex of fundic
caecum), connective tissue especially well represented within rugal folds;
collagenous fibers most abundant within rugae, unusually abundant within
muscularis mucosae but less well represented in outer wall; ganglia of
Auerbach's plexus abundant. Depth longitudinal layer 15 to 45, circular
layer 30 (caecum) to 220 (pylorus), muscularis mucosae 15 to 25.

Organization of gastric mucosa. — Mucosa gathered longitudinally into
deep, broad rugae throughout all of stomach except along midline of greater
curvature within non-tubular area, in which rugae are short, discontinuous
projections that are directed antero-posteriorly (Fig. 12) ; slight "zig-zag"
orientation of folds within entire non-pyloric stomach, although uniformly
parallel and never truly interdigitating. Folds relatively few in number
and reduced locally in frequency within apical arc of fundic caecum; folds
well developed within short cardiac vestibule, although relatively few in
number owing to narrowness of tube. Gastric mucosa comparatively deep
throughout stomach; cardiac glands unusually long; fundic mucosa deepest
within cardiac vestibule and relatively deep throughout entire distribution,
also being thick along sides of rugae but slightly shallower at apexes of
folds and consistently reduced in furrows between folds; average depth
of fundic glands at apex of caecum slightly less than in remaining fundic
zone but deeper within lesser curvature compared to greater curvature;
pyloric mucosa slightly deeper than fundic and somewhat more variable

<w'

Studies of Stomachs of Selected American Bats

625

K

■

| 0.5 mm ,

I

*V

m

t 4r

s*#

ou

* i.»

fft

14

Fig. 12. A small, isolated, superiorly directed fold of gastric mucosa within the greater
curvature of Phyllostomtts discolor.

Fig. 13. Cardiac glands (CG) of Phyllostomiis discolor and gastroesophageal junction,
showing stratified squamous epithelium (SSE).

Fig. 14. Zymogenic (chief) cells (ZC) at base of a fundic gland in Phyllostomiis discolor.

Fig. 15. Gastric pit of fundic gland (Phyllostomiis discolor). Note the broad lumen and
deep mucous border.

626 The University Science Bulletin

in depth. All basic glandular types present; cardiac zone extremely narrow
(only 0.1 mm at gastroesophageal junction); true fundic mucosa occupies
all of caecal stomach, much of tubular stomach on greater curvature, with
less extensive distribution along lesser curvature resulting in correspondingly
greater representation of pyloric glands along the lesser arc; transition zone
to pylorus narrow, occupying brief zone just proximal to pyloric constriction,
equal in length on upper and lower surfaces; pyloric mucosa extensive,
asymmetrically distributed, occurring throughout short, terminal section
down through pyloric constriction to abrupt junction with pylo-fundic
transition zone and cranially to distal surface of pyloric sphincter with no
invasion into duodenal lumen.

Cardiac glands. — Extremely thin and relatively straight tubular glands
(Fig. 13) expanded only slightly at base; gradually increasing in breadth
within upper half; occupying zone only two to three glands in breadth at
gastroesophageal junction; single cell type present; cells small, long and
narrow, with small, laterally compressed nuclei in upper half (Fig. 13) and
more rounded nuclei in lower half; mucous border narrow; lumen also
narrow; transitional area to fundus as narrow as cardiac zone and the
equivalent of three glands in breadth; surface epithelium deep, cells with
large nuclei that are markedly compressed; depth of mucosa 420 to 460.

Fundus. — Glands extensively variable within stomach, both in depth and
in representation of particular cell types. Zymogenic cells in heaviest con-
centration within midregion of fundic mucosa, decreasing slightly in num-
ber in either direction laterally; parietal cells also most abundant within
midregion, decreasing substantially to apex of caecum and increasing in
direction of pylorus as zymogenic cells decrease; decrease in parietal cells
in caecum owing to substantial increase in depth of gastric pits throughout
entire fundic mucosa within fundic caecum, although overall decrease of
glandular depth slight. Gastric glands relatively short in relation to large
size of stomach, broad (excepting those of cardiac vestibule and lesser
curvature immediately surrounding cardiac vestibule). Cell types generally
restricted to specific zones within gland throughout most of fundic mucosa
(except mucous neck cells, which occur throughout where zymogenic cells
absent) ; lower 40% of fundus composed of zymogenic cells with occasional
scattered parietal cells, middle 30% with parietal cells and frequent mucous
neck cells, upper 30% represents gastric pits. Depth of mucosa 250 through-
out most of fundus to 150 within caecum.

Zymogenic cells (Fig. 14) abundant throughout fundic zone; large,
irregularly shaped, and of moderate size; nuclei spherical and always laterally
displaced; cells sometimes polyhedral or ovoid but more regularly triangular;
lumen narrow through zone of chief cells; base of gland with chief cells,
although extremely narrow, slightly expanded into bulb. Argentaffin cells

Studies of Stomachs of Selected American Bats 627

most abundant within complex of chief cells of any bat examined; extremely
small in comparison to adjacent chief cells; more or less rectangular with
large spherical nuclei; outer cytoplasm filled with fine argentaffin granules
that stain pink with standard haematoxylin and eosin procedure; cells con-
fined to lower apical arc of chief cell distribution.

Parietal cells moderate and comparatively uniform in size in relation
to those of other phyllostomatids examined; most abundant within narrow
midregion, less abundant as scattered elements among uppermost chief
cells; usually rounded, rarely ovoid; nucleus always ovoid and moderate in
size, centrally located within cell; parietal cells may be relatively few due
to large number of mucous neck cells within midregion of gland.

Mucous neck cells small, having large, round, peripheral nuclei; mucous
border deep, much as in Pteronotus; unique among phyllostomatids ex-
amined in being more abundant in area of parietal cells than parietal cells
themselves. Gastric pits extremely deep, deepest of any phyllostomatid
observed, occupying upper 30% through most of fundus to 80% in glands
at apex of fundic caecum where parietal cells all but absent. Neck cells
extremely large; mucous border deepest of any bat examined (Fig. 15);
nuclei large and ovoid in lower half of layer, prominently compressed in
upper half; lumen a broad chamber increasing only slightly in breadth to
surface. Surface epithelium deep; cells with extremely large rounded nuclei.
Lamina propria typically sparse beneath glands.

Fitndo-pyloric transition. — Zone narrow (see Fig. 11), approximately
equal in length on both curvatures; loss of parietal cells gradual with few
cells persisting into basal area of proximal-most pyloric glands; persisting
parietal cells extremely small; loss of zymogenic cells much more abrupt;
deepening of gastric pits gradual with progressive loss of parietal and chief
cells; area of transition in which parietal cells extend to base of gland
extremely narrow; depth 250 to 310.

Pyloric glands. — Pyloric glands extensive and reflect generally well-
developed mucous cell system in Phyllostomus; glands variable in depth,
local thickenings represent deepest portions of gastric mucosa with glands
prominently deeper through greater in contrast to lesser curvature; glands
long and extremely broad (Fig. 16); composed of single cell type (cells
small with large nuclei in basal half of gland), increasing substantially
within upper half and becoming giant columnar type with relatively small,
progressively more ovoid nuclei; lumen extremely broad, hence glands
uncommonly large; origin of pyloric glands at terminus of fundic mucosa
fairly abrupt, extending laterally to distal edge of pyloric sphincter; lamina
propria beneath glands reduced as in remainder of mucosa; pyloric zone
especially well developed in length and depth in Phyllostomus as compared
to other phyllostomatids examined; depth of mucosa 200 to 310.

628

The University Science Bulletin

Fig. 16. Pyloric mucosa of Phyllostomits discolor.

Glands of Brunner. — Glands well developed at pyloric sphincter and as
moderately deep layer through proximal 3 mm of duodenum; always sub-
mucosal, never occur within stomach mucosa; distribution along distal
surface of sphincter only slightly deeper than that within submucosa of
duodenal wall (former region markedly deepened layer in all other bats
examined having !arge complement of Brunner's glands) ; glands occur
as single mass in duodenum with isolated patches not observed; glands
virtually identical in cross-sectional morphology to those of Pteronotus;
orifices between crypts of Lieberkiihn; lumen relatively broad.

Remarks. — The stomach of Phyllostomns discolor exhibits a mosaic of
characters of which some are common to other phyllostomatids and others
are common to carnivorous species irrespective of systematic affiliation. The
tubular nature of the stomach is consistent with findings for obligate insect
feeders, and suggests at least a gross morphological adaptation to a diet rich
in proteinaceous foods. Recent studies (Arata et at., 1967) have alluded to
the strongly insectivorous feeding habits of Phyllostomus discolor and P.
hastatus, species belonging to a family predominantly composed of fruit and
nectar feeders.

Other features observed in Phyllostomus include a more or less tubular
undic caecum with only a slight dorsolateral dilation at the apex, short

Studies of Stomachs of Selected American Bats 629

tubular cardiac vestibule, well-developed musculature, and lack of pro-
nounced cranial recurvature of the pylorus.

Although clearly showing features normally observed in carnivores,
P. discolor retains conditions common to most other phyllostomatids. These
include a narrow relatively abrupt pylo-fundic transition zone, dilated fundic
caecum (although only slightly on dorsal wall), an elongate terminal
stomach, and a relatively thin muscular complement within the pyloric
sphincter.

Mucous production must be extensive owing to the marked development
of pyloric glands and notably deepened gastric pits of the fundic glands,
as well as numerous mucous neck cells within the middle portion of the
fundic glands.

Subfamily Glossophaginae
Glossophaga soricina (Pallas, 1766)

Specimens examined (2). — Two specimens from Nicaragua of Glos-
sophaga soricina leachii (KU 105984, 105993).

Overall dimensions. — Greatest length 12.0-13.5; greatest breadth 4.7-4.9;
greatest length fundic pouch 4.3-4.5; greatest breadth fundic pouch 3.7;
gastroesophageal junction to pyloric sphincter 4.8-5.3; gastroesophageal junc-
tion to apex of fundic caecum 5.0-5.2; breadth pylorus at sphincter 1.7.

General structure. — Large, saccular, bean-shaped, with extensive develop-
ment of fundic caecum; both saccular and terminal parts well developed
(Fig. 17) ; marked curvature of stomach in two planes — pronounced curva-
ture in frontal plane and less pronounced curvature in transverse plane
(planar orientation of organ in situ). Fundic caecum extensively developed
(Fig. 18) and set off by sulcus on dorsal surface; pronounced invagination
of muscularis externa on greater and lesser curvatures resulting in distinc-
tive "pouch" and semi-ruminate condition; caecum dilated on dorsal side,
hence caecum spacious; apex blunt; sulcus delimiting fundic "pouch" at
incisura cardiaca on lesser curvature extending posteriad to midway along
greater curvature between apex of fundic caecum and pyloric bend (see
Fig. 17). Region of stomach distal to gastroesophageal junction tapering
gradually to midway between cardiac and pyloric junctions (where pro-
nounced crook in tubular stomach occurs, with terminal part recurved
craniad and terminal end-piece then approximately parallel to esophagus);
sulcus intermedius present but poorly developed, stomach narrows abruptly
just distal to "bend" and remains uniformly narrow to pyloric sphincter.

Incisura cardiaca pronounced and accompanied by noteworthy local
thickening of muscularis externa; thickened muscle mass dense and partially
compartmentalized into thin sheets radiating from incisura; sheets separated
by thin fibrous septa similar to those described by Dearden (1966) in micro-
tine rodents; cardiac orifice relatively narrow. Cardiac vestibule exti

630

The University Science Bulletin

2mm

Fig. 17. Mid-longitudinal representation of the stomach of Glossophaga sorkina. For ex-
planation of symbols, see Figure 3.

short and with no reduction of included gastric mucosa; longer on pyloric
than caecal side owing to angular entry of esophagus into stomach.

Pyloric sphincter asymmetrical (Fig. 19) ; musculature of pyloric sphinc-
ter substantially more massive on greater than on lesser curvature, muscle
mass generally thinnest at junction with stomach wall, increasing gradually
in breadth and terminating at rounded apex on luminal surface; muscle of
sphincter on greater curvature several times more massive than on lesser
curvature; relatively shallow; gastric mucosa (pyloric glands) covering
pyloric sphincter; glands of Brunner completely covering distal edge of
sphincter muscle mass; only circular and muscularis mucosae layers contrib-
ute to pyloric sphincter. Gastric mucosa lines entire surface of stomach,
little reduced anywhere; stomach semi-compartmentalized, possibly adapted
for storage of food material, presumably to extend length of time during
which food is exposed to gastric digestion .

Musculature. — Extensive variation in depth of component muscle layers
within stomach; outer longitudinal layer always considerably thinner than
inner circular layer; circular layer thickened at incisura cardiaca and some-
what compartmentalized into sheets of fibers by thin fibrous septa (re-
stricted to incisura cardiaca proper).

Outer longitudinal layer extensively dilated throughout lesser and greater

Studies of Stomachs of Selected American Bats

631

1mm

Fig. 18. Longitudinal section of the fundic caecum (within dorsal half) oi Glossophaga
soricina showing fold in body wall (FO) and fundic caecum (FC).

Fig. 19. Pyloric sphincter of Glossophaga soricina. Note the asymmetry of the sphinctc
musculature, showing Brunner's glands (BG) and pyloric sphincter (PS).

632 The University Science Bulletin

curvatures of fundic stomach distal to cardiac junction; remaining fundus
(fundic caecum) with relatively thickened longitudinal layer on superior
and inferior surfaces, whereas circular layer in these two areas only slightly
thicker than longitudinal muscle and clumped into many fibrous bundles
along inferior surface; both layers notably dilated in apex of caecum, oc-
curring there as extremely thin sheets. Pyloric tube musculature extensively
thickened on greater curvature (both longitudinal and circular layers) (Fig.
19), narrowing again to a thin remnant just inferior to pyloric valve; muscu-
lature of lesser curvature of pyloric tube less than one-third breadth of
corresponding tissue of greater surface at its maximum.

Muscularis mucosae narrow; filamentous sheets throughout stomach,
most prominent in pyloric tube. Depth of longitudinal layer 10 to 45,
circular layer 30 to 140, muscularis mucosae 7.

Organization of gastric mucosa. — Mucosa gathered longitudinally into
parallel rows throughout all of pyloric area and tubular fundus up to and
including cardiac vestibule; folds of saccular fundic caecum oriented longi-
tudinally on sides (dorsal and ventral surfaces in situ) and radially on
upper and lower surfaces, as well as apex of caecum and effacing one another
in spoke-like fashion within caecum; folds broad and deep in tubular
stomach and reduced somewhat in breadth (although not in frequency)
in caecum due to general reduction in depth of mucosa in this region;
folds in mid-longitudinal plane rarely broadly expanded at apex in contrast
to those of Plecotus; mucosa only slightly deeper at apex of rugae than at
bases and sides of furrows between rugae; fundic mucosa slightly deeper
on lesser than greater curvature, reduced in caecum and extreme oral por-
tion of cardiac vestibule; pyloric mucosa deep in proximal area of pyloric
zone, reduced substantially in depth at orifices and along pyloric sphincter.

All basic glandular areas represented; cardiac zone relatively broad
with transition to fundus abrupt; pyloric glands occupy all of pyloric end-
piece (that portion distal to point of narrowing) ; transition from fundus to
pylorus narrow on lesser curvature, twice as broad on greater curvature,
loss of basal zymogenic cells gradual with deepening of gastric pits and
establishment of mucous complement of pyloric glands relatively abrupt;
all of fundic caecum and non-pyloric stomach occupied by fundic type
glandular mucosa (see Fig. 17 for map of glandular areas).

Cardiac glands. — Band moderate in breadth (1 mm), extending approxi-
mately to base of muscular fold of incisura; glands relatively numerous in
comparison to most other bats examined; demonstrable transition to fundic
mucosa present, although transition relatively abrupt and characterized
by scattered parietal cells in basal half of gland; cardiac glands extremely
short and broad (Fig. 20), of fairly uniform height throughout zone and
with broad lumens; basal cells small, compact, upper cells larger and

Studies of Stomachs of Selected American Bats

633

?\

20

Fig. 20. Zone of cardiac glands at the gastroesophageal junction of Glossophaga soricina,
showing cardiac glands (CG), esophagus (ESO), and tunica muscularis (TM).

Fig. 21. Brunner's glands in the duodenum of Glossophaga soricina.

634 The University Science Bulletin

usually with large circular nuclei; gastric pits of cardiac glands broad, in-
creasing greatly in breadth toward mucosal surface where breadth so great
as to lend villi-like appearance to glands (Fig. 20) ; loose lamina propria
extensive between slightly bulbular bases of glands, less extensive with
onset of fundic mucosa; abrupt transition aborally from stratified squamous
epithelium of esophagus; depth of mucosa 125.

Fundus. — Variation in morphology within stomach due to reduction
in numbers or loss of certain cell types in specific regions; mucosa within
and adjacent to fundic caecum slightly shallower than that of remaining
fundus and gastric pits broader, "stunted" condition of glands in caecum
due to reduction in number of all three cell components (zymogenic, pa-
rietal, mucous neck) ; gastric pits deep in fundic caecum, narrow and shallow
in remainder of fundus; lamina propria restricted mostly to narrow belt
beneath bases of glands. Depth of fundic mucosa 140 to 180 in fundic
caecum and 170 to 220 in tubular portion of fundus.

Zymogenic cells usually occupy lower third of gland and occasionally
basal half in midregion of fundus near gastroesophageal junction, reduced
or absent only in extreme apex of fundic caecum; rectangular to polyhedral
with large, spherical, peripherally-displaced nuclei; granulation pronounced
in specimens examined; parietal cells rarely observed within zymogenic cell
complex.

Parietal cells normally in middle third of gland; extremely large, often
four times breadth of adjacent mucous neck cells and circular to slightly
ovoid; nuclei large (breadth equals one-third breadth of cell) and circular;
distinct and marked increase in size of cells from bottom to top of gland
(cells vary in diameter from 7 microns at base to 18 in upper gland);
basal cells irregular in shape due to more compact cellular elements than
within upper portions.

Mucous neck cells confined to upper two-thirds of tubule, continuing
into gastric pits (as in all other bats examined) from origin among parietal
cells; cuboidal or trapezoidal with large spherical nuclei (nearly twice as
large as those of parietal cells); extending to base of glands in fundic
caecum along with parietal cells (replace zymogenic cells). Argentaffin cells
extremely rare, found mostly among zymogenic cells in basal arc.

Gastric pits of fundic glands distal to caecum shallow, those within
caecum deeper, broader, 30-60% length of glands; surface epithelial cells
short and with elongate, laterally-compressed nuclei, decrease markedly in
size from luminal surface to bottom of gastric pit; mucous border of epi-
thelium not extensive.

Pylo-fundic transition. — Extremely narrow zone on lesser curvature,
more extensive on greater arc, occupies approximate region of pyloric bend;
depletion of chief cells gradual; broad area with absence of chief cells yet

Studies of Stomachs of Selected American Bats 635

with shallow gastric pits; gastric pits deepen abruptly in terminal portion
of zone; parietal cells numerous and extend into base of glands, relatively
small throughout glands; depth 200.

Pyloric glands. — Occupy tubular portion of stomach approximately distal
to pyloric bend and up to superior surface of pyloric sphincter; glands as
deep or slightly deeper than those of fundus, although notably broader and
with substantially thicker, more compact surface epithelium; composed of
single cell type (with exception of occasional basal argentaffin cell); basal
mucous cells extremely compact, base of gland bulbular; nuclei large, spheri-
cal; from midregion upward cells increase in size, nuclei enlarge and be-
come progressively more laterally flattened and "club-shaped"; lamina
propria more extensive than in fundic mucosa, areolar connective tissue
prominent beneath glands; gastric pits broad and lumen of gland extremely
broad with two glands opening into single pit; pyloric glands well devel-
oped, extensively distributed when compared to many other bats examined;
depth of pyloric glands 150 to 250.

Glands of Brunner. — Limited to proximal-most 2 mm of duodenum
and never found superior to muscularis mucosae except in limited area at
apex of pyloric valve; narrow tubules highly coiled with broad lumen;
cells irregularly shaped with large, spherical, peripheral nuclei (Fig. 21);
smooth muscle fibers and lamina propria observed only rarely between
glands.

Remarks. — The stomach of Glossophaga soricina is most distinctive in
the presence of a rounded, spacious fundic caecum prominently dilated on
the dorsal surface, and unique among bats examined with the presence of
a distinct sulcus deliminating the fundic pouch from the remaining stomach.
A figure depicting the gastrointestinal tract of Glossophaga included in
accounts of gross morphology of several bat stomachs by Park and Hall
(1951) illustrates the dilated portion of the caecum, and corresponds well
with my findings. The stomach cannot be considered truly ruminate, as no
rigid partition exists, although the stomach clearly approaches a two-chamber
condition (Fig. 17).

Rouk (1968) described and illustrated the stomach of Leptonycteris
sanborni, a North American nectivore of the subfamily Glossophaginae. The
stomach of this species agrees well (as illustrated by Rouk) with that of
G. soricina in general shape and configuration of gross features, particularly
in orientation and morphology of the terminal stomach, including the
pyloric sphincter. Also, the orientation and location of the gastroesophageal
junction as reported by Rouk is virtually identical to that of Glossophaga,
although the vestibule is somewhat more extensive in L. sanborni. Rouk
states that the stomach of L. sanborni is "more globular or saccular'
those of the insectivorous types he examined, although an included :

636 The University Science Bulletin

longitudinal drawing does not conclusively illustrate this point. Rouk also
noted that the pyloric sphincter was thinner than that of insectivorous types
examined, a feature observed in Glossophaga and several other phyllo-
stomatids I have examined. Rouk's map of the distribution of glandular
types shows the glands of Brunner to be extremely restricted in distribution,
unlike the rather large mass at the pyloric sphincter of Glossophaga. The
relative depths of the tunica muscularis are similar in the two species,
although transverse gathering of circular fibers into bundles extends craniad
more in the fundic caecum of Glossophaga.

The gastric mucosa of G. soricina is unique among herbivorous phyllo-
stomatids examined in being rather generalized, composed of relatively
short, broad glands somewhat resembling those of Plecotus; the three
cellular components each are more or less confined to one-third the length
of the tubule and overlap little in distribution.

Rouk's description of the mucosal folds of Leptonycteris sanborni sug-
gests existence of some degree of anastomosing, as observed in the fundic
caecum of Glossophaga, although the description is unclear as to the exact
location of anastomosing folds, and difficult to interpret comparatively. He
also reported the presence of what appeared to be clumps of cells identical
to glands of Brunner within the pylorus. No glands positively identified
as Brunner's (histologically or histochemically) were observed in the stom-
ach of Glossophaga.

Subfamily Carollinae
Carollia perspicillata (Linneus, 1758)

Specimens examined (2). — Two specimens from Nicaragua of Carollia
perspicillata azteca (KU 106036, 106064).

Overall dimensions.— Greatest length 10.1-11.4; greatest breadth 5.0-5.5;
gastroesophageal junction to apex of fundic caecum 4.0-4.7; gastroesophageal
junction to pyloric sphincter 4.1-6.2; breadth of fundic caecum 4.0-4.1;
breadth pylorus at sphincter 3.9.

General structure. — Stomach club-shaped and decidedly asymmetrical
with moderately well-developed cardiac vestibule (Fig. 22) ; tubular stomach
markedly elongate; marked dorsolateral flexure affecting entire stomach,
dorsal surface concave; stomach most closely resembling that of Glossophaga
among bats examined in general stomach outline; gastroesophageal junction
abrupt, muscularis externa continuous with muscularis of esophagus, muscu-
laris mucosae thickened slightly although in continuous sheets; no ana-
tomical cardiac sphincter. Cardiac vestibule moderately developed (Fig.
23), entering stomach at pronounced angle to lesser curvature creating
marked incisura cardiaca; expanding gradually from origin at gastroesopha-
geal junction and disappearing gradually within lesser curvature; extremely
thin-walled compared to remaining areas of stomach although mucosal

Studies of Stomachs of Selected American Bats

637

2mm

Fig. 22. Mid-longitudinal representation of the stomach of Carollia perspicillata. For ex-
planation of symbols, see Figure 3.

membrane deep within vestibule; incisura cardiaca relatively broad and
deep (with thickening of circular muscle layer only, no comparable condi-
tion in other musculature); incisura best expressed within dorsal half of
stomach as dilated dorsal wall of fundic caecum tends to expand somewhat
craniad.

Fundic caecum saccular, appearing as dilated circular bulb lateral to
vestibule (Fig. 24), heavily dilated and rounded on dorsal (back) surface,
comparatively small and less distinctive as "chamber" than caecum of
Glossophaga; caecum distinctive not only externally but by pattern of mu-
cosal gathering within, never truly separated anatomically by folds of
musculature or gastric mucosa; musculature of caecum not reduced ap-
preciably in depth but muscle layers loosely organized with unusually heavy
complements of elastic fibers within muscularis; complex, oblique orienta-
tion of longitudinal layer throughout caecum, extensive gathering of circu-
lar fibers into large numbers of small narrow sheets.

Sulcus intermedius present, broad and shallow; sulcus angularis rela-
tively deep, occurring just proximal to pyloric bend; tubular region of
stomach extremely long and comprising nearly three-fourths total leng
of the stomach; pyloric bend considerably distant from cardiac junction,
terminal third of stomach sharply recurved craniad and laterally toward

638 The University Science Bulletin

cardiac vestibule; terminal stomach tapering gradually from cardiac vesti-
bule to sulcus angularis just proximal to bend and remaining narrow to
pyloric sphincter, with musculature moderate and relatively uniform in depth
throughout tubular portion. Pyloric sphincter thin, slightly more massive
on greater than lesser curvature, directed into lumen of duodenum in both
specimens observed; sphincter composed of circular layer and muscularis
mucosae; pyloric mucosa lining sphincter deep; external constriction at
sphincter slight; pyloric tube distal to bend approximately parallel to cardiac
vestibule, hence direction of gastric orifices approximately equal with both
directed anteriorly in situ.

Entire inner surface area of stomach covered with mucosal lining with
pattern of rugal folds similar to that found in Glossophaga; submucosa
scanty throughout, included within rugae and best developed within pyloric
zone.

Musculature. — Both major muscular layers shallower throughout lesser
curvature than elsewhere (excepting prominent thickening of wall at
pyloric bend within lesser arc) ; muscularis externa extremely thin through
cardiac vestibule along left wall, circular layer the deepest layer throughout;
muscle wall moderate in depth through lesser curvature to pyloric bend,
thickened slightly at bend to pyloric sphincter. External layers generally
equal in depth throughout fundic caecum with circular layer thickened
slightly at bases of rugae; wall of fundic caecum slightly thicker than that
of lesser curvature. Longitudinal layer deeper relative to circular layer
through more of stomach than in any other bat examined; fibers unusually
oriented and prominently compartmentalized as discussed previously
throughout fundic caecum, caecum appearing unusually elastic; muscularis
externa of greater curvature lateral to fundic caecum relatively uniform in
depth to pyloric sphincter, with no substantial thickening of either external
layer in pyloric tube and no gathering of circular fibers into bundles in
lateral two-thirds of tubular stomach. Muscularis mucosae thin, double
layered, slightly deeper within rugae in comparison to stomach wall proper.
Depth longitudinal layer 10 (cardiac vestibule) to 80, circular layer 25
(cardiac vestibule) to 150, muscularis mucosae 10 to 15.

Organization of gastric mucosa. — Orientation of mucosal folds similar
to that of Glossophaga but longitudinal orientation better developed along
ventral and dorsal walls of caecum; folds gathered longitudinally within
most of tubular stomach and all of cardiac vestibule; folds more or less
longitudinal throughout upper two-thirds of caecum, with pronounced
zig-zag orientation and directed transversely into many short folds along
posterior wall within lower third; dorsal surface of caecum dilated with
radially arranged rugae similar to those in Glossophaga. Zones of gastric
mucosa distributed in pattern similar to those of other frugivorous and necti-

Studies of Stomachs of Selected American Bats 639

vorous phyllostomatids, but transitional zone relatively long and confined
to narrow area proximal to pyloric bend on lesser curvature and more
extensively distributed through pyloric bend on greater curvature; reduc-
tion in number of parietal cells abrupt within transition, but loss of
zymogenic cells more gradual, occurring within proximal half; fundic
zone more extensive on greater than lesser curvature with morphology
of fundic glands highly variable from one region to another, and unusual
in several structural features (refer to section on fundic mucosa); extensive
variation in depth, mucosa slightly shallower within fundic caecum than
remainder of zone; pyloric zone longer than in Glossophaga, occupying
all of terminal segment of pyloric tube to rim of pyloric sphincter. Mucosa
consistently deeper than that of Glossophaga relative to overall size of
stomach; deepest within tubular portion of fundic mucosa and basal half
of cardiac vestibule, decreasing only slightly in depth toward lateral ex-
tremities; mucosa shallowest within furrows between rugae (most apparent
in caecum) ; mucosa along lesser curvature about same depth as that along
greater curvature.

Cardiac glands. — Similar to those of Glossophaga in cellular structure
and morphology of gastric pits; glands confined to narrower zone than in
Glossophaga, tubules somewhat narrower, and lamina propria less abundant
at bases of glands; junction with stratified squamous epithelium abrupt;
glandular zone consistent in length throughout vestibule; depth approxi-
mately 90.

Fundus. — Fundic mucosa characterized by distinctive variability in sur-
face epithelium and by extensive variability in numbers of various cellular
components beneath surface epithelium from apex of fundic caecum to
pyloric mucosa. Depth of fundic mucosa 220 to 370.

Zymogenic cells in marked abundance only within fundic caecum and
proximal-most portion of tubular fundus (occupy up to basal 40%, gradu-
ally reduced in frequency laterally toward pylorus); parietal and mucous
neck cells increase in abundance toward pylorus with parietal cells increas-
ing mostly within middle portion of gland and mucous cells progressively
increasing in abundance throughout entire length of gland; parietal cells
nearly absent within apical arc of fundic caecum (only three or four cells
persisting within central area of gland at extreme apex of fundic caecum),
increasing tremendously from apex of caecum to termination within py-
lorus (heaviest concentration in mid-stomach and in upper midportion
of tubule throughout most of fundic mucosa), progressively decrea
toward base of gland, occasionally scattered among basal chief cells withi
tubular fundus only.

Chief cells unusual in structure and arrangement; small and in
in size; nuclei large, spherical, usually centrally located; cells more or

640

The University Science Bulletin

, 0,5mm ,

Fig. 23. Cardiac vestibule (CV) of Carollia perspicillata.

Fig. 2-1. Fundic caecum of Carollia perspicillata.

Fig. 25. Surface epithelium of the fundic mucosa in Carollia perspicillata.

Studies of Stomachs of Selected American Bats 641

columnar in shape, inner borders forming narrow but straight lumen (not
observed in complexes of chief cells of other bats examined); basal portion
of gland containing only slightly bulbular zymogenic cells.

Parietal cells exhibiting cline in size within gland, as well as cline in
abundance within fundic zone; lower cells smallest, increasing in size to
neck of gastric pit although never truly large relative to tubule breadth;
parietal cells usually ovoid with large spherical to ovoid nuclei (nuclei
generally decreasing in size toward lumenal surface).

Argentaffin cells observed infrequently among chief cells; triangular
with clear cytoplasm excepting small amount of peripheral granulation
at one corner of triangle.

Mucous neck cells substantially more abundant than chief cells in caecum,
decreasing sharply in abundance toward pylorus (corresponding increase
in parietal cells, coupled with progressive decrease in depth of gastric pits):
cellular morphology of mucous neck cells located among parietal cells
identical to that described for Glossophaga. Gastric pits extremely deep
and broad within fundic caecum owing to unusually thickened surface-
epithelium and gastric pit cells; surface epithelium deep throughout stomach
(Fig. 25), although especially so within caecum; pits 50-60% length ot
glands at apex, decreasing abruptly within middle portion of caecum,
shallow (15-20% length of glands) through remainder of fundic mucosa
and transitional area; pit cells large and cuboidal, increasing in size toward
surface of mucosa, nuclei enormous and spherical. Surface epithelium com-
posed of extremely elongate columnar cells in single layer with large,
laterally compressed nuclei; surface cells decreasing in depth through tubular
portion of fundus and becoming again thickened within pyloric mucosa.
Lamina propria beneath glands typically scant; numerous smooth muscle
fibers between glands within fundic caecum arising from muscularis
mucosae, and distinctively more abundant than in other bats studied.

Pylo-jundic transition. — Narrow zone, although relatively broad among
phyllostomatids studied; extensive on greater arc (parietal cells decrease
gradually in number toward sphincter with occasional cells lingering through
pyloric bend); parietal cells that persist in pylo-fundic region not confined
to specific region within subepithelial portion although usually basal;
loss of parietal cells more abrupt on lesser arc than greater curvature;
chief cells disappear abruptly throughout lateral margin of fundic zone,
gastric pits gradually increase in depth with changes in complements oi
cellular components; pyloric glands notably decrease in depth (in compari-
son to fundic mucosa) with reduced depth of terminal transitional mucosa
reflecting this decrease; surface epithelium deepens abruptly at termir
margin of zone with no evidence of thickened epithelium within tran-
sition zone; complement of lamina propria as in fundic mucosa; dept
ISO to 250.

642 The University Science Bulletin

Pyloric glands. — Distribution similar to that found in other frugivorous
and nectivorous phyllostomatids although somewhat more extensive than
in Glossophaga; occupying pyloric tube distal to bend; glands short, rela-
tively straight and tubular, shallowest within inter-rugal furrows. Extensive
differential gradient in cell morphology from base to surface of mucosa;
basal cells small with large nuclei, essentially like those within glands of
Glossophaga and upper cells progressively larger with slight decrease in
nuclear size and nuclei occasionally compressed laterally. Lumen broad
and increasing in breadth toward surface; surface epithelium unusually
thick, comparable to that above fundic mucosa of caecum; unusually ex-
tensive layer of areolar connective tissue underlying pyloric glands, re-
stricted to pyloric mucosa and with large number of interspersed mast cells
and slight increase in interposed collagenous fibers compared to remaining
mucosa (depth of connective tissue layer approximately 30). Depth of
pyloric glands 130 to 200.

Glands of Brunner. — Extremely limited in distribution, occurring as
single narrow clump on distal surface of pyloric sphincter, and with oc-
casional scattered glands within mucosa along proximal surface of sphincter
within pylorus but never within mucosa of stomach wall; glands extremely
short and broad, highly convoluted; in highest concentration along lip
of sphincter and decreasing gradually in depth to lateral junction of stomach
and duodenum; cells large, pyramidal with small ovoid, peripheral nuclei;
glands most limited in distribution of any bat examined.

Remarks. — The stomach of Carollia perspicillata somewhat resembles in
general outline that of Glossophaga soricina. The extent of development of
the fundic caecum is comparable to that of Glossophaga, although the cae-
cum is less obviously distinguished from the remaining stomach by a sulcus.
The distribution and orientation of mucosal folds is similar to that ob-
served in Glossophaga, but there is somewhat more tendency toward longi-
tudinal orientation within the dorsal and ventral surfaces of the fundic
caecum. Carollia and Glossophaga also share the deep surface epithelium
within the fundic caecum, and a marked variation in distribution of parietal
cells within the fundic mucosa.

Carollia is unique in the patterns of distribution of zymogenic and
parietal cells within the fundic stomach. Chief cells are commonest within
the fundic caecum (often reduced in this area in other bats) and decrease
progressively in number laterally, toward the terminal portion. Correspond-
ingly, parietal cells are nearly absent in the fundic pouch, increasing in
abundance to the termination of the pylo-fundic transition zone.

The tendency within the Phyllostomatidae toward elongation of the
terminal stomach is marked in C. perspicillata. Recurvature of the relatively
long terminal endpiece is so sharp that it curves back on itself. Speculations

Studies of Stomachs of Selected American Bats 643

regarding the possible functional significance of this recurvature follow in
a later section.

Structural features at the aboral end of the stomach are strikingly like
those of Artibeus. The pyloric valve is long and thin, directed into the duo-
denal lumen in both species. The glands of Brunner differ slightly histo-
logically but in both species are confined to small masses at the distal
surface of the pyloric sphincter with few within the submucosa of the
duodenal wall.

Subfamily Sturnirinae
Sturnira lilium (E. Geofrroy-St. Hilaire, 1810)

Specimens examined (3). — Three specimens of Sturnira lilium parvidens
from Nicaragua (KU 106086, 106102, 106104).

Overall dimensions. — Greatest length 9.0-10.2; greatest breadth 6.0-6.3;
gastroesophageal junction to pyloric valve 4.7-4.9; gastroesophageal junction
to apex of fundic caecum 4.8; breadth of pylorus at sphincter 1.2-1.3.

General morphology. — Small, slight curvature along back (dorsal) sur-
face, hence concave dorsally; fundic caecum saccular; cardiac vestibule
long and tapering; elongate, narrow pylorus recurved craniad; mucosal folds
gathered in unusual direction. Gastroesophageal junction markedly su-
perior to pyloric sphincter and junction with cardiac mucosa abrupt; no
evidence of anatomical cardiac sphincter; cardiac vestibule (Fig. 26) ex-
tremely long, thin walled, expanding gradually from gastroesophageal
junction to terminus in lesser curvature, oriented in S-configuration with
cardiac orifice directed antero-laterad approximately as pyloric orifice; wall
of cardiac vestibule more extensive along surface toward pylorus than that
adjacent to upper caecum. Fundic caecum extremely saccular and thin
walled; expanded on both upper and lower surfaces into spacious chamber
with rounded apex; dorsal surface expanded; fold of stomach wall along
rear (dorsal) surface that distinguishes cardiac vestibule and caecum. In-
cisure cardiaca deep and narrow; occurring as distinctive depression only
within posterior half of stomach with slight thickening of circular muscle
layer at base of incisura, and with orientation of circular muscle fibers
separated into sheets by septa; prominence of incisura attributable to notable
anterior dilation of fundic caecum. Tubular stomach distal to fundic caecum
exceptionally long and narrow; terminal half of tubular region at 90-degree
angle to remaining stomach and distal-most half with pyloric sphincter
recurved craniad; pyloric bend gradual; tapering gradually to pyloric bend
then relatively uniform in breadth to pyloric sphincter. Musculature of
pyloric sphincter poorly developed throughout, least well represented on
lesser curvature (reduced to simple "ridge" of circular muscle in mos
places and occasionally totally absent in portions of its circumfer
circular muscle and muscularis mucosae contributing to sphincti

644

The University Science Bulletin

2mm

j

Fig. 26. Mid-longitudinal representation of the stomach of Sturnira lilium. For explana-
tion of symbols, .sec Figure 3.

overlying mucosa deep along muscular ridge, hence total sphincter mass
extensive; external constriction at sphincter slight; sulcus intermedius pres-
ent at pyloric bend not prominent. Lamina propria extremely rare every-
where within stomach wall (except beneath muscularis mucosae within
pyloric mucosa), prominent within rugae (greatest within terminal portion
of pyloric tube); rugal folds and gastric mucosa well developed throughout,
no apparent secondary reduction of either; mucosa especially deep within
cardiac vestibule.

Musculature. — Longitudinal and circular layers thin throughout fundic
caecum and cardiac vestibule, with local thickening of circular layer at bases
of rugae; gathering of circular fibers into bundles pronounced only within
lower wall of caecum and occurring to a lesser degree within midregion of
greater curvature. Circular layer always thicker than longitudinal layer,
with regions of thickened circular muscle confined to tubular stomach
(originating on lesser curvature at base of cardiac vestibule, then increasing
gradually in depth to maximum at pyloric bend, then decreasing slightly
again to second thickening at pyloric sphincter); essentially identical pattern
as above within greater curvature, although musculature consistently thicker

Studies of Stomachs of Selected American Bats 645

than in lesser curvature in corresponding areas. Longitudinal layer a thin
sheet throughout stomach, including pyloric tube, not contributing to
sphincter. Muscularis mucosae a thin double layer throughout with little
submucosa beneath and most prominent in pyloric area. Ganglia oi Auer-
bach's plexus abundant only within pyloric tube and greatest within greater
curvature. Depth longitudinal layer 15 to 20, circular layer 20 to 150,
muscularis mucosae less than 10.

Organization of gastric mucosa. — Mucosa gathered uniformly into longi-
tudinal, parallel folds throughout tubular portion of stomach and cardiac
vestibule, with slight interdigitation of basal-most rugae in vestibule;
mucosa of fundic caecum arranged in interdigitating network of deep
rugal folds with net direction of orientation more or less longitudinal (apex
of fundic caecum toward pyloric tube), but some secondary transverse
orientation apparent; interdigitation of folds confined to saccular fundus
proper, fundic mucosa of mid-stomach gathered longitudinally with little
or no gathering into short discontinuous folds along arc of greater curva-
ture; remaining mucosa consistently gathered on long, continuous ridges
traversing greater part of length of stomach.

Fundic mucosa uniformly shallow on apexes of rugae within fundic
caecum and mid-stomach, no local reduction within apex of caecum and
notably deepest throughout cardiac vestibule where glands are consistently
longest and narrowest; pyloric glands notably shallower than fundic or
cardiac mucosa, although mucosal membrane of pyloric tube exceptionally
thick owing to underlying layer of glands of Brunner; both fundic and
pyloric mucosa thicker on lesser curvature than greater.

Four glandular types recognized within mucosal membrane including
glands of Brunner (submucosal in duodenum) beneath certain portions;
cardiac glands confined within typically narrow belt at gastroesophageal
junction, zone equal in length throughout; fundic mucosa occupying all
of cardiac vestibule and saccular fundus, and mid-stomach up to depletion
of chief cells proximal to pyloric bend; transitional area narrow and un-
equally distributed on lower surfaces, confined to narrow region within
pyloric bend; pyloric mucosa occupying entire surfaces of short pyloric
tube distal to pyloric bend and extending slightly beyond pyloric sphincter
into duodenum; glands of Brunner present in typical distribution at superior
end of duodenum but also occurring within gastric mucosa beneath entire
pyloric gland complement and most of pylo-fundic transition area, gradt
ally decreasing in abundance from gastroesophageal junction througl
pylorus to termination within pyloric bend; no intergraduation of
Brunner's glands with zymogenic cells of fundus, hence junction abrupt.

Cardiac glands.— Short, thin and tubular within narrow zone at ga
esophageal junction; junction with stratified squamous epithelium

646 The University Science Bulletin

glands narrow and composed of single cell type; glands two to three times
breadth of adjacent fundic glands; mucous cells extremely large (twice size
of surface epithelial cells of remaining stomach); cells pyramidal with
small, ovoid nuclei within lower half of cell (not juxtaposition to basal
membrane), cells increasing in size from base to surface; mucous border
deep; lumen extremely broad and expanded at base along with gland;
transitional zone to fundic mucosa absent; glands distinctive in morphology,
easily distinguished from remainder of gastric mucosa; depth 60, breadth
of zone 30.

Fundus. — Glands extend throughout all of saccular and tubular stomach
up to pyloric bend; glands uniform in basic morphological pattern, but
varying slightly in depth and breadth in cardiac vestibule in comparison to
remainder of fundic stomach. Glands generally short and broad with ex-
tremely shallow gastric pits; chief cells extensively distributed within gland,
succeeded by thin but distinctive zone of primarily mucous neck cells
superior to chief cell complex, with parietal cells confined mostly to upper
half and most prominent immediately beneath gastric pits; no distinctive
reductions in depth of mucosa anywhere within mucosa including fundic
caecum; mucosa of cardiac vestibule considerably deeper than remainder of
fundus, owing to substantial increase in relative numbers of parietal cells,
and also gastric pits somewhat lengthened. Depth of fundic mucosa 120 to
170 within tubular stomach and fundic caecum and up to 220 in cardiac
vestibule.

Basal zymogenic cells extensive in numbers within glands and generally
throughout fundic region, may occupy up to 60% length of gland (normally
basal 40-50%); zymogenic cells large, polyhedral, irregular in size, little
apparent size gradient within gland; nuclei small and ovoid to spherical,
peripheral although not necessarily juxtaposed to basal membrane; large
and extensive chief cells result in notable basal expansion of fundic glands,
chief cell complex most conspicuous element within glands. Cells strictly
definable as argentaffin cells not observed but presumed present.

Mucous neck cells particularly abundant within medial 20% of gland,
large and with small peripheral, spherical nuclei; clear, relatively extensive
cytoplasm; upper mucous neck cells among parietal cells smaller with larger
nuclei and wedged between adjacent parietal cells.

Parietal cells long, narrow, ovoid and with oval nuclei; nuclei vary con-
siderably in size and often displaced peripherally; size gradient within
cell distribution present but not pronounced; cells confined mostly to upper
half of gland and increase progressively in frequency to maximum directly
beneath gastric pits; parietal cells occurring infrequently among zymogenic
cells; relatively uniform in numbers throughout fundic mucosa although
somewhat more abundant within cardiac vestibule than remainder of area.

Studies of Stomachs of Selected American Bats 647

Gastric pits shallow, although somewhat variable in depth, usually oc-
cupying upper 10-15% total length of gland; cells small with large con-
sistently spherical nuclei throughout pit (no lateral compression); lumen
narrow; surface epithelium shallow and cells small with spherical nuclei,
mucous border not deep.

Pylo-fundic transition zone. — Narrow zone at pyloric bend, more ex-
tensive on greater than lesser curvature; unusual in having glandular portion
in form of cellular elements identical to glands of Brunner; transitional
glands generally intermediate in height between fundic and total pyloric
mucosa (inclusive of basal cells of Brunner's glands); depletion of zymo-
genic cells abrupt at beginning of zone, with parietal cells persisting through-
out as small remnants confined to area above basal mucous elements and
decreasing in number gradually toward pyloric sphincter; parietal cells
not observed among basal mucoid cells (presumed glands of Brunner) with
no region of gastric mucosa existing in which parietal cells extend to base;
gastric pits increase gradually in depth with progressive loss of parietal
cells; surface epithelium identical to that of pyloric area, shallow, with
small cells having small and spherical nuclei; depth 180 to 200.

Pyloric glands. — True pyloric mucosa restricted to short pyloric tube
distal to pyloric bend; glands stunted owing to layer of underlying glands
of Brunner; narrowly expanded into bulb at base, sharply increasing in
breadth from midregion to surface epithelium, resulting in funnel-shaped
appearance; depth variable, generally increasing gradually from origin
within pyloric bend to termination at pyloric sphincter; composed of single
cell type; cells small and with extremely large spherical nuclei which de-
crease slightly in size to surface epithelium and never are laterally flattened;
cells narrow and elongate with extensive mucous border; lumen broad,
increasing gradually toward surface; thin layers of connective tissue and
smooth muscle fibers derived from muscularis mucosae separating pyloric
glands from underlying glands of Brunner; surface epithelium relatively
thick when compared to that of adjacent transitional glands; lamina propria
thin layer beneath pyloric glands; basal glands of Brunner discussed in
following section; pyloric glands do not penetrate into lumen of duodenum;
depth 100 to 240 (not including basal glands of Brunner).

Glands of Brunner.— Moderate in abundance within duodenum, occu-
pying proximal 1.5 mm and occurring in single mass continuous with com-
plement in stomach; heaviest concentration at sphincter; tubules decreasing
progressively in length toward termination in duodenum; submucos
within duodenum.

Glands occupy basal position within mucosal membrane of pylor
to lesser extent, pylo-fundic transition area (Figs. 26, '.
and submucosal within pylorus (beneath muscularis mucosae)

648

The University Science Bulletin

0.1mm

^** %<;

*

*

^

*V

>*

'^-v*

''/r

Fig. 27. Glands of Brunner (BG) within the pyloric mucosa (PM) of Sturnira lilium; the
pyloric sphincter (PS) also is seen.

Fig. 28. Glands of Brunner (BG) within the pylorus of Sturnira lilium showing reaction
to I'AS; the pyloric constriction (PC) marks the limit of the pyloric stomach (left of con-
striction).

Studies of Stomachs of Selected American Bats 649

usually observed superior to muscularis mucosae (Fig. 28); increase pro
gressively in depth from origin at pyloric bend to distribution within duo-
denum.

Morphology of glands identical throughout (Fig. 27); tubules highly
convoluted, branching into two or three units and consistently large through-
out entire distribution; tubules open lumenally by way of pores between
adjacent pyloric glands or crypts of Lieberkuhn depending upon location
within gastrointestinal tract; tubules never an anatomical subunit of any
other glandular type; cellular morphology of Brunner's glands consistent
throughout distribution; cells large, pyramidal, with light granulation;
nuclei extremely small, juxtaposed to basement membrane, moderately to
heavily compressed laterally; glands large in comparison to most other
bats examined and with expansive lumen; muscularis mucosae and lamina
propria comparatively well developed beneath pyloric complement of Brun-
ner's glands.

Remarks. — Sturnira demonstrates the most pronounced invasion of the
glands of Brunner into the pyloric stomach of any bat examined. Although
this condition is not restricted to this species, it is most obvious and best
developed here. It is interesting to note that in addition to having large
numbers of Brunner's glands beneath the entire distribution of pyloric
mucosa, the pyloric sphincter is distinctively reduced, and indeed virtually
absent in portions of the gastroduodenal junction. This condition suggests
that the invasion orally of Brunner's glands may be partially a developmental
by-product, in the absence of a prominent barrier conceivably produced
by a muscular pyloric sphincter. Although Molossus, with its prominent
pyloric sphincter, also apparently has Brunner's glands beneath the pyloric
mucosa, the condition is qualitatively and quantitatively much less distinctive
than in Sturnira. Investigations of gastrointestinal development in bats
should prove useful in determining any relationship between the size of
the pyloric sphincter and the extent of Brunner's glands. To my knowledge,
the studies herein reported include the first mention of a distinctive comp-
plement of the glands of Brunner in the stomach of a placental mammal.

The musculature of the stomach wall of Sturnira is similar to that
observed in other non-carnivorous phyllostomatids— thin and with the
circular layer arranged locally into bundles enclosed by thin, tissue-likc
septa.

Subfamily Stenoderminae
Artibeus lituratus (Olfers, 1818)
Specimens examined (3).— One specimen from Chiapas (uncatalogi
and two from Nicaragua of Artibeus lituratus palmarum (KU 10i

106236).

Overall dimensions.— Greatest length 11.6-13.5; greatest breadt

650

The University Science Bulletin

2mm

Fig. 29. Mid-longitudinal representation of the stomach of Artibeus lituratus. For explana-
tion of symbols, see Figure 3.

greatest breadth of fundic caecum 5.5-6.0; gastroesophageal junction to py-
loric valve 6.0-6.1; gastroesophageal junction to apex of fundic caecum
8.4-9.1; breadth of pylorus at sphincter 1.6-1.7.

General morphology. — Similar in many gross morphological features to
stomach of Sturnira lilium, although considerably larger and extremely
saccular, with unusually large fundic caecum and long, broad cardiac
vestibule; dorsal surface concave and with marked dorsolateral curvature
of entire stomach.

Pyloric tube relatively long and narrow (Fig. 29), recurved sharply
anteriorly distal to pyloric bend; musculature thick; pronounced sulcus
intermedius in central portion of tube; marked external pyloric constriction
at sphincter; nearly entire internal surface of terminal endpiece lined with

Studies of Stomachs of Selected American Bats

ANTERIOR

651

RIGHT

Fig. 30. Diagram showing anastomosing rugal folds with extensive interdigitation of
ridges in the fundic stomach of Artibeus lituratus. Note the anterior-posterior, rather than
longitudinal (left to right) direction of the folds.

pyloric-type mucosa; rugae (Fig. 30) and inclusive muscularis mucosae and
lamina propria unusually well developed along greater arc of pyloric tube
beneath pyloric glands, underlying submucosa unusually well developed
in pylorus; phyloric sphincter (Fig. 31) of circular muscle and muscularis
mucosae with slight contribution of longitudinal muscle layer at extreme
base of sphincter, gastric mucosa well developed on musculature of valve,
sphincter muscle broad at duodenal lumen in preserved specimens, sphincter
muscle broad at origin on stomach and tapering to narrow endpiece me-
dially, sphincter symmetrical; sphincter funnel-shaped rather than disc-
shaped and must produce effective seal upon contraction of circular muscle
if functional; pyloric orifice displaced considerably inferior (4-5 mm) to
cardiac orifice with both directed in nearly same direction.

Incisura cardiaca present but poorly developed. Cardiac vestibule (Fig.
32) massively developed in all directions into distinct chamber (externally)
originating at narrow gastroesophageal junction and increasing gradually
in breadth to lower terminus within lesser curvature; vestibule in one spec
men with slight S-configuration, anteriormost region reflected laterally and
parallel to pyloric tube (probably result of slight "tilt" of stomach in sit

652 The University Science Bulletin

vestibule expanded frontally yielding greater external as well as internal
distinction to region than in Sturnira; cardiac orifice substantially superior
to pyloric orifice.

Vestibule emptying into greatly dilated saccular fundus divisible into-
two general areas, 1) central portion between pyloric bend and fundic
caecum and 2) extensively dilated fundic pouch; fundic caecum extremely
large, more rounded on dorsal surface {in situ) than in Sturnira; tapering
laterally to narrow apex, distinguished from remaining fundus by fold in
stomach wall on dorsal surface as in Sturnira; gastric mucosa not decreasing
in depth within caecum, thin walled.

Lamina propria somewhat more abundant than in most other bats
examined but non-muscular submucosa generally thin throughout (with
exception of pyloric tube, and within and at base of rugae in fundic caecum,
where loose connective tissue often dense and moderately abundant).
Stomach clearly adapted for acceptance of, and perhaps storage of, large
amounts of food.

Musculature. — Muscle layers of fundic caecum and fundus of mid-
stomach irregular in depth and morphology as in Sturnira; stomach wall
extremely thin throughout most of fundic stomach and cardiac ampulla
(except for circular layer at origin of rugal folds where internal muscle layer
notably thickened and grouped into large bundles surrounded by thin septa
of connective tissue); musculature especially thin in cardiac vestibule; circu-
lar muscle also thickened at base of rugae, single large blood vessel ob-
served within lamina propria at base of, and ascending into, each rugal
fold with system best developed in fundus, to lesser degree in pyloric tube.

Longitudinal muscle layer uniformly thin throughout most of stomach,
slightly thickened in pyloric tube in comparison to remaining stomach.

Circular muscle greatly thickened thoughout greater curvature of pyloric
tube, decreasing gradually in depth through pyloric bend; circular muscle
of lesser wall of pylorus thickened locally proximal to pyloric sphincter,
thin throughout remainder of pylorus; lamina propria above circular layer
relatively thick within greater arc of pylorus.

Muscularis mucosae forming generally thin sheet beneath gastric mucosa
but thicker beneath gastric mucosa and within rugal folds of pyloric tube.
Musculature of gastroesophageal junction thin, lacking local thickening
except for definite bundling of circular layer; no conclusive evidence sug-
gesting presence of active cardiac sphincter, although local segregation of
circular muscle fibers reveals presence of questionable morphologic speciali-
zation. Depth longitudinal layer 20 to 30, circular layer 20 in fundic caecum
to 1S5 in pyloric tube, muscularis mucosae 10.

Organization of gastric mucosa.— Mucosa arranged throughout fundic
stomach into large, complex, interdigitating networks of folds (more com-

Studies of Stomachs of Selected American Bats

653

Fig. 31. Pyloric sphincter (PS) of Artibeus lituratus.
Fig. 32. Cardiac vestibule of Artibeus lituratus showing fold in body wal
denum (DUO), and cardiac vestibule (CV). Anterior is to the left in this photograph.

654 The University Science Bulletin

plex than in Sturnira) with more or less regular pattern as illustrated in
Figures 30, 32; general direction of orientation antero-posterior, angled
such that folds progress antero-posteriorly, but directed upward and to the
right; interdigitation more pronounced in fundic caecum than remainder
of fundus; folds in pyloric tube irregular in shape, clearly directed more
longitudinally than those of fundus and not interdigitating. Mucosal folds
generally extremely well developed without local reductions anywhere in
fundus, although slightly shallower pyloric glands on musculature of pyloric
sphincter than elsewhere; mucosa unusually uniform in depth throughout
fundus, including furrows between folds, less uniform in pyloric tube
owing to thickening on sides of rugae and reduction in depth within fur-
rows between adjacent folds.

All glandular types present, although pylo-fundic transitional glands
reduced to remnant; cardiac glands occupying narrow belt at gastroesopha-
geal junction; fundic glands occupying all of remaining surface area except
pyloric tube; fundic glands long and narrow, exhibiting notable specializa-
tion in distribution of cellular types; pylo-fundic transition zone extremely
narrow, basal zymogenic cells of fundus extending to pyloric bend; pyloric
glands extensive, unusual in appearing to consist of two cellular types, form-
ing mass of enlarged pyloric sphincter.

Cardiac glands. — Narrow zone inferior to gastroesophageal junction;
similar to pyloric glands in basic morphology, but only single cell type
present; cardiac glands longer and broader than adjacent fundic glands,
broadly expanded basally, cells long and narrow in single dense layer;
nuclei spherical with marked size gradient of decrease from lower to upper
part of glands; luminal mucous border extensive (over twice breadth of
nucleus) ; broad foveolae plus extensive mucous border yielding relatively
broad glands; junction with fundic mucosa abrupt; transitional element
absent as zymogenic cells present within glands adjacent to lower cardiac
elements; junction with esophageal epithelium equally abrupt; lamina
propria beneath glands thin; depth 210 to 280.

Fundus.— Distribution of glands similar to that of Sturnira; mucosa deep
and of rather uniform depth throughout stomach; extends from posterior
limit of cardiac zone through cardiac vestibule and fundic caecum laterally to
pyloric bend; arrangement of mucosal folds complex; glands long and ex-
tremely narrow, tightly compact, little or no apparent convolution, expanded
at upper and lower ends but narrowing in midregion. Depth of fundic
mucosa 160 to ISO.

Zymogenic cells small, pyramidal with small, spherical nuclei (usually
in outer periphery of cell but occasionally displaced medially); cells rela-
tively limited in distribution within basal 15-20% of gland, never in complete

Studies of Stomachs of Selected American Bats 655

absence of parietal cells; uniform in distribution within all glands containing
parietal cells. Argentaffin cells not observed but assumed present.

Parietal cells remarkably abundant, occurring throughout virtually en-
tire length of gland to just beneath surface epithelium, most heavily con-
centrated in medial 60% where abundance virtually excludes mucous neck
cells; cells unique among bats examined in being abundant among chief
cells at base of glands (although less abundant than in medial portion)
and somewhat less common immediately beneath gastric pits; cells generally
large, circular, uppermost cells often four times breadth of those at base,
occasionally as broad as gland itself; nuclei also large, usually ovoid, most
often peripheral with nuclei of basal-most cells extremely large and ovoid.

Mucous neck cells extremely small and sparsely distributed among
parietal cells; less frequent than parietal cells in medial half of gland where
usually located along inner surface of tubule; nuclei spherical and smaller
than those of parietal cells.

Gastric pits extremely short and thick-walled, up to 15% length of
gland but usually less; cells lining pits large, densely packed, with large
nuclei and lumen narrow; surface epithelium extremely thick, represented
by dense clumps of epithelial cells with large, club-shaped nuclei.

Pylo-jundic transition zone. — Region so narrow as to appear nearly
absent; replacement of parietal cells by mucous neck cells abrupt in pyloric
bend; zymogenic cells present in base of glands in proximity of pyloric mu-
cosa, although reduced in abundance; gastric pits abruptly deepened at
terminal border of transition zone; depth about 170.

Pyloric glands. — Extensive and well-developed area occupying most of
tubular stomach distal to pyloric bend and terminating aborally at apex
of lip of pyloric sphincter; differing distinctively in morphology and dimen-
sions from remaining gastric mucosa including cardiac glands; pyloric glands
consisting of two apparent cellular types; approximately lower third occu-
pied by small light-colored cells with small round nuclei and slight cyto-
plasmic granulation; lower portion of gland bulbular; cells of upper two-
thirds larger and more ovoid, mucous border twice as broad as in lower
cells; basal cells clearly unique within mucosa and notably similar to cells
of Brunner's glands; pyloric glands long and broad, nearly twice as broad
as cardiac glands; surface epithelium thicker than that of fundus, nuclei
laterally flattened, mucous border extensive. Lamina propria unusuall)
developed along with submucosal elements in general, especially
muscularis mucosae thick in rugae as well as stomach wall prope
submucosa relatively abundant beneath muscularis mucosae,
frequent above; extensive deposits of adipose tissue occurring
of pyloric tube (beneath muscularis mucosae) and to lesser ex
stomach; connective tissue immediately beneath pyloric gl

656 The University Science Bulletin

tensive than in fundus. Glands more irregular in depth than those of fundic
glands and reduced in depth within furrows between rugae and along
surface of pyloric sphincter. Depth of pyloric mucosa 140 to 185.

Glands of Br unner. — Restricted to immediate vicinity of pyloric sphinc-
ter and fewest in number of any bat examined (excepting Carollia) ; occur-
ring in narrow belt along distal surface of sphincter but not extending be-
yond lip of sphincter; large (more than twice breadth of pyloric gland),
compound tubular, not highly convoluted; cells long and narrow with
diffuse granulation, more like mucous cells of gastric mucosa than in other
bats examined; nuclei small, spherical, basal, never laterally compressed;
lumen large, as broad as single gastric gland, empties between crypts of
Lieberkuhn; glands unique in morphology among bats examined.

Remarks. — The stomach of Artibeus lituratus clearly is adapted to accept
large quantities of food material and may well serve to retard, secondarily,
its transport through the digestive system. Extensive development of the
cardiac vestibule and fundic caecum can only suggest modifications to
assimilate a bulky diet, probably normally fruit of trees and shrubs (Good-
win and Greenhall, 1961). Thin musculature of the entire non-pyloric
stomach would allow for expansion, and thus for consumption of large
quantities of food.

Several specializations occur that probably are related to retarding
gastric flow through the stomach, a phenomenon not uncommon among
mammals that feed primarily on herbaceous material. The orientation of
mucosal folds (Fig. 30), for example, would provide a mechanism for di-
recting some food material laterad into the unusually spacious fundic cae-
cum, upon contraction of the stomach wall, with the thickened circular
muscle bands within the bases of the rugal folds providing the necessary
contraction. Unusually extensive development of the pyloric sphincter and
tube may be modifications providing an effective closing mechanism against
large quantities of food, although the importance of an anatomical pyloric
sphincter in this regard may be in question in mammals (Selkurt, 1966).
The pyloric tube is extremely narrow in relation to total stomach size and
probably offers resistance to an escaping bolus.

The fundic mucosa exhibits reduction in numbers of zymogenic cells
within individual glands, although presence of these cells is uniform
throughout the entire region. Inasmuch as chief cells are the suspected site
of secretion of pepsin, it is not surprising that they are less abundant in this
primarily fruit-eating bat than in most other species studied. However, the
unusual abundance of HCl-producing parietal cells is not clear. In addition
to its function of providing the proper pH for activity of gastric proteases,
hydrocloric acid also functions in prevention of bacterial growth and stimu-
lation of secretion of secretin and pancreozymin in the duodenum, as well

Studies of Stomachs of Selected American Bats 657

as enzymes by the pancreas (Glass, 1968). Perhaps the extensive parietal cell
complement in Artibeus can be linked to one or more of these secondary
functions. No other species of bat examined exhibited such extensive num-
bers of parietal cells and, for the present, this feature must be considered
unique in Artibeus.

Although unique with respect to certain features, the stomach of
Artibeus is similar to that of other phyllostomatids in gross morphology.
Artibeus resembles phyllostomatids generally in possession of a saccular,
asymmetrically formed, muscular, and recurved pylorus, and thin muscu-
lature throughout a well-developed globular fundic caecum.

Although distinctive, the stomach of Artibeus lituratus is clearly similar
topographically to that of Sturnira lilium. Similarities include the broad,
elongate, slightly S-shaped cardiac vestibule, well-developed saccular fundic
caecum, short and sharply recurved terminal endpiece, and (internally) the
interdigitating, diagonally-oriented mucosal folds. Differences include a
broad pylo-fundic transition zone and predominance of chief cells in the
fundic mucosa in Sturnira, in contrast to an extremely narrow transitional
element and persistence of parietal cells as the most conspicuous cell type
of the fundic mucosa of Artibeus.

It is of some interest to note that Grasse (1955), in a classification of
stomachs of bats into 1) megachiropteran, 2) microchiropteran and 3)
Desmodus-type, included as primary features of the megachiropteran
("regime frugivore") type a conical cardiac region (cardiac vestibule of
Grasse), an extensive pyloric tube, and a spacious fundic caecum. Addi-
tionally, Grasse included an illustration (after E. Pernkopf) of the stomach
of Pteropus vampyrus, an Old World frugivorous megachiropteran, that
bears a resemblance to that of Sturnira and Artibeus in the features discussed
above. Also, Grasse (1955) stated that the genus Nyctimene, a group of fig-
eating megachiropterans of the subfamily Nyctimeninae, is exceptional
among members of the suborder in possessing a simple globular micro-
chiropteran-like stomach; Koopman (1967) and Walker (1968) noted that
Nyctimene may include insects in the diet.

Examination of more species and numbers of individuals of each kind
will be required before generalizations regarding coevolution of gastric
structure and food habits are truly conclusive. Further study may shed
light on the possibility of parallel evolution in gastric morphology in the
Megachiroptera and Microchiroptera, as suggested above.

Subfamily Desmodontinae
Desmodus rotundus (E. Georrroy-St. Hilaire, 1810)
Specimens examined (3).— Three specimens from Nicaragua of
dus rotundus minimis (KU 106260, 106262-63).

Overall dimensions.— Greatest length 75.0-86.0; greatest breadt

658 The University Science Bulletin

gastroesophageal junction to pyloric sphincter 1.9-2.1; gastroesophageal junc-
tion to apex of fundic caecum 72.5-83.0; breadth of pylorus at sphincter
2.3-2.5.

General morphology. — Stomach modified into narrow, elongate caecum
of relatively uniform breadth originating at gastroesophageal junction, with
diverticulation of gastrointestinal tract into stomach and duodenum (Fig.
33); caecal pouch initially directed laterally (to the left), bending sharply
caudad approximately 3 mm lateral to esophageal junction, continuing
aborally to approximately midway along length, then recurving sharply
craniad, and recurving again posteriorly just proximal to apex (Fig. 34);
stomach uniformly narrow throughout most of length, dilated slightly
near apex and slightly constricted within terminal 15 mm. Cardiac vesti-
bule absent, squamous epithelium terminating symmetrically and abruptly
at lesser curvature; aboral end of esophagus slightly expanded at junction;
no evidence of anatomical cardiac sphincter; muscularis externa and muscu-
laris mucosae of esophagus unmodified in dimensions or structure at junc-
tion with lesser curvature.

Pylo-duodenal junction and pyloric sphincter in close proximity and lo-
cated at right angle to gastroesophageal junction; pylorus occupying region
between pyloric sphincter and beginning of tubular fundic caecum (inclusive
of mucous glands at gastroesophageal junction); valves of pyloric sphincter
short and extremely thin, expanded slightly at apex, sphincter portion of
muscularis externa composed only of circular muscle layer; remaining
muscularis externa well developed (both layers); circular layer extremely
complex and somewhat variable in depth; tunica propria extensively devel-
oped throughout stomach as dense layer between muscularis mucosae and
circular layer (heavy complements between loosely organized bundles of
circular fibers, elastic fibers rare among collagenous fibers) ; stomach most
heavily vascularized among species examined with extensive numbers of
arterioles and venules within areolar connective tissue.

Gastric mucosa unusual in that most glands reduced to simple alveolus
or tubuloacinar type (Rouk, 1968); cellular complements reduced in number
as compared to other bats examined.

Musculature. — Wall relatively uniform in depth throughout caecum
but slightly reduced in depth along lesser curvature in vicinity of gastro-
esophageal junction. Longitudinal layer moderate in depth throughout
gastric pouch and narrowest within dilated portion and at apex, occasionally
observed broken into tightly packed clumps of fibers and sometimes ap-
pearing as two layers. Circular layer always deeper than longitudinal layer
at given point; fibers of circular layer organized throughout stomach
(excepting incisura cardiaca) into normally large, loosely arranged, trans-
versely-oriented bundles (Fig. 35) surrounded on upper surface and sides

Studies of Stomachs of Selected American Bats

659

■^

Fig. 33. Diagram of the cardiac and pyloric junctions of the stomach of Desmodus
rotitndtis. Dotted lines indicate the continuation of the fundic caecum; it was not possible to
include it to scale in this figure; GEJ, gastroesophageal junction. For explanation of oth
symbols, see Figure 3.

Fig. 34. Semidiagrammatic representation of the stomach of Desmodus rotundas.
the local dilation of the caecum near the apex.

660 The University Science Bulletin

by extensive amounts of connective tissue; circular muscle bundles slightly
variable in depth, no bundling at incisura cardiaca and gastroesophageal
junction. Muscularis mucosae in extremely thin sheets and no deeper within
rugae than stomach wall proper. Lamina propria mucosae extensive superior
to circular layer. Depth longitudinal layer 15 to 30, circular layer 30 to 100
(usually 60 to 75), muscularis mucosae less than 10.

Organization of gastric mucosa. — Rugal folds numerous in longitudinal
parallel rows throughout entire length of stomach with infrequent trans-
verse ridges connecting single pairs of rugae; folds in compressed zig-zag
configuration throughout stomach; extensive variation in height of rugae
but individual folds uniform in height throughout length.

Zone of cardiac mucosa here considered absent as mucous glands at
pyloric sphincter and gastroesophageal junction indistinguishable morpho-
logically and histochemically; pyloric mucosa extending 3 to 4 mm laterally
from pyloric sphincter to beginning of fundic tube (inclusive of mucosa at
cardiac junction); transitional zone narrow, symmetrical, occupying proxi-
mal-most 3.0 mm of fundic tube; remaining 95% of stomach lined with
markedly reduced fundic mucosa unlike any other bat examined; gastric
mucosa deepest within pyloric zone where glands truly tubular, decreasing
progressively through pylo-fundic transition to shallow acinar glands of
fundic tube.

Cardiac glands. — Considered absent as mucous glands at gastroesopha-
geal junction broad, relatively straight and tubular, indistinguishable from
those at pyloric sphincter.

Fundus. — -Fundic-type mucosa occupying nearly all of inner lining to
apex of caecum; uniformly shallow throughout; expression '"tubuloacinar"
as applied to fundic glands by Rouk (1968) appropriate; glands pouched,
short and broad (Fig. 36) ; simple branched tubular glands characteristic of
mammalian fundic mucosa not observed. Depth of fundic mucosa 60 to 70.

Chief cells normally within basal third of gland, maximally to midpoint;
cells pyramidal, narrow and elongate; nuclei small, rounded, and always
located basally. Argentaffin cells present although extremely rare; more or
less triangular, broader than those of other species examined; nucleus small
and ovoid; peripheral granulation not observed.

Parietal cells not abundant, sparsely distributed from base (occurring
among chief cells) to gastric pit; uniformly large and usually circular; not
abundant enough to allow examination of intraglandular size gradient
although no small parietal cells observed; nuclei greater in diameter than
those of adjacent mucous neck cells and uniformly circular.

Mucous neck cells most abundant element; cells small, more or less
cuboidal, with small antero-posteriorly compressed nuclei; cells extending
throughout middle 30-50% of gland. Gastric pits necessarily short and

Studies of Stomachs of Selected American Bats

661

T»V'.<V : • e:. . ***** .-. * 4^** . k mtk

my*'. > ~tf

"

Fig. 35. Stomach wall of Desmodus rotundas (lower fundus) showing lamina propr
anil circular muscle (CM); note the tubuloacinar-type mucosal glands, extensive
propria, and bundling of circular fibers.

Fig. 36. Tubuloacinar fundic glands of Desmodus rotundas. The longitudin.i
stomach is left to right.

662 The University Science Bulletin

:■ ■ ■-■:■ :■:■. if- / :■ :■■■ ■

■
if

■i

-<

20u

#'

37

Fig. 37. Surface epithelium of fundic mucosa in the stomach of Desmodtts rotundtis.

sometimes simple pore within surface epithelium; distinctive and rather
broad lumen throughout length of gland. Surface epithelium extremely
thick (Fig. 37) throughout all fundic mucosa; cells narrow and elongate,
with small slightly ovoid nuclei within basal third of cell, extensive mucous
border along lumenal surface. Lamina propria extensive beneath and
between fundic glands, collagenous fibers infrequent, however.

Pylo-fitndic transition zone. — Occupying narrow band within uppermost
portion of tubular stomach immediately to left of gastroesophageal junction;
acquisition of zymogenic cells abrupt at terminal margin; definition of
initial onset of sparse parietal cells difficult, although cells appear immediately
within glands and are notably reduced in depth (latter character signifying
beginning of transitional zone in this bat) ; reduction in glandular depth at
initial margin fairly abrupt, most of transitional area composed of tubulo-
acinar elements; gastric pits also decreasing abruptly in depth; lamina pro-
pria consistent with that of fundus; depth of transitional glands 70 to 150.

Pyloric glands. — Originating at gastroesophageal junction and extending
aborally to distal edge of pyloric sphincter and laterally to initiation of
caecal tube; short, broad, tubular glands composed of single cell type; not
notably coiled, broadly expanded at base, slightly expanded at upper limit.
Cells narrow and elongate, numerous in cross-section of tubule; nuclei

Studies of Stomachs of Selected American Bats 663

small, spherical, located basally. Lumen extremely broad in base of gland,
narrowing only slightly toward upper exit; glands deepest and narrowest
at cardiac junction, shallowest at onset of oxynctic mucosa and at pyloric
sphincter. Surrounding lamina propria and smooth muscle fibers con-
siderably less abundant than within transitional or fundic mucosa. Depth
of pyloric mucosa 150 to 210.

Glands of Brunner. — Restricted to proximal 2.5 mm of duodenum and
totally submucosal; distributional pattern similar to that in several other bats
examined; pronounced external constriction at pyloric sphincter resulting
from lateral expansion or bulge of glandular mass at proximal end of duo-
denum; glands similar dimensionally and morphologically to those oi
Artibeus; tubules broad, typically highly coiled; cells relatively numerous;
nuclei slightly smaller and more nearly ovoid than those of Artibeus;
lumen broad, increasing slightly in breadth toward crypts of Lieberkuhn;
lamina propria mucosae and smooth muscle fibers of muscularis mucosae
most extensive between tubules among bats examined.

Remarks. — Several authors including Huxley (1865), Grasse (1955),
Schultz (1965), and Rouk (1968) have previously illustrated and briefly
discussed the topography and gross morphology of the stomach of Desmodus.
Additionally, Schultz illustrated the gastroesophageal and pyloric areas of
a closely related species of vampire, Diphylla ecaudata. Grossly, the stomach
of Diphylla differs from that of Desmodus primarily in having a more
laterally-displaced pyloric sphincter relative to the gastroesophageal junction
(after Schultz), with a resulting probable increase in the distribution of
pyloric mucosa. Robin (1881) noted that the caecum of Diphylla averages
somewhat shorter than that of Desmodus. Of the previous investigators,
only Rouk has endeavored to describe the composition of the gastric mucosa;
my account differs somewhat in content from his.

The elongate tubular stomach and the tubuloacinar-type mucosal glands
are unique among bats examined. Because partly or wholly digested food
materials that are present in the plasma of the ingested blood meal may
represent the essential food source of vampire bats, it is not surprising to
find a secondary reduction of mucosal cellular elements responsible for
secretion of materials active in digestion of albumunous foods. Nutritional
elements utilized by vampires and the mechanisms of intestinal absorption
are little understood, thus retarding complete evaluation of this striking
reduction of gastric mucosa.

Pronounced segregation of circular muscle fibers into loosely arrange
bundles, with prominent amounts of interposed loose connective I
ably permits adequate stretching of the caecal wall to allow
rather large blood meals (see Wimsatt and Guerriere, 1962).
these authors noted an extensive condensation of the blood

664 The University Science Bulletin

stomach, with copious urine excretion following ingestion. The extensive
vascularization of the stomach observed in Desmodtts accounts morphologi-
cally for the rapid and distinctive fluid absorption from the stomach of this
species.

Family NATALIDAE
Natalus stramineus Gray, 1838

Specimens examined (1).— One specimen from Nicaragua of Natalus
stramineus saturatus (KU 106280).

Overall dimensions— Greatest length 7.3; greatest breadth 3.2; gastro-
esophageal junction to pyloric sphincter 3.9; gastroesophageal junction to
apex of fundic caecum 33\ breadth pylorus at sphincter 0.9.

General morphology. — Stomach small, markedly comma-shaped; gen-
erally similar to stomach of Plecotus in gross morphology, but strikingly
like stomach of soricids described by Myrcha (1967); fundic caecum mod-
erate in size, not directed anteriorly (but laterally), apex narrowly rounded
with no partitioning of caecum by folds of musculature. Well-developed
cardiac vestibule (Fig. 38) formed partially of dilated lower terminal end
of esophagus (Fig. 39), hence uppermost portion of vestibule lined with
stratified squamous epithelium (unique among bats examined); esophagus
entering stomach directly anteriorly, hence no angle of entrance and no
incisura cardiaca; cardiac vestibule grades gently into lesser curvature;
junction with stratified squamous epithelium abrupt and inferior to dila-
tion of terminal end of esophagus, cardiac orifice thus extremely broad;
stratified squamous epithelium smooth at junction with gastric mucosa;
no evidence of active cardiac sphincter.

Tubular stomach distal to cardiac junction long and narrow; muscula-
ture as well as pyloric sphincter better developed on greater than on lesser
curvature; pronounced sulcus intermedius proximal to pyloric sphincter;
stomach only slightly curved on the whole, tubular region relatively straight
as compared to other bats examined and lying in plane nearly perpendicular
to plane of esophagus; slight antero-lateral inflection of terminus originating
at sulcus intermedius. Marked external constriction at pyloric sphincter;
sphincter includes muscularis mucosae and circular layer; sphincter longer
and narrower on greater curvature. Lamina propria well represented be-
neath muscularis mucosae in rugae, otherwise scant, never abundant above
muscularis mucosae layers.

Mucosal folds well developed throughout stomach; musculature sub-
stantially thicker in greater than lesser curvature.

Musculature. — Stomach wall extremely thin throughout lesser curvature
from apical arc of fundic caecum to just proximal to pyloric sphincter.
Oblique muscle layer (muscularis mucosae) unusually thick and dense in
midregion between gastroesophageal junction and pyloric sphincter on

Studies of Stomachs of Selected American Bats

665

2mm

Fig. 38. Mid-longitudinal representation of the stomach of Natalus stramineus. For ex-
planation of symbols, see Figure 3.

lesser curvature, at one point as thick as circular and longitudinal layer
combined, formed of typically thin sheets throughout remainder of stomach.
Sheets of muscularis mucosae within esophagus continuous with those of
stomach with no alteration of morphology. Musculature (externa) of apex
and lower surface of fundic caecum three times thickness of that of lesser
curvature wall; longitudinal layer as deep as circular layer in much of
fundic caecum. Circular muscle layer of greater curvature deep throughout,
with some "bundling" in area adjacent to fundic caecum; longitudinal
layer shallow; marked local thickening of circular muscle layer just proxi-
mal to pyloric sphincter (only within greater curvature). Depth longitudi-
nal layer 10 (lesser curvature) to 45, circular layer 30 to 140, muscularis mu-
cosae 4 to 15 (lesser curvature).

Organization of the gastric mucosa. — Mucosa gathered uniformly into
parallel, longitudinal rows throughout all of stomach with no interdigita-
tion of ridges; mucosa deepest along lesser curvature especially in mid-
region at base of cardiac vestibule; only slightly reduced in depth at apex
of fundic caecum and considerably less so than in Plecotus with no reduc-
tion in frequency of rugae at apex; pyloric mucosa deep (deeper than most
of fundic complement); fundic mucosa deep at junction with esophagus,
uniform in depth throughout most of cardiac vestibule (in contrast to
shallow, variable cardial mucosa of Plecotus); cardiac glands unusual
both morphology and distribution (found beneath esophageal epi
at cardiac junction as well as in stomach proper); fundic mucosa occup)

666

The University Science Bulletin

, 1 m m ,

39

* r

l

;

'*

«

• -
-

. > -

40

Fig. 3(). Photograph showing cardiac vestibule of Natalus stramineus. Note the narrow
wall of the lesser curvature in contrast to the greater curvature.

Fig. 40. Large cardiac glands at the gastroesophageal junction of Natalus stramineus.

Studies of Stomachs of Selected American Bats 667

most of internal surface area of stomach (as in other bats) but unusual in
being most highly developed within fundic caecum; pylo-fundic transi-
tion zone extensive as zymogenic cells limited in distribution; pyloric
mucosa well represented and more extensive in distribution than most
species examined.

Glands of Brunner relatively sparse, present in small noncontinuous
clumps in proximal 3 mm of duodenum and superior surface of pyloric
sphincter.

Cardiac glands. — Small band of highly convoluted tubular glands ad-
jacent to, and beneath, lower limit of stratified squamous epithelium at
gastroesophageal junction; mostly superior to muscularis mucosae but
found also between the two layers of this tissue in uppermost area of
distribution under esophageal lining. Glands extremely large (Fig. 40),
two to three times breadth of adjacent gastric glands; lumen extremely
broad, superior-most glands appear to empty through pits formed of modi-
fied stratified squamous epithelium; glands composed of single cell type
indistinguishable in morphology and staining reactivity (refer to section
on histochemistry of cardiac glands) from cells of glands of Brunner;
cardial cells large and with extensive mucous border; nuclei small, spherical,
juxtaposed to basement membranes; cytoplasm agranular and stains ex-
tremely pale pink with eosin Y; connective tissue not abundant but more
so than that observed between glands of fundus; transitional glands be-
tween cardiac and fundic mucosa difficult to identify, at best an extremely
narrow zone characterized by occasional parietal cell in base of gland.
Depth of cardiac glands 100.

Fundus. — Extensive variation in depth and complements of cellular
elements; true fundic glands best developed within fundic caecum and
along lesser curvature near gastroesophageal junction; extensive variation
in depth with deepest glands along (1) upper wall of fundic caecum, (2)
within cardiac vestibule, and (3) at apexes of rugal folds; fundic mucosa
slightly reduced in depth at extreme apex of fundic caecum with marked
reduction in furrows between adjacent fundus 95 to 230.

Zymogenic cells in highest concentration within, and in areas immedi-
ately adjacent to fundic caecum; occupy lower 50% of gland in most
areas, some areas of fundus with chief cells in lower 20% of gland (through-
out much of greater curvature); chief cells gradually reduced in number
lateral to plane of gastrosophageal junction in direction of pyloru
pyramidal to trapezoidal or polyhedral, with large, basal, spherica
basal region of gland including chief cells bulbous, parietal cells
mon among chief cells. Argentaffin cells positively identified
among basal-most cells; morphology essentially that as describee

668 The University Science Bulletin

field (1953) for argentaffin cells of human small intestine; cells long and
narrow with nucleus displaced medially; dense peripheral granulation.

Parietal cells abundant throughout most of fundus and transitional
area; notably abundant in upper portion of glands resulting in extremely
shallow gastric pits; cells circular to ovoid in lower areas of gland, becoming
smaller and more rounded in upper portions of tubules; parietal cells re-
duced substantially in frequency throughout apical arc of fundic caecum
accounting for slight reduction in glandular height.

Mucous neck cells typically small with unusually large nuclei; occur
in small, tightly compressed clumps wedged among adjacent parietal cells;
little cytoplasm. Gastric pits extremely shallow throughout all of fundus
as well as transitional areas; pits rarely more than fifth total length of
gland, even in fundic caecum; parietal cells occasionally found just be-
neath surface epithelium suggesting essential absence of gastric pit and
presence of simple "pore"; cells lining pits having typically large nuclei of
lower mucous neck cells, nuclei becoming slightly laterally flattened toward
surface of mucosa. All mucous cell components of glands of fundic caecum
have more spherical nuclei than found in remainder of fundus. Surface
epithelium thick; nuclei large, basal, club-shaped; cells in single layer.
Connective tissue beneath glands nearly absent, glands often resting on
mucularis mucosae.

Pylo-fundic transition zone. — Region extensive owing to relatively lim-
ited distribution of zymogenic cells as discussed previously; glands excep-
tionally narrow and long, parietal cells abundant throughout most of length;
parietal cells sometimes in continuous column within glands and in sub-
stantially greater numbers than mucous neck cells; loss of zymogenic cells
at beginning of zone gradual; loss of parietal cells at terminus of area also
gradual with one or a few cells persisting in proximal-most pyloric glands;
gastric pits as shallow as those of fundus proper; surface epithelium un-
changed over that of fundus; little or no progressive decrease in depth
except for slight decrease at point where parietal cells finally disappear;
depth ISO to 220.

Pyloric glands. — Zone of pyloric glands relatively extensive in comparison
to total internal surface area of stomach, best developed on lesser curvature;
glands generally narrow at origin of distribution, increasing in breadth toward
termination at onset of glands of Brunner in duodenum. Glands composed
of single cell type in single layer; nuclei large and spherical throughout.
Gastric pits deep to base of gland (as characteristic of pyloric glands, mu-
cous border of glands narrow resulting in broad glandular lumen; cells
considerably larger than mucous neck cells of fundic glands; surface epi-
thelium deeper than that of fundus owing to much larger epithelial cells;
glands terminate at beginning of Brunner's glands on distal surface of

Studies of Stomachs of Selected American Bats 669

pyloric sphincter. Lamina propria more abundant at base of, and between,
pyloric glands than in fundic mucosa. Depth of pyloric mucosa 200 to 230.

Glands of Brunner. — Occur in discontinuous patches beneath muscularis
mucosae in proximal 3 mm of duodenum; ring of glands occupy entire distal
surface of pyloric sphincter, decreasing gradually in abundance to terminate
0.5 mm from sphincter; other smaller patches of three to six glands occur
distal to main group; proximal-most glands extremely large with spherical
and peripheral nuclei and extensive cytoplasmic complement; more distal
isolated clumps considerably smaller and nuclei always small and markedly
flattened, lumen much narrower. Cells and glands indistinguishable from
previously described "cardiac" glands. Brunner's glands clearly different in
both distribution and morphology from other bats examined.

Remarks. — The stomach of Natalus stramineus is unique among bats
examined (and notably similar to stomachs of soricid insectivores studied
by Myrcha, 1967) in possessing a terminal pyloric section oriented nearly
directly laterally with little recurvature craniad. Also, N. stramineus pos-
sesses the largest and most distinctive cardiac vestibule among carnivorous
bats studied.

The cardiac gland zone, although typically narrow, is distinctive in
several respects. These glands are highly convoluted and extremely broad,
a combination of features observed only in Natalus. Several tubules were
observed to open through pores within the squamous epithelium at the
gastroesophageal junction; thus, the cardiac glands lie partially beneath
esophageal epithelium. Additionally, tubules of cardiac glands in cross-
section are nearly indistinguishable from the glands of Brunner of the
same bat. Bensley (1905) and others have noted unusual similarity between
cardiac glands and Brunner's glands in several species of mammals, although
the present case appears to be the first reported for bats. In all other species
examined, the two glandular types were clearly distinctive in morphology
within each bat.

The fundic mucosa resembles most closely, among bats examined, that
of Plecotus in distributions and morphology of cellular components, as well
as in pattern of mucosal reduction in the fundic caecum. Natalus and
Plecotus differ, however, in distribution of different mucosal elements; the
pyloric zone is more extensive in Natalus, shifting the transitional area
medially, thus reducing the extent of the fundic mucosa relative to that
observed in Plecotus. Additional specimens of Natalus must be studied in
order to verify the tentative conclusions presented here.

Family VESPERTILIONIDAE
Plecotus townsendu' Cooper, 1837
Specimens examined (3).— Three specimens of Plecotus townsen
lescens from South Dakota.

670

The University Science Bulletin

2mm

Fig. 41. Mid-longitudinal representation of the stomach of Plecotus townsendii. For ex-
planation of symbols, see Figure 3.

Overall dimensions. — Greatest length 8.9-9.5; greatest breadth 3.1-3.8
(inclusive of cardiac vestibule); gastroesophageal junction to pyloric sphinc-
ter 2.5-2.7; gastroesophageal junction to apex of fundic caecum 4.6-5.1;
breadth pylorus at sphincter 2.0-2.1.

General morphology. — Stomach saccular and reniform (when empty),
notably symmetrical with cardiac vestibule at midpoint of lesser curvature
(Fig. 41); caecum extensive but unpouched, occupying nearly half length
of stomach; apex of caecum relatively blunt and directed antero-laterad,
upper limit of caecum lying approximately in same plane as pyloric sphinc-
ter and posterior limit of cardiac vestibule; caecum unusual in being long,
rather pointed, directed somewhat craniad, and extending a distance equal
to, or greater than, tubular stomach distal to cardiac junction; cardiac
vestibule moderate in height, expanding gradually into lesser curvature;
distance from esophageal junction to apex of fundic caecum greater than
that to pyloric sphincter; junction having stratified squamous epithelium
of esophagus abrupt, no suggestion of active muscular cardiac sphincter;
incisura cardiaca absent (esophagus entering stomach perpendicularly).
Terminal stomach distal to esophageal junction relatively short and broad;
musculature of pyloric sphincter (Fig. 42) relatively uniform in mass
throughout, but slightly more massive on greater than on lesser curvature,
thicker along luminal edge than elsewhere; musculature of sphincter com-
posed exclusively of inner circular layer plus thin sheets of muscularis

Studies of Stomachs of Selected American Bats 671

mucosae that line sphincter mass; notable external constriction circum-
scribing pyloric sphincter; glands of Brunner moderate in abundance and
restricted to distal surface of pyloric valve musculature and proximal-most
2 mm of duodenum. Esophageal junction in a plane distinctly anterior to
that of pyloric junction owing to moderately well-developed cardiac vesti-
bule (Fig. 41); cardiac vestibule and pyloric orifices directed in approxi-
mately same direction, terminal stomach at nearly right angle to plane
of remaining stomach; stomach generalized in gross morphology with
little apparent specialization of structure.

Musculature. — Muscle layers generally moderate in depth when com-
pared to other bats examined; total muscularis externa slightly thicker
within lesser than greater curvature, circular layer notably thickened in
terminal half of pyloric tube (especially thick in greater curvature) ; circular
muscle layer always substantially thicker than longitudinal layer except
in extreme apical tip of fundic caecum where layers approximately equal in
depth; ganglia of Auerbach's plexus rare; extensive bundling of circular
muscle occurs throughout proximal two-thirds of greater curvature; no
suggestion of a muscular cardiac sphincter; depth longitudinal layer 15 to
25, circular layer 40 (caecum) to 160 (pyloric tube), muscularis mucosae
less than 10 to 15.

Organization of gastric mucosa. — Mucosa clearly gathered longitudinally
throughout sides of stomach (dorsal and ventral surfaces of organ in situ) ;
rugal folds exceptionally numerous, long, narrow, and notably high; folds
gathered uniformly longitudinally throughout all of terminal portion; folds
clearly oriented transversely at apex of fundic caecum (Fig. 43) and within
greater curvature up to onset of terminal stomach, and occurring as deep
"pyramids" of mucosa lying directly along arc of greater curvature and
wall of fundic caecum; folds of cardiac vestibule directed longitudinally;
folds greatest in breadth in pyloric tube, decreasing gradually in breadth
toward apex of caecum; folds of fairly uniform depth throughout sides,
zig-zag orientation in lower half of caecum (Fig. 44), notably deeper along
arc of greater curvature; slight increase in depth of folds in lesser curvature;
substantial reduction in gastric mucosa only in apex of fundic caecum where
folds are short, narrow, and relatively sparsely distributed; folds along arc
of greater curvature in unusual configurations (extremely long, narrow
"fingers" often expanded at their apexes) (Fig. 43).

Well-defined cardiac, fundic, and pyloric glands present; cardiac zone
extremely narrow (one or two glands in breadth); pyloric zone rel;
narrow; fundo-pyloric transition area relatively broad as transition to pylon
gland mucosa is gradual; fundic mucosa occurring throughout
stomach and cardiac vestibule.

Cardiac glands.— Junction with stratified squamous epithelium

672

The University Science Bulletin

,0,5mm ,

• -

'"f-7

'43

Fig. 42. Pylo-duodenal junction of Plecotus townsendii.
Fig. 43. Mid-longitudinal section of stomach of Plecotus townsendii showing short, broad,
rugal folds of greater curvature.

Studies of Stomachs of Selected American Bats 673

glands long and narrow, with extremely limited distribution, composed of
single cell type; mucous cells of somewhat variable morphology throughout
gland as basal cells of lower third of tubule small, compact, rectangular,
more numerous than upper cells and with nuclei peripheral and spherical;
cells of upper two-thirds with progressively more laterally flattened nuclei
and cells more elongate than in basal elements; gastric pits always broad
and increase in breadth toward luminal surface; glands increase slightly
in depth from cardiac junction to transitional fundic zone. Transitional
zone from cardiac to fundic mucosa narrow with infrequent parietal cells
in lower half; gastric pits relatively deep until onset of true fundic glands.
Depth of cardiac glands 200 to 300, breadth of cardiac zone 60 to 100.

Fundus. — Glands relatively short and broad throughout most of fundus,
long and narrow only on lesser curvature within (and lateral to) cardiac
vestibule; fundic mucosa best developed on lesser curvature but notably
reduced within extreme apex of fundic caecum; all cellular components
large and relatively few in number. Depth of fundic glands 80 to 180.

Basal zymogenic cells large, pyramidal with spherical nuclei; may
occupy up to 60% length of gland but usually about 40%; cells rather
uniform in size throughout stomach and reduced in size and number only
in apex of elastic stomach; chief cells form basal "bulbous" expansion of
gastric gland.

Parietal cells relatively large and usually spherical (but occasionally
ovoid) with large, spherical nuclei centrally located in cell; greatest density
of cells in medial half of gland occurring extremely infrequently among
chief cells and most often in clumps of three or four but occasionally as
many as 20; marked increase in cell size from bottom to top of distribu-
tion (uppermost cells extremely large), whereas nuclear size remains
constant; only marked decrease in numbers in terminal half of fundic
caecum (with corresponding increase in cell size), slight increase in rela-
tive numbers in pylo-fundic transition area; intercellular caniculi pro-
nounced.

Mucous neck cells interspersed singly or in small clumps among parietal
cells and most common just above chief cells (where parietal cells less
numerous) and, of course, in necks of gastric pits; little if any size variation
within gland; nuclei spherical, slightly smaller than those of parietal cells.

Argentaffin cells extremely rare in base among chief cells; small and
with peripherally located granulation but otherwise clear cytoplasm; not
observed elsewhere in stomach.

Gastric pits relatively deep; mucous pit cells large with spherical
which become slightly flattened laterally in upper portion ol pii
usually occupy 40-50% length of gland and occasionally three-foi
length. Surface epithelium thick; nuclei extremely large and ovoid, often
"club-shaped."

674

The University Science Bulletin

l

0,1mm

m?'„ £

Fig. 44. A longitudinal section showing mucosal folds (rugae) in stomach of Plecotus
townsendii.

Fig. 45. Pyloric glands of Plecotus townsendii.

Studies of Stomachs of Selected American Bats 675

Pylo-fundic transition zone. — Transition to pyloric glands gradual and
zone relatively broad; abrupt loss of zymogenic cells, whereas parietal cells
reduced in numbers only gradually, those remaining in terminal portion
of transition large; mucous cells increase in frequency as parietal cells
reduced; no change in glandular depth over that of fundus; transition zone
occupies approximately angular portion of stomach; depth 120 to 170.

Pyloric glands. — Extremely broad and massive (Fig. 45), each gland
expanded into giant bulb at base; lamina propria between adjacent glands
substantially better developed than in fundic glands with smooth muscle
fibers extending between glands within lower half of layer; extensive vari-
ation in nuclear shape between glands, nuclei spherical, nuclear structure
usually consistent within a given gland; single cell type present with
exception of occasional argentaffin cell in extreme base; surface epithelium
identical to that of fundus. Glands near pyloric sphincter with narrow,
sharply-compressed cells resulting in extremely broad openings of gastric
pits into lumen, hence surface epithelium reduced to narrow "clumps."
Glands terminate abruptly at distal edge of pyloric sphincter; pyloric glands
in relatively narrow zone substantially distal to pyloric bend. Depth of
pyloric mucosa 100 to 230.

Glands of Britnner. — Moderate in abundance and in single solid mass on
entire distal surface of pyloric sphincter and proximal 1.5 mm of duodenum;
highly convoluted, large with broad lumen; cells long, narrow; nuclei ex-
tremely small, markedly flattened, and always positioned against basal
surface (peripheral); cellular morphology uniform throughout gland; glands
always submucosal and restricted to duodenum; muscle fibers of muscularis
mucosae prominent between glands; open luminally between crypts of
Liberkiihn.

Remarks— The stomach of Plecotus townsendii is generalized in gross
morphology, being symmetrical in shape, and having a simple fundic
caecum and gradual curvature throughout. These features along with dis-
tribution of zones of gastric mucosa are similar to those reported for several
species of Insectivora by Myrcha (1967), particularly several species of Old
World talpids and erinaceids. The pyloric and pylo-fundic transition zones
are well defined and approximately equal in length. The contributing
muscle layers in the stomach wall are relatively thick with increasing thick-
ness in both upper and lower walls of the pyloric tube toward the pyloric
sphincter.

Rouk (196S) described the stomachs of Myotis velijer and Antrozous
pallidas as being of the "simple type," and his figure of the stom
Myotis velijer bears striking similarity to that of P. townst
slightly in orientation of the cardiac vestibule and in distribution of p>
fundic transitional mucosa. All three genera {Antrozous, Myoti.

676 The University Science Bulletin

have more or less symmetrical stomachs with a fundic caecum hooked
craniad, smooth overall curvature, a gradual (rather smooth) pyloric bend,
a moderately thick pyloric sphincter, and a short, narrow cardiac vestibule.
A photograph published by Fischer (1909:124) of a "Situtspraparat der
Baucheingegeweide von Plecotus aunt us" showed the stomach of this
species to be virtually identical topographically to that described here for
P. townsendii.

Vizonyx vwesi (Menegaux, 1901)

Specimens examined (1). — One specimen from Isla Partida, Gulf of
California (KU 113998).

Overall dimensions. — Greatest length 10.0; greatest breadth 7.1; gastro-
esophageal junction to pyloric sphincter l.S; gastroesophageal junction to
apex of fundic caecum 6.0; breadth pylorus at sphincter 2.6.

General morphology. — Stomach large and generally tubular but ex-
tremely rounded through fundic caecum and greater curvature (Fig. 46);
expanded substantially in antero-posterior direction; dorsal surface convex;
pyloric tube short.

Cardiac vestibule short and, although cardiac glands comparatively
abundant within vestibule, thick-walled and extremely narrow in relation
to breadth of pylorus; vestibule increasing gradually and slightly in breadth
from gastroesophageal junction to incorporation into lesser curvature, situ-
ated approximately midway along lesser curvature; esophagus entering
stomach at slight angle to lesser curvature, creating incisura between vesti-
bule and caecum; incisura cardiaca short and broad with muscle layer
prominently thickened locally and longitudinal layer unchanged; gastro-
esophageal junction abrupt with all muscular complements continuous
between esophagus and vestibule but with slight local thickening of muscu-
laris mucosae and lamina propria mucosae at junction; no specific evidence
of an anatomical sphincter. Gastric orifices in relatively close proximity to
each other and generally equivalent in direction anteriorly, although
vestibule slightly acute to pyloric tube.

Fundic caecum extremely large and spacious but not truly saccular,
gently rounded, nearly spherical; musculature reduced in depth within
extreme apical portion of caecum only with no external or internal features
suggesting anatomical demarcation from remaining stomach.

Tubular portion at terminal point distal to plane of vestibule more than
twice breadth of gastroesophageal junction, directed anteriorly to the right;
pyloric curvature gradual and continuous with musculature prominently
thickened throughout tubular portion (which tapers little to pyloric sphinc-
ter). Musculature of pyloric sphincter short and massive, appearing more
or less uniform in depth throughout; sphincter a simple, rounded projection
of circular muscle from wall of pylorus, valve displaced more craniad

Studies of Stomachs of Selected American- Bats

677

2mm

Fig. 46. Mid-longitudinal representation of the stomach of Pizonyx vivesi. For explanation
of symbols, see Figure 3.

on greater than on lesser curvature; mucosa lining sphincter slightly re-
duced in depth over that of remaining pylorus; deep but narrow external
constriction at pyloric sphincter; glands of Brunner not abundant and
restricted to duodenum; sulcus intermedius absent, sulcus angularis present
but not pronounced.

Pattern of depth of muscular wall similar to that observed in Plecotus;
gastric mucosa deep throughout sides and apexes of rugal folds with pat-
tern of local reductions of mucosa similar to that found in Plecotus; lamina
propria mucosae considerably more prominent than in Plecotus, loose lamina
with abundant collagenous fibers best developed throughout greater curva-
ture and within rugae (including fundic caecum to apex), much less pro-
nounced through lesser arc; elastic fibers generally infrequent (including
fundic caecum) and lamina propria between mucosa and muscularis mu-
cosae typically infrequent throughout.

Musculature. — Longitudinal layer of uniform breadth throughoi
of greater curvature and through fundic caecum to just lateral of :
cardiaca, reduced to thin remnant sheet through incisura cardiaca, ca:

678 The University Science Bulletin

vestibule, and all of remaining lesser curvature to pyloric sphincter. Vari-
ation in depth of circular layer within stomach basically similar to pattern
in Plecottis; circular layer differing from that in Plecotits in being approxi-
mately equal in depth within much of greater and lesser curvatures (con-
sistently deeper in lesser as compared to greater curvature in Pelcotus) with
proportionally greater thickening within pyloric tube, gathering of circular
fibers into transverse bundles more frequent and bundles larger (although
restricted to greater curvature and apex of fundic caecum as in Plecotits),
and in having more extensive inter-bundle complements of lamina propria.
Muscularis mucosae uniform in depth throughout stomach wall but notably
deeper within rugae. Depth longitudinal layer 15 to 45, circular layer 35
to 250, muscularis mucosae less than 10 to 25.

Organization of gastric mucosae. — Pattern of gathering of mucosal folds
notably similar to that of Plecotits; folds oriented strictly longitudinally
throughout tubular portion, longitudinally directed through dorsal and
ventral walls of fundic caecum with pronounced zig-zag orientation within
lower one-half of caecum (as in Plecotits) ; short, transversely directed folds
originating along greater curvature not nearly so abundant as in Plecotits,
but transverse folds along lesser curvature of fundic caecum as in Plecotits.
Rugae numerous, nowhere locally reduced in number; rugal folds com-
paratively broader and deeper than those of Plecotus, probably as function
of increased stomach size.

Zones of gastric mucosa well defined with relatively few transitional
elements from one major glandular type to another; region of cardiac
glands extensive relative to all other bats examined, occupying major
portion of short cardiac vestibule; fundic mucosa locally variable in mor-
phology within extensive fundic stomach, lining all of remaining stomach
except pyloric tube; transition from fundic to pyloric mucosa relatively
abrupt, with zone narrow as compared to Plecotits and two other vesper-
tilionids previously examined (Rouk, 1968), confined to narrow area within
pyloric curvature; pyloric glands extremely large and prominent, occupying
most of tubular stomach distal to pyloric bend. Mucosa consistently deep
throughout stomach, reaching greatest depth within pyloric zone; pro-
nounced reduction in depth of glands within furrows of fundic zone only
(similar to condition of Plecotits), pattern of reduction of mucosal depth
on sides and surfaces of rugae within apical arc of fundic caecum similar
to pattern in Plecotits.

Cardiac glands. — Glands occurring in broad, deep zone 1 mm in length
at gastroesophageal junction (Fig. 46); glands extremely broad, being three
times breadth of adjacent fundic glands; cells within basal fourth small,
elongate and tightly packed within basal arc; nuclei usually peripheral, ex-
tremely small, spherical, mucous border narrow; cells within upper three-

Studies of Stomachs of Selected American Bats (>7{)

fourths becoming progressively more elongate (increase in mucous border)
toward surface epithelium (with nuclei unchanged in size although modi-
fied to prominently flattened discs located in juxtaposition to basement
membrane) ; lumen always extremely broad, especially within bulbous
basal expansion as mucous border of basal cellular framework narrow;
glands relatively uniform in breadth superior to basal portion; transitional
zone to fundic mucosa a narow belt of two or three cardiac-type glands
with infrequent basal parietal cells; areolar connective tissue extensive basal
to (and between) glands with large numbers of inclusive collagenous
fibers (restricted to cardiac zone) ; depth increasing slightly from gastro-
esophageal junction to junction with fundic mucosa; depth of cardiac
mucosa 250 to 300.

Fundus. — Extensive variation in morphology and depth of glands; best
developed in both depth and complements of cellular elements within lesser
curvature; significantly reduced in depth between rugae of greater curva-
ture but only slightly between rugae within all of fundic caecum; reduction
in depth usually involving reduced numbers of cells, rarely reduction in
dimensions of cellular components; glands usually long and narrow, rela-
tively straight and tubular, distinctively expanded basally, thereafter uni-
form in breadth to surface epithelium; gastric pits also highly variable in
depth and breadth. Depth of fundic mucosa 150 to 280 (lesser curvature).

Zymogenic cells occupying lower 40% of gland throughout lesser curva-
ture (especially abundant within lesser curvature) and most of upper
two-thirds of tubular portion of fundus; notably reduced in number
throughout glands of lower wall of fundic caecum and midregion of greater
curvature, increasing again in number in greater curvature within tubular
portion; cells large with small, ovoid nuclei located variably within basal
third of cell; areas in which zymogenic cells reduced usually having cells
only within lower 10-15% of gland (argentaffin-type cells not observed
but assumed present).

Parietal cells most abundant in lesser curvature and cardiac vestibule;
most common within upper 60% of gland, but scattered and less frequent
among chief cells and moderately abundant through apical arc of fundic
caecum (within glands on rugae) ; reduced in abundance and density on
dorsal and ventral surfaces in upper portion of gland throughout
body of stomach (excluding tubular portion); reduction corresponds
1) increase in depth of gastric pits, 2) reduction of overall glandula
and 3) marked drop in number of chief cells; marked lower to upper
size gradient in parietal cells, ranging from small among basal ch
to those nearly breadth of tubule just beneath necks of gastric pit:
cells averaging large and spherical with relatively small, usual
nuclei; intercellular caniculi not observed although assui

580 The University Science Bulletin

Mucous neck cells infrequent among closely packed and numerous
parietal cells within upper portions of most fundic glands, but more nu-
merous in midregion where parietal and chief cells less abundant; mucous
neck cells identical in morphology to those of Plecotus except nuclei small
(smaller than those of adjacent parietal cells). Gastric pits varying from
10-25% length of gland throughout lesser curvature to up to 75% within
glands of greater arc; increased depth of pits partially attributable to
absolute increase in pit depth, as well as relative decrease in glandular depth
as compared to lesser curvature; pit cells identical in structure to cells of
cardiac and pyloric glands of this species, lumen broad. Surface epithelium
approximately two-thirds depth of gastric pit epithelium, cells in tightly
packed clumps at interim between adjacent gastric pits; clumps not ex-
tensive in mass owing to great breadth of gastric pits; basal lamina propria
sparse and with less abundant intertubule matrix than observed in either
cardiac or pyloric mucosa.

Pylo-fundic transition zone.— Narrow zone (Fig. 46) illustrating rela-
tively abrupt transition to pyloric mucosa; reduction and subsequent loss
of parietal cells fairly abrupt with parietal cells persisting in large numbers
throughout nearly all of transitional area; loss of zymogenic cells even
more abrupt (within breadth of two or three glands) at initial edge of
transitional area; increase in depth of gastric pits gradual from about 30 to
50% length of gland within transitional area, then abrupt increase to base
at margin with pyloric mucosa; lamina propria mucosae prominent basally
and between adjacent glands and confined in abundance to transitional belt,
uncommon in adjacent fundic mucosa; depth 210 to 310.

Pyloric glands. — Occurring within zone equivalent to that of Plecotus
but notably different in morphology; glands usually large in both length
and breadth (Fig. 47), similar to cardiac glands in dimension although
slightly broader; cells within basal 10% similar to mucous neck cells, small
and with spherical, centrally located nuclei and narrow inner mucous
border; remaining 90% of length composed of giant, elongate, columnar
cells with small, disc-like, crescent-shaped nuclei and a vast mucous border
similar to upper cells of cardiac glands; glands distinctively bulbular at
base, narrowed above basal portion, increasing in breadth only slightly to
lumenal surface; lumen extremely broad throughout, broadest within basal
region; pyloric mucosa restricted aborally to proximal surface of pyloric
sphincter, not extending into duodenal lumen; lamina propria especially
pronounced between glands, moderate basally. Pyloric, along with cardiac,
glands representing most massive single mucosal-type glands observed in
entire study. Depth of pyloric glands 250 to 340.

Glands of Brunner. — Not abundantly distributed, occurring in narrow
band within proximal 1.1 mm of duodenum and along distal surface of

Studies of Stomachs of Selected American Bats

68]

"Sri

V.

<

'W-«'

X

0.1mm

V

fckf*

X

V

■

' /

47

Fig. 47. Pyloric glands of Pizonyx vivesi.

sphincter; always submucosal and restricted to duodenum; typically con-
voluted, uniform in breadth and relatively narrow; cross-sectional structure
similar to that found in other bats examined, differing from Plecotus only
in location and structure of nuclei; nuclei peripheral in cell (but not basal),
compressed ovals (usually markedly compressed in Plecotus); tubules more
heavily enclosed by surrounding smooth muscle fibers originating from
muscularis mucosae than in other bats examined; exit from lumen narrow.

Remarks. — The stomach of Pizonyx appears to be slightly altered from
the basic insectivorous type. The general outline or configuration is similar
to that of Plecotus and Rhynchonycteris, although the stomach oi Pizonyx is
considerably more rounded and has a shorter and broader terminal portion
than in either of the other two latter bats. The spaciousness of the fundic
portion suggests a modification to assimilate large quantities of food at
a single feeding. A noteworthy feature is the extensive distribution oi
cardiac glands at the gastroesophageal junction. These glands are
relatively straight tubular, and strikingly similar in morpholog;
pyloric glands.

The structure and local variation in depth of the tunica mus
externa closely resembles that observed in Plecotus, although compl

(,S2 The University Science Bulletin

of loose lamina propria within the muscularis externa and superior to the
muscularis mucosae are substantially greater in Pizonyx. Although the
valves of the pyloric sphincter of Plecotus are longer than those of the
single specimen of Pizonyx examined, the asymmetrical orientation of the
flaps is identical in the two species.

Both the pylo-fundic transitional glands and the glands of Brunner are
distinctively reduced in distribution as compared to Plecotus. Perhaps the
narrow transitional zone is a consequence of advancement distally of basal
zymogenic cells.

The only feature in which Pizonyx clearly resembles Noctilio leporinus,
another fish-catching bat (of which the stomach was observed macro-
scopically only), is in the short, broad, gently arched terminal stomach, and
this feature is shared also with Pteronotus parnellii.

Family MOLOSSIDAE
Molossus ater E. GeofTroy-St. Hilaire, 1805

Specimens examined (2). — Two specimens from Nicaragua of Molossus
ater nigricans (KU 106285, 106287).

Overall dimensions. — Greatest length 12.1-13.7; greatest breadth 4.1-4.5;
gastroesophageal junction to apex of fundic caecum 4.2-4.4; breadth of
pylorus at sphincter 1.3-1.4.

General morphology. — Stomach tubular with relatively short, narrowly
rounded fundic caecum; generally slight anterior flexure of the stomach
resulting in slight convexity of greater curvature; pyloric and cardiac orifices
oriented so that pyloric sphincter substantially inferior to cardiac sphincter
in stomach in situ; terminal portion tapers gradually to pyloric sphincter.
Incisura cardiaca present but not well developed, esophagus entering stomach
at angle; stratified squamous epithelium extending posteriorly to lesser
curvature.

Stomach extensive distal to cardiac junction (see Fig. 48) ; marked pyloric
bend just proximal to pyloric sphincter, endpiece sharply recurved craniad;
sulcus angularis present as constriction within greater curvature immedi-
ately proximal to pyloric bend, corresponding constriction in lesser curva-
ture less pronounced; pyloric constriction at sphincter pronounced and sym-
metrical. Pyloric sphincter asymmetrical, mass of valve considerably larger
on greater than on lesser curvature but sphincter not unusually large overall;
valve funnel-shaped and directed into duodenum; contributing muscle in-
cluding circular layer and muscularis mucosae.

Musculature well developed throughout, except in fundic caecum; ex-
tensive grouping oi fibers into bundles within circular layer; unusual and
prominent "notch" in muscle wall of greater curvature; substantial reduc-
tion in depth of wall in fundic caecum. Fundic caecum with apex directed

Studies of Stomachs of Selected American Rats

683

2mm

Fie. 48. Mid-longitudinal representation of the stomach of Molossus ater. For explanation
of symbols, see Figure 3.

slightly postero-laterad in contrast to the antero-laterad orientation in ves-
pertilionids resulting in suggestion of general symmetry to stomach; caecum
tapering uniformly laterally, never expanded on dorsal surface; nonmuscular
submucosa extremely rare (with exception of prominent loose lamina
propria overlying midventral "notch" within greater curvature), elastic
fibers extremely sparse in most places and nearly absent within fundic
caecum.

Glands of Brunner well developed at, and directly proximal to pyloric
sphincter, occurring in single mass at sphincter and as narrow layer within
pyloric stomach; terminal end of bile duct dilated at junction with duo-
denum, resulting in unusual ampule containing duodenal mucosa and
submucosal glands of Brunner; stomach in general different in external
morphology from those of other carnivorous species examined (see Fig. 48).

Musculature. — Circular layer notably thickened at incisura cardiaca
(thickened locally and confined to immediate region of gastroesophageal
junction with muscle fibers gathered into long narrow sheets enclosed by
septa radiating inwardly from depression at incisura); longitudinal layer noi
correspondingly thickened at incisura; opposing musculature adjacent to
incisura slightly thickened and gathered into bundles locally although ana-
tomical cardiac sphincter probably not present.

Circular layer generally thinner in lesser than in greater curva
longitudinal layer always extremely thin, except in wall ol great*
ture; circular layer irregular in depth, depth moderate at pyloric spf
on lesser curvature, decreasing sharply at pyloric bend, narrow

684

The University Science Bulletin

0.5 mm"

WJt*

?\A"

.•»•■»

#Sf Trt' > ***" - -ar yaw

^t.

*# *

>•

-^-**/. DP.

A*J*

■ft * >- 9

f;; IK*- -• it ^^v" 50

Fig. 49. Midventral cleft in greater curvature of stomach of Molossus ater.
Fig. 50. Pyloric sphincter of Molossus ater showing Brunncr's gland cells. (BG) within the
stomach.

Studies of Stomachs of Selected American Bats 685

lesser curvature (excepting cardiac junction) and a thin sheet with some
bundling through fundic caecum, increasing to moderate depth through
greater curvature to a point proximal to sphincter where layer again nar-
rows; extensive bundling of circular layer occurring throughout most of
greater curvature from apex of fundic caecum to midventral "notch."
Large, distinctive depression in musculature of stomach wall of greater
curvature in frontal half, marked by unusually well-developed submucos;'
beneath muscularis mucosae and atypically deep fundic mucosa along fold;
midventral notch clearly not an artifact as structural disfiguration of muscle
layer present (Fig. 49). Musculature generally irregular in depth through-
out stomach, especially through fundic caecum and greater curvature, par-
tially owing to extensive gathering of circular fibers into irregularly shaped
bundles.

Muscularis mucosae typically double-layered, relatively well developed;
uniform in depth throughout stomach. Ganglia of Auerbach's plexus
abundant. Depth longitudinal layer 8 to 15, circular layer 15 to 250, muscu-
laris mucosae 15.

Organization of gastric mucosa. — Mucosa consistently gathered into
straight, parallel, longitudinal rows throughout most of stomach, with
slight zig-zag pattern in walls of fundic caecum lateral to plane of esopha-
geal junction; only major transverse gathering is in greater wall of fundic
caecum lateral to midventral notch. Extensive variation in depth of mu-
cosa, generally deeper along lesser than along greater curvature and in
tubular stomach (in contrast to caecum); also notable variation in depth
within fundic mucosa which is deepest along lesser curvature and along
musculature of midventral notch of greater curvature, but is shallow along
apex of caecum with extensive variation in depth on both dorsal and ventral
walls; marked reduction in number of folds occurring throughout fundic
caecum as well; mucosa shallower in furrows than on sides ot rugae
throughout stomach.

Cardiac glands present and confined to typically narrow zone at gastro-
esophageal junction; fundic mucosa extensive with marked variation in
morphology and depth from one area to another, extending distally to
midway between pyloric bend and cardiac junction on lesser curvature,
somewhat further on greater arc; extremely narrow transition zone from
cardia to fundus; pylo-fundic transition glands (Fig. 48) occupy area
between true fundus and pyloric bend, although somewhat deeper
pyloric tube on greater than on lesser curvature; transition zone approxi
mately equal in total length on greater and lesser surfaces; pyloric gl
in well-defined region and with unusual basal cell component
identical to glands of Brunner; pyloric glands not extending di
yond proximal surface of pyloric sphincter.

686

The University Science Bulletin

■ *■*"* . i ^^

< >"

j

>*" Ss/:-i&* ft t1

iff n sR

*

ii
0.1mm

*VM

•,

I

Fig. 51. Glands oi Brunner (BG) within the duodenum and pylorus of Molossus ater
(Periodic Acid SehifT Leucofuchsite Reaction).

Fig. 52. Transitional zone of Molossus ater. Note the infrequently scattered parietal cells

(PC).

Studies of Stomachs of Selected American Bats 687

Glands of Brunner well developed as submucosal glands on distal sur-
face of pyloric sphincter and proximal 2 mm of duodenum (Figs. 50-51),
occurring also in pyloric glands as mucosal elements, indistinguishable in
morphology from Brunner's glands within duodenum (Fig. 50), also occu-
pying dilated chamber at entrance of bile duct into duodenum; mucosa
generally shallow throughout stomach.

Fundic glands deepest in midregion between fundic caecum and tran-
sition zone; cardiac glands longer than fundic glands; fundic glands
abruptly decreasing in depth in comparison to adjacent cardiac glands;
pyloric mucosal membrane slightly deeper than most of fundus, perhaps
owing to underlying cells of Brunner's glands, and relatively uniform in
depth (deepest at bases of rugae and between adjacent folds).

Cardiac glands. — Glands contained within extremely narrow zone (100
to 200 wide) at gastroesophageal junction; composed of single cell type,
cells similar to those of surface epithelium of remaining stomach; glands
simple, tubular, of branched type; cells differing from those of fundus in
having consistently spherical nuclei throughout; cells small, long and
narrow, extremely compact within gland with small, spherical nuclei;
glands easily distinguishable from pyloric glands of this species on account
of nuclear morphology; glands longer than fundic and pyloric glands and
with extremely broad lumen; basal lamina propria scanty; depth of cardiac
glands 300 to 360.

Fundus. — Fundic region occupying all of saccular stomach and tubular
area up to approximately midway between cardiac junction and pyloric
bend. General pattern in distribution of fully developed fundic glands
that of a parallelogram with upper facies directed toward fundic caecum;
glands with well-developed chief cells extending further into fundic caecum
along lesser than along greater curvature (possibly a developmental phe-
nomenon due to flexion of the pyloric tube); glands relatively uniform in
cell distribution throughout fundic zone, except glands in apex of fundic
caecum which show marked structural reductions; fundic glands usually
moderate in length and narrow, extensively convoluted; glands expanded
greatly at lower end, less so in upper portion. Most fundic glands uni-
formly and consistently devisible into three zones, defined by types of cells
present; upper third composed of columnar epithelium lining deep gastric
pits, middle third composed of parietal cells and mucous neck cells, lower
third composed of chief and argentaffin cells; glands of middle and uppe
fundic caecum marked by substantial increase in length and breadt
gastric pits to 75% length of gland with lower fourth occupied by
infrequent parietal cells mixed with mucous neck cells and occa
chief cells at base; gradual reduction in number of chief ce
gradual reduction in number of chief cells from midregion of
apex of fundic caecum. Depth of fundic mucosa 140 to 260.

688 The University Science Bulletin

Zymogenic cells abundant basally in most of noncaecal fundic mucosa;
large, pyramidal with large and spherical nuclei; cells always greater in
length than breadth; argentaffin cells commonly observed among chief cells
at base and narrowly triangular with extremely large, spherical nuclei, and
peripheral argentafile granules.

Parietal cells large and almost always circular with moderately sized
spherical nuclei always centrally located; cells usually restricted in distri-
bution to medial third and never abundant; size of cells uniform through-
out distribution in given gland with little size increase from bottom to top
of zone; found in basal fourth of tubules in fundic caecum but rarely ever
observed among zymogenic cells.

Mucous neck cells infrequent beneath base of gastric pit and scattered
among parietal cells; nuclei larger than those of parietal cells, cytoplasm
minute, cells irregularly shaped. Gastric pits deep; cells lining pits large,
with nuclei becoming increasingly more laterally flattened toward surface
of mucosa; surface epithelium extremely thick, cells large with extremely
large nuclei that are only slightly compressed. Lamina propria always
sparse beneath glands but somewhat more extensive beneath muscularis
mucosae along greater curvature, owing to buildup of connective tissue
around bundles of circular muscle fibers.

Pylo-fundic transition zone (Fig. 52). — Broad zone occupying approxi-
mately same area on both curvatures, although distributions not exactly
comparable; loss of zymogenic cells gradual within fundus, zone of gastric
mucosa lacking chief cells and maintaining parietal cells extending slightly
distal to pyloric bend; loss of parietal cells in terminal area also gradual,
but more abrupt than termination of chief cells proximally; parietal and
mucous neck cells extending to base of glands, although parietal cells
less abundant in base than in middle third; gastric pits, surface epithelium,
and density of lamina propria equivalent to conditions observed within
fundic mucosa; depth 220.

Pyloric glands. — Glands occupy zone within most of stomach between
pyloric bend and pyloric sphincter; deepening of gastric pits gradual through
terminal portion of transition zone with parietal cells disappearing gradu-
ally; basic pyloric gland morphology similar to that observed in other bats,
but unusual in presence of apparent layer of glands of Brunner beneath
entire distribution of pyloric mucosa and hence contributing to the depth
of the mucosal membrane (see section on glands of Brunner) ; mucous
cells of pyloric glands large; nuclei extremely large, becoming progressively
more flattened laterally from bottom to top of gland; pyloric glands in-
creasing in breadth from base of gland as cells increase in size, lumen also
becoming progressively broader upwards; depth of pyloric glands irregular
but generally deeper on lesser than on greater curvature; surface epithelium

Studies of Stomachs of Selected American Bats 689

extremely thick, slightly thicker than that of fundus; depth of pyloric glands
120 to 200.

Glands of Brunner. — Glands well represented as a single mass in sub-
mucosal distribution along distal surface of pyloric sphincter and proximal
2 mm of duodenum; also found in terminal dilation of bile duct (also
submucosal) and directly beneath entire complement of pyloric mucosa
(usually above muscularis mucosae) as thin layer of glands reduced in size;
duodenal glands long and narrow, highly convoluted with small, pyramidal
cells; nuclei large, peripheral, ovoid to substantially flattened laterally; cyto-
plasm pale pink and diffusely granulated when stained with Harris'
haemotoxylin and eosin; glands always submucosal; those within mucosa of
pylorus approximately half diameter of duodenal complement (Fig. 51),
highly convoluted, both mucosal and submucosal (observed between two
layers of muscularis mucosae) although usually within mucosa; occurring
also at base of mucosal membrane between adjacent pyloric glands; cellular
morphology of pyloric complement identical to that in duodenum, lumen
narrow in both.

Remarks. — The tube-like stomach of Molossits presents a perplexing
array of features; most features represent conditions observed in other
insect-feeding bats. The stomach, however, although tubular, is less sym-
metrical in outline than is that of other carnivorous bats and has the elon-
gate terminal stomach usually observed in bats that are primarily or entirely
herbivorous. The pyloric sphincter is moderately developed, asymmetrical,
as in Plecotus.

The gastric mucosa is uniformly gathered longitudinally through the
simple, unpartitioned fundic caceum. The patterns of distribution of cellular
elements within the fundic mucosa and variations within the fundic caecum
are essentially as observed in Plecotus. The most noteworthy feature of
the gastric mucosa is an extremely thin layer of narrow, convoluted, tubular
glands which are histologically indistinguishable from the glands of Brunner
in this species and which lie immediately beneath the zone of pyloric mu-
cosa. These serous tubular elements are less conspicuous and presumably
less abundant than the corresponding layer of glands observed in Sturnira.
In one specimen of Molossits, the sub-pyloric glands appeared to be in dis-
continuous masses near the proximal surface of the pyloric sphincter, sug-
gesting less complete invasion into the stomach than observed in specimens
of Sturnira.

One unusual feature observed only in Molossits is a deep and promi
transverse cleft in the stomach wall that is located along the greater cu
ture. This depression is clearly not an artifact of fixation or
or a contractional wave within the stomach musculature, because
is accompanied internally by a distinctive break in the muscle

690 The University Science Bulletin

heavy local concentration of lamina propria, and slightly deeper gastric
mucosa than in adjacent areas. It is further of interest that Rouk (1968),
in his description of the stomach of a molossid, Tadarida brasiliensis, de-
scribed and illustrated a single depression midway along the greater curva-
ture that appears to be identical to that described here for Molossus.

Rouk's description of the stomach of T. brasiliensis suggests further
noteworthy similarities between that species and Molossus ater. The gen-
eral stomach outline is similar, although curvature of the stomach anteriorly
is greater in T. brasiliensis. The fundic caecum in both species tapers to a
narrow apex, and the orientation and position of the gastroesophageal
junction and structure of the pyloric sphincter appear to be identical in
the two. Neither species has a cardiac vestibule, and both have an elongate,
cranially reflected, terminal pyloric endpiece.

HISTOCHEMISTRY OF THE GASTRIC MUCOSA
AND DUODENAL JUNCTION

The results that follow are limited to a survey of basic types of reactions
for mucosubstances, and are to be regarded as necessarily preliminary to
more detailed studies. To my knowledge, the present study (excepting
the brief PAS-Alcian blue account of surface epithelial mucus of Rouk,
1968) is the first discussion of comparative mucopolysaccharide histochem-
istry of the gastric mucosa of bats. The following accounts are directed
to the identification and discussion of neutral and acid mucopolysaccharides
as demonstrated by toluidine blue O metachromasia, PAS, and Alcian blue
8GX 300. No attempt was made to identify specific chemical end-groups,
because the objective of the study was to examine and characterize mucin
specificity in species of bats relative to each other. More detailed discussion
of specific categories of mucin are to be found in the accounts of techniques
to follow, where a particular technique was most appropriate in mucin iden-
tification.

Periodic Acid Schiff Reaction

The PAS reaction is generally thought to indicate the presence of
neutral mucopolysaccharides, and, because most mucins are apparently
colored with this procedure, it also provides a general indication of overall
mucous activity. Mucins stain red or magenta with this procedure.

In any test for the presence of complex carbohydrate material, it is
necessary to account for the potential reactivity of lipids and simpler carbo-
hydrates. As a check for the presence of distinctive concentrations of gly-
cogen, sections of stomachs of each species were subjected to digestion by
salivary amylase for 10 minutes at room temperature, prior to oxidation
by periodic acid. In no case was glycogen, or other material presumably

Studies of Stomachs of Selected American Bats 691

removed by exposure to amylase, found to be a contributing factor to the
intensity of PAS reactivity in sites of production of mucin material.

Using PAS, staining cells of the glands of Brunner always was uniform
throughout the cytoplasm, with no apparent local areas of especially high
intensity. Staining of surface epithelium was restricted to a narrow border
of mucous droplets along the luminal surface (surface opposite to base),
except for slight additional staining of fine cytoplasmic granules just be-
neath the luminal mucous material in one species (Plecotus). Parietal cells
stained uniformly throughout the cytoplasm, but mucous neck cells stained
only within the cytoplasmic material between the nucleus and lumen of
the tubule (outer portion of cytoplasm). All cells within the cardiac and
pyloric glands, irrespective of location, stained only with the luminal portion
of the cytoplasm. The above statements apply equally well to those cells
that stained with Alcian blue, except for an apparently more complete
staining of the mucous neck cell cytoplasm with Alcian blue in the case
of Phyllostomits.

A uniformly strong reaction characterizes the surface epithelial mucus of
all species of bats with the exception of Desmodus, in which the reaction
is only slightly reduced in intensity in comparison to the other species
examined. An equivalent statement may be made concerning the gastric
foveolae of the fundic, pyloric, and cardiac glands of Desmodus, excluding
the lower portions of cardiac and pyloric tubules to be discussed later.
Staining was noticeably stronger in the foveolae within the fundic caecum
in contrast to the remainder of the fundus in two species, Glossophaga and
Carollia. Duodenal goblet cells of all species examined reacted with a uni-
formly high degree of intensity.

The glands of Brunner reacted relatively weakly in Rhynchonycteris,
Plecotus and Pizonyx, moderately in Artibeus and Sturnira, and strongly to
extremely strongly in the remaining species. Identical qualities of reactivity
in the glands of Brunner of the duodenum and the basal cell complexes
within the pyloric stomachs of Sturnira (Fig. 28) and Molossus (Fig. 51)
help to confirm their previous identification, on morphological grounds, as
glands of Brunner. Genera of the family Phyllostomatidae (excluding
Sturnira and Artibeus) and the single noctilionid consistently revealed the
greatest degree of reactivity in the glands of Brunner.

Deeper cellular elements of the fundic mucosa revealed highly varied
reactivity to PAS. Mucous neck cells among parietal cells react consistently
less than surface or foveolar epithelial cells, except for equivalent i
in the case of Noctilio. The frugivores (Carollia, Sturnira, and Artibeu
as a group, showed less reactivity within the mucous neck cells
parison to carnivorous species, except for extremel;
Desmodus and virtually no positive coloring at all in Plecotus i

692 The University Science Bulletin

Little mention has been made in the literature of PAS reactivity, positive
or negative, of gastric parietal cells. Leblond (1950) noted that the parietal
cells of albino rats stain "weakly" with a periodic acid-fuchsin-sulphurous
acid procedure, approximately equivalent to PAS, and suggested that
some mucin activity might be attributable to parietal cells. A clearly positive
staining in parietal cells, over and above that of several other gastric elements
that are clearly PAS positive, was observed in several species of bats. Only
a trace of reaction was observed in carnivorous bats, except for a weak
reaction in the blood-feeding Desmodus, and moderate staining in Molossits.
The presumably partially nectivorous and partially carnivorous Glossophaga
revealed moderately stained parietal cells, as did three frugivorous species.

Although Lillie (1965) recorded positive PAS reactivity in the zymogenic
granules within the salivary glands of man and several species of rodents,
and in the pancreatic zymogen granules of man, rabbits, and some mice
(but not in rats or guinea pigs), little or no information appears to be
available regarding such reactivity in the zymogenic granules of gastric
chief cells. Little or no staining was observed in gastric zymogenic cells of
the bats examined, with the exception of a consistently strong (+4) staining
of the zymogenic granules in two specimens of Noctilio. The significance
of this result is unknown.

As noted previously, the major portions of the cardiac and pyloric
foveolae stain intensely with PAS, with little variability among species.
In contrast, the lower one-third to one-tenth, depending upon the species,
of the cardiac and pyloric tubules were found to always stain less intensely
than the remainder of the tubule (excepting the equivalent reaction in the
pyloric glands of Molossits). A reasonable explanation for this might be
that less total mucous material is present in the basal portions of the
tubules than within the upper border. However, little physiological evi-
dence is available to confirm such a statement, and results of staining with
Alcian blue, discussed beyond, tend to refute this idea as well. Loss of
staining capacity within the basal portion is abrupt from one cell to the
next, suggesting a specific loss of some cellular material within the basal
area. The degree to which the basal area stains in contrast to the remainder
of the tubule depends upon the glandular area under consideration, and
the species examined.

In the case of the cardiac glands, basal reactivity is only slightly less
(+3 or +4) than in the remainder of the gastric tubule (+5), except for
marked decreases (+1 or trace) in the cases of Plecotus and Pizonyx. A
similar pattern was observed in the pyloric glands, but several additional
species, all carnivorous, revealed marked reduction in basal reactivity. Basal
portions of the pyloric glands of Plecotus and Pizonyx revealed little or no
reactivity. Also, the pyloric glands of Rhynchonycteris and Molossits had

Studies of Stomachs of Selected American Bats 693

little reactivity in the basal area. All the phyllostomatids and Natalus re-
vealed only slightly less basal staining in the pyloric glands in comparison
with the remainder of the tubule.

Overall production of neutral mucins, as revealed by PAS, appears to
be relatively uniform among bats examined. There is clearly generalized
reduction in intensity of PAS reactivity throughout the stomach and upper
duodenum of Desmodus, although relatively slight in comparison to other
bats. Species exhibiting especially high levels of secretion of gastric mucins
demonstrated by PAS include Noctilio, Pizonyx, Molossus, and particularly
by Phyllostomus.

Because results of PAS are uniform among species for a number of
glandular and cellular elements, it is difficult to draw conclusions regarding
relationships between and among the bats studied. It is of interest to note,
however, that certain pairs or groups of bats are nearly identical, yet unique,
in reactivity in the spectrum and combinations of glandular or cellular
types showing particular degrees of staining capacity. Pkcotus and Pizonyx
are nearly identical in reactivity, differing slightly only in the basal portions
of the cardiac glands. Sturnira and Artibeus are equally similar, as are
Pteronotus and Phyllostomus. Members of the Phyllostomatidae are not
equivalent in reactivity of all gastric regions examined. Carnivorous and
partially carnivorous phyllostomatids can be separated from frugivorous
kinds by the intensive staining in the lower mucous neck cells and weaker
staining in the parietal cells. Glossophaga, principally a nectivore that also
eats some insects and fruit, evidences staining of the lower mucous neck
cells and nearly equally intensive staining of parietal cells. The most un-
usual bats in terms of patterns of staining are Noctilio with heavily stained
zymogenic granules, and Desmodus, which exhibits overall reduction in
presence of PAS positive mucus.

Reaction to Toluidine Blue O

Toluidine blue O stains mucous material either orthochromatically
(blue) or metachromatically (reddish-purple). Toluidine blue metachro-
masia is thought to indicate the occurrence of sulfated acid mucopoly-
saccharides.

To test for the presence of hyaluronic acid which may stain metachro-
matically, tissue sections were exposed to an aqueous medium of testiculai
hyaluronidase, which presumably digests both hyaluronic acid and sev
chrondroitin sulfates. No demonstrable alteration of local metachrom
was observed in any species after 18 hours digestion at 40°C. The
suggest absence of hyaluronic acid and chrondroitin sulfates
distinctive components of certain metachromatically stainin
study.

A summary of the results of toluidine blue staining is show

694

The University Science Bulletin

Table 1. Toluidine blue metachromasia.

V

a >

Species O <-£

Rhynchonycteris naso + (tr)

Noctilio labialis —

Pteronotus parnellii —

PhyUostomus discolor +

Glossophaga soricina +

Carollia perspicillata —

Sturnira lilium —

Artibeus lituratus

Desmodus rotundus —

Natalus stramineus —

Plecottts townsendii + (tr)

Pizonyx vivesi +

Molosst/s ater —

Control (Peromyscus) .... —

Area of Reactivity

c c

■ — ! u

O PQ

-f-(tr)

+ (tr)
+ (tr)
+ (tr)

+

c c
cC u

ca v.

S^

+
+
+

+ (tr)

—

-f(tr)

+

+

+

—

+

+

+

—

+ (tr)

+

—

+

o —

O 8

+
+
+
+
+
+
+
+
+
+
+
+
+
+

Some slight variation in staining was noted in the glands of Brunner be-
tween individual specimens of two species (Plecotus townsendii and Carollia
perspicillata), but the reaction to this stain of specimens of all other species
was uniform for a given gland. This observation complements the negative
results of sulfate autoradiography of Jennings and Florey (1956) in surface
epithelia of several species of rodents. Decidedly positive reactions were
widespread in the foveolae of fundic glands only in Rhynchonycteris and
PhyUostomus, and in the fundic caeca of Plecotus and Pizonyx, with slight
reactivity in the fundic caecum of Glossophaga. Sections of the stomachs
of two rodents used as controls {Peromyscus leucopus and laboratory-raised
Mus musculus) yielded no metachromatic staining in either surface epithelia
or gastric foveolae.

The most consistently positive metachromatic staining was observed in
the goblet cells of the upper duodenum. All species examined yielded a
distinctly positive metachromasia.

Well-defined patterns of positive metachromasia within systematically
related groups or among bats with similar food preferences were not ob-
served. It is of interest to note, however, that only PhyUostomus yielded
a positive metachromasia in the deeper mucous neck cells. Also, the two
specimens of P. discolor examined demonstrated the greatest overall meta-
chromasia within the seven regions of the stomach considered, apparently
indicating atypically well-developed sulfate mucous material (s) in that
species. The results also suggest that noncarnivorous species, especially

Studies of Stomachs of Selected American- Bats 695

frugivores like Sturnira and Artibeus, contain comparatively less meta-
chromatic material, and hence mucous material with sulfate groups, than
do most carnivorous species. There are exceptions, however, and additional
studies involving more species and numbers of species are required to test
this point. Desmodus shows definite overall lack of gastric metachromasia,
although perhaps for different reasons than the frugivores. Little is known
regarding the physiological role of gastric mucins in general, and less with
specific regard to those with sulfate ester linkages (if indeed toluidine blue
metachromasia demonstrates such substances).

All stomachs examined were devoid of gastric contents and, therefore,
supposedly were in relatively equivalent physiological states with respect
to mucus production; the variability observed in response of mucins to
toluidine blue O probably reflects true differences among the species studied.

Reaction to Alcian Blue

The Alcian group of dyes, including Alcian blue, green, and yellow are
thought to denote the occurrence of acid mucopolysaccharides by way of
coloration of carboxyl groups. Mowry (1963) noted that Alcian blue closely
approximated the results of Hale's reaction (Hale, 1946) for acid polysac-
charides.

The Hale colloidal iron procedure was tried twice on sections of stomachs
of bats, but results were entirely negative for every species, perhaps a result
of inappropriate fixation.

Two different, although similar, Alcian blue tests were conducted. In
the bats studied, reactions to the two procedures utilizing Alcian blue were
generally comparable excepting where cardiac and pyloric glands were in-
volved.

Exposure to periodic acid prior to staining with Alcian blue was reported
by Mowry (1963) to demonstrate the presence of glycogen. No distinctive
concentrations of glycogen were found in my study by the use of this
procedure.

Goblet cells of the upper duodenum reacted strongly to both Alcian blue
procedures in every species examined. Spicer (1960) noted that acid dif-
ferentiation appeared to increase staining of sulfated mucins by Alcian blue;
however, in light of the clearly positive reactivity in the absence of periodic
acid oxidation, it appears safe to suggest that both sulfated and non-
mucosubstances occur in the goblet cells of most, if not all, bats examined.
These results are entirely consistent with those obtained by Spicer
for three species of rodents and a rabbit.

The duodenal glands of Brunner revealed strongly positive
with Alcian blue in only one species, Carollia, in contrast to cor
more widespread toluidine blue metachromasia in these glands ;
species of bats (see Table 1). Staining of mucins wit

696 The University Science Bulletin

Brunner not only occurred in two specimens of Carollia but was exceedingly
strong. Although positive reactivity might have been the result of sulfated
mucosubstances known to be present by the results of toluidine blue tests,
this prospect seems unlikely in the absence of positive results in any other
species. A trace of Alcian blue staining was observed in Brunner's gland
cells of Sturnira and Artibeus, but it is difficult to justify this as resulting
from the presence of mucosubstances because there was no contrast with the
surrounding tissue. The basal-most cell complex within the pyloric stomach
of Sturnira revealed but a trace of staining, and that of Molossus revealed
none, perhaps further justifying the identification of glands of Brunner
within the stomachs of these two bats.

A weak to strong reaction was recorded in the gastric surface epithelium
of the cardiac stomach of all species examined excepting Sturnira, Artibeus,
and Desmodus. The epithelium of Sturnira and Artibeus stained only with
a trace, that of Desmodus not at all. Results were nearly identical between
the two Alcian blue procedures, with no distinctive differences requiring
evaluation.

Surface epithelial cells of the fundus reacted positively to Alcian blue,
although consistently slightly less than the cardiac epithelium in most
species. Rhynchonycteris and Phyllostomus would seem to be exceptions
to this trend by showing equality of reactivity in the two glandular regions,
and Desmodus revealed moderate staining in the fundic epithelium in con-
trast to none at the surface of the cardia. Several species demonstrated
decreasing reactivity of the surface epithelium with Alcian blue through the
terminal portion of the fundus and through the pyloric zone to the pyloric
sphincter. These included all carnivorous species and Glossophaga, but
not the three frugivorous species or the sangivorous Desmodus. This reduc-
tion in staining quality probably represents a progressive decrease in reactive
material from the midstomach toward the pyloric sphincter. This condition
is especially interesting in light of the often observed increase in depth
of the surface epithelium from the fundic mucosa through the pylorus.

Reactions of cells within the cardiac glands are highly variable between
portions of a single tubule, among species, and between the two procedures
for certain species. As in the case of PAS, the upper neck as well as various
relative percentages of the basal portion of tubules stain differentially with
Alcian blue. Whereas reaction to Maxwell's Alcian blue tends, in most
species, to correspond with results observed with PAS (that is, staining
within the basal portion is less than that observed in the neck region), a
simple exposure to aqueous Alcian blue (Steedman's procedure) revealed
the reverse situation in Natalus, Plecotus, Pizonyx, and Molossus. In these
genera the basal area stained more intensely than the neck. The cardiac
glands of Sturnira yielded no positive staining with either procedure; those

Studies of Stomachs of Selected American Bats 697

of Artibeus stained weakly and equally throughout the tubule with both
procedures. The aqueous reaction (Steedman's) for the remaining seven
species was moderate to strong, and either equal in strength throughout
the tubule, or less in intensity in the basal area (seen also in these seven
species using Maxwell's procedure). The pattern of overall relative re-
sponse to Alcian blue between cardiac glands of different species was roughly
equal between the two procedures, although Steedman's was consistently
slightly more intensive than Maxwell's.

In summary, the results of staining cardiac glands with Steedman's
Alcian blue revealed a pattern of distribution into one of three categories,
depending upon the presence or absence of reversed differential staining
within the tubular glands of the cardia. One group that had differential
staining with the aqueous Alcian blue procedure reversed to that of PAS
is composed of Natalus, Plecotus, Pizonyx, and Molossus. A second group,
composed only of Sturnira and Artibeus, showed little or no reactivity
throughout the cardiac tubules. The last group included all remaining
species in which the cardiac glands reacted moderately to strongly, either
in a pattern similar to that of PAS (basal staining less than upper portion)
or with uniform staining throughout the length of the tubule.

Results for Alcian blue staining of pyloric glands were somewhat similar
to those obtained in the cardiac mucosa. Reversed (to that of PAS) differ-
ential staining occurred after exposure to Steedman's Alcian blue procedure
in three (Natalus, Plecotus, Pizonyx) of the four species in which this was
demonstrated for the cardiac glands (and with increased intensity relative
to Maxwell's procedure, also seen in the cardiac glands). The basal portions
of the pyloric glands of these three species stained moderately to intensely,
whereas the necks gave only a weak reaction (Molossus revealed only
slight staining throughout the pyloric glands). Using Maxwell's procedure,
basal portions of the pyloric glands of these species were consistently un-
stained and reactions in the necks were weak to moderate. Sturnira re-
vealed no staining of pyloric glands with either procedure; Artibeus stained
slightly in the neck region, but only when Maxwell's procedure was used.

None of the remaining species showed reversed differential staining
in the pyloric glands. Except for Glossophaga and Carollia, the pyloric
glands of which reacted uniformly and strongly only to Steedman's pro-
cedure (and with complete lack of staining after periodic acid oxidation),
there is a clear trend within remaining genera studied toward more
staining of mucus using Maxwell's procedure. Rhynchonycteris, Pteroi
Phyllostomus, and Desmodus reacted only with Maxwell's procedure,
of the species of the Emballonuridae, Noctilionidae or
demonstrated the reverse differential staining discussed pr

In summary, it may be said that the cardiac and pyloric gl

698 The University Science Bulletin

considerably more similar histochemically among species in the Natalidae,
Vespertilionidae and Molossidae, than among the Emballonuridae, Noc-
tilionidae, Desmodontidae and Phyllostomatidae, with the exception of the
uniformly poor reactivity of the phyllostomatids Sturnira and Artibeus. The
results included here would seem to refute contentions by some workers
that pyloric glands and glands of Brunner may be equivalent gastrointestinal
entities.

The mucous neck cells of the fundic mucosa of Rhynchonycteris, Noctilio,
and species of the Phyllostomatidae (excluding Phyllostomus, Sturnira and
Artibeus) stained weakly to intensely with Steedman's procedure, but
stained only slightly or not at all with Maxwell's. Phyllostomus stained
strongly with both procedures, whereas Sturnira revealed no staining in the
mucous neck cells and Artibeus showed only a trace of staining with
Steedman's method. The mucous neck cells of Desmodus did not react to
Alcian blue. The remaining four species of the Natalidae, Vespertilionidae,
and Molossidae revealed an interesting pattern of coloration. The mucous
neck cells of Natalus reacted moderately to Maxwell's procedure, but did
not stain with a simple aqueous method. In contrast, those of Plecotus,
Pizonyx and Molossus all stained moderately with the aqueous procedure;
the first and last species showed a trace of staining with Maxwell's pro-
cedure but Pizonyx did not.

In comparing the results obtained with PAS and Alcian blue, the most
notable immediate contrast is that observed in the glands of Brunner.
Although Lillie (1%5) recorded these glands as staining only red in a
combined PAS-Alcian blue procedure, suggesting negative results in terms
of Alcian blue, Spicer (I960) reported that although these glands in mice
and rats react as indicated by Lillie, Brunner's glands of guinea pigs and
rabbits yielded a moderate to strong reaction to an aqueous solution of
Alcian blue. Carollia appears to represent an example of the latter condition
among bats.

It is apparent that variation exists among species of bats in the presence
or absence and degree of reactivity to Alcian blue in various mucous cells,
hence in relative amounts of acid mucosubstances, and is substantially
greater than the reactivity of relatively neutral materials stained with SchifT's
reagent. The fact Alcian blue at pH 7.1 may show greater specificity in
reactivity with various substances than does SchifT's reagent may be a con-
tributing factor to this variability.

Summary. — In summary, acid and neutral mucopolysaccharides of epi-
thelial, gastric and Brunner's gland secretions react variably to a greater
or lesser degree to four histochemical procedures that demonstrate the
presence of mucin materials. Variability in reactivity of gastric and duo-
denal mucins to toluidine blue is extreme among carnivorous species;

Studies of Stomachs of Selected American Bats 699

herbivores and a single sanguivore are uniformly unreactive in all but
goblet cells mucins. The results of PAS and Alcian blue procedures reveal
interesting patterns of staining in certain portions of the gastric mucosa
that suggest unique combinations of various mucin materials within closely
related families or subfamilies of bats, and reveal evidence that bats with
markedly different food preferences have different qualitative and quanti-
tative mucopolysaccharide complements.

Mucin Histochemistry of the Stomach in Relation to Food Habits

The utilization of four histochemical procedures, thought to be specific
for complex mucopolysaccharide substances, has revealed interesting trends
and similar patterns of reactivity among certain phylogenetically and nu-
tritionally related groups of bats.

The most extensive gastric mucous production of any bat examined
was that of Phyllostomus discolor. Mucins of this species reacted consistently
and with high intensity to all three procedures employed, and these results
reflect the presence of a particularly well-developed gastric mucosa in this
bat. The omnivorous diet and seemingly ravenous feeding habits of this
species evidently require abundant surface mucus to replenish a rapid loss
of large quantities in escaping chyme. The omnivorous habits of Phyllo-
stomus probably are related to comparatively extensive secretion of meta-
chromatically staining gastric mucus and to equally extensive neutral epi-
thelial mucins. In addition to the extensive mucous material in the stomach
of Phyllostomus, as revealed by tests with PAS, the strongly positive results
from both Alcian blue procedures indicate that chemical diversity of acid
mucins may be more extensive in this bat than in most other species.

Histochemical results indicate that the three frugivorous species and the
sangivorous Desmodus are comparatively low in production of gastric
mucins, including surface as well as deeper mucous cells. It is clear, how-
ever, that Carollia far exceeds Sturnira and Artibeus in production of gastric
mucus, and approaches the moderate condition of Glossophaga. Moderate
reactions to PAS were recorded for the surface epithelium of Artibeus and
Sturnira, although the quantity of mucus present was low for each. To-
luidine blue and Alcian blue reacted weakly in these two species.

In all bats but the frugivores, the surface epithelial mucin demonstrated
by Alcian blue was found to decrease progressively in reactivity through
the terminal portion to the pyloric sphincter; this epithelium stained con-
sistently to the sphincter in the obligate fruit-eaters. This suggests the
possibility of more widespread secretion of mucin in the stomachs of the
latter group.

In overall mucous production, the primarily nectivorous Glossophaga
tends to resemble insectivorous and carnivorous species more than frugi-

700 The University Science Bulletin

vores, although it is notably similar to Carollia in several features. Mucous
material, as revealed by PAS and Alcian blue, is more abundant throughout
the stomach of Glossophaga than the three frugivorous kinds, although less
abundant than in Pteronotus or Phyllo stomas. Glossophaga also reveals
greater diversity of cell types yielding a demonstrable metachromasia than
do any of the frugivorous species (diversity also is the tendency in insecti-
vorous species in this regard). Glossophaga and Carollia are similar and
unique, however, in the greater intensity of staining of the gastric foveolae
within the fundic caecum in contrast to the remainder of the fundic mucosa
(absent in other species, even in those with deepened gastric pits in the
fundic caecum). Additionally, Glossophaga shows intensive staining of
the lower mucous neck cells, otherwise restricted to insectivorous and
carnivorous species, and this, in combination with the moderate reactivity
of parietal cells to PAS noted above, clearly suggests the presence of an
intermediate condition relevant in mucin production in this species. It is
of interest to note that the parietal cells of the three frugivorous species
(also Glossophaga) react moderately strongly with PAS, but there is no
comparable result in any purely insectivorous or carnivorous species. The
possibility exists, although admittedly remote, that the parietal cells of
these frugivores may assume a mucus-producing function in the absence
of the surface epithelial mucus found to occur in most other species of bats,
in addition to their function as the probable source of hydrochloric acid.

It has been fairly well established that mucous production is slight in
the stomach of Desmodiis, in surface and deeper mucous material, as well
as in secretions from the glands of Brunner. In the absence of numerous
parietal and chief cells, there is little probable need for surface mucus as a
protective device against autodigestion, and the non-abrasive nature of the *
primary food source of this species would suggest little need for epithelial
mucus as a lubricant against mechanical injury.

Most insectivorous and carnivorous species of bats yielded moderate to
strong reactions (PAS) for neutral epithelial mucins, and for acid muco-
polysaccharides as demonstrated by Alcian blue. As regards overall mucous
production, relatively little information emerges, using the general tech-
niques employed in this study, to suggest differences between insectivorous
and piscivorous species. However, the mucins of the cardiac and pyloric
glands of Pizonyx revealed particularly strong metachromasia with toluidine
blue, and the epithelial mucins in these two glandular areas may be gen-
erally thought of as uncommonly abundant. In comparison to its relative,
Plecotus, and from what can be gathered from the work of Rouk (1968)
on two other vespertilionid bats, the cardiac and pyloric glands of Pizonyx
are morphologically and physiologically distinctive elements in this bat,
and for reasons that are unclear provide particularly abundant mucus at
the two gastric orifices.

Studies of Stomachs of Selected American Bats 701

Some considerable variation exists in reactivity of the deeper cells of
the cardiac and pyloric glands to particular procedures among insectivorous
species, particularly for the two Alcian blue techniques. Because this vari-
ation appears to have greater systematic than morphologic application, a
discussion of it is deferred to a later section. Insectivorous, piscivorous, and
nectivorous species reveal heavier complements of gastric mucus in the neck
cells of the cardiac and pyloric glands than do obligate fruit-eaters and the
sangivorous Desmodus.

The unique presence, among the bats studied, of PAS-reactive zymogenic
granules in the stomach of Noctilio is not readily explainable.

Toluidine blue metachromasia is considered by most, including Spicer
(1960, 1963), to demonstrate the presence of sulfated acid mucopolysac-
charides. This belief is supported by autoradiographic analyses of 3-JS ab-
sorption in the gastric mucosa of several species of rodents and one rabbit
as presented by Jennings and Florey, 1956; these show general conformity in
;!'"'S04= uptake in areas correspondingly demonstrating toluidine blue meta-
chromasia. Metachromasia may occur in some neutral mucopolysaccharides
such as those of the glands of Brunner but Wislocki et al. (1947) suggested
that hyaluronic acid may account for some metachromasia in tissue sec-
tions. Tests for hyaluronic acid were negative in all species examined sug-
gesting a lack of this substance in the gastric mucus of bats.

Reactivity to toluidine blue O in the form of observable metachromatic
staining reveals extensive variability between species and families, and
among carnivorous species generally. Non-carnivorous species within the
Phyllostomatidae tend to show less reactivity than do the carnivorous
members examined. A principally nectivorous phyllostomatid, Glossophaga,
which apparently often takes insects in the diet, had greater diversity of
areas showing positive reactivity (as in the strictly carnivorous types) than
did the three frugivorous phyllostomatids. The two species examined of
the Vespertilionidae were notably similar, although not identical, in pattern
of metachromatic staining.

The glands of Brunner, although not an element of the stomach in
most species, are morphologically and physiologically closely associated
with the stomach, and therefore were examined histochemically. The glands
of Brunner of all species revealed some PAS-positive material, whereas
toluidine blue gave mixed results, and only Carollia yielded a positive
staining with Alcian blue. It would appear that the glands of Brunner of
Carollia, although extremely limited in distribution, are unusually active
in the secretion of an abundance of neutral, as well as acid mucopolysac-
charides.

Species variability in reactivity of Brunner's glands to PAS has been
shown to exist in rodents by Spicer (1960). He noted that Brunner's

702 The University Science Bulletin

glands of albino mice stained weakly, those of rats stained moderately, and
those of guinea pigs stained more intensely than either rats or mice. The
physiological significance of such variability is unknown. It is of interest
to note that Lillie (1965:511) recorded the glands of Brunner as negative
to Alcian blue, and it would appear that Carollia and perhaps Sturnira and
Artibeus may represent genera within subfamilies of bats unusually adapted
for mucous production by the glands of Brunner. Relatively neutral mucin
material appears to be most consistently represented in the Brunner's glands
of the insectivorous or partially insectivorous Phyllostomatidae, Noctilioni-
dae, and Molossidae. The results of PAS, as noted previously, confirm the
presence of Brunner's glands in the stomachs of Sturnira and Molossus.

GASTRIC MORPHOLOGY IN RELATION TO FEEDING

Food Habits of Species Examined

Relatively little as yet is known concerning the specific nature and
geographic variability of food habits for most species of bats. Therefore,
we can only generalize regarding feeding preferences of certain species.

Little or no information appears available regarding the specifics of food
habits of Rhynchonycteris naso or, indeed, emballonurid bats in general.
Goodwin and Greenhall (1961) noted that R. naso on Trinidad and To-
bago, along with all other species of sac-winged bats on these islands, feed
exclusively on insects. Perhaps mostly small, soft-bodied prey species are
taken by this bat and are macerated to the point that identification is
difficult.

Noctiho labialis, unlike the fish-eating N. leporinus, appears to be an
obligate insect-feeder as noted by Bloedel (1955) and Hooper and Brown
(1968). According to Hooper and Brown, beetles, Lepidoptera, and bugs
comprise a large portion of the diet of N. labialis. Numerous species of the
Phyllostomatidae, long thought to be almost completely herbivorous or
nectivorous, are now known to be in part insectivorous or carnivorous
as a result of recent food habit analyses. A number of kinds including
Pteronotus parnellii are as yet poorly studied. P. parnellii has been reported
to be an obligate insect-feeder. Phyllostomus discolor was reported by
Goodwin and Greenhall (1961) to eat only fruit, but Arata et al. (1967)
observed insect remains in the stomach of P. discolor, noting that the genus
Phyllostomus is omnivorous. The omnivorous habits of a closely related
species, P. hastatus, are well documented (Dunn, 1933; Arata et al., 1967).
Glossophaga soricina is a nectar-feeder, but feeds additionally on soft fruits
(Arata et al., op. cit.) and is known also to take large insects, thus dis-
counting the hypothesis that some insects are taken only incidently with
ingested fruit. Although Carollia perspicillata is primarily a consumer of
soft, pulpy fruit, Arata et al. (1967) noted that the stomach contents of

Studies of Stomachs of Selected American Bats 703

more than a fifth of the specimens of this species examined by them
contained quantities of insect material. Also noted was a single instance
of either cannibalism or predation on another bat. It would appear, on the
basis of present information, that C. perspicillata is not truly an obligate
fruit-eater, although it probably is less an omnivore than Glossophaga or
Phyllostomus. Sturnira lilium is clearly an obligate fruit-eater. Artibeus
lituratits may occasionally take some insects in the diet, although it con-
sumes mostly soft fruits, apparently in great quantities, as food material
passes through the digestive tract quite rapidly (in 15 to 20 minutes as
reported by Goodwin and Greenhall, 1961).

The three vampires of the Desmodontidae are generally thought of as
being obligate blood-feeders, with the exception of occasional instances of
insect-feeding by Desmodus. Considering the marked structural and be-
havioral specialization for blood-feeding in this group, relative obligation
to sanguivorous habits appears sufficiently verified. Wimsatt and Guerriere
(1962) noted the rather large daily blood meals taken by Desmodus in
captivity, averaging over 15 ml per day over a period of a year, and
Wimsatt (1969) recently confirmed these observations in the field.

Three of the remaining four species (Natalus, Plecotus, and Molossus)
may be considered obligate insect-feeders. Ross (1967) noted that P. town-
sendii is especially fond of small Lepidoptera. According to Pine (1969),
Molossus ater in Costa Rica takes many small insects, especially ants, other
small Hymenoptera, and a variety of beetles. Little information is available
regarding the habits of the Natalidae; Goodwin and Greenhall (1961)
stated only that the diet of Natalus is restricted to insects.

Pizonyx vivesi is presently recognized as a regular fish-eater, although
insects and small crustaceans have been observed in quantity in the stomachs
of several specimens (Reeder and Norris, 1954). Small sardines and an-
chovies may comprise a substantial portion of the fish diet of Pizonyx.

It is of interest to note that many species of a number of families, al-
though principally restricted to a non-insect diet, occasionally take insects.

Gross Morphological Features

Although variation in general morphology among the species of bats
examined is extensive and each species is clearly unique in one or more
ways, it is apparent that trends in similarity do exist among both closely
related genera and among those with similar food habits. Several broad
categories of general structure can be seen in the 13 species examined.

The first of these groups, which includes Noctilio, Pteronotus, Natalus,
Plecotus, and Pizonyx, is characterized by a stomach that is relatively short
in relation to breadth, and that has a convex dorsal surface with no dilation
of the dorsal surface of the fundic caecum, and is generally symmetrical,
with the terminal portion (that region distal to the gastroesophageal junc-

704 The University Science Bulletin

tion) equal (or nearly so) in length to that of the fundic caecum. Also,
the terminal stomach is relatively short and never sharply recurved craniad,
and the esophagus enters the stomach approximately midway along the
lesser curvature, never at a marked angle to the stomach. Two additional
vespertilionids, Myotis velifer and Antrozous pallid us (both insect-feeders)
examined by Rouk (1968), would seem to fit this category. This simplified,
tubular gastric form is restricted to, although not universal among, carniv-
orous species. Pizonyx vivesi deviates most from the basic plan discussed,
because its esophageal entrance is slightly angled and its stomach is con-
siderably more rounded than in the other four genera. The apparent spe-
cializations in gross morphology observed in Piznoyx cannot be attributed
necessarily to fish-eating, however.

Generally speaking, the stomachs represented by this first group of bats
closely resemble those of the Insectivora discussed by Myrcha (1967) and
by Allison (1948). This similarity is greatest between bats (excluding
Natalus) and hedgehogs of the family Erinaceidae, with which there is
striking correspondence in configuration and relationship of the fundic,
pyloric, and gastroesophageal areas, as well as similar distribution of gastric
mucosa. Natalus, with its relatively straight pyloric tube, most closely
resembles soricids. It is suggested that this first category of species is repre-
sented by those with stomachs that topographically and grossly resemble
the probable simple primitive condition in bats, species with a more spe-
cialized fundic caecum and an elongate terminal portion being most ad-
vanced. This statement is not intended to infer that the stomachs are
necessarily primitive in all characteristics, but only that the basically simple
"insectivore-type" prevails.

Two species of obligate insectivorous bats included in this study clearly
do not conform to the category described above. While distinctively tubu-
lar and convex throughout on the dorsal surface, the stomachs of Rhyncho-
nycterts naso and Molossus ater have notably elongate terminal portions in
comparison to those of the five species previously discussed, as well as a
sharply angled entrance of the esophagus into the stomach. Pyloric recurva-
ture is moderate in Rhynchonycteris and pronounced in Molossus; regarding
topography, these two species represent a second category of more special-
ized insectivorous kinds. Possibly they are adapted for consumption of
relatively large quantities of food during a single feeding, resulting in
extended exposure to proteolytic digestion in the elongated pyloric portion
of the stomach. It will also be noted that the transitional zone from fundic
to pyloric mucosa is most extensive in these two species among all those
examined with elongated terminal portions. This condition would tend to
expose the food bolus to HCl-producing mucosa for a greater distance than
in other species. A more complete analysis of quantities of food consumed
is required to verify this suggestion.

Studies of Stomachs of Selected American Bats 705

A third group, which includes Curollia perspiallata, Sturnira lilium,
and Artibeus lituratus, have stomachs with a moderately to extensively
developed cardiac vestibule, well-developed fundic caecum that is dilated
and hooked on the dorsal surface, and a narrow, elongate terminal portion
that is recurved sharply craniad. Notable related specializations of the
stomach wall are discussed in preceding sections. The conditions prevailing
in these species would seem to relate to accommodating large quantities of
bulky food material, consistent with the frugivorous habits of these species.
The specializations of, and general increase in size of, the fundic caecum
in frugivorous bats would seem to parallel a similar trend noted by Voron-
tzov (1960) in passing from primarily albuminous (animal) to principally
cellular (plant) feeding in muroid rodents. A chamber-like cardiac vesti-
bule, especially well developed in Sturnira and Artibeus, increases the avail-
able surface to which ingested material is exposed in the gastric phase of
digestion. It is not possible at this time to conclude whether the vestibule
is to be considered histogenetically a portion of the esophagus or a dilated
portion of the lesser curvature, although in light of the well-developed
gastric mucosa within the vestibule the latter explanation seems the more
probable. As noted previously, Curollia perspicillata may be a less obligate
fruit-eater than either Sturnira or Artibeus and some features of gross
morphology that tend to distinguish Sturnira and Artibeus from remaining
kinds are, indeed, partially reduced in Carolha. The cardiac vestibule is
distinctive as a conical ampule externally in Carolha, although noticeably
shorter and relatively narrower than in either of the other two species. The
fundic caecum is well developed and dilated dorsally, although somewhat
more rounded as in Glossophaga, a species to be discussed shortly. The
pyloric sphincter of Carollia combines features observed in Artibeus (such
as being long and symmetrical) with a structural feature of insectivorous
or partially insectivorous kinds in which the sphincter valves increase in
breadth toward the inner limit (narrowly tapered in Artibeus and reduced
to a vestige in Sturnira) and terminate in a broadly rounded apex. Carolha
clearly is a species that normally consumes mostly pulpy plant material and
fruit juices, but perhaps has turned to such habits only relatively recently
inasmuch as its stomach appears intermediate in many ways between
insectivorous and frugivorous types.

The fourth category of stomachs represented by those of Phyllostomus
discolor and Glossophaga soricina suggests an additional type of inter-
mediacy between strictly herbivorous and carnivorous conditions. The
cardiac vestibule is either reduced to a vestige (Phyllostomus) or absent
(Glossophaga), and, whereas the fundic caecum is dilated and expanded
on the dorsal surfaces, the caecum appears relatively smaller in comparison
to the remaining stomach than in the two obligate fruit-eaters discussed

706 The University Science Bulletin

previously. Only short endpieces of the terminal portions are recurved
cranially in both Phyllostomus and Glossophaga, although the terminal
tubular stomach is extensive in both. Also, whereas the pyloric sphincter
retains its narrowness, as observed in other phyllostomatids, the valve in
these two species is asymmetrical (larger on the greater than the lesser
curvature). It is suggested that asymmetry of the pyloric sphincter is in
some way related to carnivorous feeding in bats (additional discussion on
this point follows in a more detailed comparison of the pyloric sphincters
examined). In addition, the stomachs of Glossophaga and Phyllostomus
appear to be "intermediate" in degree of symmetry, revealing more gradual
antero-posterior curvature than observed in Carollia, Sturnira or Artibeus.
The significance of such general arching of the stomach is obscure, although
an hypothesis regarding the function of sharp recurvature of the pylorus is
presented below.

In Sturnira and Artibeus, which presumably consume rather large quan-
tities of bulky plant material, mechanisms to retard gastric emptying, thus
exposing food material that is difficult to digest to longer periods of gastric
enzyme exposure, would theoretically be of selective advantage. Such
mechanisms are well known in other kinds of herbivorous mammals (for
example, ruminate and semi-ruminate stomachs of bovines and many herbiv-
orous rodents). A sharply angled pyloric tube such as found in Artibeus,
Sturnira or Carollia likely would have a distinctive crook at the point of
recurvature when the stomach is filled to the point of distention. This
crook in the terminal stomach might well help to retard the rapid transport
of food through the stomach. A single insectivorous species, Molossus ater,
was found to have a similarly sharp cranial recurvature, although the
terminal recurved endpiece is extremely short, and the degree of constric-
tion of the pylorus upon complete distention of the stomach with food
is probably not great. It is apparent that the degree of recurvature of the
pyloric portion of the stomach is to some degree a function of angulation
of the stomach with respect to a transverse plane through the animal (that
is, the more anteriorly displaced is the fundic caecum, generally, the greater
the pyloric bend). In the case of the extremely large and elongate stomach
of Molossus ater, apparent compensation for size has been made by a shift
of the stomach up and to the left, underneath the left lobe of the liver.
Thus, the degree of pyloric recurvature is also, most probably, a function
of relative size of the stomach; large stomachs may be the result of in-
creased nutrient requirements or reduced energy per unit of food or both.
Members of the family Molossidae are strong, fast fliers, which represent
the most advanced development of the forelimb for flight among bats.
Superior development of flight allows at least one species, Tadarida bra-
sihensis, to migrate, and the relatively large size of the stomach in this

Studies of Stomachs of Selected American Bats 707

species (Rouk, 1968), to accommodate relatively large amounts of food,
is not surprising in this regard.

It would appear then, that elongation of the stomach is related to in-
creased food intake in the evolutionary history of the Chiroptera. An
interesting parallel in this regard can be found in the Insectivora. Myrcha
(1967) has found that the stomachs of some insectivores, notably Soricidae
and Talpidae, reveal elongation of the terminal portion of the stomach,
likely also in response to increased food requirements.

The last category to be considered, containing only Desmodus, repre-
sents the vertex of specialization in gross gastric morphology in bats thus
far examined. It is difficult to observe any major feature of the stomach
of Desmodus that closely resembles the condition in any other species thus
far studied (excluding, of course, Diphylla as discussed by Grasse, 1955).
The elongate and tubular fundus is clearly adapted to hold relatively large
amounts of blood. Extremely elastic bands of thin, loose, bundled, muscula-
ture and distinctive complements of loose areolar connective tissue allow
extensive expansion of the fundic caecum. The extremely close proximity
of the cardiac and pyloric orifices (Fig. 33) suggests, in the absence of
apparent structural modifications of the cardia to assure passage of ingested
blood to the left into the caecum, that the pyloric sphincter must be at least
partially functional, or as suggested by Selkurt (1966), localized contrac-
tions of the proximal area of the duodenum may provide the means of
gastric closure at the lower end. The pyloric sphincter is extremely thin,
symmetrical, and appears to be structurally most similar to that in several
phyllostomatids, especially Carollia and Artibeus; in Desmodus it is reduced
secondarily to a thin flap.

The long, thin, tubular nature of the stomach of Desmodus undoubtedly
serves to facilitate utilization of a wholly liquid meal in several ways.
Elongation of the stomach coupled with extensive vascularization of the
submucosa provides increased surface area for the rapid absorption and
excretion of fluids as noted by Wimsatt and Guerriere (1962). Modifica-
tions of the gastric mucosa, to be discussed later, may also provide for
increased surface area. The tubular stomach, when folded on itself midway
along its length and displaced longitudinally or transversely within the body
cavity, would provide a mechanism for balancing the extremely large
blood meal equally on the left and right sides of the bat, providing for
more stable flight from the host after feeding. Vampires have been ob-
served to engorge themselves to the point that they are unable to take
flight after feeding (Goodwin and Greenhall, 1961); a means to provide a
"balanced load" would seem to be advantageous to these bats. Studies
of the intravisceral configurations of the stomach of Desmodus just after
feeding should be carried out to test this suggestion.

708 The University Science Bulletin-

Table 2. Classification of pyloric sphincters of several species of bats.

Valve flaps absent Sturnira

Valve flaps present, broad (apex blunt)

short, nearly symmetrical Pizonyx

short, asymmetrical Natalus

long, asymmetrical Noctilio

Plecotus

Molossus

Rhynchonycteris

Valve flaps present, narrow

short, symmetrical Desmodits

long, symmetrical Carollia

extremely long, tapering, symmetrical Irtibctis

long, asymmetrical Pteronotus

Phyllostomus
Glossophaga

Extensive variation in morphology of the pyloric sphincter in bats (Table
2) presents an interesting physio-morphological problem. Without excep-
tion, every species of carnivorous or partially carnivorous bat examined in
this study revealed some degree of asymmetry of the pyloric sphincter
(that is, the valve was not uniform in length or breadth, or both, through-
out the circumference of the gastroduodenal junction, the larger portion of
the valve usually being on the greater curvature). This asymmetrical con-
dition is in contrast to the pyloric sphincters of herbivorous and sanguivo-
rous species, which, although highly variable in length among species, are
equal throughout their circumference within a given bat. Dearden (1969)
examined the pyloric sphincters of four genera of microtine rodents, noting
two types of asymmetry in the sphincter — that in which the larger sphinc-
teric mass was on the greater curvature, and conversely, that in which the
valve was larger on the lesser arc. Dearden suggested that in those species
of voles with the greater mass of muscle on the greater curvature, sphincteric
action would seem to be of a milking nature, whereas symmetrically circu-
lar closure would likely result from the other condition. Although a well-
developed membrane of gastric mucosa comprises an important part of the
pyloric sphincter in bats, which lack the pediculated squamous flaps ap-
parently important to the closure mechanism in some species of voles, the
anatomical variability observed in pyloric sphincters in bats suggests that
these mammals also have functional differences in gastroduodenal closure
mechanisms. A relatively effective gastroduodenal closure mechanism, such
as would theoretically be produced in those species examined with sym-
metrical sphincters (Sturnira, Artibeus, Carollia, Desmodits), would pre-
sumably be of greatest advantage in bats feeding on materials that are
difficult to assimilate, such as pulpy plants and blood. Nothing is known

Studies of Stomachs of Selected American Bats 709

regarding the physiology of the pyloric sphincter in bats, although condi-
tions in primarily non-carnivorous bats suggest that additional studies on
pyloric sphincter morphology and physiology in bats might prove most
interesting in relation to feeding ecology.

Fundic Caecum

Comparative examination of the fundic caeca of bats reveals little spe-
cialization that can be directly related to feeding habits. An exception is
Desmodus, in which the entire stomach is essentially a thin-walled caecum
adapted to accept and retain a meal of blood that requires no physical
breakdown; the complex and normally thick muscular wall observed in
fundic caeca of other bats is lacking. No species examined lacked a fundic
caecum, and all species (excluding Desmodus) revealed fundic caeca that
were but slight alterations of a basic plan. All species of the Phyllostomati-
dae, with the possible exception of Pteronotus, possess a fundic caecum
dilated at the apex and hooked dorsally to a greater or lesser degree. This
feature is restricted to phyllostomatids; it is best developed in Artibeus,
Sturnira and Curollia, to a lesser degree in Glossophaga. It is reasonable
to suggest that this condition is also related to increasing effective gastric
volume in those species that consume large quantities of plant bulk and
liquid material, being absent (or reduced) in those kinds that have retained
(or returned to) insectivorous and carnivorous habits. The fundic caecum
of all carnivorous species consists of a simple, usually short, blind, lateral
extension of the mid-stomach, excepting the slight terminally hooked apex
observed in that of Phyllostomits (an omnivore), and functions to hold a
portion of the food material during the period of gastric churning.

Cardiac Vestibule

The cardiac vestibule and its probable functional significance in three
fruit-feeding or at least predominantly fruit-feeding kinds (Sturnira, Arti-
beus, and Carolha) has previously been discussed. The vestibule in these
three species is unique among those examined in that it represents a nearly
distinct gastric entity. It is of interest that fruit-eating megachiropteran
bats may have an equally well-developed cardiac vestibule, implying further
a relationship between consumption of large complements of fruity material
and invasion of gastric mucosa craniad into the lower portion of the
abdominal esophagus. All insectivorous and partially carnivorous species,
as well as the one sanguivorous species studied, have either a relatively
short vestibule or lack one entirely, with the possible exception of Natalus
stramineus. The vestibule of Natalus is rather extensive in breadth, although
not as distinctively delimited from the lesser curvature as the vestibule of
the three frugivorous species. Although nutritional requirements of in-
sectivorous and carnivorous bats likely are equivalent to, or greater than,

710 The University Science Bulletin

those of fruit-eaters, usable caloric substrate is likely greater per weight
of food material ingested. Insect-eating bats macerate their food into fine
fragments, and the few carnivorous kinds whose digestive efficiency has
been examined have been noted to leave little recognizable animal material
other than matted hair and hard fragments in the fecal material, even
when whole mice are consumed. This efficiency in the digestive process
suggests relatively lower requirements of food mass than in fruit- and blood-
feeders; hence, comparatively less modification of gastric proportions are
required to increase internal volume and surface area. It is of interest to
note that Natahts, the single insect-feeder possessing a large cardiac vesti-
bule, also has a relatively small fundic caecum and a terminal stomach
that is moderate in length, but extremely narrow. It is probable that the
enlarged cardiac portion in this species has developed in response to an
otherwise rather limited gastric volume.

Relatively little is known concerning the mechanism of gastroesophageal
closure in bats, which is surprising in light of their interesting and almost
unique habit among mammals of hanging upside down while roosting
(after and sometimes even during the period of feeding). An anatomically
distinctive cardiac valve is unknown in bats with the exception of that
reported herein for the genus Noctilio. This structure evidently has been
derived independently in this genus, perhaps in response to the acquisition
of fish-eating habits later abandoned by N. /tibialis. The stomachs of both
species of noctilionids appear comparatively small in relation to body
size, and the unique cardiac valve, assumed functional to some degree as
specialization of the mucularis mucosae is apparent, may prevent back-up
of large quantities of ingested fish from a relatively small stomach. Should
a cardiac closure mechanism exist in other bats, as seems likely, it must
be represented by a physiologically specialized muscular component of
circular fibers and muscularis mucosae at or near the gastroesophageal
junction, perhaps consisting partly of locally thickened musculature as in
those species with a prominent incisura cardiaca. Anatomical specialization
of muscle layers at the gastroesophageal junction in bats other than Noctilio
is restricted to a slight breakup of fibers of the muscularis mucosae, the
significance of which is unclear, and some notable local thickening of the
circular muscle layer just below the gastroesophageal junction in Rhyncho-
nycteris, Glossop/nigu, and Molossus. Botha (1958a) examined the mucosal
folds in the cardia of an unspecified species of bat, probably an insectivorous
vespertilionid, and concluded that the mucosa-lined ridges at the orifice
formed a watertight papilla, effectively closing the gastroesophageal junc-
tion. It is possible that those species with local thickening of the circular
muscle layer in this area {Rhynchonycteris , Glossophaga, and Molossus) have
efficient closure of the stomach, although no physiological evidence is avail-

Studies of Stomachs of Selected American Bats 711

able to support this point. It will be noted that the three species of frugi-
vores all have extremely thin musculature at the gastroesophageal junction
and throughout the cardiac vestibule. This condition does not, of course,
dispel the possibility of the existence of an active closure mechanism,
although the strength and efficiency of a sphincter may be more limited
in these species.

Musculature

The walls of all stomachs examined were composed of an outer longi-
tudinal and inner circular layer, and two extremely thin intra-submucosal
sheets of smooth muscle, a pattern which is common to other kinds of
mammalian stomachs. Variability between species in corresponding depths
of various portions of the stomach frequently was found to be less than
variation within a single stomach, especially between tubular and saccular
portions of several species.

Most of the species examined revealed some degree of increased thick-
ening of the greater curvature wall, in contrast to that of the lesser curva-
ture. The relative amount of, or specific region of, the greater curvature that
shows this condition is, again, quite variable and rarely expressed through-
out all of the greater arc, or within both outer layers. In Phyllostomus ,
Natalus, and Molossus, the muscular wall is consistently thicker through
the greater arc, by virtue of the circular layer, than most of the correspond-
ing wall of the lesser curvature. This condition generally holds for
Pteronotus, although the terminal portion is only slightly thicker in the
greater curvature than in the lesser (hence uniformly shallower than in
the lesser arc in the three species mentioned above). The circular layer
of the greater curvature wall in Noctilio is distinctive in being only
slightly thicker than the lesser wall, whereas the longitudinal layer is
markedly increased in depth in the medial and terminal portions of the
greater curvature.

The stomachs of Glossophaga, Carollia, Sturnira, and Artibeus all pos-
sess some degree of sacculation of the fundic caecum, with a greater or
lesser degree of reduction of the muscular wall within this portion of the
stomach. The musculature of the greater curvature of the fundic caecum
is thicker than the facing wall on the lesser curvature in Glossophaga and
Carollia, whereas these two surfaces are essentially equal in depth and thin
throughout in Sturnira and Artibeus. The terminal pyloric portions in these
four bats are always thicker than the proximal wall and generally thickest
on the greater curvature, although the difference is not noticeable in
Carollia.

The functional significance of much of this irregularity in the depth
of muscle layers is unclear, although conjectures regarding answers to such
questions follow where appropriate. In the absence of any pertinent physio-

712 The University Science Bulletin

logical information concerning gastrointestinal tracts of bats, it is difficult
to explain such extensive variability.

It would seem likely that differential muscle development that results
in greater masses of circular fibers on the greater curvature would likely be
a response to increased food intake, which would demand additional basal
support in the gastric sac to accommodate the increase in food mass. Al-
though marked modifications in this condition of differential musculature
are found to exist in localized portions of the stomach in many species,
most still express the condition, at least in the terminal portion of the
stomach. Exceptions to this trend, although rare, do exist and much addi-
tional work is needed to clarify the role of these conditions.

Especially in Plecotus and less so in Pizonyx, some marked uniformity
in depths of corresponding portions of the lesser and greater curvatures is
apparent. Musculature of the fundic caecum in each of these two species
is moderate and uniform in depth, except for reduction at the extreme
apex; thickening in the pylorus is pronounced in Piznoyx and slight in
Plecotus. Plecotus and Rhynchonycteris most closely approach the condition
of equal thickness of the muscle wall throughout the stomach among
the carnivorous and herbivorous species examined. The muscular wall in
the stomach of Rhynchonycteris is remarkably uniform, Plecotus somewhat
less so; all remaining species show some local thickening, usually in the
pyloric endpiece. It is suggested that the condition of nearly uniform depth
of the stomach musculature in these two species is a relatively generalized
condition.

Decrease in depth of the caecal wall is apparent among herbivorous kinds,
whereas this region remains deep or even increases in depth over other
areas in obligate insectivores and carnivores. It would seem likely that
gastric volume is of greater importance in herbivorous kinds, hence loss
of musculature in the caecum permits an increase in elasticity. On the
other hand, prominent musculature in the proximal portion is of greater
advantage to insectivorous kinds in increasing the area over which large,
rythmical contractions take place, hence the physical mechanism of gastric
mixing. It is interesting that Ph\llostomus discolor, a probably omnivore,
approaches the condition of equal muscular depth throughout the stomach
to a greater degree than any other phyllostomatid examined, with only
slight and nearly symmetrical thickening in the terminal portion of the
stomach: this suggests a retention of generalization in this feature in
Phyllostomus and perhaps in other phyllostomatines.

Quite localized decrease in caecal musculature at the extreme apex
(owing to decrease in depth of the circular layer), as occurs in Rhynchonyc-
teris, Noctilio, Phyllostomus, Plecotus, and Piznoyx, probably reflects a
compromise to obtain increasing flexibility and yet remain the substantial
musculature required for gastric mixing in carnivorous bats.

Studies of Stomachs of Selected American Bats 713

Asymmetrical pyloric thickening, discussed previously, is noted to be
widespread among bats examined. Where present this does not necessarily
correspond with, nor is it restricted to, those kinds with similar asymmetry
of the nearby pyloric sphincter, nor is it correlated with relative size of the
sphincter. The significance of this feature remains obscure, and its wide-
spread occurrence suggests that studies of gastric motility and the physio-
logical role of the pyloric sphincter in bats need be undertaken to help
resolve these questions.

As regards the outer musculature in Desmodus, the extensively bundled
fibers within both outer layers provide a suitable mechanism to explain
prominent expansion of the stomach to hold the large quantities of blood
taken in individual feedings. Some suggestion of increased depth of the
greater curvature over that of the lesser curvature is to be found near the
gastroesophageal junction in Desmodus, the significance of which, as in
other bats revealing this feature, remains obscure. Although the bundles
of circular fibers are variable in size, the total depth of the outer wall is
fairly uniform throughout the fundic caecum in Desmodus, suggesting
symmetrical expansion of the caecum upon filling. The mechanism to
allow extensive expansion is not, apparently, reduction of muscular tissue
in depth, but an increase in the elastic portion of the wall by way of in-
creased complements of areolar connective tissue.

An index for determining differences and degree of variability in ar-
rangement of circular muscle fibers was taken to be the extent of bundling
of this layer as viewed in a mid-longitudinal section of the stomach. This
condition, as observed in mid-longitudinal section, is restricted for the
most part to the lateral and lower wall of the fundic caecum, but in some
is present up to the medial area of the greater curvature. There are,
however, several notable exceptions to this trend. The stomach of Desmodus,
as discussed previously, illustrates extensive bundling throughout all of the
fundic caecum, unique in this regard among bats examined. Most species
examined had some degree of bundling of the circular layer in the fundic
caecum, usually in the lower portion and within the greater curvature.
The only exception to this trend was found in Natalus, in which the
bundling present in a mid-longitudinal section was a small area just
proximal to the pyloric bend. The function of gathering of circular fibers into
bundles appears to be a gain in degree of elasticity of the stomach, expressed
to some degree in the fundic caecum or greater curvature (or both) of most
species, especially the phyllostomatids and Molossus. The apparent lack of
this feature in the fundic caecum of Natalus is particularly interesting in
light of the unique feature in this bat of increased muscular depth within
the fundic caecum in comparison to the remainder of the stomach. This
suggests that the stomach of Natalus is relatively non-plastic, with little

714 The University Science Bulletin

capacity for expandability of the fundic portion. Whether or not this condi-
tion restricts the relative amount of food consumed in comparison to other
species cannot at this time be concluded, but seems unlikely to be the case
in light of the rather extensive cardiac vestibule near the esophagus in
Natalus, providing some increase in gastric volume.

Relatively little reference has thus far been made to the longitudinal
layer of the stomach, because it consists of only a thin sheet (in comparison
to the overlying circular complement) throughout most portions of stomachs
of nearly all of the 13 species studied. However, in at least three species of
insectivorous bats {Pteronotus, Noctilio, Natalus), representing three dis-
tinctive families, the longitudinal layer is prominently developed locally
throughout part or all of the fundic caecum. In the case of Natalus it is
as deep as, or deeper than, the thick circular layer throughout most of the
caecum. Where such hyper-development occurs, the longitudinal layer of
the caecum becomes several times the depth of the remaining portion of
this layer, which as noted is usually extremely thin. Except in Pteronotus,
the omnivorous Phyllostornus discolor reveals the best development of the
longitudinal layer among the primarily frugivorous phyllostomatid group,
and some correlation between this condition and carnivorous habits in bats
seems apparent. An increase in longitudinal muscle fibers would increase
the strength of the shortening component for a rhythmical-segmentation
contraction, which would presumably increase the effectiveness of mixing
caecal contents with that of the midportion of the stomach. This suggestion
must, of course, be tested by examining additional species with similar
feeding habits, and examining the mode of gastric motility in these mam-
mals.

Gastric Mucosa and the Glands of Brunner

orientation of folds

In general, only two patterns of directional orientation of mucosal folds
were observed in the 13 species examined. The first condition, that of
longitudinal orientation of parallel folds, was found in the tubular portions
of all stomachs irrespective of the feeding type represented. Longitudinal
orientation to a greater or lesser degree was also observed in the fundic
caeca and cardiac vestibules of all insectivorous species, the omnivorous
Phyllostornus, the piscivorous Pizonyx, and sanguivorous Desmodus. Within
the non-tubular portions of stomachs of several species (including Phyllo-
stornus, Glossophaga, Plecotus, Pizonyx, Molossus, and Desmodus), which
have mostly longitudinal orientation of folds, there occurs slight to moderate
wrinkling (or zig-zag arrangement) of the rugae, although no true inter-
digitation of the folds is to be found. This condition is lacking in several
other carnivorous species, most of which (Rhynchonycteris excepted) also
reveal relatively thick musculature in the fundic caecum. The condition

Studies of Stomachs of Selected American Bats 715

of wrinkling of the rugal folds appears related to increased ability to
stretch the caecal portion of the stomach. Wrinkling of the mucosal folds
provides for additional muscular support in the form of folded sheets of
muscularis mucosae to compensate for the extremely dilated outer muscu-
lature that would result upon marked distention of the stomach. It is not
surprising to find that this feature is lacking in those species with deep
caecal musculature, which probably restricts the degree of expansion of
the caecum.

The second condition of orientation of mucosal folds is that of trans-
verse anterior-posterior orientation. This situation occurs 1) infrequently
among a few species of insectivorous kinds (here found restricted to
infrequent, short transverse folds in brief areas of the fundic caecum,
usually the upper wall), 2) in the fundic caecum of Glossophaga and
Carollia as narrow transverse ridges along the sides, and 3) in an extreme
condition of highly complex, heavily interdigitating ridges in the non-
tubular fundus of Artibeus and to a lesser degree that of Sturnira. A possible
function of these interdigitating folds, that of helping to secure food ma-
terials within the fundic caecum for extended periods of time, has been
previously discussed for Artibeus. An alternative, or at least supplementary,
hypothesis would be to suggest that these folds provide for increased elas-
ticity in addition to the already thin outer wall of the fundic caecum to
assist in accommodation of large meals of fruit such as are known to be
taken by Artibeus and Sturnira.

LOCAL REDUCTIONS IN MUCOSAL DEPTH

Two conditions of local modification in mucosal depth in comparison
to the remainder of the stomach seem worthy of mention, as they are
widespread among the species examined. In the stomachs of all species
studied, except for Carollia and Sturnira, the gastric mucosa is clearly
deeper in the lesser than in the greater curvature. This condition does not
appear correlated with muscular depth as it is found to exist in species
with an hypertrophied wall along the lesser arc and along the greater arc,
as well as in species with uniform or nearly uniform muscular depth, and
is usually restricted to within the zone of fundic mucosa. Perhaps no rela-
tionship exists between mucosal depth and digestive efficiency or strength
of the gastric wall, although this seems unlikely. Evidence that the condi-
tion is best developed and, therefore, most important in carnivorous kinds
is hardly conclusive, and thorough study of these patterns should be under-
taken in other kinds of mammals with diverse food habits.

A second condition is that of reduced depth of the mucosa within the
fundic caecum in comparison to the remainder of the stomach wall. Gen-
eralized reduction throughout the fundic caecum occurs in only one species,
Glossophaga soricina, whereas localized reduction only at the lateral apex

716 The University Science Bulletin

of the caecum (usually slight) occurs in the stomachs of Rhynchonycteris,
Phyllo stomas, Pkcotas, Pizonyx, and Molossus. Reduced glandular depth
in the caecum is generally the result of reduction in number of all types of
cells present, but particularly the zymogenic cells. It would appear that
this condition is again part of the general hypertrophy observed in other
layers at the apex of the caecum in these species, related to increased local
elasticity. The herbivorous species and the insectivorous and carnivorous
species with thick musculature within the caecum do not demonstrate this
condition of local reduction of mucosal depth. They have other mechanisms
for increasing the volume of the caecum or have caeca that are relatively
non-elastic.

distributions of types of gastric mucosa

Extensive variation between species and surprisingly little variation
within species was found in distribution of most of the types of gastric
mucosa. The zone of cardiac glands at the gastroesophageal junction is a
relatively narrow band in all species examined, but approaches a condition
of being a distinctive element of the mucosa in Pizonyx vivesi, unequaled
in this feature by any other bat examined. The cardiac glands of Pizonyx
are extremely large and occupy the entirety of the short cardiac vestibule.
These glands are composed entirely of mucus-producing cells, and are pre-
sumed to be a source of lubricating substance applied to the food material
as it enters the stomach (Bensley, 1903). All species examined possessed at
least a few of these glands, but their extensive development in Pizonyx
suggests that the food consumed by this species requires an especially
abundant coat of mucus. Whether the fish consumed by Pizonyx present
in especially abrasive mass for the stomach is unknown. It is of some in-
terest to reflect on the fact that the greatest degree of specialization at the
gastroesophageal junction among bats is to be found in two fish-eating
kinds studied, the other being Noctilio leporinus. It is indeed apparent,
however, that these specializations are of a wholly different order — hyper-
development of cardiac glands in Pizonyx and an anatomical cardiac
sphincter lined with squamous epithelium in Noctilio. Although seemingly
unrelated anatomically, it can be argued that these specializations produce
similar conditions at the entrance of the esophagus into the stomach, that
of a smooth or heavily mucus-coated surface. The import of this feature
as it relates to fish-eating habits should be investigated further.

Although not unique in spatial distribution within the stomach, the
cardiac glands of Natalus are large in size and are unusual in that not many
occur beneath the stratified squamous epithelium at the extreme upper
limit of the cardiac vestibule. These glands are nearly identical morpho-
logically to (and about the same size as) the glands of Brunner in the
duodenum of this bat. These two types of glands are also identical in

Studies of Stomachs of Selected American Bats 717

reactivity to two histochemical procedures for mucins (unique among bats
examined). The unique similarity in the two glandular types in Natalus
should be investigated further. Study of the ultrastructire of these cells,
along with those of the glands of Brunner, might provide additional evi-
dence regarding the apparent parallelism in structure, as well as shed light
on their poorly understood physiological roles in the mammalian stomach.
The pyloric glands are easily distinguished from the cardiac glands in
Natalus and the marked similarity of the latter to Brunner's glands is in-
deed unusual.

Myrcha (1967) examined the distribution of zones of gastric mucosa
in 27 species of insectivores representing three families. He concluded that
the often-observed elongation of the terminal portion in many species was
a direct result of increased food consumption, and that an extension of the
transitional zone between fundic and pyloric mucosa was a consequence
of this elongation. Among the bats studied, general correspondence was
found between elongation of the terminal portion and elongation of the
zone of transitional mucosa except in the Noctilionidae and Phyllostomati-
dae. In the latter groups the pattern not only breaks down but shows some
reversals of the trend among the included species.

An example of this breakdown may be seen in two obligate insectivorous
bats, Pteronotus parnellii and Noctilio labialis. Both have relatively short
terminal portions of the stomach, yet have extremely long transitional
zones, apparently developed at the expense of the pyloric mucosa. This
pattern becomes further complicated upon examination of the transitional
areas of Phyllostomus and Artibeus, two species with somewhat elongate
terminal portions, yet extremely narrow transition zones. The distribution
of pyloric mucosa may be an important controlling influence in determining
the amount of pylo-fundic transitional mucosa present in bats, because
those species with extremely broad transitional zones (for example, Rhyn-
chonycteris, Noctilio, and Pteronotus) have small complements of pyloric
mucosa, and, conversely, those species with extremely narrow transitional
zones (for example, Phyllostomus, Artibeus, and Pizonyx) have rather
extensive regions of pyloric mucosa, as compared with their closest rela-
tives. It is thus apparent that the phenomenon of coelongation of the
terminal stomach and zone of transitional mucosa, as outlined by Myrcha
(1967) for the Insectivora, is not found among the frugivorous Micro-
chiroptera. Specific information concerning the relationship of feeding habits
to the physiological significance of the pyloric mucosa is needed to help
resolve this question.

The gastric glands of bats are, for the most part, similar to those of
other kinds of mammals in types of cells present and in the distributional
order of cells within the tubules from one end to the other. However, the

718 The University Science Bulletin

proportion of individual tubules occupied by various types of cells, and
the depth of various mucosal and submucosal layers differ from the arrange-
ment in many other mammals. The gastric glands generally may be re-
garded as shallow in comparison to those of rodents and insectivores,
animals of similar size. The cause or effect of the shallow nature of these
glands appears primarily to be a loss of parietal and mucous neck cells in
medial areas of the tubules. The result of this loss is greater equality in
length among regions occupied by zymogenic, parietal, and mucous neck
cells in bats, because the zones of parietal and mucous neck cells tend to
be much longer in rodents, insectivores, and many other mammals with a
mucosal membrane. In addition, there is a corresponding reduction in the
amount of submucosal connective tissue throughout much of the gastro-
intestinal tract of bats, resulting in a presumably highly elastic intestine
and equally elastic local areas of the stomach.

As regards the distribution of pyloric mucosa, no clear-cut patterns
emerge with respect to similarities among species with similar food prefer-
ences. It is evident however, that frugivorous leaf-nosed bats, as well as
the omnivorous Phyllostomus, tend to have extensive complements of these
glands, with rather narrow transitional zones. This condition, although
much less apparent, is found also in Pizonyx. Only Natalus among the
insectivorous bats examined has pyloric glands in an unusual length of the
stomach. Because pyloric glands histogenetically represent a type of gastric
mucosa in which the cells responsible for secretion of materials engaged in,
or complementary to, proteolytic breakdown are replaced by mucous pro-
ducing types, it is understandable that these highly active digestive cells
are less widespread in distribution in frugivorous bats. Several species of
carnivorous bats (Rhynchonycteris, Pteronotus, and Nocttlio) reveal a near
loss of true pyloric glands with marked proximal expansion of parietal cells
toward the pyloric sphincter. A consequence of the elongation of the
stomach in frugivorous species has been an extension of the pyloric zone
only, rather than corresponding expansion of the transitional zone as seen
in most insectivorous kinds. Phyllostomus has acquired a stomach con-
figuration and muscular wall more like those of carnivores, and yet, has
retained the zonation of gastric mucosa most often observed in frugivorous
species. The rather extensive distribution of pyloric mucosa in Natalus,
the stomach of which is unusual in many features, is unique among in-
sectivorous species examined, a probable consequence of notable elongation
of the terminal portion. Glossophaga, an apparent "intermediate" in the
true sense of the term, more closely approaches the "insectivorous" condi-
tion in distribution of pyloric and fundic mucosa than does Phyllostomus.

These conclusions do not refute the possibility that relatively numerous
pyloric glands at the proximal end of the stomach may be of some con-

Studies of Stomachs of Selected American Bats 719

siderable physiological advantage in one or more of the bats studied, par-
ticularly the frugivorous species in which they are so abundant. Numerous
physiological roles have been suggested for the mucous producing cells of
these glands, including their being sites of anti-autodigestion enzymes and
neutralizes for pepsin among other things. However, until the exact physio-
logical role of these glands is understood, it is difficult to draw meaningful
conclusions regarding their relative abundance in the stomachs of bats.

Fundic Mucosa

The fundic mucosa comprises the most morphologically specialized and
presumably the most physiologically specialized portion of the gastric
mucosa in bats. Fundic glands always occupy the majority of the internal
surface area of the stomach and are characterized by the presence of both
parietal and zymogenic cells within the tubules. That portion of the gastric
mucosa that contains numerous parietal cells but no zymogenic cells is here
not considered fundic, but rather transitional mucosa, mostly for the sake
of convenience. It is therefore to be assumed that when distributions of
parietal cells in the stomach as a whole are considered, the transitional
mucosa is included as a part of these distributions.

The zymogenic cells are considered to be the site of secretion of the
enzyme pepsin, activated to proteolytic action in the presence of HC1.
Although the greatest numbers of chief cells were found in several insec-
tivorous species, other insectivorous and carnivorous bats revealed smaller
complements of chief cells than did frugivorous species. There is no clear
relationship, therefore, between food habits and relative numbers of zymo-
genic cells. The substantial reduction in the number of these cells in
Desmodus is taken to be a consequence of general hypertrophy of fundic
glands in this bat. The striking depletion of zymogenic cells and especially
parietal cells in Desmodus elicits interesting questions concerning the ability
of Desmodus to assimilate food material, particularly in light of the high
percentage of solids found to be utilized metabolically by Desmodus (Wim-
satt and Guerriere, 1962).

It is clear that wherever zymogenic cells are notably abundant, the dis-
tribution of parietal cells is correspondingly limited to a narrow portion
of the tubule. The relative proportion of the gastric tubule occupied by
chief cells seems to be a stronger controlling force afTecting the numbers
of parietal cells present than the varieties of foods taken in the diet. Frugiv-
orous, insectivorous, and piscivorous species were observed with both re-
stricted and extensive complements of parietal cells in the fundic and transi-
tional areas. Such extensive variation remains unexplained. Allison (1948)
has suggested that distribution and numbers of parietal cells may remain
static in the gastric mucosa of insectivores, and Rouk (1968) observed no dif-

720 The University Science Bulletin

ferences in the distributions of parietal cells in active and hibernating Myotis
velifer. These facts suggest that differences between species in complements
of at least some cell types in the gastric mucosa is, in all likelihood, real and
definable interspecific variation, the significance of which requires addi-
tional investigation for adequate explanation.

It will be noted that in three species examined (Rhynchonycteris, Noctilio,
and Pteronotits), nearly all of the gastric mucosa contains some parietal
cells and these cells are, for the most part, extremely small and numerous.
Perhaps large numbers and extensive distribution of parietal cells in these
species compensates for small cell size to provide the needed concentration
of HC1.

As in the case of zymogenic cells in bats, no clear-cut patterns emerge
with regard to relative abundances of parietal cells in comparison to types
of foods taken. The especially numerous parietal cells found to occur in
Artibeus may be related in some way to the rapid food transport in this
species, with a substantial flow of HCl being required to counterbalance
the presumed short exposure to digestive fluids. This conclusion is only a
supposition and requires further analysis. It may be that parietal cell
abundance is related to rate of food passage in other species, or that the
abundance of other cell types in the gastric mucosa is also related to time
of exposure of the food to gastric digestion, although little information is
available on this subject.

With the exception of Artibeus, in which the parietal cells are both
abundant and of large size, there is a trend among the bats examined to
suggest that these cells are relatively large in species in which they are few
in number; conversely, the cells are smaller in those bats in which they are
more numerous. To my knowledge, no such suggestion of an inverse
correlation between size and abundance of parietal cells is to be found
for any other kind of mammal. It is of interest to note that Allison (1948)
observed the parietal cells of species of the Insectivora to be progressively
smaller from the lower to the upper boundary of their distribution within
an individual fundic tubule, a condition found to occur only in Rhynchonyc-
teris among the 13 species of bats examined herein. At least 10 of the
remaining 12 species (the condition is unclear in Carollia and Desmodus)
revealed the reverse situation — the largest cells were toward the surface
of the mucosa. The parietal cells closest to the surface of the gastric lining
are presumed to be the "oldest." Differences in growth capacity and physio-
logical activity in parietal cells in various species of bats, as well as between
major subgroups of mammals, seems probable.

It is apparent that greater local variation within the stomach in distri-
bution of parietal cells and zymogenic cells, as well as in depth of gastric
pits, is found among insectivorous or partially insectivorous bats as a group,

Studies of Stomachs of Selected American Bats 721

but not among frugivorous, piscivorous, or sanguivorous kinds. The fundic
mucosa of insectivorous kinds is consistently better developed in the lesser
curvature than in the greater curvature; in terms of development longitu-
dinally it usually reaches its maximum within the middle portion, decreas-
ing in richness of cell types in either direction laterally. The reduction
within or at the apex of the fundic caecum in many species appears also
to be related to increasing the elasticity of the stomach wall.

Few comments have thus far been made specifically regarding relative
amounts of gastric mucus produced via the combined secretions of pyloric
and cardiac glands, surface epithelium, cells of the gastric pits, and the
deeper mucous neck cells. Considerable variation exists among species in
depth of the surface epithelial layer, as well as depths of the gastric pits
within the fundic mucosa. Those species with relatively deep gastric pits
or particularly thick surface epithelium might theoretically produce the
most epithelial mucus. The three primarily frugivorous species of the
Phyllostomatidae along with Phyllostomus have both extensive pyloric
mucosa and relatively deep surface epithelium, suggesting some probable
advantage to abundant mucous secretion in relation to frugivorous habits.
The surface epithelium of Glossophaga is comparable in depth to most
insectivorous species. The extremely well-developed gastric epithelium of
Desmodus may be traceable to its close phylogenetic relationship with the
Phyllostomatidae; however, it is most likely related in some way to the
extensive capacity for absorption of water from the stomach into the
vascular network, as comparatively little surface mucus was observed in
Desmodus in histochemical tests for mucopolysaccharides.

The gastric pits of the fundic tubules, formed by an apparent involution
of the surface epithelium, are unusually well developed in Phyllostomus.
This feature, in combination with the extensive complement of deep, mas-
sive pyloric glands and notably thick surface epithelium, presents Phyllo-
stomus as an unusually heavy producer of gastric mucus. The extensive
quantities of food known to be consumed in rather short periods of time by
species of Phyllostomus would demand high levels of enzyme and acid.
Rather complete decomposition of materials such as bone (Dunn, 1933)
must produce gastric environments that threaten the mucosal lining unless
a deep mucous coat is present. Finally, as regards gastric mucous produc-
tion, several species had differential increase in the depth of the gastric
pits within the fundic caecum in comparison to the remainder of the
stomach. This increase is most often accompanied by corresponding re-
duction in numbers of other cell types in these glands; the advantage of
having relatively more surface mucus in this area than in the remainder
of the fundic portion is unclear.

There exists a trend among certain groups of mammals for a relatively

722 The University Science Bulletin

greater distribution of the glands of Brunner in the upper duodenum of
herbivorous species as opposed to carnivorous species. However, an im-
pressive group of exceptions to the trend may be found without exhaustive
search, as in bats. One general statement regarding the distribution of
these submucosal glands in bats can be made: these highly convoluted
elements are always restricted to the immediate vicinity of the pyloric
sphincter, usually in a small mass, with little or no representation in the
form of smaller, isolated clumps inferior to the main mass, as is common
in other kinds of mammals. Some variation exists, however, in relative
distribution of these glands among bats, in the morphology of glands, and
in at least two species in degree of abundance in the terminal portion of
the stomach. Two of the three frugivorous species, Artibeits and Carollia,
have Brunner's glands so reduced in distribution as to be nearly lost; those
of Sturnira are only moderate in abundance in the duodenum, although
they are extensive in the pyloric stomach proximal to the pyloric sphincter
(shared apparently only with Molossus). Among the phyllostomatids ex-
amined, the greatest masses of Brunner's glands were observed in Pteronotus
and Phyllostowus, two species that likely consume insects in quantity.
As the functional role of secretions of Brunner's glands remains obscure,
it is difficult to comment on the meaning of variation in their distribution.
It would appear that certain frugivorous kinds have secondarily lost most
of these glands. Carleton (1935) noted that the location of the common
bile or pancreatic duct in some mammals may determine the limit of dis-
tribution of Brunner's glands into the duodenum, although there are excep-
tions. The common bile duct enters the duodenum only a few millimeters
distal to the pyloric sphincter in bats, and may be related in fact somehow
to distribution of the glands of Brunner. However, it must be noted that
in Molossus the glands occur slightly beyond the entrance of the common
bile duct into the duodenum, and, in fact, an ampulla is formed at the
junction as a result of invasion of Brunner's glands into the duct. The
general consensus now is that the glands of Brunner function to protect
the lining of the initial portion of the duodenum from the damaging
action of acid material escaping from the stomach (Grossman, 1958). In
this respect, the near loss of Brunner's glands in a species such as Artibeus
with its rather extensive number of large parietal cells is indeed perplexing.
The invasion of the glands into the pyloric portion of at least two distantly
related species, the food habits of which are entirely different, suggests
that this condition might be widespread among certain families. Probably
it confers additional protection against damage to the pyloric mucosa of
Molossus and also especially to Sturnira, although the particular need for
such a mechanism in these species is not readily apparent.

In concluding the discussion of gastric structure in relation to feeding,

Studies of Stomachs of Selected American Bats 723

it seems appropriate to suggest that control over morphology may be two-
fold, one factor affecting the general shape and relative proportions of
various regions of the gastric sac and the other controlling the distribution
and abundance of elements of the gastric mucosa. Each factor acts sepa-
rately in gastrointestinal evolution to accommodate particular types or
quantities of food. In this regard, it appears that the quantity of food
consumed may be an especially important criterion in the evolution of
gastric structure. Elongation of the stomach has been a particularly im-
portant event in the acquisition of frugivorous habits in bats, whereas most
insectivorous and all piscivorous species examined have retained a relatively
simplified stomach. Those species thought to be non-obligate feeders and
that take substantial quantities of both animal and plant matter, were
found to be somewhat intermediate in the structure of the stomach. Gastric
structure as observed in bats does not usually correspond with the conditions
found to exist in the Insectivora, although Natalus and Rhynchonycteris
closely parallel the Insectivora in several features. In general, evolution
away from insectivorous feeding in bats has resulted in noteworthy spe-
cialization in one or more structural features of the stomach.

SYSTEMATIC RELATIONSHIPS AS REVEALED BY
GASTROINTESTINAL STUDIES

Although the stomach of each of the 13 species examined was clearly
distinctive in one or more major features, clear-cut relationships were
found within families as well as among closely related families. Modifica-
tion of gastric structure to accommodate and assimilate foods other than
insects has led to marked changes in external configuration of the stomach.
Upon further investigation, however, various features of internal gastric
structure emerge to reveal more precisely, relationships among various
groups of bats. Because the Phyllostomatidae represents the family with
the most genera examined, and also has the greatest intrafamilial variability
in food habits of included species, it presents the most interesting body
of information.

Relationships of bats with the Insectivora, both phylogenetic and as
regards food habits, bring to mind consideration of which species of bats,
if any, seem to conform most closely to gastric structure as presently known
for the Insectivora. The stomach of Natalus stramineus, as noted previously,
closely resembles those of soricid insectivores in stomach configuration, dis-
tribution of mucosal elements, and particularly in external and internal
morphology of the terminal or tubular portion. In addition, the stomachs
of Rhynchonycteris and Plecotus also are generalized externally and internally
and are strikingly like those of several talpid and erinaceid insectivores,
respectively. These relationships suggest a generalized condition in early

724 The University Science Bulletin

chiropterans, most, if not all, of which were probably insect-feeders. The
specializations observed in other insectivorous bats such as Molossus and
Noctilio are modifications of the relatively primitive plan seen in the Ves-
pertilionidae, Natalidae, and Emballonuridae.

Within the Vespertilionidae, all species thus far examined reveal marked
symmetry of the stomach (esophageal junction located nearly midway along
the lesser curvature) and a short, slightly recurved pyloric tube. The list
of examined species showing these features includes Myotis vehjer, Plecotus
townsendii (plus the unnamed species of Plecotus discussed by Fischer,
1909), Antrozous pallidas, and Pizonyx vivesi. Eisentraut (1950) included
a drawing of the stomach of Eptesicus serotinus that, although appearing
somewhat distended as do most of the stomachs illustrated by that author,
is in general correspondence with those of other vespertilionids. Pizonyx
and Antrozous represent the greatest deviants from the basic plan of overall
stomach configuration. The stomachs of both are more rounded than those
of other vespertilionids, and Pizonyx differs further in its atypical distribu-
tion of mucosal zones. Hence, Pizonyx is even less like Myotis, with which
it recently has been merged on the basis of karotypic evidence (Baker and
Patton, 1967) than the other vespertilionid genera that have been studied.

The stomach of Natalus stramineus is much like that of Plecotus town-
sendii in general topography and in distributions of zones of gastric mucosa,
but differs in degree of elongation of the pyloric tube and cardiac vestibule.
It is difficult to determine whether similarity in structure is due to parallel
adaptation to similar feeding characteristics or close phylogenetic association,
or both. It has been suggested that the Natalidae may be systematically
closely associated with the vespertilionids through the genus Kerwoula
(see Dobson, 1878, for a discussion of relationships), and the results of
gastrointestinal studies support, to some degree, the contention of close
relationship between the two families.

Although highly variable among themselves, the genera of the Phyllo-
stomatidae examined have several features in common that tend to dis-
tinguish them as a group from the other families considered in this study.
In all species examined, the muscular portion of the pyloric sphincter tends
to be narrow in relation to size of the stomach, or may be partially lacking
(Pteronotus) or nearly completely lost (Sturnira). An extremely narrow
sphincter was also found to occur in Desmodus. Also among phyllostom-
atids, with the possible exception of Pteronotus, there occurs some degree
of dilation of the dorsolateral terminus of the fundic caecum, although this
expansion is not restricted to the dorsal surface.

Recent investigations of the systematic relationships of the nominal
subfamilies Stenoderminae and Sturnirinae have led most investigators to
the conclusion that Sturnira should be considered a stenodermine. The

Studies of Stomachs of Selected American Bats

725

B

F

Fig. 53. A diagram illustrating the gradual increase in specialization in stomachs of some
phyllostomatid bats. Pteronotus parnellii (A); Phyllostomus discolor (B); Glossophaga soricina
(C); Carollia perspicillata (D); Stiirnira lilium (E); Artibeus litwatus (F). Stomachs art-
drawn to scale.

726 The University Science Bulletin

results of my gastrointestinal studies strongly support the proposal that
Sturnira should be placed in the Stenoderminae.

In addition to the specific features that tend to group the phyllostomatids
(including the Desmodontinae), it is also of interest that certain genera
in the family resemble one another or a representative of another family
in unique ways. As noted, the stomachs of Artibeus and Sturnira (Fig 53)
are similar, resembling each other in gastric morphology to a much greater
degree than either resembles representatives of the other subfamilies studied.
Additionally, Phyllostomus and Glossophaga are notably similar, whereas
Carollia is more or less intermediate between this pair and the previously
discussed pair, possibly resembling Glossophaga more than the other genera.
The stomach (described by Rouk, 1968) of another glossophagine bat,
Leptonycteris sanborni, closely resembles that of Glossophaga in the structure
of the terminal portion and differs generally only in the size of the cardiac
vestibule (larger in Leptonycteris). Although each is specialized in some
ways, Pteronotits and Noctilio are strikingly similar in several features,
particularly the distributions of parietal and zymogenic cells, morphology
of the gastric glands, and general external topography, suggesting the possi-
bility of closer systematic relationship than has been previously supposed.

A systematic revision of the subfamily Chilonycterinae recently was
completed by James D. Smith. The results of my study indicate that this
group likely deserves familial recognition, complementing his findings based
on other characters.

It is of interest to note that if one arranges six species of phyllostomatids
examined (excluding Desmodus), each of which represents a presently
recognized subfamily according to phylogenetic arrangements such as those
of Hall and Kelson (1959), Koopman (1967), or Simpson (1945), there is
a gradual transition from insectivorous feeding (Pteronotus) to frugivorous
habits (Artibeus) with variously omnivorous kinds interposed in sequence
(see Fig. 53). Corresponding closely with this change in food habits is a
progressive alteration of gastric structure from the simplified condition of
Pteronotus to greater elongation and compartmentalization of the stomach,
culminating in the condition found in Sturnira and Artibeus. Trends in
other morphological features complement the trends in specialization ob-
served in stomachs of phyllostomatid bats. The sequence presented in
Figure 53 is not intended to imply that one species is necessarily older than
another in an evolutionary sense, but that modifications in food habits
from the presumed early insectivorous condition have resulted in progressive
modifications in the morphology of the stomach as well. The desmodontines
are clearly so highly specialized in gastric morphology that it is difficult to
relate them to other North American groups, although close relationship
with the Phyllostomatidae has been shown by Forman et at. (1968) using
other characters.

Studies of Stomachs of Selected American Bats 727

Stomachs of molossids thus far examined reveal that this family consists
of relatively specialized insectivorous bats with stomachs that are clearly
different from those of other insectivorous species. The stomach of Tadarida
brasiliensis (examined by Rouk, 1968) differs from that of Molossus ater
only slightly, containing the same degree of elongation of the terminal
stomach and a prominent mid-longitudinal notch in the greater curvature.

The histochemistry of deeper cells of the cardiac and pyloric glands of
bats reveals trends difficult to interpret in light of morphological patterns
observed; yet the results are of interest with respect to interfamilial rela-
tions. It will be recalled that the results of the PAS procedure showed less
reactivity within the basal arc or bulbular portion than in the upper, neck
region of the cardiac and pyloric glands of all species, except for an equiva-
lent reaction from both glandular areas in the case of Molossus. When Max-
well's Alcian blue technique was employed for acid mucopolysaccharides in
these glands, the same pattern of differential staining as occurred with PAS
was obtained. However, employment of the aqueous Alcian blue (Steed-
man's method) yielded a similar pattern of results to those of PAS and
Maxwell's in all species but Sturnira and Artibeus, which showed little
reactivity at all, and a reversed condition (in which the basal cells in the
cardiac and pyloric glands stained considerably darker than the upper
cells) in a group including Natalus, Plecotus, Pizonyx, and Molossus. Al-
though the physiological import of such results is unclear, the consistency
of this differential staining in both cardiac and pyloric glands among repre-
sentatives of the Natalidae, Vespertilionidae, and Molossidae supports the
present placement of these families in the superfamily Vespertilionoidea.
Winge (1941) suggested that the Molossidae may have arisen from primitive
vespertilionids (based on similar structure of the femur and ulna), and
the close relationship of the latter group with the Natalidae has been
previously discussed. My results suggest that a broader application of more
specific histochemical procedures upon the gastric mucosa might prove useful
in resolving questions regarding relationships among higher taxonomic
categories of bats.

LITERATURE CITED

Allison, A. C. 1948. The stomach in South African Insectivora, with notes on the organiza-
tion of mammalian gastric glands. J. Anat. 82:249-261.

Arata, A. A., J. B. Vaughn and M. E. Thomas. 1967. Food habits of certain Colombian bats.
J. Mammal. 48:653-655.

Baker, R. }., and J. L. Patton. 1967. Karyotypes and karyotypic variation of North American
vespertilionid bats. J. Mammal. 48:270-286.

Bensley, R. R. 1903. The structure of the glands of Brunner. The Decennial Publ., Univ.
Chicago, 10:279.

. 1905. The cardiac glands of mammals. Amer. J. Anat. 2:105-156.

Bloedel, P. 1955. Hunting methods of fish-eating bats, particularly Noctilio leporinus. J.
Mammal. 36:390-399.

728 The University Science Bulletin

Botha, G. S. M. 1958a. Mucosal folds at the cardia as a component of the gastrooesophageal

closing mechanism. British J. Surg. 45:569-580.
. 1958/'. Histological observations of the gastrooesophageal junction in the rabbit. J.

Anat. 92:441-446.

1958<~. A note on the comparative anatomy of the cardio-oesophageal junction. Acta

Anat. 34:52-84.
Carleton, A. 1935. Distribution of Brunner's glands in the duodenum of mammals. Proc.

Zool. Soc. London, pp. 385-390.
Clara, M. 1940. Untersuchungen iiber den farberischen Nachweis des Schleimes in den

Driisenzellen beim menschen. Z. mikr. anat. Forsch. 47:183-246.
Cuvier, G. 1805. Organes de la digestion. Lecons d'anatomie comparee. Paris, Volume 4

(see pp. 31-32).
Davenport, H. A. 1960. Histological and Histochemical Techniques. W. B. Saunders, Phila-
delphia. xxii+401 pp.
Dearden, L. C. 1966. Histology of the gastro-esophageal junction in certain microtine rodents.

J. Mammal. 47:223-229.
. 1969. Stomach and pyloric sphincter histology in certain microtine rodents. J.

Mammal. 50:60-68.
Dobson, G. E. 1875. Conspectus of the suborders, families, and genera of Chiroptera arranged

according to their natural affinities. Ann. Mag. Nat. Hist. ser. 4, 16:345-357.
. 1878. Catalogue of the Chiroptera in the . . . British Museum. Taylor and Francis,

London, xlii+567 pp.
Dunn, L. 1933. Observations on the carnivorous habits of the spear-nosed bat, Phyllostomns

discolor panamensis Allen, in Panama. J. Mammal. 14:188-199.
Eisi ntraut, M. 1950. Die Ernahrung der Fledermausen (Microchiroptera). Zool. Jahrb. Jena.

79:114-177.
Fischer, H. 1909. liber funktionelle Anspassungen am Fledermausmagen. Arch. ges. Physiol.

129:113-137.
Forman, G. L., R. J. Baker and J. D. Gerber. 1968. Comments on the systematic status of

vampire bats (family Desmodontidae). Syst. Zool. 17:417-425.
Glass, G. B. J. 1968. Introduction to Gastrointestinal Physiology. Prentice-Hall, Englewood

Cliffs, N.J. 207 pp.
Gomori, G. 1952. Histochemical studies of the gastric mucosa. J. Nat. Cancer Inst. 13:

1039-40.
Goodwin, G. G., and A. M. Greenhall. 1961. A review of the bats of Trinidad and Tobago.

Bull. Amer. Mus. Nat. Hist. 112:189-301.
Grasse, P. P. 1955. Traite de Zoologie. Mammiferes. Masson et Cie, Paris, 17:1758-1804.
Grossman, M. I. 1958. The glands of Brunner. Physiol. Rev. 38:675-690.
Hale, C. W. 1946. Histochemical demonstration of acid polysaccharides in animal tissues.

Nature 157:802.
Hall, E. R., and K. R. Kelson. 1959. The Mammals of North America. Ronald Press, New

York. l:xxx+l-546+79.
Hooper, E. T., and J. Brown. 1968. Foraging and breeding in two sympatric species of neo-
tropical bats, genus Noctilio. J. Mammal. 49:310-312.
Huxley, T. H. 1865. On the structure of the stomach in Desmodus rufus. Proc. Zool. Soc.

London, pp. 386-390.
Jarvi, O., and P. Lauren. 1951. On the role of heterotopias of the intestinal epithelium in

the pathogenesis of gastric cancer. Acta Path. Microbiol. Scand. 29:26-44.
Jennings, M. A., and H. W. Florey. 1956. Autoradiographic observations on the mucous

cells of the stomach and intestine. Quart. J. Exp. Physiol. 41:131-152.
Kolb, A. 1954. Biologishe Beobachtungen an Fledermausen. Saugetierkundl. Mitt. 2:15-26.
Koopman, K. F. 1967. Bats. In Recent Mammals of the World . . . (S. Anderson and J. K.

Jones, Jr., eds.), pp. 109-150. Ronald Press, New York.
Langley, J. N. 1880-82. On the histology of the mammalian gastric glands, and the relation

of pepsin to the granules of the chief cells. J. Physiol. 3:269-291.
Leblond, C. P. 1950. Distribution of periodic acid-reactive carbohydrates in the adult rat.

Anat. Rec. 86:1-49.
Lillie, R. D. 1965. Histopathologic Technic and Practical Histochemistry. McGraw-Hill,

New York, xii+775 pp.
Mathis, J. 1928tf. Beitrag zur Kenntnis des Fledermausdarmes. Z. mik. anat. Forsch. 12:

594-647.
. 1928Z>. Uber die Brunnerschen Drusen and iiber Kornchenzellen einiger Fledermause.

Anat. Anz. 65:1-7.

Studies of Stomachs of Selected American Bats 729

Miller, G. S. 1907. The families and genera of bats. Bull. U.S. Nat. Mus. 57:1-282.
Mowry, R. W. 1963. The special value of methods that color both acidic and vicinal hydroxyl

groups in the histochemical study of mucins . . . Ann. New York Acad. Sci. 106:402-423.
Myrcha, A. 1967. Comparative studies on the morphology of the stomach in the Insectivora.

Acta Theriol. 12:223-244.
Owen, R. 1868. On the Anatomy of Vertebrates. Mammals. Longman, Green and Co.,

London, vol. 3:915 pp.
Pantin, C. F. A. 1948. Notes on Microscopical Technique for Zoologists. Cambridge Press,

New York. 79 pp.
Park, H., and E. R. Hall. 1951. The gross anatomy of the tongues and stomachs of eight

new world bats. Trans. Kansas Acad. Sci. 54:64-72.
Pine, R. H. 1969. Stomach contents of a free-tailed bat, Molossus ater. J. Mammal. 50:162.
Reeder, W. G., and K. S. Norris. 1954. Distribution, type locality, and habits of the fish-
eating bat, Pizonyx vivesi. J. Mammal. 35:81-87.
Robin, H. A. 1881. Recherches anatomique sur les Mammiferes de l'ordre des Chiropteres.

Ann. Sci. Nat. Zool. 12:1-180.
Rollett, A. 1871. Bemerkungen zur Kenntiss der Labdrusen und der Magenschlcimhaut.

Unters Inst. Physiol. Hist., Graz.
Ross, A. 1967. Ecological aspects of the food habits of insectivorous bats. Proc. West. Founda-
tion Vert. Zool. 1:205-264.
Rouk, C. S. 1968. Comparative gastric histology of selected American bats. Master's thesis,

Oklahoma State Univ., Stillwater. 51 pp.
Rouk, C. S., and B. P. Glass. 1970. Comparative gastric histology of American bats. J.

Mammal. 51:455-472.
Schofield, G. 1953. The argentaffin and mucous cells of the human intestine. Acta Anat.

18:256-272.
Schultz, W. 1965. Studien iiber den Magen-Darm-Kanal der Chiropteren. Ein Beitrag zum

Problem der Homoltfisierung von abschnitten des Saugetierdams. Zeit. Wissenschaft., Zool.

171:241-391.
. 1970. Einige Bemerkungen zum Bau des Verdauungstraktes und der systematischen

Stellung des Spitzzahnflughundes, Hurpionycteris ivlnteheadi Thomas, 1896 (Megachirop-

tera). Sonderdruck auf Z. f. Siiugetierkunde. 35:81-89.
Selkurt, E. E. (ed.) 1966. Physiology. Little, Brown and Co., Boston. 2nd ed., xii-f-796 pp.
Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bull.

Amer. Mus. Nat. Hist. 85 :xvl+ 1-350.
Spicer, S. S. 1960. A correlative study of the histochemical properties of rodent acid muco-
polysaccharides. J. Histochem. Cytochem. 18:18-33.
. 1963. Histochemical differentiation of mammalian mucopolysaccharides. Ann. New

York Acad. Sci. 106:379-388.
Vorontzov, N. N. 1960. The ways of food specialization and evolution of the alimentary

system in Muroidea. Symp. Theriol. Brno. pp. 360-377.
Walker, E. P. 1968. Mammals of the World. Johns Hopkins Press, Baltimore. 2nd ed.,

vol. 1 :xlviii+644 pp.
Wimsatt, W. A. 1969. Transient behavior, nocturnal activity patterns, and feeding efficiency

of vampire bats (Desmodtts rotundas) under natural conditions. J. Mammal. 50:233-244.
Wimsatt, W. A., and A. Guerriere. 1962. Observations on the feeding capacities and ex-
cretory functions of captive vampire bats. J. Mammal. 43:17-27.
Winge, H. 1941. The interrelationships of the mammalian genera. C. A. Reitzels Forlag,

Copenhagen, 1:418 pp. (translated from Danish by E. Deichmann and G. M. Allen).
Wislocki, G. B., H. Bunting and E. W. Dempsey. 1947. Metachromasia in mammalian

tissues and its relationship to mucopolysaccharide. Amer. J. Anat. 81:1-37.

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLIX Pages 731-874 April 7, 1972 No. 11

TABLE OF CONTENTS

ABSTRACT 733

I INTRODUCTION 734

MATERIALS AND METHODS 734

ACKNOWLEDGMENTS 737

II SKELETAL MORPHOLOGY OF PHANAEUS VINDEX

MACLEAY 737

The Head

Head Capsule 738

Appendages of the Head 746

The Cervix 758

The Thorax

The Prothorax 758

The Pterothorax '. 762

The Pterothoracic Endoskeleton 769

The Legs 770

The Wings and their Articulations 772

The Abdomen

The Abdominal Terga 774

The Abdominal Sterna 776

External Genitalia 776

III COMPARATIVE SKELETAL MORPHOLOGY

OF PHANAEINES 776

The Head Capsule 779

Appendages of the Head 787

Thoracic Segments 793

Legs 806

Wings - 811

Abdomen 812

IV ADAPTIVE SIGNIFICANCE OF ASPECTS OF

PHANAEINE SKELETAL MORPHOLOGY 813

V CLASSIFICATION OF THE PHANAEINES 815

Artificial Key to the Genera of Phanaeine Scarabaeinae 819

Diabroctis Gistel 820

Sulcophanaeus Olsoufief? 821

Phanaeus MacLeay 826

Oxystemon Castlenau 835

Coprophanaeus Olsoufieff 839

Dendropuemon Perty - 843

Tetramereia Klages 851

Homalotarsus Janssens 852

Megatharsis Waterhouse 854

VI EVOLUTION AND PHYLOGENY OF PHANAEINES 855

Some Preliminary Considerations 855

Phylogenetic Speculation 857

Generic Affinities 861

Parallelisms 864

Evolutionary Implications of Phanaeine Ecology and Distribution .. 867

LITERATURE CITED 871

APPENDIX: Material Studied 874

FRONTISPIECE

Phanaetis iindex MacLeay. Upper left, dorsal view of female; upper right, dorsal view of male.
Lower left, lateral view of female; lower right, lateral view of male. (About twice natural size.)

Comparative Skeletal Morphology, Systematics and Evolution
of the Phanaeine Dung Beetles (Coleoptera: Scarabaeidae)1

w. d. Edmonds-

abstract

The principal purpose of this paper is a re-examination of the supraspecific
classification of phanaeine Scarabaeinae based largely upon a comparative study
of the skeletal morphology of adults. Part II is a detailed morphological study of
Phcinaeus vindex MacLeay, which was chosen as a "model" species. In this part,
which is essentially an independent study of scarabaeine morphology, a termi-
nology is devised which is used in subsequent parts. Part III is an account of the
observed variation in skeletal morphology among phanaeines. It is largely upon
this variation that the revised classification is based. Part IV is a short resume
of the presumed adaptive significance of skeletal features of phanaeine morphol-
ogy concerned with fossorial habits. The revised supraspecific classification of
phanaeines presented in part V incorporates several significant changes from
older systems: 1) the phanaeine genera (not collectively referred to as the sub-
tribe Phanaeina) include Diabroctis (-Taurocopris), Sulcophanaeus, Phanaeus,
Oxysternon, Coprophanaeus, Dendropaemon, Tetramereia, Homalotarsus and
Megatharsis; Oruscatus, Gromphas and Rolbites are excluded from "Phanaeina";

2) two names, Sulcophanaeus and Coprophanaeus, proposed by G. d'Olsoufieff
for subgenera of Phanaeus are now used as names of genera quite different in
composition from the subgeneric taxa to which the names originally referred;

3) three new subgenera are proposed, Mioxysternon for some species of Oxy-
sternon, and Paradendropaemon and Coprophanaeoides for species of Dendro-
paemon; 4) extensive use is made of the informal categories of species-group and
species-complex where appropriate in all larger genera. Part VI is a discussion
of phanaeine evolution and phylogeny, including generic affinities arrived at by
analysis of the taxonomic distribution of derived character states as well as dis-
cussions of parallelisms and the evolutionary implications of phanaeine ecology
and distribution. Phyletically, the phanaeine Scarabaeinae, which are presumed
to comprise a monophyletic group, appear to consist of two distinct phyletic
lineages (the Oxysternon-Phanaeus and Coprophanaeus-Dendropaemon genus-
groups), together comprising the bulk of the known species, in addition to a
number of small phyletically isolated groups representing relicts of several older
phyletic elements.

1 Contribution No. 1478 from the Department of Entomology, The University of Kansas,
Lawrence, Kansas.

2 Present address: Department (if Biological Sciences, California State Polytechnic College,

Pomona, California 91768.

734 The University Science Bulletin

PART I: INTRODUCTION

Phanaeines are perhaps one of the best generally known groups of
Scarabaeidae. Their striking metallic colors and the often bizarre appearance
of "horned" males have made them showy collectors' items for some 200 years.
So well noted were they by early collectors that over 80% of all known species
(about 90) had been described by 1875. The only serious attempt to date
(and a generally very good one) to examine the group in toto was made by
Gregor d'Olsoufieff (1924), who, in an extensive monograph, provided the
first comprehensive keys to species and proposed many new supraspecific
groupings.

The primary intent of this paper is a re-examination of Olsoufieff's
supraspecific classification of phanaeines based primarily upon the compara-
tive skeletal morphology of adults. Before this could be done critically,
however, it was necessary to examine in detail the skeletal morphology of a
"moder species to provide a base for the comparative study. Phanaeus vtndex
was arbitrarily selected as this model. Part II of this paper is a treatment of
the skeletal morphology of P. vindex as well as those aspects of musculature
necessary for interpretation of skeletal features. Here a terminology is de-
vised which is used in other sections. Since an effort was made to compare
P. vindex with other, non-phanaeine dung beetles, this is in effect an inde-
pendent study of scarabaeine morphology; as such, it is hoped that it will be
useful to other students of Scarabaeidae.

Part III is an account of some of the observed variation of skeletal mor-
phology; the conclusions in subsequent parts are based largely upon the data
assembled here. Part IV is an examination of the adaptive significance of
certain morphological features; although this deals specifically with pha-
naeines, many of the statements there are applicable to other dung beetles as
well. Part V is the systematic section. The concluding section (VI) is a
discussion of the evolutionary aspects of phanaeine skeletal morphology,
biology and distribution and the bearing of these aspects on their presumed
phylogeny.

MATERIALS AND METHODS

The morphological study of Phanaeus vindex is based upon examination of
both dried museum specimens and specimens preserved in liquid. Like material
of Coprophanaeus lancifer, a much larger species, was studied simultaneously as
a check on several interpretations of the morphology of P. vindex, particularly
those dealing with muscles. All other material used in this study was dried. For
examination of skeletal features, all dried material was first soaked overnight in
cold 10° Cl KOH solution. In addition, it was found that the normally much
folded hind wings could be extended for easy examination by detaching and soak-
ing them briefly in the same solution.

The majority of the illustrations are line drawings made mostly with the aid

Evolution of the Phanaeine Dung Beetles

735

of an ocular grid. Because of the great variation in size of structures, they are not
all drawn to the same scale. Photographs are used to illustrate those structures
not easily depicted in ink drawings. Efficient illustration of many features requires
several views; for orientation in these instances a large, solid dot connects guide-
lines indicating the same point in drawings of the two or more aspects.

Following is a list of abbreviations used in the figures and not defined in
captions:

dorsal portion of eye
dorsal process of tentorium
dorsal surface of labrum
elytron
. mesepimeron (anterior, lateral
and posterior surfaces)
metepimeron

posterior margin of mesepimeron
elytral epipleuron
epipleural shelf
mesepisternum
metepisternum (anterior and
lateral surfaces)
fronto-clypeal ridge
fronto-clypeal sulcus
femur

profemoral tuft
labio-gular fimbria
flagellum of antenna
flexor sclerite of elytron (and
hind wing)
metafurcasternum
galea
glossa
glossal bar
glossal flap
gula

gular ridge
gular suture

horizontal lamella of mesonotum
head wall
hypostoma

inflexed margin of basisternum
(of presternum)
invaginated anterior margin of
mesepisternum
inflexed margin of gula
inflexed margin of sternellum
(of presternum)
incisor lobe of mandible
interspiracular pad
lateral arm of hypopharyngeal
suspensorium
lacinia

labral muscles
labial palp
lateral clypeal carina

abas2

mesobasalar apophysis

deye

acx

articulation of coxa

dprt

afr

apical fringe of labrum

dsula

afur2, 3

meso-(meta-)furcal arm

ely

afur2P

lateral process of mesofurcal arm

epim2a,

amcly

anterior margin of clypeus

ampr

antcro-median process of labral

epim3

suspensorium

epimpm

amt

anterior margin of tentorium

epip

anf

antennifer

epips

ans

antennal socket

epis2

ap

aedeagal apodeme

epis3a,i

apely

articular process of elytron

apglb

apodeme of glossal bar

fcr

apmcly

apical margin of clypeus

fes

apmmi.2

apodemes of mandibular muscles

fe

apst

prosternal apophysis

fet

arl

articular line (between clypeus

fim

and labrum)

fla

armx

articulation of maxilla

fs

at

acrotergite of abdominal tergum

ata

anterior tentorial arm

fst

atibc

anterior protibial carina

gal

awp

anterior metanotal wing process

gl

ax

axillary sclerite

gib

ba

basalis of mandible

glf

bas2, 3

meso(meta-)basalare

gu

bfur2, 3

base of meso(meta-)fiirca

gur

br

median brace of mesofurca

gus

bst

basisternum (of presternum)

hi

bstc

basisternal crest

hw

ca

canthal area

hyp

car

cardo

imbst

cbr

cephalic brachium

cly

clypeus

imepis

clypr

clypeal process

cmb

comb of incisor lobe of mandible

imgu

conj

cnniunctivus of mandible

imstn

cr

transverse tibial carina

cpr

cephalic process

inlb

ex

coxa

ip

cxd

coxal depression of sternum

la

da

dorsal arm of hypopharyngeal

suspensorium

lac

dacx.3

dorsal articulation of metacoxa

Iami.2

darm

dorsal articulation of mandible

lbpl

darsgal

dorsal articular sclerite of galea

lcrcly

736

The University Science Bulletin

If
lfos

libs
lobe

Ipc

Ishpx

mac

mbfcs

mbr

mcon

md

mem

mgpmt

mlhpx

molb

morg

mrhpx

mshpx

mt

mth

mxpl

nt,,i

oa

ocar

ocfor

oclb

ocr

oes

oford

oforv

ogs

paar

par

pbs

ped

pfos

phb

plrg2, 3

pls2, 3

pm

pmpr

pmts

pntm

pocr

pocs

pol

polb

lateral file of labrum

lateral pronotal fossa

lateral labial sclerite

postero-ventral lobe of

metapleuron

lateral pronotal carina

lateral hypopharyngeal sclerite

mandibular acetabulum

mesal branch of fronto-clypeal

sulcus

median brush of labrum

mandibular condyle

muscle disc of metepisternum

membrane

median groove of prementum

median lobe of hypopharynx

molar lobe of mandible

molar ridges of mandible

median ridge of hypopharynx

median hypopharyngeal sclerite

mentum

mouth

maxillary palpus

posterior inflection of mesonotum

ora

arm

occipital area of parietals

occipital foramen

occipital lobe

occipital ridge

oesophagus

optic foramen of dorsal portion

of eye

optic foramen of ventral portion

of eye

oculo-gular space

paraocular area of parietals

paraocular ridge

pleurobasisternal suture

pedicel

posterior pronotal fossa

phallobase

meso(meta-)pleural ridge

meso(meta-) pleural suture

paramere

postero-median process of labral

suspensorium

premental sclerite

circumnotal ridge

postoccipital ridge

postoccipital suture

postocular line

postocular lobe

pors paraoral sclerite

ppha paraphragmal area of prescutum

(of metanotum)

ppl propleuron

pr process

prcar articular process of cardo

prmx articular process of maxilla

proc preocular carina

pscu prescutum (of metanotum)

pstn metapostnotum

ptbr posterior tentorial bridge

ptd posterior tentorial depression

ptibc posterior carina of proti'oia

pwp posterior metanotal wing process

phi. 2. j phragma (first, etc.)

rg ridge

rmgl receptacle of mandibular glands

scu2,3 meso(meta-)scutum

sct2,3 meso(meta-)scutellum

sen scnsilla

smt submentum

sp scape

spfr scapal fringe
ssi, ii, in, iv stipital sclerite I, etc.

st stria

stn sternellum (of prosternum)

sts mesometasternal suture

sub3m<i metasubalar muscle disc

su hpx hypopharyngeal suspensorium

Si-8 abdominal sternum 1, etc.

T].-s abdominal tergum 1, etc.

tab metepisternal tab

tar tarsus

tcrcly transverse clypeal carina

tib tibia

tibs tibial spur

tibt tibial teeth

topr tormal process of labral

suspensorium

tr trochanter

trn trochantin

vacx3 ventral (sternal; articulation of

metacoxa

varm ventral articulation of mandible

varsgal ventral articular sclerite of galea

veye ventral portion of eye

vl vertical lamella of mesonotum

w w-suture of metanotum

wp2, 3 meso(meta-)pleural wing

process

y articular surface of mesepimeron

z articular surface of mesepimeron

Evolution of the Phanaeine Dung Beetles 737

ACKNOWLEDGMENTS

I would like to express my thanks to Dr. C. D. Michener for the many
manifestations of his support, encouragement and guidance throughout the course
of this study. I am also especially indebted to Dr. G. W. Byers for his thoughtful
consideration of the many problems and questions brought to him.

I also offer very special thanks to Padre F. S. Pereira, CMF, who, besides
suggesting this study in the first place, was my gracious host during a year's stay
at the Departamento de Zoologia ("Museu Paulista") of the Secretaria da Agri-
cultura in Sao Paulo, Brazil, and to Dr. Gonzalo Halffter, who on several occa-
sions has in many ways facilitated my study of Mexican phanaeines. It is largely
because of the generous offer of time, material and encouragement by these dis-
tinguished students of Scarabaeinae that this study was possible. My sincerest
thanks also to Dr. Paulo Emilio Vanzolini, Director of the Departamento de
Zoologia, who in many ways facilitated my research activities in Brazil.

My thanks also go to the following individuals who generously loaned material
for this study from the collections in their care: O. L. Cartwright (United States
National Museum), C. L. Hogue (Los Angeles County Museum), H. B. Leech
(California Academy of Sciences), J. F. Lawrence (Museum of Comparative
Zoology, Harvard University) and Roger Damoiseau (Institut Royal des Sciences
Naturelles de Belgique). Dr. Damoiseau provided one of the two known speci-
mens of Homalotarsus impressus for use in this study.

Support of research activities in Brazil was provided by a Fulbright Scholar-
ship (1965-66) awarded by the United States Department of State. Support of the
final year of this study at the University of Kansas was provided by a Systematics
and Evolutionary Biology Traineeship funded by the National Science Foundation
and administered by the University of Kansas (Grant GB-4446X, Dr. William A.
Clemens, principal investigator). The support of these agencies is most gratefully
acknowledged.

Finally, a special vote of appreciation to Mrs. Judy Keller, who typed the final
version of the manuscript.

PART II: SKELETAL MORPHOLOGY OF PHANAEUS V INDEX

MACLEAY

This a robust, compact dung beetle common in the eastern United States.
The compactness of the body is accentuated by the strong sclerotization of
nearly all external surfaces. Compared with related species, P. vindex is
moderate in size, varying from 14 to 22 mm in length and 8 to 14 mm in
width (at the bases of the elytra). Unlike most other scarabaeines, phanaeines
are commonly brilliantly metallic in color. In this species most of the upper
surface of the head and sides of the pronotum are yellow-green, the remainder
of the pronotum is coppery-red, and the elytra, pygidium, exposed pteropleura
and greater part of the lower body surface are green to yellow-green.

The sculpturing of the integument varies in nature and degree over the

738 The University Science Bulletin

body surface. In general, the upper surface is strongly rugose; the lower
surface is smooth to punctate but never rugose. Vestiture is restricted to
pleural and sternal surfaces and to the appendages. Its variation is described
below when appropriate.

The Head
head capsule

As in other Scarabaeidae, the head is prognathous; the ventral and lateral
walls are expanded, bringing the mouthparts into a horizontal plane with
their apices directed forward. The capsule is very heavily sclerotized; as is
common in beetles, the thickness of the integument tends to obscure many
of the surface features (sutures, etc.) observed in other insects.

The most conspicuous feature of the upper surface of the head of large
(so called "major") males is a long, massive tapering projection, the cephalic
process (Fig. 1, cpr), which arises between the eyes and curves posteriorly
over the pronotum. The development of the cephalic process (as well as
pronotal processes) is allometric and varies considerably within local popu-
lations. In smaller ("minor") males, the cephalic process is considerably
reduced in length (Fig. 2) or may be nearly absent. The cephalic process of
the female, on the other hand, is less variable in development, being a short,
erect, apically emarginate projection (Fig. 3). Large, often bizarre corniform
outgrowths of the head wall (and pronotum) are very common in the
Scarabaeidae, particularly in the subfamilies Scarabaeinae and Dynastinae
(see Arrow, 1951).

The flattened, anterior portion of the head dorsum is divided in front of
each eye by an oblique lateral clypeal carina (Figs. 1, 4, lcrcly) extending
from the head margin toward the base of the cephalic process. These carinae
are almost always present in phanaeines but seldom in other dung beetles.

Directly posterior and more or less parallel to each lateral clypeal carina
is a faint line interpreted here as the fronto-clypeal sulcus (- f ronto-genal
sulcus of authors) (Figs. 1, 4, fcs). Mesally this sulcus curves toward the eye
and divides, one short, simple branch continuing toward but not reaching
the eye, and the other, Y-shaped mesal branch of the fronto-clypeal sulcus
(mbfcs) extending toward the base of the cephalic process. Medially the
fronto-clypeal sulcus is incomplete. Ventral segments of the fronto-clypeal
sulcus (Fig. 9) follow the anterior margins of the antennal sockets and
terminate at the dorsal articulations of the mandibles (darm). A medially
incomplete fronto-clypeal sulcus with a mesal branch is found in all
Scarabaeinae.

The fronto-clypeal sulcus marks an internal fronto-clypeal ridge (Figs. 6-7,
fcr) joining the head wall above with that inflexed beneath it. Except via a
small sinus, the transverse septum thus formed on each side of the head

Evolution of the Phanaeine Dung Beetles

739

amcly
clypr

deye

ocar

Fics. 1-4. Phanaetis vindex. 1, head, lateral view, well developed, "major" male; 2,
cephalic process of "minor" male, A — lateral view, B — frontal view; 3, cephalic process of
female, A and B same as 2; 4, head of male, dorsal view (cephalic process removed).

740 The University Science Bulletin

prevents direct flow of hemolymph between the clypeus and paraocular areas
of the parietals. The mesal end of each fronto-clypeal ridge forms a dorso-
ventral brace between the upper head wall and the dorsal articulation of the
mandible. Even though it is not an independent structure, for descriptive
purposes this mesal bracing portion of each ridge will be called a cephalic
brachium (head brace). Figure 5 shows the cephalic brachium (cbr). Be-
cause its thickened vertical margins resemble two arms extending dorsally
from the dorsal articulation of the mandible, the brachium resembles an in-
verted bipod. Between its "arms" extends a vertical, triangular plate bent
along its midline (height) such that a V-shaped line is inscribed where it is
fused with the dorsal head wall. This line of fusion follows the arms of the
Y-shaped mesal branch of the fronto-clypeal sulcus. The "bipodal" cephalic
brachium and almost always coincident Y- or T-shaped mesal branch of the
fronto-clypeal sulcus are characteristic of Phanaeus and Oxysternon. Other
phanaeines and non-phanaeines examined have a "monopodal" brachium
(see part III, character 7).

Arising directly behind each dorsal articulation of the mandible is a
slender, spatulate anterior tentorial arm (Fig. 5, ata). Originating on the
expanded distal end of this process is a muscle inserting on the base of the
scape. No anterior tentorial pits are visible. The anterior tentorial arms are
very delicate and highly susceptible to loss if dissection is not careful. Very
similar anterior tentorial arms were observed in all beetles examined.

The clypeus is greatly expanded and, because of the ventral inflection of
its margin, consists of two distinct areas: the flattened dorsal surface of the
clypeus (Fig. 4, cly), anterior to the fronto-clypeal sulcus above, and the
ventral surface of the clypeus (Fig. 9), anterior to the ventral portions of
the same sulcus. Literature reference to the "clypeus" of dung beetles is
usually to the dorsal surface. The dorsal surface of the clypeus of phanaeines
is transversely ridged except for a smooth, narrow anterior border (Fig. 79,
cly) ; the ventral surface bears setigerous punctures.

The fold of the clypeus, separating dorsal and ventral surfaces and de-
scribing the generally semicircular leading margin of the head is here referred
to as the anterior margin of the clypeus (Figs. 4, 9, amcly). This margin is
slightly upturned and, in vindex, slightly truncate or rounded medially. The
morphological anterior limit of the clypeus, or apical margin of the clypeus
(Fig. 9, apmcly), is the U-shaped margin of the ventral surface of the clypeus
exposed when the labrum is removed. The dorsal articulations of the man-
dibles (Figs. 5, 9, darm) are spherical condyles at the postero-lateral ex-
tremities of the apical margin.

The ventral surface of the clypeus is slightly depressed medially as a dis-
tinctly crescent-shaped area which partially receives the labrum. Between
this depression and the anterior margin of the clypeus is the clypeal process

Evolution of the Phanaeine Dung Beetles

741

ata

cbr

apmcly

arm

deye

cbr

darm

deye

ocar

ocfor

Figs. 5-8. Phanaeas rindex, male. 5, eye region of head capsule, internal view; 6, trans-
verse section through right side of head capsule anterior to fronto-clypeal sulcus (fs=plane of
frontal section in Fig. 7); 7, oblique frontal section of fronto-clypeal ridge; 8, head capsule,
caudal view (cephalic process removed).

(= "estructura ventral del clipeo" of HalfTter, 1958 and elsewhere) (Figs. 9-
10, clypr). The clypeal process of most dung beetles examined is, as in
vindex, a rather strongly produced transverse carina.

Interpretation of the surface areas of the head immediately posterior to the
fronto-clypeal sulcus is difficult. The spatial relationship between the anten-
nal sockets and this sulcus, however, suggests that a distinct frontal region
(frons) has been much reduced or eliminated by anterior and/or posterior
expansion of other cephalic areas (e.g., the clypeus). Consequently, a vertex

742 The University Science Bulletin

and genae are not distinguishable. These ideas are those of DuPorte (1960);
it seems logical to adopt his designation of the dorsal and lateral areas of the
head between the fronto-clypeal sulcus and the occipital area as the parietal
area, or parietals.

The expanded areas of the parietals posterior to the fronto-clypeal sulcus
and lateral to the eyes are the paraocular areas of the parietals (Fig. 1, 4, 9,
paar). Like the clypeus, they are inflected ventrally. Apparently because of
their position, these areas are commonly referred to as "genae," a name with
doubtful morphological connotations as applied to dung beetles. The para-
ocular areas of most dung beetles are not "areas," but, rather, each is a finger-
like extension of the head wall onto the eye surface. Such a prolongation is
referred to as a canthus.

In the phanaeines and a few other Scarabaeinae (e.g., Onitis and
Scarabaeus), the paraocular areas of the parietals are expanded through the
eyes, dividing them completely. Posteriorly each area firmly contacts, but is
not fused with, a swelling of the parietals behind the eye, the postocular lobe
of the parietals. The line of contact between these two expansions is a vertical
"suture" connecting the posterior angles of both portions of the eye; this
suture is the postocular line (Figs. 1, 5, pol). Figure 5 shows the internal
structure of the head capsule of vindex, in this region. Here it can be seen
that the internal margin of the paraocular area completely divides the optic
foramen into upper and lower parts (oford, oforv). This division in vindex,
and presumably in other dung beetles with similar structure, is correlated
with a division of each optic lobe of the brain into two branches, each passing
through a separate foramen.

The shape of the posterior margin of the paraocular areas varies among
phanaeines. As seen from above in vindex (Fig. 4), this margin curves to the
posterior angle of the eye. In other phanaeines (e.g., Diabroctis mimas, Fig.
115) it extends to the middle of the lateral margin of the eye, leaving, as seen
from the side, a vertical flattened area (Fig. 115A, ca) completing the division
of the eye. This flattened area will be referred to later as the canthal area. It
suffices to note here that vindex and similar species do not possess a distinct
canthal area.

As seen from above (Fig. 4), strong depressions in the dorsal head wall
behind the eyes set off the postocular lobes of the parietals (polb). These
prominent lobes fit into notches in the anterior pronotal margin. Except for
the dorsal portions of the postocular lobes which protrude from these notches,
the dorsal surface of the head between the postocular lobes and posterior to
the cephalic process is concealed by the anterior pronotal margin upon re-
traction of the head. The cephalic-prothoracic interlocking mechanism
achieved by accentuation of the postocular lobes of the parietals and notching
of the anterior pronotal margin is characteristic of Phanaeus. Although it is

Evolution of the Phanaeine Dung Beetles 743

approached in Oxystemon, this mechanism is apparently not found in other
phanaeines or other scarabaeines.

The cephalic process of both sexes is a hollow outgrowth of the head wall.
Because the fronto-clypeal sulcus is incomplete medially, it is impossible to
know the exact morphological identity of the area of the head bearing the
cephalic process. That is, it cannot be known with reasonable certainty
where the clypeus ends and the parietals begin in this area.

In some phanaeines, including vindex, there appears dorsally and laterally
on the caudal portion of the head capsule a strong, bowed setose ridge (Figs.
4, 8, ocr). It marks no corresponding internal ridge and probably represents
a secondary thickening rather than a true occipital sulcus. Primarily because
of its position, I call it the occipital ridge and the dorsal and lateral areas of
the head wall between it and the postoccipital suture the occipital area
(ocar). Similar occipital ridges occur in some other dung beetles; Matthews
(1961) first referred to them as transverse occipital carinae in Copris. In
American Canthonina the occipital ridge provides important characters at
the generic level (HalfTter, in litt.). The morphological implications of "occip-
ital" designations are justifiable since the occipital area is presumably, at
least in part, composed of true occiput. Dorsally the occipital area is nearly
suppressed because of the close approximation of the occipital ridge and post-
occipital suture; laterally it is expanded, flattened and slightly depressed.

The postoccipital suture (Fig. 8, pocs) is visible only dorsally as a fine
line closely following the posterior margin of the head. Laterally it is ob-
literated where it apparently coincides with the lateral margins of the occipital
foramen. Thus, a very narrow postocciput appears only dorsally. Dorsally
the postoccipital suture marks a deep, plate-like invagination continuous
laterally with much narrower, rim-like inflections of the posterior margin of
the head wall. This continuous inflection is the postoccipital ridge (Fig. 8,
pocr). The ventral extremities of the postoccipital ridge are continuous with
the tentorial structure described below.

The occipital foramen is a dorso-ventrally elongate opening in the pos-
terior head wall (Fig. 8, ocfor). Its ventro-lateral angles are rounded, and in
them articulate the apices of the cervical sclerites.

The ventral surface of the anterior portion of the head wall formed by
the clypeus and paraocular areas of the parietals has been mentioned above.
The antennal sockets are located in depressions of the paraocular areas
marked posteriorly by the preocular carinae (Fig. 9, proc). A laterally-
directed, conical antennifer (anf) projects from the mesal margin of each
antennal socket. Antero-mesally the ventral portion of each eye closely ap-
proaches the base of the mandible, from which it is separated by a narrow,
vertical portion of the head wall, the oculo-malar space (= pleurostoma?).
In some dung beetles (e.g., Scarabaeits) this area is separated by a weak

744

The University Science Bulletin

armx

ocar

Figs. 9-15. Phanaetis vindex, male. 9, head, ventral view; 10, median portion of anterior
margin of clypeus, frontal view; 11, ventral head wall, internal view; 12, dorsal process of
tentorium, lateral view; 13-15, maxillary muscles originating on tentorium, diagrammatic:
13, longitudinal (parasagittal) section (xs— plane of cross section in Fig. 14); 14, cross section;
15, dorsal view (cf.. Fig. 11; ls=plane of longitudinal section in Fig. 13).

suture extending from near the dorsal articulation of the mandible (i.e., the
supposed position of the anterior tentorial pit) to the eye. This suture, which
is obscured in phanaeines, is possibly part of a subgenal (pleurostomal)
suture.

Ventrally the posterior portion of the head capsule of scarabaeids is
strongly convex. Its most conspicuous element is the gula (Fig. 9, gu), a
large, central, bulging sclerite. The gula is located between the ventral ex-
tensions of the postoccipital suture, which for convenience are referred to as

Evolution of the Phanaeine Dung Beetles 745

gular sutures (gus) (a cervical origin of the gula is assumed). The anterior
external limit of the gula is approximated by a dense, transverse row of setae
comprising the labio-gular fimbria (Fig. 9, nm). Removal of these setae from
dung beetles reveals no suture, ridge or other line marking the junction be-
tween the gula and submentum; it reveals only a line of change in sculpture
between the smooth gula and setigerously punctate submentum. In other
Scarabaeidae (e.g., Phyllophaga), however, an entirely similar, presumably
homologous row of setae occurs along a definite labio-gular suture. It appears
reasonable, then, to consider the labio-gular fimbria of Scarabaeinae the
definitive line of fusion of the gula and submentum. The posterior margin
of the gula (Fig. 8, imgu) is strongly inflexed and consolidated with the
ventral extremities of the postoccipital ridge. The submentum is fused later-
ally with the oculo-gular spaces and forms an integral part of the ventral
head wall.

The remaining portion of the head capsule mesal and posterior to the
ventral portions of the eyes can best be considered as three indistinct paired
areas (Fig. 9). The first two are the ventral extensions of the occipital area
(ocar) and postocular lobes (polb). The third is the oculo-gular space (ogs),
that area between the margin of the eye and the lateral margins of the gula
and submentum. Because of the lack of ventral extensions of the occipital
sulcus, the morphological composition of the oculo-gular space is indetermin-
able; this space could be fused parts of the subgena or facial area (sensu
DuPorte, 1960). If DuPorte is correct, the extent of the areas occupied by the
ventral portions of the eyes of such scarabaeids as Phyllophaga indicates that
the oculo-gular spaces are formed mostly by an invasion of the parietals onto
the ventral surface of the head at the expense of the postgenae, which are
greatly or completely replaced. This suggestion is further supported by the
absence of an occipital ridge in many phanaeines and other dung beetles so
that occipital and postgenal areas appear so consolidated with the parietals as
to lose their identity.

The antero-mesal margin of each oculo-gular space bears a solidly fused,
dorsally flexed sclerite adjacent to the anterolateral angle of the submentum;
this raised area is a hypostoma (Fig. 9, hyp). Dorsolaterally each hypostoma
bears a small acetabulum, the ventral articulation of the 77iandible (Fig. 11,
varm), which receives the mandibular condyle.

The posterior tentorial pits (Fig. 9, ptd) are represented by wide depres-
sions in the ventral wall of the head at the sides of the gula. Internal exami-
nation of the gula, however, reveals a well developed tentorium supported
laterally by invaginations along the entire lengths of the gular sutures rather
than along the restricted areas indicated by the depressions. As DuPorte
(1960) concludes for Phyllophaga, the gular sutures of dung beetles are
probably best interpreted morphologically as elongations of the posterior

746 The University Science Bulletin

tentorial pits, and the gular ridges as expanded bases of the posterior tentorial
arms. For descriptive clarity, however, use of the terms gular ridge and gular
suture will be continued.

The tentorium of vindex conforms to the very uniform structure found
in the Scarabaeinae studied. The invaginations of the gular sutures, the gular
ridges (Figs. 11, 14, gur) fuse medially to form a thinly sclerotized roof over
the gula.'^ A strong antero-dorsal flexion of the posterior portion of this roof
has produced what resembles a dorsally projecting platform supported
medially by a vertical, longitudinal lamella. This platform is the dorsal
process of the tentorium (Fig. 12, dprt). Posteriorly it extends as a very
delicate, E-shaped posterior tentorial bridge (ptbr). Extremely delicate fila-
ments connect the antero-lateral angles of the dorsal process with the anterior
tentorial arms; in Figures 5 and 11 only incomplete portions of these fila-
ments are shown. Originating on the dorsal process of the tentorium of
vindex, and presumably other dung beetles, are two paired muscles. One
(Figs. 13-15, c) inserts on the tendon of a much larger muscle (d) origi-
nating posteriorly on the gula and inserting on the ventral posterior angle of
the stipes; the other (b) divides apically into two sub-bundles, both of which
insert on the cardo mesal to the maxillary articulation (Fig. 23). A third pair
of maxillary muscles (a) originates posteriorly on the gular ridges and inserts
on the articular process of the cardo lateral to the maxillary articulation.
Muscles b, c and d are adductors of the maxillae; a is an abductor.

The anterior margin of the tentorium (Fig. 11, amt) is thickened and
resembles a bar arching between the internal antero-lateral angles of the gula.
Associated with the ends of this "bar" are two mesally-directed processes, the
maxillary articulations (armx), on which articulate the cardines.

The hypopharynx forms with the prementum a compound structure
treated below along with the labium.

APPENDAGES OF THE HEAD

As is characteristic of Scarabaeinae, the lamellate antennae are nine-seg-
mented (Fig. 16-17). The posterior margin of the first segment, or scape
(sp), bears a row of long setae, the scapal fringe (spfr), which reposes over
the ventral surface of the eye. The second antennal segment is the pedicel
(ped); the remaining seven segments comprise the Hagellum (fla).

The last three antennal segments, the lamellae, together constitute the
antennal club (Fig. 17). The form of the lamellae and antennal club is very
characteristic of phanaeines and may be described as the Phanaeus-type. The
basal lamella (seventh antennal segment) is scoop-shaped because of the
concave distal surface which receives the smaller, similarly shaped second
lamella. The terminal lamella (ninth segment) is received by the second; it

3 In Anomiopsoides, a curious South American cucraniine, the gular ridges are not fused
medially but are widely separated as in Phyllophaga.

Evolution of the Phanaeine Dung Beetles 747

is rounded and only slightly hollowed distally. Unlike the rest of the antenna,
the lamellae are soft and fleshy except for the proximal, convex surface of the
first, which is lightly sclerotized.

Although such modification of the antennal club is characteristic of
phanaeines, it is not restricted to them. An instance of convergence is the
Phanaeus-typc antennal club of the South African genus Seepages. The an-
tennal club of the male of the Bornean Onthophagus egregius Arrow is
similar to the Phanaeus-type, while that of the Palaearctic geotrupine genus
Lethrus is modified in a similar, though distinct manner.

While the following account deals specifically with P. vindex, its mouth-
parts are illustrative of peculiar, though quite uniform structures character-
istic of the whole subfamily. The mouthparts of adult Scarabaeinae (and
some Aphodiinae) are quite unlike their usually more generalized counter-
parts in other beetles. As pointed out by Halffter and Matthews (1966) and
others, they are highly modified for the manipulation of soft food, such as
dung. The earliest discussion of adult scarabaeine mouthparts (those of
Dichotomius [as Copris] carolinus) is that by Smith (1892), which deals
primarily with the mandible. An excellent early study by Hardenberg (1907)
compared the mouthparts of D. carolinus (as Pinotus) with those of other
dung beetles, including Phanaeus carnijex {-vindex, not sulcatus), and with
those of many other Scarabaeidae. Other general discussions include Mac-
Gillivray (1923; Pinotus), Mohr (1930; Canthon, Bolboceros and Aphodius),
Madle (1934; Aphodius), Halffter (1961; Canthon) and Halffter and
Matthews (1966; Phanaeus).

The mouthparts are here considered to include the labrum (= labrum-
epipharynx of authors). Such consideration is supported partly by the mor-
phological postulate that the labrum of insects is appendicular in origin but
primarily by the curious morphology of the labrum of dung beetles, which
suggests that it functions more as a mouthpart than a "preoral flap." The
labrum ("labro-epifaringe") of Canthon is discussed at length by Halffter
(1961).

Except for the suspensorium, the labrum is a dorso-ventrally flattened,
squared flap lying in a horizontal plane on the lower surface of the head.
Figure 18 shows diagrammatically its position relative to the head wall. It is
joined by a narrow membrane (arl, representing the labro-clypeal sulcus) to
the apical margin of the clypeus. The median portion of this articular line is
seen dorsally on the extracted labrum as a transverse sclerotic thickening
(Fig. 19A, arl). Anterior to this line is the dorsal surface of the labrum
(dsula),4 which in other beetles is usually visible externally.

4 Reference to position in discussion of mouthparts will be made in the definitive (prog-
nathous) sense rather than the morphological (hypognathous) one. Hence, the morphological
anterior surface of the labrum is called the dorsal surface, etc.

748

The University Science Bulletin

ampr

-mim^

ampr

topr

pmpr

Figs. 16-19. Phanaeus vindex, male. 16, left antenna, ventral view (last two segments not
visible); 17, same, side view of antennal club (7, 8 and 9 indicate segment number); 18,
labrum and portion of clypeus, longitudinal (parasagittal) section; 19A, extracted labrum,
dorsal view (muscles shown unilaterally; ls=plane of longitudinal section in Fig. 18); 19B,
same, ventral (oral) surface.

Evolution of the Phanaeine Dung Beetles 749

Posteriorly the labrum extends internally as a strongly sclerotized sup-
porting structure, the labral suspensorium (Figs. 18-19A). Posterior exten-
sions of the labral suspensorium form three processes, two lateral formal
processes of the labral suspensorium (topr) and a postero-median process
(pmpr) which from above is strongly bent to the right. Extending to the
apex of the labrum is the antero-median process (ampr, = "vastago clavi-
forme" of HalfTter, 1952), a slender projection lying between the dorsal and
ventral labral walls and adherent to the latter (Fig. 18). Apically the
antero-median process expands dorso-ventrally, joining the upper and lower
labral walls; ventrally this expansion bears the large setae of the median
brush (mbr). Inserting laterally on each tormal process is a large muscle
(lami) originating on the dorsal head wall directly behind the cephalic
process; these muscles presumably depress the apex of the labrum. The
insertion of these muscles suggests that the tormal processes represent the
tormae (sensu Snodgrass, 1935) of other insects; indeed, the labral sus-
pensorium might be the highly modified product of fusion of the tormae.
A much shorter, second pair of muscles (lam2) extends between the anterior
edge of a large round hole in the suspensorium and the membrane between
the tormal and postero-median processes; they presumably pull this mem-
brane forward. Band-like thickenings of the membrane posterior to the
tormal processes connect the labral suspensorium laterally to the bases of the
mandibles, maintaining a close appression of the labrum upon them.

The lumen between the labral suspensorium and the ventral wall of the
labrum (Fig. 18, lumen) contains an extensive network of exocrine glandular
material, the secretions of which may serve such ends as brood ball con-
struction (as an adhesive) or brood ball protection (as an antibiotic prevent-
ing bacterial and fungal growth, etc.).

The ventral surface of the labrum (Fig. 19B) is membranous and bears
several types of regularly distributed setae. The median brush (mbr) is a
longitudinal clump of thick, closely set setae inserted in the antero-median
process of the labral suspensorium. Lateral to the median brush are similar
setae scattered in a thick mat of short, fine hair. Two lateral, apically di-
vergent longitudinal rows of prostrate setae comprise the lateral files (If).
The space between the lateral files is the median area of the labrum ("zona
central" of HalfTter, 1961); those spaces outside them, the lateral areas
("zonas laterales"). The lateral areas are clothed by long, silky hairs similar
to the longer ones along the apical margin comprising the apical fringe (af r) .

The membrane extending posteriorly from the labral suspensorium (con-
tinuous with the ventral surface) covers the upper ends of the molae of the
mandibles and extends farther backward to form the dorsal wall of the
buccal cavity. The ventral surface of the labrum rests against the mandibles
with its median brush between the flattened incisor lobes of the mandibles.

750

The University Science Bulletin

molb

varsgal

miac

prmx

Figs. 20-24. Phanaens vindex, male. 20, left mandible, lateral view; 21, same, ventral
view, 22, molar lobes of mandibles, caudal view; 23, left maxilla, ventral view; 24, same,
dorsal view.

Of the mouthparts, the mandibles are perhaps the most curiously modi-
fied (Figs. 20-21). Smith (1892), followed by Hardenberg (1907) and others,
recognized several more-or-less distinct parts which he considered homologs
of maxillary sclerites. For descriptive purposes, it is convenient to recognize

Evolution of the Phanaeine Dung Beetles 751

four mandibular parts: the basalis (Smith), incisor lobe, conjunctivus
(Smith) and molar lobe.

The basalis, or basal piece, is heavily sclerotized and bears the articu-
lations and muscle insertions of the mandible (Fig. 20, ba). The mandibular
acetabulum (mac, dorsal articulation) fits over the condyle on the postero-
lateral extremity of the apical margin of the clypeus. The mandibular
condyle (Figs. 20-21, mcon; ventral articulation) fits into the acetabulum in
the dorso-lateral angle of the hypostoma. Directly lateral to the mandibular
condyle is a small swelling on which is inserted the fan-shaped apodeme
(apmni2) of the abductor muscle of the mandible; the larger apodeme of
the adductor muscle (apmmi) is inserted ventro-mesally on a swelling
adjacent to the molar lobe. Both of these muscles originate on the postero-
ventral head wall.

The receptacle of the mandibular glands (Figs. 20-21, rmgl) is thin-
walled and cup-shaped; it is attached rather loosely to the posterior surface
of the basalis. The receptacle contains a large mass of exocrine glandular
tissue similar to that in the lumen of the labrum. Inserted medially on the
receptacle is a large muscle (not shown in Figs. 20-21) with no conspicious
apodeme which originates beside the adductor and abductor muscles on the
head wall. Since its insertion is on a line between the dorsal and ventral
mandibular articulations, it presumably does not function, at least efficiently,
in gross mandibular movements. Rather, it likely moves the receptacle,
perhaps to force discharge of the products of the mandibular glands.

The molar lobes (Figs. 20-21, molb) are dorso-ventrally elongate, mesal
lobes of the basalis. Their transversely ridged mesal surfaces are asym-
metrical (Fig. 22), the left concave and the right convex. Miller (1961)
found each molar lobe to consist of a series of juxtaposed microscopic
processes ("tritors") capable of extreme trituration of food particles. Dorsally
the ridging of both molar lobes becomes very coarse and projects beyond the
posterior margin, which appears serrate (Fig. 20, morg). In situ the molar
lobes fit against the vertical posterior wall of the hypopharynx, which forms
the anterior portion of the deeply concave floor of the buccal cavity. Only
the dorsal ends of the molar lobes project above the posterior hypopharyngeal
wall. Presumably, food entering the buccal cavity is first masticated by the
coarser dorsal ridging and then more finely ground by the microscopic ridg-
ing after entering.

The basalis extends anteriorly as a dorso-ventrally flattened lobe thought
to function as a food scraper and to correspond to the incisor lobe of the
more generalized chewing mandible. Laterally this incisor lobe of the
mandible (inlb) is strongly sclerotized, but it becomes decreasingly sclero-
tized and more translucent mesally. The apex of this lobe is repeatedly
narrowly incised mesally and bears many long superficial setae (not shown

752 The University Science Bulletin

in Fig. 21). A double row of much shorter, closely-set setae comprising the
comb of the incisor lobe (cmb; see Miller, 1961) occurs along the mesal
margin of a weakly sclerotized, triangular area. This fringed area is referred
to by Smith and several later authors as the "prostheca."

The con]unctivus (Smith) is a convex, soft and flexible "sclerite" on the
ventral surface of the mandible between the comb of the incisor lobe and
molar lobe (Fig. 21, conj). It appears longitudinally ridged, but as Halffter
(1961) has pointed out, each "ridge" consists of a row of thick adherent
setae. In situ the conjunctivus fits over the thick, lateral margin of the
posterior hypopharyngeal wall lateral to the molar lobe, possibly functioning
as a seal or strainer preventing circumvention of the coarsely ridged dorsal
ends of the molar lobes by large food particles. Miller (1961) believes the
conjunctivus ("flexible area of the mandible") cushions the grinding move-
ments of the molar lobes and, possibly, permits independent molar move-
ments "while the mandibles are in the closed position."

The maxillae articulate in a plane parallel to that of the mandibles. As
seen ventrally (Fig. 23), they are well sclerotized appendages not markedly
different from a generalized type, as are the labrum and mandibles. Each
cardo (car) is rounded basally, and projecting mesally from its basal margin
is the elongate articular process of the cardo (prcar). This process bears a
forked lobe (prmx) which articulates with the maxillary articulation (Fig.
11, armx).

The stipes constitutes the bulk of the body of the maxilla. It is quadri-
partite, each division being separated from the others by membranous sulci
(i.e., the parts can be disjoined quite easily). What are here considered
three distal sclerites of the stipes have been considered two subdivisions of
the galea and a palpifer in other studies of scarabaeine morphology (e.g.,
Mohr, 1930). The primary basis for the present interpretation is the in-
trinsic musculature of the maxilla, pertinent aspects of which will be brought
out below. Although appropriate names for the parts of the stipes could be
devised using descriptive prefixes, I have used numerical designations to
avoid confusion with other maxillary terminology.

The basal sclerite of the stipes, stipital sclerite I (Fig. 23, ssi; = stipes of
authors) articulates with the cardo. On its internal surface originates a
muscle inserted on the ventral articular sclerite of the galea (varsgal).
Stipital sclerite II (ssn; = subgalea of authors) is a large, rectangular piece.
Distally it bears the galea, laterally the lacinia; its ventral surface bears a
wide oblique ridge, mesal to which the surface is strongly concave. The con-
cave area fits upon the apex of the labium and is hidden when the maxillae
are closed. Stipital sclerite III (ssm; = proxagalea of authors) is an elongate
narrow piece interposed between the apices of sclerites II and IV. Stipital
sclerite IV (ssiv; = palpifer of authors) is also elongate and forms the lat-

Evolution of the Phanaeine Dung Beetles 753

eral portion of the maxilla. Only sclerite IV is well developed dorsally (Fig.
24). Originating on its base are two muscles, one inserting on the basal seg-
ment of the maxillary palpus, the other on the dorsal articular sclerite of the
galea (darsgal).

The lobate galea (gal; = distagalea of authors) is articulated with the
stipes by two transverse sclerotized bands, the articular sclerites of the galea.
Both sclerites are free, one in the dorsal, the other in the ventral membrane
of the base of the galea. Except for the articular sclerites, the galea is a thick,
membranous lobe which is dorso-ventrally flattened and flexed ventrally
over the apex of the labium. Dorsally it bears a dense pad of short, erect
setae (Fig. 24).

The lacinia (lac) is an apically widened and truncate, flattened lobe.
Since it is united laterally with stipital sclerite II and not intrinsically mus-
culated, it is presumably not moved independently. Dorsally can be seen a
sclerotized strip which basally bears a dense tuft of rather long setae. The
apex of the lacinia bears a dorsal, densely setose pad like that on the adjacent
surface of the galea.

The maxillary palpus (mxpl) articulates in a large socket in the apex of
stipital sclerite IV. It is composed of four cylindrical segments which are
successively longer distally. The apex of the distal segment bears a small,
supposedly sensory, membranous spot.

The cardo and stipital sclerites I and IV bear numerous long, stiff setae
(not shown in Figs. 23-24). The remainder of the appendage, including the
palpus, is either bare or with small, inconspicuous setae.

The four extrinsic muscles of the maxilla insert on conspicuous apodemes.
Those three originating on the tentorium (and gula) have been described
above. The letters designating apodemes in Figures 23 and 24 correspond to
those designating muscles in Figures 13-15. The fourth (Fig. 24, mlac)
originates on the head wall and inserts on the base of the lacinia.

As is generally true of beetles (Dorsey, 1943), the labial musculature of
scarabaeines is considerably reduced and modified, making interpretation
of the labium difficult by standard morphological criteria. Extrinsic muscles
are absent and intrinsic ones are reduced in number. Likewise, the gross
structure of the labium is curiously modified and difficult to compare with
the generalized plan.

The submentum is generally identified as that postlabial sclerite on which
originate the median muscles of the labium (Snodgrass, 1935; = submento-
premental muscles of Matsuda, 1965). This pair of muscles is either absent
or considerably modified functionally in dung beetles. My designation of
the submentum (smt) in Figure 9 and elsewhere is based primarily upon its
position relative to the posterior tentorial pits (gular sutures) and its pre-
sumed correspondence with similar, definitely premental plates in other

754

The University Science Bulletin

27

pmts

mlbhpx
mshpx

Ishpx

29

mrghpx

Figs. 25-29. Phanaeus vindex, male. 25, mcntum and prementum, ventral view (setae
shown unilaterally); 26, prementum-hypopharynx, oral surface (setae shown unilaterally; floor
of buccal cavity not shown, cf., Fig. 30); 27, right glossa, mesal surface; 28, median groove of
prementum (glossae torn away; only apex of median lobe of hypopharynx shown); 29,
hypopharyngeal sclerites, dorsal view.

beetles. Snodgrass (1935:294) called the corresponding sclerite in Phyllo-
phaga (Melolonthinae) the mentum, the submentum being represented by
an internal plate on which originate the median muscles of the labium. He
did not suggest the possibility that this internal plate represents only a pos-
terior apodemal inflection of the submentum, the remainder of the sclerite
being the exposed area distal to the gula. Kadic (1902), followed by Harden-
berg (1907) and Mohr (1930), referred to the present submentum as the pos-
terior plate of the submentum, the present mentum being the anterior plate.
As described above, the submentum forms a part of the ventral head wall;
distally it projects freely between the bases of the maxillae.

The mentum (Fig. 25, mt) articulates with the straight distal margin of

Evolution of the Phanaeine Dung Beetles 755

the submentum. It is emarginate anteriorly and densely clothed with long,
stiff setae. Its postero-lateral angles are inflexed mesally; on them articulate
the ventral extremities of the hypopharyngeal suspensorium (Fig. 33).

The remaining labial structures are borne on the dorsal surface of the
mentum and, except for the palpi, are hidden. The prementum is repre-
sented by a pair of premental sclerites (Fig. 30, pmts), the wedge-shaped
ventral angles (vapmts) of which articulate with the antero-lateral angles of
the mentum. In many dung beetles, including some phanaeines, the pre-
mental sclerites are membranous medially and only a C-shaped margin is
sclerotized (Fig. 150).

The three-segmented labial palpi (Fig. 25, lbpl) articulate on the ventral
angles of the premental sclerites (Figs. 25, 30). The two proximal segments
bear many long, stiff setae, while the distal segment bears a single, very
small seta on the outer surface. As does the distal segment of the maxillary
palpus, the distal segment of the labial palpus bears a small, presumably
sensory, membranous apical spot. On the basal segment of each is inserted
a single muscle originating on the posterior margin of the adjacent pre-
mental sclerite.

Projecting anteriorly from between the premental sclerites are two elon-
gate, membranous lobes, the glossae (Fig. 26, gl). The use of the term glossa,
however, is loose and not meant to imply either that paraglossae are present
or that they have presumably been lost. The glossal structure of all dung
beetles and other scarabs examined is problematic.

In all dung beetles examined, except Anomiopsoides, the mesal surfaces
of the glossae are covered by a thick mat of short setae (Fig. 27) like that on
the galeae and laciniae. Projecting mesally from the lateral margin of each
glossa is the glossal flap (Figs. 26-27, glf), a translucent membranous flap
fringed with long hairs. Seen mesally (Fig. 27), the glossae of phanaeines
and some other dung beetles are apically bibbed. The slit separating these
lobes is usually obscured by setae. One could assume that this slit is an
incomplete division (latarima of MacGillivray, 1923) between the glossa and
paraglossa, which are fused basally. However, consideration of the glossae
of such scarabaeines as Heliocopris, which are trilobed, and of Catharsius,
which are not notched, considerably weakens such an assumption.

The glossae are supported laterally by the glossal bars (Figs. 26, gib). Pos-
teriorly these bars project internally (Fig. 32, apglb) and are closely connected
but not fused with the antero-dorsal angles of the lateral labial sclerites
(Figs. 26, 30, libs).5 Inserting on the connection of each of these two pieces is

5 In Heliocopris the position of the lateral labial sclerites is occupied by a heavily sclerotized
posterior extension of the glossal bar. This condition suggests that where lateral labial
sclerites occur, they represent disjunct portions of the glossal bars. The strength of this sug-
gestion, however, rests on a knowledge of the condition of the glossal bar in a wider variety
of scarabaeines than that used for this study.

756

The University Science Bulletin

pors

mrghpx

mrghpx

vapmts

Figs. 30-33. Phanaeus vindex, male. 30, extracted labium-hypopharynx (without sub-
mentum), lateral view (broken outline refers to Fig. 32); 31, posterior wall of hypopharynx;

32, suspensorium of hypopharynx and oral arms, lateral view (broken outline refers to Fig. 30);

33, same, dorsal view (hypopharynx and prementum removed).

a muscle originating medially on the anterior margin of the submentum
(Fig. 13, e). Since each muscle diverges from its origin, I assume that, as a
pair, they function primarily to spread the glossae. They could, however, be
interpreted either as glossal muscles, whose origins have shifted from the
presumably primitive prelabial position to the postlabium (submentum), or
as the median muscles of the labium, which functionally have become ab-
ductors of the glossae rather than levators of the prementum.

In certain phanaeines, including vindex and some other dung beetles, the
glossae come together at least basally (Fig. 26). In others, their bases are
separated to varying degrees (Fig. 149). If the glossae are spread apart, the
median groove of the prementum (Fig. 28) is exposed. The raised sides of
this groove bear long, mesally-directed setae. Its floor is weakly sclerotized
and bears dark, generally scattered sensory spots (sen) which, on internal
examination, resemble sensilla coeloconia.

Posteriorly the glossae are continuous with the hypopharynx, and together

Evolution of the Phanaeine Dung Beetles 757

they form a large, quadrate lobe on the dorsal surface of the mentum (Fig.
30). The membranous median lobe of the hypopharynx (Figs. 26, 28, mlbhpx)
is above the apex of the median groove of the prementum and slightly cov-
ered laterally by the bases of the glossae. In species with separated glossae,
the median lobe is completely exposed (Fig. 149) and covered with fine hair.
Directly beneath the dorsal surface of the hypopharynx are three hypo-
pharyngeal sclerites. A Y-shaped lateral hypopharyngeal sclerite (Fig. 29,
lshpx) is on each side of the median sclerite (mshpx). Figure 29 shows their
positions relative to the median lobe. Inserted on a hook-shaped ventral
process of the median sclerite is a single (unpaired) muscle originating
medially on the mentum (Fig. 13, f).° It supposedly depresses the median
lobe of the hypopharynx. The vertical posterior wall of the hypopharynx is
concave on each side of the postero-median ridge of the hypopharynx (Fig.
31, mrghpx). This ridge is bowed to the left to fit between the asymmetrical
molar lobes of the mandibles, which fit against the hypopharynx.

Associated with the hypopharynx is a system of sclerotized bars com-
prising the suspensorium of the hypopharynx (Figs. 32-33). This supporting
structure is well developed in dung beetles and consists of two lateral arms
(la; = suspensoria of Snodgrass, 1935) joined by a thicker, transverse bar.
The lateral arms also support the membranous floor of the buccal cavity;
their ventral ends articulate with the mentum. Extending dorsally from the
transverse bar is the dorsal arm of the suspensorium (da). It projects freely
through the body of the hypopharynx to beneath the dorsal extremity of the
postero-median ridge (Fig. 32) and supposedly prevents collapse of the
otherwise soft, membranous hypopharynx.

The oral arms (Figs. 32-33, oa) are connected end-to-end with the lateral
arms by thickened membranes. Ventrally they are bent mesally beneath the
mouth (Fig. 33, mth). The dorsal ends of the oral arms are greatly expanded
muscle discs ("tendon caps") on which insert short, thick muscles originating
on the dorsal head wall behind the cephalic process. Since the arms are
linked with the lateral arms of the suspensorium, which in turn articulate
with the mentum, their muscles presumably effect elevation of the mentum
(and prementum). From a functional morphological point of view, such
function of these muscles, which are generally retractors of the hypopharynx
(Snodgrass, 1935), is logical since median muscles (elevators) of the pre-
mentum are apparently absent (see above).

The paraoral sclerites (Fig. 30, pors) are oval, sclerotized thickenings in
the lateral walls of the buccal cavity. On each is inserted a muscle from the
head wall which presumably dilates the cavity.

"Dorsey (1943) describes a similar unpaired muscle ("median muscle of the labium") in
Popillia japonica (Rutelinae), but he finds it in no other beetle examined (p. 32), including
Macrodactylas (Melolonthinae) .

758 The University Science Bulletin

The Cervix

The cervix is the membranous region extending from the margin of the
occipital foramen to the inflexed anterior margin of the prothorax. It is
fairly long and allows the head to be forcibly withdrawn beyond the pro-
thorax, although the head is probably never protracted this far by the living
beetle. The cervix allows free movement of the head, but movement is
largely restricted to a vertical plane by the tight fit of the posterior portion
of the head within the prothorax.

Located latero-ventrally in each side of the cervix is a single elongate
cervical sclerite (Fig. 34). It articulates basally with the inflexed antero-
lateral angle of the propleuron and apically with the rounded ventro-lateral
angle of the occipital foramen. The setose external surface of the sclerite is
concave; the internal surface is ridged for muscle attachment. Upon re-
traction of the head, the apices of the sclerites fold dorso-mesally against the
recessed anterior prothoracic margin. Situated dorsally in the cervix near
the posterior head margin is a pair of lightly sclerotized, setose pads. Halffter
(1961) has referred to similar pads in Canthon as anterior cervical sclerites.
In view of their dorsal position and the double articulation of the cervical
sclerite, it seems doubtful that they represent anything more than thickenings
of the cervical membrane.

The Thorax
the prothorax

The large, box-like prothorax accommodates the strongly developed
muscles effecting movement of the head and front legs, the principal digging
organs. Its anterior margin (to which is connected the cervical membrane)
is recessed. The sclerites contributing to the anterior margin are all strongly
inflexed and shaped to receive the posterior part of the head.

Like the head, the pronotum of dung beetles and other "horned"
scarabaeids often bear outgrowths (protuberances, ridges, etc.) producing
forms characteristic of species. As are those of the head, such outgrowths are
allometric and usually only the males bear well developed "armature." The
pronotum of a well developed male of vindex is shown in Figures 35 and
36; that of the female in Figures 3>7 and 38. In less well developed ("minor")
males, its angulate appearance is less well marked. Sexual differences as
exhibited by the prothorax are pronotal; the propleurae and sterna of both
sexes are respectively alike.

In phanaeines, a distinct postero-median angle of the pronotum (Fig. 38,
5) fits between the bases of the elytra and covers the mesonotum. In some
species this angle bears two distinct impressions, the posterior pronotal fossae
(Fig. 194, pfos). On each side of the pronotum of all dung beetles (except
some Canthonina— Halffter in litt.) there is a shallow, usually smooth de-
pression, or lateral pronotal fossa (Fig. 36, lfos), marking an internal tubercle

Evolution of the Phanaeine Dung Beetles

759

Figs. 34-38. Phanaeus vindex. 34, right cervical sclerite, lateral view (post, indicates
posterior end); 35, prothorax of well developed male, dorsal view; 36, same, lateral view; 37,
prothorax of female, lateral view; 38, same, dorsal view (1-5, sections and angles of circumnotal
ridge: 1 — anterior portion, 2 — antero-lateral portion, 3 — lateral portion, 4 — antero-lateral
angle, 5 — postero-median angle).

on which articulates the base of the procoxa (Fig. 40, acx). In both sexes of
vindex the pronotum is very strongly rugose (Fig. 165).

The circumnotal ridge (Figs. 35-36, pntm) is a sharp ridge separating the
definitive pronotum from the rest of the prothorax. For descriptive con-
venience, various sections and angles of this ridge are defined in Figure 38.
The sinuous lateral portion (3) is characteristic of the majority of phanaeines.
The transverse anterior portion (1) is widened and in most dung beetles is
continuous with the antero-lateral portions (2). In Phanaeus, however, the
anterior portion is interrupted on each side by a distinct, rounded emargina-

760 The University Science Bulletin

tion (Fig. 35, emarg) into which fits the prominent postocular lobe of the
parietals. The morphological significance of the circumnotal ridge is unclear.
It is present in all scarabaeids and separates a dorsal area of the prothorax
(pronotum) which differs usually in integumental sculpturing and often in
color from the ventral areas (pleura and sternum).7

A very large, round opening in the flattened posterior wall of the pro-
thorax (Fig. 39) fits over a protruding rim formed by parts of the meso-
thorax. The lateral portions of this wall are widest and shaped so as to fit
snugly against the opposing areas of the mesopleura.

Except postero-laterally, the propleura (Figs. 36, 39, ppl) are ventral areas
of the prothorax. Each propleuron is setigerously punctured except for a
smooth, raised area lateral to the procoxal cavity. As seen ventrally (Fig. 39),
their mesal margins are the curved lateral margins of the coxal cavities. A
postcoxal extension (postcoxale) of each pleuron is joined with the lateral
angle of the sternellum (stn) by a double pin-and-socket mechanism shown
in Figure 42. Each propleuron is fused anteriorly with the lateral margin of
the basisternum along the pleurobasistemal suture (Fig. 39, pbs). The
propleura of a few phanaeines (Dendropaemon) and some other dung
beetles (e.g., Copris and Earysternus) are each divided by a transverse
propleural carina (Figs. 179-180, arrows). This carina extends laterally from
the coxa, separating a "proepisternum" and "proepimeron." Since there is
never an internal ridge corresponding to it, the carina is probably best in-
terpreted as a secondary integumental feature rather than as a primary
division between pleurites. In many dung beetles, including some pha-
naeines, the anterior portion of each propleuron ("proepisternum") is
strongly concave or otherwise set off from the "proepimeron."

Coincident with the expansion of the prothoracic box, the procoxae have
become transversely oriented, cylindrical structures whose apices nearly touch
each other mesally. The propleura have expanded ventro-mesally and cover
the bases of the coxae, which articulate internally with the lateral pro-
notal walls. Their transverse orientation has affected the structure of the
prostemum, a portion of which is depressed and hidden by the apices of the
coxae. Thus, only a portion of the prosternum is visible with the coxae in situ.

The exposed portion of the basisternum (Fig. 39, bst) is anterior to the
front coxae and between the pleurobasistemal sutures. Its anterior margin
(Fig. 41, imbst) is reflexed posteriorly and shaped to receive the gula and

7 Butt (1944) has not differentiated notal and pleural areas in the prothorax of Amphimallon
(Melolonthinae), but, rather, considers all prothoracic areas except the sternum as pronotal (p.
11). There seems to be some validity to his designation (which could be applied equally well
to dung beetles) if, for example, one negates the supposedly "homodynamous" evolution of all
thoracic segments, particularly their pleura. Nevertheless, mainly because of its established
use in beetle taxonomy, I choose to maintain at least descriptive distinction between pronotum
and propleura.

Evolution of the Phanaeine Dung Beetles

761

cxd

ppl

apst

imbst

imstn

Figs. 39-43. Phanaeus tindex, male. 39, prothorax, ventral view (right coxa in place;
setation shown on beetle's right only; xs=plane of cross section in Fig. 40); 40, same, cross
section through right side; 41, prosternum, internal view (right prosternal apophysis removed;
setation shown on left side only); 42, interlocking mechanism of sternellum and propleuron,
right side (broken outline indicates cavities in sternellum receiving pleural lobes); 43, sternellum
and adjacent parts, posterior view (mem=intersegmental membrane).

adjacent occipital areas of the head. The basisternal crest (Figs. 39-40,
bstc) is a median, longitudinal ridge extending posteriorly between the apices
of the coxae to the anterior angle of the sternellum. A tuberculate swelling
on each side of this crest articulates in a groove (Fig. 60, groove) in the apex
of the coxa. These tubercles are secondary sternal articulations of the front
coxae, primary articulations (with the lateral margins of the sternum) being
hidden by the extensive overlap of the coxae by the pleura. The exposed
portion of the sternellum (Fig. 39, stn) is triangular, the apex directed be-

762 The University Science Bulletin

tween the coxae and fused with the basisternal crest. Its posterior margin
(Figs. 41, 43, imstn) is reflexed anteriorly and forms a shelf fitting over the
mesosternum.

Removal of the front coxae reveals the hidden portion of the presternum,
a smooth trough on each side of the basisternal crest in which rotates the
apex of the coxa (Fig. 39, cxd). Projecting from the lateral margin of each
trough is a prosternal apophysis, a petiolate structure composed of three
vertical lamellae (Figs. 39, 41, apst). To the apophyses are attached muscles
from the posterior head margin, from the procoxae and from phragma I.
The positions of these apophyses indicate that the depressed portions of the
prosternum are composed of roughly equal parts of basisternum and sternel-
lum, although there is no visible line between the two.

Located latero-ventrally in the pro-mesothoracic intersegmental mem-
brane on each side of the sternellum is a very large, spindle-shaped thoracic
spiracle (Fig. 43, tspr). There is only one pair of thoracic spiracles. Between
them is a V-shaped, setose interspiracular pad (ip), which, along with the
spiracles, is probably never exposed during normal activity. Ritcher (1969a,
b) discusses the spiracles of vindex and other scarabaeids.

THE PTEROTHORAX

Restriction of flight power to the hind wings is correlated with a pro-
portionately large metathorax, which contributes the bulk of the volume of
the pterothoracic box. Except for the mesepimera, the mesothorax forms a
constricted rim over which fits the prothorax. Except in Anomiopsoides,
which lacks hind wings, the pterothoracic structures of all scarabaeines ex-
amined are quite similar.

As seen dorsally, in situ (Fig. 44), the mesonotum appears as a single,
pentagonal sclerite between the articular processes of the elytra. Much of it,
however, is inflected and hidden by the close approximation of adjacent
sclerites. Figures 45 and 46 show these external aspects of the mesonotum.
Interpretation and naming of the mesonotal areas are difficult. A "V-suture"
(Snodgrass, 1935 and elsewhere) separating scutum and scutellum is either
lacking or unrecognizable. Furthermore, internal ridges, while well de-
veloped, are probably not indicative of basic notal divisions. Primarily be-
cause standard taxonomic terminology is based on it, it seems best to follow
Snodgrass' interpretation (1909 and elsewhere) of the pterothoracic nota.
Accordingly, the dorsally exposed, pentagonal mesonotal area is the scutel-
lum (Fig. 44-46, sct2). In a few dung beetles (e.g., Eurysternus, Megathopa
and Onitis) and in many other scarabaeids, a posterior, triangular portion of
the mesoscutellum (usually very small in dung beetles) is permanently ex-
posed between the bases of the elytra and not covered by the pronotum. The
mesoscutum (Fig. 45, scu) is divided by the scutellum and represented on

Evolution of the Phanaeine Dung Beetles

763

set 2

epis2

H£g^

Figs. 44-49. Phanaeus vindex, male. 44, pterothorax, dorsal view (elytra and hind wings
removed; setation shown on right only); 45, extracted mesonotum, lateral view; 46, same,
posterior view; 47, extracted mesopleuron, internal view: 48, extracted metanotum, anterior
view; 49, same (stippling indicates internal structures).

764 The University Science Bulletin

each side of it by a rounded inflection of the notum on which articulates the
first elytral axillary sclerite. A longitudinal, mid-dorsal line on the scutellum
marks a deep, vertical plate (Fig. 46, vl) fused anteriorly to phragma 1. This
plate intersects a narrower, horizontal plate (hi) extending between the
scutal areas and also fused to phragma 1.

The posterior infection of the mesonotum (Figs. 45-46, nt,,i) is strongly
downflexed and widely emarginate. The origin of this margin is not clear;
it may be scutellar, a remnant of the mesopostnotum, or of another origin.
Functionally, however, it appears to be a mechanism allowing firm contact
of the pteronota. The rounded (lower) margin of the inflection is joined
very closely to the margin of phragma 2 by the intersegmental membrane.
Articulating at each side of the inflection is a vertical, triangular plate corre-
sponding to that which Snodgrass (1909) refers to as sclerite q (Fig. 46, q)
in other beetles. Snodgrass suggests that sclerites q might represent the re-
mains of a mesopostnotum. Inserted on it is a large muscle originating on
phragma 1.

At each side of the mesonotum, in front of the articular processes of the
elytra, is a rectangular mesobasalare (Figs. 44, 50, 52, bas2). Each is produced
subintegumentally as an acute apodeme (Fig. 52, abas2) on which originate
two muscles, one inserting on the mesocoxal margin, the other on the base
of the trochanter.

The metanotum more nearly conserves typical pterothoracic structure
than does the mesonotum. It is irregularly convex and much wider than the
mesonotum. Figure 44 shows its surface areas and sutures. Like the meso-
scutum, the metascutum (scu3) is divided medially by the metascutellum
(sct3), from which it is separated by the V-suture (v). An oblique suture
(s) on each side of the scutum sets off a triangular, postero-lateral scutal area,
the lateral angle of which is the posterior metanotal wing process (pwp).
Originating on each side of the metascutum are two large muscles, both of
which insert on the anterior margin of the metacoxa. The prescutum (Fig.
48, pscu) forms an anterior surface of the metanotum hidden in situ by the
mesothorax and separated from the scutum on each side by a weak, trans-
verse w-suture (Figs. 44, 48, w). Lateral projections from the prescutum are
the anterior metanotal wing processes (awp). The paraphragmal areas of
the prescutum (Fig. 48, ppha) are fused with the lateral margins of phragma
2 along the line of attachment of the intersegmental membrane, and the
composite structure (i.e., paraphragmal areas + phragma 2) is separated
from the body of the metanotum by a narrow, transverse membranous area
(mem). For clarity, Figure 48 is modified in Figure 49 to contrast internal
(apodemal) metanotal areas (stippled in Fig. 49) with external ones. In-
serted on the paraphragmal areas are muscles originating on the metasternum
and comprising the tergo-sternal indirect flight muscles. The metapostnotum

Evolution of the Phanaeine Dung Beetles

765

epim

st 3

S1+2

Figs. 50-51. Phanaeus vindex, male. 50, pterothorax, lateral view (elytra and coxae in
place; mem=intersegmental membrane); 51, same, ventral view (right coxae removed).

(Figs. 44, 53A, pstn) is reduced to a narrow ridge united with the posterior
metanotal margin and bearing phragma 3.

The mesopleura are compressed. The mesepisterna (Figs. 50, 52, epis2)
form the sides of the projecting mesothoracic rim fitting within the pro-
thorax; basally they are united with the mesosternum. A ridge -like inflection
of the anterior margin of each mesepisternum (Fig. 47, imepis) is fused
dorsally and ventrally with the mesopleural ridge. On the dorsal end of this
inflection is a disc (md) on which originates a muscle inserting on the
antero-lateral angle of the mesonotum. The mesopleural suture (Fig. 52,
pls2) is an impressed, vertical groove separating the mesepisternum and
mesepimeron. Its internal invagination is the mesopleural ridge (Fig. 47,
plrg), the prominent dorsal and ventral ends of which are, respectively, the
mesopleural wing process (wp2) and pleural articulation of the mesocoxa
(acx 2). The pleural ridge is the origin of a large muscle inserting on the
mesocoxal margin. Each mesepimeron has three distinct surfaces, anterior

766

The University Science Bulletin

st3

scu3

sct3

Figs. 52-53. Phanaeus vindex, male. 52, pterothorax, anterior view (elytra and mesocoxae
in place; setation shown on right only); 53A, same, posterior view (elytra and metacoxae re-
moved); 53B, dorso-posterior angle of metasternum, lateral view.

(Fig. 52, epim2a), lateral (Fig. 50, epim2i) and posterior (Fig. 47, epimi,),
which are set off by sharp, vertical bends in the body wall. Of the three,
only the lateral surface is permanently exposed and colored and sculptured
like adjacent exposed sclerites. The posterior surface is pressed close to the
anterior surface, such that in Figure 47 care must be taken to note the
vertical posterior margin of the mesepimeron (epimPm) to distinguish in-
ternal and external surfaces. A smooth, elongate area (y) on the posterior

Evolution of the Phanaeine Dung Beetles 767

surface receives the opposing margin of the metepisternum; a similarly
smoothed, concave area of the lateral surface (z) contacts the base of the
mesocoxa.

The metapleurites are oriented horizontally (see Snodgrass, 1909). The
metepisternum has two distinct surfaces, anterior (Fig. 57, episa) and lateral
(Figs. 50, 54, epis3i). The lateral surface is trapzoidal, and its dorso-posterior
angle (Figs. 54, 56, tab) is produced as a rounded metepisternal tab which
fits in a depression in the elytral epipleuron. The metepisternal tab is char-
acteristically present in phanaeines. The metepimeron (Fig. 54, epim) is a
narrow, horizontal sclerite along the dorsal edge of the episternum. It is
completely hidden beneath the epipleuron of the closed elytron. Posteriorly,
a finger-like portion of the epimeron (Figs. 53A, 56) projects mesally and
unites with the lateral extremity of the metapostnotum. Separating the
metapleurites is a faint horizontal line, the metapleural suture (Fig. 54, pis).
Its invagination, the metapleural ridge (Fig. 55, plrg) is a very strongly
developed, oblique plate; Figure 55B shows the origin of this ridge relative
to the metapleurites. The antero-dorsal extremity of the metapleural ridge is
the metapleural wing process (Figs. 54-57, wp), which supports the second
axillary sclerite and base of the Sc vein of the hind wing. Posteriorly the
metapleural ridge unites with a vertical ridge (Fig. 55A, rg) supporting the
mesal projection of the metepimeron, and the two converge on a rounded
lobe (Figs. 54-55, lobe) which inserts in a slot in the dorsal margin of the
metasternum. This lobe is not the pleural articulation of the metacoxa. The
functional "pleural" or dorsal articulation of the metacoxa (Figs. 53A-53B,
dacx3) is a tubercle borne on the dorso-posterior angle of the metasternum.
Whether or not this tubercle is the dissociated extremity of the pleural ridge
or a secondary articular structure is not clear.

The metabasalare (Fig. 57, bas) is fused with the anterior surface of the
metepisternum, and together the two form an irregular vertical wall fitting
against the posterior surface of the mesothorax. The stippled area in Figure
58 (c.f., Fig. 57) is the internal portion of this anterior wall. (In some
beetles [e.g., Carabidae] the metabasalare is free, but in Scarabaeidae it is
apparently always fused to some degree with the metepisternum [see Snod-
grass, 1909 j.) The apex of the basalare (Fig. 59, basp) is elongate and closely
associated but not fused to the metapleural wing process (wp); it provides a
second fulcral point for the base of vein Sc of the hind wing. Internally the
base of the basalare bears an enormous, oval muscle disc (Figs. 56-57, basmd).
From this disc extend two muscles, a very large one to the metasternum and a
much smaller one to the posterior margin of the metacoxa. Since the basalare
is rigidly fused with the body wall, it is quite likely that the basalar-meta-
sternal muscles augment the action of the tergo-sternal indirect flight muscles.

The mesosternum (Fig. 51, st2) is a narrow, transverse sclerite between

768

The University Science Bulletin

Figs. 54-59. Phanaeus vindex, male. 54, right metapleuron, lateral view; 55A, same, in-
ternal view (basalare removed; xs=p!ane of cross section in 55B); 55B, same, cross section;
56, same, dorsal view (basal membrane of hind wing removed); 57, same, anterior view;
58, same as 57 (stippled area is internal); 59, apices of metabasalare and metapleural wing
process, inner view.

the apices of the mesocoxae. It is convex and bears a transverse, brace
( ) -shaped ridge which contacts the inflexed posterior margin of the

prosternellum. The mesosternum is fused with the median area of the meta-
sternum along the mesometasternal suture (sts).

The very large metasternum (Fig. 51, st3) forms most of the ventral wall
of the pterothorax. It is superficially divided into three parts by the elongate
mesocoxae, which rotate in depressions of the metasternum (cxd) similar to

Evolution of the Phanaeine Dung Beetles 769

the procoxal depressions of the presternum. The shape of the median area
of the metasternum (between the mesocoxae) is characteristic of phanaeines.
Slightly posterior to the mesometasternal suture, it is strongly and abruptly
depressed medially as a vertical angulate prominence. Consequently, as seen
laterally (Fig. 50), the metasternum is flat and squared anteriorly; seen ven-
trally (Fig. 51), it is pentagonal and acuminate anteriorly. Since this depres-
sion of the median area is normally viewed from below, the metasternum of
phanaeines is often described as "raised."

THE PTEROTHORACIC ENDOSKELETON

The pterothoracic structures of interest here are the phragmata and
sternal apophyses; other internal skeletal features are discussed above.

Phragma 1 (Figs. 45-46, 52, phi) is a deeply invaginated plate borne by
the anterior margin of the mesonotum. Medially its ventral margin is bent
posteriorly, resulting in the bilobed appearance seen anteriorly (Fig. 52). On
its anterior surface originate a pair of muscles inserting on the prosternal
apophyses; each side of the posterior surface is the origin of two muscles, a
large one inserting on the mesocoxal margin and a smaller one on sclerite q.
There are apparently no well developed muscles extending between phrag-
mata 1 and 2. Phragma 2 (Fig. 48, ph2) was mentioned previously.

Phragma 3 (Fig. 53A, ph3) is borne by the metapostnotum. Like
phragma 1, it is emarginate medially for passage of the aorta and digestive
tract. At the sides of the emargination are two dark, oblique lines which
separate a median area; from this area extend short muscles (the dorsal
longitudinal indirect flight muscles) to phragma 2. It is noteworthy that
these muscles, which effect the downstroke of the wing, are so poorly de-
veloped compared to the extensive development of their antagonists, par-
ticularly the tergo-sternals and basalar-sternals. From the lateral areas of
phragma 3 extend muscles, the "lateral obliques," inserting along the anterior
margin of the metascutum. On the lateral processes of phragtiia 3 (ph3ip)
attach muscles from the metafurcal arms that probably stabilize the meta-
furca.

Both the mesosternal and metasternal apophyses are fused basally. The
resulting compound structures, the mesofurca and metafurca, are quite dif-
ferent in shape. The base of the mesofurca (Fig. 52, bfur2) is a transverse,
vertical plate marked externally by the mesometasternal suture. In phanaeines
and some other scarabaeines, it is braced by a vertical, longitudinal plate, the
mesof ureal brace (br). The mesof ureal arms (afur2) are cup-shaped, open-
ing laterally; within them and on their posterior surfaces originate muscles
inserting on the mesocoxal margin. A ribbon-like, lateral extension of each
arm (afur2p) serves as the origin of a muscle extending through the coxa
and inserting on the base of the trochanter.

770 The University Science Bulletin

The metafurca (= metendosternite of Crowson, 1938, 1944) is an an-
teriorly inclined, T-shaped invagination of the metasternum (Fig. 53A). It
is borne by a rounded, sternal inflection, the metafurcasternum, (fst3), be-
tween the bases of the hind coxae. The tubercle on each side of the fur-
casternum is a ventral (sternal) articulation of the metacoxa (vacx3). The
base of the metafurca (bfur3) is elongate and, in cross section, T-shaped (as
a steel T-beam). The metaf ureal arms (afur3) are subapical (the apex of
the metafurca is not visible in Fig. 53). On the metafurca, as a whole, orig-
inate five pairs of muscles, all of which insert on the metacoxa.

THE LEGS

The fossorial front leg of vindex is shown in Figures 60 and 61. The coxae
(ex) are cylindrical and articulate laterally with the internal, tuberculate
lateral articulation of the procoxa (Fig. 40, acx), which fits into a small
notch (Figs. 60-61, n) in the coxal margin. The concealed trochantin (Fig.
62) articulates anteriorly on the coxal margin. The trochanter (Figs. 60-61,
tr) is fused to the base of the femur and largely hidden in the apex of the
coxa; the femur-trochanter line of fusion is visible only posteriorly. The
inner margin of the femur (fe) is flattened to allow close appression of the
tibia. The profemoral hair plate (Fig. 60, fet) is a circular group of closely
packed, presumably proprioceptive setae on the anterior surface of the femur.
When the femur is retracted, this hair plate slips behind the apex of the coxa.
The tibia (tib) is flattened antero-posteriorly and bears three lateral protibial
teeth (tibt). The anterior (atibc) and posterior tibial carinae (ptibc) (Figs.
60-61, respectively) extend the length of the tibia. Abrupt, lateral turns of
the anterior carina extend onto the surface of the two distal tibial teeth. In
the discussions that follow, tibial teeth bearing such extensions of the anterior
tibial carina will be called carinate. A single protibial spur (tibs) articulates
on the inner apical angle of the tibia; in vindex and most other phanaeines,
it is pointed and only slightly curved. In other dung beetles, the spurs may
show sexual differences (e.g., some Canthon) or extraordinary modifications
(Onitini), but they are apparently rarely absent.

In most dung beetles protarsi are well developed and found in both sexes.
In phanaeines, however, they are found only in the females of certain groups,
including Phanaeus, where they are always reduced in size and clawless
(Fig. 63). When present, it inserts on the inner apical angle of the tibia,
mesal to the tibial spur. No known male phanaeines possess front tarsi; their
curious lack in phanaeines and some other dung beetles has been known for
many years (Brulle, 1837; Dejean, 1838).

The middle and hind legs differ principally in length and in articulation
and orientation of the coxae. The middle leg (Fig. 65) is shorter and
stockier than the hind leg (Fig. 66). The cylindrical mesocoxae (Figs. 51,

Evolution of the Phanaeine Dung Beetles

771

groove

Figs. 60-66. Phanaeus vindex, female. 60, left foreleg, anterior view; 61, same, posterior
view; 62, trochantin, anterior view; 63, front tarsus; 64, right middle tarsus, ventral ( =
plantar) surface; 65, right middle leg, ventral view; 66, right hind leg, ventral view (coxa
removed; see Fig. 50). (Setae are shown only in Figs. 63-64.)

cx2, and 65, ex) are oriented longitudinally and approximately parallel in
most phanaeines. The metacoxae (Fig. 51, c\5) are elongate and always
oriented transversely. The middle coxa has but a single, pleural articulation
(to allow rotation along its long axis), while the hind coxa has two. Neither
posterior pair of legs has a recognizable trochantin (contrary to Snodgrass,

772 The University Science Bulletin

1909: 563). The trochanter (tr) of both pairs is triangular and fused to the
base of the femur. Their femora differ essentially only in length. The tibiae
(tib) are only slightly shorter than their respective tarsi. They differ in
shape, the middle tibia being round in cross section, the hind tibia being
approximately rectangular because of its flattened dorsal surface. The mid-
dle tibiae each bear two unequal apical tibial spurs (tibs) ; each hind tibia,
only one. The lateral surface of the hind tibia of a few phanaeines bears a
subapical transverse tibial carina (Fig. 262, cr), which is absent in vindex
(Fig. 261). This carina is uncommon in phanaeines but is characteristic of
certain other groups, such as Copris. Middle and hind tarsi are well de-
veloped but always flattened and clawless in phanaeines. Figure 64 is a
posterior tarsus of vindex; as in all phanaeines, the ventral (plantar) surface
of each tarsomere except the last is longitudinally carinate, while the dorsal
surface is smooth.

THE WINGS AND THEIR ARTICULATIONS

The elytra (Fig. 67; 50, ely) completely cover the metanotum, metepi-
mera and all but the last abdominal terga. The humeral angle (h) of each
elytron is swollen; beneath it lies the raised base of the folded hind wing.
The elytral epipleitron (Fig. 50, epip) is the downflexed lateral margin of
the elytron; in phanaeines it is narrow. The elytra of phanaeines are held
tightly in place when at rest by two complementary mechanisms, namely,
by the interlocking grooves in the opposing elytral margins along the elytral
suture and by containment of each elytron in the epipleural shelf by the
metepisternal tab. The epipleural shelf (Fig. 56, epips) is formed by the
recessed dorsal margin of the metepisternum, on which rests the epipleuron.
The anterior margin of each elytron is strongly downflexed (Fig. 52), form-
ing a vertical wall fitting against the prothorax. The articular process of
the elytron (Fig. 67, apely) is a lobate extension of the flexed anterior margin.
The elytral striae (Fig. 67, st) are longitudinal, impressed lines. The elytra
of phanaeines each have eight striae, including the often overlooked one (8)
adjacent to the epipleuron. The striae are numbered as in Figure 67 (1-8).
Each longitudinal zone between two striae is an inter stria (= interstice, in-
terspace). The interstriae of vindex are flat and coarsely rugose (Fig. 284).

There are two elytral axillary sclerites (axillaries) beneath the hollowed
underside of the articular process of the elytron (Fig. 68). The first (Figs.
68-69, axl) articulates with both the mesoscutum (at point a in Fig. 68) and
the mesopleural wing process (along ridge b). This double articulation sug-
gests that the first axillary is the functional equivalent of the first two axil-
laries found in other scarabs (e.g., Phyllophaga; Snodgrass, 1909), which
may be fused in phanaeines. The concave second axillary (sod) fits over the
first and is connected by a membranous thickening with the small flexor
sclerite of the elytron (Fig. 69, fs), which projects internally as a conical

Evolution of the Phanaeine Dung Beetles

773

apely

apely

axi

fs

ax2

>c;

/\

69

ax2

axi

ax3

71

j ug a I lobe

vannal lobe

Figs. 67-71. Phanaerts i index, male. 67, left elytron, dorsal view (strial number, 1-8, to
left of corresponding stria); 68, articular process of elytron and associated elytral axillary
sclerites, ventral view (a=lobe of axillary 1 articulating with mesoscutum; b=ridge of axillary
1 articulating on mesopleural wing process); 69, extracted elytral axillary sclerites, dorsal view;
70, base of right hind wing and adjacent metanotal area, dorsal view; 71, right hind wing,
dorsal view (bc=basal cell: other symbols indicate veins).

apodeme. Inserted proximally (nearest the body wall) on the flexor apodeme
is a thin "tendon" (no muscle) originating on an apodeme (Fig. 57, pr)
from the lower angle of the metabasalare. Inserting distally is a muscle-
bearing, flexible apodeme originating on a process (Fig. 47, pr) from the
mesepimeron. The insertions of the direct muscles of the elytron are difficult
to trace, but all originate on the mesopleural ridge.

The membranous hind wings are well developed; dung beetles are gen-

774 The University Science Bulletin

erally strong flyers. The reduced venation is of the scarabaeoid type discussed
by Stellwaag (1914), Forbes (1922), Graham (1922), d'Orchymont (1920),
Balthasar (1942), et al., no two of whom fully agree on vein terminology.
The system used here (Fig. 71) is largely a compromise of several systems.
The vein and area designations used are not meant to imply homology with
those of other insects, including beetles; this terminology is adopted here
strictly for comparative purposes within the group.

There are three axillary sclerites of the hind wing (Fig. 70). The first
(axl) hinges laterally on the anterior metanotal wing process and articulates
anteriorly with the base of vein Sc. The small second axillary (ax2) articu-
lates ventrally on the metapleural wing process. With the wing extended (as
in Fig. 70) the apex of the third axillary (ax3) is directed upward, beyond
the wing surface. When the wing is flexed, the third axillary rotates on its
long axis, the apex swinging mesally to above the posterior metanotal wing
process. Mesal to the third axillary and connected to it by a thickening in
the basal wing membrane is the flexor sclerite of the hind wing (fs). On the
flexor sclerite insert all three of the direct flight muscles of the hind wing,
two of which originate on the metapleural ridge, the third on the dorsal side
of the metabasalar muscle disc.

As external sclerites, subalares are lacking in the mesothorax, unless
represented by the flexor sclerites. In the metathorax, however, the subalare
is definitely represented by a stalked muscle disc (Fig. 53A, sub3md) inserted
on the basal membrane of the hind wing posterior to the third axillary
(point i in Fig. 70). From this disc extends a muscle to the anterior margin
of the metacoxa.

The Abdomen

the abdominal terga

The dorsum of the abdomen is composed of eight terga (Fig. 72, Ti~-
and pyg). The first seven are covered by the elytra and are thinly sclerotized
except for the acrotergites (at), thin bands across the anterior margins of
each. The ridge-like invagination of each acrotergite is an antecosta, an at-
tachment for dorsal longitudinal muscles. The acrotergite of tergum 1 is the
metapostnotum; its antecosta, phragma 3. Associated latero-ventrally with
each of the first seven terga is a pair of abdominal spiracles (spri-7). With
few exceptions, these spiracles of phanaeines are located in the lateral mem-
branous region separating terga and sterna.8 The large elliptical spiracles of
tergum 1 (Fig. 73) open anteriorly; the much smaller succeeding six pairs
open dorsally.

8 This arrangement of abdominal spiracles is characteristic of "Scarabaeidae — Laparosticti,"
a commonly employed but informal taxonomic division of the family including Scarabaeinae
(see Ritcher, 1969b).

Evolution of the Phanaeine Dung Beetles

775

(post.)

Figs. 72-78. Phanaeus vindex, male. 72, abdomen, dorsal view; 73, first abdominal spiracle,
anterior view; 74, abdomen, caudal view; 75, same, lateral view (terga 1-7 removed); 76,
genital capsule, lateral view; 77, same, dorsal view; 78, ninth sternum, ventral view (post, in-
dicates posterior margin).

The last (eighth) abdominal tergum, or pygidium, is a thickly sclero-
tized, vertical plate covering the anal and genital openings9; it is only par-
tially covered by the elytra. The covered portion, the base of the pygidium
(Figs. 72, 74-75, bpyg) is short and grooved medially. The permanently ex-
posed apex of the pygidium (apyg) is rounded and, in phanaeines, bears a
continuous, unbroken raised margin.

9 The suggestion arises that the pygidium comprises two terga, 8 and 9; however, a critical
study of the musculature of the pygidium and adjacent structures (which is quite intricate) is
in order before the possibility can be seriously considered.

776 The University Science Bulletin

the abdominal sterna

The abdominal venter is composed of eight sterna in the female, nine in
the male. The first two (Figs. 51, 75, Si + i;) are inflexed behind the meta-
coxae and, except for the dorso-lateral extremities of the second, permanently
hidden.10 Sterna 3-8 are ventrally exposed (Fig. 75, S.s-s). Although clearly
separated by impressed lines, they are rigidly fused. Sterna 1-3 are acutely
produced antero-medially between the apices of the metacoxae (Figs. 51, 75)
and form a prominence fitting against the metafurcasternum. The dorso-
lateral extremities of sterna 2-7 (= abdominal pleurites of authors) are im-
pressed and covered by the elytra. Longitudinal ridges (Fig. 75, rg) separate
these impressed areas from the ventral portions. The dorso-lateral ex-
tremities of sternum 8 are acute and do not extend beneath the elytra.

Sternum 9 (Fig. 78) is found only in the male. It is a thinly sclerotized,
internal plate (i.e., "cryptosternite") fitting beneath the apex of the genital
capsule. Though no such plate is present in the female, it may be repre-
sented by a pair of poorly sclerotized plates below the opening of the
common oviduct.

EXTERNAL GENITALIA

The male intromittant organ is enclosed in the genital capsule (Figs. 76-
77), which consists of an elongate tube, the phallobase (phb), bearing api-
cally a pair of lateral appendages, the parameres (pm). Within the phallo-
base is a well developed, membranous lobe, the aedeagus, which is presum-
ably extruded from the phallobase between the movable parameres during
copulation. Basally the aedeagus bears a large aedeagal apodeme (ap), on
which insert the retractor muscles of the aedeagus. Sclerotized external
genitalic structures exist only in the male of dung beetles. The terminology
applied here is that of Snodgrass (1957). It differs considerably from that
commonly applied in beetle taxonomy, in which, for example, the genital
capsule is often referred to as the aedeagus, and the present aedeagus as the
median lobe, phallus or penis. Lindroth (1957) has reviewed the various
terminologies applied to beetle genitalia.

PART III: COMPARATIVE SKELETAL MORPHOLOGY

OF PHANAEINES

The purpose of this part is to present a resume of the morphological
variation largely upon which is based the classification proposed in part V
and the evolutionary discussion in part VI. In addition, this section intro-

111 The second abdominal sternum, when exposed only dorso-lateral ly, characterizes the
haplogastran type beetle abdomen; Jeannel and Paulian (1944) stressed this character when
grouping Scarabaeoidea and Staphylinoidea as the suborder Haplogastra (but see Crowson,
1955).

Evolution of the Phanaeine Dung Beetles 777

duces some characters which, to my knowledge, have heretofore not been
discussed in systematic studies of Scarabaeinae.

The characters treated here are admittedly only a small portion of those
potentially available from an unlimited study of dung beetle skeletal mor-
phology. Practicality, however, demands adoption of some reasonable
criteria to limit them to a manageable number, which, unless computer
techniques are used, is apt to be quite small. The following criteria were
stressed in selecting characters discussed below:

(1) that they (the characters) be supraspecific characters; that is, that
their variation in general be less within a species group (or higher category)
than among several. This criterion, however, does not strictly limit charac-
ters to those with wide taxonomic application since some character states
may be constant only among the species of a single taxon while varying
widely among the species of other taxa of similar size or rank;

(2) that as nearly as possible they represent all tagmata and appendages;

(3) that they be studied in at least one representative species of each
recognized supraspecific taxon;

(4) that they not require extensive quantitative analysis. This restriction
was imposed primarily because of the lack of series of a sufficiently diverse
group of representative species large enough to support a reasonable degree
of quantitative rigor;

(5) that, except in cases of monobasic genera, they not be characters
unique to single species.

Even though most are easily observed, characters were not avoided be-
cause they require special effort to assess.

About 90% of all known species representing all phanaeine genera were
examined during the course of this study, including most of the rarer ones.
However, not all species were examined with the same thoroughness; the
appendix indicates the extent to which each was examined. Since a revision
of species is not the present problem, no special effort was made to obtain
large series (more than 20 individuals). Such large series were studied in
some instances only to aid in determining what characters tended to vary
greatly intraspecifically. Except in cases of extremely rare species (e.g.,
Homalotarsus impressus), the minimum number of specimens examined of
each was an individual of each sex; usually more specimens were available.

In general, the sequence of characters treated here is the same as in the
foregoing section. The name of some characters is preceded by a number
(1-22) to permit easy cross reference in following sections. In a few in-
stances (e.g., character 6) certain highly correlated features are treated
together even though they may be in quite distinct anatomical locations.
Numbered characters are "generic characters" in that they generally have
wide taxonomic application at the genus level; hence, they form the nucleus

778

The University Science Bulletin

Figs. 79-82. 79, Phanaeus vindex, female, frontal view of head; 80, Oxysternon curvispinum,
male, dorsal view of head; 81, Phanaeus haroldi, female, dorsal view of head; 82, P. melampus,
female, dorsal view of head.

of the generic descriptions given below. It should be said, however, that
description of genera is limited neither to states of these numbered characters
nor to the variation discussed in this section.

The states of many characters are designated by a, b, etc. In these in-
stances the discussion of each state of the character is introduced by a few
"key" words followed, if necessary, by a complete description of the state.
For brevity, these key words, which are not necessarily self-explanatory,
will be used in subsequent discussions (keys, descriptions, etc.). If a given
character state is reasonably regularly distributed at or above the subgenus
level, its definition is followed by the names of those taxa possessing it. For
this purpose, it is necessary to use names not formally introduced and de-
fined until later; to gain some taxonomic perspective, one can refer to the
classification outline presented at the beginning of Part V. Names of taxa
used in this section may be followed by (ex) or (inf). When (ex), ex-
ception (s), is noted, one or more species (but usually only a small per-
centage of all species) of the cited taxon possess another defined state; unless
they are pertinent to the discussion of the character, these exceptional species

Evolution of the Phanaeine Dung Beetles

779

94

Figs. 83-97. Anterior clypeal margins, dorsal views. 83, Copiophanaetis lancijer; 84, C.
pertyi; 85, C. saphirinus; 86, C. dardanus; 87, C. pinto; 88, C. jasius; 89, Dendropacmon
renatii; 90, D. viridis; 91, D. ganglbaueri; 92, D. amyntas; 93, D. bahianus? 94, D.
denticollis; 95, Tetramereia convexa; 96, Megatharsis buchjeyi; 97, Homalotarsus impresstts.

are not mentioned. When (inf), inferred, is noted, the character state in
question is one requiring dissection to observe, for which expendable
material of the cited taxon was not available.

Discussion of some characters is omitted from this part. Those concern-
ing color and size will be mentioned in the systematic section. For com-
parison of the genital capsules of males, which were not examined for this
study, the reader is referred to Olsoufieft's monograph (1924), particularly
to Plates III-VII which illustrate the genital capsules of a large number of
phanaeines.

The Head Capsule

Various aspects of the variation in structure of the head capsule provide
important taxonomic characters at supraspecific levels.

clypeal sculpturing: The entire dorsal surface of the clypeus of most
phanaeines is completely transversely ridged (Figs. 79, 82). In some species

780

The University Science Bulletin

Figs. 98-110. Anterior clypeal margins, dorsal views. 98, Diabroctis mimas; 99, D.
mirabilis; 100, Oxystemon conspicillatum; 101, 0. palaemon; 102, Sulcophanaeus imperator;
103, S. velutinus; 104, S. carnifex; 105, S. menelas; 106, Phanaeus splendidiilus; 107, P.
endymion; 108, P. palaeno; 109, P. demon; 110, P. amethystinns.

of Oxystemon (including all Mioxysternon, Fig. 80) and in isolated species
in other genera it is punctate posteriorly.

1. shape of the anterior margin of the clypeus.

a. emarginate medially — interrupted medially by three (rarely one,
Fig. 90) usually acute emarginations, middle one being deepest; emargina-
tions setting off two large, acute median teeth (Figs. 83-89, 91-92). Copro-
phanaens, Dendropaemon. In some species of Dendropaemon the clypeal
teeth and emarginations are rounded (Figs. 93-94). The emarginate clypeus
is usually coincident with the U-shaped clypeal process (see character 2a,
below).

Evolution of the Phanaeine Dung Beetles

781

111 Xf

113

B

114

Figs. 111-114. Clypeal processes, A — frontal view, B — lateral view, C — cross section. Ill,
Phanaeus mexicanus; 112, Oxysternon conspicillatum; 113, Sulcophanaeiis jaitntts; 114,
Coprophanaens spitzi (A — ventral view; B — frontal view).

b. not emarginate medially — not interrupted medially by acute emar-
ginations; medially rounded (Fig. 109) or distinctly bidentate (Figs. 95-108,
110). All other genera not mentioned in a. This state is coincident with
clypeal processes of several types (character 2, b-d) but never with a U-shaped
process.

Character 1 is the first of several (see section on phylogeny) indicating a
presumed phyletic separation of Coprophanaeus and Phanaeus (and allied
genera). At least in Coprophanaeus an emarginate clypeus, strengthened by
a unique U-shaped clypeal process, as well as strongly toothed front tibiae
(see character 18) are probably related to necrophagy (see part IV). The
unique clypeal (and tibial) dentition of Coprophanaeus is approached by
Phanaeus endymion and closely related species (Figs. 108, 256).

2. clypeal process:

a. U-shaped — curved ridge extending onto ventral surfaces of median
clypeal teeth (Fig. 114, u). Coprophanaeus, Dendropaemon. Although usu-
ally distinct, the process of Dendropaemon is only weakly produced.

b. transverse— straight, transverse ridge usually wider than high (Figs.
9-10, clypr; 111) but sometimes tooth-like (Fig. 113) or acute. Diabroctis,
Sulcophanaeiis, Phanaeus (ex), Mioxystemon (very weakly produced).

c. spiniform — laterally compressed, acute spine (Fig. 112). Some Oxy-
sternon, sen. str.

d. reduced — no distinct ridge, etc., present; process at most represented
by a weak gibbossity, raised area or poorly defined ridge. Tetramereia,
Homahtarsits, Megatharsis. In all three genera the ventral surface of the

782 The University Science Bulletin

clypeus is strongly swollen in the area occupied by the clypeal process in other
groups; this swelling is similar to that bearing the U-shaped process.

It should be noted that since the head functions as a digging instrument,
such features as the anterior margin of the clypeus and clypeal processes are
subject to wear which can alter their shape.

lateral clypeal carinae : These ridges are usually present although their
extent and distinctness is subject to variation which cannot meaningfully be
partitioned into states. These carinae are characteristically absent in only a
few supraspecific taxa {Mioxystemon, Tetramereia and Homalotarsus) as
well as isolated species in nearly all other groups.

The transverse clypeal carina is unique among phanaeines to Diabroctis
(Figs. 194, 197, tcrcly); a similar carina is characteristic of Onthophagus
(Onthophagini).

3. cephalic process of the male: The shape of the male cephalic process
varies greatly among species, in addition to intraspecific allometric variation.
The following three types are the most common; they do not represent
mutually exclusive groups since intermediate and unique types are not con-
sidered. Except when otherwise indicated, the following apply only to large,
well developed ("major") individuals; the type of process borne by smaller
("minor") individuals, while reduced in size, is usually recognizable.

a. lamellate — antero-posteriorly flattened process with or without two
or three distal projections of equal or unequal size and shape; usually pre-
senting flattened anterior surface (Figs. 186, 188, 192-193, 305). This type is
characteristic of Coprophanaeus, sen. str., some Dendropaemon, and Dia-
broctis mimas and cadmas.

b. carinate — a transverse carina, wider than high. Most Dendropae-
mon, Tetramereia, Homalotarsus (inf), Megatharsis (inf).

c. corniform — a long, usually gradually tapering "horn" erect or curv-
ing posteriorly over pronotum (Figs. 1, 181, 183, 189, 198, 203-204, 209-211,
216, 218, 221, 226). This is by far the most common type and only type
found in large males of the following taxa: Sidcophanaeus, Oxysternon, sen.
str., Phanaeus, Megaphanaeus, Metallophanaeus. In small males of Oxy-
sternon, sen. str., all Mioxystemon and two Phanaeus (palaeno and \irbyi)
the process is reduced to a pair of closely set, median tubercles (Fig. 80).

4. cephalic process of the female: In general, the female cephalic process
is subject to much less interspecific and intraspecific variation than is that of
the male.

a. corniform — a long, tapering "horn" (Figs. 182, 199) or large, horn-
like process (Fig. 201). This is an uncommon condition among phanaeines
and other dung beetles; in general, these females are of species the males of

Evolution of the Phanaeine Dung Beetles

783

Figs. 115-124. Head and prothoracic structure; A — lateral view of head region around eye,
B — same, dorsal view, C — left side of anterior portion of pronotal margin, dorsal view. 115,
Diabroctis mimas; 116, Coprophanaens jasius; 117, Sulcophanaeus jaunus; 118, Oxysternon
conspicillatum; 119, Phanaeus haroldi; 120, Coprophanaetis saphirinus; 121, Sulcophanaeus
chryseicoUis; 122, S. imperator; 123, Dendropaemon renatii; 124, D. jractipes.

784 The University Science Bulletin

which also have corniform processes; since the pronotal structure of the fe-
male is usually also masculine, the sexes tend to closely resemble each other
in general appearance. Corniform or corniform-like processes are character-
istic of Megaphanaeus and the faunas group of Sulcophanaeus.

b. carinate — a straight or anteriorly bowed, transverse carina, usually
trituberculate (Figs. 81, 185, 197, 207, 213). This is the usual type of process
in phanaeines and other dung beetles; it is characteristic of all phanaeine
taxa except those mentioned above. In many species of Phanaeus it is a
short thickened ridge which is more-or-less isolated between the eyes (Fig.
82) and, in a few species (e.g., vindex, Figs. 3, 79), distinctly raised.

paraocular areas of the parietals : These vary in many respects, among
them sculpturing (smooth to strongly rugose) and relief. In Dendropaemon
they are always strongly raised medially (above) as a large bump or thick
transverse ridge (Fig. 134) ; they are flat in Tetramereia, Homalotarsus and
Megatharsis. In other taxa they vary among species (convex, flat or con-
cave). In most phanaeines the paraocular areas are carinate lateral to the
upper portion of the eye; this carina is characteristically lacking in few multi-
specific taxa (e.g., Coprophanaeus, sen. sir.). These areas also vary in width,
the basis of the following character:

5. width of the paraocular areas:

a. narrow — width of area, along its posterior margin, less than three
times (usually about two times) width of adjacent dorsal portion of eye
(Figs. 115B-123B). All taxa except those mentioned in b.

b. wide — width at least three times (usually more) width of eye (Figs.
124B, 134-135). Dendropaemon, sen. str., (ex), Tetramereia, Homalotarsus,
Megatharsis.

6. postocular lobes of the parietals : The shape of the posterior portion
of the head capsule varies in a number of rather subtle ways, most of which
indicate no regular taxonomic pattern. The most easily described of these
variables is the angle at which the postocular lobes of the parietals are de-
pressed posterior to the dorsal portions of the eyes (as seen dorsally). Al-
though intermediates occur, the state of the members of most taxa is clearly
one of two extremes, each correlated with a distinct shape of the anterior
portion of the circumnotal ridge:

a. depressed transversely — lobes depressed straight behind eyes in plane
perpendicular to long axis of body (Figs. 115B-117B, 120B-124B) ; anterior
portion of circumnotal ridge not broken by emarginations receiving the
postocular lobes (same Figs., C). All genera except Phanaeus and Oxy-
sternon. In the dardanus group of Coprophanaeus, sen. str., the circumnotal
ridge is effaced behind the eyes but not distinctly emarginate. In all Copro-

Evolution of the Phanaeine Dung Beetles

785

¥ A

~l 130

f ^fu

131

Figs. 125-131. Head structure; A — inner view of cephalic brachium and adjacent structures
(stippling indicates surrounding intracephalic space), B — cross section through brachium,
C — outer view of head wall in region of mesal branch of fronto-clypeal sulcus (indicated by an
arrow). 125, Sulcophanaeiis f annus; 126, Coprophanaetis jasius; 127, Phanaeus demon (female);
128, Oxysternon conspicillatum; 129, O. palaemon; 130, Phanaeus palaeno; 131, P. chalcomelas.

phanaeus and Dendropaemon the postocular lobes are conspicuously flat-
tened behind (compare polb in Figs. 115A and 116A).

b. depressed obliquely — lobes depressed behind eyes in plane oblique to
long axis of body; depressions setting ofT prominent, posteriorly directed
portions of the lobes (Fig. 4) received by distinct emarginations (Fig. 35,
emarg) interrupting anterior position of circumnotal ridge. Phanaeus (ex).
Oxysternon, sen. str., characteristically shows a state intermediate to a. and
b. above (Fig. 118); the circumnotal margin, while usually weakened by
shallow emarginations, is not always distinctly broken; in Mioxysternon the
entire anterior portion of the circumnotal ridge is effaced.

canthal area: The shape and extent of this area varies more or less
continuously with most taxa possessing one of two extremes:

786 The University Science Bulletin

a. distinct — posterior margin of paraocular area (as seen dorsally)
approaching middle of eye, leaving, as seen laterally, a flattened or concave
canthal area (Figs. 115A, ca, 116A-118A, 121A-124A) between upper and
lower portions of eye. Most members of all taxa except Phanaeus and
Mioxystemon.

b. indistinct — posterior margin of paraocular area approaching poste-
rior angle of eye, not leaving a distinct canthal area (Figs. 1, 4). Mioxy-
stemon and Phanaeus.

Figure 120A shows an intermediate state.

7. SHAPE OF CEPHALIC BRACHIUM:

a. bipodal — presenting two distinct "arms" (Figs. 5, cbr; 128, 131);
mesal branch of fronto-clypeal sulcus Y- or T-shaped (Fig. 1, mbfcs).
Phanaeus, Oxystemon, sen. str., Mioxystemon (inf). In a few species of both
genera the bipodal nature is obscured by a presumed reduction in size of one
or both "arms" (Figs. 127A, 129A, BOA) and coincident shortening or loss
of the anterior extension of the sulcus (same Figs., C) ; in these cases only
cross sections of the respective brachia (same Figs., B) show that they are
bipodal.

b. monopodal — presenting a single, usually straight and vertical "arm"
(Fig. 126A); mesal branch of fronto-clypeal sulcus not forked, usually
straight (Fig. 126C). All genera except Phanaeus and Oxystemon; inferred
for Homalotarsus. The brachia of Sulcophanaeus faunus (Fig. 125A) and
S. carnijex are apparently unique monopodal types which appear super-
ficially as bipodal (a few Sulcophanaeus do have bipodal brachia). As indi-
cated by the course of the mesal branch of the fronto-clypeal sulcus (Fig.
125C) and cross section of the brachium (Fig. 125B), the brachium of these
species is a single, anteriorly bent vertical plate.

Since the shape (forked or not) of the mesal branch of the fronto-clypeal
sulcus can be considered a projection of a cross section of the brachium, type
of brachium can usually be determined accurately without internal exami-
nation. The sulcus is regularly obscured only in the auricollis group of
Sulcophanaeus.

8. size of eyes:

a. small — dorsal interocular width more than four times width of dor-
sal portion of an eye (Figs. 4, 134-135) ; width of lower portion of an eye less
than twice (usually about equal) width of adjacent oculo-gular space (Fig.
9). With few rare exceptions, all taxa except Megaphanaeus, Coprophanaeus,
sen. str., and Coprophanaeoides.

b. large — dorsal interocular width less than four times width of dorsal
portion of an eye (Fig. 132) ; width of lower portion of an eye at least twice

Evolution of the Phanaeine Dung Beetles

787

*rW- — ^fy-

133

Figs. 132-135. 132, Dendropaemon renatii, dorsal view of anterior portion of head; 133,
same, ventral view of posterior portion of head; 134, D. viridis, dorsal view of anterior portion
of head; 135, Megatharsis buckjeyi, dorsal view of anterior portion of head.

(usually more) width of adjacent oculo-gular space (Fig. 133). Mega-
phanaens, Coprophanaens, sen. str. and Coprophanaeoides.

9. occipital ridge:

a. complete — usually setose ridge extending dorsally over posterior po-
sition of head (Fig. 1, 4, 8, ocr) ; small medial section occasionally obscured
by punctures. Oxysternon, Phanaeits.

b. incomplete — ridge absent or, at most, represented by usually weak
lateral segments, never distinct dorsally. All genera except Oxysternon and
Phanaetts.

In all phanaeines except Megaphanaeus the occipital areas are flattened or
weakly concave; in this subgenus, however, the areas are abruptly raised
postero-laterally as prominent bumps, anterior to which they are strongly
concave.

Appendages of the Head

antennae: The antennae are subject to some subtle variation which is
exceedingly difficult to render as distinct character states without the use of
statistical procedures (e.g., relative lengths of antennomeres, etc.) ; very little
antennal variation is striking enough to assess otherwise. Noteworthy among
this latter variation is the tendency in Dendropaemon for elongation of first

788

The University Science Bulletin

137

138

139

Figs. 136-139. 136, Dendropaemon jractipes, flagellum of right antenna, dorsal view; 137,
D. denticollis, A — apical view of left antennal club, B — apical five segments of left antenna,
ventral view; 138, Homalotarsus impressus, apical five segments of right antenna, dorsal view;
139, Megatliarsis buchleyt, flagellum of right antenna, dorsal view.

lamella (Fig. 136) and dorse-ventral flattening of the entire club (Fig. 137)
and the near hemispherical club of Megatliarsis (Fig. 139) and Homalotarsus
(Fig. 138).

labrum: Variation in the labrum offers a number of important taxo-
nomic characters, both unique to few taxa and with wide application.
Among the unique features are two characteristic only of Dendropaemon:
(1) apical margin deeply emarginate (Fig. 145) (that of other phanaeines
may be weakly emarginate), and (2) antero-median process of suspensorium
forked anteriorly, within fork insert setae of median brush (Fig. 145). Labral
characters of potential significance which were not studied include various
aspects of chaetotaxy (e.g., number of setae in lateral files) ; study of these
was not approached since their efficient interpretation would have required
statistical methods (see introduction to this part).

One of the principal aspects of labral variation is the definition of two
principal labral types based on the correlation of three characters: lateral
files, type of median brush and type of labral suspensorium.

10. LABRAL TYPE:

a. Coprophanaeus-type — lateral files of labrum parallel or nearly so
(divergent in Coprophanaeits), distance between anterior ends less than 1.25
(usually about 1.0) times distance between posterior ends (Figs. 140, 142,
144) ; central setae of median brush much longer than setae posterior to
them (Fig. 148) or median brush comprising only few, very long setae (Fig.
144) ; length of opening in suspensorium 0.5 times or less length of dorsal
surface of labrum, opening usually very transverse, formal arms narrow and
often converging posteriorly (Figs. 141, 143, 145-146). Coprophanaeits,
Homalotarsus (inf), Tetramereia, Megatliarsis, Dendropaemon. These

Evolution of the Phanaeine Dung Beetles

789

140

142

% • «y v$

■

144

141

143

145

146

Figs. 140-148. Labra. 140, Megatharsis bachjeyi, ventral view; 141, same, dorsal view;
142, Tetramereia convexa, ventral view; 143, same, dorsal view; 144, Dendropaemon viridis,
ventral view; 145, same, dorsal view; 146, Coprophanaeus jasius, dorsal view; 147, same, ventral
view; 148, same, lateral view of median brush of labrum.

genera, in the foregoing order, show progressively a reduction in the number
of setae comprising the median brush, from a large number in Copro-
phanaeus to a few in Dendropaemon; the labral suspensorium of the last four
genera is weakly sclerotized compared to that of other phanaeines.

b. Phanaeus-type. — lateral files divergent, distance between anterior
ends more than 1.25 (usually about 1.5) times distance between posterior
ends (Figs. 19A, 147) ; central setae of median brush not much longer than
setae posterior to them (Fig. 18), brush never reduced to a few setae, long or

790

The University Science Bulletin

Figs. 149-158. 149, Sttlcophanaeus batesi, dorsal view of labium (floor of buccal cavity not
shown); 150, same, lateral view (slightly smaller scale than Fig. 149; only apical portion
shown in detail); 151, Dendropaemon viridis, lateral view of labium (only apical portion
shown in detail); 152, same, ventral view of mentum and palpi; 153, Tetratnereia convexa,
ventral view of mentum and palpi; 154, Megatharsis buckjeyi, ventral view of mentum and
palpi; 155, same, dorsal view of right palpus; 156, Homalotarsits impressus, ventral view of
mentum and palpi; 157, Coprophanaens jasitis, ventral view of mentum and palpi; 158,
Sulcophanacus jaunus, ventral view of mentum.

Evolution of the Phanaeine Dung Beetles 791

not; length of opening in suspensorium more than 0.5 (usually nearly 1.0)
times length of dorsal surface of labrum, opening rounded, not distinctly
transverse, tormal arms thick (Fig. 19A). Diabroctis, Sulcophanaetis,
Phanaeus, Oxysternon.

mandibles and maxillae: These mouthparts are remarkably similar in
all taxa studied; they yielded no taxonomic characters useful at the supra-
specific level.

labium: Unique variation of the labium is restricted to Megatharsis
buc^leyi, which, unlike any other phanaeine, has one-segmented labial palpi
(Fig. 154) ; the dorsal surface of each segment bears a membranous slit (Fig.
155, arrow) possibly indicating an incomplete line of fusion. Other labial
characters, with wider taxonomic application, include the following three:

11. SHAPE OF MENTUM:

a. weakly emarginate — anterior emargination shallow or scarcely dis-
tinct, its depth less than l/4 (with few exceptions l/8 or less) greatest length
of mentum; greatest width of mentum usually less than 1.5 (approaching 1.0
in Dendropaemon) times greatest length (Figs. 152-154, 156-157). Copro-
phanaeus, Dendropaemon, Tetramereia, Homalotarsus, Megatharsis.

b. strongly emarginate — anterior emargination deeper, depth at least
!/4 (usually Yi or more) length of mentum; width of mentum almost always
1.5 or more times length (Figs. 25, 158). Diabroctis, Sidcophanaeus,
Phanaeus, Oxysternon.

12. MEDIAN LOBE OF HYPOPHARYNX :

a. present — large, distinct lobe at summit of premental groove (Fig.
26, mlhpx). All genera except those mentioned below.

b. absent — no distinct lobe present. Dendropaemon, Tetramereia,
Homalotarsus (inf), Megatharsis.

13. PREMENTAL SCLERITES :

a. completely sclerotized — entire sclerite darkly sclerotized (Fig. 30,
pmts). Phanaeus, Oxysternon.

b. marginally sclerotized — only a C-shaped portion of posterior mar-
gin darkly sclerotized, otherwise weakly sclerotized (Figs. 150-151, pmts).
All genera except Phanaeus and Oxysternon.

There exists another labial character — of great interest but little taxo-
nomic value at the supraspecific level — concerning the position of the glossae,
which varies between the following extremes:

a. appressed. — glossae closely appressed along mesal margins for at
least basal half of their lengths, completely closing over median groove of

792

The University Science Bulletin

Figs. 159-162. 159, Phanaeus pyrois, male, left side of pronotum; 160, Oxystanon conspicilla-
tum, male, same; 161, Phanaeus haroldi, female, same; 162, Dendropaemon viridis, caudal
view of pygidium.

prementum and enclosing between their bases median lobe of hypopharynx
(if present) (Fig. 26).

b. separated — glossae widely separated basally, not touching mesally
at any point; median groove of prementum exposed, median lobe of hypo-
pharynx (if present) free (Fig. 149).

Species showing the above extremes and intermediate positions are in-
cluded in each of the multispecific genera except Oxysternon, all examined
species of which have appressed glossae. The following, for each genus, lists
three common species showing, respectively, appressed, intermediate and
separated glossae :

Evolution of the Phanaeine Dung Beetles

793

Figs. 163-166. 163, Phanaens damocles, female, antero-lateral portion of pronotum; 164,
P. triangularis, female, same; 165, P. vindex, female, postero-median portion of pronotum;
166, Coprophanaeus jasius, female, left antero-lateral portion of pronotum.

Phanaens: vindex, palaeno, qitadridens
Sidcophanaens: f annus, imperator, batesi
Coprophanaeus: ensifer, jasius, milon
Dendropaemon: viridipennis, denticollis, renatii
The glossae of Megatharsis are appressed; those of Tetramereia and
Homalotarsus (inf ) are of intermediate position.

Thoracic Segments
The shape of the pronotum of most phanaeines differs markedly between
the sexes and among species. Like those of the head, pronotal outgrowths
usually vary allometrically within single local populations. The form of the

794

The University Science Bulletin

Figs. 167-180. 167, Dendropaemon amyntas, male, lateral view of pronotum; 168, D.
renatii, female, same; 169, same, dorsal view of pronotum; 170, D. riridis, male, lateral view
of pronotum; 171, same, dorsal view of pronotum; 172, Tetramereia convexa, lateral view of
pronotum; 173, Dendropaemon jractipes, same; 174, D. denticoUis, same; 175, D. haroldi, same;
176, D. nitidicollis , same; 177, same, dorsal view of pronotum; 178, Megatharsis buc\leyi, lateral
view of pronotum; 179, D. haroldi, ventral view of left side of prothorax and base of leg; 180,
D. viridipennis ,, same.

Evolution of the Phanaeine Dung Beetles

795

Figs. 181-189. 181, Coprophanaetts lancifer, male, lateral view of head and pronotum; 182,
same, female; 183, C. saphirinus, male, same; 184, same, female, lateral view of pronotum; 185,
C. pertyi, female, dorsal view of head and anterior portion of prothorax; 186, C. pluto, male,
lateral view of head and pronotum; 187, same, dorsal view of median pronotal prominence;
188, same, anterior view of cephalic process; 189, C. pertyi, male, lateral view of head and
pronotum.

male pronotum varies greatest among species and is often distinctive of a
species. Female pronotal form varies less among closely related species;
indeed, the females of some groups of species (e.g., Phanaeus daphnis and
related species) are impossible to distinguish using secondary sexual charac-
ters of the head and pronotum, even though the males may be quite distinct.
It is difficult to arrange pronotal variation in shape into reasonably distinct

796

The University Science Bulletin

Figs. 190-197. 190, Coprophanaeus jasius. female, dorsal view of anterior portion of prono-
tum; 191, same, lateral view of pronotum; 192, same, male, lateral view of head and pronotum;
193, same, male, anterior view of cephalic process; 194, Diabroctis tnimas, male, dorsal view
of head and pronotum; 195, same, lateral view; 196, same, female, lateral view of pronotum;
197, same, female, dorsal view of head.

Evolution of the Phanaeine Dung Beetles

797

Figs. 198-203. 198, Sulcophanaeus faunus, male, lateral view of head and pronotum; 199,
same, female: 200, same, female, dorsal view (incomplete) of pronotum; 201, S. carnifex,
female, lateral view of head and pronotum: 202, same, female, anterior view of cephalic
process; 203, same, male, lateral view of head and pronotum.

character states. Obvious types do exist, although their taxonomic reliability
is reduced by a fairly high incidence of presumed convergence, particularly
among males. In general, however, closely related species have similar
pronotal shape.

Characters 14 and 15 treat pronotal types based on the shape and position
of various projections, concavities, etc. Character 14 is based on well de-

798

The University Science Bulletin

Figs. 204-209. 204, Sidcophanaeits imperator, male, lateral view of head and pronotum;
205, same, female; 206, S. velutinus, male, lateral view of pronotum; 207, S. chrysekollis,
female, dorsal view of head and anterior portion of pronotum; 208, same, female, lateral view
of pronotum; 209, same, male, lateral view of head and pronotum.

veloped ("major") individuals; less well developed males, while most often
presenting reduced versions of the same features, may appear quite distinct.
Like cephalic structure, variation in pronotal shape among conspecific fe-
males is much less pronounced than among males; thus, the types defined
in character 15 should apply equally well to large and small individuals. The
following do not include unique types; furthermore, since the variation
within types is sometimes quite wide, the defined types should not be
considered strictly mutually exclusive.

14. SHAPE OF PRONOTUM OF MALE:

a. type A — postero-lateral angles produced laterally as rounded lobes,
giving pronotum a decidedly triangular or trapezoidal shape as seen dorsally;
usually flattened dorsally, though sometimes convex or concave and bearing
one or more tubercles or stronger prominences on dorsal surface (Figs. 35-36,
218-219,223-226,228).

b. type B — bearing two latero-median processes separated by con-
cavity (Figs. 210-211, 216-217).

c. type C — strongly and widely raised antero-medially, bearing trans-

Evolution of the Phanaeine Dung Beetles

799

Figs. 210-215. 210, Oxystemon conspicillatum, male, dorsal view of head and pronotum;
211, same, lateral view; 212, same, female, lateral view of pronotum; 213, same, dorsal view
of head and anterior portion of pronotum; 214, O. curvispinum , male, dorsal view of head and
anterior portion of pronotum (separated); 215, same, lateral view of pronotum.

verse angulate prominence or carina, anterior to which pronotal surface is
vertical or concave (Figs. 167, 186, 189, 192, 194-195).

d. type D — bearing four tubercles or prominences in transverse row-
near posterior margin of pronotum; tubercles may differ little in size or
inner two or outer two may be strong projections (Figs. 198, 203-204).

e. type E — pronotum deeply and widely concave medially, posterior
margin of concavity bearing massive, saddle-shaped prominence (Fig. 181).

f. type F — flattened and sloping anteriorly with pair of closely set
postero-median prominences which may be crest-like or elongate projections
(Figs. 183, 206, 209).

800

The University Science Bulletin

Figs. 216-222. 216, Phanaeus dejeani, male, lateral view ol head and pronotum; 217, same,
dorsal view; 218, /'. endymion, male, lateral view of head and pronotum; 219, same, male,
dorsal view of pronotum; 220, same, female, lateral view of head and pronotum; 221, P.
haroldi, male, same; 222, same, female, dorsal view of head and anterior portion of pronotum.

15. SHAPE OF PRONOTUM OF FEMALE:

a. type A — convex except for antero-median concavity bordered an-
teriorly by transverse carina or crest (Figs. 184-185, 190-191).

b. type B — convex except for three closely set antero-median tubercles,
median one of which may be dentate (Fig. 220).

c. type C — convex except for low, antero-median prominence bordered
anteriorly by unevenly bowed carina (Figs. 207-20S).

d. type D — convex but bearing an antero-median transverse carina
or wide ridge anterior to which pronotal surface sometimes nearly vertical
(Figs. 37-38, 196, 205, 213, 227).

e. type E — evenly convex (Figs. 178,310-315).

Evolution of the Phanaeine Dung Beetles

801

Figs. 223-228. 223, Phanaeus lugens, male, lateral view of pronotum; 224, same, dorsal
view; 225, P. demon, male, lateral view of pronotum: 226, P. mexicanus, male, lateral view of
head and pronotum; 227, same, female, lateral view of pronotum: 228, P. daphnis, male, lateral
view of pronotum.

16. proxotal sculpturing: The texture of the pronotal surface varies con-
siderably. Although intermediate conditions occur in some species most
supraspecific taxa are characterized by one of the following states:

a. smooth-punctate — completely smooth or with limited weakly
roughened areas, but entire surface punctate (Figs. 159-161).

b. roughened — surface entirely granulate (Fig. 164), granulo-rugose
(Fig. 163) or rugose (Fig. 165); if distinct puncturing present, usually re-
stricted to postero-median area. In most Coprophanaeus and Coprophanae-
otdes the roughening, at least anteriorly, is in the form of transverse ridging
(Fig. 166).

The following are other taxonomically restricted variations of the
prothorax :

posterior proxotal fossae: These fossae are usually evident as puntiform
pits or large rounded depressions on the postero-medial angle of the pro-
notum (Fig. 194, pfos) ; they are characteristically absent only in Oxysternon,
the postero-median angle of the pronotum of which is a strongly produced,
acute process fitting between the bases of the elytra (Fig. 210).

802

The University Science Bulletin

Figs. 229-245. 229, Dendropaemon denticollis, dorsal view of mesoscutellum; 230, same,
lateral view of antero-median angle of metasternum; 231, same, ventral view of pterothorax;
232, same, lateral view of metepisternum; 233, D. haroldi, same as 230; 234, same, same as 231;
235, same, same as 232; 236, D. nitidicollis, same as 230; 237, same, same as 231; 238, same,
same as 232; 239, Homalotarsits impressus, dorsal view of mesoscutellum; 240, same, ventral
view of mesosternum and median area of metasternum; 241, Dendropaemon renatii, lateral
view of metepisternum; 242, D. bahianus?, same; 243, D. hirticollis, same; 244, Oxysternon
conspicillatum , lateral view of pterothorax (elytra in place, coxae removed); 245, same, ventral
view of mesosternum and median area of metasternum.

Evolution of the Phanaeine Dung Beetles

803

Figs. 246-259. Front tibiae. (All Figs, except 247B are of anterior surfaces.) 246,
Coprophanaetts jasins; 247, Dendropaemon viridis (A — anterior surface, B — posterior surface) ;
248, Megatharsis buchj.eyi; 249, Tetramereia convexa; 250, Diabroctis mimas (A — male, B —
female); 251, Sulcophanaeus noctis, male; 252, S. columbi, female; 253, S. imperator, male;
254, Homalotarsus impressus; 255, Phanaeus haroldi, female; 256, P. blanchardi, female; 257,
P. \irbyi, male; 258, Oxysternon festivum, male; 259, Phanaeus mexicantts (A — female, B —
male).

804

The University Science Bulletin

Figs. 260-262. 260, Homalotarsus impressus, female, dorsal view of apex of femur, tibia
and tarsus of right hind leg; 261, Phanaeus vindex, male, ventro-lateral view of left hind tibia
and tarsus; 262, Sulcophanaeus j annus, male, same.

lateral pronotal carinae : These carinae (Figs. 199, 201, lpc) are dis-
tinct only in the Sulcophanaeus jaunus group. In other Sulcophanaeus, in
Megaphanaeus and Diabroctis (particularly males) the pronotum is thick-
ened along a non-carinate line in the same relative position as this carina.
The lateral pronotal carinae of S. columbi are produced postero-laterally as
small crests. These angles produce a pronotal shape similar to type A (char-

Evolution of the Phanaeine Dung Beetles

805

263

265

Figs. 263-269. 263, Tetramereia convexa, ventral view of left hind tibia and tarsus; 264,
Homalotarsus impressus, dorsal view of right hind tarsus; 265, same, ventral view of first
segment of right mesotarsus; 266, Megatharsis buc\leyi, dorsal view of left hind tibia and
tarsus; 267, same, dorsal view of middle left tibia and tarsus; 268, Dendropaemon viridis, dorsal
view of right hind leg (except tarsus) ; 269, same, ventral view.

acter 14), suggesting that the rounded postero-lateral lobes characteristic of
type A are homologs of lateral pronotal carinae.

transverse propleural carinae: These carinae are present only in Den-
dropaemon, sen. str. When complete, each extends along the posterior mar-
gin of the procoxal cavity onto the propleural surface lateral to the coxa (Fig.
180, arrows); when incomplete, each is restricted along the posterior margin
of the coxal cavity (Fig. 179, arrow).

Variation in the pterothoracic segments of taxonomic significance is lim-
ited to the spinate prolongation of the metasternum of Oxysternon (Figs.

806

The University Science Bulletin

Figs. 270-279. 270, Dendropaemon denticollis, ventral view of right hind tibia and tarsus;
271, D. haroldi, ventral view of right hind tarsus; 272, D. smaragdinum , same; 273, D.
lobatum?, same; 274, D. rcnat'u, same; 275, D. amyntas, same (worn); 276, D. tiridipennis ,
same; 277, D. viridis, same; 278, D. ganglbaueri, same (arrow points to small second segment);
279, same, ventral view of left middle tarsus.

244-245) and to various aspects of body shape variation in Dendropaemon
(Figs. 229-238, 241-243), which are discussed more fully in the systematic
section.

Legs
17. front tarsi of female:

a. present — short, five-segmented tarsi inserted on inner apical angle
of tibia. Diabroctis, Sulcophanaeus (ex), Oxysternon, Phanaeus, Mega-
phanaeus.

b. absent — Coprophanaeus, sen. str., Metallophanaeus, Dendropae-
mon, Tetramereia, Homalotarsus, Megatharsis.

Evolution of the Phanaeine Dung Beetles

807

Figs. 280-283. Left elytron, dorsal views. 280, Oxysternon (M.) curvispinum; 281, O.
conspicillatum (arrows point to second stria); 282, Sulcophanaeus noctis; 283, S. columbi
(arrow points to third stria).

18. FRONT tibiae:

a. Coprophanaeus-type — very strongly and acutely quadridentate,
teeth separated basally by very narrow spaces; apical three teeth always
carinate (Fig. 246). Coprophanaeus.

b. Dendropaemon-type — acutely quadridentate, teeth not separated

808

The University Science Bulletin

Figs. 284-287. Lett elytron, dorsal views. 284, Phanaeus vindex; 285, P. mexicanus; 286,
/'. damocles; 287, /'. endymion.

by narrow spaces basally; apical three teeth always carinate; inner basal angle
expanded, covering apex of femur (Fig. 247). Dendropaemon (except
ganglbaueri).

c. Phanaeus-type — tridentate or quadridentate, teeth acute or rounded,
separated basally by wide, usually rounded spaces; apical two or three teeth

Evolution of the Phanaeine Dung Beetles

809

Figs. 288-291. Left elytron, dorsal views. 288, Coprophanaeus lancijer; 289, C. jasius;
290, Dendropaemon vindis; 291, same (higher magnification to show carinulate striae).

carinate (Figs. 248-259) ; tibiae of male sometimes distinctly narrower than
those of female (Figs. 250, 259). All genera except above two.

19. MIDDLE AND HIND TARSI :

a. two-segmented — some Dendropaemon, sen. sir., Paradendropae-
mon (Figs. 272, 277-279).

810

The University Science Bulletin

Figs. 292-299. (Stippling indicates areas for which detail not shown.) 292, Coprophanaetis
jasius, right hind wing; 293, Sulcophanaeus chryseicollis , base of left hind wing; 294,
Dendropaemon renatii, same; 295, D. piceum, same; 296, Megatharsis buckleyi, same; 297,
Tetramereia convexa, same; 298, Dendropaemon haroldi, same; 299, D. amyntas, same.

b. three-segmented — some Dendropaemon, sen. str., Coprophanae-
oides (Figs. 270-271, 273-276).

c. four-segmented — Tetramereia (Fig. 263).

d. five-segmented — all other genera; relative shapes and sizes of seg-
ments usually as in vindex (Fig. 64) ; shapes of tarsal segments of Mega-
tharsis (Figs. 266-267) and Homalotarsus (Figs. 264-265) unique.

Evolution of the Phanaeine Dung Beetles

811

300

spr s

Figs. 300-303. 300, Dendropaemon ganglbaueri, female, ventral view of abdomen; 301,
Megatharsis buc\leyi, female, caudal view of abdomen (setae shown on left side only); 302,
Sulcophanaeits carnifex, lateral view of right dorso-lateral extremity of abdominal sternum 7
and adjacent areas; 303, Coprophanaetts dardanus, same.

transverse tibial carina: In addition to a few isolated species (e.g.,
Sulcophanaeits imperator), this carina (Fig. 262, cr) is restricted to the Sulco-
phanaens faunus group, Diabroctis, and the Oxysternon festivum group.

Other variation of the legs includes the following, relatively isolated spe-
cializations: The setigerous granulation of the dorsal surface of the hind legs
of Homalotarsus (Fig. 260) and the characteristic expansion of the lower
apical angle of the middle and hind femora of Dendropaemon (Figs. 268-
269).

Wings
20. sculpturing of elytral interstriae :

a. smooth — no distinctly punctured or rugose areas; if punctures pres-
ent, restricted to minute punctures clearly visible only with high (20X)
magnification (Figs. 281-282, 285, 2S7, 289).

b. punctate — distinctly punctured, punctures visible with low (10X)
or no magnification (Figs. 280, 286, 290).

c. rugose — granulate, rugose or very heavily punctured with co-
alescing of punctures (Figs. 284, 288).

812 The University Science Bulletin

21. relief of elytral interstriae :

a. flat— (Figs. 282, 284, 286).

b. convex— usually evenly (Figs. 280-281, 285, 287, 290-291) but some-
times raised only medio-longitudinally such that striae appear to lie in broad,
shallow troughs (Fig. 283).

22. striae:

a. carinulate — bordered laterally by fine carinae which may be straight
(Figs. 290-291) or undulate (Fig. 288). Megaphanaeus, Dendropaemon (ex-
cept ganglbaueri) and isolated instances in other genera.

b. simple — distinct or very weak, sometimes punctate lines with un-
defined lateral margins (Figs. 280-287, 289). All taxa except above.

Of the characters of the hind wing only one is reasonably consistent at
the supraspecific level: a distinct notching of the wing between the vannal
and jugal lobes (Figs. 292, 294-295, 298) characteristic of Coprophanaeus,
sen. str., Coprophanaeoides and some Dendropaemon, sen. str.; all others
(except Phanaeus bispinus) have no emargination (Figs. 71, 293, 296, 299).

Abdomen
Only the following features comprise marked departure from the other-
wise striking consistency in abdominal structure among the phanaeines
studied :

position of last (7th) abdominal spiracle: The opening of this spiracle
assumes three positions.

a. marginal — spiracular opening located in membrane connecting
tergum 7 and sternum 7 (Fig. 72, spr?); usually visible only dorsally; this is
the most common position.

b. intermediate — located in a notch in dorsal margin of sternum 7;
usually visible laterally (Fig. 303).

c. intrasternal — located within dorso-lateral portion of sternum 7 (Fig.
302, spr), faint suture (s) extending from spiracle to dorsal margin of
sternum 7; visible only dorsally.

Supposedly the evolution of position of this spiracle has included its ven-
tral migration and/or dorsal extension of sternum 7, ultimately becoming
isolated by closure of the sternum along a distinct suture. The intrasternal
position is the least frequent and characteristic only of Megaphanaeus and
Mioxysternon as well as isolated species-pairs in Sidcophanaeus. States a. and
b. occur in a random taxonomic pattern.

The only other noteworthy abdominal variations are the lateral depres-
sions of the base of the pygidium of Dendropaemon (Fig. 162, Id), the
sternal fringes of Megatharsis (Fig. 301) and the abnormally elongate
sternum 7 of Paradendropaemon (Fig. 300, s?).

Evolution of the Phanaeine Dung Beetles 813

PART IV: ADAPTIVE SIGNIFICANCE OF ASPECTS OF
PHANAEINE SKELETAL MORPHOLOGY

What is known about the habits of phanaeines, as well as other Scara-
baeinae, has been well summarized by Halffter and Matthews (1966). One
important group of adaptive modifications of dung beetles, those related to
coprophagy, has been discussed by them (pp. 241-248). As they and others
have pointed out, adult scarabaeines are basically fossorial animals. Al-
though they seek food on the surface, a large part of the feeding, mating
and reproductive activities is carried on in underground tunnels and cham-
bers excavated by the beetles. Efficient fossorial habit requires a highly
specialized morphology. The purpose of this discussion is to bring out
certain aspects of adult scarabaeine morphology, as illustrated specifically
by the phanaeines, presumably related to digging habits. Certain characters
here deemed fossorial are no doubt also important in other behavioral con-
texts; indeed, a portion of this discussion is the examination of just such a
case. Nevertheless, it is assumed that the initial selective factors involved in
their evolution from antecedent types favored morphological complements
of fossorial behavior.

The fossorial features of phanaeines, and dung beetles in general, fall
into two general categories: (1) structural modifications of direct use as
digging tools, and (2) strengthening devices for increased body resistance
to the generally compressive forces exerted on the body during digging and
manipulation of soil (and food). Most of the features to be mentioned are
shared with other Scarabaeinae and some other laparostict scarabaeids.

Structural Tools
The primary earth-moving structures are the head and front legs. The
following modifications of the head capsule and legs of phanaeines pre-
sumably contribute to the efficiency of these organs as digging tools:

a. Differential expansion and ventral folding of the head sclerites — This
trend has two very important adaptive consequences, (1) increase in the
digging surface area of the head, principally through dorsal expansion of
the clypeus, and (2) elimination of dorsal and lateral exposure of the head
appendages, which are thereby protected from potential injury.

b. Clypeal dentition and transverse ridging — While the dorsum of the
head (clypeus) is primarily employed in pushing and packing soil (and
food), its anterior margin functions primarily as a scraper. Clypeal teeth
and other irregularities produce sharper, thus more effective cutting instru-
ments. In addition to teeth, the transverse ridging of the dorsal surface
contributes to its overall abrasiveness.

c. Antero-lateral flattening and lateral toothing of the front tibiae — The
front tibiae are principally scraping structures whose effectiveness is in-
creased by dentate cutting edges.

814 The University Science Bulletin

d. Elongation, double articulation and transverse orientation of the front
coxae — Elongation has effectively increased the surface area of the coxal
margin available for muscle insertion. If increased musculation is to func-
tion efficiently to rotate the coxa, the axis of rotation of a subcylindrical coxa
must be fixed by two articulations, one at each end. Transverse orientation
of the coxa preserves an antero-posterior movement of the distal portion of
the leg effective both for walking and digging movements (dirt is pushed
under body).

e. Strong development and musculation of middle and hind legs — These
legs, while not directly involved in digging, propel the body forward and,
to a large extent, determine the effectiveness of digging movements of the
head; they are strongly musculated. The tarsi are articulated so that, during
propulsion of the body by middle and hind legs, they project laterally to
allow firm implantation of the expanded distal ends of the tibiae.

Strengthening Devices
To effectively aid digging movements, the body wall must provide rigid
mechanical support. The following features (in addition to thickly sclero-
tized integument) presumably contribute to the overall body resistance
necessary to withstand the stresses imposed by digging:

a. Cephalic brachium — While the brachium no doubt imparts necessary
rigidity to the dorsal mandibular articulation, it also increases the resistance
of the head capsule to compressive forces exerted on the expanded anterior
portion of the head dorsum.

b. Complete division of the eye — The firm contact (along the postocular
line) of the paraocular and postocular lobes of the parietals apparently re-
lieves the eye from pressure transferred posteriorly from the head dorsum
and from antagonistic pressure transferred anteriorly by resistance of the
prothorax.

c. Intersegmental inflection of the body wall — Longitudinal compressive
forces are borne by opposed inflected surfaces which distribute resistant
forces over much greater areas than would be afforded by the mere opposing
edges of sclerites. The most extensive inflections occur between the head
and prothorax and between the prothorax and pterothorax. The narrowed
posterior portion of the head capsule fits snugly against the prothorax in a
receptacle formed by the inflection of the anterior prothoracic margin.
Maintenance of this head-prothorax connection is enhanced in Phanaeus by
the interlock mechanism formed by the postocular lobes of the parietals and
emarginations in the anterior portion of the circumnotal ridge. The inflected
posterior margin of the prothorax is opposed dorsally by anterior inflections
of the elytra and mesepimera and ventrally by the antero-median promi-
nence of the metasternum.

d. Rigidity of posterior part of body — Overall strengthening is achieved

Evolution of the Phanaeine Dung Beetles 815

by two complementary devices, firm interlocking and anchoring of the elytra
and fusion of the abdominal sterna. Elytral interlocking is by means of an
opposing ridge-groove mechanism along the elytral suture; anchoring of the
elytra is affected by the metepisternal tabs, which maintain close elytral
appression along the elytral suture by restricting lateral slippage of the elytra.
Concealment of all spiracles is apparently another important fossorial
adaption. The thoracic spiracles are hidden by the overlap of the mesothorax
by the prothorax; the abdominal spiracles are covered by the elytra. Directly
exposed spiracles would no doubt be highly susceptible to obstruction by
soil particles.

Secondary Modification of Fossorial Features
As mentioned above, certain basically fossorial characters have equally
important adaptive significance in other behavioral contexts, such as feeding
habit. Coprophanaeits is apparently exclusively a necrophagous genus while
most other phanaeine genera are known to be principally coprophagous. In
the context of necrophagy, the strong development of clypeal and front tibial
dentition in Coprophanaeits is especially noteworthy. While dentition in the
present sense is basically a fossorial adaptation, its degree of development in
phanaeines is closely related to feeding habit. Strongly developed dentition
in Coprophanaeits is probably directly related to the principal food of the
genus, carrion. The resiliency of flesh is obviously much greater than that of
dung. Hence, structural tools facilitating its collection would presumably
be selectively advantageous for a necrophagous beetle. I have observed
Megaphanaeits in Brazil using the foretibiae, employed with unmistakable
cutting-slicing motions, and clypeus to shear flesh from animal carcasses, in
the manner suggested by HalfTter and Matthews (1966: 93). That Copro-
phanaeits is efficient at removing carcasses is well evidenced by the obser-
vation of Martinez (1959) that four individuals of Coprophanaeits (Mega-
phanaeits) ensifer cut into pieces and buried a 20 kg (44 lb) dog carcass in
one night! I observed a somewhat lesser feat performed by five C. (C.)
jasius, which disposed of an armadillo weighing approximately 2 kg in a
similar period of time.

PART V: CLASSIFICATION OF THE PHANAEINES

That the phanaeine genera form a quite cohesive, natural assemblage has
been assumed for many years. Early authors referring new taxa to the
"Phanaei," "group of Phanaeus" etc., recognized the group primarily by the
form of the antennal lamellae and lack of tarsal claws. These same features
were stressed by MacLeay (1819) in proposing Phanaeits for certain Linnean
species of Scarabaeits.

The first formal taxonomic recognition of a phanaeine assemblage was

816 The University Science Bulletin

Kolbe's "unterfam. Phanaeinae" (1905: 550), which included Phanaeus,
Gromphas, Bolbites, Oruscatus, Megatharsis and Dendropoemon (sic).
Kolbe apparently overlooked Oxysternon and Diabroctis, both of which were
proposed for species of Phanaeus long before 1905. Gillet's checklist (1911)
included the aforementioned genera (except Diabroctis) along with En-
nearabdus and Pteronyx in Phanaeides, one of three subdivisions (though
not called subtribes by him) of Coprini. Olsoufiefr (1924) also applied the
name Phanaeides to essentially the same assemblage as Gillet except that he
included Diabroctis (as Taurocopris) and Tctramereia (as Boucomontius)
and excluded (p. 16) both Pteronyx and Ennearabdus. The latter genus has
since been placed in a tribe of its own, Ennearabdini (Pereira and Martinez,
1956). Moreover, Olsoufiefr divided Phanaeus into five subgenera, Sulco-
phanaeus, Coprophanaeus, Metallophanaeus, Megaphanaeus and Phanaeus,
sen. str. Blackwelder's checklist (1944) follows Olsoufiefr except that the
above subgenera are each raised to generic rank and Pteronyx and Ennearab-
dus are readmitted to Phanaeides.

Most modern workers have adopted a classification of Scarabaeinae based
upon the keys given by Janssens (1946, 1949), in which the larger tribes are
subdivided into subtribes. Accordingly, the phanaeine genera comprise
Phanaeina, one of the three subtribes of Coprini. (The subtribal names of
Janssens are generally emended to terminate in -ina rather than -ides; thus,
Phanaeides is now written Phanaeina.) Since Janssens cited neither the
constituent genera of his subtribes (except those in disagreement with the
keys) nor bibliography, one can only assume that their taxonomic compo-
sition agrees with that cited by previous authors.

For the present I have avoided formal nomenclatural reference to the
phanaeine genera as Phanaeina or as its nomenclatural equivalents
(Phanaeini, etc.). Thus, my collective references to this group of genera
(phanaeines, phanaeine Scarabaeinae, etc.) are intended to be informal
recognition of a taxon roughly equivalent taxonomically to Phanaeina
(== Phanaeides = Phanaeini) of authors. My principal reason for at least
temporarily side-stepping adoption of a formal name is the serious deficiency
in our understanding of the interrelationships of other genera currently em-
braced by Coprini as they bear upon the taxonomic position of phanaeines.
I have, thus, reluctantly concluded that Janssens' classification, at least as
applicable to Coprini, can at best be used with caution, pending its intensive
re-examination. Presently it suffices to say that I consider the phanaeine
genera discussed below a monophyletic taxon characterized by the following
features :

I. Antennal club three-segmented; basal lamella more-or-less scoop-shaped
and enclosing, at least partially, the two distal lamellae; club usually
large and spherical, but in Dendropaemon often flattened, elongate.

Evolution of the Phanaeine Dung Beetles 817

2. Dorsal surface of clypeus at least partially (usually wholly) transversely
ridged.

3. Eyes (compound) completely divided into upper and lower portions by
extension of paraocular areas of the parietals ("genae").

4. Front tarsi lacking in males; present or absent in females.

5. Middle and hind tarsi present, usually five-segmented, claws absent.

6. Metepisternal tab present.

7. Medial portion of metasternum acuminate anteriorly, often flattened and
pentagonal ("rhomboidal").

8. Elytral interstriae usually convex, smooth or punctate.

9. Secondary sexual dimorphism usually strongly pronounced, sexes being
easily distinguished by configuration of cephalic and pronotal protuber-
ances, ridges, etc.

10. Size variable, usually 10-30 mm; typically robust; colors often partially or
wholly metallic.

11. Eggs laid singly in earth-covered brood balls (see HalfTter and Matthews,
1966).

12. Restricted to New World, mainly neotropical.

OlsoufiefPs monograph (1924) is still the only comprehensive treatment
of the phanaeine species. It has, moreover, been the basis of all subsequent
treatments (including this one). OlsoufiefT evidently had access to most of
the known phanaeine type material, and his species descriptions and keys to
species are the standards. There are, however, a number of works postdating
his which complement it or enlarge upon its reliability as an identification
manual. Among them are these: Barattini and Saenz (1961, 1964;
Gromphas), Blackwelder (1944), Blut (1939), Martinez (1944, 1959),
Martinez and Pereira (1967), Pereira (1949), Pereira and Martinez (1956),
Pessoa (1934, 1935), Pessoa and Lane (1936, 1941), Robinson (1947) and
Vulcano and Pereira (1967:565 ft). The supraspecific classification presented
here differs from that of OlsoufiefT in the following respects:

1. Splitting of Phanaetts into three genera. The following outline shows
the general changes, details of which are brought out below :

Olsoufieff (1924) Present

Genus Phanaeus

Subg. Sulcophanaeus Genus Sidcophanaeus

Subg. Phanaeus, sen. str. Genus Phanaeus

Genus Coprophanaeus
Subg. Megaphanaeus Subg. Megaphanaeus

Subg. Metallophanacus Subg. Metallophanaeus

Subg. Coprophanaeus Subg. Coprophanaeus, sen. str.

818 The University Science Bulletin

2. Division of Oxystemon and Dendropaemon into two and three sub-
genera, respectively.

3. Recognition, where practical, of species-groups within generic cate-
gories. (This was attempted formally by OlsoufiefT only for Phanaeus,
sen.str.)

4. Removal of Bolbites, Gromphas and Oruscatus from the phanaeine
assemblage.

Removal of Bolbites, Gromphas and Oruscatus from "Phanaeina" is done
confidently but, at the same time, reluctantly since their alternative positions
in the current classification of Coprini are not in the least clear. The rather
strict definition of phanaeines given above definitely excludes Gromphas and
Oruscatus, at least morphologically. Both are probably more closely related
to dichotomies or, perhaps, to Old World elements which I have not
studied. Nevertheless, as far as the South American fauna is concerned, both
are taxonomic isolates, closely related neither to each other nor to other ex-
tant groups with which I am familiar. Bolbites onitoides Har. (monobasic),
on the other hand, may later prove to be a phanaeine in some wider taxo-
nomic sense than that presently used. It appears to show features of both
phanaeines (some aspects of head structure, general body shape, metepisternal
tab, etc.) and nonphanaeines (presence of tarsal claws, antennal club not of
Phanaeus-type, etc.). It could very well happen that this species will prove
to be highly important in assessing the position of phanaeines relative to the
rest of the tribe. Nevertheless, firm placement of these three genera must
await a thorough re-examination of Coprini on a world-wide basis.

The following classification is based upon phenetics and, therefore, sub-
ject to an inherent degree of inadequacy. The implicit hope is, of course,
that the characters used, although a small sample of those potentially available,
have been selected and studied sufficiently well that results based upon them
will be supported by newly discovered characters. The terminology used in
generic descriptions, keys, etc., is that established above; in doubtful in-
stances, readers should refer to the two preceding sections. Some character
states describing genera are numbered as are their definitions in part III;
salient features are italicized in generic descriptions.

A potential source of inconvenience is that I was able to examine very
little type material. The result could be perpetuation of a certain amount of
nomenclatural confusion attending the names of some of the more poorly
known species. I suspect that the greatest potential for error lies in the genus
Dendropaemon, most of the species of which are only poorly known and
subject to conflicting descriptions and usage in the literature. Nevertheless,
on the whole I do not anticipate a great deal of subsequent need for cor-
rection since the majority of species names have been used quite consistently
for many years. Moreover, during the course of this study, I consulted vari-

Evolution of the Phanaeine Dung Beetles

819

ous workers who together have examined a great deal of the available type
material; among them is Padre F. S. Pereira of Sao Paulo.

Below is an outline of the classification proposed in the remainder of this
part:

Genus DIABROCT1S Gistel
Genus SULCOPHANAEUS Olsoufieff
faunus-group
imperator-group
auricollis-growp
Genus PHANAEUS MacLeay
splendidulus-gxoup
palaeno-complex
endym ion-complex
chalco m elas-com plex
splendidnlus-complex
bispinus Bates (an isolate)
vindex-group

f index-complex

mexicanus-complex (comprising
several species-clusters)
Genus OXYSTERNON Castlenau
Subgenus OXYSTERNON, sen. str.
jestivum-group
sericen m -grou p
Subgenus MIOXYSTERNON

NEW SUBCENUS

Genus COPROPHANAEUS Olsoufieff
Subgenus MEGAPHANAEUS
Olsoufieff
Subgenus M ET A LLO PHANAEUS
Olsoufieff

Subgenus COPROPHANAEUS,
sen. str.

jasius-group
dardanus-group
Genus DENDROPAEMON Perty
Subgenus PARADENDRO-

PAEMON NEW SUBGENUS

Subgenus COPROPHANAE-

OIDES NEW SUBGENUS

Subgenus DENDROPAEMON,
sen. str.

dentico/lis-group
viridipennis-groxip
piceum-group
Genus TETRAMERE1A Klages
Genus HOMALOTARSUS Janssens
Genus MEG ATH ARSIS Waterhouse

Artificial Key to the Genera of Phanaeine Scarabaeinae

The following key is meant principally for purposes of identification. An
effort has been made to select those character states most easily seen. Infrageneric
taxa are separated in discussions of genera.

1. Middle and hind tarsi each with fewer than five segments; front tarsi

always absent in female 2

-Middle and hind tarsi five-segmented; front tarsi present or absent in

female 3

2. Middle and hind tarsi four-segmented, fourth (distal) segment very

small (Fig. 263); clypeus not emarginate medially (Fig. 95); body

robust (Figs. 312-313), color red-brown Tetramereia Klages

-Middle and hind tarsi either two or three-segmented (Figs. 270-279);
clypeus emarginate medially (Figs. 89-94); body robust to flattened
(Figs. 304-309); color usually at least partially metallic blue or
green Dendropaemon Perty

3. Dorsal surfaces of hind tibiae and basal four hind tarsal segments cov-

ered by large, setigerous granules (Fig. 260); tarsal segments as in
Figures 264 and 265; color dull metallic green; body robust (Figs.
310-311) Homalotarsus Janssens

820 The University Science Bulletin

- Dorsal surfaces of hind tibiae and tarsi glabrous or with small

setigerous granules; body robust, color variable 4

4. Labial palpus comprising a single oval segment (Figs. 154-155); dorsal

surface of hind tibiae (Fig. 266) and abdominal sterna (Fig. 301)

bearing dense fringes of rather long setae Megatharsis Waterhouse

-Labial palpi three-segmented; tibiae and abdominal sterna without

dense fringes 5

5. Anterior margin of clypeus emarginate medially (Figs. 83-89), clypeal

process U-shaped; front tarsi usually absent in female; front tibiae

as in Figure 246 Coprophanaeus Olsoufieff

-Anterior margin of clypeus not emarginate medially (Figs. 98-110),
clypeal process never U-shaped; front tarsi rarely absent in female;
front tibiae variable (Figs. 250-259) but not as in Figure 246 6

6. Transverse clypeal carina present (Figs. 194, 197); transverse hind

tibial carina present; large, green species Diabroctis Gistel

-Transverse clypeal carina absent; hind tibial carina rarely present; size

and color variable 7

7. Antero-median angle of metasternum prolonged as a long, sharp spine

extending between apices of the front coxae (Figs. 244-245); pos-
tero-median angle of pronotum acute, distinctly prolonged between

bases of elytra (Fig. 210) Oxy station Castlenau

-Antero-median angle of metasternum not spinate; postero-median

angle of pronotum not distinctly prolonged 8

8. Occipital ridge absent or incomplete; anterior portion of circumnotal

margin not broken by emarginations behind eyes (Figs. 117C, 12 1C-

122C) Sulcophanaeus Olsoufieff

-Occipital ridge complete (Figs. 1, 4); anterior margin of circumnotal
ridge distinctly interrupted by emarginations receiving postocular
lobes of parietals (Figs. 35, emarg; 36-38) Phanaeus MacLeay

Genus Diabroctis Gistel

Diabroctis Gistel, 1857: 604. Type: Scarabaeus mimas L., original designation.
Tattrocopris Olsoufieff, 1924:18. Type: Scarabaeus mimas L., present designation.

Until recently this genus has been known as Tattrocopris Olsoufieff; the
generic synonymy, Taurocopris = Diabroctis, was established by Martinez
and Pereira (1967). Olsoufieff designated no type-species for Tattrocopris.
In order to stabilize the use of these names by precluding future question
about their synonymy, it seems appropriate to hereby designate as the type-
species of Tattrocopris Olsoufieff (1924) the following species: Scarabaeus
mimas L. By so doing, Tattrocopris automatically becomes a junior objective
synonym of Diabroctis Gistel.

DESCRIPTION

(1) anterior margin of clypeus not emarginate medially, strongly toothed
(Figs. 98-99); (2) clypeal process transverse; lateral clypeal carinae well
developed; transverse clypeal carina present (Figs. 194, 197, tcrcly); (3)

Evolution of the Phanaeine Dung Beetles 821

cephalic process of male lamellate, a transverse ridge prolonged on each side
into acute "horn" (Figs. 194-195) (except mirabilis where lateral prominences
greatly prolonged); (4) cephalic process of female carinate (Fig. 197); (5)
paraocular areas of parietals narrow; (6) postocular lobes of parietals de-
pressed transversely (Fig. 115B), anterior portion circumnotal margin not
emarginate (115C); canthal area distinct (Fig. 115A, ca) ; (7) cephalic
brachium monopodal; (8) eyes small; (9) occipital ridge incomplete, repre-
sented only by lateral segments; (10) labrum Phanae us-type; (11) mentum
strongly emarginate; (12) median lobe of hypopharynx present; (13) pre-
mental sclerites marginally sclerotized; (14) pronotum of male type C (Figs.
194-195); (15) pronotum of female type D (Fig. 196); (16) pronotum
punctate (rugose antero-laterally in mimas) ; posterior pronotal fossae present
(except mirabilis), large; (17) front tarsi present in female; (18) front tibiae
type C (Fig. 250); hind tibiae carinate laterally (as in Fig. 262); (19) mid-
dle and hind tarsi five-segmented; (20) elytral interstriae smooth and (21)
evenly convex; (22) striae simple; hind wing not notched; color dar\, dull
green to metallic yellow-green; size large (length 25-40mm).

The three species included in Diabroctis {mimas [L.], cadmus [Har.],
mirabilis |Har.]) are easily recognized by their green color, large size and a
transverse clypeal carina anterior to the cephalic process. D. mimas and
cadmus are among the largest phanaeines (40 mm) and appear to be closely
related; the male prothorax of these species (Figs. 194-195) is exceptionally
massive. D. mirabilis differs from its congeners in the features brought out
above, the most striking of which is the massive, antler-like development of
the male cephalic process. All species have a distinct fringe of setae projecting
from the elytral suture (the setae insert on the elytral margins), which is
very prominent in mirabilis. Some species of Phanaeus (e.g., demon') have
such fringes.

Diabroctis is restricted to South America; although cadmus has been
cited (erroneously?) from Costa Rica (Gillet, 1911; Black welder, 1944), it
is probably only South American (Colombia). D. mimas occupies an enor-
mous range essentially determining the distributional limits of the genus
while mirabilis is common only in southern Brazil and adjacent regions. A
map of the range of Diabroctis is not given since it would look essentially
like that given for Oxysternon, sen. str. (Map 3). Little is known about the
biology of Diabroctis but it appears basically coprophagous. Both mimas and
mirabilis have been collected in large numbers from human and bovine feces;
mimas is often collected in swine feces in Manaus, Amazonas (Brazil).

Genus Sulcophanaeus Olsoufieff

Sulcophanacas Olsoufieff, 1924: 23 (as subgenus of Phanaeus). T\pe: Scarabaetis sulcatus
Drury = Scarabaetis carnijex L., 1758, not 1767, original designation.

Olsoufieff (1924: 23) designated Phanaeus sulcatus (Drury) as the type-

822 The University Science Bulletin

species of the subgenus Sulcophanaeus. The name sulcatits Drury has since
been recognized as a junior subjective synonym of carnijex L., 1758 (Mat-
thews, 1966: 39). Consequently, the valid name of the type-species of this
genus becomes Phanaeus carnijex (L.), 1758. It should be noted that, pend-
ing a ruling by the International Commission on Zoological Nomenclature,
the name Sulcophanaeus could later be ruled a junior objective synonym of
Onthurgus Gistel (see discussion below in introduction to Phanaeus).

description

(1) anterior margin of clypeus not emarginate medially (Figs. 102-105);
(2) clypeal process transverse, either tooth-like or ridge-like; lateral clypeal
carinae usually present; (3) cephalic process of male corniform (Figs. 198,
203-204, 209); (4) cephalic process of female corniform or carinate (Figs.
199, 201-202, 205, 207); (5) paraocular areas of parietals narrow; (6) post-
ocular lobes of parietals depressed transversely (part B of Figs. 117, 121-122),
circumnotal ridge not interrupted by emarginations (parts C of same Figs.) ;
canthal area usually distinct (Fig. 117); (7) cephalic brachium monopodal
(bipodal in few species); (8) eyes small; (9) occipital ridge incomplete;
(10) labrum Phanaeus-typc; (11) mentum strongly emarginate (Fig. 158);
(12) median lobe of hypopharynx present; (13) premental sclerite marginally
sclerotized (Fig. 150); (14) pronotum of male variable (Figs. 198, 203-204,
206, 209); (15) pronotum of female type C or D (otherwise in faunus
group, Figs. 199-200); (16) pronotum smooth (rugose in laeander and
columbi); anterolateral angles of pronotal ridge often angulate; posterior
pronotal fossae present (except actaeon)\ (17) front tarsi present in female
(except carnijex and, reportedly, steinheili)\ (18) front tibiae Phanaeus-typc;
transverse hind tibial carinae sometimes present; (19) middle and hind tarsi
five-segmented; (20) elytral interstriae smooth and (21) flat or convex; (22)
striae simple; hind wing not notched.

The name Sulcophanaeus is used here in a much different sense from
that used by Olsoufiefr (1924). In the subgenus Sulcophanaeus Olsoufiefr
included only two species, carnijex (as sulcatits) and stcinheili, which re-
portedly resembled each other in the structure of the male genital capsule
and lack of front tarsi in the female; but he was doubtful about the sound-
ness of his grouping. In the present sense Sulcophanaeus brings together a
rather diverse group of species including not only the above two but also a
large number included by Olsoufiefr in Phanaeus, sen. str. (all of species
group 1, species group 2 [except bispinus) and part of species group 3). The
genus forms three species groups which, because of their distinctness, would
likely warrant consideration as subgenera if the genus were larger. While
these species groups are certainly distinct, it is noteworthy that each, inter

Evolution of the Phanaeike Dung Beetles 823

se, comprises more or less readily recognizable isolated elements (phenetically
and often geographically) mostly as single species or species pairs.11

Key to the Species Groups of Sulcophanaeus

1. Both sexes with well developed corniform cephalic processes, that of

the male very long, curved, tapering (Fig. 198), that of the female
either similar to male {jaunus, Fig. 199, and columbi), strongly
bifurcate (laeander) or as in Figures 201 and 202 (carnijex); para-
ocular areas of parietals each strongly carinate above lateral to eye;
postocular lobes of parietals, seen laterally, extended anteriorly as
finger-like process (Fig. 117A) approaching upper portion of eye
(except carnijex); clypeal process tooth-like (Fig. 113); lateral
pronotal carina present (Figs. 199, 201, 203, lpc); elytral interstriae
unevenly convex ("sulcate") (Fig. 283); large beetles, either uni-
form dull black brown or metallic green jaunus group

-Cephalic process of male as above; that of female a simple or tuber-
culate carina (Figs. 205, 207); paraocular areas carinate or not;
postocular lobes of parietals not projecting anteriorly; clypeal
process ridge-like (except rhadamantus); lateral pronotal carina
absent (except male imperator); elytral interstriae convex or flat;
moderate sized (20-30mm), color variable 2

2. Paraocular areas of parietals not strongly carinate lateral to eye; front

tibiae tridentate, teeth concentrated in apical 1/3 of outer margin
(Fig. 251); elytral interstriae flat (Fig. 282), though sometimes
with irregular wrinkling; pronotum of male flat or shallowly con-
cave, sloping downward anteriorly from pair of closely set postero-
median processes (type F); pronotum of female convex except for
anterior transverse ridge shaped as in Figure 207; elytra dull black
(except actaeon), pronotum completely or marginally metallic.

auricollis group

-Paraocular areas of parietals carinate lateral to eye; apical three teeth
of front tibiae occupying more than 1/3 outer margin; elytral inter-
striae convex; pronotum of male variable; pronotum of female
variable; color variable imperator group

the jaunus group (f annus |F.|; columbi [MacL.]; laeander [Water-
house | ; carnijex [L.|): All these species are large insects (30+mm) and, in
addition to the features mentioned in the key, resemble each other quite
closely in general appearance of the head capsule. Although they are easily
distinguishable, columbi and laeander form a loose species pair (rugose
pronotum, metallic green color, bipodal cephalic brachium). Both, however,
are clearly related to jaunus (see Key), which at the same time shares a
number of unique features with carnijex (strong transverse hind tibial
carinae, uniform dull black-brown color, intrasternal position of last ab-

11 In this and subsequent discussions a species pair is meant to refer to two species which
are presumed more closely related to each other than either is to other members of a given
taxon. In general the members of a pair are very closely related and sometimes difficult to
separate (i.e., "sister species"), but this need not be so.

824 The University Science Bulletin

dominal spiracle). S. jaitnus, then, is an annectant species connecting other-
wise quite different species. The general resemblance of fannus and carnijex
is striking and was mentioned, but not stressed by Olsoufieff (p. 79).

S. carnijex is noteworthy in that, being endemic to Jamaica, it is the only
phanaeine not a continental species. The most reasonable explanation for its
presence there is that, like a number of other Jamaican scarabaeines (Mat-
thews, 1966), it is an insular relict of probably very long standing residence.
The relict status of carnijex as well as its ecology is discussed by Matthews
(1966). The other species in this group are South American (see Map 1).
The ranges of laeander and columbi are poorly known due to the rarity of
these species in collections; the former is generally considered Venezuelan
(a savanna species ?) while columbi has been collected in the southern part
of the Amazon Basin. S. jaitnus, which is definitely a forest species, occupies
a wide range in the Amazon Basin. Except for carnijex, the biology of these
species is largely unknown. Matthews (1966) reports that carnijex is copro-
phagous, occurring in large numbers in cow dung, and likely restricted to
certain soil types. This is the only instance where the possible limiting effects
of edaphic factors upon phanaeines has been reported; the distributional
effects of soil types on Scarabaeinae has so far been largely ignored. S.
columbi has been collected from tapir feces (B. Malkin, pers. comm.) while
jaunus is reportedly necrophagous( Pereira, pers. comm.). I know nothing
of the biology of laeander, but if it truly is a savanna species, I would expect
it to be coprophagous.

the imperator group {imperator [Chevr.|; stein heili [Har.]; rhadaman-
tus [Har. j; batesi [Har.]; menelas [Cast.]): This group consists of two
species pairs and one species which appears taxonomically isolated. The
latter species, rhadamantus, is extremely rare; only because it somewhat re-
sembles imperator do I place it here. S. batesi and menelas (splendididus of
Olsoufieff and often so identified in collections) are very similar (unique
shape of pronotum of both sexes, "sulcate" elytra), except for color differing
only in minor details. A study of the geographic variation of this pair is in
order since they could be subspecies. A rather close relationship between
imperator and steinheili can be supported here only on the basis of the males
(very similar pronota; presence of hind tibial carinae). This similarity was
mentioned both by Harold (1875) and Olsoufieff (p. 79). I have not seen
the female of steinheili, which, according to Olsoufieff, lacks front tarsi. This
species pair only slightly resembles menelas-batesi in general facies. The
presence of hind tibial carinae and lateral pronotal carinae link it with
jaunus-carnijex, which it resembles even less in general appearance. More-
over, imperator has a bipodal cephalic brachium, known elsewhere in the
genus only in laeander-columbi. Such a mosaic appearance of features can

Evolution of the Phanaeine Dung Beetles

825

lY,

*5*£

iX

-o

MAP 1

Sulcophanaeus
SJgs? carnifex group
:::$:$ imperator group
| auricol I is group

V

\

J%?

Map 1. Approximate distributions of the species groups of Sulcophanaeus. Arrows connect
widely disjunct portions of distributions.

be interpreted as either parallel evolution or a closer phyletic proximity than
indicated by the classification; I have chosen the former interpretation.

The imperator group comprises largely temperate species. Except for
steinheili, a rare species known only from the Bogota region of Colombia,
they occur in southern South America (Map 1). S. imperator, the southern
relative of this species, is also a high altitude species inhabiting the higher
Andean plateau regions of Argentina and Bolivia. The range of batesi
approximates that of imperator while menelas occurs in the grassland regions
of eastern and central Argentina and Bolivia and southern Brazil. S. rhada-
mantus is known from very few specimens, some from the state of Buenos

826 The University Science Bulletin

Aires, Argentina. Martinez (1959), reports that imperator, batesi and
menelas are strictly coprophagous, preferring herbivore (livestock) excre-
ment. The latter species, however, is evidently at least occasionally necro-
phagous (J. Foerster, pers. comm.).

the auricollis group {auricollis [Har.]; chryseicollis [Har.]; noctis
[Bates]; velutinus [Murray]; actaeon [Erichson]): Of the species groups
comprising Sukophanaeits this is the most cohesive, phenetically and geo-
graphically. Three of the five included species, auricollis, chryseicollis and
noctis, appear quite closely related, differing most obviously in color. While
the pronotum of the male of velutinus is strikingly different from that of the
above three species (cf., Figs. 206, 209), the pronotum of the females of
all four are very similar, principally because of the unique shape of the
anterior carina (Fig. 207). S. actaeon, the female of which I have not seen,
differs from the above species in several respects, the most striking of which
is its all metallic color. While the elytral interstriae are flat and dull black
in other species, in actaeon they are irregularly raised (appearing "blistered")
and metallic blue or green. To a much lesser degree, the "blistering" is seen
in other species of the group, particularly velutinus. A curious feature of this
group, seen also in some Phanaeus, is a tendency for the propleura to be
hollowed for the reception of the posterior margins of the paraocular areas
of the parietals. The depth of these concavities varies as does the size of the
tubercle adjacent to each which fits beneath the paraocular area.

The center of distribution of this group appears to be the lowlands and
foothills along the eastern slopes of the Andes from Panama southward
through Peru (Map 1). One species, chryseicollis, is largely restricted to
neotropical southern Mexico and is the only continental Sulcophanaeus not
occurring also in South America. Both noctis and velutinus occur from
Panama to Peru; auricollis is largely restricted to Colombia while actaeon is
known to me from only a few Peruvian localities. None of these species,
however, are well represented in recent collections with precise locality
designation. Accordingly, in view of the tremendous physiographic variation
of the general range of this group, very little is known of their ecological
distribution within it. I have no ecological data on actaeon and auricollis.
S. chryseicollis has been collected in southern Mexico at night using traps
baited with human excrement (Halffter and Matthews, 1966). Both noctis
and velutinus have been collected in the same manner from the Chiriqui
region of Panama (D. Viers, pers. comm.).

Genus Phanaeus MacLeay

Phanaeus MacLeay, 1819: 124. Type: Scarabaeus camijex L., 1767, not 1758 = Phanaeus
vindex MacLeay, subsequent designation by Olsoufieff, 1924: 23, see discussion below.

Lonchophorus Germar, 1824: 126. Type: Scarabaeus camijex L., 1767, not 1758 = Phanaeus
vindex MacLeay, present designation.

Evolution of the Phanaeine Dung Beetles 827

Onthurgus Gistel, 1857: 602. Type: Scarabaetis carnifex L., 1767, not 1758 = P/ianaeus vindex
MacLeay, original designation (objective generic synonymy subject to opinion of the
International Commission on Zoological Nomenclature, see below).

Phanaeus, sen. str., OlsoufiefT, 1924: 23 (pars).

The following nomenclatural melodrama has emerged in determining
:he valid name of the type-species of Phanaeus. MacLeay (1819) designated
no type-species for Phanaeus, nor were any of the included species indicated
as type-species. Subsequently, OlsoufiefT (1924: 23) designated as the type-
species of Phanaeus, sen. str., a North American species which he referred to
as P. carnifex (L.). Although he rightly suspected that Linnaeus had used
the name carnifex in at least two different senses (one in 1758, another in
1767), OlsoufiefT nevertheless synonymized the names carnifex L. and vindex
MacLeay (pp. 95 and 151). Matthews (1966: 41), in order to rectify the
confusion attending the name carnifex, designated a lectotype fixing the
name to a Jamaican species and thereby preserved Linnaeus's original usage
of the name carnifex (1758). Consequently, carnifex becomes the valid name
of the type-species of Sulcophanaeus (see above), and vindex, the only
available synonym for carnifex sensu Linnaeus (1767) and OlsoufiefT (1924),
becomes the valid name of the type-species of Phanaeus. The following
synonymies are based on several sources :

Sulcophanaeus carnijex (L.)

= Scarabaeus carnifex L., 1758: 346 (not 1767)

= S. siilcatus Drury, 1770

= S. sphinx Fab., 1772

= S. belzebul Fab., 1775

= S. nicanor Fab., 1781

= S. /arHerbst, 1789

= Phanaeus moloch MacLeay, 1819: 129 (species inquirenda)

= P. siilcatus (Drury), OlsoufiefT, 1924: 23
Phanaeus vindex MacLeay

zz Scarabaeus carnijex L., 1767: 546 (not 1758) (junior primary homonym)

= Phanaeus vindex MacLeay, 1819: 133 (species inquirenda)

— P. vindex MacLeay, 1833: 47 (from Gillet, 1911: 82)

= P. carnifex (L.), Gistel, 1857: 602

= P. carnifex (L.), OlsoufiefT, 1924: 23, 39, 95, 151

The mistaken identity of carnifex L., 1767, by Linnaeus himself and
subsequently by OlsoufiefT, has resulted in a case of mis-identified type-species
of a genus. In accordance with Articles 67 (j) and 70 (a) of the International
Code of Zoological Nomenclature, I have requested that the International
Commission on Zoological Nomenclature rule Scarabaeus carnifex L., 1767,
not 1758, = Phanaeus vindex MacLeay the type -species of Phanaeus Mac-
Leay. The use of the name carnifex by OlsoufiefT is unquestionably in the
sense of Linnaeus, 1767, hence in the sense of vindex MacLeay (see Olsou-
fiefT, 1924: 95).

As for the name Onthurgus Gistel, the situation is not quite so simple.

828 The University Science Bulletin

Gistel (1857: 602) designated as the type -species of this genus "Phanaeus
camifex (L.)", leaving uncertain in which sense of Linnaeus (1758 or 1767)
the name was used. The Code states (Art. 70) that one should assume that
names designated as type-species of genera are used correctly unless there is
evidence to suggest otherwise. To assume correct usage in this case would
be to assume that Gistel used the name carnijex in the sense of Linnaeus,
1758, thereby requiring that Onthurgus be regarded a senior objective
synonym of Sulcophanaeus. Since there are several reasons favoring the re-
tention of the name Sulcophanaeus (particularly the taxonomic relationships
implicit in the "phanaeus" part of the name), I have requested that the Com-
mission 1) exercise its plenary powers to suspend Article 70 of the Code and
rule that the type-species of Onthurgus Gistel is Scarabaeus carnijex L., 1767,
not 1758, = Phanaeus vindex MacLeay and thereby effect the objective syn-
onymy of Onthurgus and Phanaeus and/or 2) rule the name Onthurgus
Gistel a nomen oblitum in accordance with Article 23 (b) of the Code since,
to my knowledge, the name has never been used since 1857. A positive ruling
by the Commission on either (1) or (2) will preserve the present usage of
the name Sulcophanaeus.

DESCRIPTION

(1) anterior margin of clypeus not emarginate medially, usually toothed
(Figs. 4, 105-110); lateral clypeal carinae almost always present; (2) clypeal
process transverse (rarely spiniform); (3) cephalic process of male corni-
form (Figs. 1, 216, 218, 221, 226); (4) cephalic process of female carinate
(Figs. 3, 79, 81-82, 220, 222); (5) paraocular areas of parietals narrow; (6)
postocular lobes of parietals depressed obliquely (Fig. 4), prominent, fitting
in emarginations of circumnotal ridge (Fig. 35, emarg; not distinct in
haroldi, Fig. 119); canthal areas not distinct; (7) cephalic brachium bipodal
(anterior arm reduced in some species, Figs. 129-130); (8) eyes small; (9)
occipital ridge complete (Figs. 1, 4, 8, ocr) ; (10) labrum Phanaeus-type
(Figs. 18-19); (11) mentum strongly emarginate (Fig. 25); (12) median
lobe of hypopharynx present; (13) premental sclerites completely sclero-
sed; (14) pronotum of male type A or B (Figs. 35-36, 216-219, 221, 223-
228); (15) pronotum of female type B or D (Figs. 37-38, 220, 227); (16)
pronotal sculpturing punctate or roughened (Figs. 159, 161, 163-165) ; pos-
terior pronotal fossae usually present; (17) front tarsi present in female
(Figs. 61, 63); (IS) front tibiae tridentate or quadridentate (Figs. 60-61,
256-259); (19) middle and hind tarsi five-segmented (Figs. 64-66); (20)
elytral interstriae smooth (Figs. 285-287) or rugose (Fig. 284); (21) inter-
striae flat or evenly convex; (22) striae usually simple; hind wing not
notched (except bispinus); last abdominal spiracle marginal; color only
rarely not at least partially metallic.

The name Phanaeus is used here in a more restricted sense than has

Evolution of the Phanaeine Dung Beetles 829

heretofore been customary. It is roughly equivalent taxonomically to species
groups 3 and 4 of the subgenus Phanaeus, sensn Olsoufieff (1924). Taxo-
nomically Phanaeus is the most difficult phanaeine genus because of the high
incidence of irregular distribution of character states and of continuous vari-
ation of characters reliable (i.e., with discrete states) in other genera. Thus,
there exist in the genus (particularly evident in the mexicanus complex)
extremes which definitely warrant taxonomic recognition; but, on the
other hand, separation of the extremes is precluded by an array of inter-
mediate forms. In the interest of a balanced classification, I have concluded
it advisable to recognize only informally what appear to be the constituent
taxa of this genus. I have, therefore, used freely the following categories:
species-group, species-complex and species-cluster. Further study using more
characters will be necessary before the applicability of more formal categories
(e.g., subgenus) can be determined.

Key to the Species Groups and Species Complexes of Phanaeus

1. Pronotum evenly, strongly roughened (granulate to rugose) at least

antero-laterally but often entirely so; punctures, it present, re-
stricted postero-medially; cephalic process of female an isolated
carina or thick ridge (Fig. 82) (except homes cluster); pronotum
of male with prominent postero-lateral angles, usually appearing

more-or-less triangular from above vindex group 6

- Pronotum completely or largely punctate, often appearing glassy
smooth and shining, at most with irregular gibbosities (except fe-
male of meleagris) or minutely granulate (males of endymion com-
plex); cephalic process of female a transverse, usually trituberculate
carina extending fully between mesal ends of lateral clypeal carinae
(Fig. 81); pronotum of male variable splendidulus group 2

2. Pronotum of male with close-set pair of parallel, elongate processes (as

in Figs. 214-215); that of female with a single such process project-
ing forward from middle of posterior margin of an antero-median
concavity; elytral striae carinulate, distinctly punctured; hind wing

deeply notched; dorsum dull greenish-brown bispinus Bates

-Pronotum of male as in Figures 216-219, 221; that of female as in Fig-
ures 220, 222; elytral striae variable; hind wing not notched; color
variable 3

3. Clypeal process small, spinate; cephalic process of small males a short,

bituberculate process; elytral striae carinulate, not punctate; dorsum
dark, highly shining green; anterior margin of clypeus strongly
toothed medially; venter often bearing long, whitish hair .. palaeno complex
-Clypeal process variable, not spinate; cephalic process of small males

never a bituberculate process; other characters variable 4

4. Front tibiae of female quadridentate (Fig. 256), resembling Copro-

phanaeus-type; anterior margin of clypeus strongly toothed medi-
ally (Fig. 108); pronotum of male very flat dorsally, produced
postero-laterally and appearing triangular from above (Figs. 218-
219), sometimes very finely and irregularly granulate dorsally but

830 The University Science Bulletin

usually appearing smooth without magnification; pronotum of fe-
male smooth, without granulation endymion complex

-Front tibiae of female tridentate; anterior margin of clypeus at most

only weakly toothed medially; other characters variable 5

5. Elytral striae carinulate, distinctly punctate; pronotum bearing irregu-

larly spaced gibbosities dorsally, punctate antero-laterally (more-or-
less evenly, though not strongly rugose in meleagris); clypeal
process not tooth-like; pronotum of male very flat dorsally, pro-
duced postero-laterally and appearing triangular from above;
dorsum yellow-brown ("bronze ) or nearly black .... chalcomelas complex
-Elytral striae simple, not punctate; pronotum punctate or smooth;
clypeal process tooth-like (as in Fig. 113); pronotum of male deeply
concave dorsally (Fig. 217); color variable splendidulus complex

6. Elytral interstriae rugose (Fig. 284) or (seen magnified) distinctly

heavily punctured, punctures coalescent vindex complex

-Elytral interstriae smooth or only very finely punctured (Figs. 285-

286) mexicanus complex

The splendidulus group: This is essentially a South American group
(Map 2) the members of which fall rather neatly into the species complexes
separated above. The palaeno complex is a species pair (palaeno Blanch, and
kjrbyi Vigors) ; palaeno bears a striking resemblance to Oxysternon palae-
mon, the two species forming the only potential intergeneric mimic-model
complex among phanaeines. Both species of this complex occur (sym-
patrically ?) in the savanna-turned-pasturelands of southern Brazil and ad-
jacent regions; palaeno (and also \irbyi ■) is very common in cattle feces.

The endymion complex includes three closely related species (endymion
Har.; pyrois Bates; blanchardi Ols.) with quadridentate front tibiae re-
sembling those of Coprophanaeus. Both endymion and pyrois occur in
tropical lowland forests of Central America and southern Mexico and are
the only yet described members of the splendidulus group occurring outside
mainland South America. These two species have been collected from traps
baited with human feces. I know blanchardi from a single locality in Ecua-
dor ("Pucay"). There is a fourth, yet undescribed species likely belonging
in this complex which I know from a single, well developed male collected
near Temascaltepec, Mexico (state of Mexico) by H. E. Hinton. The male
in question is evidently one of several individuals collected by Hinton from
a decomposing hymenomycete fungus (probably a toadstool) (Hinton,
1935). Although referred to by him as endymion, which the new species
closely resembles, Hinton's illustration is undoubtedly of a well developed
male of the same new species, not endymion. (P. funereus Balths. is an
unassigned name probably belonging in this complex.)

The splendidulus complex includes three rather closely related species
(splendidulus [Fab.], dejeani Har. and melibaeus Blanch.) in addition to
haroldi Kirsch. The first two species have been collected from cow dung in

Evolution of the Phanaeine Dung Beetles

831

MAP 2

Phanaeus

vindex group

vindex complex '

mexicanus complex ■;

splendidulus group -z<

Map 2. Approximate distributions of the species groups of Phanaeus.

the Serra do Mar region of southern Brazil hut neither appears restricted to
the area. Some confusion exists with locality data under the name splendi-
dulus since until recently the name was erroneously applied to Sulcophanaeus
menelas. P. melibaeus is restricted to the southern Amazon Basin where it
has been collected in dung. The shape of the pronotum and aspects of head
structure set haroldi apart from other members of the complex; I have in-

832 The University Science Bulletin

eluded it here, perhaps artificially, on the basis of its closer resemblance to
melibaeus than to any other species in the group. The range of haroldi
appears concentrated in the tropical lowland valleys along the eastern slopes
of the Andes (Colombia, Ecuador and Peru) .

The chalcomelas complex is a species pair (chalcomelas Perty and
meleagris Blanch.). The former is a widespread Amazon Basin species
which is reportedly coprophagous (Halffter and Matthews, 1966: 31). It and
meleagris, a necrophagous species (he. eit., 31), bear distinct, though usually
irregular gibbosities over the surface of an otherwise punctate pronotum
which suggest the pronotal roughening of the vindex group.

Phanaeus bispinus Bates (an isolate of complex rank) resembles both
chalcomelas (carinulate-punctate elytral striae, color) and haroldi (shape of
cephalic process and other aspects of head structure), but several unique
features support its taxonomic isolation with the splendidulus group: shape
of the pronotum of the male (not unlike that of Oxysternon curvispinum,
Figs. 214-215) and of the female (which is unlike any other phanaeine, a
single elongate process in place of the paired processes of the male) and
notching of the hind wing (seen otherwise only in some Coprophanaeus
and in Dendropaemon). I have seen few specimens of bispinus, all from the
Amazon Basin; Pereira and Martinez (1956) say it is necrophagous.

The vindex group: This species group includes the bulk of those species
here assigned to Phanaeus (32 of 44 species). Most are North American
species (Map 2).

The vindex complex includes four species with either rugose or heavily
punctured elytral interstriae. The rugosity of the elytra is strongest in vindex
MacL. (Fig. 284) and difformis Lee; in igneus MacL. (= floridanus Ols.)
and triangularis Say ( = torrens Lee), however, the elytral rugosity tends to
be effaced and evident only with magnification as dense, confluencent punc-
turing. All four species are predominantly coprophagous and commonly
collected in cattle feces in pastures. P. igneus and triangularis are endemic to
the United States (particularly the southeast) ; the ranges of vindex and
difformis, while centered in the United States, extend into northern Mexico.

The mexicanus complex brings together a relatively large number of
species whose combined ranges extend from northwestern South America
to the southwestern United States (Map 2). Most are endemic or largely
endemic to tropical Mexico and adjacent Central America. All included
species whose feeding habits are known are normally coprophagous al-
though their ecological distribution (horizontal and vertical) appears quite
wide. The mexicanus complex remains in need of revision at the species
level using more characters than those employed here; a detailed study of
these species including geographic variation and reference to type material

Evolution of the Phanaeine Dung Beetles 833

could well result in reduction to subspecific rank or synonymy of some of
the species cited below. I have little doubt that the mexicanns complex is a
natural group in spite of the fact that my study so far has produced no sound
basis for assessing the interrelationships among the constituent clusters of
obviously closely related species. I find that character states that are normally
fixed in other phanaeine taxa of comparable size and/ or rank (e.g., various
aspects of integumental sculpturing) vary in a seemingly irregular pattern
among these clusters. Still the only satisfactory way, albeit taxonomically
unsatisfying to me, of distinguishing closely related species is by differences
in secondary sexual features of well developed males (particularly pronotal
shape). The very striking phenetic similarity of small males and females of
closely related species attests to a degree of virtual siblingness unapproached
in any other phanaeine taxon. Identification of potentially sibling indi-
viduals can at best only be inferred, and then only if they form parts of a
sample including well developed males.

Following are diagnoses of the species-clusters comprising the mexicanns
complex; included are two isolated species treated as of cluster rank. In
general, the species of each cluster resemble each other in at least the follow-
ing respects: shape of male pronotum, relief of elytral interstriae (flat or
convex); pronotal sculpturing (granulate or rugose) and great similarity of
females. I have avoided compiling a key to species-clusters since variation
of characters used would render it largely inefficient and little better than
that of OlsoufiefT (1924: 37-46).

Cluster 1 (lugens Nev., prasinus Har., hermes Har.) : Prothorax evenly
granulate; male pronotum as in Figures 223-224; female pronotum convex,
with three flat antero-median tubercles; elytral interstriae weakly convex,
striae distinctly punctate (seen under 10X magnification). Besides striking
interspecific female similarity, the males very closely resemble each other;
the most conspicuous variation distinguishing most individuals of either sex
is color. Both prasinus and lugens occur in Venezuela and adjacent regions
of Colombia; hermes ranges from Colombia into Central America.

Cluster 2 (mexicanns Har., amythaon Har., wagneri Har., lunaris
Faschb.) : Prothorax rugose, less heavily so in female; male pronotum as in
or similar to Figure 226; female pronotum as in Figure 227, with low-set,
transverse antero-median ridge flanked by shallow, usually smooth con-
cavities; elytral interstriae convex, striae simple or very weakly punctate. I
cannot reliably identify females or small males of these species unless they
are collected along with well-developed males of a single species; well-
developed males differ in subtle features of pronotal shape. The distribution
of this cluster is unique, the composite range extending from Peru to Ari-
zona: lunaris occurs in eastern Peru (and Ecuador ?), wagneri is largely

834 The University Science Bulletin

Central American, mexicanus is very common in southern Mexico while
amythaon has been collected from southwestern Mexico northward along the
western seaboard of Mexico into Arizona (Prescott).

Phanaeus achilles Bohart: I have seen one well developed male of this
species; it resembles the males of cluster 2 in all respects except color (bright
red above) and strong puncturing of the elytral striae. It is known from
Peru and Ecuador.

Cluster 3 {demon Har., obliquans Bates) : Pronotum rugose except dor-
sally; male pronotum (see below); female pronotum as in Figure 227; elytral
interstriae weakly convex, striae simple; leading margin of head distinctly
notched at lateral extremity of each postclypeal carina (Fig. 109). The
pronotum of the male of demon is unique in the genus (Fig. 225); the
postero-lateral angles are produced as massive projections (which are acute
and slanted posteriorly in obliquans) flanking a deep, smooth median con-
cavity. Both species are ordinarily of a green color much more highly shining
than the usually dull metallic colors of other species in the complex. P.
demon is very common in the hot tropical valleys (e.g., of Rio Balsas) of
southern Mexico; I know obliquans only from a few Central American
localities.

Cluster 4 (qimdridens Say, damocles Har.): Pronotum rugose; prono-
tum of female convex, with transverse anteromedian ridge; elytral interstriae
flat, shining, striae nearly effaced (Fig. 286). The pronotum of male
quadridens is flat above with large, flaring postero-lateral angles; that of
male damocles closely resembles that of cluster 5 (see below). Both species
are evidently restricted to higher altitudes in Mexico; quadridens is character-
istic of the Mexican Plateau (see HalfTter, 1964), where few other Phanaeus
are encountered, while damocles is found in pine-oak forests in southern
Mexico (Halffter, pers.comm.).

Cluster 5 (palliatus Sturm, eximius Bates) : Pronotum rugose; male
pronotum shaped as in Figures 223-224 except that dorsal surface concave
with median tubercle; pronotum of female as in cluster 4; elytral interstriae
convex, striae simple. This species pair is perhaps to be regarded as an
annectent among clusters 5, 7 and 8. I have seen very few eximius, all from
Central America; palliatus is also generally regarded as rare although it can
be found in numbers in isolated localities in the mountains near Mexico City.
G. Halffter and I collected many specimens of palliatus from burro and
human dung near Ocoyoacac (state of Mexico); HallTter feels that it is
narrowly distributed altitudinally. The variation in color of palliatus is
striking; in the single Ocoyoacac sample are bright metallic red, green and
dark blue individuals of both sexes.

Cluster 6 (pilatei Har., sallei Har., beltianus Bates, flohri Nev.) : Pro-
notum rugose; shape of pronotum of well developed males similar to cluster

Evolution of the Phanaeine Dung Beetles 835

1 but with very smooth, triangular area basally; pronotum of female convex,
with small transverse antero-median ridge; elytral interstriae convex, striae
simple. This cluster is evidently restricted to Central America and adjacent
regions of Mexico. Although I have seen only a few specimens of each, I
suspect that of the four names listed, only two will remain full species upon
serious study. I know nothing of the biology of these species.

Cluster 7 (guatemalensis Har., amethystinus Har., [probably = martinezi
HalfTter|, scutijer Bates, melampus Har.): Pronotum rugose; pronotum of
well developed males similar to Figures 223-224, with smooth basal area
bearing small acute tubercle anteriorly; pronotum of female convex, with
medially dentate ridge anteriorly; elytral interstriae convex, striae simple.
These species are very similar; all except scutijer (likely a color variation of
guatemalensis} are dull or blue-black. As in cluster 6, these species are re-
stricted to southern Mexico and Central America.

Phanaeus adonis Har.: This species, relatively common in valleys along
the southern periphery of the Mexican Plateau (Halftter, pers. comm.), re-
sembles triangularis more than any Mexican species. As in triangularis, the
male pronotum is flat dorsally and strongly produced postero-laterally and
the pronotum of both sexes is coarsely granulate. Unlike the vindex com-
plex, however, the elytral interstriae are at most weakly punctured.

Cluster 8 (daphnis Har., tricornis Ols., widens Cast., nimrod Har.,
jurcosus Bates): Pronotum rugose; pronotum of well-developed males with
rounded postero-lateral angles, concave dorsally with usually large median
or basal process of varying shape, sides of concavity serrate and/or dentate
(Fig. 228); pronotum of female convex, with small antero-median ridge;
elytral interstriae convex, striae simple. All these species are relatively small
(to 15mm in length) and usually wholly green or yellow-green. I cannot
reliably separate females and most small males of this cluster without cor-
roborating evidence; of the subtle pronotal features distinguishing well
developed males, the shape of the median process is most conspicuous. This
cluster occurs from southern Mexico northward along the western coast of
Sinaloa.

Unassigned names probably belonging in the Phanaeus mexicanus com-
plex: goidanichi Balths., pseudojurcosus Balths., substriolatus Balths.,
jran\enbergeri Balths., babori Balths.

Genus Oxysternon Castlenau

description

Dorsal surface of clypeus completely transversely ridged or punctate

posteriorly; (1) anterior margin of clypeus not emarginate medially, at most

weakly toothed (Figs. 210, 213-214); (2) clypeal process transverse, spini-

form (Fig. 112) or reduced; lateral clypeal carinae present or absent; (3)

836 The University Science Bulletin

cephalic process of male corniform (Figs. 210-211) or reduced to a short
bituberculate process or two simple tubercles; (4) cephalic process of female
a simple or tuberculate carina; (5) paraocular areas of parietals narrow; (6)
shape of postocular lobes of parietals as described in part III, anterior portion
of pronotal ridge effaced ( Mioxy sternon) or weakened or broken by shallow
emarginations; canthal area distinct or not; (7) cephalic brachium bipodal
(not observed for Mioxy stern on) \ (8) eyes small; (9) occipital ridge com-
plete; (10) labrum Phanaeus-type; (11) mentum strongly emarginate; (12)
median lobe of hypopharynx present; (13) premental sclerites completely
sclerotized; (14) pronotum of male as in Figures 210-211 or 214-215; (15)
pronotum of female type D (Figs. 212-213); (16) pronotum punctate (Fig.
160), often appearing completely smooth; postero-median angle of pronotum
distinctly produced as acute process fitting between inner basal angles of
elytra (Fig. 210), posterior pronotal fossae absent (except ? biimpressum
Ols.) ; antero-median angle of metastemum strongly produced as acute spine
fitting between front coxae (Figs. 244-245) ; (17) front tarsi present in female;
(18) front tibiae tridentate or quadridentate; (19) middle and hind tarsi five-
segmented; (20) elytral interstriae smooth to punctate and (21) evenly
convex (Figs. 280-281); (22) striae simple, sometimes nearly effaced; hind
wing not notched; seventh abdominal spiracle marginal or intrasternal.

In addition to the above features there is a tendency among species of
Oxysternon for abdominal sternum 7 to be almost vertical relative to sternum
6, such that, seen in profile, the abdomen has a decidedly angulate appearance.

Key to the Subgenera and Species Groups of Oxysternon

1. Cephalic process of male two closely-set median tubercles (Fig. 80),

that of female either absent or a weak, sometimes tuberculate
carina; clypeal process reduced to small tubercle; lateral clypeal
carinae absent; occipital ridge nearly effaced; metasternum, along
inner margin of each mesocoxa, with a single row of irregularly
placed large punctures; elytral striae distinct, usually strongly
punctate; last abdominal spiracle intrasternal; color metallic blue

or green; length 12-15mm Mioxysternon

-Cephalic process of male horn-like, often massive or, in small indi-
viduals, reduced to a pair of closely set median tubercles; that of
female a strong, trituberculate carina; clypeal process transverse
or spiniform; lateral clypeal carinae present, at least partially;
occipital ridge strongly developed; metasternum not strongly punc-
tate adjacent to mesocoxae; elytral striae present but often nearly
effaced, never strongly punctate; last abdominal spiracle marginal;
color variable, usually bright; length 10-35mm Oxysternon , sen. str 2

2. Dorsal surface of clypeus black; clypeal process spiniform; vertical,

postero-lateral portion of circumnotal ridge effaced by a broad,
oblique thickening (Fig. 211, arrow); each mesocoxa paralleled
along outer side by wide, raised ridge-like portion of metasternum

Evolution of the Phanaeine Dung Beetles

837

Map 3. Approximate distributions of the subgenera of Oxysternon. (?'s refer to occurrence
of Oxysternon, sen. str.)

which does not bend dorsally at anterior end; front tibiae tridentate,
only apical two teeth carinate on anterior surface; hind tibiae each
with weak transverse carina on outer face near apex; hind apical

tibial spur truncate distally jestivum group

Dorsal surface of clypeus at least partially metallic; clypeal process
transverse (spiniform in palaemon); circumnotal ridge not effaced
postero-laterally; thickened outer border of mesocoxal depression
narrow, bent dorsally at anterior end; front tibiae quadridentate,
apical three teeth carinate anteriorly, basal tooth small; hind tibiae
not carinate on outer apical surface; hind tibial spur acute .... silenum group

838 The University Science Bulletin

Subgenus Oxysternon Castlenau, sensu stricto

Oxysternon Castlenau, 18-10: 82. Type: Scarabaeus festivus L., present designation.

Sternaspis Hope, 1837: 52 (junior homonym of Sternaspis Otto. Type: Scarabaeus festivus L.,

present designation.
Strombodes Gistel, 1857: 602. Type: Scarabaeus festivus L., original designation.

Among the members of this subgenus are the better known Oxysternon.
All are large, usually conspicuously colored South American phanaeines
(Map 3). Both constituent species groups are quite cohesive, although
palaemon is a phenetic annectant between them (see below).

The festivum group (festivum |L.|; macleayi Nev.; lautum |MacL.];
conspicillatum [Web.]; ebeninum Nev.): Various structural details (par-
ticularly the shape of the pronota of well-developed males) can be used to
subgroup the constituents of this taxon. Thus, festivum and macleayi form
a species pair, and the remaining three species constitute a second subgroup.
All species have characteristic coloration, which evidently varies little. Two,
festivum and conspicillatum, are quite common in collections. Surprisingly
little is known of the biology of this group, although collection data indicate
all are amazonian. The range of conspicillatum is enormous, covering the
entire Amazon Basin south of the Amazon and Rio Negro and certain non-
basin areas (e.g., Quito, Ecuador). 0. festivum is also amazonian but occurs
north of the Amazon River (? into the Guianas); festivum has been collected
from traps baited with human feces. The known range of ebeninum is
restricted along the upper Rio Negro. Both lautum and macleayi are known
only from the Amazon Basin.

The silenum croup (silenum Cast.; smaragdinum Ols.; sericeum Ols.;
palaemon Cast.): The most common species, palaemon, is noteworthy in
that it shows features (e.g., spiniform clypeal process) otherwise unique to
the festivum group. This species is coprophagous and occurs in great num-
bers in the savanna regions ("cerrados") of south-central Brazil. I have seen
very few specimens of silenum, smaragdinum and sericeum; they appear so
similar to me as to suggest conspecificity. All specimens of these three species
that I have seen were collected at scattered localities in the western portions
of the Amazon Basin.

Unassigned names probably Oxysternon, sen. str.: biimpressum Ols.,
aeneum Ols.

Mioxysternon new subgenus

Type: Oxysternon spinijerum Cast. (Derivation of subgeneric name: duos (dwarf) -f- Oxy-
sternon, neuter; reference to relatively small size.)

This subgenus brings together three small-sized, rare species of Oxy-
sternon (spiniferum Cast.; curvispinum Ols.; pteroderum Nev.) clearly
distinguished from their congeners by those features mentioned in the key.

Evolution of the Phanaeine Dung Beetles 839

Most noteworthy of these are the apparent lack of a corniform cephalic
process on otherwise well developed males, small size, obliteration of the
occipital ridge and shape of the pronotum of the male (Figs. 214-215), which
in the type-species resembles that of Phanaeus bispinus. The pronotum of
the single male pteroderum I have seen bears two parallel, longitudinal crests
rather than the parallel prolongations of spiniferum. Of interest is the fact
that the cephalic process of the male of Mioxystemon resembles that charac-
teristic of the "minor" males of Oxysternon, sen. str., two closely-set tubercles.
I have seen very few specimens of these species, all from scattered South
American localities (Map 3).

Judging by original description, striatopunctatum Ols. would also fall in
this subgenus.

Jb^

Genus Coprophanaeus Olsoufiert

DESCRIPTION

(1) anterior margin of clypeus emarginate medially (Figs. 83-88); (2)
clypeal process U-shaped (Fig. 114); lateral clypeal carinae present; (3)
cephalic process of male corniform or lamellate (Figs. 181, 186, 188, 192-193) ;
(4) cephalic process of female corniform (Fig. 182) or carinate; (5) para-
ocular areas of parietals usually narrow (Fig. 116); (6) postocular lobes of
parietal depressed transversely, pronotal margin entire (Fig. 116) or weakly
effaced but never distinctly broken by emarginations; canthal area usually
distinct (Fig. 116A); (7) cephalic brachium monopodal (Fig. 126); (8)
eyes almost always large; (9) occipital ridge incomplete; (10) labritm
Coprophanaeus-/y/?<? (Figs. 146-148); (11) mentum weakly emarginate (Fig.
157); (12) median lobe of hypopharynx present; (13) premental sclerites
marginally sclerotized (as in Fig. 150); (14) pronotum of male type C, E or
F (Figs. 181, 183, 186, 189, 192); (15) pronotum of female type A (Figs.
184-185, 190-191) or like type E of male (Figs. 181-182); (16) pronotum
usually at least partially rugose or distinctly ridged (Fig. 166) ; posterior
pronotal fossae usually present; (17) front tarsi present or absent in female;
(18) front tibiae Coprophanaeus-/y/7<? (Fig. 246); (19) middle and hind tarsi
five-segmented; (20) sculpturing of elytral interstriae variable, usually
smooth; (21) interstriae flat or convex; (22) striae sometimes carinulate;
hind wing sometimes notched basally (Fig. 292) ; position of last abdominal
spiracle variable.

This genus brings together three of the five subgenera of Phanaeus estab-
lished by Olsoufiert (1924): Coprophanaeus, Megaphanaeus and Metallo-
phanaeus. The subgeneric rank of these three taxa is here maintained. All
Coprophanaeus are immediately recognizable by the presence of emarginate
clypeus, strongly quadridentate front tibiae and usually distinct ridging of

840 The University Science Bulletin

the pronotum. The only other phanaeines likely to be confused with this
genus are the few robust species of Dendropaemon. Coprophanaeus is South
American although a few species occur in Mexico (Maps 4 and 5) ; all species
whose habits have been observed are exclusively or ordinarily necrophagous,
usually nocturnal and tend to occur in forests.

Key to the Subgenera and Species Groups of Coprophanaeus

1. Front tarsi present in female; striae conspicuous, always distinctly

carinulate, carinulae straight or undulate (Fig. 288); hind wing
not notched basally; last abdominal spiracle located within sternum
7 (though not visible laterally in bellicosus); posterior surface of
each front tibial tooth with oblique, conspicuous brush-like clump
of short setae; female cephalic process (except of bellicosus) and
pronotal shape very similar to those of male such that sexual
dimorphism, especially among smaller individuals, is not pro-
nounced; pronotum of both sexes with an often massive, more-or-
less saddle-shaped or quadrate process extending anteriorly from
posterior portion of large median concavity; paraocular areas of
parietals dorsally distinctly carinate lateral to eye; occipital areas of
parietals each with angulate prominence; size large to very large
(40-60mm); dorsum dark metallic blue, green or deep violet
Megaphanaeus

-Front tarsi absent in female; striae often nearly effaced, at most only
weakly carinulate in some individuals with green elytra; hind
wing notched or not basally; last abdominal spiracle marginal or
set in distinct notch in edge of sternum 7; posterior surface of front
tibial teeth each with single, oblique row of setae; sexual di-
morphism, as expressed by cephalic and pronotal processes, very
pronounced or not; pronotum of female with shallow antero-
median concavity bordered anteriorly by transverse, crest-like
ridge; pronotum of male variable; paraocular areas of parietals
with no distinct carina lateral to eye (except in pertyi and pessoai);
occipital areas of parietals more-or-less flattened, without distinct
angulate prominence; size moderate (20-40mm); color variable 2

2. Hind wing not notched basally; width of lower portion of each eye less

than twice (usually about equal) width of adjacent oculogular
space; cephalic process of male a simple, tapering "horn"; color of
pronotum and usually also elytra metallic, shining Metallophanaeus

- Hind wing distinctly notched basally, vein V2 usually conspicuously
hooked apically (Fig. 292); lower portion of eyes large, bulging,
width of each more than twice width of adjacent oculogular
space; cephalic process of male lamellate, as in Figures 192, 193 or
188, never a simple, tapering horn; dorsum usually black or dull
blue-black with metallic areas restricted to margins of pronotum,
etc., rarely nearly wholly metallic green above (C. spitzi and some
specimens of C. milon and pluto) Coprophanaeus, sen. str 3

3. Anterior portion of circumnotal ridge not distinctly effaced behind

Evolution of the Phanaeine Dung Beetles 841

eyes; last abdominal spiracle marginal, not set in distinct notch in
sternum; posterior margin of paraocular areas of parietals (seen
dorsally) straight, not bending toward posterior angle of eye;
pronotum of large males never with strongly produced, quadrate
median process; cephalic process of larger males as in Figures 192-
193, its sides distinctly convergent apically jasius group

- Anterior portion of circumnotal ridge distinctly effaced behind eyes,
gaps so produced each as wide or wider than adjacent portion of eye
(as in Fig. 119C); last abdominal spiracle set in distinct notch in
dorsal margin of sternum 7; posterior margin of paraocular areas
seen dorsally) distinctly bent toward posterior angle of eye; pronotum
of well-developed males with large quadrate median process (Figs.
186-187); cephalic process of male with sides parallel or divergent,
usually plate-like with two or three apical tuberculate processes
(Fig. 188) dardanus group

Subgenus Megaphanaeus OlsoufiefT

Megaphanaeus Olsoufieff, 192-1: 23 (as a subgenus of Phanaeus). Type: Scarabaeus lancijer
L., original designation.

This subgenus includes some of the largest known Scarabaeinae; the
largest phanaeines, C. ensijer and lancijer, regularly exceed 50mm in length;
only some species of Heliocopris are larger. Also noteworthy is the strong
resemblance of males and females in development of secondary sexual fea-
tures, the most unusual aspect of which is the presence of a massive cephalic
"horn" on the female. The sexual resemblance is only superficial since sepa-
ration of sexes can always be immediately accomplished by examination for
front tarsi, which are present only in females.

The name Megaphanaeus is used here as established by Olsoufieff. It
includes four species {lancijer [L.|; ensijer Germ.; bonariensis Gory; belli-
cosus Oliv.), the first three of which are evidently rather closely related.
These three species have strongly sculptured elytra with undulate strial
carinulae; the pronotum is likewise strongly sculptured, bearing flattened
granules which coalesce to form extensive transverse ridging. C. bellicosus
differs from them in several respects: non-corniform cephalic process of the
female (which closely resembles that of Sulco phanaeus carnijex, Figs. 201-
202), smooth, flat elytral interstriae, straight elytral carinulae and deep violet
color. (The above species are bright metallic blue, green or blue-green.)
Among other features, elytral sculpturing can be used to separate the species
of this subgenus.

With the possible exception of bellicosus, whose habits I do not know,
Megaphanaeus is a necrophagous group. I have collected lancijer in sur-
prisingly large numbers from carrion-baited traps along the Amazon River
(Belem and Manaus) and ensijer in a similar manner in the state of Sao
Paulo, Brazil. Martinez (1959) reports bonariensis to be necrophagous,

842

The University Science Bulletin

&

MAP 4

> Metal lophanaeus

::::l]:::: . Megaphanaeus

Map 4. Approximate distributions of the subgenera Megaphanaeus and Metallophanaeus.

preferring bird carcasses. All species are South American (Map 4), where
they appear to be restricted to forests.

Subgenus Metallophanaeus OlsounefT

Metallophanaeus Olsourteff, 1024: 22 (as subgenus of Phanaeus). Type: Phanaeus saphirinus
Sturm, original designation.

This subgenus brings together two species-pairs, saphirinus Sturm-
horus (Waterhouse) and pertyi Oh.-pessoai Pereira. According to Martinez
and Pereira (1967), thalassinus Perty, which I have not seen, is closely re-

Evolution of the Phanaeine Dung Beetles 843

lated to the latter pair. Both pairs are quite distinct and perhaps brought
together only artificially here, the only alternative (which I consider inade-
quate) being to erect a separate subgenus for pertyi-pessoai. Olsoufieft (1924)
and Martinez and Pereira (1967) include the latter pair in Coprophanaeus,
sen. str. The general appearance of the pertyi group suggests an annectant
taxon between Metallophanaeus and Coprophanaeus, sen. str. All species
except pertyi occur in southern South America (Map 4) ; pertyi is limited to
the arid regions ("caatinga") of northeast Brazil. Contrary to Halftter and
Matthews (1966:30) saphirinus is at least sometimes necrophagous (Ed-
monds, 1967) ; the same authors cite horns as necrophagous.

Subgenus Coprophanaeus Olsoufieft", sensu stricto

Coprophaneaus Olsoufieft, 1924: 22 (as subgenus of Phanaeus). Type: Scarabaeus jasius Oliver,
original designation.

The name Coprophanaeus is used here in a somewhat more restricted
taxonomic sense than by Olsoufieft. The included species, which comprise
the bulk of the genus, are all somber colored and robust and at a glance
rather closely resemble each other.

The jasius group includes at least four species (jasius [Oliver]; cerberus
|Har.]; parvulus [Ols.]; spitzi [Pessoa]) all restricted to South America
(Map 5) ; jasius occupies an enormous range coincident with all but the
southernmost regions of the composite range of the group. I have collected
all four included species from carrion-baited traps in Brazilian forests.

The dardanus group also comprises at least four species (milon [Blanch.];
telamon [Er.]; dardanus [MacL.]; pluto [Har. |). The composite range of
this group is much wider than the former. This is the only coprophanaeine
taxon species of which occur outside South America; i.e., the Central Amer-
ican distribution indicated in Map 5.

Unassigned names probably belonging in Coprophanaeus, sen. str.: cos-
tatus Ols., arrowi Ols., jasion Fels., boucardi Nev., punctatus Ols., cono-
cephalus Ols., ohausi Fels., sericeus Fels., rex Balths., strandi Balths., roubali
Balths.

Genus Dendropaemon Perty

DESCRIPTION

(1) anterior margin of clypeus emarginate medially (Figs. 89-94, 132,
134) ; (2) clypeal process U-shaped, sometimes reduced; lateral clypeal
carinate present or absent; (3) cephalic process of male usually carinate,
sometimes lamellate (Fig. 305); (4) cephalic process of female carinate;
paraocular areas of parietal s usually raised medially into large bump or
transverse ridge (Fig. 134) ; (5) paraocular areas of parietals usually wide
(Fig. 134) ; (6) postocular lobes of parietals depressed transversely, circum-

844

The University Science Bulletin

i/&K-

'tr

^>

■%> .

MAP 5

Coprophanaeus, sen. str.

&F

Map 5. Approximate distribution of Coprophanaeus, sen. str.

notal ridge not interrupted by emarginations (Figs. 123C, 124C) ; canthal
area distinct (Figs. 123 A, B, 124A, B); (7) cephalic brachium monopodal;
(8) eyes small or large; (9) occipital ridge incomplete; (10) labnim Copro-
phanaeus-ty/?^ (Figs. 144-145), apical margin deeply emarginate, median
brush reduced to jew long setae inserting in apical for\ of antero-median
process of suspensorium; (11) mentum wealth emarginate (Fig. 152); (12)

Evolution of the Phanaeine Dung Beetles 845

median lobe of hypopharynx absent; (13) premental sclerites marginally
sclerotized (Fig. 151); (14-15) pronotum of both sexes usually similar, more-
or-less flattened (see discussion below); (16) pronotum partially roughened
or entirely punctate; posterior pronotal fossae present; transverse propleural
carina present {Dendropaemon, sen. str., Figs. 179-180) or absent (Copro-
phanaeoides and Paradendropaemon); (17) front tarsi absent in female;
(18) front tibiae Dendropaemon-/y/7(? (Fig. 247) (except ganglbaueri);
lower apical angle of middle and hind femora expanded, covering base of
tibia (Figs. 268-269) (except ganglbaueri); (19) middle and hind tarsi two-
or three-segmented (Figs. 270-279) ; (20) elytral interstriae smooth or punc-
tate; (21) same evenly convex; (22) striae carinulate (Figs. 290-291) (very
weakly so in ganglbaueri); hind wing notched (Figs. 294-295, 298) or not
(Fig. 299) ; base of pygidium impressed on each side of median furrow (Fig.
162, Id) (very weakly so in ganglbaueri).

In addition to the above features, Dendropaemon is characterized by a
remarkably great variation in body shape. Included in the genus are two
species (amyntas and ganglbaueri) of the generally robust shape character-
istic of other phanaeines (Figs. 304-305). Most Dendropaemon, however,
are narrowed and dorso-ventrally flattened to some degree (Figs. 306-309).
Following is a list of some of those observed skeletal features judged directly
related to the flattening-narrowing trend among these species. As indicated,
most vary continuously among progressively flatter, narrower species.

a. Reduction in size of cephalic process of male; prominent in robust
species (Fig. 305), carinate in flattened species (Figs. 307, 309).

b. Dorso-ventral flattening of antennal club (Fig. 136A) in all species
except ganglbaueri; in some (Fig. 136) flattening accompanied by elongation
of first lamella (segment 7).

c. Maximal elongation of mentum, approximately as long as wide in all
species (Fig. 152).

d. Progressive dorso-ventral flattening of prothorax, from robust (Figs.
167-168, 174) to successively more flattened (Figs. 170, 173, 175-176); flatten-
ing is accompanied by progressive narrowing and squaring of dorsal outline,
from rounded laterally (Figs. 169, 304) to more parallel-sided (Figs. 171, 177,
306, 308) and by reduction of pronotal processes (cf., Figs. 167, 176).

e. Narrowing of mesoscutellum; in all Dendropaemon length at least
equal to, but usually greater than width (Fig. 229); in other phanaeines,
width always greater (usually much greater) than length (Fig. 44).

f. Flattening and elongation of metepisternum; in other phanaeines and
in robust Dendropaemon, length (along dorsal margin) less than 1.5 times
height (along anterior margin) (Figs. 54, 241, 244); in flattened species,

846

The University Science Bulletin

306

308

Figs. 304-309. 304, Dendropaemon amyntas, male, dorsal view; 305, same, lateral view;
306, D. viridis, male, dorsal view; 307, same, lateral view; 308, D. nitidicollis, male, dorsal
view; 309, same, lateral view.

length progressively proportionately greater, but always more than twice
height (Figs. 232, 235, 238, 242-243).

g. Narrowing of metasternum; in robust species, length of its median
area (from antero-median angle to posterior margin) only slightly greater
(less than 1.25 times) than width (smallest distance between mesal margins

Evolution of the Phanaeine Dung Beetles 847

of coxal depressions) (Fig. 51); in narrowed species, length progressively
proportionately greater but always twice or more width (Figs. 231, 234, 237);
accompanying narrowing is progressive reduction in prominence of antero-
median angle of metasternum (Figs. 230, 233, 236).

h. Narrowing of elytra; in robust species, combined widths of elytra (at
humeral angles) greater than length (along elytral suture) (Fig. 304); in
narrowed species, length greater than width (Figs. 306, 308); narrowing
accompanied by proportionate narrowing of abdomen, which is covered by
elytra.

Primarily because of the great variation in general form it is difficult to
recognize Dendropaemon immediately using superficial "sight" criteria.
Most species are more-or-less flattened, elongate and parallel-sided; all have
a reduced number of tarsomeres (two or three). The extreme of the elon-
gation-flattening trend observable in smaller species of the genus is paralleled
in Onthocharis Westwood which superficially resembles small Dendro-
paemon also in color (metallic green) and clypeal emargination, and in
Stiptopodiits Har.

A meaningful evolutionary interpretation of the remarkable variation in
form of Dendropaemon is largely precluded by a lack of reliable data on its
biology, which likely will prove to be quite aberrant. The genus name
means "tree dweller," an appellation of doubtful accuracy. OlsoufiefT (1924:
121) and Blut (1939:263) state emphatically that Dendropaemon inhabit
rotting wood; Halffter and Matthews (1966:40) suggest they may feed on
fungi in rotting wood. Blut (p. 266) goes so far as to say that given species
of Dendropaemon feed selectively in given species of hardwood trees!
Evidently, however, the wood-dwelling habits of the genus have yet to be
positively demonstrated. All specimens about which I could learn circum-
stances of capture (from collector or label) were collected at lights or lying
dead on the ground. Moreover, neither I nor any other collector whom I
know personally has been able to collect adults from rotting logs, etc., in
localities where they are known to occur. To my knowledge, Dendropaemon
has likewise never been collected along with other phanaeines from carrion
or feces of any kind. That some (or all) species could be dendrophilous is,
of course, possible; but other, equally possible habitats have not yet been
adequately explored, such as vertebrate burrows and nests, nests of ground-
dwelling termites and ants, and fungi (particularly hymenomycetes).

Most museum specimens of Dendropaemon are quite old and, hence, apt
to bear non-precise locality data. Consequently, Map 6 is at best a tenuous
approximation of the range of the genus. The detailed distributional com-
ments of Blut (1939:265) need confirmation.

A detailed revision of Dendropaemon species is greatly needed. The

848

The University Science Bulletin

A Megatharsis
• Tetramerei

x'f.y.: Dendropaem

Map 6. Approximate distributions of Megatharsis, Tetramereia and Dendropaemon. (?'s
refer to occurrence of Dendropaemon.)

descriptions and keys of Olsoufieff (1924), Pessoa and Lane (1936) and Blut
(1939), which in general were based on too few specimens, need enlarge-
ment and clarification. Blut lists twenty-five species, fifteen of which I have
examined in addition to two others of uncertain name. Heretofore the genus
has been informally divided into two species groups based on number of
tarsal segments (two or three). In the following treatment tarsomere num-
ber is de-emphasized since other, more reliable variation quite clearly
indicates an altogether different systematic picture.

Evolution of the Phanaeine Dung Beetles 849

Key to the Subgenera and Species Groups of Dendropaemon

1. Abdominal sternum 7 long, as long as combined lengths of sterna 5

and 6 (Fig. 300); inner basal angle of front tibiae not expanded
as in Figure 247, tibiae gradually narrowing basally; hind tarsi two-
segmented, basal segment cylindrical, nearly as long as tibia, apical
segment minute, hemispherical (Fig. 278, arrow indicates apical

segment); color dull black-brown Paradendropaemon

-Abdominal sternum 7 no longer (usually shorter) than sternum 6
alone; front tibiae as in Figure 247; hind tarsi two- or three-seg-
mented, basal segment flattened, its length never approaching that
of tibia; color variable 2

2. Transverse propleural carina absent; eyes large (Figs. 132-133); tarsi

three-segmented (Fig. 274) Coprophanaeoides

-Transverse propleural carina present (Figs. 179-180, arrows); eyes
small (Fig. 134; large in denticollis); tarsi two- or three-segmented
Dendropaemon, sen. str 3

3. Transverse propleural carina complete, extending laterally beyond coxa

(Fig. 180); emarginations of anterior clypeal margin variable (Figs.
90, 92-93); tarsi three-segmented (Figs. 273, 275-276); hind wing

not notched (Fig. 299); length 8-15mm viridipennis group

-Transverse propleural carina incomplete, not extending laterally from
covn (Fig. 179); emarginations of clvpeal margin as in Figures 94
and 134; tarsi two- or three-segmented; hind wing notched or not 4

4. Middle and lateral emarginations of clypeus rounded, teeth rounded

(Fig. 94); tarsi three-segmented (Figs. 270-271; two-segmented in
smaragdinum , Fig. 272); hind wing at most only slightly notched
(Fig. 298); length 6-9mm; flattened, elongate, metallic green
species denticollis group

- Middle emargination of clypeus acute, lateral emarginations absent,
teeth acute (Fig. 134); tarsi two-segmented (Fig. 277), length of
basal segment more than three times length of apical segment; hind
wing deeply notched (Fig. 295); length 7-12mm; flattened, solid
black or metallic green species plceum group

Paradendropaemon new subgenus

Type: Dendropaemon ganglbaueri Felsche. (Derivation of subgeneric name: para [beside] +
Dendropaemon, neuter; reference to taxonomic isolation within genus.)

I have seen only one specimen of Dendropaemon ganglbaueri (a female),
but it is so unlike its congeners that I must conclude that it warrants place-
ment in a distinct subgenus. The following are those features which con-
tribute to its uniqueness: (1) very large, triangular clypeal teeth (Fig. 91),
(2) evenly tapering front tibiae (not strongly expanded basally as in other
Dendropaemon, Fig. 247), (3) middle and hind femora not expanded
apically as in other Dendropaemon (Figs. 268-269), (4) convex, evenly
granulo-rugose pronotum (very similar to that of Tetramereia and Homalo-
tarsus), (5) unique hind tarsi (Fig. 278), (6) exceptional length of ab-

850 The University Science Bulletin

dominal sternum 7 (Fig. 300), which is proportionately much greater than
in any other phanaeine, and (7) only very weak, scarcely visible lateral de-
pressions of base of the pygidium (distinctly strongly depressed in all other
Dendropaemon). It is, of course, possible that any of the above features may
be limited to the female. This robust species is known so far only from a
few localities in the state of Sao Paulo, Brazil.

Coprophanaeoides new subgenus

Type: Dendropaemon renatii OlsoufiefF. (Derivation of subgeneric name: Coprophanaeus +
-aides [resembling], neuter; reference to resemblance to Coprophanaeus, sen. sir.)

This subgenus brings together at least three species, the names of only
two of which I can use confidently: renatii Ols. and hirticollis Ols. I have
studied a third species under the name of bahianum Har., considered by
Blut (1939: 295-296, q.v.) as of incertae sedis because of the confusion attend-
ing its original description and subsequent usage. These three species bear
striking resemblance to Coprophanaeus, sen. str., particularly because of the
shape of the medial emargi nation of the clypeus (Fig. 132), large eyes (Figs.
132-133), ridged pronotal sculpturing and deeply emarginate hind wings.
Unlike any other phanaeine, the entire upper surface of hirticollis is covered
by scattered, erect setae. The distributional data that I have indicate that the
range of Coprophanaeoides is centered in southern Brazil.

Subgenus Dendropaemon Perty, sensu stricto

Dendropaemon Perty, 1830: 30. Tvpe: Dendropaemon piceum Perty, subsequent designation by

Blut, 1939: 267.
Enicotarsus Castlenau, 1831: 35. Type: Enicotarstis riridipennis Cast., by monotypy.
Onthoecus Dejean, 1837: 156. Type: Dendropaemon amyntas Har., present designation.

The bulk of the species studied fall into this subgenus, the most salient
feature of which is the presence of propleural carinae. The species included
here express the entire range of body shape variation described above, the
extremes of which are shown by D. amyntas (Figs. 304-305) and D. nitidi-
collis (Figs. 308-309).

The viridipennis group comprises four rather large (15-22mm) species
bearing complete transverse propleural carinae and three-segmented tarsi
(viridipennis [Cast.]; amyntas Har.; jractipes Felsche; lobatum Waterh.).
D. amyntas, a robust species, superficially resembles "normal" phanaeines,
and, besides ganglbaueri, it is the only Dendropaemon likely to be so con-
fused. D. lobatum is a subcylindrical species which, although parallel-sided,
presents cephalic and pronotal processes similar to those of amyntas. The
other species, while large, are distinctly flattened and parallel-sided. D.
viridipennis bears a striking resemblance to viridis in size, shape and color.

The denticollis group here includes five, very similar small (7-llmm)
metallic green species (denticollis Felsche, nitidicollis Ols., fasces Blut,

Evolution of the Phanaeine Dung Beetles 851

haroldi Ols., smaragdinum Waterhouse). All have distinctly rounded clypeal
teeth and emarginations (Fig. 94) and incomplete propleural carinae. D.
nitidicollis (Figs. 308-309) is the smallest known phanaeine (length 7mm).

The piceum group comprises three species with incomplete propleural
carinae, single emargination of clypeus with large, acute teeth (Figs. 90, 134)
and two-segmented tarsi (Fig. 277). The hind wing of these species is dis-
tinctly deeply notched (Fig. 295) (piceum Perty, telaphum Waterhouse,
viridis Perty).

Unassigned names probably Dendropaemon sen. str. (based on Blut,
1939) : redugens Waterhouse, angustipennis, Har., silvanum Blut, batrachites
Blut, similis Blut, subcylindricum Blut, temtitarsis Felsche, crenato striatum
Felsche, quadratum Cast.

Genus Tetramereia Klages

Tetramereia Klages, 1907: 141. Type: Tetramereia jredcrichji Klages = Dendropaemon con-

vextim Har., original designation.
Enrypodea Klages, 1906: 1. Type: Eurypodea jrederickj Klages = Tetramereia jrcderickji

Klages = Dendropaemon convexum Har., by monotypy.
Boucomontins OlsoufiefT, 1924: 120. Type: Dendropaemon convexum Har., by monotypy.

Tetramereia convexa (Har.) (Figs. 312-313) is a robust, brown species
with an involved nomenclatorial history during which it has been placed in
four genera (as type-species of three!). Klages (1906) proposed the name
Eurypodea frederic\i (n. gen., n. sp.) in a single page article printed at his
own expense. A series of errors resulted in the publication of another de-
scription of the same species proposing the name Tetramereia fredericlui
(Klages, 1907). Both descriptions were based on the same specimen. Olsou-
fiefT (1924: 120), suspecting the duplication, nevertheless proposed still an-
other genus name, Boucomontins , for Dendropaemon convexum Har., later
found to be a senior synonym of T. frederic\ii. I have elected for now to
continue usage of the name Tetramereia for two reasons: 1) I have judged
Eurypodea does not comply with Articles 7 and 8 (Chapter IV, Criteria of
Publication) of the International Code of Zoological Nomenclature and,
hence, is unavailable (Art. 10) (see also Janssens, 1940) ; 2) Tetramereia is
the name now of common usage. Nevertheless, I have requested that the
International Commission on Zoological Nomenclature render an opinion
on the availability of the name Eurypodea.

DESCRIPTION

(1) anterior margin of clypeus not emarginate medially, distinctly toothed
(Fig. 95); (2) clypeal process reduced, a small, very weak transverse ridge;
lateral clypeal carinae absent; (3-4) cephalic process of both sexes a simple
transverse carina; paraocular areas of parietals flat and (5) wide; (6) post-
ocular lobes of parietals depressed transversely, anterior portion of circum-

852 The University Science Bulletin

notal ridge not interrupted by emarginations; canthal area distinct; (7)
cephalic brachium monopodal; (8) eyes small; (9) occipital ridge incom-
plete; (10) labrum resembling Coprophanaeus-type (Figs. 142-143); (11)
mentum weakly emarginate (Fig. 153); (12) median lobe of hypopharynx
absent; (13) premental sclerites marginally sclerotized; (14-15) pronotum
of both sexes evenly convex except for small, antero-median transverse carina;
(16) pronotum evenly granulo-rugose; posterior pronotal fossae absent;
mesometasternal suture angulate (as in Fig. 240); mesoscutellum as long as
wide (as in Fig. 239); (17) front tarsi absent in female; (18) front tibiae
quadridentate, apical three teeth carinate (Fig. 249); (19) middle and hind
tarsi jour-segmented (Fig. 263) ; (20) elytral interstriae weakly punctured,
(21) evenly convex; (22) striae simple; hind wing not notched (Fig. 297);
last abdominal spiracle marginal; length 13-15 mm.

So far this species is known only from a few scattered South American
localities (Map 6) ; the type locality of frederic{ii is Suapure, Cuara (River)
Valley, Venezuela. Tetramereia convexa very closely resembles Homalo-
tarsus impressus (see discussion of Homalotarsus below). It is the only
known phanaeine with four-segmented tarsi, a feature which indicated to
Blut (1939) and Olsouneff (1924) an intermediate phylogenetic position of
the genus between Dendropaemon and Megatharsis (see phylogenetic dis-
cussion in the last part of this paper). Because of the lack of developed
cephalic and pronotal protuberances, the sexes of this species very closely
resemble each other. The biology of Tetramereia is unknown.

Genus Homalotarsus Janssens

Homalotarsus Janssens, 1932: 119. Type: Homalotarsus impressus Janssens (by original desig-
nation and monotypy).

To my knowledge only two specimens of this curious phanaeine are
known, both of which are deposited in the Institut Royal des Sciences
Naturelles de Belgique (Brussels). The holotype is labeled "South Amer-
ica"; the second specimen, "Brazil." I was able to examine only the latter
(sex unknown; Figs. 310-311); it bears Janssens' determination label and
agrees with his original description of the genus. Homalotarsus impressus is
remarkably similar to Tetramereia convexa, and the two should likely be
considered congeners. Nevertheless, until more is known about both species,
I choose to maintain the generic distinction.

I was unable to examine the mouthparts and endoskeletal features of H.
impressus; I have assumed that they are similar to their counterparts in
Tetramereia. Moreover, because of their great similarity, I have restricted
my description of Homalotarsus to the following observed features which
differ markedly from their counterparts in Tetramereia; those features not
mentioned are as in T. convexa (see above) :

Evolution of the Phanaeine Dung Beetles

853

312

313

315

Figs. 310-315. 310, Homalotarstis impressus, female?, dorsal view; 311, same, lateral view;
312, Tetramereia convexa, female, dorsal view; 313, same, lateral view; 314, Megatharsis
buckjeyi, female, dorsal view; 315, same, lateral view.

854 The University Science Bulletin

1. shape of mentum and labial palpi as in Figure 156;

2. shape of front tibia as in Figure 254, apical tooth not carinate;

3. middle and hind tarsi five-segmented (Figs. 264-265);

4. dorsal surface of apex of hind femur, of tibia and of first four tarso-
meres bearing large setigerous granules (Fig. 260);

5. color metallic yellow-green.

Genus Megatharsis Waterhouse

Megatharsis Waterhouse, 1891: 59. Type: Megatharsis buckleyi Waterhouse, by monotypy.

Although Waterhouse (1891) likened this species to Bolbites onitoides
Har., Megatharsis buc/(leyi (Figs. 314-315) is a unique phanaeine which, to
my eye, resembles Bolbites very little if at all. I have examined only two
specimens (females), both collected in Macas, Ecuador; the type locality is
nearby Chiquinda, Ecuador (Map 6). The following description is based on
females although I fully expect the male to agree closely with it; Waterhouse
failed to mention the sex of the holotype.

description
(1) anterior margin of clypeus not emarginate medially, obtusely biden-
tate (Figs. 96, 135); (2) clypeal process reduced to small gibbosity; lateral
clypeal carinae present; (4) cephalic process of female (and male?) a trans-
verse, trituberculate carina (Fig. 135); paraocular areas of parietals flat, (5)
wide; (6) postocular lobes of parietals depressed transversely, anterior por-
tion of circumnotal ridge not emarginate; canthal area distinct, small; (7)
cephalic brachium monopodal; (8) eyes small; (9) occipital ridge incom-
plete; (10) labrum resembling Coprophanaeus-type (Figs. 140-141); (11)
mentum weakly emarginate (Fig. 154); (12) median lobe of hypopharynx
absent; labial palpi of one segment (Fig. 154) with membranous slit on
dorsal surface (Fig. 155, arrow); (13) premental sclerites marginally sclero-
tized; (15) pronotum of female (and male?) evenly convex, (16) evenly
granulo-rugose and margined posteriorly by fringe of long setae behind a
sharp, transverse carina (Fig. 178); posterior pronotal fossae absent; meso-
metasternal suture angulate medially (as in Fig. 240) ; (17) front tarsi absent
in female; (18) front tibiae quadridentate (Fig. 248); (19) middle and hind
tarsi five-segmented, segments of characteristic shape (Figs. 266-267) ; dorsal
surface of hind tibiae with dense, longitudinal fringe of hair (Fig. 266) ;
(20) elytral interstriae punctate, (21) weakly convex, nearly flat; (22) striae
simple; hind wing not notched (Fig. 296); last abdominal spiracle marginal;
abdominal sterna 3-7 each produced as small acute tooth projecting from
beneath elytral margin; abdominal sterna 3-7 with dense fringes of long
hair, completely traversing sternum 7 (Fig. 301) while restricted laterally on
sterna 3-6; length 10mm; color dull blue or green.

Evolution of the Phanaeine Dung Beetles 855

PART VI: EVOLUTION AND PHYLOGENY OF PHANAEINES

Fossil phanaeines are unknown. Frenguelli (1938, 1939) attributed to
Phanaeus ("o un genero afin") brood balls recovered from various Tertiary
deposits (Oligocene through Pleistocene) in southern Argentina.12 Pierce
(1946), describing Palaeocopris from the La Brea tar pits (Los Angeles,
California; Pleistocene), suggests a morphological resemblance to this genus
with Phanaeus; I have not seen the type material of Palaeocopris.

Lacking meaningful paleontological data, systematists concerned with
evolutionary and phylogenetic questions are forced to base conclusions upon
neontological data. The purpose of this discussion is to consider (1) the
phylogenetic (cladistic) relationships among phanaeines indicated by an
evolutionary interpretation of the comparative skeletal morphology and the
biology of contemporary forms, and (2) the evolutionary implications of
the ecology and distribution extant phanaeines.

Some Preliminary Considerations

Data for this discussion are essentially the same available for any
approach to inferences about relationships, whether intended to be strictly
phenetic or strictly cladistic. Since the data are the same, the conclusions of
phenetic vs. phylogenetic inferences about relationships differ only as a result
of the way they are treated. The evolutionary and phylogenetic conclusions
arrived at here are based upon phenetic relationships indicated primarily by
morphological characters and upon presumed cladistic relationships inferred
by stressing (weighting) certain characters and by making judgments about
evolutionary direction.

Phylogenetic speculation requires some a priori notions about evolution-
ary direction; that is, it requires decisions about which state of a character
employed in assessing relationships is "primitive" (has undergone little or
no evolution since the origin of a group; = plesiomorphous character of
Hennig, 1965, 1966) and which is (are) "derived" (= "modified" = "special-
ized"; apomorphous character of Hennig). Table 1 lists for phanaeines the
presumed primitive and derived states for some of the characters used in this
study which lend themselves to such judgment. Table 1 is, in effect, a list of
evolutionary trends observable in phanaeines. It is to be noted, however,
that while certain character states are designated as primitive for phanaeines
(e.g., striking sexual dimorphism, metallic color), they may be specialized
ones when considering Scarabaeinae as a whole. In these instances, the
derived state in phanaeines can be a secondary (convergent) condition and
phenetic equivalent of the primitive state for the entire subfamily. An im-

12Gonza!o Halffter (pers. comm.) has examined the brood balls used in Frenguelli's studies
and concludes that they are essentially of the type presumed characteristic of phanaeines and
isolated species of Duhotomius (see Halfiter and Matthews, 1966:130-139).

856

The University Science Bulletin

Table 1. Primitive and derived features of phanaeines.

Character*

Primitive State

Derived State (s)

1. Shape of anterior margin of

Not emarginate

Emarginate medially

clypeus

medially

2. Shape of clypeal process

Transverse

Spiniform, U-shaped or reduced

A Lateral clypeal carinae

Present

Absent

3. Shape of cephalic process of

Corniform

Lamellate, carinate

male

4. Shape of cephalic process of

Carinate

Corniform

female

5. Width of paraocular areas

Narrow

Wide

6. Shape of postocular lobes of

Depressed trans-

Depressed obliquely —

parietals — anterior portion of

versely —

emarginate

circumnotal ridge

not emarginate

7. Shape of cephalic brachium

Monopodal

Bipodal

8. Size of eyes

Small

Large

9. Occipital ridge

Incomplete

Complete

10. Type of labrum

Phanaeiis-typc

Copro p/uinaeus-iy pe

1 1 . Shape of mentum

Strongly emarginate

Weakly emarginate

12. Median lobe of hypopharynx

Present

Absent

13. Premental sclerites

Marginally sclerotized

Completely sclerotized

16. Pronotal sculpturing

Smooth-punctate

Roughened

1 7. Front tarsi of female

Present

Absent

18. Front tibiae

P/ianaeus-type

Copro p/ianaeas-type
Dendropaem o«-type

19. Middle and hind tarsi

Five segmented

With fewer than five segments

22. Striae

Simple

Carinulate

B Hind wing

Not notched basally

Notched basally

C Position of last abdominal

Marginal

Intermediate, intrasternal

spiracle

D Color

Bright (metallic)

Sombre

E Size

Length under 30 mm

Length over 30 mm

F Secondary sexual dimorphism**

Striking (sexes very
different)

Subtle (sexes similar)

G Feeding habit

Coprophagous

Necrophagous

H Time of activity

Diurnal

Nocturnal

Numbering of characters corresponds to that in part III: reference should be made to this
part for complete definition of morphological character states and their taxonomic dis-
tribution. Those characters designated A-H are characters either discussed, but not num-
bered, in part III or not discussed in that part.
Based principally upon cephalic and pronotal shapes of well developed individuals.

portant group of characters is not included in Table 1 : those which, because
of their restricted taxonomic occurrence, are considered phanaeine special-
izations (e.g., transverse clypeal carina of Diabroctis, the several unique
features of Detidropaemon and its subtaxa, etc.). These specializations, which
are obviously most useful in "keying" the taxa to which they are restricted,
are taken into account in the following discussions.

The direction of evolution is sometimes intuitively self-evident. Thus, a
tarsal complement of fewer than five segments in phanaeines (and other

Evolution of the Phanaeine Dung Beetles 857

dung beetles) can (to me) be construed only as arising from the loss (or
fusion) of one or more segments. In many instances, however, logical de-
cisions depend largely upon a knowledge of variation in non-phanaeine
taxa; for this reason it was necessary to examine other, non-phanaeine dung
beetles (see Appendix for a list of those examined). When highly subjective
judgments were called for in Table 1, the following points were considered:
Since evolution as it is now understood tends to be a conservative process,
a character state of widespread occurrence among otherwise quite diverse
taxa is probably a primitive one. Accordingly, in character 7 the designation
of a monopodal brachium as primitive was based principally upon the
presence of monopodal brachia in all non-phanaeine scarabaeines examined
as well as in the majority of phanaeines. Furthermore, the evolutionary
status of the morphological character states regarded as correlates of be-
havorial trends are most logically considered the same as the status of the
behavioral phenomena in question. Thus, if necrophagy in phanaeines is a
specialized behavioral pattern, character states regarded as morphological
correlates of necrophagy (e.g., emarginate clypeus) are no doubt also derived.
In making phylogenetic inferences I have stressed a fundamental hy-
pothesis of "phylogenetic systematics" as most comprehensively approached
by Hennig (1965, 1966): that, when practical, taxa should be recognized on
the basis of shared derived character states, i.e., by synapomorphy or joint
possession of apomorphous features (Hennig), since to do so increases the
probability that taxa recognized are monophyletic. I do not, however, con-
sider myself a phylogenetic systematist in the rigid sense of Hennig since I
have chosen to temper the process of classifying and subsequent ranking of
taxa with some measure of "taxonomic practicality." A "true" phylogenetic
systematist (if such a person exists) might object to the phrase, "when prac-
tical," above. Thus, the classification of phanaeines proposed above does not
completely reflect my ideas about the phylogenetic relationships among
phanaeines, which are discussed below.

Phylogenetic Speculation
Table 2 shows the taxonomic distribution of the character states of the
26 characters listed in Table 1. For each taxon listed in Table 2 the per cent
frequency of derived states (%D) is given at the right.13 Using these per-
centage figures it is possible to generalize about the relative overall "primi-
tiveness" (or "derivedness") of the various taxa based on those 26 characters.
Thus, taxa 3 and 4, the auricollis and imperator groups of Sulcophanaeus,

13 In calculating %D, "O" entries were counted twice, once as primitive, once as derived;
"X" entries were disregarded. Ordinarily %D = no. D's/26; but for, say, Oxysternon, sen. str.,

no. dots -j- no. O's

%D = X 100%.

26 + no. O's — no. X's

858

The University Science Bulletin

V 4-j

V u

-o c

G

'rt

CO

N

CO

4— »

G

4-J

-a

CO

o

HJ

3

_C

w

+-<

G

4-J
CO

s

~

c

O

O

<-W

-o

bfl

^-^

C
O

0 I

a;

r-

4— »

Ll

rt

ai

1)

oo

CO

-a

4— '

cs

4— >

<u

.•*

CO

—

IT,

a;

T3

CO

~!

IU

«;

II

g;

TD

c

J— 1

O

T3

c

G

k.

-o

G*

-o

aj

II

o

4—1

on

o

£

_Q

00

Ul

s

Uh

a-

?^

<d

CO

4^

M-l

u

O

;>>

u

G

O
G

IT

u

«j

n

vo

3

rt

H 9>

4-»

a:
G

u

4->
CO

<u

G

_G

4-J

II

i-
o

o

o

G

G

^ — s,

-

o

t-l

G

—

IU

-^

3

(J

_Q

a!

o

!-.

rt

G

—

_G

II

U

T3

u

X

a

o

G
O

00

G

3

X

—

«

00

H

4-1

«

4-1

rs

00

w
m

<u
>

• «— <

<

Ui

H

Evolution of the Phanaeine Dung Beetles 859

emerge as most primitive (%D = 4), hence most obvious choices as the
closest phenetic (and phyletic) approaches to "palaeophanaeines," while
Dendropaemon (average %D = 58) appears to be the most highly evolved,
hence most different from some hypothetical common ancestor.

Table 2 is somewhat misleading since some groups (particularly taxa
12-17) can actually be considered more highly evolved relative to the others
than indicated by %D's. This is so since most of those character states unique
to these taxa and, therefore, considered derived, are not enumerated per se
in Tables 1 and 2. The incidence of such unique characters is especially high
in Dendropaemon (e.g., body shape variation, lateral depression of base of
pygidium, forked antero-median process of labral suspensorium, reduction
in number of setae comprising median brush, etc.; see systematic section).
The incorporation of these characters into Table 2 in the case of Dendro-
paemon and other taxa would only intensify the phenetic gaps indicated
there; the relative positions of all taxa on an "evolutionary scale" would not
change. Incidence of "concealed synapomorphy" will be discussed more
fully below.

Of the 26 characters listed in Table 2, I have judged some intrinsically
more useful than others for purposes of deducing phylogenetic relationships.
Those presumed high in phyletic information content are shaded in Table
2. Those characters not shaded (omitted from phylogenetic considerations)
form two principal groups: (1) some characters (12, 17, 19, A) were ex-
cluded because they are regressive, that is, their evolution involves loss or
"reduction" (character 2, which is partially regressive, was not altogether
ignored). I have accepted the hypothesis that regressive features (character
states) are inherently more susceptible than progressive ones to parallel or
convergent development (see, for example, Mayr, 1969: 222-223). Summary
exclusion of regressive characters has probably resulted here in little loss of
phyletic information since their inclusion would not significantly alter my
phylogenetic inferences about phanaeines. (See discussion of parallelisms
below.) (2) Characters D-H were excluded for discretionary reasons. Sepa-
ration of largely discrete states of characters D, E and F is highly subjective;
the designation of primitive states of characters G and H is based on unequal
knowledge (e.g., the feeding habits of relatively very few Sitlcophanaeus are
known, but these are coprophagous; on the other hand, relatively a large
number of Phanaens are known to be coprophagous; as indicated by avail-
able, but unequally representative data, I assume both groups to be basically
coprophagous) .

Taxonomic data on the phylogenetic (shaded) characters in Table 2 have
been reorganized in Table 3 as a phyletic "affinity matrix" showing the
frequency of shared derived character states (= frequency of synapomorphy)
between all possible pairs of taxa. One can observe directly in Table 3 pat-

860

The University Science Bulletin

Table 3. Affinity matrix (taxon X taxon) based on frequency of shared derived
character states (synapomorphy) among pairs of phanaeine taxa. Taxa are

numbered as in Table 2.

Taxa

1

2
3
4

5
6

9
10
11

12
13

14

15
16
17

— Diabroctis
0 ,

1 - Sulcophanaeus

0 0-

0 1 10-

Oxysternon

12 10 13-

0 1 10 3 3 1-

Phanaeus

0 2 10 3 3 14-

0 3 0 0 0 10 1
0 10 0 0 0 0 1
10 0 0 10 1

6 — Coprophanaeus

7 6-

10 0 0 10 16 5 7
10 0 0 10 15 5 6
10 0 0 10 17 5 8

7 - Dendropaemon
10 7-

10001012 2 3 4 4 4 - Megatharsis

100010 12 2 3 4 4 4 4 - Tetramereia

10 001012 2 3 4 4 4 4 4 - Homalotarstts

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Taxa

terns of high and low frequency of synapomorphy between, among and
within genera. All taxa (1-17) are presumed monophyletic, the constituent
species (one or more) of which are themselves related by synapomorphous
states described fully only in the classification proposed above. Open triangles
along the diagonal in Table 3 embrace intrageneric comparisons for those
genera with two or more subgeneric groupings.

If, as is assumed, frequency of synapomorphy is a valid measure of in-
ferred phyletic (cladistic) relationship, one should expect the relative fre-

Evolution of the Phanaeine Dung Beetles 861

quency of synapomorphy to complement the classification of the groups for
which comparisons are made. That is, in the case of phanaeine taxa, one
should expect (where applicable) the /«/rageneric frequency of synapomor-
phy to be greater than intergeneric frequency, and that groups thus defined
should be assignable to hierarchical groups largely corresponding with those
of the classification proposed above. With some significant exceptions to be
discussed below, this seems to be the general case. Inspection of Table 3
reveals two rather distinct phyletic generic groupings and two enigmatic
groupings together with some inferences about evolutionary rates. Phylo-
genetic "conclusions" are presented below under two headings, generic
affinities and parallelisms, and summarized in Figures 316 and 317.

Generic Affinities

Genus-group 1: Phanaeus-Oxysternon. A relatively close phyletic re-
lationship between these genera is characterized by the presumed evolution-
ary fixation of three derived features, the first two being unique to the genus
group: a complete occipital ridge, completely sclerotized premental sclerites
and bipedal cephalic brachia (seen otherwise only in some Sulcophanaeus).
As indicated by %D in Table 2, Oxysternon (particularly Mioxysternon,
%D = 28) is the more highly evolved of the genera. Tables 2 and 3, how-
ever, do not account for synapomorphous states unique to these genera, most
notable of which are the spinate metasternum and acute postero-median
pronotal angle of Oxysternon and the head-prothorax interlocking mecha-
nism of Phanaeus.

The relatively low frequencies of synapomorphy (0-1) between this genus
group and other taxa indicate phyletic isolation; that is, they support the
hypothesis that this genus group can be regarded as a distinct evolutionary
lineage. On the other hand, the relatively low incidence of synapomorphic
states (3) characterizing the group (i.e., relative overall primitiveness) sug-
gests a relatively slow evolutionary rate (compared to, say, genus group 2).

A rather long phyletic history is indicated by the taxonomic pattern, i.e.,
the distinctness of taxonomic gaps. Thus, intermediate types minimizing
gaps between genera, among subgenera and species-groups of Oxysternon
and among the species complexes of the Phanaeus splendid ulus group are
virtually non-existent. Conversely, the taxonomic difficulties arising because
of the phenetic overlap among constituent subgroups of the P. vindex group
suggests the hypothesis that this species group, (1) underwent relatively
recent (and rapid ?) differentiation and/or (2) its differentiation, whether
recent or remote, has been characterized by a comparatively low incidence
of extinction. The geographic distribution of the vindex group (see below)
supports the first alternative, that this species group underwent extensive

862 The University Science Bulletin

radiation in an initially restricted, though highly partitioned geographic
theater (Mexico).

Genus-group 2: Coprophanaeus-Dendropaemon. Relative to other
phanaeines, these genera are highly evolved. Indeed, the extreme phenetic
divergence of Dendropaemon makes it very difficult to conclude intuitively
that there exists a close phyletic relationship between such different taxa as
C. (Megaphanaens) lancijer (Figs. 181-182) and D. (Dendropaemon) nitidi-
collis (Figs. 308-309) ! However, a close phyletic affinity between these
genera is evidenced by a high frequency of synapomorphy (see Table 3) :
emarginate clypeus and U-shaped clypeal process, Coprophanaeus-type la-
brum, weakly emarginate mentum, and roughened pronotal sculpturing.
Moreover, several other tendencies tend to reinforce the relationship (al-
though they are not all considered in Table 3) : usual absence of front tarsi
in females, similarly developed front tibiae, tendency for secondary sexual
dimorphism to be less pronounced than in other phanaeines, specialized
feeding habits (assumed of Dendropaemon), common occurrence of pro-
notal ridging, as well as the conservation of a number of primitive aspects
of head structure in otherwise quite highly evolved groups (monopodal
brachia, persistence of distinct canthal areas, transverse depression of post-
ocular lobes, incomplete occipital ridge, etc.).

On the basis of the high degree of synapomorphy between Copro-
phanaeus, sen. str., and Coprophanaeoides, I have concluded that the latter
taxon is an annectant between the two genera. Moreover, it appears that
Coprophanaeus, sen. str., is more closely related phyletically to Copro-
phanaeoides than to either Megaphanaeus or Metallophanaeus. That Copro-
phanaeoides and Dendropaemon, sen. str., are undeniably congeneric is
supported by their sharing of a number of uniquely dendropaemine features
(see above). Consequently, I must conclude that Coprophanaeus is a para-
phyletic taxon (sensu Henning, 1966) and not a monophyletic one (Fig.
317).

A relative long history of the Coprophanaeus-Dendropaemon lineage is
indicated by the phenetic isolation of infrageneric taxa as well as the occur-
rence of "problematic" species. The origin of these taxonomic isolates is
more logically interpreted as relict survival of more primitive types; these
isolates include C. (Megaphanaeus) bellicosus, Metallophanaeus (which is
itself likely not monophyletic; see systematic section) and Paradendro-
paemon. The latter is most problematic since, because it lacks many of the
salient features of Dendropaemon, it cannot be accommodated easily by a
strict taxonomic delimitation of "Dendropaemon." Indeed, the superficial
appearance of D. ganglbaueri is more like Tetramereia than any of its con-
geners. Were it not for its possession of unmistakably dendropaemine
mouthparts, I would be inclined to rank Paradendropaemon as a genus.

Evolution of the Phanaeine Dung Beetles

863

phyletic

distance ■

Fig. 316. Phylogram of the inferred phyletic relationships among species included in
Sidcophanaeus.

Moreover, the subgeneric rank of Paradendropaemon imposes the possibility
that Dendropaemon , like Coprophanaeus, is paraphyletic.

Taxa of uncertain affinities: The phyletic disposition of other phanaeine
taxa is not as easily accomplished as that for the aforementioned genus-
groups. Relatively speaking, Diabroctis and Sidcophanaeus are primitive
groups most logically interpreted as comprising relict elements. (The dis-
proportionately high %D of 28 of the S. faunas group is attributable to the
differential possession of derived features [O's in Table 2] by the two con-
stituent species-pairs, and not to apomorphous states shared by all four spe-
cies; consequently, the %D of any given species in the group is less than 28).
There are no derived states shared by all Sidcophanaeus; and, as measured
by those characters considered in Tables 2 and 3, all species-groups are vir-
tually isolated phenetically and, presumably, phyletically. Moreover, the
intragroup variation is such that the phenetic gaps among most constituent
species of the species groups of Sidcophanaeus are themselves quite distinct.
I have concluded, therefore, that Sidcophanaeus is very likely a paraphyletic

864 The University Science Bulletin

(or polyphyletic) assemblage which brings together a number of phyletic
isolates representing the remnants of numerous older, largely primitive
phyletic elements (Fig. 316). This conclusion is supported by the known
distribution of the genus (see below). What has been said of Sulcophanaens
is also true of Diabroctis, which is evidently also a small relict group of
phyletic rank equal to that of the species-groups of Sidcophanaeus.

A somewhat similar situation exists with Tetramereia, Homalotarsits and
Megatharsis, all of which are monobasic. For reasons discussed in the sys-
tematic section, the two former genera should probably be merged. As a
group, these species appear no more closely related to each other than either
is to Dendropaemon, which they resemble only slightly more than Copro-
phanaens (Table 3). In view of their extreme isolation within the phanaeines
and of their often curious, unique specializations (e.g., one-segmented labial
palpus of Megatharsis, unique shape of the tarsomeres of all three species,
etc.), I have concluded that they represent highly divergent relicts of doubt-
ful affinities with the Coprophanaeus-Dendropaemon lineage. A relictual
status of Megatharsis and Tetramereia is also tentatively suggested by their
distributions.

Summary of phylogenetic speculation (Fig. 317): Phyletically, the
phanaeine Scarabaeinae, which are presumed to comprise a monophyletic
group, appear to consist of two distinct phyletic lineages, together comprising
the bulk of known species, as well as a number of phyletically isolated, small
groups representing relicts of several older elements. Such phyletic (phe-
netic) fragmentation is at least compatible with the hypothesis that the
phanaeines comprise an ancient group of origin and long-standing residence
in South America.

Parallelisms

As can be seen in Table 2, for example, given apomorphous character
states are not necessarily restricted to presumably monophyletic taxa. In these
instances one must ordinarily assume that the derived state arose inde-
pendently in two or more phyletic lines. The origin of the similarity of two
(or more) taxa resulting from joint possession of the same, independently
acquired derived state (s) can be interpreted as the result of either parallel
evolution or convergent evolution. Differences between these concepts are,
to me, more in degree (taxonomic frame of reference) rather than in sub-
stance since overemphasis of either kind of similarity will result in the recog-
nition of polyphyletic groups. For present purposes, the presumed inde-
pendent expressions of a given apomorph among phanaeine taxa are here
cited as parallelisms while those few cited instances of synapomorphy be-
tween phanaeines and non-phanaeines are considered convergences.

Since I am not familiar with a large enough taxonomic variety of

Evolution of the Phanaeine Dung Beetles

865

U

c

o

c

TO

—

n
u

•s

i!

c

oo
.C

°C
u

£

3

X

a

o

a

'v
a
c
a

c

-j

'G

3

•Si

bo •
C 2

o g

.£* a
IS «

c

Q

-a
.s c

■2 .S

u
a. j«

.9 -a

u

"3 XI

O 3
_ C

2 S

CO

S'-S

ro

o -a
- «j
b -a

CO

866 The University Science Bulletin

Scarabaeinae to compile a comprehensive list, I have limited a discussion of
convergent synapomorphy to mention of the following few instances:

a. transverse clypeal carina: Diabroctis and Onthophagits.

b. apical hind tibial carina: various phanaeines (see below) and "Co-
prina" (Copris and related genera).

c. lack of tarsal claws: all phanaeines, Eucranini and Onitini.

d. lack of front tarsi in males: all phanaeines, some Onitini, some Can-
thonia.

e. reduction in tarsomere number: Dendropaemon, Heteroditopus (=
Pinacotarsus) .

f. elongate, flattened body: some Dendropaemon, Onthochans, Stipto-
podius.

g. necrophagous habit: Coprophanaeus and other species (see below),
Deltochilum, some Dichotomius, some Cant/ion.

h. P ha nae its-type antennal club: all phanaeines, Sceliages.

Instances of presumed parallelisms among phanaeines are quite numer-
ous. While the following list is not exhaustive, it will serve to illustrate the
variety of characters affected and taxa involved. The name of each derived
state is followed by those phanaeine taxa within each of which the apomorph
is assumed to have arisen.

a. presence of transverse hind tibial carina: Oxystemon jestivum group;
Diabroctis; Sidcophanaeits carnijex-jaunus pair; S. imperator-steinheili

pair.

b. large size (length greater than 30mm): Diabroctis; Megaphanaeus;
Sulcophanaeus f annus group.

c. reduction of tarsomere number: Dendropaemon; Tetramereia; (?
Paradendropaem on ) .

d. bipodal cephalic brachium: Phanaeus-Oxysternon genus-group; Sidco-
phanaeits imperator; S. laeander-cohtmbi pair.

e. corniform female cephalic process: Megaphanaeus; Sidcophanaeits
faunits group.

f. intrasternal position of last abdominal spiracle: Megaphanaeus; Sulco-
phanaeus faunus-carnifex pair; Mioxysternon.

g. loss of front tarsi in female: Metallophanaeus; Coprophanaeus, sen.
str.; Tetramereia-Homalotarsus; Sidcophanaeits carnifex; S. steinheili.

h. flat elytral interstriae: Sidcophanaeits auricollis group; Megaphanaeus;

Phanaeus damocles-quadridens; P. vindex-difformis.
i. rugose elytral interstriae: Megaphanaeus; Phanaeus vindex complex,
j. carinulate striae: Megaphanaeus; Dendropaemon; Phanaeus bis pin us;

P. \irbyi-palaeno pair,
k. notching of hind wing: Coprophanaeus, sen. str. -Dendropaemon;

Phanaeus bispinus.

Evolution of the Phanaeine Dung Beetles 867

1. roughened pronotum: Coprophaneus-Dendropaemon genus group;

Phanaeus vtndex group; Sulcophanaeus laeander-columbi pair; Tetra-

m ereia-Ho m alo tar sits ( -Megath arsis ) .
m. spiniform clypeal process: Oxystemon festivum group; Phanaeus

kjrbyi-palaeno pair,
n. necrophagous habit: Coprophanaeus; isolated species of Phanaeus

(e.g., bispinus, meleagris); isolated species of Sulcophanaeus (e.g.,

faun us).

Evolutionary Implications of Phanaeine Ecology and Distribution

Halftter and Matthews (1966) discuss quite fully various aspects of the
ecology and behavior of Scarabaeinae. Rather than redevelop their discussion
as it applies to phanaeines, comment here will be limited to the following
summary statements:

1. The principal food sources of phanaeines are animal feces (particu-
larly the dung of large herbivores such as cattle) and carrion.

2. Necrophagy as such is likely a behavioral adaptation for forest ex-
istence and a derived behavioral pattern in Scarabaeinae (see Halfifter and
Matthews for a detailed argument). Dung beetles in South American
forests are very often necrophagous; conversely, few necrophagous species
inhabit grasslands. This behavorial pattern usually cuts across established
taxonomic lines, such that very few taxa can be delimited by necrophagous
feeding behavior as well as by morphology. Coprophanaeus is one of these
few. Ecological experiments designed to test the relationship between nec-
rophagy and forest habitat have not been done. Among other things, these
might investigate (a) the apparent scarcity of carrion in open grassland
compared to a forest, (b) rates of desiccation of carrion in a forest vs. grass-
land, and (c) the frequency and severity of competition between necro-
phagous animals in a forest vs. in a grassland (e.g., between beetles and
vultures in grasslands).

3. Within the phanaeines the taxonomic distribution of principally
diurnal-coprophagous-grassland species and nocturnal-necrophagous-forest
species is largely generic. Thus, most Phanaeus are coprophagous-grassland
forms while Coprophanaeus consists entirely (as now known) of necro-
phagous-tropical forest species. These two genera include about 60% of all
phanaeines. Diabroctis, Sulcophanaeus and Oxystemon are poorly known
ecologically, but all appear to be largely coprophagous groups with both
grassland and forest species. Nothing is known about the biologies of
Dendropaemon, Tetramereia, Homalotarsus and Megatharsis.

Phanaeines are restricted to the New World, occupying South America
(except Chile and southern Argentina), Central America, Mexico and east-
ern and extreme southwestern portions of the United States (Maps 1-6). The

868 The University Science Bulletin

Table 4. Summary of the distribution and relative abundance of phanaeines.*

Region

Taxon

1

2

3

4

5

Diabroctis

(+)

—

—

—

—

Sulcophanaeus

( + + + + )

( + )

+ ( + )

+

—

Oxysternon

( + + + + + )

—

—

—

—

Phanaeus

spletididulus complex

( + + )

(++)

(1 sP)

(2 sps)

—

vindex complex

—

—

(+) +

( + + + + ) + +

+<+)

Coprophanaens

Megaphanaeus

(+)

—

—

—

—

Metallophanaeus

(+)

—

—

—

—

Coprophanacus, s.s.

( + + + + + ) +

1 sp

+

+

—

Dendropaemon

(+++++)

—

—

—

—

Megatharsis

(1 sp)

—

—

—

—

Tetramereia

(lsp)

—

—

—

—

Homalotarsus

(1 sp)

—

—

—

—

* See text for explanation.

only insular species is Sulcophanaeus carnijex, which is endemic to Jamaica.
Table 4 summarizes the distribution according to the following numbered
regions :

1. Tropical South America (except northwestern sector and areas west of
the Andes)

2. Temperate South America

3. Northwestern South America (including areas to the Pacific side of
the Andes) and Panama

4. Central America (except Panama) and tropical Mexico

5. Temperate Mexico and the United States.

In Table 4 presence in one or more regions is denoted by one or more + 's,
or by the number of species if very few; — denotes absent. The number of
+'s indicates relative abundance in a given region; parentheses enclose that
portion endemic to the region.

The tropical South American fauna includes representatives of all
phanaeine genera, eight of which are endemic or largely so. Such a high
degree of generic endemism strongly suggests South America as the place of
origin of the group. Differentiation of all genera (as presumably could also
be the case for other South American scarabaeine groups) was probably
before or during Tertiary isolation of South America. If during Tertiary
isolation, the greatest development of phanaeines could have coincided with
that of South American herbivores (e.g., the ungulates). During periods of

Evolution of the Phanaeine Dung Beetles 869

isolation inter-American water gaps presumably made difficult northward
dispersal of both dung beetles and large land animals with which they
were associated. If the initial development of scarabaeine groups in all geo-
graphic areas were closely tied to the presence of a large food supply pro-
vided primarily by the dung of large animals, the extinction from South
America of grassland herbivores must have greatly affected the evolution of
the scarabaeine fauna. Necrophagous, forest dwelling groups, such as Copro-
phanaeus, could have resulted from movement of grassland groups into
forest habitats in response to extinction of herbivores. The coprophagous
groups which survived were probably largely restricted to small, geograph-
ically isolated refugia where they differentiated as small, phenetically quite
dissimilar groups.

The disjunct patterns of distribution within the faunas and imperator
groups of Sulcophanaeus are indicative of relict groups and, hence, com-
patible with the hypothesis presented above that the genus comprises the
relicts of several older phyletic elements. The following disjunct patterns
between closely related species are not evident in Map 1:

a. Sulcophanaeus f annus, Amazon Basin — S. carnife.x, Jamaica.

b. S. laeander, Venezuela (savannas?) — S. columbi, southern Amazon
Basin.

c. S. imperator, Andean highlands of northern Argentina and of Bolivia —

S. steinheili, high mountain valleys of Colombia.
The distribution of the auricollis group (Map 1) is evidently continuous.
This appears to be the only sulcophanaeine group that has undergone even
limited northward dispersal and differentiation. The close resemblance of
the Andean species auricollis and the Mexican chryseicollis and the occur-
rence of two species (noctis and velutinus) in both South and Central
America suggest that the movement was relatively recent, perhaps accom-
panying the equally limited northward dispersal of Coprophanaeus, sen. str.
(Map 5) and the Phanaeus splendidulus group (Map 2) (see below).

While the relict status of Megatharsis and Tetramereia (and Homalo-
tarsus) is supported by the extreme phenetic and geographic isolation of
these species (Map 6), both are very poorly represented in collections and
may later prove to be more widely distributed (in the case of Megatharsis)
or more evenly distributed (Tetramereia). Likewise, the distribution of
Dendropaemon is poorly known and, therefore, not amenable to serious
geographic analysis.

Within other genera distributional patterns appear more regular, i.e.,
with closely related species occupying largely adjacent regions. There is a
tendency, however, for phenetic isolation to be coincident with geographic
isolation (e.g., in the cases of Coprophanaeus pertyi and pessoai, C. belli-
cosus, and the species of Mioxysternon). The northward dispersal and

870 The University Science Bulletin

accompanying linear differentiation (speciation) of the S. auricollis group
is paralleled in Coprophanaeus, sen. str., by the dardanus group (C. telamon
extending from South America into Central America and southern Mexico;
C. pluto endemic to Mexico) and in the Phanaeus splendid iilus group by the
endymion complex (P. blanchardi in northwest South America, pyrois in
Central America and endymion in Central America and southern Mexico).

The Phanaeus find ex group is the only phanaeine taxon centered geo-
graphically outside South America. This large species-group is centered in
Central America and adjacent southern Mexico, where it almost certainly
originated and underwent extensive, probably rapid radiation (as indicated
by the narrow, taxonomically inconclusive phenetic gaps between clusters of
closely related species). From Mexico the group dispersed both northward
and southward, differentiating further in both directions, although most
clusters remained in or near the locus of radiation. Northward movement
of Mexican species undoubtedly gave rise to a temperate fauna (the vmdex
complex) largely endemic to the United States. Although the possibility
needs to be examined more closely, the distributions of U.S. species indicate
that their history includes disruptions by Pleistocene glaciation (suggested by
my somewhat premature suspicion that P. floridanus and isolated popula-
tions of P. vindex in Arizona are products of differentiation in refugia from
glaciation). The hermes cluster, conversely, likely arose from South Amer-
ican immigrants moving southward from Mexico. Apparently the only
group to disperse successfully in both directions was the mexicanus cluster.
The combined range of the constituent species of this cluster extends from
Peru (lunaris) through Central America (ivagneri), southern Mexico (mexi-
canus), and along the western coast of Mexico into Arizona (amythaon).
The ecological-geographical distribution of Phanaeus in Mexico is quite
complex and in need of continued analysis.

In trying to assess the significance of geographical patterns of phanaeine
distribution, two points come to mind. First, field work in areas like the
Amazon Basin is restricted logistically to efficient avenues of penetration,
i.e., mostly large rivers. Thus, one can gather data yielding sound ideas
about beetles collected along and near rivers but no ideas about what occurs
in areas between rivers, which in the Amazon Basin are immense pieces of
real estate. Consequently, inferences about distributions of largely Ama-
zonian groups are subject to the hypothesis that, because the area appears
monotonously "homogenous" from an airplane, the unworkable inter-river
areas may be assumed to be faunistically similar to areas along and near
encompassing rivers. This is a logical, and probably largely accurate ex-
pectation. But just how "homogenous" are the 2.5 billion or so acres called
the Amazon Basin? Second, an important consideration is the possible ef-
fects of human activity on the recent movements and abundance of dung

Evolution of the Phanaeine Dung Beetles 871

beetles. The livestock industry in all parts of the world has undoubtedly
resulted in the geographic expansion and increased abundance of a great
many coprophagous forms. Seemingly, the faunas least affected by such
activity would have been those supported prior to the introduction of live-
stock by indigenous herbivores, such as the bison in the North American
Great Plains and the rich herbivore fauna of African savannas. The positive
effects of the introduction of livestock has no doubt been great in Latin
America, which in recent, pre-European times supported a very depauperate
indigenous herbivore fauna. Many coprophagous species of phanaeines and
other dung beetles now considered very common were perhaps quite scarce
in those times. A small scale example is the case of Sulcophanaeus carnijex,
which today can be collected in incredible numbers in cow dung on Jamaica
(Matthews, 1966) ; Jamaica has presumably never supported large herbivores
other than cattle, the only known mammalian fossils being bats, monkeys
and small rodents. On the other hand, the same livestock producing activ-
ities, which always involve the cutting of forests for pasturage (or timber,
crops, etc.), has undoubtedly had some negative affect on the distribution
and abundance of stenotopic forest species, whose habitat is shrinking at a
tremendous rate in many parts of Latin America.

LITERATURE CITED

Arrow, G. J. 1951. Horned Beetles. Junk, Lochem. 154 pp.

Balthasar, V. 1942. Uber die Nervatur des Hinterfliigels tier Lamellicornien. Ceske

Sploecnosti Nauk Trida Matematicko-Prirodovedecka Vestni'k 1942, Part XI, 1-44.
Barattini, L. P., and A. Saenz. 1961. Contribucion al conocimiento de las especies del

genero Gromphas Brulle. lero Congr. Sudamericano Zool. 3:21-29.
. 1964. Contribution a l'etude des especes du genre Gromphas Brulle. Rev. Franc.

d'Entom. 31:173-181.
Blackwelder, R. E. 1944. Checklist of the coleopterous insects of Mexico, Central America,

West Indies and South America, part 2. United States National Museum, Bulletin 185,

pp. 189-341.
Blut, H. 1939. Beitrag zur Verbreitung and Svstematik der Gattung Dendropaemon.

Archivs fur Naturgeschichte (N.S.) 8:263-300. '
Brulle, M. 1837. Observations sur l'absence des tarses dans quelques insectes. Ann. Sci.

Natur., Zool., ser. 2, 8:246-249.
Butt, F. H. 1944. External morphology of Amphimallon majalis (Razoumovski) (Coleoptera,

the European chafer). Cornell Univ. Agric. Exp. Sta. Memoir No. 266. 18 pp.
Castlenau, Gomte de. {■=. F. de Laporte). 1831. (Description of Enicotarsus vrndipennis) .

Mag. Zool. 1 (2) :plate 35. (Reference from Blackwelder, 1944.)

. 1840. Historie Naturelle des Insectes, Coleopteres, vol. 2. Paris. 564 pp.

Crowson, R. A. 1938. The met-endosternite in Coleoptera; a comparative study. Trans. Rov.

Ent. Soc. London 87:397-416.
. 1944. Further studies on the met-endosternite in Coleoptera. Trans. Roy. Ent. Soc.

London 94:273-310.
. 1955. The Natural Classification of the Families of Coleoptera. N. Lloyd & Co.,

London. 187 pp.
Dejean, P. 1837. Catalogue des Coleopteres de la Collection de M. le comte Dejean. 3rd ed.

(rev. corr. et augm.). M.-M. Pere et Fils, Paris, 384 pp.
. 1838. (. . . observations . . . sur l'absence ou la presence des tarses dans plusieurs

insectes de la famille Lamellicornes . . .) (from the minutes of a society meeting.)

Ann. Soc. Ent. France 1838, pp. viii-ix.
Dorsey, C. K. 1943. The musculature of the labrum, labium and pharyngeal region of adult

and immature Coleoptera. Smiths. Misc. Coll. 103(7): 1-42.

872 The University Science Bulletin

DuPorte, E. M. 1960. Evolution of cranial structure in adult Coleoptera. Canad. J. Zool.

38:655-675.
Edmonds, W. D. 1967. The immature stages of Phanaeus {Coprophanaeus) jasius Oliver and

Phanaeus (Metallophanaeus) saphirinus Sturm (Coleoptera: Scarabaeidae). Coleopt.

Bull. 21:97-105.
Forbes, W. T. M. 1922. The wing venation of the Coleoptera. Ann. Ent. Soc. America

15:328-343.
Frenguelli, J. 1938. Bolas de escarabaeidos y nidos dc vespidos fosiles. Physis 12:348-352.
. 1939. Nidos fosiles de insectos en el Terciario de Neuquen y Rio Negro. Notas Museo

La Plata, LV (18-Paleontologia) :379-393.
Germar, E. F. 1824. Insectorum Species Novae aut Minus Cognotae, Descriptionibus Illus-

tratae. Hendelii et Fillii, Halae. 624 pp.
Gillet, J. J. E. 1911. Coleopterorum Catalogus, pars 38:Scarabaeidae: Coprinae I. Junk,

Berlin. 100 pp.
Gistel, J. N. F. X. (Also spelled Gistl.) 1857. Achthundert und zwanzig neue oder

unbeschriebne wirbellose Thiere. Vacuna, 1857, pp. 513-606. (Also as separate with

pagination 1-94.)
Graham, S. A. 1922. A study of the wing venation of the Coleoptera. Ann. Ent. Soc.

America 15:191-200.
Halffter, G. 1952. Notas sobre el genero Phanaeus I — Phanaeus quadridens Say. Ciencia

(Mexico) 12:79-86.

.1958. Dos nuevos generos de Canthonina. Ciencia (Mexico) 17:207-212.

. 1961. Monoe;rafia de las especies norteamericanas del genero Canthon Hoffsg. Ciencia

(Mexico) 20:225-320.
. 1964. La entomofauna americana — ideas acerca de su origen y distribution. Folia

Ent. Mexicana, num. 6, 108 pp.
Halffter, G., and E. G. Matthews. 1966. The natural history of dung beetles of the sub-
family Scarabaeinae (Coleoptera, Scarabaeidae). Folia Ent. Mexicana, num. 12-14.

312 pp.
Hardenberg, C. B. 1907. Comparative studies of the trophi of the Scarabaeidae. Trans.

Wisconsin Acad. Sci. Arts Letters 15:548-602.
Harold, E. von. 1875. Neue Coprophagen. Deutsche Ent. Zeitschr. 19:209-217.
Hennig, W. 1965. Phylogenetic systematica. Ann. Rev. Ent. 10:97-116.

. 1966. Phylogenetic Systematica. Univ. Illinois Press, Urbana. 263 pp.

Hinton, H. E. 1935. Anotaciones acerca tie las costumbres micetofagicas de las especies de

Phanaeus. An. Instit. Biol., Mexico 6:129-130.
Hope, F. W. 1837. The Coleopterist's Manual, Containing the Lamellicorn Insects of Lin-
naeus and Fabricius. London. 121 pp.
Janssens, A. 1932. Description d'un phanaeide inedit (Coleoptere Copridae). Bull. Ann. Soc.

Ent. Belgique 72:119-122.
. 1940. Contribution a l'etude des coleopteres lamellicornes coprophages II. — Remarques

sur quelques Phanaeides. Bull. Mus. roy. d'Hist. natur. Belgique, vol. 16, no. 34.

7 pp.
-. 1946. Ibid. XL — Table synoptique et essai tie classification practique des Scarabaeidae

Laparosticti. Loc. cit., vol. 22, no. 12. 13 pp.
. 1949. Ibid. XIII. — Table synoptique et essai de classification practique des coleopteres

Scarabaeidae. Bull. Instit. roy. Sci. natur. Belgique 25(15): 1-30.
Jeannel, R., and R. Paulian. 1944. Morphologie abdominale des Coleopteres et systematique

de l'ordre. Revue Francaise d'Ent. 1 1 :65-l 10.
Kadic, O. 1902. Studien iiber das Labium der Coleopteren. Zeit. nat. Wissensch. 36:207-228.
Klages, E. A. 1906. On the Scarabaeidae of southern Venezuela. A wonderful new form of

the group Copies. Published privately at Crafton, Penna. 1 p.
. 1907. A wonderful new beetle of the group Copris. Proc. Ent. Soc. Washington

8:141-142.
Kolbe, H. J. 1905. Uber die Lebensweise und die geographische Verbreitung der Coprophagen

Lamellicornier. Zool. Jahrb., Suppl. 8:475-594.
Lindroth, C. H. 1957. The principal terms used for male and female genitalia in Coleoptera.

Opuscula Ent. 22:241-256.
Linnaeus, C. 1758. Systema naturae . . . ed. 10, vol. 1. Holmiae. 823 pp.

. 1767. Ibid ed. 12 (rev.), vol. 1, pars 2, Holmiae, pp. 533-1327.

MacGillivray, A. D. 1923. External Insect Anatomy. Scarab Co., Urbana.

MacLeay, W. S. 1819. Horae entomologicae: or essays on the annulose animals, vol. 1, part

1. Bagster, London. 524 pp.

Evolution of the Phanaeine Dung Beetles 873

. 1833. Annulosa Javanica or an Attempt to Illustrate the Natural Affinities of Insects

Collected in Java by T. Horsfield. London. 50 pp. (Reference from Gillet, 1911;

cited MacLeay, 1825 in Blackwelder, 1944.)
Madle, H. 1934. Zur Kenntris der Morphologic Okologie und Physiologie von Aphodius

rtifipes und einigen verwandten Arten. Zool. Jarhb. 58:303-396.
Martinez, A. 1944. Insectos nuevos o poco conocidos. Rev. Soc. Ent. Argentina 12:184-189.
. 1959. Catalogo de los Scarabaeidae argentinos. Rev. Mus. Argentino de Cien. Nat.

"Bernardino Rivadavio" 5:1-126.
Martinez, A., and F. S. Pereira. 1967. Notas escarabaedologicas III. (Col. Scarabaeidae —

Scarabaeinae). Rev. Soc. Ent. Argentina 29:53-69.
Matsuda, R. 1965. Morphology and Evolution of the Insect Head. Amer. Ent. Inst. Memoir

No. 4. 334 pp.
Matthews, E. G. 1961. A revision of the genus Copris Muller of the western hemisphere

(Col. Scarab.). Entomologica Americana 41 (N.S.) :1-137.
. 1966. A taxonomic and zoogeographic survey of the Scarabaeinae of the Antilles

(Coleoptera: Scarabaeidae). Mem. Amer. Ent. Soc. No. 21. 134 pp.
Mayr, E. 1969. Principles of Systematic Zoology. McGraw-Hill, New York. 428 pp.
Miller, A. 1961. The mouthparts and digestive tract of adult dung beetles (Col. Scarab.),

with reference to the ingestion of helminth eggs. J. Parasitol. 47:735-744.
Mohr, C. O. 1930. Morphological comparisons of Coprinae, Aphodiinae and Geotrupinae.

Trans. Illinois St. Acad. Sci. 22:263-284.
d'Olsoufieff, G. 1923 (1924). Les Phanaeides (Col.: Lame!.), Famille Scarabaeidae — Tr.

Coprini. Insecta 13:4-172.
d'Orchymont, A. 1920. La nervation alaire des coleopteres. Ann. Soc. Ent. France 89:1-50.
Pereira, F. S. 1949. O subgenero Metallophanaeus. (Coleop. -Scarabaeidae). Arquivos do

Museu Paranaense 7:217-230.
Pereira, F. S., and A. Martinez. 1956. Algunas notas sinonimicas en Phanaeini. Rev.

Brasileira Entomologia 5:229-240.
Perty, }. A. M. 1830. De insectorum in America . . . observationes nonnullae. In Delectus

Animalium Articulatorum . . . Maximiliani Josephi I . . . Fasc. 1:1-44. Monachii,

London.
Pessoa, S. B. 1934. Contribucao para o conhecimento das especies brasileiras do genero

Phanaeus (Col. Scarabaeidae). Annaes Faculdade Medicina Sao Paulo 10:279-314.
. 1935. Notas sobre o genero Taurocopris, com a descrip^ao de uma especie nova (Col.

Scarabaeidae). hoc. cit. 11:33-35.
Pessoa, S. B., and F. Lane. 1936. Sobre os coleopteros do genero Dendropaemon Perty, de

S. Paulo e regioes visinhas. Revista Biologia e Hygiene 7:89-93.
. 1941. Coleopteros necrofagos de interesse medico-legal. Ensaio monografico sobre a

familia Scarabaeidae de Sao Paulo e regioes visinhas. Revista do Museu Paulista

25:309-504.
Pierce, W. D. 1946. Descriptions of the dung beetles (Scarabaeidae) of the tar pits. Bull.

So. California Acad. Sci. 45:119-131.
Ritcher, P. O. 1969a. Spiracles of adult Scarabaeoidea (Coleoptera) and their phylogenetic

significance, I: The abdominal spiracles. Ann. Ent. Soc. America 62:869-880.

. 1969b. Ibid. II: Thoracic spiracles and adjacent sclerites. hoc. cit. 62:1388-1398.

Robinson, M. 1947. A review of the genus Phanaeus inhabitating the United States. Trans.

Amer. Ent. Soc. 73:299-305.
Smith, J. B. 1892. The mouthparts of Copris Carolina and notes on the homology of the

mandible. Trans. Amer. Ent. Soc. 19:83-87.
Snodgrass, R. E. 1909. The thorax of insects and the articulation of the wings. Proc. U.S.

Nat. Mus. 36:511-595.

. 1935. Principles of Insect Morphology. McGraw-Hill, New York. 667 pp.

. 1957. A revised interpretation of the external reproductive organs of male insects.

Smiths. Misc. Publ. 135, no. 6. 60 pp.
Stellwaag, F. 1914. Der Flugapparat der Lamellicornier. Zeit. wiss. Zool. 108:359-429.
Vulcano, M. A., and F. S. Pereira. 1967. Sinopse dos Passalidae e Scarabaeidae s. stro. da

regiao amazonica (Insecta-Coleoptera). Atas do Simposio sobre a Biota Amazonica,

vol. 5 (Zoologia):533-603.
Waterhouse, C. O. 1891. New Scarabaeidae in the British Museum: 5th contribution. Ann.

Mag. Nat. Hist., ser. 6, 8:53-60.

874 The University Science Bulletin

APPENDIX: MATERIAL STUDIED

Following is a list of species examined for this study arranged according to the classification
proposed above. After each species name is noted parenthetically the degree to which each
was studied by the following letter code:

c — completely; all skeletal features studied in detail.

m — examined externally but mouthparts extracted; internal features not observed,
e — examined externally only; mouthparts not extracted.
If only one or two specimens were studied, their sex and number are also given. In addition,
a list of the non-phanaeine beetles follows that of phanaeines; all these species were examined
completely.
Diabroctis

mimas (c); mirabilis (m); cadmus (m, 1 male).
Sulcophanaeus

j annus (c); columbi (m); laeander (m); carnifex (c); imperator (c); steinheili (m, 1 male,

compared with type by Pereira); rhadamantus (c, 1 pair); batesi (c); menelas (c);

auricollis (e); chryseicollis (c); noctis (m); velutinus (m); actaeon (m).

Oxysternon

jestivum (c); macleayi (e); lautum (c); conspicillatum (c); ebeninum (e); palaemon (c);
sericeum (m); smaragdinum (m); silenum (m); curvispinum (m); pteroderum (e, 1
male).

Phanaeus

splendidulus (c); mclibaeiis (m); dejeani (m); palatno (c); \irbyi (m); haroldi (c);
bispintis (c); chalcomelas (c); meleagris (m); tndymion (c); pyrois (m); blanchardi (e, 1
pair); lunaris (m); achilles (m, 1 male); lugens (m); prasinus (m); hermes (m); mexicanus
(c); amythaon (m); wagneri (m); demon (c); obliquans (m); qtiadridens (c); damocles
(m); palliatus (m); eximius (m, 1 pair); adonis (c): pilatei (m); *«//«' (m); beltianus (m);
guatemalensis (m); amethystinus (m); scitlijcr (e, 1 male); tnelampus (m); daphnis (c);
tricornis (m); tridens (m); ntmrod (m); furcosus (m); vindex (c); difjormis (c); />/-
angitlaris (c); floridanus (c).

;'rt^/?/j (c); cerberus (c); parvtdtis (c); *p//2/' (c); milon (c); telamon (c); dardanus (c);
p/«//o (m); saphirinus (c); /;o;v« (m, 1 pair); /w/y/ (m); pessoai (e); lanrifer (c); ensijer
(c); bonariensis (c); bellicosus (c).

denticollis (c); haroldi (m); /tf/r« (e); nitidicollis (m); smaragdinum (e); viridipennis
(m); amyntas (m); fractipes (in); lobatum? (m); piceum (m); viridis (c); telaphum
(m); renatii (m); hirticollis (m); bahianus? (m); ganglbaueri (m, 1 female).

Tetramereia convexa (c).

Homalotarsus impressus (e, 1 female).

Megatharsis buchjeyi (m, 2 females).

Non-Phanaeines examined (all are Scarabaeinae unless indicated otherwise)

Anomiopsoides sp.; Heliocopris sp., Canthon pilularis (L.); Dichotomitis sp.; Omto sp.;
Eurysternus sidcifer Balths.; Deltoclulum brasiliense (Cast.); Scarabaeus sp.; Catharsius
molossus (L.); Copris sp.; Pliyllophaga sp. (Melolonthinae) ; Pseudolucanus sp. (Lu-
canidae).

THE UNIVERSITY OF KANSAS

SCIENCE BULLETIN

Vol. XLIX Pages 875-988 June 5, 1972 No. 12

TABLE OF CONTENTS

ABSTRACT 877

INTRODUCTION 877

ACKNOWLEDGMENTS 879

MATERIALS AND METHODS 879

Illustrations 881

GENERAL AND TAXONOMIC CHARACTERS 881

MORPHOLOGY AND TAXONOMIC USE OF GENITALIA 886

SEXUAL DIMORPHISM 888

INDIVIDUAL VARIATION 889

LIFE HISTORIES 890

FOOD PLANTS 893

SYSTEMATIC POSITION AND RELATIONSHIPS 894

SYSTEMATIC TREATMENT 896

Chlamisinae Gressitt, 1946 896

Review of Taxomony 896

Key to the Genera of North American Chlamisinae 902

Pseudochlamys Lacordaire, 1848 902

P. semirufescens n. sp. 905

Chlatnisus Rafinesque, 1815 908

Key to the North American species of Chlamisus 911

C. flavidus n. sp 911

C. arizonensis (Linell) 914

C. texanus (Schaeffer) 915

C. joveolatus (Knoch) 917

C. quadrilobatus (Schaeffer) 919

C. huachacae (Schaeffer) 921

C. nigromacidatus n. sp 923

C. maculipes (Chevrolat) 924

Diaplacaspis Jacobson, 1924 928

D. prosternalis (Schaeffer) 930

Neochlamisus n. gen. 932

Key to species groups of Neochlamisus 935

velutinus group 935

Key to the velutinus species group 936

N. velutinus n. sp 936

N. memnonius (Lacordaire) 940

N. moestificus (Lacordaire) 942

N. scabripennis (Schaeffer) 947

N. subelatus (Schaeffer) 950

gibbosus group 951

Key to the gibbosus species group 951

TV. gibbosus (Fabricius) 952

N. bimaculatus n. sp 956

N. insularis (Shaeffer) 958

N. platani (Brown) 960

N. bebbianae (Brown) 963

N. eubati (Brown) 966

N. fragariae (Brown) 969

N. comptoniae (Brown) 970

N. assimilis (Klug) 972

N. chamaedaphnes (Brown) 973

N. cribripennis (LeConte) 976

N. tuberculatus (Klug) 977

LITERATURE CITED 979

MAPS 982

A Revision of the Subfamily Chlamisinae of America
North of Mexico (Coleoptera: Chrysomelidae)1

Jay B. Karren

ABSTRACT

In this revision 29 species are figured and described with three Mexican
species figured and described to a limited extent. Keys to the species and genera
of this area are included. Although primarily taxonomic, this work brings to-
gether available data concerning variation, distribution, host plants, parasites and
predators. Five new species, Pseudochlamys semirujescens, Chlamisus flavidus,
C. nigromaculatus, Neochlamisus velutinus, and N. bimaculatus are described.

A study of the genitalia has made possible a better understanding of the
group and has indicated more reliably the evolutionary relationships among
species and genera, so that meaningful statements about zoogeography and
possible dispersal patterns in the group can be made. On the basis of genitalic
characters, a new genus is proposed for some species formerly included in the
genera Chlamisus and Diplacaspis. Pseudochlamys is recorded in North America
for the first time by the description of one new species.

A study of the food plants revealed that most have a wider range and host
distribution than was initially supposed. Although a particular species may be
limited to one host in a small geographical area, in other localities the host
preference can be quite different. Some food plants are acceptable to several
species. Field observations and museum labels show that some North American
chlamisines feed on a wide variety of host plants, while others are host specific.

INTRODUCTION

This is a revision of the species of Chlamisinae occurring in America
north of Mexico, together with records of biological observations on these
species. It is based on a study of 4,878 specimens, field studies of two species,
and a preliminary study of many Mexican and Central American species.

Most insect collectors easily overlook the Chlamisinae because both adults
and larvae of many kinds of these small beetles superficially resemble
caterpillar droppings or other bits of debris. As a result, chlamisines are
not common in collections and have not attracted the level of taxonomic
attention given the other chrysomelid subfamilies. Although some of the
North American species were among the first known in the subfamily, the
few revisionary studies of Chlamisinae have not treated them in much detail
but have dealt primarily with the more numerous and more strikingly
colored Neotropical chlamisines. This revision was undertaken in the hope
of clarifying the taxonomic status of the various recognizable forms oc-
curring in America north of Mexico, making known something of the
biology of the group, and describing the supposedly phylogenetic relation-

1 Contribution no. 1489 from the Department of Entomology, The University of Kansas,
Lawrence, Kansas 66044.

878 The University Science Bulletin

ships among species and species groups, both in the region treated and
between these groups and their Mexican and more southern relatives.

Of the 32 names that have been proposed in this group, seven are placed
in synonymy, two are misidentincations of Mexican species, and two are
names that cannot be associated with any species because of lost type material
and inadequate descriptions. The last two could be junior or senior syn-
onyms of recognized species; but for inability to identify species involved
and in the interest of nomenclatural stability, neotypes will not be designated.
Five species are described as new. Descriptions and figures of 26 North
American species are included and three Mexican species are figured and
described to a limited extent. These latter are included because the informa-
tion about them will aid in identifying other North American species and
will contribute to an understanding of the species groups involved.

The inadequacy of Lacordaire's (1848) classification of Chlamisinae has
been recognized for a long time, although no one has attempted to revise the
entire group. A preliminary study indicated that genitalic characters are
useful in determining the species and their relationships. A study of the
genitalia has made possible a better understanding of the group and has
indicated more reliably the evolutionary relationships among species and
genera, so that meaningful statements about zoogeography and possible
dispersal patterns in the group could be made. On the basis of genitalic
characters, a new genus is proposed for some species formerly included in
the genera Chlamisus and Diplacaspis. Pseudochlamys is recorded in North
America for the first time by the description of one new species in that genus.
Chlamisus is probably still heterogeneous, but further study is needed of the
Mexican and Central American species before any further generic distinc-
tions can be drawn.

My previous study of the genus Exema (Karren, 1966) indicated certain
monophagous tendencies within the subfamily. Brown's study of the
Canadian Chlamisinae (1943) indicated that each species is restricted to one
particular host. Further study of the species described by Brown revealed
that most have a wider range and host distribution than was initially sup-
posed. Some food plants are acceptable to several related species. Although
a particular species may be limited to one host in a small geographical area,
in other localities the host preference can be quite different. Field observa-
tions and museum labels show that some North American chlamisines
feed on a wide variety of host plants, while others are host specific. There
is a need for more accurate and extensive field observations of host
preferences.

The life histories of a very few species have so far been made known.
Through laboratory rearings and field observations, additional informa-
tion has been obtained on several species. Data on habits and life history

A Revision of the Subfamily Chlamisinae 879

stages increase the understanding of the relationships among the species.
Laboratory rearing of specimens has also revealed considerable parasitism
of eggs, larvae and pupae by Ichneumonidae, Chalcididae and other
Chalcidoidea.

ACKNOWLEDGMENTS

Sincere thanks go to the following institutions and individuals who kindly loaned material
for this study.

Dr. N. L. Anderson, Montana State Univ.; Dr. R. H. Arnett, Jr., Purdue Univ.; Dr. W. T.
Atyeo, Univ. of Nebraska; Dr. G. E. Ball, Univ. of Alberta; Dr. E. U. Balsbaugh, Jr., South
Dakota State Univ.; Dr. L. J. Bayer, Univ. of Wisconsin; Dr. W. W. Boyle, Pennsylvania State
Univ.; Dr. H. R. Burke, Agricultural and Mechanical College of Texas; Dr. E. F. Cook, Univ.
of Minnesota; Dr. R. Damoiseau, Institut Royal des Sciences Naturelles de Belgique; Dr. W.
R. Enns, Univ. of Missouri; Dr. R. Fischer, Michigan State Univ.; Mr. S. Frommer, Univ. of
California, Riverside; Dr. J. L. Gressitt, Bernice P. Bishop Museum; Dr. W. Hackman,
Museum Zoologicum Universitatis, Helsinki; Dr. W. J. Hanson, Utah State Univ.; Mr. H. A.
Hespenheide, Univ. of Pennsylvania; Dr. C. L. Hogue, Los Angeles County Museum; Dr. P.
Hunter, Univ. of Georgia; Dr. M. T. James, Washington State Univ.; Dr. K. V. Krombein,
U.S. National Museum; Dr. J. L. Laffoon, Iowa State Univ.; Dr. R. L. Langston, Univ. of
California, Berkeley; Dr. I. Lansbury, Hope Department of Entomology, Univ. Museum,
Oxford; Dr. J. D. Lattin, Oregon State College; Mr. A. R. Moldenke, Stanford Univ.; Dr. T. E.
Moore, Univ. of Michigan; Dr. C. S. Moss, Escuela Nacional de Agricultura, Chapingo,
Mexico; Mile. B. Nicole, Museum National d'Histoire Naturelle, Paris; Dr. M. Nielsen, U.S.D.A.,
Mesa, Arizona; Mr. F. Pacheco, Institute Nacional de Investigaciones Agricolas, Obregon,
Mexico; Dr. J. C. Pallister, American Museum of Natural History; Dr. L. L. Pechuman,
Cornell Univ.; Dr. R. L. Post, North Dakota Agricultural College; Dr. C. W. Rettenmeyer,
Kansas State Univ.; Dr. M. W. Sanderson, Illinois Natural History Survey; Dr. G. Scherer,
Museum G. Frey, Munich; Dr. J. B. Schmitt, Rutgers Univ.; Dr. R. O. Schuster, Univ. of
California, Davis; Dr. C. A. Triplehorn, Ohio State Univ.; Dr. G. E. Wallace, Carnegie
Museum; Dr. R. L. Wenzel, Field Museum of Natural History; Dr. F. G. Werner, Univ. of
Arizona; Dr. S. L. Wood, Brigham Young Univ.; Dr. R. E. Woodruff, Florida Department
of Agriculture; Dr. D. A. Young, North Carolina State Univ.

Special thanks go to the following individuals who loaned specimens and made important
type material available for study: Dr. W. J. Brown, Canada Department of Agriculture; Dr.
F. H. Heike, Humboldt Universitat zu Berlin; Dr. J. L. Lawrence, Museum of Comparative
Zoology, Harvard Univ.; Dr. R. T. Thompson, British Museum (Natural History); Dr. R. E.
White, U.S. National Museum; Dr. J. A. Wilcox, New York State Museum.

Thanks also go to my wife, Sarah W. Karren, for the many hours of typing, drawing and
constructive criticism devoted to the preparation of this paper.

Finally, I wish to thank Dr. George W. Byers for his guidance and direction during this
study and for reading and editing the manuscript. I am also indebted to Drs. Charles D.
Michener and R. H. McGregor for reading the manuscript and offering suggestions.

This research was aided by a travel grant and summer traineeship from the Systematics
and Evolutionary Biology Program of the University of Kansas, sponsored by the National
Science Foundation (Grant GB-4446X). The manuscript was written while the author held
a Dissertation Fellowship from the University of Kansas.

MATERIALS AND METHODS

Specimens for this study were borrowed from all of the major collections in
North America and several European museums, as listed in the Acknowledg-
ments. Type specimens have been examined whenever possible and compared
with other material used in this study. It has not seemed necessary at this time to
designate neotypes for those types that are lost. Some collections and field ob-
servations were made to supplement the data obtained from museum specimens.

880 The University Science Bulletin

Unfortunately, biological data, even names of host plants, are entirely lacking
for a few of the rare species. Only more careful field studies in the future will
provide such information.

Label data on specimens were recorded in a uniform manner, in some cases
requiring interpretation where numbers represent dates or standard abbreviations
are not used. Complete label data are recorded for type material. For the dis-
tribution of species in the United States and Canada, only the county is given
(for type specimens, placed in parentheses if not originally included on the
locality label). There is a distribution map for each species on which the spots
usually represent counties in the United States, except more specific localities are
indicated in the large western counties and political divisions of countries other
than the United States. More definite information is recorded, when available,
for localities outside the United States and Canada. Major political divisions of
Canada, Mexico and the Central American countries are cited in the same way
as states. Localities that are not definitely identified are placed in quotation
marks. Inclusive collecting dates are designated whenever two or more labels
indicate a collecting span of ten days or fewer.

For detailed study, especially of poorly mounted material on cards and points,
the specimens were first treated with relaxing fluid (95% ethanol, 106 cc; distilled
water, 98 cc; benzol, 14 cc; ethyl acetate, 38 cc) for about 15 minutes or until the
legs were pliable. The legs and antennae were then spread into more extended
positions for study, the specimen mounted on an insect point and allowed to dry.
Broken parts were mounted on the upper surface of the point. The male genitalia
or female spermathecae were removed and placed in hot KOH for about 10
minutes, then transferred to 70% alcohol to which a drop or two of acetic acid
had been added. Removal of the ejaculatory guide from the aedeagus was usually
done in 70° G alcohol in a Syracuse dish. Minuten pins and small cover slips were
helpful in orienting the parts in a proper position for study and drawing.
Genitalia were later placed in small vials which were attached to the pin of
the mounted specimen.

All drawings and measurements were made with the aid of an ocular grid
in a steroscopic microscope, using 15X, 60 X, arid 120X magnifications.

Terminology is taken from "A Glossary of Entomology" by Torre-Bueno
(1937). More detailed information on special terms and their use in the species
descriptions will be found in the discussion of the general and taxonomic char-
acters of the subfamily. Terms used for the male genitalia are those used by
Powell (1941). The female spermathecae are described using the terms of Spett
and Lewitt (1926).

The following abbreviations are found in parentheses with the listings of
types and paratypes indicating the private collection, museum or university in
which the types are to be found.

AMNH American Museum of Natural History

CAS California Academy of Sciences

CDA Canada Department of Agriculture

CU Cornell University

ENA Escuela Nacional de Agricultura, Chapingo, Mexico

A Revision of the Subfamily Chlamisinae 881

HFH H. F. Howden collection

INHS Illinois Natural History Survey

ISU Iowa State University

JAW J. A. Wilcox collection

JBK J. B. Karren collection

MCZ Museum of Comparative Zoology, Harvard University

MZU Museum Zoologicum Universitatis, Helsinki

OSU Ohio State University

SDSU South Dakota State University

TAMU Texas A & M University

UAR University of Arizona

UK University of Kansas

UM University of Missouri

USNM United States National Museum

Illustrations

Structures of special taxonomic value are illustrated. Where there is variation
in an illustrated structure, this is described in the treatment of the species but the
drawing is of the most common form.

Illustrations for each species usually include the entire beetle (dorsal, lateral
and cephalic aspects), presternum, aedeagus (dorsal and lateral aspects), ejacula-
tory guide (dorsal and lateral aspects) and the spermatheca. A few drawings
have been made of antennae and tarsal claws when deemed necessary to differen-
tiate two or more forms. Comparable illustrations have been made to the same
scale for all species.

In the drawings of the dorsal and cephalic views of the beetles, the right side
is used to indicate color patterns and the left side punctation and sculpturing.

Each pair of drawings of the entire beetle, aedeagus, and ejaculatory guide is
numbered as one illustration and represents first the dorsal then the lateral aspect
of each structure. The dorsal aspect of the aedeagus is drawn with the large apical
portion as nearly horizontal as possible, reducing the visible amount of the curved
section over the basal foramen. An accurate representation of the size of this
section is found in the lateral aspect. In a few cases the ventral aspect of the
ejaculatory guide is shown to illustrate details not shown in the other two aspects.

GENERAL AND TAXONOMIC CHARACTERS

A formal description or diagnosis of the subfamily will not be given;
instead, a discussion of the characteristics which have been used in the
descriptions and those which otherwise warrant some explanation will be
given. Any structure that is of the same generalized shape throughout the
subfamily will be referred to as normal here and omitted in the descriptions.

North American species of the subfamily Chlamisinae are small insects
ranging from 2.1 to 6.0 mm long. The species of Exema and Pseudochlamys
are among the smallest forms in North America. The genera Neochlamisus
and Chlamisus exhibit a wide range of variation, with species among the
smallest and the largest in North America.

882 The University Science Bulletin

Color in North American chlamisines varies from yellow to metallic
black. The beetles may be either uniformly colored or with various color
patterns that are distinctive for the species. The colors of most North
American species are metallic to some extent. This condition is obvious
without the aid of a microscope in most species, but in others only small
areas under high magnification reveal their metallic nature.

The head is usually the same color as the rest of the insect, but it may
be differently colored, in which instances the color pattern is a secondary
sexual characteristic. The antennae are entirely yellow or light brown, or
the distal segments may be dark brown. The color change can be abrupt
or gradual toward the apical segments. The labrum is yellow to light
brown, the other mouth parts darker. The undersurface is generally similar
to the rest of the body in color except that it is variously marked with black
in the yellowish forms from the Southwest. The legs are generally uniformly
colored like the rest of the body, although in a few species they are yellow
to light brown or yellow banded.

As seen from above, the general body shape in the group is subquadrate,
a little narrowed behind. Many species are distinctly more elongate than
others. The cuticle has a characteristic finely granulated appearance, more
obvious in some species and on some structures than in others. Such
granulation is here described as strongly granulate as opposed to vermiculate,
punctate or striate. Any structure lacking punctures, tubercles or carinae is
described as smooth although it still may have the familiar granulate ap-
pearance.

Faint, shallow or deep depressions called foveae are common in most
species on various body structures but are most common and usually deepest
on the metasternum. Often the closely set foveae and punctures produce
a reticulate condition. The foveae are usually more strongly granulate than
the surrounding area, and the center of each is occupied by a faint to deep
puncture. The foveae on a structure may be so confluent that only small,
irregular carinae remain. The elongate tubercles and connecting ridges
between two or more tubercles on the elvtra are also called carinae.

J

Many different sizes and shapes of parallel impressed lines and ridges
are found on many parts of chlamisines. The former are called striae, al-
though they are not the usual longitudinal lines on the elytra referred to
by coleopterists. They are found on the pronotum, on the elytral tubercles
and carinae, inside large punctures and foveae and on almost any other
part of the body. Some striae become so convoluted and irregular that the
surface is called vermiculate. Ridges are referred to as carinae.

There is a great deal of variation in the punctures of the cuticle. They
can be so small or faintly impressed that they can only be perceived with
difficulty under high magnification. Others are small but very deep and

A Revision of the Subfamily Chlamisinae 883

some are deep with a large impressed surrounding area. Most punctures
have a single seta in the center, ranging from so small that it is difficult to
see to large, long and obvious. In at least one species {Chlamisus maculipes)
there are several hairs in each puncture. Three punctate conditions are
commonly described: finely punctate, when the punctures are more widely
separated than their diameters; moderately punctate, when the distance
between punctures and their diameters are approximately equal; and
coarsely punctate, when punctures are separated by less than their diameters.
In addition, deep punctures are those deeper than their diameters; shallow
punctures are shallower than their diameters. Faint punctures are barely
visible under high magnification. The foveolate, striate and punctate con-
ditions described above in several states are actually a continuum. The terms
are used only for convenience of description and their usefulness depends
heavily on the observer's judgment.

The head is inserted inside the anterior section of the pronotum. It
appears circular when viewed from the front. The large lateral eyes are
deeply emarginate just above the antennal sockets. In the dark or metallic
species the area of the frons in the emargination of the eye usually has a
large yellow spot. Three areas of the frons are commonly referred to in
the descriptions. The lower frons is that portion below the level of the
antennal sockets. Above this and between the emargination of the eyes is
the middle frons. The area above the middle frons and below the vertex
is the upper frons.

The mouthparts are received between the clypeus and the large pro-
sternum. The antennal grooves are located lateral to the prosternum, be-
tween it and the procoxae. In some species there is an antennal fovea on the
lower frons just ventral and medial to each antennal socket. The undif-
ferentiated clypeus and lower frons between and below these foveae are
often depressed and punctate. The middorsal area of the upper frons below
the vertex is also often depressed or longitudinally grooved.

The antennae have been used for a long time as one of the basic char-
acters for the classification of chlamisines. Less significance is placed on
them in the present paper because of interspecific variation in the antennal
segments and their noncorrelation with other characters. There are, how-
ever, some general patterns of antennal characters that are of use in estab-
lishing relationships.

The scape is similar in most of the chlamisines. It is the longest
antennal segment, slightly curved and enlarged apically. The small sub-
globular pedicel is attached to its apex. The remaining segments are
gradually flattened toward the apex. The width of these flattened segments
refers to their greatest width. The third antennal (first flagellar) segment
is usually subcylindrical and a little longer than the preceding. In some

884 The University Science Bulletin

cases it has a slight apical enlargement on the ventral side. The next 1-3
segments can be similar to each other and subequal, or each may be sub-
conical with a slightly broader and deeper enlargement than that of the
preceding segment. The apical 4-5 segments are rectangular in outline and
usually a little broader than long. The apical segment is always subcorneal,
a little larger than the preceding segments and rounder at the apex. In
previous treatments of the chlamisines the antenna is described as serrate,
but it is not clearly understood by coleopterists which segment is to be con-
sidered the beginning of the serrate portion of the antenna. I have avoided
this problem by calling segments subtriangular when they are apically as
wide as long, but narrowly constricted at the base. Segments that are wider
than long but still not rectangular are called subrectangular. The basal
flagellar segments which are longer than their apical widths are called
subcylindrical.

The pronotum is faintly to highly gibbose, usually with a longitudinal
groove that partially divides the gibbosities. The pronotum can be either
punctate, striate or both, depending upon the species or species group. The
pronotum is either nontuberculate or possesses tubercles according to a
definite pattern.

The tubercles on the pronotum and elytra follow the same general
pattern as discussed by Karren (1966), except that they are usually more
highly developed and in some cases more numerous than the entire comple-
ment suggested. Tubercles sometimes are shifted slightly in position, but
never so much that they cannot be recognized. The terminology used in
the above mentioned paper with a few additions is adequate for purposes
of this study. The term marginal carina refers to a variable ridge extending
from the umbone to just before the apex of the elytron near the lateral
margin. Carinate tubercles are found in the positions of named tubercles,
but extend more or less to other tubercles.

The mesoscutellum is always visible as a subquadrate structure behind
and between the posterior lobes of the pronotum. The sides converge more
or less anteriorly, depending upon the group. The anterior margin, which
consists of three rounded processes, is usually covered by the pronotum.
The central process may develop between the posterior processes of the
pronotum and extend as a rounded carina along the center of the scutellum
for various distances. The posterior margin may be straight, depressed,
broadly emarginate or both depressed and emarginate. The posterolateral
angles are reflexed when they project above the elytra. The metascutellum
or second scutellum may or may not be exposed just behind the mesoscutel-
lum and between the elytra. In those species where it is visible, it is
usually sharply carinate.

Behind the mesoscutellum the elytral suture is usually straight for a

A Revision of the Subfamily Chlamisinae 885

short distance. The remainder is strongly serrate to the apex. The propor-
tion of the suture that is straight varies among the species, except that it
is completely serrate in all species of Exema and in C. quadrilobatus. Several
Neotropical species have an entirely straight suture.

The pygidium is always completely exposed beyond the ends of the
elytra. It is carinate, tuberculate, vermiculate or foveolate with various
degrees of punctation. The apex is narrowed, rounded, not truncate, and
extends slightly anteriorly beneath the abdomen. Two large lateral de-
pressions are usually evident on the apical one-third of this structure. Four
similar depressions are also found near the base.

The most common form of the prosternum is coarsely punctate and
goblet-shaped. The anterior half is broad and the sides are gradually or
suddenly convergent to a long prosternal process, which fits tightly between
two anteriorly projecting processes of the mesosternum. There is actually
a third central process on the mesosternum, which fits into the hollow,
dorsal, apical portion of the prosternal process. Variations in the prosternum
include almost straight and convergent lateral margins resulting in a sub-
triangular shape. In other species the prosternal process is enlarged,
flattened, evenly rounded or longitudinally grooved toward the apex. The
large anterior half of the prosternum is either flat or with a large central
depression. Just behind the anterior margin, which flanges beneath the
mouthparts, there may be a transverse groove.

The broadly exposed sclerites of the sternum are the large medial
metasternum, divided by a complete longitudinal groove, and the smaller
lateral metepisternal plates. These are usually deeply foveolate or roughly
sculptured. The five visible segments of the abdomen, especially the first
and the last, are similarly sculptured. The intervening segments 2-4 are
shortened so that only the posterior margins are exposed and sculptured.
Any reference to the abdomen in a description refers to one or more of
these ventral segments.

The anterior and middle pairs of legs in repose are received in broad
grooves on the ventral surface, bordered by the lateral pronotal margin, the
prosternum, and the anterior margin of the metasternum. The posterior
pair of legs is received in depressions formed in the first visible abdominal
segment just posterior to the metasternum. Each femur has a small longi-
tudinal groove to receive the tibia. An acute, usually inwardly curved,
apical process on the posterior angle of each front tibia is usually found in
males that I have studied, though in a few species it is paired. A similar
process is found on the middle tibia of some males and the front and middle
tibiae of some females. These are called tibial spines in the descriptions
and keys.

The tarsal segments are of the generalized chrysomelid type. The first

886 The University Science Bulletin

two segments are subtriangular, the third is strongly bilobed, the fourth is
hidden between the lobes of the third segment, and the fifth is long and
slender. The tarsal claws are simple, broadly toothed or bifid; the form
usually has generic or specific significance.

MORPHOLOGY AND TAXONOMIC USE OF GENITALIA

The male genitalia provide many characters of taxonomic and evolu-
tionary significance. Previous study of chrysomelid genitalia and their im-
portance has been of a general nature. Powell (1941) studied the male
genitalia of most of the subfamilies but included only a few representatives
of each group and only one chlamisine species. Spett and Lewitt (1926)
did a similar study of the female spermatheca that included one species of
Chlamisus. Several other less extensive studies of minor importance have
also been undertaken. The most important contribution to an understand-
ing of the male genitalia was included in a revision of the Chlamisinae of
Argentina by Monros (1952). This included a number of line drawings of
the entire or apical half of the male aedeagus. Very little has been said
about the structure and sclerotization of the internal sac of the aedeagus,
but important indications of the relationships of species can be obtained
from a comparative study of these sclerotized parts, as will be shown below.
Taxonomic use has been made of such information for the first time in
this study.

The largest structure among the male copulatory organs is the long,
tubular aedeagus. Closely associated with the aedeagus are the tegmen,
the spiculum gastrale, two elongate lateral plates and the first post-pygidial
tergite. This tergite and the lateral plates are similar in all the species
studied and therefore will not be considered further. The Y-shaped spiculum
gastrale is also of generalized form in the material studied and indicates a
certain degree of affinity of Chlamisinae to the closely related subfamilies.
More variable but still rather uniform within the subfamily is the Y-shaped
tegmen, the lateral forks of which partially embrace the aedeagus just
posterior to the basal foramen. A basal strut of the tegmen extends forward
below the aedeagus and partially covers the large foramen. A continuous
membrane attaches along the entire lateral edge of the tegmen, extends
dorsally around the posterior aedeagus, and forms a membranous sleeve
through which the aedeagus passes.

Important specific and generic characters are found in the size and shape
of the aedeagus. The most generalized form, found in many chlamisines,
many closely related subfamilies, and to a lesser extent throughout the
chrysomelids, is the short, tubular aedeagus with an apical orifice and basal
foramen. Various modifications in the shape of the apex and the form and
placement of apical hairs furnish characters for specific identification. More

A Revision of the Subfamily Chlamisinae 887

specialized chlamisines have the "apical" orifice (Powell, 1941) more cen-
trally situated on the aedeagus. The apical portion of the aedeagus is also
variously modified, usually flattened dorsoventrally with a ventrally curved
apex. The dorsal wall of the anterior half of the aedeagus is curved ventrally
and less sclerotized than the posterior portion. The basal foramen is found
on the ventral side of this anterior portion.

An internal sac lies anteriorly within the aedeagus and is continuous
with its apical orifice. The walls of this sac are either membranous or
sclerotized to varying degrees, offering further diagnostic characters. A
sclerotized plate in the dorsal wall of the sac is continuous with the median
dorsal plate attached to the anterior edge of the apical orifice. Internally
this plate forms a projecting hook of various shapes that is exserted during
copulation. Attached to the anterior end of the sac is the ejaculatory guide,
a sclerotized structure with numerous modifications of taxonomic use.
The generalized tubular shape of the guide is found in Pseudochlamys,
Exema, and some species of Chlamisus. In Neochlamisus the structure is
more specialized, consisting of two closely fitting asymmetrical plates with
a ventral sheath that varies in size and shape. The ejaculatory duct enters
the sac through a basal orifice near the anterior end. In those forms with a
sheath, the basal orifice is just behind the base of the sheath and between
the asymmetrical plates near their proximal end. The posterior edge of this
orifice can be seen in cleared material. Its position in relation to the total
length of the ejaculatory guide is of taxonomic value. Muscles for retracting
the sac attach to its anterior end and lie in the anterior aedeagus. The long
ejaculatory duct passes through this end of the aedeagus and through the
large basal foramen.

The female genitalia are more uniform throughout the subfamily than
the male structures, and the slight differences are hard to evaluate except
for those of the spermatheca. This structure has several characters that can
be used taxonomically and in phylogenetic studies. The highly sclerotized
spermatheca is attached to the common oviduct by a sclerotized spermathecal
duct, which varies from long and highly convoluted to short and straight.
The spermatheca proper is divided, by the attachment of the spermathecal
duct, into a short basal section, and a longer U-shaped or J-shaped muscular
section, respectively the collum and the cornua (Spett and Lewitt, 1926). The
collum may be very small and bead-like or large and bulbous. The
spermathecal gland, a nonsclerotized elongate sac, attaches to the apex of the
collum. This accessory gland is sometimes damaged or difficult to find in
preserved specimens, so that it is not always completely illustrated with the
spermatheca, though at least the point of attachment is indicated. The
muscles of the cornua extend from the apex across the curved portion. There
is a closed nonsclerotized sac inside the cornua. Sperm that are forced into

888 The University Science Bulletin

the spermatheca partially collapse this internal sac. Contraction of the
muscles extend the sac and force sperm through the spermathecal duct to
the common oviduct. Care must be used in removing the spermatheca from
the cavity of the abdomen or it will be broken. It is somewhat isolated from
the other genitalic structures and often remains inside the abdomen when
they are removed.

SEXUAL DIMORPHISM

The various characters exhibiting sexual dimorphism will be included
with generic and specific treatments. The general dimorphic characters as
well as those restricted to certain species will be pointed out for the North
American species. When used alone these characters are diagnostic in most
cases and when correlated with other characters, they are useful in showing
the relationships of species within a genus. In some instances there is a
continuum from one dimorphic character state to the other.

The most obvious sexually dimorphic characteristic of the group is size.
Females are generally much larger than males, although the largest males
of a species are the same size as or slightly larger than the smallest females.
In a few cases the smallest female specimen may be similar in size to the
smallest male, but never is the largest male as large as the largest female.
This size difference is less obvious in the smaller species than in large ones.

Common to the entire group is a small to large midventral fovea on the
last abdominal segment, especially of the female. There is considerable
interspecific and some intraspecific variation among females in this char-
acter, but the fovea in the male is always smaller and less distinct than in
the corresponding female. In most females it is a distinct, large, circular
depression, whereas in the male the area is smooth and flat or only faintly
impressed. In females this fovea is usually surrounded by long and short
hairs, the longer ones covering the two sides and the short ones the anterior
and posterior borders of the fovea. In males these hairs take the form of
two groups on the foveal area separated by a shiny, smooth, or punctate sec-
tion. Each group is usually made up of long anterior hairs gradually be-
coming shorter toward the posterior margin of the segment. Very short
hairs are usually present along the margins of the foveal area.

Single or double apical spines occur on the front tibiae of most male
chlamisines. An additional spine is found on the middle tibiae in some
species or species groups. In females they may or may not be present de-
pending upon the species group. Their arrangement in some species is the
same in both sexes, but in most the females have none or fewer and smaller
spines than in males. The number and sizes of tibial spines in each sex is
usually constant for a given species, but in some there is variation.

Other kinds of sexual dimorphism restricted to certain genera, species or

A Revision of the Subfamily Chlamisinae 889

species groups are found throughout the subfamily. Males of Pseudochlamys
have enlarged and sometimes asymmetrical mandibles, while the females
are normal. The head and pronotum of males also differ in size and shape
from those of females in this genus.

Spines and spinulae as sexually dimorphic characters found on the
ventral abdominal segments have already been discussed in my revision of
Exema (Karren, 1966). These occur in some chlamisines but not in any of
the species presently under consideration.

Several species of Neochlamisus have a wider presternum in females
than in males. In N. gibbosus and less so in a few other species, the pro-
sternal process is strongly serrate in males and normally long and narrow
in females. The pygidium in males is also elongate and narrowed at the
apex in these same species. In some species, especially gibbosus, the ab-
dominal process which partially covers the apex of each posterior femur is
usually narrower and the sides more nearly parallel in females. There are
differences in texture of the pygidium between males and females of
Neochlamisus, but these are difficult to describe because of variation within
species.

Only one species (Chlamisus joveolatus) shows any sexual dimorphism
in color pattern, but it is so slight that it is of questionable value in dis-
tinguishing the sexes. The face of both sexes is partially yellow on the
clypeus and lower frons. This coloring usually extends higher on the frons
in males than in females. A much more striking characteristic is the wide
presternum and prosternal process found in females compared to the
narrow subtriangular one in the males.

In Chlamisus maailipes the male antennae reach to the middle of the
metasternum and have elongate segments. The female antennae barely
reach the posterior margin of the presternum. In addition, the presternum
and the metasternum between the coxae are covered with long dense hairs
in the male, while in the female the hairs are short and scattered. The hairs
on the presternum of the male are decumbent, while those on the meta-
sternum are erect.

A further discussion of sexual dimorphism in Neotropical species will
be found in Monros (1952).

INDIVIDUAL VARIATION

Moderate size and color variation among individuals is common in
most species and is easily recognized. Size variation as sexual dimorphism
was discussed in the previous section. Small color variations in the
"metallic" species are very difficult to describe since changes in the direction
of the light source can cause certain areas or spots to appear different. Oc-
casionally major variation is related to the method of preservation or the

890 The University Science Bulletin

age of the specimens. Some variation may be due to the insects' having fed
on different host plants, but there is no conclusive evidence of this. The
extent of coloration in maculate species is subject to considerable variation,
but the species can be recognized to some extent by coloration alone.

The tubercles and carinae follow the generalized pattern outlined for
Exema (Karren, 1966), but it is difficult to describe accurately a particular
tubercle or carina for a species because of extensive variation in sculpturing.
The commonest condition is described for each species, and the reader is
cautioned to realize that there may be considerable variation. A particular
tubercle can be small or large, blunt or sharply carinate, and may be con-
nected to another tubercle.

The pattern of punctures and foveae on two individuals is never exactly
the same, although some specific uniformity can be found. For example,
those in one species may be in general deeper, shallower, larger or smaller
than those in a related species. When described as part of a complex of
characters of a species, punctation can be used in specific differentiation.
In a few species some variation has been found between males and females
in the fovea on the last abdominal segment.

Intraspecific variation in the genitalia is usually very slight, while the
genitalia of many closely related species are so similar that they are difficult
to use in species diagnosis. An interesting exception has been studied by
Brown (1944) in Neochlamisus bebbianae. He found a case of striking
dimorphism in the male aedeagus in one population, the apex being either
evenly rounded or bluntly pointed. In addition, differences were found in
the ejaculatory guide and other sclerotized structures of the internal sac.
An additional case of dimorphism of this type is found in N. moestificus.

Each species could be described as having characters varying normally.
Most individuals of a species would fall fairly near the mean of variation
of a character, but a few individuals with extreme conditions of the
character under consideration might overlap with extreme individuals of
another species. In a few cases I have been able to relate variation in a species
to geographical distribution. This has been noted in the discussions.

In the descriptions, I have tried to describe for each character the com-
monest form, based on the available samples. Variation and exceptions are
described in the discussions that accompany the descriptions. Accurate
identification and understanding of the variability of a species is always
easier when a large series is studied, especially when there are many in-
dividuals from one locality and preferably from one host.

LIFE HISTORIES

Life history studies have been investigated for species of the genera
Neochlamisus and Exema. All species appear to have similar life histories

A Revision of the Subfamily Chlamisinae 891

and case-bearing habits in the larvae. The same is true for the Neotropical
species that have been studied. Each has a definite season of the year in
which oviposition takes place. Short periods in which certain species can
be taken in abundance and the simultaneous rarity of other species suggest
that the season of adult activity for one species may be quite different from
that of another in the tropics. Activity is probably correlated with growth
and maturity of food plants. North American species are more similar in
their seasonal activity, probably because climatic conditions to a certain
degree correlate plant growth. Some seasonal differences in peak abundance
have been noted for southern species which appear to have a more extended
season of activity, and are found during the entire year.

The following information about life histories is taken from published
data and from field and laboratory observations. Temperate North Ameri-
can species hibernate as adults and begin their egg laying in late April
and early May. I have observed oviposition throughout the summer months
in species of Exerna in Kansas. Brown (1943) noted that the species he
studied reached a peak of ovipositing the first week in June. One species
in Kansas, Neochlamisus platani, is just beginning its egg laying in late
May and early June. It can lay 10-15 eggs per day for short periods of time.
Species of Exema lay only 1-3 eggs per day but may continue for a month
in the laboratory.

Eggs are laid by first attaching a short stalk to the leaf. The female
appears to use this stalk to pull the elongate yellowish or orange egg from
the tip of the abdomen. The stalk is probably continuous with the chorion
of the egg (Fig. 1). She then covers the eggs with her feces. Small flat-
tened plates of greenish fecal material are squeezed between genital sclerites
so that several are used to build the first row encircling the base of the egg.
The stalk is twisted (Fig. 2) when the egg is rotated by the hind legs
during this process. As one row is completed another is begun until the
entire egg is covered and a flat top is added. Slight variations are found
in the different species and genera. Shortly after the covering is completed
the case hardens and turns a darker color. The process requires 15-20
minutes in the North American species that have been observed.

From six to ten days are required for the eggs to hatch. Humidity is
an important factor in determining the hatching time in the laboratory.
Upon hatching the larva chews away the flat top of the egg case. Con-
trary to my previously published observations (1964), the larva usually
also chews away the end of the stalk inside the egg covering. The covering
is then pulled away from the stalk and becomes the beginning of the larval
case. Occasionally the larva leaves the case and chews away the stalk from
the outside. During mid-summer, larvae of all stages can be taken on
the host plants.

892

The University Science Bulletin

A.

Figs. 1-8. Neochlamisus platani, immature stages. 1, egg on stalk; 2, fecal covered egg; 3,
2nd instar larva and additions to case; 4, mature larva in case; 5, prepupal stage; 6, pupal stage;
7, front leg of mature larva; 8, antenna of mature larva. Upper scale Figures 1-3, middle scale
Figures 4-6, lower scale Figures 7-8.

A Revision of the Subfamily Chlamisinae 893

The larva is somewhat U-shaped inside the case (Fig. 4). The legs
are long and protrude from under the edge of the case (Fig. 3) as the larva
moves over a surface. Feces, plant hairs, and other debris are used to
enlarge the case. At each molt the case is attached firmly to the substrate
for 1-2 days before the larva again begins to feed. The mature larva (Fig.
5, 7-8) also attaches its case firmly to the substrate before pupation (Fig. 6),
usually in a protected area such as the axil of a leaf.

Late in the summer, adults begin to emerge from pupal cases that have
been attached to plants for 2-3 weeks. The adults are fully colored and
sclerotized, since they remain in the pupal case several days before emerg-
ing. There is apparently only one generation a year in the North American
species. Adults collected early in the summer readily copulate, while those
collected in August and September have never been so observed. More
detailed information about specific life histories will be given, when avail-
able, with the species treatments.

Eggs, larvae and pupae collected in the field frequently produce para-
sitic Ichneumonidae, Chalcidoidea and Chalcididae. Available information
on specific parasites is included in the species treatments.

Adult chlamisines are sometimes used to provision the nests of certain
wasps of the genus Cerceris. Scullen (1965) cited two examples. The
beetles have been identified as Neochlamisus insularis and N. bebbianae.

FOOD PLANTS

Earlier studies, especially by Brown (1943, 1946, 1952, 1961), indicated
that most North American species have strong monophagous tendencies
and limited geographic distributions. My studies have shown that distribu-
tions are much more extensive than was thought by Brown and that several
food plants are used by some of the species. A few are generally mono-
phagous throughout their range, but others are commonly found on sev-
eral host plants. More complicated are those which occur on a particular
host in one locality but a different host in another locality. In each genus
or species group there seem to be species of each type.

Brown has published reliable host plant records for all the species in-
cluded in his studies. To these, I have added a considerable number in
this paper.

Previous to the work of Brown, a number of authors had reported
host plants for several species usually under the names Chlamys plicata or
C. gibbosa. Care must be used in referring to these records because of the
previously confused nomenclature of the North American Chlamisinae.

Specimens are often taken on plants that are not used as food. Larvae
especially wander to neighboring plants to pupate. The most reliable
information comes from actual observation of feeding by adults and larvae
or collection of a series of adults or larvae on the same plant. Food plants

894 The University Science Bulletin

from which a series of at least three specimens have been collected at one
time and place are hereinafter identified with the word "series" in paren-
theses after the name of the plant. Those reported in the literature but not
otherwise verified are identified with the word "literature" in the same
manner. Plants on which larvae have been taken are also indicated.

SYSTEMATIC POSITION AND RELATIONSHIPS

The systematic position of the subfamily has been discussed by several
authors (Spett and Lewitt, 1926; Boving and Craighead, 1930-1931; Zia,
1936; Chen, 1940; Powell, 1941; Monros, 1952; Varma, 1954, 1955; de Monte,
1957), using different bases, including female genitalia, male genital cap-
sules, larvae, and wings. In general, they agree that Chlamisinae have
definite affinities with the other camptosomates, those chrysomelids in which
a larval case is constructed for all of the immature stages.

Many different opinions exist as to the taxonomic rank of the chlami-
sines. There has been a general tendency toward raising the taxon from tribe
to subfamily to family. There are the usual advantages to retaining sub-
family status for this group, as pointed out by Monros (1952). I have
followed his classification because it reflects more closely the present under-
standing of the phylogenetic relationship of the group to the rest of the
chrysomelids.

For the most part, the elytral sculpturing is very distinctive for the
group, but several Australian genera of Cryptocephalinae (Brachycaidus,
Lacnobothra, etc.) have similar sculpturing. The enlarged asymmetrical
mandibles in males of Pseudochlamys resemble those found in Clytrinae
(Clytrini, Megalostomini). Although the adults of Lamprosominae differ
in many respects from Chlamisinae, the larvae are very similar, and the
lamprosomines are considered part of the camptosomates closely related
to the Eumolpinae and Chlamisinae. They are intermediate in many re-
spects between the two groups.

Powell's (1941) study of the male genitalia of chrysomelids showed a
close relationship between the subfamilies Cryptocephalinae and Clytrinae.
Powell was less certain about the relationship of the Chlamisinae to these
two groups but believed that further study would prove their close rela-
tionship to Cryptocephalinae and Clytrinae. He examined only one species,
Exema dispar, which proved to have several specializations. Powell used
the presence of subapical hairs of the aedeagus, a characteristic found only
in the camptosomates, to show the relationship of Chlamisinae to the other
two subfamilies of the section. The Y-shaped spiculum gastrale and teg-
men are found in many genera of Eumolpinae, Chrysomelinae, Donaciinae,
Orsodacninae, and Criocerinae, but there is a closer similarity among those
of the camptosomate genera studied by Powell.

Further study has indicated that the species of Exema found in North

A Revision or the Subfamily Chlamisinae

895

Figs. 9-20. Male aedeagus, dorsal and lateral aspects. 9, Pseadochlamys semirujescens; 10,
P. megalostomoides; 11, Chlamisus flavidtis; 12, C. arizonensis, 13, C. texanus; 14, C.
fovelatus; 15, C. qttadrilobatus; 16, C. huachucae; 17, C. nigromactdatus; 18, C. maculipes;
19, Diplacaspis prosternalis ; 20, D. paradoxa.

896 The University Science Bulletin

America are some of the most specialized chlamisines with regard to the
genitalia. The genus Pseudochlamys and some species of Chlamisus show
a much closer relationship to the other camptosomates and better support
the conclusions reached by Powell.

The terminal position of the apical orifice of the male aedeagus is a
more generally occurring characteristic throughout the chrysomelids than
the Y-shaped spiculum gastrale and tegmen. It is also found in most of the
chlamisines that I have studied, however, as pointed out by Powell, its
position is subcentral in Exema. I have also found this modified condition
in many species groups of the chlamisines, including the genus Neochlami-
sus. This is interpreted as a specialized state. This characteristic, correlated
with other morphological similarities between species, is one of the im-
portant criteria for the species groups treated in this paper.

SYSTEMATIC TREATMENT
CHLAMISINAE Gressitt, 1946

Chlamydeae Lacordaire, 1848:636-643.

ChlamydaeBdy, 1865:58.

Chlamydes Chapuis, 1874:193.

Chlamydinae Gemminger and Harold, 1874:3301.

Chlamydidae Jacoby, 1881:73-74.

Chlamydini LeConte and Horn, 1883:240-241.

Fulcidacina Jacobson, 1924:239.

Fulcidacinae Schaeffer, 1926:181-182.

Fulcidacini Leng and Mutchler, 1927:44, footnote 3.

Fulcidacidae Navajas, 1944a:75-76.

Chlamisinae Gressitt, 1946:84.

Chlamisidae Brues, Melander and Carpenter, 1954:577.

Type genus: Chlamisus Rafinesque, 1815 (=Chlamys Knoch, 1801, not
Bolten, 1798). Chlamys is recognized as the type of the subfamily since
Lacordaire (1848) used it as the basis of the family group name, Chlamy-
deae. The replacement name, Chlamisus becomes the type genus to retain
the original application of the taxon. The name Fulcidacinae, based on
the generic name Fulcidax, as proposed by Jacobson (1924) is invalid
since it is not based on the valid name of the original nominal type genus.

Review of Taxonomy
The chlamisine classification proposed by Lacordaire (1848) has been
used by entomologists for over 120 years. It consisted of seven genera, with
the large genus Chlamys divided into 14 species groups mainly on the basis
of antennal form, body shape, and color. Although several authors have
pointed out the shortcomings of this system, no one has attempted to
completely revise Lacordaire's classification. During the years that Monros
worked with this group, he established two new genera for species that
Lacordaire had included in Chlamys and Exema, leaving Lacordaire's basic

A Revision of the Subfamily Chlamisinae

897

Figs. 21-37. Neochlamisus, male aedeagus, dorsal and lateral aspects. 21, memnonhis;
22, moestificus (dorsal aspects show extremes in variation) ; 23, velutintis; 24, scabripennis ; 25,
subelatus; 26, gibbosus; 27, bimaculattis; 28, insularis; 29, platani; 30, bebbianae; 31, ettbati;
32, jragariae; 33, comptoniae; 34, assimilis (dorsal aspect only); 35, chamaedaphnes; 36,
cribripennis; 37, tubercidatus.

898 The University Science Bulletin

classification unchanged. As new species have been described, it has been
found that the characters used by Lacordaire sometimes resulted in un-
related forms being placed close together in the classification. Additional
difficulty has resulted because the descriptions and terminology were not
always clearly understood.

A major problem with Lacordaire's classification is overlooked or dis-
regarded characters. Ftdcidax was characterized by having simple tarsal
claws. Although this is a valid character for this genus, several species of
North American Exema examined by Lacordaire also have the same condi-
tion. The one species of Diplacaspis was characterized by the unique meta-
scutellum, although at least one species, and perhaps more, of Chlamisus
examined by Lacordaire also have an exposed second scutellum. Less sig-
nificant is the case of Pseudochlamys, which is characterized by large asym-
metrical mandibles. A species of Chlamisus with similar mandibles was
recognized by Lacordaire but thought to be a deformed specimen. Addi-
tional material and a study of the genitalia indicate that the mandibular
condition in Pseudochlamys is more variable than originally supposed.
Eventually several other species with buccal modification will be included
in this genus.

The present revision includes all of the species found in America north
of Mexico. Lacordaire's classification is followed where I feel it reflects
the relationships of the species. A study of the male genitalia of North
American as well as many Central and South American species indicates
relationships among genera and species groups that are not evident in a
study of antennae, body shape and color, upon which characters the classi-
fication of Chlamisinae had previously been based. It has been my purpose
to show correlation of these characters with as many others as possible,
especially the genitalia. The resulting classification probably reflects more
closely the phylogenetic relationship of the groups within the subfamily.

Two related species groups formerly included in Diplacaspis and Chlami-
sus are found in North America; one is characterized by an exposed meta-
scutellum and the other by a striate pronotum. Similarities in the male
and female genitalia, antennae, and sculpturing suggest that the two groups
are congeneric and a new genus Neochlamisus is proposed. Intermediate
species and similar genitalic characters in the genus suggest species groups
rather than subgeneric status for the two groups. It was felt that any other
treatment of the species would be contrary to the present understanding
of the relationships.

The metascutellum of Neochlamisus shows several degrees of exposure.
It may be always broadly exposed, within a species, or variable from nar-
rowly exposed, to rarely exposed, to never exposed. Exposure of the meta-
scutellum is closely correlated in the velutinus species groups with the
presence of two velvety spots on each elytron and single tibial spines in both

A Revision of the Subfamily Chlamisinae

899

Figs. 38-49. Male ejaculatory guide, dorsal and lateral aspects. 38, Pseudochlamys semiru-
jescens; 39, P. megalostomoides (also ventral aspect); 40, Chlamisus flavidus; 41, C.
arizonensis; 42, C. texanus (also ventral aspect); 43, C. joveolatus; 44, C. quadrilobatas; 45, C.
huachucae; 46, C. nigromaculatus; 47, C. maatlipes; 48, Diplacaspis paradoxal 49, D.
prosternalis.

900 The University Science Bulletin

sexes. The size of these spots also varies continuously from very large to
absent. Both exposed metascutellum and velvety spots are found occasionally
in the gibbosus group but never both in the same species.

The male and female genitalia respectively are more similar from species
to species in Neochlamisus than in any other North American chlamisine
genus. The specialized male structures are distinctive. The female sperma-
theca is similar throughout the genus and distinct from those of related
genera.

The North American species left in Chlamisus belong to more than
one genus according to the above standards, but until there has been
further study of the Mexican species it is better not to propose new genera.
It would be preferable to establish new genera when there is a clearer
understanding of the included species and their taxonomic position in the
subfamily.

Pseudochlamys is represented by one new species in America north of
Mexico. Exema has been treated previously (Karren, 1966). Diplacaspis is
represented by one species quite different in shape and sculpturing from
the type species, but the two have several characters in common including
the genitalia that suggest the relationship. The male genitalia in species placed
in Diplacaspis are less similar than those of species of Neochlamisus to
which Diplacaspis is related, but several characters indicate relationship
close enough for the species assigned to Diplacaspis to be included in the
same genus as the type species. Several Neotropical species of Diplacaspis
seem to be intermediate between the type species and the one species from
our area.

I have been unable to associate two names proposed for North American
species of Chlamisinae with any identified species. These are:

Chlamys tuberosa Knoch, 1801. Type locality: North America (prob-
ably Pennsylvania, since it came from Melsheimer's collection). Found on
chinquapin (probably Castanea pumila). There are no other records of
North American Chlamisinae on this plant, although several species have
been found on related plants.

Chlamys polycocca Lacordaire, 1848. Type locality: United States.
Lacordaire stated that his species was closely related to but distinct from
his phcata. I am unable to recognize which one of the gibbosus group it is.

The published descriptions are inadequate, the type localities are too
general, and no type specimens have been located. Both tuberosa and poly-
cocca have for many years been considered synonyms or varieties of gib-
bosus (or plicatus) by most authors. The names gibbosus or plicatus have,
however, been used for a number of closely related species for many years,
so that the treatment of tuberosa and polycocca as synonyms of gibbosus
is of questionable value. In the interest of stability and common usage I
have not referred these names to any of the North American species.

A Revision of the Subfamily Chlamisinae

901

Figs. 50-66. Neochlamisus, male ejaculatory guide, dorsal and lateral aspects. 50, veltitinus;
51, memnonius ; 52, tnoestifiais; 53, scabripennis; 54, subelatits; 55, gibbosus; 56, bimaculatus;
57, instilaris; 58, platani; 59, bebbianae; 60, cubati; 61, fragariae; 62, comptoniae; 63, assimilis
(dorsal aspect only) ; 64, chamaedaplwes; 65, cribripennis; 66, tuberculatus.

902 The University Science Bulletin

Key to the Genera of North American Chlamisinae

la. Tarsal claws bifid (Fig. 126); prosternal process widened between

mesocoxae; metascutellum broadly exposed Diplacaspis

b. Tarsal claws simple or broadly toothed; sides of prosternal process
parallel, convergent or serrate, if slightly widened between meso-
coxae then metascutellum not exposed 2

2a. Mandibles of males enlarged, normal in female; prosternum strongly
and abruptly constricted behind anterior margin, posternal process

more than % as long as prosternum Pseudochlamys

b. Mandibles of males and females similar; prosternum gradually con-
stricted behind anterior margin or prosternal process about l/2
length of prosternum 3

3a. Male ejaculatory guide asymmetrical, with sheath; frons in emargina-
tion of eye with yellow spot (except vein tin us which has two

velvety spots on each elytron) Neochlamisus

b. Male ejaculatory guide symmetrical, without sheath; frons in emargi-
nation of eye without yellow spot (face may be all or partly yellow,
with yellow area extending into emargination, but elytra do not
have velvety spots) 4

4a. Males with spines or spinulae on first ventral abdominal segment,

sutural serration of elytra complete Exema~

b. Males without spines or spinulae on first ventral abdominal segment,
sutural serration of elytra incomplete, suture straight immediately
behind mesoscutellum (except quadrilobatus) Chlamisus

Pseudochlamys Lacordaire, 1848

Pseudochlamys Lacordaire, 1848:644.

Type species: Pseudochlamys megalostomoides Lacordaire (monotypic).

History: This genus was established by Lacordaire for a unique species
with striking characteristics. Two more species were added when Bowditch
(1913) transferred Chlamys seminigra Jacoby to the genus and described
as new P. rujescens. At the time there was some doubt about the proper
generic designation for both species. I have examined the type of rujescens,
which is very similar to megalostomoides , and agree with those authors
who place it in this genus. There has probably been some mistake con-
cerning the drawing of rujescens supposedly made by Bowditch, in Achard
(1914). It has very little resemblance to the specimens I have examined
or to any other members of the genus. The specimens of seminigra in the
Bowditch collection are congeneric with Pseudochlamys. The only other
species described is P. bellicosus Monros. At least two other species of
Chlamisus that I have examined, from Mexico, probably should be placed
in this genus, but definite action will not be taken until these species are
studied further.

Further treatment of Exema is excluded from this paper. For a revision of it see
Karren, 1966.

A Revision of the Subfamily Chlamisinae

903

Figs. 67-93. Female spermatheca. 67, Pseudochlamys semirufescens; 68, P. megalostomoides;
69, Chlamisus flavidus; 70, C. arizonensis ; 71, C. joveolatus; 72, C. quadrilobatus; 73, C.
hiiachncae; 74, C. nigromaculatus ; 75, C. maculipes; 76, Diplacaspis prosternalis; 77, Neo-
chlamisus velutinus; 78, N. memnonins; 79, A7, moestificus ; 80, A/, scabripennis ; 81, A7.
subelatus; 82, A7, gibbosus; 83, A7, bimaculatus; 84, A7, insttlaris; 85, A7, platani; 86, N.
bebbianae; 87, A7. eubati; 88, AT. fragariae; 89, A7, comptoniae; 90, A7, assimilis; 91, A7.
chamaedaphnes; 92, A/, cribripennis; 93, A7. tuberctdatus.

904 The University Science Bulletin

Diagnosis: The enlarged mandibles, genae and clypeus in males, shape
of the prosternum, male and female genitalia and color pattern serve to
identify this genus in North America.

Description: Males: 3.7-4.8 mm long; 2.1-3.0 mm wide. Females: 4.1-5.0
mm long; 2.4-3.0 mm wide. Coloration nonmetallic, yellowish or reddish
brown with darker markings; underside and legs may be partially black.

Head enlarged or not, but usually visible from above (especially in
males); pronounced sexual dimorphism: males with enlarged, sometimes
asymmetrical mandibles, genae enlarged and elongated, clypeus enlarged
and excavated; females more normal for subfamily, with very little enlarge-
ment of structures. First antennal segment (scape) in males sometimes en-
larged and flattened, segment 2 (pedicel) subglobular, 3 smaller than 2 and
subtriangular, 4 larger than 3 and subtriangular or subrectangular, 5-10
subrectangular, 11 subconical. Prothorax appearing reticulate, coarsely
punctate with deep round punctures; evenly rounded or slightly gibbose,
medial sulcus faint, carinae small or faint, tubercles absent, posterior lobes
normal. Mesoscutellum normal, faintly granulate or shiny, concave or
slightly carinate, posterolateral angles slightly reflexed (posterior elonga-
tions in seminigra). Metascutellum not exposed. Elytra moderately to
coarsely punctate, tubercles and carinae faint or absent, with some of normal
sculpturing visible. Pygidium moderately to coarsely punctate, nontubercu-
late and noncarinate, normal depressions faint, slightly convex. Prosternum
expanded beneath enlarged mandibles in male, normal in female. Ventral
thoracic segments foveolate, abdominal segments moderately to coarsely
punctate, mid-ventral fovea of last segment small or absent in male, small
to moderate in female. Tibial spines present or absent.

Aedeagus long and tubular, slightly curved, with apical orifice terminal;
ejaculatory guide variable, in one piece, short, tubular. Spermatheca varia-
able but usually long and slim without enlargements; duct not long and
convoluted.

Distribution: This genus has been reported from Brazil, Paraguay,
northeastern Argentina, Colombia, and Costa Rica. I have additional
records from Panama, Honduras and southern Mexico. With the descrip-
tion of P. semirujescens n. sp. the distribution extends into the southern
United States.

Systematic position: Many characteristics of the genus suggest that it
is one of the more generalized of the subfamily. Species of this group
have a definite affinity with the Clytrinae (especially Megalostomini) be-
cause of the similar sexually dimorphic characters. The poorly defined
tubercles and carinae and absence of the pronotal gibbosity are generalized
characteristics seen in closely related subfamilies. The female genitalia are
of the generalized camptosomate type, while the male genitalia have many
generalized characteristics found throughout the Chrysomelidae.

A Revision of the Subfamily Chlamisinae 905

Discussion: Four species have already been described in this genus or
referred to it: megalostomoides, seminigra, rufescens, and bellicosus. The
only species found in North America is new and extends the distribution of
the genus from southern Mexico into the southern United States.

The above generic description is based upon specimens of megalosto-
moides and semintfescens and information of other species given in the
literature. Figures of megalostomoides are provided (Figs. 10, 39, 68) for
comparison with those of semintfescens.

The taxonomic value of enlarged mandibles and other sexually di-
morphic characters to identify members of this genus is not completely
understood at this time. There are at least two Mexican species apparently
of Chlamisus with enlarged mandibles, and further study may prove them to
belong to Pseudochlamys. Several species of Chlamisus have many other
characteristics similar to Pseudochlamys, but the males lack the enlarged
mandibles. Many of the characteristics ascribed to Pseudochlamys by
Lacordaire, Jacoby, Achard, and Monros have already proven to be specific
characters rather than generic ones. As further species of Pseudochlamys
are described and the genus becomes better understood, its place in rela-
tion to other genera in the classification of Chlamisinae should become
more evident.

Pseudochlamys semintfescens new species

Diagnosis: Since this is the only member of the genus in North America,
it can be identified easily by the generic characteristics, and by the reddish
brown markings on the yellowish dorsum. Both the male genitalia and
female spermatheca are diagnostic for the species.

Description: Male holotype: 4.1 mm long; 2.4 mm wide. Color of
dorsum yellow with two long, broad, reddish brown spots covering most
of the pronotum leaving a central, longitudinal, narrow, yellow stripe,
yellow anterior border and lateral spots; depressed areas of elytra including
punctures reddish; vertex of head faintly reddish; mandibles dark brown;
prosternal margins reddish; undersurface including legs mostly pale red-
dish with faint yellowish red areas; antennal segments 1-3 reddish, 4-6
yellow, 5 reddish, apical four segments dark brown; tarsal claws and tibial
spines black.

Head moderately to coarsely punctate, punctures large and round;
clypeus and genae tumid, less distinctly punctate than frons; mandibles
large, mouthparts symmetrical; antennal segment 3 subtriangular and
smaller than 2, 4 subtriangular with greatest width equal to width of fol-
lowing segments, 5-10 subrectangular, 11 subconical (Fig. 96). Pronotum
coarsely and deeply punctate, gibbosity broad, faintly developed; tubercles
and carinae absent; posterior lobes smooth. Mesoscutellum shiny, slightly
concave, its posterolateral corners slightly reflexed, posterior margin con-

906

The University Science Bulletin

Figs. 94-96. Pseitdochlamys semirujescens. 94, dorsal and lateral aspects; 95, presternum;
96, antenna. Upper scale Figure 94, lower scale Figures 95-96.

cave, anterior margin tuberculate, lateral margins slightly convergent. Ely-
tra (Fig. 94) moderately to coarsely and deeply punctate, punctures round
and setigerous; tubercles and carinae small or absent; small transverse
su-2 carina joins long elevated area containing two subequal rows of punc-
tures beginning at su-la and extending to small su-3 tubercle; small sub-
marginal carina extends from umbone to lateral area of su-3b, small carina
extends from su-2a to submarginal carina; slightly depressed discal area,
small areas anterior and posterior to su-2 depressed; anterior 1/2 of suture
straight, posterior 4/5 serrate. Pygidium slightly convex, finely punctate,
depressions faint, punctures round and deep. Presternum (Fig. 95) broadly
expanded behind mouthparts, sides strongly and abruptly convergent, proc-
ess long and slim, about 3/4 as long as entire presternum, apex curving
toward base of metasternum, central area slightly depressed. First and last
abdominal segments moderately punctate; small mid-ventral fovea on last

A Revision of the Subfamily Chlamisinae 907

segment; foveal hairs few and scattered. Anterior and middle legs with
small tibial spines.

Aedeagus (Fig. 9) short, curved, tubular, with fine scattered, lateral and
ventral hairs near apex, apical orifice subterminal; ejaculatory guide (Fig.
38) short and tubular.

Female allotype: 4.3 mm long; 2.5 mm wide. Similar to male except
mandible small as in other chlamisines; mid-ventral fovea of last abdominal
segment larger than in male and with many hairs; no tibial spines; antennal
segment 7 yellow, 8 light brown, 9-11 dark brown.

Spermatheca (Fig. 67) irregularly S-shaped, cornua evenly tubular to
apex, apical 1/2 uniformly curved; collum suddenly recurved and expanded
before apex; spermathecal duct short and explanate at attachment to com-
mon oviduct.

Types: Male holotype, Arivaca (Pima Co.), Arizona, 26 July 1941, E. L. Todd; female
allotype, same locality and date, R. H. Beamer (both UK). Paratypes, all 17 from Arizona:
2 $ , Arivaca, Pima Co., 14 July 1966, A. G. Hamilton, grass and mesquite, (JBK); 1$ &
2$, Patagonia Mts. (Santa Cruz Co.), 15 July 1952, 20 Aug. 1940, 4 Aug. 1953, D. J. &
J. N. Knull (OSU, JAW); 15, Patagonia (Santa Cruz Co.), 29 July 1945, F. H. Parker
(UAR); IS & 1 9, Santa Rita Mts. (Pima Co.), 12 June 1933, R. H. Beamer (UK); 19,
same locality, July, F. H. Snow (UK); 1 9, same locality, June, D. K. Duncan (UAR); 1 <$,
Oak Creek Canyon (Coconino Co.), 9 July 1941, L. H. Banker (UK); 1 6\ Baboquivari Mts.
(Pima Co.), F. H. Snow (UK); 1 S , same locality, 8 July 1935, E. D. Ball (UAR); 1 $ ,
Globe (Gila Co.), 15 July 1943, F. H. Parker (UAR); 1 9, Santa Catalina Mts., Pepper Sauce
Canyon (Pima Co.), 16 Aug. 1924, E. P. Van Duzee (USNM); 1 9 , Nogales (Santa Cruz Co.),
25 June 1933, J. Russell (UK); 19, same locality, Sept., A. Kocbele (CAS). (Map 1.)

Discussion: The type and paratypes were the only specimens available
for study, except 1 6 and 1 9 in the Fall Collection (MCZ) (Santa Rita
Mts., July, F. H. Snow). They are labeled Chlamys arizonensis Linell. The
name semintfescens describes the reddish markings of the beetle.

In shape, sculpturing and color pattern, this species has many similarities
to megalostomoides. The most striking similarity is in the females; the
males of semintfescens appear to be specimens of megalostomoides with
poorly developed mandibles and a slightly excavated clypeus. Characters
differentiating the two species are found in the male and female genitalia.

Specimens of megalostomoides were thought to be very uncommon
until recently. Recognition of the host plant and additional collecting reveals
that they are common from southern Brazil to southern Mexico on Wal-
theria americana and possibly one other species of Waltheria in Brazil
(Bokermann, in correspondence). The strict monophagy observed for this
species may be some indication of a similar relationship of semirufescens
to its possible food plant.

This beetle has been misidentified as Chlamisus arizonensis in most of
the borrowed material. The superficial similarity of semirufescens to the
pale species of Chlamisus has also caused some confusion of it with those
species, but the male and female genitalia are quite different in each species.

908 The University Science Bulletin

A close examination of the size, shape, sculpturing and coloration also
furnishes numerous characters for distinguishing semirujescens from the
pale species of Chlamisus, even if males are not available or the genitalia
are not examined.

The small number of specimens available from a limited area reveals
less variation than is normally found in species of the subfamily. Indi-
viduals vary in the darkness of the color markings but not in the pattern.
There is little variation in size, even between sexes, among the specimens
examined.

There are no host data for this species except a vague record of two
specimens collected by A. G. Hamilton from sweepings of "grass and
mesquite."

Chlamisus Rafinesque, 1815

Chlamys Knoch, 1801:122-128 (not Holten, 1798:').

Chlamisus Rafinesque, 1815:116.

Arthrochlamys H. Ihering,4 1905:642.

Boloschesis Jacobson, 1924:239.

Bolochesis Leonard, 1928:463, footnote 55a (misspelling).

Type species: Chlamys foveolata Knoch, designated by Navajas, 1946.
Navajas did not select tuberosa, the first species described by Knoch as the
type species because there is a question about its identity. Two other species,
Clythra monstrosa Fabr. and Bruchus gibbosus Fabr. had been designated
earlier as types of the genus, but are invalid because neither was an originally
included species.

History: Chlamys was generally recognized as the valid name of the
genus for about 130 years, during which time nearly 400 new species were
described. Three authors during this period discovered that Chlamys was
a junior homonym of Chlamys Bolten (179cS), a mollusk (available accord-
ing to Opinion 96 of the International Commission on Zoological Nomen-
clature).

Boloschesis, the youngest of the synonyms of Chlamys Knoch, was the
first name used in a major revision (Schaefifer, 1926). Brown (1943), in his
revision of the Canadian species, recognized Arthrochlamys as the valid

3 Sometimes listed as Roding or Bolten in Roding, 1798. The homonyms listed in Neave,
1939, p. 703, are as follows:

"Chlamys Bolten 1798, Mus. Bolten., 2, 161.— Moll. Cf. Myochlamys Ihering 1907.

Chlamys Knoch 1801, N. Beitr. Insectenkunde, (I) 122. — Col. (See Arthrochlamys

Ihering 1904 & Boloschesis Jacobson 1942).

Chlamys Young & Young 1877, Ann. Mag. nat. Hist., (4) 20, 429. — Spong.

Chlamys Beneden 1892, Bull. Acad. Sci. Belgique, (3) 23, 227.— Crust."

4 Sometimes listed as Ihering and Ihering. Although they coauthored the article, the
authorship of the new names is clearly stated on page 679.

A Revision of the Subfamily Chlamisinae 909

name. The oldest available name, Chlamisus, was not used until 1946 by
Gressitt. Neave (1939), previously recognized the name Chlamisus Rafi-
nesque (1815), and incorrectly stated that it was an emendation of Chlamys
Berthold (1827); however, Berthold was presumably merely using Knoch's
name of 1801.

It is not entirely clear that Rafinesque (1815) was proposing Chlamisus
as a substitute name for Chlamys. Other genera of Chrysomelidae for which
Rafinesque appears to have proposed replacement names in the same
manner have been entirely overlooked by subsequent authors. They are
not even recognized as junior synonyms as far as I can determine. Never-
theless, I am accepting Chlamisus Rafinesque as a replacement name for
Chlamys Knoch (1801). Chlamisus has been generally accepted since
Gressitt (1946) used it as the valid name for the largest chlamisine generic
group.

In 1940 and 1942, Chujo synonymized the name Chlamys under Exema
because of the difficulty in separating these taxa using Lacordaire's key.
He later (1955, 1956) recognized the validity of Chlamisus but retained his
concept of one taxon for the two genera. In these same papers he listed
Myochlamys Ihering, 1905 (this should be 1907) as a synonym of Chlamisus
or Exema, but Myochlamys was proposed as a replacement name for
Chlamys Bolten in Mollusca and should not appear as a synonym of any
coleopterous group.

The name Chlamisus has been used for over 450 of the 577 species of
the subfamily. In many respects it has been a "catch-all" genus rather than
a clearly defined morphological taxon. I have used the name Chlamisus
in a provisional sense pending revision of the Neotropical fauna. In this
study, the genus includes those species of North America still remaining in
Chlamisus after removal of the gibbosus species group. The reasons for
this change are given in the treatment of Neochlamisus, new genus. The
eight species that remain clearly represent two distinct lines of evolution
on the basis of male genitalia. Numerous other differences characterize
these species so that there is very little difficulty in identifying them. Gen-
eral statements about the genus are difficult. It is especially difficult to make
a description of the genus that will apply to all the included species. The
following diagnosis and description are based only on the known North
American species now included in Chlamisus.

Diagnosis: The presence of a tibial spine on the front legs and some-
times middle legs of males, lack of any tibial spines in females, and absence
of the diagnostic characters of the other genera serve to distinguish the
Chlamisus in North America.

Description: Males: 2.5-6.0 mm long; 1.4-3.9 mm wide. Females: 2.6-6.0
mm long; 1.5-4.1 mm wide. Color varying from yellow with dark markings

910 The University Science Bulletin

to slightly metallic black; underside and legs completely or in part dark
brown.

Head finely to coarsely and deeply punctate, usually depressed or grooved
on upper irons; antennae variable. Pronotum finely to coarsely and deeply
punctate, often reticulate, gibbosity slightly developed and faintly divided
to large and deeply divided; nontuberculate or with rounded tubercles and
carinae. Mesoscutellum subquadrate, granulate to striate or carinate. Elytra
moderately to coarsely punctate, with small to large tubercles and carinae;
anterior suture usually straight (serrate in quadrilobatus) , posterior suture
serrate. Pygidium convex, surface evenly rounded or tuberculate, medial
carina present or absent. Presternum subtriangular to goblet-shaped, coarsely
punctate, concave or slightly convex. First abdominal segment punctate,
foveolate, carinate or reticulate; last abdominal segment moderately or
coarsely punctate or foveolate; mid-ventral area flat to deeply foveolate with
surrounding foveal hairs. One tibial spine on front and sometimes middle
legs of male, no tibial spines in female.

Aedeagus variable; long, curved, tubular, with blunt or rounded apex,
to flattened and explanate, with or without apical hairs; apical orifice termi-
nal or subcentral; ejaculatory guide tubular, variable, diagnostic for each
species, never with a sheath; dorsal wall of internal sac with curved, sclero-
tized hook. Spermatheca variable, diagnostic for each species.

Distribution: The genus Chlamisus as presently understood is found
in all zoogeographical realms, but only in extreme southern and eastern
parts of the Palearctic and the northern tip of Australia. It is most abundant
in the Neotropical realm. In North America it is found from the extreme
Southwest to the southern New England states.

Systematic position: Chlamisus as treated here includes a few species
which resemble Pseudochlamys, a few which share characteristics with
Diplacaspis and Neochlamisus, and a few intermediates. The genus in North
America is much more variable than any of the other genera and its
species show a continuous spectrum of variation in several characters be-
tween Pseudochlamys and Diplacaspis. An example of this is the apical,
subapical, or subcentral position of the apical orifice of the aedeagus. The
loss of apical hairs on the aedeagus in a few species is a specialized condi-
tion; apical hairs are present in most camptosomates.

Discussion: Further study of the numerous Mexican and Neotropical
species of Chlamisus will probably prove the North American species to be
northern representatives of several generic groups. The present evidence
indicates that North American species are products of specialization at the
fringe of chlamisine radiation. Additional information is needed especially
about Neotropical species groups and their relationships before a more
meaningful classification can be proposed for the present genus Chlamisus.

A Revision of the Subfamily Chlamisinae 911

As this knowledge is obtained, new genera will be named for species groups
that deserve generic status.

Lacordaire's groups do not seem to reflect the phylogenetic relationships
between groups. A preliminary examination of examples from most of his
species groups shows that some species of each group have similar genitalia,
supporting the other group characters, hut a few forms exist that are poorly
associated under his groups. Further study is needed to determine the
relationship of species within each group and the phylogenetic relationship
of the groups.

Key to the North American Species of Chlamisus

la. Pronotum and elytra pubescent maculipes

b. Pronotum and elytra glabrous 2

2a. Elytra uniformly black 3

b. Elytra pale with reddish brown markings 4

3a. Head in part yellow; pronotum coarsely punctate joveolatus

b. Head entirely black; pronotum tuberculate and finely striate .... quadnlobatus
4a. Posterolateral slopes of pronotum striopunctate; upper frons with
wide, black inverted V; pronotum and elytra with extensive black

markings nigromaculatus n. sp.

b. Posterolateral slopes of pronotum not striopunctate; head yellow with
pale or small dark spot or stripe on upper frons; pronotum and

elytra with pale markings 5

5a. Pronotal gibbosity not divided; pygidium finely or moderately

punctate on basal half 6

b. Pronotal gibbosity faintly divided; pygidium coarsely punctate on

basal half 7

6a. Head deeply punctate; elytra faintly tuberculate, with small irregular

longitudinal carinae arizonensis

b. Head not deeply punctate; elytra with large rounded tubercles .. flavidus n. sp.

7a. Pygidium coarsely punctate at base to finely punctate at apex, punc-
tures deep; apical 1-2 segments of antennae dark brown; pronotum
and elytra yellow to reddish brown with irregular dark mark-
ings huachucae

b. Pygidium coarsely, deeply and evenly punctate, reticulate; apical 5-6
segments of antennae dark brown; pronotum yellowish green;
elytra pale with large dark spot near scutellum texanus

Chlamisus flavidus new species

Diagnosis: The pale color and dark markings, large and rounded tu-
bercles on the elytra, punctation of the head, pronotum and pygidium are
distinctive features of this species. The aedeagus, ejaculatory guide and
spermatheca are also diagnostic.

Description: Male holotype: 33 mm long; 2.3 mm wide. Color yellow
with black to light brown markings; antennae mostly yellow, gradually

912

The University Science Bulletin

Figs. 97-99. Chlamisus flavidus. 97, dorsal and lateral aspects; 98, presternum; 99, antenna.
Figs. 100-102. Chlamisus arizonensis. 100, dorsal and lateral aspects; 101, presternum; 102,
antenna.

becoming light blown on apical seven segments; faint brown spot on center
of vertex; pronotum with irregular light brown area extending along
lateral slopes of gibbosity, becoming darker on posterior slopes; posterior
margin of pronotum and anterior margins of elytra brown to black; pro-
sternum with light brown to black margins, center of metasternum light

A Revision of the Subfamily Chlamisinae 913

brown, anterior processes black; posterior margins of abdominal segments
light brown; small central area of each leg segment light brown.

Head coarsely punctate, less so on upper frons; antennal segment 3
subtriangular, 4 subrectangular, 5-10 rectangular (Fig. 99). Pronotum
moderately and deeply punctate, with coarse depressed punctures on gib-
bosity, tubercles absent; gibbosity not divided, evenly rounded, postero-
lateral slopes deeply punctate, posterior lobes smooth, tumid. Mesoscu-
tellum carinate, posterior margin slightly depressed and emarginate, sides
convergent anteriorly, posterolateral corners slightly reflexed. Elytra (Fig.
97) with large rounded tubercles, ae, su-1, su-la, su-2a, su-2b, su-3, su-3a
and ap; sc small, others faint or absent; su-1 posterior to normal position;
suture closed anteriorly, posterior 3/4 serrate. Pygidium moderately punc-
tate, depressions faint. Prosternum (Fig. 98) about twice as long as wide,
central area slightly depressed. First abdominal segment roughly foveolate,
last segment finely punctate, with small mid-ventral fovea on its posterior
half. Well developed tibial spine on front and middle legs.

Aedeagus (Fig. 11) tubular, curved, sides constricted before middle,
v/ith fine apical hairs; apical orifice subterminal; median dorsal plate wide;
ejaculatory guide (Fig. 40) tubular, with anterior enlargement below ori-
fice, slightly narrowed behind.

Female allotype: 3.8 mm long; 2.6 mm wide. Coloring and sculpturing
as in male except short dark stripe on center of vertex, posterior slopes of
pronotal gibbosity darker than in male. (Scutellar angle of left elytron
incompletely formed in this individual.) Large, deep mid-ventral fovea
on last abdominal segment. No tibial spines. Spermatheca (Fig. 69) long,
S-shaped; base of cornua long and straight, apical one-third slightly en-
larged; collum half as long as cornua, slightly enlarged and recurved next
to base of cornua; duct long and coiled, not convoluted; spermathecal gland
sac-like, as long as collum.

Types: Male holotype, Hermosillo, Sonora, Mexico, 18 June 1961 (KU); female allotype
and 1 $ paratype, Arizona, Andreas Bolter Collection (INHS, UK). Three additional para-
types from Sonora, Mexico: 1 $ , Guaymas, 3 June 1967, on Larrea divaricata (JBK); 2$,
Huatabampo, 29 May 1955, F. Pacheco (ENA). (Map 1.)

Discussion: The genitalia of flavidus have several similarities to those
of Pseudochlamys semirufescens, including the curved, tubular aedeagus,
the subterminal apical orifice, the tubular ejaculatory guide and irregular,
S-shaped spermatheca. However, the genitalia are more similar to those
of several other species of Chlamisus. This species has been placed with
the species of Chlamisus also because of the small mandibles in the males
and the similar sculpturing.

The host plant of flavidus is uncertain since there has been a plant
associate reported for only one specimen, taken on Larrea divaricata. Only
the six types were available for study.

914 The University Science Bulletin

Chlamisus arizonensis (Linell)

Chlamys arizonensis Linell, 1898:479-480.
Chlamisus arizonensis (Linell); Monros, 1954:24.

Diagnosis: This species is characterized by an elongate form, coarse
and deep punctations, head and pronotum faintly tuberculate, elytra with
small irregular carinae, and moderately, evenly punctate pygidium. The
aedeagus, ejaculatory guide and spermatheca are also diagnostic.

Description: Male: 3.2-3.3 mm long; 1.7-1.8 mm wide. Female: 3.5-3.8
mm long; 2.0-2.1 mm wide. Color yellowish brown with light brown to
black markings, head with dark stripe or spot on upper frons; apical 4-5
antennal segments dark; pronotum with black spot on top of anterior
slope, upper sides of gibbosity darkened by large Y-shaped band extending
back over posterolateral slopes to posterior margin, small dark area anterior
to ae tubercle; posterior margin of pronotum and anterior margins of elytra
black; central area of sterna brown to black; anterior margin of presternum
black, sides brown.

Head coarsely and deeply punctate; antennal segment 3 subtriangular,
4-10 rectangular (Fig. 102). Pronotum coarsely and deeply punctate, punc-
tures depressed and evenly distributed; gibbosity rounded, not divided;
posterior lobes small. Mesoscutellum slightly convex, posterior margin de-
pressed and broadly emarginate, sides slightly convergent anteriorly, postero-
lateral angles acute. Elytra (Fig. 100) faintly tuberculate, ae small, su-2
and su-3 small and broad; faint, irregular, longitudinal carinae separated
by large, deep and depressed punctures; discal depression distinct, small
depression with 4-7 punctures posterior to su-2; anterior elytral suture
closed, posterior 3/4 serrate. Pygidium moderately and evenly punctate,
depressions faint, carinae lacking. Presternum (Fig. 101) less than twice
as long as wide, sides slightly emarginate, gradually converging to apex
of process; central area depressed. First abdominal segment rough, mod-
erately punctate; last segment coarsely punctate medially to finely punctate
laterally; large shallow mid-ventral fovea in both sexes, lateral hairs short
in male, long in female; mid-ventral margin of segments 1-4 with long hairs
in both sexes. Tibial spine on front and middle legs of male; no tibial
spines on female.

Aedeagus (Fig. 12) enlarged before middle, curved and tubular, with
no apparent apical hairs; apical orifice subterminal; median dorsal plate
wide; ejaculatory guide (Fig. 41) tubular with medial anterior projection.
Spermatheca (Fig. 70) L-shaped, cornua narrowed at apex, collum about
twice as long as wide, length equal to width of cornua.

Types: Linell described arizonensis from two females collected in
Arizona by Morrison, one from the Riley collection, the other from the
Hubbard and Schwarz collection. Both of these are in the United States

A Revision of the Subfamily Chlamisinae 915

National Museum type collection (USNM, no. 1298). The former is here
designated the lectotype. H. S. Barber (unpublished) considered these
specimens holotype and paratype respectively. According to the type label,
Linell named this second specimen after Morrison, the collector of the
specimens; but the name was never published. He must have changed the
name after returning the specimens to Schwarz.

Discussion: It was necessary to examine the two type specimens of
anzonensis to distinguish it from several other species in the Southwest
with which it has been confused. Two misidentified specimens in the Fall
collection (MCZ) are P. semirufescens. Borrowed collections contained
six species under the name arizonensis. Monros (1954), after examining the
types of arizonensis and huachucae, synonymized the two, but a study of
additional specimens and the male and female genitalia of the two forms
indicates two species.

Specimens examined (6 males and 10 females): (Map 1)

UNITED STATES: ARIZONA: Cochise Co., 14 June, 8-20 July, 7 Aug.; "Granite Dell,"
11 July.

Chlamisus texanus (Schaeffer)

Chlamys texana Schaeffer, 1906:227-228.

Diagnosis: The unusual greenish coloration, large markings and small
carinae on the elytra, and dark brown five or six apical segments of the
antennae will identify this species. The aedeagus, ejaculatory guide and
spermatheca are also diagnostic.

Description: Male (based on the type and one other male specimen,
both in the United States National Museum) : 3.2 mm long; 2.1 mm wide.
Head reddish brown, yellow between bases of antennae, dark spot on
middle of vertex; antennae light brown, apical 5-6 segments dark brown;
yellowish green pronotum marked on posterolateral slopes with reddish
brown spot joined to light brown arch across anterior summit; posterior
margin of pronotum and anterior margin of elytra black. Elytra yellow to
light brown, each with large dark reddish brown spot anterior to su-2
tubercle. Pygidium greenish yellow; presternum yellow with light brown
margins; metasternum dark brown, becoming lighter brown laterally;
ventral abdominal segments light brown, becoming greenish yellow laterally.

Head moderately and deeply punctate, lower frons and clypeus with
recessed punctures producing a reticulate appearance. Antennal segment 3
subtriangular, 4 subtriangular but much larger than 3, 5 subrectangular,
6-10 rectangular, 11 evenly rounded apically (Fig. 105). Pronotum coarsely,
evenly, and deeply punctate, punctures recessed producing a reticulate
appearance, tubercles and carinae absent; gibbosity evenly rounded, with
very faint median longitudinal groove; posterior lobes granulate. Meso-
scutellum slightly carinate between posterior lobes of pronotum, posterior

916

The University Science Bulletin

106

Figs. 103-105. Chlamisus texanus. 103, dorsal and lateral aspects; 104, presternum; 105,
antenna. Figs. 106-109. Chlamisus foveolatus. 106, dorsal and lateral aspects; 107, female
presternum; 108, male presternum; 109, antenna. Figs. 110-112. Chlamisus quadrilobatus.
110, dorsal and lateral aspects; 111, presternum; 112, antenna.

A Revision of the Subfamily Chlamisinae 917

margin slightly depressed and straight except for acute posterolateral angles,
sides slightly convergent anteriorly. Metascutellum slightly exposed but
undeveloped between anterior elytra. Elytra (Fig. 103) coarsely and deeply
punctate, most punctures recessed, producing partially reticulate surface;
large, dark spots not reticulate; tubercles lacking or present as small, faint
carinae; su-2 small and broad; three small carinae present extending from
ae to su-3, from anterior inside margin of umbone to su-2a and from umbone
along lateral margin to before apex; anterior elytral suture open, posterior
4/5 serrate. Pygidium evenly, coarsely and deeply punctate, punctures
recessed, producing reticulate, slightly convex surface, with small depres-
sions. Prosternum (Fig. 104) twice as long as wide, subtriangular, sides
slightly emarginate, center depressed. Metasternum foveolate. First ab-
dominal segment coarsely and deeply punctate, punctures recessed, pro-
ducing reticulate appearance; last abdominal segment foveolate; mid-ventral
fovea broad and deep, surrounded by few short hairs. Well developed
tibial spine on front and middle legs.

Aedeagus (Fig. 13) curved, slightly constricted before middle, tubular,
with fine apical hairs; apical orifice subterminal, median dorsal plate less
than 1/2 width of aedeagus; ejaculatory guide (Fig. 42) tubular, with
ventral anterior enlargement; constricted at middle. Female unavailable
for dissection.

Types: Female holotype, Brownsville (Cameron Co.), Texas, Charles Schaeffer collection
(USNM).

Discussion: Four specimens were available for study. The one from
Mexico agrees with the type and the original description. The greenish
coloration of the type has faded and the specimen is covered with museum
dust, but the identifying characteristics can still be seen. No information is
available about food plants.

Specimens examined (3 males and 1 female): (Map 1)

UNITED STATES: TEXAS: Cameron Co., 15 May; Hidalgo Co., 7 Apr.

MEXICO: YUCATAN: Merida, 29-30 July.

Chlamisns joveolatus (Knoch)

Chlamys joirolata Knoch, 1801:130.
Chlamisns joveolata (Knoch); Fattig, 1948:8.

Diagnosis: This species is elongate, black with pale legs, and has a
deeply divided pronotal gibbosity, sharp medial carina on pygidium, and
a broad prosternal process in females. The aedeagus, ejaculatory guide and
spermatheca are also diagnostic.

Description: Male: 2.5-3.0 mm long; 1.4-1.8 mm wide. Female: 2.6-3,3
mm long; 1.5-1.9 mm wide. Color dark brown to black with yellow to
light brown appendages; clypeus, lower and middle frons similarly colored;
pronotum usually black and elytra dark brown.

918 The University Science Bulletin

Head coarsely punctate, middle frons moderately punctate; medial groove
from vertex to middle frons. Antennal segment 3 subcylindrical, 4 sub-
triangular, 5-10 rectangular, 11 evenly rounded (Fig. 109). Pronotum deeply
punctate to foveolate, with obscure foveae laterally; gibbosity with broad
punctate groove from anterior margin to posterior lobes; posterolateral
slopes nearly perpendicular, posterior lobes tumid and punctate. Meso-
scutellum smooth, posterior margin curved and slightly depressed, sides
convergent anteriorly; posterolateral angles acute. Elytra (Fig. 106) mod-
erately to coarsely punctate, with sharp tubercles and carinae; most tuber-
cles connected by irregular carinae, ae large and acute; punctures deep,
most posterior punctures depressed; straight anterior section of suture short
and open, posterior suture serrate. Pygidium moderately to coarsely punc-
tate, with sharp medial carina; apex strongly bent forward, large rounded
hump lateral to carina, shallow depressions groove-like from apex to base.
Presternum less than twice as long as wide; prosternal process narrow
in male (Fig. 10S), wide in female (Fig. 107). First abdominal segment
coarsely punctate to foveolate, last abdominal segment coarsely punctate;
mid-ventral area with long lateral hairs, flat in male, deeply and broadly
foveolate in female. Well developed tibial spine on front and middle legs
of male, no tibial spines on female.

Aedeagus (Fig. 14) tubular, strongly curved, without apical hairs; apex
with ventral longitudinal keel, slight constriction behind middle; apical
orifice round and subterminal, median dorsal plate small and narrow;
ejaculatory guide (Fig. 43) tubular, flattened dorsoventrally, with large
posterior groove, small dorsal subapical emarginations, apex curved ven-
trally. Spermatheca (Fig. 71) with long curved cornua, apical 1/2 J-shaped;
collum short, tubular, its length equal to basal width of cornua; duct long
and coiled; spermathecal gland short and sac-like, about 1/4 as long as
cornua.

Type; No type specimen was found in the Berlin Museum where
Knoch's types should be located, and none was found elsewhere. The
identity of the species is not in question so a neotype is not designated.

Discussion: C. joveolatits has seldom been confused with any other
species. The available specimens show a wide distribution in eastern
North America and a central Mexican locality. Very little variation is
found in specimens from widely separated localities. Several specimens are
dark brown rather than black. These are probably teneral or poorly pre-
served. There is less variation in the sculpturing in this species than in most
of the other members of the subfamily in North America.

The species has some resemblance to species of Exema, but the genitalia
are very different. Detailed examination of sculpturing also reveals the
superficiality of this resemblance. There are several South American species

A Revision of the Subfamily Chlamisinae 919

of Chlamisus with which joveolatus has a very close relationship, even
though this is not brought out by Lacordaire's treatment of the species.
There are no close relatives in North America.

Plant records: Qtterctis bicolor (series with larvae), Rhus toxicodendron.

Specimens examined (36 males and 74 females): (Map 2)

UNITED STATES: ALABAMA: Mobile Co., 17 June. ARKANSAS: Garland Co., 29
May. DISTRICT OF COLUMBIA: Washington, 4 July. FLORIDA: Bay Co., 20 May;
Liberty Co., 15 May. ILLINOIS: state record, no date. INDIANA: Clark Co., 13 June;
Marshall Co., 11 June. KANSAS: Riley Co., May; Shawnee Co., 12-25 June. KENTUCKY:
state record, no date. LOUISIANA: Natchitoches Par., 28 May. MARYLAND: Baltimore Co.,
24 June, 6 July; Montgomery Co., 23 June, 10 July, 24 July, 4 Aug.; Prince Georges Co., 5 Aug.
MISSISSIPPI: George Co., 7 May. MISSOURI: Ozark Co., 9 June; St. Louis Co., 24 Aug.-20
Sept. NEW JERSEY: Camden Co., 30 May; Cape May Co., 28 May, 11 June; Gloucester Co.,
15 June. NORTH CAROLINA: Harnett Co., no date; Madison Co., no date; Moore Co.,
12-18 May, 10 June; Watauga Co., 7 Aug. OHIO: Greene Co., 19 Aug.; Hocking Co., 2 July,
20 July. PENNSYLVANIA: Allegheny Co., no date. SOUTH CAROLINA: state record, no
date. TENNESSEE: Montgomery Co., 17 July. TEXAS: Cherokee Co., 3 May; Dallas Co.,
no date; Gillespie Co., 7 May; Harrison Co., 22 May, 7 June; Leon Co., 18 May; Sabine Co.,
27 Mar. VIRGINIA: Nansemond Co., 11 June; Nelson Co., 28 June.

MEXICO: HIDALGO: Jacala, 20 Aug.; 10 mi. NE Jacala, 1-2 Aug.

Chlamisus quadrilobatus (Schaeffer)

Bolosc/iesis quadrilobata Schaeffer, 1926:186-187.

Chlamys quadrilobata (Schaeffer); Leng and Mutchler, 1933:44.

Chlamisus colima Monros, 1951:257-259 (new synonymy).

Diagnosis: This species is recognized by its small size, black color, highly
tuberculate dorsum and completely serrate elytral suture. The aedeagus,
ejaculatory guide and spermatheca are also diagnostic.

Description: Male: 2.6-3.3 mm long; 1.7-2.2 mm wide. Female: 2.7-3.5
mm long; 1.9-2.4 mm wide. Color shiny black to brownish bronze; antennae
yellowish brown, apical 3-4 segments dark brown; tarsi and inner surface
and apex of tibiae yellowish brown.

Head generally moderately punctate, coarsely and deeply punctate
above and between antennal sockets, vertex finely and faintly striate; deep
depression between eyes below vertex. Antennal segments 3-4 subcylindrical,
5 subtriangular (slightly wider in males than females), 6 subrectangular
(Fig. 112). Pronotum finely and deeply punctate, with moderate striae
radiating from punctures; gibbosity broadly divided; tubercles well devel-
oped, mp and la large and round, anp divided by oblique carina, sm
carinate; several other smaller tubercles present; posterolateral slopes deeply
punctate, posterior lobes tumid. Mesoscutellum finely striate, posterior
margin slightly depressed and curved anteriorly, posterolateral angles acute,
sides convergent anteriorly. Elytra (Fig. 110) moderately and deeply punc-
tate, tuberculate and carinate; entire suture serrate; ae large and tuberculate,
all other anterior tubercles sharply carinate; posterior tubercles rounded;
many posterior punctures with fine radiating striae. Pygidium coarsely

920 The University Science Bulletin

and deeply punctate, with well developed central carina and four central
rounded tubercles; basal depression deep. Prosternum (Fig. Ill) subtri-
angular, sides convergent, process less than 1/2 total length. First abdom-
inal segment foveolate with single, rounded tubercle at each side; last
segment coarsely punctate, mid-ventral area with few hairs, flat in male,
with shallow, broad fovea in female. Large tibial spine on front legs of
male only.

Aedeagus (Fig. 15) long and tubular, slightly curved, broadly con-
stricted near middle, apex evenly rounded in dorsal view, pointed in lateral
view; lightly sclerotized section over basal foramen less than 1/3 total
length of aedeagus; apical orifice subterminal, median dorsal plate narrow;
ejaculatory guide (Fig. 44) complex, ventral elongate process attached to
two semicircular plates loosely enclosing ejaculatory duct. Spermatheca
(Fig. 72) long and slim, cornua elongate, curved, J-shaped, apex slightly
narrowed; collum short, length less than width of cornua; spermathecal
gland short, sac-like.

Type: San Antonio (Bexar Co.), Texas, Hubbard and Schwarz, USNM
type 29466; probably female. The specimen is poorly mounted so that the
sex cannot easily be determined. Another specimen, probably a male, same
locality, 22 June 1895, H. Sotau, is labeled paratype.

Discussion: Two paratypes of Chlamisus colima Monros from Colima,
Mexico, in the United States National Museum, are conspecific with quad-
nlobatus, a species widely distributed in Mexico. On the basis of the para-
types colima is synonymized under quadrilobatus. Two species closely
related to quadrilobatus are found in southern Mexico and Central America;
further study may reveal others.

The species has no close relatives in North America, but it has a slight
resemblance to species of Exema. The combination of characters used in the
diagnosis and the key may show this species and its relatives are sufficiently
distinct to merit being placed in a genus separate from Chlamisus. Any
such treatment which would divide the North American species of Chlami-
sus would necessitate the naming of at least two or three more new genera.

Plant records: Waltheria americana (series), Hyptis pectinata, Salvia albida (series),
Mimosopsis aculeaticarpa, Melampodiam divaricatum and Sida glutinosa.
Specimens examined (39 males and 72 females): (Map 1)
UNITED STATES: KANSAS: Douglas Co., 21 Mar. LOUISIANA: Natchitoches Par.,

16 Aug. TEXAS: Bastrop Co., 7 Sept.; Bexar Co., Sept.; Kerr Co., 17-23 Apr.; Sabine Co.,
27 Mar.; San Patricio Co., 15 Apr.; Uvalde Co., 6 July.

MEXICO: CHIAPAS: 12 mi. N Ocozocoautla, 26 July; 19 mi. W Ocozocoautla, 25 June.
COLIMA: Colima, 9 Feb.; 31 mi. E Colima, 19 July. JALISCO: 10 mi. SW Autlan, 22 July;
40 mi. E Guadalajara, 13 June; 11 mi. N Pihuamo, 19 July; Volcan de Colima, no date.
MICHOACAN: Morelia, 22 June; 20 mi. N Nueva Italia, 21 June; 7 mi. W Zamora, 15 June.
MORELOS: 3 mi. N Alpuyeca, 30 Mar.; Canon de Lobos, Yautepec, 18 Mar.; Cuernavaca,

17 Mar., May, 4 June, July; 3 mi. N Cuernavaca, 14-21 Mar., 10 Apr.; 5 mi. E Cuernavaca,
29 Mar.; Hueyapan, 8 July; 7 mi. SSW Yautepec, 14 June; "Las Estacas," 6 Apr. NAYARIT:

A Revision of the Subfamily Chlamisinae 921

2 mi. S Acaponeta, 21 Mar.; Jesus Maria, 7 July; San Bias, 20 June; Tepic, 10 June, 13 Oct.
SAN LUIS POTOSI: 20 mi. NE Ciudad del Maiz, 19 June; "6 mi. S Matalpa," 7 July.
SINALOA: 32 mi. S Mazatlan, 22 June; 20 mi. N Villa Union, 19 Aug. SONORA: Alamos,
1 Aug.; "Guirocoba," 6 July. TABASCO: 4 mi. N Teapa, 14 June. TAMAULIPAS:
Matamoros, no date. VERACRUZ: Atoyac, 16-17 June; Cordoba, 15 May; Cotaxtla Exp. Sta.,
Cotaxtla, 23 June; 70 km. W Veracruz, 1 Aug.

GUATEMALA: ALTA VERAPAZ: Trece Aguas, 27 Apr. ESQUINTLA: Finca El Zapote,
7 July; Finca San Victor, 10 Sept.

EL SALVADOR: CUSCATLAN: El Rosario, 24 Mar. SAN SALVADOR: 2.5 mi. W
Quezaltepeque, 27 June.

HONDURAS: ATLANTIDA: La Ceiba, 23 Sept.

COSTA RICA: ALAJUELA: Ciruelas, 28 Apr. CARTAGO: Turrialba, no date; 4 mi. E
Jicotea, 2 July. GUANACASTE: 3.5 mi. NW Arenal, 28 July; 7 mi. W Arenal, 29 July; 4 mi.
NW Canas, 7-9 July; 7 mi. SE Irma, 31 July; 1-6 mi. SW Los Angeles, 31 July. HEREDIA:
Barba, 23 May. PUNTARENAS: Baranca near Puntarenas, 25 Aug.

Chlamisns huachiicae (SchaefTer)
Chlamys huachiicae SchaefTer, 1906:228-229.

Diagnosis: Specimens of huachiicae are recognized by the unevenly
punctate pygidium, dark brown apical 1-2 segments of the antennae, and
brown to black color pattern of the elytra and pronotum. The aedeagus,
ejaculatory guide and spermatheca are also diagnostic.

Description: Male: 3.1-3.5 mm long; 1.8-2.0 mm wide. Female: 3.5-4.1
mm long; 2.1-2.5 mm wide. Color yellow to reddish brown with irregular
markings of light brown to black; antennae yellow with apical 1-2 seg-
ments dark brown, posterior margin of pronotum and anterior margins of
elytra black; elytral depression and basal 1/3 of pygidium brown; underside
except legs generally brown, sides yellow to light brown; margins of pro-
sternum brown.

Head coarsely to finely punctate, vertex coarsely punctate; antennal
segment 3 subtriangular, 4-10 rectangular, 11 subconical (Fig. 115). Pro-
notum coarsely and deeply punctate, punctures depressed causing a reticu-
late appearance; tubercles absent; gibbosity rounded, faintly divided by
broad longitudinal groove, posterior lobes faintly punctate. Mesoscutellum
flat; posterior margin emarginate and depressed, sides convergent anteriorly,
posterolateral angles acute. Elytra (Fig. 113) coarsely and deeply punctate;
tubercles generally carinate, carinae broken and irregular; groups of large,
deep punctures in depressions, anterior, posterior and lateral to broad su-2
tubercle; su-3 broad and elongate; complete marginal carina extending from
umbone nearly to su-3a. Pygidium coarsely punctate at base to finely punc-
tate at apex, punctures deep; medial carina and depressions faint. Pro-
sternum (Fig. 114) twice as long as wide, with elongate process 1/2 or
more of total length, sides gradually convergent posteriorly, more abruptly
convergent in females. First abdominal segment coarsely punctate, punc-
tures depressed, producing reticulate appearance; last abdominal segment
moderately punctate, mid-ventral area flat with long lateral and anterior

922

The University Science Bulletin

ns

Figs. 113-115. Chlamisus huachucae. 113, dorsal and lateral aspects; 114, presternum; 115,
antenna. Figs. 116-118. Chlamisus nigromaculatus. 116, dorsal and lateral aspects; 117,
presternum; 118, antenna.

hairs in male, deeply foveolate with long lateral hairs in female. Large
tibial spine on front and middle legs of male, no tibial spines on female.
Aedeagus (Fig. 16) curved, tubular, strongly bent at middle and gradu-
ally flattened toward apex; apex bluntly pointed in dorsal view; ejaculatory
guide (Fig. 45) tubular, slightly narrowed behind, with anterior medial
process. Spermatheca (Fig. 73) simple, J-shaped; collum thin, tubular,
longer than width of cornua.

A Revision of the Subfamily Chlamisinae 923

Types: A male syntype (USNM no. 42-603) is here designated the lecto-
type. The specimen is labeled as follows: Palmerlee, Cochise Co., Arizona,
9 July, Brooklyn Museum Coll. 1929. There are also three female syntypes
(USNM no. 42305) with the same locality data, collected on 14 June and
9 July.

Discussion: C. huachucae is one of the pale southwestern species. It
can easily be separated from the other species, especially arizonensis with
which it is most easily confused, by the diagnostic characters.

This species is similar to Exema and Neochlamisus in the structure of
the aedeagus and spermatheca. It has coloring and sculpturing similar to
those of the pale species of Chlamisus and Pseudochlamys. The shape of
the aedeagus and spermatheca and similar antennal and elytral punctations
indicate a close relationship of huachucae to nigromaculatus n. sp.

The small number of specimens studied show little variation except in
the degree of darkness of the color pattern and the usual variation in punc-
tures and tubercles. No information on host plants is available.

Specimens examined (6 males and 21 females): (Map 2)

UNITED STATES: ARIZONA: Cochise Co., 14 June, 2-29 July; Pima Co., 10-20 June,
11 Aug.; Santa Cruz Co., 9 June, 20 June, 4-20 July, 3-4 Aug., 23 Aug.

MEXICO: COAHUILA: nr. Jame, 33 mi. SE Saltillo, 18 July. SONORA: Yecora, 20-22
May.

Chlamisus nigromaculatus new species

Diagnosis: The extensive dark markings on the pronotum and elytra,
unevenly rounded pronotal gibbosity with striopunctate posterolateral slopes,
and wide black markings on the upper frons are distinctive. The aedeagus,
ejaculatory guide and spermatheca are also diagnostic.

Description: Male holotype: 3.5 mm long; 2.1 mm wide. Color yellow
to light brown with extensive black markings covering most nontubercu-
late areas; ventral surface black with lateral yellow spots; legs light brown
with dark central areas on tibiae and femora; head with wide black inverted
V extending from antennal sockets to vertex, narrow transverse black band
across vertex and along dorsal margin of eye to inverted V; pronotum
irregularly marked with large black spots, posterolateral slopes black;
posterior margin of pronotum and anterior margins of elytra black. Meso-
scutellum yellow with continuous black margins. Elytral tubercles, lateral
and sutural margins and apex yellow. Pygidium with several irregular
basal black spots, many black central and lateral punctures.

Head coarsely punctate, clypeus and lower frons moderately punctate;
antennal segment 3 subtriangular, 4 subrectangular, 5-10 rectangular, 11
subconical (Fig. 118). Pronotum coarsely and deeply punctate, tubercles
absent; gibbosity faintly divided posteriorly, its summit rough with irregu-
lar depressed punctures; posterolateral slopes striopunctate, posterior lobes
poorly developed and punctate. Mesoscutellum flat, subquadrate, with

924 The University Science Bulletin

projecting posterolateral angles. Elytra (Fig. 116) coarsely punctate, with
well developed tubercles and carinae; su-2 broad, su-3 broad and elongate,
anterior tubercles rounded or acute; anterior elytral suture open, but meta-
scutellum not visible, posterior 3/4 of suture serrate. Pygidium coarsely
punctate basally to finely punctate at apex; depressions and medial carina
faintly developed or absent. Prosternum (Fig. 117) subtriangular, slightly
longer than wide. First abdominal segment coarsely punctate (large trans-
verse groove on left side only in this specimen); last abdominal segment
coarsely punctate; mid-ventral fovea shallow on posterior half of segment,
with long anterior and lateral hairs; numerous long hairs on mid-ventral
posterior margin of first four abdominal segments.

Aedeagus (Fig. 17) curved, tubular, flattened dorsoventrally toward
apex, with subapical hairs on ventral surface; apical orifice subcentral,
median dorsal plate wide; apex evenly rounded in dorsal view; ejaculatory
guide (Fig. 46) long and complex, with anterior ventral keel, posterior
dorsal margins broadly separated.

Female allotype: 3.7 mm long; 2.2 mm wide. Color and sculpturing as
in male; prosternal process narrower than in male; numerous short hairs
on mid-ventral posterior margin of first four abdominal segments: mid-
ventral fovea of last abdominal segment large and deep; no tibial spines.
Spermatheca (Fig. 74) with cornua nearly U-shaped; collum short and sub-
triangular, length equal to width of cornua; spermathecal duct long and
convoluted.

Types: Male holotype: Chiricahua Mts. (Cochise Co), Arizona, 22 July 1953, D. }. and
J. N. Knull (OSU). Same data on 1? paratype (JAW). Female allotype with same data
except collected on 20 July 1953 (OSU). (Map I.)

Discussion: Only the short type series was available for study. The
name is chosen to describe the numerous black markings on the dorsum of
the beetle. The species is more closely related to huachucae than to other
members of the genus but differs in the characters mentioned in the diag-
nosis. It likewise has some similarities to species of Exema and Neochlami-
sus in the form of the aedeagus and spermatheca. No host plants are known.

Chlamisus maculipes (Chevrolat)

Chlamys maculipes Chevrolat, 1835:120.

Chlamisus maculipes (Chevrolat); Blackwelder, 1946:649.

Diagnosis: This is the only highly pubescent North American Chlami-
sus. In addition it is characterized by its large size and yellow ventral and
black dorsal surfaces. The males have elongate antennae and long de-
cumbent hairs on the prosternum and metasternum. The aedeagus, ejacu-
latory guide and spermatheca are also diagnostic.

Description: Male: 3.6-6.0 mm long; 2.5-3.9 mm wide. Female: 4.4-6.0

A Revision of the Subfamily Chlamisinae

925

Figs. 119-123. Chlamisus mactdipes. 119, dorsal and lateral aspects; 120, male antenna;
121, female antenna; 122, male prosternum; 123, female prosternum.

mm long; 3.0-4.1 mm wide. Color of dorsum black, venter yellow with
black markings, lateral and anterior margins of pronotum yellow; head
yellow, with light brown to black spot on vertex; legs yellow with light
brown to black central spot, tarsi yellow, claws black; prosternum black,
coxal articulation black, pygidium black with lateral and apical margins
yellow; abdomen black with yellow border.

926 The University Science Bulletin

Head finely and faintly punctate, curved hairs on upper frons near eyes;
antennae in male (Fig. 120) extend beyond posterior margin of prosternal
process, segment 3 subconical, 4 subtriangular, 5-10 rectangular, but length
greater than width; antennae of female (Fig. 121) short, not reaching
posterior margin of prosternal process, segment 3 subcylindrical, 4 sub-
triangular, 5 subrectangular, but longer than wide, 6-10 subquadrate; seg-
ment 11 in both sexes elongate, bluntly pointed. Pronotum nontuberculate,
coarsely punctate, punctures obscured by 3-5 curved hairs from each punc-
ture, two nonpubescent areas on upper anterior slope of gibbosity, gibbosity
feebly divided, posterior lobe punctate. Mesoscutellum carinate between
posterior lobes of pronotum, lateral margins finely striate, carinate and
slightly convergent, posterior margin curved and depressed, posterolateral
angles acute. Punctation and pubescence of elytra (Fig. 119) similar to
pronotum except scattered tubercles smooth; ae, su-la, su-2, su-2a, ma-b
and su-3a usually present and well developed, ma faintly carinate; elytral
suture serrate except for short distance behind mesoscutellum. Pygidium
punctate and pubescent, resembling elytra, depressions obscure or lacking;
medial carina faint, smooth. Prosternum tumid, with numerous long hairs
in male (Fig. 122), short curved hairs in female (Fig. 123). Punctures
and pubescence of venter similar to those of dorsum, lateral punctures of
thorax foveolate and widely separated; large central area of metasternum
of male with numerous long straight hairs; central punctures of last ab-
dominal segment, each with a single hair. Mid-ventral area of last abdom-
inal segment in male flat with short lateral hairs, mid-ventral fovea of
female large and shallow with long lateral hairs. Well developed tibial
spine on front and middle legs of male; no tibial spines on female.

Aedeagus (Fig. 18) long, curved, tubular, flattened dorsoventrally toward
apex; apex explanate, with posterolateral triangular process; apical orifice
subcentral, median dorsal plate narrow and forked; large lateral and
small ventral subapical hairs; ejaculatory guide (Fig. 47) long, tubular,
with dorsal and ventral anterior processes; posterior, long, curved, ex-
planate plates open. Spermatheca (Fig. 75) widely J-shaped; collum
shorter than width of cornua; duct long and convoluted.

Type: The type cannot be found or recognized in material from the
British Museum where it should be located. The original description is
brief, but since the species is easily characterized there is no doubt about
its identity.

Discussion: The many unique characters found in this species suggest
that it should have generic status, together with several related Mexican
species. The present name of the species should remain unchanged until
there is a better understanding of its relationships with other species and
species groups.

A Revision of the Subfamily Chlamisinae 927

The species is placed in Chlamisus but nearer Diplacaspis and Neo-
chlamisus because of the similarities in the aedeagus, ejaculatory guide and
spermatheca. Although in some ways specialized, these structures in macu-
hpes could represent intermediate steps in the evolution of their more
highly specialized counterparts in the latter genera.

There is considerable variation in the ventral color pattern of macirfipes.
For example, the last abdominal segment may range from completely
yellow to completely black. The central black spot on the pygidium varies
from small to large, leaving only a narrow yellow margin. Females are
generally darker than males because of larger black markings. I cannot
correlate any color variation with geographic distribution.

Size variation is probably greater in this species than in any other in
North America. The specimens from Texas are among the smallest
members of the species.

Finally, the variation in the size of elytral tubercles is especially notice-
able. It is not any greater than that found in other North American species,
but it is more obvious because of the large size of most specimens and the
small number of tubercles on the elytra.

Specimens have been collected over a wide range in Mexico and Central
America, as well as in southernmost Texas. Individual beetles have been
collected on a number of unidentified plants. One, together with larvae,
was collected on Ipomaea pedicellaris, and one female was observed ovi-
positing on Verbesina fraseri. The species probably feeds on a variety of
plants.

Specimens examined (52 males and 76 females): (Map 2)

UNITED STATES: TEXAS: Cameron Co., 25 May-9 June, 12 July.

MEXICO: CAMPECHE: 10 mi. N Hopelchen, 17 Apr. CHIAPAS: 4 mi. NE
Pichucalco, 14 June; 1 mi. SW Simojovel, 19 Mar. COLIMA: Colima, no date.
GUERRERO: Acapulco, 10 Aug.; Iguala, 2 June; 8 mi. N Iguala, 23 Aug.; Taxco,
2 Aug. JALISCO: Volcan de Colima, 6 June. MORELOS: Cuernavaca, July; 7 mi.
SSW Yautepec, 14 June, 2 July, 28 July. NAYARIT: Jesus Maria, 26 June; Tepic, 10
June, 20-24 July; 10 mi. NW Tepic, 1 Sept. OAXACA: Almoloya, no date; 44 mi. N
Matias Romero, 19 June; Rincon Antonio, no date; Tehuantepec, 29 June. PUEBLA:
Acatlan, 17 June; Huauchinango, no date. SAN LUIS POTOSI: Ciudad de Valles, July, 29
Aug.; Salto del Agua, 28-30 June; Tamazunchale, 19 June. TABASCO: Teapa, no date.
VERACRUZ: Atoyac, 16-17 June; Cordoba, 8 May, 10-14 June; Cotaxtla Exp. Sta., Cotaxda,
27 June, 12 July; 2 mi. N El Cedro, 7 June; Fortfn, 8 July; Jicaltepec, 14-20 Mar.;10 mi. NE
Minatitlan, 4 July; Orizaba, June; Playa Vicente, no date; Puente National, 21 June, 3 July;
Rinconada, 29 June; San Andres Tuxtla, no date; Veracruz, no date; 20 mi. S Veracruz, 16
June. "Yucatan," no date.

GUATEMALA: ALTA VERAPAZ: Panzos, no date. RETALHULEU: Champerico, Aug.

BRITISH HONDURAS: "M-Tee Dist.," 12 June.

HONDURAS: MORAZAN: Zamorano, July.

NICARAGUA: no locality, no date.

COSTA RICA: GUANACASTE: 4 mi. NW Tilaran, 30 July. SAN JOSE: Santa Ana
(?), 12 Apr.; San Jose, no date. "Borroeta," 4-8 Apr.

928 The University Science Bulletin

Diplacaspis Jacobson, 1924

Diaspis Lacordaire, 1848:645-646.

Diaspsis Lacordaire; Jacoby, 1889:155 (misspelling).

Diplacaspis Jacobson, 1924:239 (new name for Diaspis Lac, not Costa, 1835, Coccidae).

S\warraia van Emden, 1932:9 (new name for Diaspis Lac).

Type species: Diaspis paradoxa Lacordaire (monotypic).

History: The genus Diaspis was established by Lacordaire for a unique
Mexican species having a visible second scutellum and bifid tarsal claws.
A second species was described by Baly (1878) from the Amazon region
of Brazil. Jacoby (1881) added two more Mexican species to the genus
by transferring moestificus and memnonius from Chlamys since they have
an exposed second scutellum. He also expressed some doubt about the
validity of this structure as a generic character. The name Diaspis proved
to be a junior homonym of Diaspis Costa (1835), an important genus of
Homoptera. Jacobson (1924), proposed the replacement name Diplacaspis,
now the valid name. Unaware of Jacobson's publication, van Emden (1932)
proposed the new name Skjvarraia for the genus. Navajas (1944b, 1946)
pointed out this synonymy in his review of the nomenclature and type
species of the Chlamisinae.

Monros (1954), after studying the type material in the United States
National Museum, transferred the North American species, prosternalis,
scabripennis and subelata from Chlamys and Boloschesis to Diplacaspis.
The Mexican species pici was described by Monros (1959) and as a result
of the synonymizing of pici with prosternalis and removal of moestifica,
memnonia, scabripennis and subelata from Diplacaspis, the genus now con-
tains nine species.

Diagnosis: The genus is characterized by having bifid claws, and ex-
panded prosternal process between the mesocoxae, a flattened, curved
aedeagus, and an elongate, pointed, symmetrical ejaculatory guide with
a small or partial sheath.

Description: Male: 2.7-4.5 mm long; 2.2-3.2 mm wide. Female 2.7-3.5
mm long; 2.0-2.7 mm wide. Color shiny black to reddish or greenish
bronze; antennae yellowish brown.

Head finely to coarsely punctate, vertex sometimes striopunctate, upper
frons with slight central depression. Antennal segment 1 shorter than
normal for the subfamily, 2 subcylindrical and shorter than 3, 3-4 sub-
cylindrical and subequal, 5 subtriangular, 6-10 rectangular, 11 subcorneal.
Pronotum finely and deeply to moderately and faintly punctate, sometimes
moderately and irregularly striate; gibbosity evenly rounded to highly tu-
berculate, faintly to deeply divided; posterior lobes smooth to faintly punc-
tate; tubercles well developed or absent. Mesoscutellum subquadrate, sides
faintly to strongly convergent anteriorly, posterior margin straight and
slightly depressed, posterolateral angles slightly reflexed. Metascutellum

A Revision of the Subfamily Chlamisinae 929

projecting between anterior 1/3 of elytra, rounded or sharply carinate,
broadly exposed. Elytra finely to moderately punctate, punctures deep or
depressed; tubercles faint to large and sharply carinate, granulate or finely
striate; anterior 1/3 of elytral suture broadly opened, exposing metascutel-
lum, posterior 2/3 serrate. Pygidium slightly convex and tuberculate, with
central carina or faint longitudinal groove, moderately to coarsely punc-
tate; deep punctures near apex and in lateral depressions. Presternum
coarsely punctate with lateral margins convergent anteriorly, process ex-
panded between mesocoxae, its apex bluntly pointed; central area slightly
depressed. First abdominal segment punctate to obscurely foveolate later-
ally, with two lateral depressions; last segment punctate or foveolate, mid-
ventral area in male flat or longitudinally grooved, with numerous re-
curved hairs, in female with large, deep fovea, lateral hairs short and scat-
tered. Single tibial spine on front and middle legs of both sexes, smaller
in female than in male. Tarsal claws bifid.

Aedeagus long and curved, gradually flattened dorsoventrally toward
apex; apex truncate to long-pointed, apical orifice subcentral; ejaculatory
guide symmetrical, with small to large sheath, anterior half with dorsal,
paired, long, pointed plates. Spermatheca J-shaped, apex of cornua gradu-
ally narrowed, duct long and convoluted.

Distribution: The species are most numerous in Brazil but one occurs
in the Southern United States, and several occur in southern Mexico and
Central America.

Systematic position: In sculpturing and form of the pronotum and
elytra, Diplacaspis paradoxa from Mexico resembles species of Melhto-
chlamys. Other characters, however, indicate no close relationship between
the genera. The shape of the presternum could be a modification of that
found in any of the related genera.

Most of the species of Diplacaspis resemble Neochlamisus more closely
than any other chlamisine genus. Each genus has several specialized char-
acters shared with some or all of the species of the other. A combination
of characters is necessary to distinguish all of the species.

The faintly striate pronotum and elytra resemble those commonly
found in Neochlamisus and a few species of Chlamisus. The form of the
spermatheca is similar in all three genera. The form of the ejaculatory
guide suggests a close relationship to Neochlamisus. The flattened aedeagus
is similar to those of the above genera, but several others have the same
form for this structure. The highly tuberculate sculpturing, large, deeply
divided pronotal gibbosity and foveolate abdomen are characters found in
all three genera. In spite of the close relationships mentioned above, there
is sufficient discontinuity for generic distinction in Diplacaspis. It has an

930 The University Science Bulletin

intermediate position between Neochlamisits and the more specialized
species of Chlamisus.

Discussion: The diagnosis and description above are compiled from
information about the North American and Mexican species of Diplacaspis
that I have studied. Published descriptions of several South American
species indicate the tarsal claws are not always bifid and the prosternal
process is not always expanded between the coxae. Nevertheless, the use
of these characters aids in the identification of the North American and
Mexican species.

With the transfer of four species to the new genus Neochlamisits, the
remaining nine are a more closely related group. Although the exposed
second scutellum is present in all of them, the character is not diagnostic
because it is found in other genera. However, chlamisines possessing this
character and lacking deep, closely set punctures which give the appearance
of velvety spots are all in Diplacaspis.

Drawings of the male genitalia of the Mexican D. paradoxa (Figs. 20,
48) have been included for comparison with prosternalis and species of
Neochlamisits. Reliable plant records for two different species suggest
Diplacaspis has a preference for Acacia spp. Only one species, with a wide
distribution in Mexico, is found in the United States.

Diplacaspis prosternalis (SchaefTer)

Chlamys prosternalis Schaeffer, 1906:227.
Boloschesis prosternalis (Schaeffer); Schaeffer, 1926:186.
Chlamisus prosternalis (Schaeffer); Blackwelder, 1946:649.
Diaclaspis (sic) pici Monros, 1959:3-4 (new synonymy).

Diagnosis: The species is easily recognized by the bifid claws and ex-
panded prosternal process. The absence of velvety spots on the elytra when
the metascutellum is broadly exposed distinguishes prosternalis from the
veliitinus species group in Neochlamisits. The genitalia are also diag-
nostic for this species.

Description: Male: 2.7-3.4 mm long; 2.2-2.7 mm wide. Female: 2.7-3.5
mm long; 2.0-2.7 mm wide. Color shiny black to reddish bronze; antennae
light brown. Head coarsely punctate, vertex striopunctate, clypeus and
lower frons moderately punctate, upper frons with slight central depression.
Antennal segments 3-4 subcylindrical, 5 subtriangular, 6-10 rectangular (Fig.
127). Pronotum finely and deeply punctate, moderately, irregularly striate
except laterally; tubercles well developed, su, la and mp large and rounded;
gibbosity deeply divided; posterolateral slopes with 5-7 large deep punc-
tures; posterior lobes smooth, tumid. Mesoscutellum subquadrate, sides
slightly convergent anteriorly, posterior margin straight, slightly depressed,
posterolateral angles slightly reflexed. Metascutellum well developed,
sharply carinate, broadly exposed. Elytra (Fig. 124) moderately and deeply

A Revision of the Subfamily Chlamisinae

931

^ \y^rm&

Figs. 124-127. Diplacaspis prosternalis. 124, dorsal and lateral aspects, 125, prosternum;
126, tarsal claw; 127, antenna.

punctate, with well developed, broad tubercles usually not carinate; tu-
bercles finely striate, faintly punctate; anterior 1/3 of elytral suture broadly
open, exposing metascutellum, posterior 2/3 serrate. Pygidium convex,
coarsely punctate near apex and in four lateral and two deep basal de-
pressions; faint medial carina with deep lateral punctures. Prosternum
(Fig. 125) narrowed between procoxae then expanded between mesocoxae,
apex rounded, central area slightly depressed. First abdominal segment
punctate to obscurely foveolate laterally, with two lateral depressions; last
segment foveolate, mid-ventral area with numerous recurved hairs in male,
female with large deep fovea, lateral hairs short and scattered. Tibial spine
on front and middle legs of both sexes, spines smaller in female than in
male. Tarsal claws (Fig. 126) bifid.

Aedeagus (Fig. 19) curved and tubular, gradually flattened dorsoven-
trally toward apex; apex narrowly truncate; ejaculatory guide (Fig. 49) sym-
metrical, with two long, dorsal acute processes and a partial sheath. Sperma-
theca (Fig. 76) J-shaped, cornua gradually narrowed toward apex; collum
short and narrow, shorter than width of cornua; duct long and convo-
luted.

932 The University Science Bulletin

Types: A male in the U.S. National Museum, with the following label
data, is here designated the lectotype: Brownsville (Cameron Co.), Texas;
Brooklyn Museum Coll. 1929; cotype no. 42304 USNM. There are three
syntypes from the Brownsville area with the following data: Esperanza
Ranch, July; Yucca Ridges, 6 Aug. (USNM).

Discussion: The species was first described in the genus Chlamys.
Because of the broadly exposed second scutellum it was transferred to
Diplacaspis. A comparison of prostemalis with paradoxa shows a consider-
able difference in the sculpturing and shape. The tubercles, carinae and
the gibbosity are larger in prostemalis than in paradoxa. The shape (dorsal
outline) of prostemalis is subquadrate compared to the almost oval shape
of paradoxa. Intermediate forms are represented in the South American
species. The two species are considered congeneric on the basis of simi-
larities in claws, presternum, genitalia and metascutellum.

The description of pici was taken from a manuscript of Monros. There
was no mention in his publication of prostemalis, a species that must have
been unknown to him. On the basis of the original description of pici
and specimens from the type locality, I consider it conspecific with pro-
stemalis.

Plant records: Acacia farnesiana, A. greggii, Gaillardia sp. and Stiaeda? sp.
Specimens examined (41 males and 62 females): (Map 3)

UNITED STATES: ARIZONA: Pima Co., 4 Apr., 15 May-8 June, 23 June-30 July,
10-13 Aug. TEXAS: Cameron Co., 4-13 Apr., 25 May, 20 June, 8 Aug.; Kinney Co., 27 Aug.;
Live Oak Co., 25 Mar.: Starr Co., 28 Mar., 9 Apr.; Uvalde Co., 6 July.

MEXICO: BAJA CALIFORNIA SUR: 10 mi. SW San Jose del Cabo, 9 July; Triunfo,
7 July. CHIAPAS: 28 mi. W Cintalapa, 9 Apr. GUERRERO: Acapulco, Oct. NAYARIT:
70 mi. S Acaponeta, 22 Aug. NUEVO LEON: Monterrey, 1 Jan.; 5 mi. S Monterrey, 6 July,
25 Feb. OAXACA: Oaxaca, no date; 10 mi. NE Oaxaca, 20 June. PUEBLA: 30 mi. NW
Petlalcingo, 3 Apr. SINALOA: 30 mi. E Villa Union, 7 July. SONORA: Guaymas, 10 Apr.;
San Pedro Bay, 7 July. TAMAULIPAS: Tampico, 26 Mar. VERACRUZ: St. Lucrecia, no
date; Tuxpan, 18 Aug.; Veracruz, no date.

GUATEMALA: ALTA VERAPAZ: Lanquin, 5-8 June.

NICARAGUA: GRANADA: Granada, no date.

Neochlamisus new genus

Type species: Neochlamisus velutinus new species (present designation).

This genus includes 16 North American species formerly in Diplacaspis
and Chlamisus, several of which extend their distribution into Mexico, plus
several Central American species. A few South American species also appear
to have similar characteristics.

Most if not all of the references to North American species of Chlamisus
also refer to species of Neochlamisus and will not be repeated here. Specific
references will be found under the species headings.

Diagnosis: The ejaculatory guide of the aedeagus is asymmetrical and
partially enclosed in a sheath. The spermatheca is J-shaped or U-shaped

A Revision of the Subfamily Chlamisinae 933

and enlarged near the apex of the cornua. The spermathecal duct is long
and convoluted. In addition North American species can be identified by
the presence of either velvety spots on the elytra or striations on the postero-
lateral slopes of the pronotum.

Description: Male: 2.8-4.8 mm long; 2.1-3.7 mm wide. Female: 2.9-4.9
mm long; 2.2-3.8 mm wide. Color dark brown to bright bronze, generally
with yellow, blue, green, or red metallic reflections; frons in emargination
of eye with or without yellow spot; underside, legs and tarsi similar in
color to dorsum; antennae yellow to light brown; labrum light reddish
brown; mandible and palpi dark reddish brown.

Cuticle minutely granulate. Head with slightly to deeply depressed
medial area on front, finely to coarsely punctate, sometimes striate or
vermiculate. Antennal scape and pedicel normal for the subfamily, seg-
ment 3 longer than wide, subcylindrical, segments 4-10 compressed, bluntly
serrate, subtriangular to rectangular in outline, segments finely pubescent
and with 6-8 long hairs, segment 11 subcorneal, slightly flattened. Pronotum
faintly to moderately tuberculate, nonstriate to coarsely and regularly to
highly irregularly striate, finely to moderately punctate; posterolateral slopes
striopunctate or striate, irregular striae converging on area near middle of
lateral margin, anterior slope always more finely and irregularly striate;
gibbosity slightly to deeply divided, posterior lobes smooth to striate, some-
times with scattered punctations. Mesoscutellum subquadrate, impunctate,
granulate or finely to moderately striate. Metascutellum covered by elytra
or exposed, smooth, shiny, carinate, and highly developed between anterior
1/5 of elytra when exposed. Anterior margin of each elytron crenate from
scutellum to past middle of humeral umbone except short area antero-
lateral to tubercle ae; elytra granulate, sparsely to confluently punctate or
rugose, highly tuberculate and carinate; anterior tubercle acute to blunt,
posterior tubercles usually evenly rounded; suture serrate on posterior 4/5
only; some species with two velvety spots on disc lateral to and anterior
to tubercle su-2, usually separated by oblique carinae from tubercle su-2
to umbone; tubercles, carinae and deep puncture finely striate. Pygidium
granulate, medial carina usually complete; foveae shallow to deep, faintly
to deeply punctate, mostly confluent and highly granulate, lateral areas
usually more coarsely punctate; posterolateral, elongate depressions small
to moderate, apical area moderately punctate. Prosternum goblet-shaped,
coarsely punctate, center of anterior half flat to slightly impressed; posterior
process narrow or explanate, with lateral serrations. Anterior edge of meta-
sternum between mesocoxae with two blunt processes between which
prosternal process fits; ventral surface appearing reticulate, with punctures
in shallow to deep, close or sometimes confluent foveae. First abdominal
segment with lateral process partly covering apex of femur and base of

934 The University Science Bulletin

tibia, in repose; many circular confluent foveae producing large shallow
depressions dorsal and posterior to process; Y-shaped carina between coxae
accentuated by punctures; regular row of small foveae on posterior border
of each abdominal segment; last abdominal segment with large, shallow
to deep fovea on mid-ventral area in female, always less impressed in male,
surrounded in both by short posterior and anterior hairs, lateral hairs longer,
anterior hairs sometimes lacking in female. One or two tibial spines usually
well developed, generally curved inward, on front leg of male, present or
absent in female; middle legs with or without tibial spines, which may be
obscured by the apical spurs and setae.

Aedeagus explanate and truncate to round or pointed; apical orifice sub-
central, short longitudinal row of lateral hairs before apex; ejaculatory
guide asymmetrical, short to long; spermatheca J-shaped or U-shaped, apex
of cornua slightly enlarged; collum short, sometimes slightly enlarged,
shorter than width of cornua; duct long and highly convoluted.

Systematic position: Neochlamisus is placed at the end of this revision
because of its many specialized characters, including asymmetrical ejacula-
tory guide, ejaculatory sheath, variation in tibial spines, pronotal striation,
yellow spot on the frons in the emargination of the eye (except velutinus
and memnonins) and in some species the exposed metascutellum and
velvety spots on the elytra. A close relationship to Diplacaspis is suggested
by the occurrence of several of these characters in both genera, as discussed
under that genus. Many of these similarities involve only a few species of
each genus. The structural similarity in the aedeagus, pronotal striation
and the tubercles of the elytra suggests some relationship of Neochlamisus
to North American Exema, but the resemblance is not as strong as to
Diplacaspis. The character states of Neochlamisus are probably more recent
specializations than the corresponding ones found in the other two genera.

Discussion: Two distinct species groups can be recognized in the genus:
the species formerly included in Diplacaspis and those formerly in Chlami-
sus. I have named these only as species groups since it seems that subgeneric
status would contribute nothing to an understanding of them.

On the basis of the interspecific similarity in both male and female
genital structures, the species of this genus are more closely related than
are those within any of the other genera studied. The ejaculatory guides
of the North American species are all of the same general shape. This
structure in velutinus and memnonius has some similarity to those of the
gibbosus species group and suggests the relationship of the two groups.
Specific variation is more apparent in the sculpturing of various parts of
the body.

A Revision of the Subfamily Chlamisinae 935

Key to Species Groups of Neochlamisits

la. Two velvety spots on each elytron divided by oblique carina; pro-
notum faintly striate or striopunctate; metascutellum exposed
between elytra (except in moestificus); tip of aedeagus explanate
to truncate; spermatheca J-shaped with slightly enlarged apex;
anterior and middle tibiae with one apical spine in both sexes

velutinus

b. No velvety spots or only faint traces of them on each elytron; pro-
notum faintly to coarsely striate; metascutellum seldom and only
slightly exposed; tip of aedeagus truncate to acute; spermatheca
J-shaped with noticeable knob on apex; anterior tibiae of males
with two apical spines (except only one in chamaedaphnes and
jraganae), middle tibiae with one spine, females usually with one
spine (very small in some) on anterior tibiae gibbosus

velutinus Group

The following species, formerly included in the genus Diplacaspis, are
included in the velutinus species group: memnonius, moestificus, scabri-
pennis, velutinus n. sp., and subelatus. Examination of the male and female
genitalia indicates that Diplacaspis paradoxa is not congeneric with the
above species, which are closely related to the North American species of
the gibbosus group. The supposition that the velutinus and gibbosus
groups are congeneric is supported by similarity in sculpturing, punctation,
size, and coloring, as well as in both male and female genitalic structures.

Most authors have doubted the validity of the genus Diplacaspis since
all species with an exposed second scutellum were included without regard
to other characters in which they differed from the type species. Consider-
able inter- and intraspecific variation of this character reduces its value as
a generic indicator, but it does indicate a relationship between the two
species groups and a certain degree of affinity among species. The presence
of a second scutellum remains a generic character in Diplacaspis but differ-
entiation of the genus in North and Central America is also based upon
the shape of the presternum, bifid claws, shape of the aedeagus and the
development of the ejaculatory guide.

The relationship between the velutinus and gibbosus species groups is
also suggested by two other characters, velvety spots on the elytra and the
presence of apical spines on the front and middle tibiae. One species of the
gibbosus group has a small velvety spot in the discal area, some specimens
of two other species have faint indications of velvety spots, and two other
species have the single tibial spine characteristic of the velutinus species
group.

The species of this group are most easily identified by the presence of
the visible second scutellum and the velvety spots on the elytra. One
Mexican species, memnonius is included here because of its similarity to

936 The University Science Bulletin

velutinus n. sp. The two have been known under the same name for many
years.

Key to the velutinus Species Group

la. Emargination of eye without a yellow spot 2

b. Emargination of eye with a yellow spot 3

2a. Velvety spots large and confluent, each including more than 20 deep

punctures; oblique carinae obscured by large velvety spots memnonius

b. Velvety spots small, each including about 10 deep punctures,

separated by a high, sharp, oblique carina velutinus n. sp.

3a. Pronotal gibbosity shallowly divided; color coppery; elytral sculpture

coarsely punctate-rugose; areas between punctures carinate subelatus

b. Pronotal gibbosity moderately to deeply divided; color coppery to
metallic black; elytral sculpture coarsely to moderately punctate;

areas between punctures granulate 4

4a. Metascutellum broadly exposed and projecting between elytra, ex-
posed anterior margin lA as wide as mesoscutellum; posterior
lobes of pronotum distinctly striate; mesoscutellum faintly striate
to rugose; elytra shiny granulate, slightly rugose, punctures large
and confluent, color bright reddish bronze; scutellar tubercle
small scabripennis

b. Metascutellum narrowly to scarcely exposed; posterior lobes of prono-
tum faintly striate; mesoscutellum smooth granulate, often with a
carina or tubercle in posterior part of pronotal groove; elytra
sparsely to moderately punctate, punctures usually distinct and
separated; color bright coppery to dark bronze; scutellar tubercle
faint to carinate; pronotum striopunctate moestifieus

Neochlamisus velutinus new species

Diagnosis: The absence of a yellow spot on the frons in the emargina-
tion of each eye, the high and deeply divided pronotal gibbosity, large, high,
carinate elytral tubercles, subtrunctate aedeagus, and long, slender ejacula-
tory guide together serve to distinguish this species from all others in
the genus.

Description: Male holotype: 3.8 mm long; 2.9 mm wide. Color dark
brownish bronze with small areas of metallic blue, purple, and red reflec-
tions. Antennae yellowish brown; labrum light reddish brown; mandibles
and palpi dark reddish brown; frons in emargination of eye without yellow
spot.

Head very finely and faintly striate, with irregular and often confluent
punctures especially on clypeus and lower frons; vertex strongly striopunc-
tate, central area of upper frons below vertex depressed; antennal segments 3
and 4 subcylindrical, 5 triangular, 6-10 rectangular (Fig. 131). Pronotum
rough, finely and irregularly striate, tubercles indistinct except mp broad
and rounded; gibbosity deeply divided (Fig. 129), anterior slope and sm
tubercles deeply punctate. Mesoscutellum with central longitudinal carina

A Revision of the Subfamily Chlamisinae

937

Figs. 128-131. Neochlamisus velutintis. 128, dorsal and lateral aspects; 129, cephalic aspect;
130, prosternum; 131, antenna.

on anterior half sloping to depressed lateral margins, posterolateral margins
reflexed. Metascutellum broadly exposed, longitudinally carinate. Elytra
carinate, with many high tubercles (Fig. 128), coarsely and deeply punc-
tate, anterior margins faintly crenate from scutellum to middle of humeral
umbone except in front of ae tubercle; tubercles and carinae finely striate,
high ae tubercle sloping to anterior margin, su-la sharply carinate and
connected with um-b; numerous deep punctures lateral and anterior to
high, broad, and acute su-2, area between these punctures sharply carinate
or spiny, ma carinate from humeral umbone to ma-a, su-3 high and elon-
gated, ma-b area tumid; anterior 1/4 suture broadly open, exposing meta-
scutellum, posterior 3/4 serrate. Pygidium covered with large, irregular,
confluent or isolated shallow foveae with faint setigerous punctures; an

938 The University Science Bulletin

elongate depression near posterolateral border; central longitudinal carina
broken; apical area flat. Presternum (Fig. 130) slightly depressed in center,
process slightly grooved longitudinally, posterior edge sloping steeply.
Sternum covered with large, shallow foveae with faint setigerous punctures,
foveae closer together than their diameters, producing reticulate appearance.
First abdominal segment with process partly covering apex of femur and
base of tibia in repose and forming lower edge of large depression; last
segment with large, shallow, mid-ventral fovea; many short, lateral foveal
hairs.

Aedeagus (Fig. 23) subtruncate, with transparent apical edge, slightly
enlarged at middle; ejaculatory guide (Fig. 50) except for long sheath,
slightly asymmetrical; basal orifice about 1/4 as long as ejaculatory guide.

Female allotype: 4.1 mm long; 3.2 mm wide. Color, sculpturing, and
antennae similar to those of male. Mid-ventral fovea on last abdominal
segment larger than in male, with hairs more numerous anteriorly. Sperma-
theca (Fig. 77) similar in general shape to that of other species in this
genus.

Types: Male holotype, San Bernardino Ranch, Cochise Co., Arizona, 3750 ft., August, F. H.
Snow (UK). Female allotype with the same data. Additional paratypes, all from Arizona:
1 $ , Baboquivari Mts. (Pima Co.), 24 July 1941, Burt Hodgden; 1 $, Baboquivari Mts. (Pima
Co.), 18 July 1932, R. H. Beamer; 1$, Santa Rita Mts. (Pima Co.), 12 June 1933, R. H.
Beamer; \$, Huachuca Mts. (Cochise Co.), 11 June 1933, R. H. Beamer; 1$, Sierra Ancha
Mts., Gila Co., no date, Duncan (UK). 1 $ , 13 mi. S Ajo, Pima Co., 2 Apr. 1966, L. & C.
W. O'Brien (JBK); 1 $ , 8 mi. E San Vicente, 3000 ft., Pima Co., 8 Aug. 1954, F. G.
Werner, L. Monros collection, 1959 (USNM); 3$ and 5$, Santa Rita Range Reserve, Pima
Co., 23 May 1957, 24 Apr. 1958, 29 Apr. 1958, 27 May 1958, G. Buder and F. Werner, swept
mesquite (UAR); 1$ and 45, Canelo, Pyeatt's Ranch, (Santa Cruz Co.), 28 May 1953,
A. & H. Dietrich (CU); 1 5, Portal (Cochise Co.), 17 June 1956, H. & A. Howden (HFH);
2$ and 3$, Sabino Canyon, Tucson (Pima Co), 16 May 1953, A. & H. Dietrich (CU);
3$ and 3 9, Sabino Canyon (Pima Co.), 11 July 1949, 30 July 1950, 4 June 1951, 11 July
1952, D. J. & J. N. Knull (OSU); 2$ and 2 9 , Tucson (Pima Co.), 27 July 1936, 13 Aug.
1936, J. N. Knull (OSU, JAW); 1 $ and 3 9 , Tucson (Pima Co.), 18 May 1953, 20 May 1953,
A. & H. Dietrich (CU); 1 $ , 10 mi. N Tucson (Pima Co.), 25 Apr. 1957, G. D. Buder, swept
alfalfa (UAR); 1$, 15 mi. E Tucson, Pima Co., 6 June 1952, M. Cazier, W. Gertsch, and
R. Schammel (AMNH).

Discussion: Most references to memnonuis in North America and
Mexico refer to veluUnus. An examination of the female syntype of mem-
nonius in the British Museum reveals several differences between the two
species as pointed out in the above key. The determined series in the same
collection, specimens of which were referred to in Biologia Centrali-
American (Jacoby 1881, 1889) contain both species including specimens of
another species, moestificus. The name velutinus describes the appearance
of the spots on the elytra.

Males vary in length from 3.6-3.9 mm and in width from 2.7-3.0 mm;
length of females varies from 3.8-4.2 mm and the width from 2.8-3.2 mm.

A Revision of the Subfamily Chlamisinae 939

The color is dark brownish to dark reddish bronze, with small areas of
metallic blue, purple, or red reflections.

N. velutinus more closely resembles memnonius than it does the other
three species of the group. The close relationship of these two species is
further suggested by the similar ejaculatory guides, which differ in size
and shape from the others in the group. The much smaller velvety spots
on the elytra and the shape of the aedeagus serve to separate velutinus from
memnonius. In the development of the pronotal gibbosity and of the velvety
spots of the elytra, velutinus is intermediate between moestificus and mem-
nonius. The sculpturing of velutinus varies considerably but generally fol-
lows that found in the genus as a whole as to the number, position and
size of punctures and tubercles. The coloring is similar to that of moestifi-
cus.

The highly developed metascutellum is broadly exposed, similar to that
found in the other species of the group except moestificus in which it is
definitely smaller and almost to completely covered by the often closed
elytral suture.

The specimens which I have seen from Chihuahua, Nayarit, Sinaloa,
and Sonora, Mexico, are mostly females, with tibial spines very small or
lacking on the front and middle legs. Their elytra are less tuberculate
than those of most Arizona specimens but not unlike some found in samples
from Arizona and Baja California. There is one male from Guaymas,
Sonora, Mexico, associated with these females, and since it is similar to the
Arizona specimens, both it and the females have been included in velutinus.

Plant records: Several different series without larvae have been collected
on Prosopis glandulosa. Only single specimens have been found on Acacia
constricta, Neltuma glandulosa, Prosopis chilensis and alfalfa.

Specimens examined (89 males and 140 females): (Map 4)

UNITED STATES: ARIZONA: Cochise Co., 4-17 June, Aug., 14 Sept., 29 Sept.; Gila Co.,
20 June, 18 July, 7 Aug.; Graham Co., 27 Apr., May, June, 4 July, 10 Aug.; Maricopa Co.,
13 Sept.; Navajo Co., 21 June; Pima Co., 2 Apr., 24 Apr.-12 June, 3 July-13 Aug.; Pinal Co.,
26 July; Santa Cruz Co., 3 Apr., 28 May, 28 June, 14 July, 29 July. TEXAS: Bexar Co.,
6 Apr., 23 Aug.; Cameron Co., 25 Mar., 17 Apr., 6 May, 23-25 June, 22 July-8 Aug.; Comal
Co., no date; Duval Co., 18 Mar., 23 Apr.-4 May, 31 May; Hidalgo Co., 20-26 Mar.; Kinney
Co., 27 Aug.; La Salle Co., 29 June; Presidio Co., 20 June; San Patricio Co., 28 Mar.; Starr Co.,
28 Mar.-9 Apr., 2 June; Uvalde Co., 23 May, 14 June, 30 June, 4 Aug., 25 Aug.; Val Verde Co.,
25 Apr.-5 May; Webb Co., 27 June; Zapata Co., 21 May.

MEXICO: BAJA CALIFORNIA: San Fernando, 31 July; Bahi'a de Los Angeles, 15 May;
Mesquital, 28 July. BAJA CALIFORNIA SUR: San Ignacio, 26 June; 15 mi. N El Refugio,
4 July; Triunfo, 13 July; 15 mi. N San Ignacio, 26 June; 45 mi. N San Ignacio, 27 June;
Vanancio, 17 July; Santo Domingo, 19 July. CHIHUAHUA: Chihuahua, no date; 92 km.
N Chihuahua, 30 June. COAHUILA: Boqillas del Carmen, 23 May. GUERRERO: Rio
Balsas, no date. MORELOS: Puente de Ixtla, no date. NAYARIT: Jesus Maria, 26 June;
2 mi. S Acaponeta, 21 Mar. NUEVO LEON: Apodaca, 27 July. SAN LUIS POTOSI: 46 mi.
N San Luis Potosi, 1 Sept. SINALOA: 32 mi. S Mazatlan, 22 June; 20 mi. E Villa Union,
19 Aug. SONORA: Guaymas, 3 June; Magdalena, 23 July; 4.9 mi. N Magdalena, 25 Aug.;
Yecora, 20-22 May.

940 The University Science Bulletin

Neochlamisus mem nonius (Lacordaire)

Chlamys memnonia Lacordaire, 1848:708-709 (misspelled menmonia6).
Diaspis memnonia (Lacordaire); Baly, 1878:343.

Diagnosis: The large, distinct velvety spot on each elytron, high pronotal
gibbosity with deep medial groove, acute tubercles and carinae, broadly ex-
posed metascutellum, and the absence of a yellow spot on the frons in the
emargination of the eye serve to distinguish this species.

Description: Male: 3.0-3.9 mm long; 2.7-3.3 mm wide. Female: 4.0-4.5
mm long; 2.8-3.2 mm wide. Dark to reddish bronze in color. Antennae
yellowish brown, apical half of last segment dark; labrum with dark brown
apical border; frons without yellow spot in emargination of eye.

Head finely and faintly punctate, except vertex moderately punctate;
antennal segment 4 subtriangular, 5 subrectangular, 6-10 rectangular. Pro-
notum moderately striate, with large areas nonstriate and granulate, faintly
and finely punctate; mp tubercle small, nonstriate, la and su forming high
gibbosity, moderately striate, coarsely to moderately punctate; anp absent;
gibbosity deeply divided, posterolateral slopes irregularly striate and coarsely
punctate, posterior lobes nonstriate and impunctate. Mesoscutellum granu-
late, lateral margin with small dorsal ridge, posterolateral angles reflexed.
Metascutellum broadly exposed. Elytra (Fig. 132) coarsely to moderately
punctate, anterior tubercles acute and carinae sharp; large, confluent,
discal, velvety spots separated by sharp carina from broad su-2 to small
su-la tubercles; anterior 1/5 or more of elytral suture open, broadly exposing
metascutellum, posterior 4/5 serrate. Pygidium with smooth, medial carina;
foveae confluent except in center next to carina. Presternum (Fig. 133)
about 1 1/2 times as long as wide. Last abdominal segment with large,
granulate fovea; fovea deeper in female and with fewer scattered hairs
than in male. Well developed single spine on front and middle tibiae in
male; no tibial spine in female.

Aedeagus (Fig. 21) explanate with transparent apical edge; ejaculatory
guide (Fig. 51) more than six times as long as wide, sheath slender, basal
orifice 1/5 total length of ejaculatory guide. Spermatheca (Fig. 78) J-shaped,
apex of cornua slightly enlarged.

Types: Lacordaire apparently based his description on two females, one
from the Dupont collection and the other from Pilate, who collected it in
the environs of Campeche, Yucatan, Mexico. I have not been able to locate
the Dupont specimen; it is assumed to be lost. The Dupont syntype could
be in one of a number of collections, but probably ought to be either in the
Paris Museum or in the Bowditch collection at the Museum of Comparative
Zoology, Harvard. An examination of this latter collection failed to locate

5 In the "table alphabetique," page 879 of Lacordaire's monograph, this species is indexed
as memnonia. The spelling in the text is therefore an obvious misspelling.

A Revision of the Subfamily Chlamisinae

941

Figs. 132-133. Neochlamisus memnonius. 132, dorsal and lateral aspects; 133, prosternum.

the specimen. Material from the Paris Museum marked as part of the R.
Oberthur collection to which Dupont's specimens are said to have gone,
did not contain this type.

A female specimen in the British Museum (Natural History) is without
doubt the syntype from Pilate's collection and agrees very well with La-
cordaire's description, except that he makes no mention of the broadly
exposed second scutellum. It is as obvious as the second scutellum of
Diplacaspis paradoxa; therefore, it is remarkable that it was overlooked by
Lacordaire. This syntype is designated as the lectotype. It has the follow-
ing label data: Cotype; 5; Baly coll.; Chlamys memnonia Lac. type, type
ex coll. Pilate; Chlamys memnonia Lac, 5, C. Yucatan, N. S.; Diaspis
memnonia Lac, Yucatan, ex coll. Deyolle. The specimen is 4.0 mm long
and 2.8 mm wide.

Discussion: This species is included because of its close relationship to
velutinus, with which it has been confused by earlier authors. The form
commonly recognized as memnonius in the United States is velutinus. The

942 The University Science Bulletin

reference to Chlamys mimnonius (misspelling) by Fall (1927) probably
refers to velutinus. The specimens from Mexico referred to by Pallister
(1953) as Chlamisus memnonins are moestificus. There were four mis-
identified specimens of this genus in the British Museum labeled mem-
nonms. Two of them are velutinus and two are moestificus. All but one
were referred to as mem nonius by Jacoby (1889) in the Biologia Centrali-
Americana.

Jacoby (1881), after examining the Pilate type from Baly's collection,
felt that Lacordaire overlooked the second scutellum in this species. Jacoby
and Baly placed memnonius in the genus Diaspis, but I have already dis-
cussed in the generic treatment my reasons for not including this species
and several others in that genus.

The large velvety elytral spots so distinctive in this species probably
are a specialized condition, and the other species of this group represent
intermediates. The smooth punctate condition of the corresponding area
of the elytra which is retained in most of the gibbosus species group prob-
ably is a more primitive condition. Likewise the loss of the yellow spot on
the frons in the emargination of the eye represents a specialized condition,
as compared to the more primitive one of a yellow spot or an even earlier
condition of a completely yellow face.

Plant records: The close relationship of this species to velutinus is re-
flected in the host plants, both being found on legumes. At present the
major host of memnonius is not definitely known, but it is probably a
species of Prosopis. The one female examined is mounted with the larval
case from which it emerged. This case is attached to a small thorny twig
characteristic of some species of Prosopis. The case has no indication of
plant hairs being used in its construction (cf. larvae of moestificus). It does
show various dark and light annulations which indicate that several dif-
ferent materials were used in its construction.

Specimens examined (six males and three females) : (Map 4)

MEXICO: CAMPECHE: Environs of Campeche (lectotype). GUERRERO: Acapulco,
no date; Chilpancingo, no date. JALISCO: 17 mi. W Guadalajara, 5000 ft., 11 Aug., pupa
on legume; Plan de Barrancas, 24 Aug. OAXACA: Tehuantepec, 18 July; Rincon Antonio,
no date. PUEBLA: 10 mi. SE Iziicar de Matamoros, 19 June, on legume.

Neochlamisus moestificus (Lacordaire)

Chlamys moeslifica Lacordaire, 1848:706-708.

Diaspis moeslifica (Lacordaire); Jacoby, 1881:75.

Chlamys conjusa Schaeffer, 1919:325.

Boloschesis moestifica (Lacordaire); Schaeffer, 1926:185-186.

Chlamisus moestifica (Lacordaire); Blackwelder, 1946:649.

Diagnosis: This species has a very narrowly exposed metascutellum,
small velvety elytral spots, distinct elytral punctures, striopunctate and tu-

A Revision of the Subfamily Chlamisinae 943

berculate pronotum, smooth and granulate mesoscutellum, and large, often
sharply carinate elytral tubercles.

Description: Male: 3.2-3.9 mm long; 2.3-3.8 mm wide. Female: 3.3-4.5
mm long; 2.2-3.1 mm wide. Color dark brownish to bright reddish bronze.
Antennae yellowish brown at base, becoming darker brown at apex; frons
with small to large yellow spot in emargination of eye.

Head mostly finely and faintly punctate, more coarsely punctate between
antennal sockets and on vertex, vertex finely and faintly striopunctate.
Antennal segment 4 subtriangular, 5 subrectangular, 6-10 rectangular, seg-
ments 4 and 5 slightly smaller in female than in male. Pronotum rough,
finely and regularly striate; mp, la, and su tubercles coarsely punctate,
producing rough appearance, other areas faintly punctate, anp tubercle
faint; gibbosity moderately divided (Fig. 135), posterolateral slopes strio-
punctate; posterior lobes transversely, finely striate, with scattered punc-
tures. Mesoscutellum granulate, posterior margin flat to slightly depressed,
posterolateral corners reflexed slightly. Metascutellum barely visible be-
tween elytra. Elytra (Fig. 134) with large tubercles and carinae, moderately
punctate, anterior tubercles sharply carinate; each discal velvety spot dis-
tinct, consisting of 6-12 deep punctures separated by sharp ridges; su-2
slightly crescent-shaped; elytral suture narrowly exposing metascutellum.
Pygidium with smooth medial carina, foveae usually confluent. Presternum
(Fig. 136) nearly twice as long as wide. Last abdominal segment with
large, granulate, slightly depressed mid-ventral fovea; fovea more depressed,
and with hairs more scattered in female than in male. Well developed
spine on front and middle tibiae; tibial spines usually smaller and less
curved in female than in male.

Aedeagus (Fig. 22) subtruncate to explanate, with small transparent
apical edge; ejaculatory guide (Fig. 52) short and robust, length five times
width; basal orifice 1/5 as long as ejaculatory guide. Spermatheca (Fig. 79)
J-shaped, apex of cornua slightly enlarged.

Types: There is one male specimen in the British Museum with the
following label data: E. Coll. Laferte; D.; Cotype; C. moestifica Lac;
6756; Chlamys subaenea Reiche, Mexico. This specimen may or may not be
one of Lacordaire's syntypes. He had many examples from the collection
of Dupont from which he described two varieties of moestificiis. Lacor-
daire based several species in part on material from both the Laferte and
Reiche collections, but he did not mention these collectors in the descrip-
tion of moestificiis. No other types have been found in the following
museums where Lacordaire's or Dupont's specimens might be located:
Museum National d'Histoire Naturelle, Paris; Institut Royal des Sciences
Naturelles de Belgique, Bruxelles; Museum of Comparative Zoology, Har-
vard University; Museum Alexander Koenig, Bonn. The specimen in the

944

The University Science Bulletin

Figs. 134-136. Neochlamisus mocstifiats. 134, dorsal and lateral aspects; 135, cephalic
aspect; 136, presternum.

British Museum is accordingly accepted as a type until there is evidence
to the contrary. It is not being designated as the lectotype since there is
some question about its identity, and there is no question about the identity
of the species.

Discussion: This is the most variable species in the velutinus group. It
also seems to accept a variety of host plants. The color is generally dark
with some metallic reflection. All the examples I have from Utah, however,
are dark to bright reddish bronze. They also are larger and more carinate
than the majority of the specimens seen from other localities. Many speci-
mens from this Utah locality have a small, almost imperceptible yellow
spot on the frons in the emargination of the eye. They may be confused
with velutinus, but the latter species is much darker and more tuberculate.
The prosternal process and the ejaculatory guide are much shorter in rela-
tion to the width in moestificus than in velutinus.

A Revision of the Subfamily Chlamisinae 945

Occurrence of different populations on separate host plants suggests
that two or more species are involved. The following evidence for moestifi-
cus indicates one highly variable species, possibly with some populations in
an active state of speciation. Even though samples with fairly consistent
differences have been taken from two species of Eriogonum and one species
of Polygonum, the differences are not consistent enough to warrant sepa-
rate status at the species level. For example, the shape of the aedeagus is
fairly uniform in the Utah sample of moestificus, but among males from
other localities, the shape varies, including the two examples illustrated
and many intermediates. The Arizona samples from Eriogonum and
Polygonum show this range of aedeagal variation but are fairly uniform
in sculpturing. Other samples from Arizona show a range of variation in
both sculpturing and shape of the aedeagus; so it is impossible to distinguish
more than one species.

One of the most obvious morphological variations is the size of the
scutellar tubercle. In the Utah forms it is usually large and sharply car-
inate, with tubercle su-1 present. In several Arizona samples this tubercle
is almost absent, with no trace of a carina, su-1 is small, and the other
elytral tubercles are reduced in size. In the specimens with abbreviated
tubercles, the punctures are very distinct and sparse, and the apex of the
aedeagus is usually similar to that found in the Utah examples. Although
this type of sculpturing and punctation is fairly consistent in samples
from one or two localities, some highly tuberculate specimens also have
the distinct, sparse punctures.

The character of distinct punctures used by Schaeffer (1926) to dis-
tinguish moestificus from scabripennis is valid for the majority of the
specimens examined, but the larger and more exposed metascutellum found
in scabripennis is a more useful character in separating these two species.
In moestificus the exposure of the metascutellum is always slight and in
many cases there is almost no exposure. I have seen examples from several
different localities with punctation very similar to that of scabripennis, but
the specimens are darker and the metascutellum scarcely exposed. On the
basis of these characters, the distribution, and the host information, I con-
sider them variants of moestificus.

Additional and more accurate information on host plants and distribu-
tion may reveal one or more sibling species in what I choose to recognize
as moestificus. The widely distributed eastern samples are usually only
single specimens. Information from such samples is more tentative than
that from the Southwest where more extensive samples are available. It
appears that this species is undergoing the most active evolution in the
species group. The present study suggests that subelatus and scabripennis
are more recently separated from the moestificus line of evolution than are

946 The University Science Bulletin

memnonius and velutinus. They have reached a level of differentiation that
deserves species recognition on the bases of the morphological and ecological
characters mentioned in the respective species discussions.

Plant records: Eriogonttm racemosum (series with larvae), E. wrighti? (series with larvae),
E. microthecum , Polygonum sp. (series with larvae), Ceanothus jendleri (series with larvae),
and C. buxifolia.

Parasites: No parasites have been identified, but larvae of parasitic
Hymenoptera have been found in pupal cases of this species collected in
the field.

Immature stages: The following information is derived mostly from
material collected in Utah County, Utah, on Eriogonttm racemosum.
Comparisons were made with larvae and adults taken on Ceanothus
jendleri near Flagstaff, Arizona. Adults of N. moestijicus were collected
together with larvae and pupae near Provo, Utah, on the host plant. In
a few instances pupae and adults were picked from grass or other plants
in the vicinity; however, on the basis of observations of feeding and the
type of pubescent material used in the construction of the larval case, the
Eriogonttm is considered the only host in this instance. It is quite common
for adults, and even mature larvae ready to pupate, to wander to plants
other than the hosts. Such records cause considerable confusion when one
is trying to determine actual hosts.

The larvae and pupae, like the adults, are very difficult to find, but for
a different reason. The use of plant pubescence in the fecal case makes it
very similar in texture to the plant stem and difficult to see. The con-
struction of this case indicates that the larva incorporates several types of
material into it. This reflects feeding on the leaves, stems and even the
flowers at different times during larval development. Overall, the case is
black to dark brown, with some light brown annulations as well as sections
incorporating large amounts of pubescence. The original egg case is at-
tached to the apex and is of a different texture and color, as might be
expected.

The larvae are generally found near the base of the plant where most
of the leaves occur. The mature larvae pupate by attaching the fecal case
firmly to a major stem of the plant, usually in or near a furcation. The
case is situated so that the apex points down and away from the stem.
If the original egg covering from the female has not been broken from the
larval case, it is situated parallel to the stem and appears as an apical
appendage on the case. Inside this larval case are the discarded exuviae of
the earlier instars compressed into the apex. Upon emerging from the pupal
skin the adult remains inside the case until its cuticle has darkened and
hardened. With its mandibles it then cuts a circular incision near the
apex of the case and crawls out. I suppose that mating does not take place

A Revision of the Subfamily Chlamisinae 947

until the following spring since only one generation a year has been
observed, and adults collected in the spring readily mate in the laboratory
while those collected in August do not.

Ecological notes: The adults although brightly colored are usually
difficult to see because of their small size and their habits. The Eriogonum
plant itself, although 2-3 feet in height, is difficult to see from a distance
since the leaves are basal and the stems are very slim. The adults of the
beetles usually appear as dark objects along the silvery stems. They quickly
drop to the ground when disturbed. They can be collected easily by placing
a net beneath the plant, then gently touching the stem. On warm days
the adults have a tendency to fly rather than drop to the ground; then
sweeping is an effective means of collecting.

Specimens examined (159 males and 263 females): (Map 5)

UNITED STATES: ARIZONA: Cochise Co., 3 June, 14-20 June, 5 July-19 Aug., 7 Sept.,
19 Sept.; Coconino Co., 19-25 June, 13-14 July, 8-15 Aug.; Gila Co., May, 14-17 June,
12 July-Aug., 10 Sept.; Graham Co., 10 May, 19 July, 9 Aug.; Navajo Co., 19 June, 1 May,
6 July, 18-20 July; Pima Co., 12 June, 10-26 July, 1 Aug., 18 Aug., 1-3 Sept., 24 Sept.; Pinal
Co., 4 Sept.; Santa Cruz Co., 10 June, 6-7 Sept.; Yavapai Co., 10-30 June, 29 July-3 Aug.;
Yuma Co., 20 May. ARKANSAS: Benton Co., 11 Aug. COLORADO: Archuleta Co., 22-24
June; Larimer Co., 14 Aug. FLORIDA: Broward Co., Apr.; Duval Co., no date; Volusia Co.,
no date. GEORGIA: Haralson Co., 19 June. ILLINOIS: Union Co., 7-11 Aug. KANSAS:
Riley Co., June. KENTUCKY: state record, no date. MARYLAND: Montgomery Co.,
summer. NEW MEXICO: Catron Co., 23 June; Grant Co., 23 July. NEW YORK: Albany
Co., 4 Oct.; Westchester Co., 5 July. OREGON: Wheeler Co., 9 Sept. SOUTH CAROLINA:
Oconee Co., 25 May. TEXAS: state record, no date. UTAH: Utah Co., 30 July, 22-23 Aug.

MEXICO: CHIHUAHUA: Chihuahua, no date. DURANGO: Palos Colorados, 5 Aug.;
24-36 mi. W Durango, 4-6 May, 6-9 June, 2-7 July, 21 July; 28 mi. E El Salto, 22 July.
GUERRERO: Omilteme, July. OAXACA: Juquila, no date. PUEBLA: Environs of Puebla,
no date. SONORA: Yecora, 20-22 May.

Doubtful locality: 1 female in the Paris museum labeled "Brazil."

Neochlamisus scabripennis (Schaeffer)

Boloschesis scabripennis Schaeffer, 1926:184-185.

Chlamys scabripennis (Schaeffer); Leng and Mutchler, 1933:44.

Diagnosis: Characters used to identify this species are the bright reddish
bronze color, distinctly striate pronotum, moderately divided pronotal gib-
bosity, large, confluent elytral punctures, broadly exposed metascutellum
and coarsely striate, rugose mesoscutellum. Many of the elytral punctures,
tubercles and carinae are vertically striate. The area between punctures is
shiny granulate or striate.

Description: Male: 3.5-3.8 mm long; 2.5-2.8 mm wide. Female: 3.6-4.6
mm long; 2.7-3.3 mm wide. Antennae yellowish brown, becoming darker
brown at apex; mandibles reddish brown; frons with large yellow spot in
emargination of eye.

Head finely and faintly punctate except moderately punctate on vertex.
Antennal segment 4 subtriangular, 5 subrectangular, 6-10 rectangular, seg-
ments 4-6 slightly smaller in female than in male. Pronotum finely punc-

948

The University Science Bulleti

N

Figs. 137-139
prosternum
142, prosternum

Neoc/ilamisus scabripennis. 137, lateral aspect; 138, cephalic aspect; 139,
Figs. 140-142. Neochlamistis subelatus. 140, lateral aspect; 141, cephalic aspect;

tate, finely to moderately and regularly striate along contours; mp tubercle
small, other tubercles merged with large, rough, moderately punctate and
moderately divided gibbosity (Fig. 138) ; posterolateral slopes striopunctate,
striations converging at base; posterior lobes transversely, moderately striate
with few punctures. Mesoscutellum coarsely granulate, posterior margin
depressed, posterolateral corners reflexed and faintly striate. Metascutellum

A Revision of the Subfamily Chlamisinae 949

well developed, broadly exposed between elytra. Elytra (Fig. 137) with
large blunt tubercles and sharp carinae, coarsely punctate, granulate; tu-
bercles, carinae and punctures obscurely striate; discal velvety spots obscure,
consisting of 15-20 deep punctures separated by sharp ridges, su-2 broad
and acute, carinate with su-la and umbone, sc carinate to su-1. Pygidium
with smooth medial carina, foveae largely confluent and granulate. Pro-
sternum (Fig. 139) nearly twice as long as wide. First abdominal segment
with hairs posterior to large mid-ventral carina; last segment with large
smooth depressed mid-ventral fovea, punctation mostly confluent, foveal
hairs sparse on anterior and posterior border of sternum, lateral areas im-
punctate to finely punctate. Well developed spine on front and middle tibiae
of both sexes.

Aedeagus (Fig. 24) explanate, with transparent apical edge; ejaculatory
guide (Fig. 53) short and robust, length about three times width, sheath
broad. Spermatheca (Fig. 80) U-shaped, cornua slightly enlarged; collum
bulbular with blunt-pointed apex.

Type: Male holotype, Marathon (Brewster Co.), Texas, 7 June, 1908;
Mitchell and Cushman coll.; scabripennis type, USNM 29464. The type
is not brightly colored, and the dull metallic reflections are more purplish
or reddish than in most specimens.

Discussion: Although Schaeffer listed paratypes from New Mexico,
Arizona, and California, I have not seen any examples from further west
than extreme southeastern Arizona. I also have one male and one female
labeled "Lakewood, N.J.," a locality far removed from the normal distri-
bution of this species. I consider these specimens mislabeled.

This species is fairly uniform in size, color, and sculpturing, but as in
other species in this species group, there are some differences in the size
and shape of tubercles, placement of punctures, and brightness of the color.
It is similar to subelatus in color but very different in sculpturing, in which
it more closely resembles moestificus.

The shape of the aedeagus and ejaculatory guide is fairly consistent in
the specimens examined. The aedeagus might be considered intermediate
between the extremes illustrated for moestificus, so that on this basis alone
one could not distinguish any specimens.

Plant records: "Grease- wood" (series), (probably Sarcobatus vermictdatus; Larrea is
sometimes called greasevvood in Arizona); Larrea divaricata (series).

Specimens examined (38 males and 30 females): (Map 6)

UNITED STATES: ARIZONA: Cochise Co., 27 June, 9-25 July. NEW MEXICO: Dona
Ana Co., 22 June. TEXAS: Brewster Co., 7 June, 1-2 July; El Paso Co., 15 June, 8-9 July;
Presidio Co., 4-30 Apr., 6-8 Aug.; Terrell Co., 27 Apr.; Val Verde Co., 27 Aug.

MEXICO: CHIHUAHUA: 3 mi. W Jimenez, 26 July. SAN LUIS POTOSI: 46 mi. N
San Luis Potosi, 1 Sept. ZACATECAS: 10 mi. N Fresnillo, 10 May.

950 The University Science Bulletin

Neochlamisus subelatus (Schaeflfer)

Boloschesis subelata Schaeffer, 1926:184.

Chlamys subelata (Schaeffer); Leng and Mutchler, 1933:44.

Diagnosis: This is the least tuberculate species of the group. It appears
rugose because of the broad striae and punctures on the elytra and has a con-
sistently bright reddish bronze color. The metascutellum is always widely
exposed, and the pronotal gibbosity is slightly divided.

Description: Male: 3.5-3.8 mm long; 2.4-2.8 mm wide. Female: 3.6-4.3
mm long; 2.6-3.0 mm wide. Dark reddish to bright reddish bronze. An-
tennae light brown; frons with large yellowish brown spot in emargination
of eye.

Head moderately punctate on clypeus and between antennal sockets,
coarsely punctate and finely striate on vertex, rest finely punctate; de-
pressed medial area on upper frons usually large and deep. Antennal seg-
ment 4 subtriangular, 5 subquadrate, 6-10 rectangular, segments 4-5 smaller
in female than in male. Pronotum finely to moderately and regularly
striate along contours; mp tubercle large, evenly rounded, finely punctate;
la, anp and su tubercles small and merging with evenly rounded, slightly
divided, coarsely punctate gibbosity (Fig. 141); posterolateral slopes faintly
striopunctate; posterior lobes moderately striate, with scattered punctures.
Mesoscutellum coarsely granulate and moderately striate, posterior margin
gradually depressed, posterolateral corners rounded and slightly reflexed.
Metascutellum broadly exposed between elytra. Elytra (Fig. 140) with
small rounded tubercles and sharp carinae, coarsely punctate, area between
punctures usually shiny; faint discal velvety spots consisting of 6-12 deep
punctures separated by sharp ridges, su-2 crescent-shaped and carinate to
su-la and umbone, ae broadly rounded, sc carinate with su-1; 2-3 coarse,
short, longitudinal striae between su and mesoscutellum. Pygidium with
smooth medial carina; many large shallow foveae near carina. Length of
presternum (Fig. 142) less than twice width. First abdominal segment
faintly carinate on posterior half between coxae, with large setigerous punc-
tures; last segment with flat to slightly depressed mid-ventral fovea, foveal
hairs sparse. Mid-ventral foveal area large, more depressed in female than
in male, lateral hairs numerous. Well developed spine on front and middle
tibiae in both sexes.

Aedeagus (Fig. 25) subtruncate to explanate, with short transparent
apical edge; ejaculatory guide (Fig. 54) short and robust, length about three
times width. Spermatheca (Fig. 81) slimmer than that of other species in
the genus; apex of cornua slightly enlarged, collum bluntly pointed.

Types: SchaerTer based his description on specimens reared from larval
cases taken on Larrea tridentata at Tucson (Pima Co.), Arizona. One
female collected 23 February 1897 on Larrea tridentata is the holotype

A Revision of the Subfamily Chlamisinae 951

(USNM, cat. no. 29463). There are also two female paratypes from the
Hubbard and Schwarz collection labeled Jan. and Feb. 1897 (USNM).

Discussion: Although the body coloring and one of the host plants are
the same as for scabripennis, there is little chance of confusion of the two
because of the difference in sculpturing. The geographical distributions of
these species do not overlap, except in eastern Arizona. As is usually the
case in Neochlamisus, there is considerable variation in the shape of tu-
bercles and carinae, and especially in the placement of punctures. This
variation, however, does not cause any problems in identification of the
species.

As in scabripennis, the shape of the aedeagus and the ejaculatory guide
of subelatus is intermediate between the two extremes described for moestifi-
cus. This indicates the close relationship of the three species.

Plant records: Larrea tridentata glutinosa (series with larvae) and L. divaricata (series).
Specimens examined (71 males and 102 females): (Map 6)

UNITED STATES: ARIZONA: Coconino Co., Aug.; Gila Co., no date; Mohave Co.,
28 Apr., 16-20 July, 15 Sept.; Maricopa Co., 9-11 May, 16 June, 29 June-6 July, 31 Aug., 19
Sept.; Pima Co., 16 Feb., 9 Apr., 4 May, 8-16 June, 6 July-10 Aug., 24 Sept., 31 Dec; Pinal
Co., 21 Aug., 19 Sept.; Yavapai Co., 14 June, 25 July; Yuma Co., 14 Mar., 16 June.
CALIFORNIA: Imperial Co., 10 Jan.; Inyo Co., 7 June; Kern Co., 25 Apr., 1-6 June; Los
Angeles Co., 11 Aug.; Riverside Co., Apr., 21 May, 25 June, 14 Oct.; San Bernardino Co.,
25-30 Mar., 9-11 Apr., 26 Apr., 5-11 June, 14 Sept.; San Diego Co., 25 June, 6 July, 20 July,
7-12 Aug., 17 Sept. NEVADA: Clark Co., 9 Sept., 6 Nov.; Lincoln Co., 15 July, 24 Nov.;
Nye Co., 31 May, 10-16 Aug.

MEXICO: SONORA: Altar, 9 May; Guaymas, 3 June.

gibbosus Group

This species group includes gibbosus and eleven related species formerly
included in the genus Chlamisus. One additional species is described as new.

This group includes those species of Neochlamisus with an asymmetrical
ejaculatory guide, an incompletely serrate but closed elytral suture, and
without large velvety spots on the elytra (small velvety spots on bimacu-
latus) .

The species of the gibbosus group are found from Mexico to eastern
Canada on diverse and numerous host plants. The species are much less
restricted to a particular host than those of the velutinus group. The close
relationship of the two species groups of Neochlamisus is illustrated by the
form of the aedeagus and ejaculatory guide, the number of tibial spines
and faint indications of velvety elytral spots in several species of the gib-
bosus group.

Key to the gibbosus Species Group

la. Last abdominal segment and pygidium usually roughly sculptured

but not vermicuiate 3

b. Last abdominal segment and pygidium vermicuiate 2

952 The University Science Bulletin

2a. Single spine on front tibia of male jragariae

b. Two spines on front tibiae of male eubati

3a. Pronotal gibbosity deeply divided; prosternal process usually serrate

in male 4

b. Pronotal gibbosity feebly or moderately divided; prosternal process

not serrate in either sex 8

4a. Males with two unequal spines on front tibia; females without tibial
spines; aedeagus broadly truncate; posterolateral slopes of pronotum

coarsely and regularly striate gibbosus

b. Males with two subequal spines on front tibia; females with or with-
out tibial spines; aedeagus not broadly truncate; posterolateral

slopes of pronotum finely to coarsely and irregularly striate 5

5a. Small or faint velvety spots on elytra; females without tibial spines;

abdominal segments not foveolate bimaculatus n. sp.

b. No velvety spots on elytra; females with spine on front and middle

tibiae; abdominal segments foveolate 6

6a. Elytra with numerous, usually noncarinate tubercles insularis

b. Elytra with normal tubercles and carinae 7

7a. Color bright yellowish bronze; posterolateral slopes of pronotum
usually finely striate, striae usually convergent; pygidium

foveolate platani

b. Color dark brownish bronze; posterolateral slopes of pronotum

moderately and irregularly striate; pygidium punctate bebbianae

8a. Single spine on anterior and middle tibiae of males; similar but

smaller tibial spines in female chamaedaphnes

b. Two spines on anterior tibia, single similar spine on middle tibia

of males; females with or without tibial spines 9

9a. Prosternum subtriangular, sides evenly convergent anteriorly or

slightly emarginate 10

b. Prosternum goblet-shaped, sides strongly emarginate 1 1

10a. Anterior slope of pronotum obscurely striate; first abdominal seg-
ment impunctate, pygidium obscurely foveolate tuberculatus

b. Anterior slope of pronotum finely striate, first abdominal segment

impunctate, pygidium moderately to coarsely punctate crtbrtpennis

11a. Spines absent from anterior and middle tibiae of females comptoniae

b. Spines present on anterior and middle tibiae of females assimihs

Neochlamisits gibbosus (Fabricius)

Bruchus gibbosus Fabricius, 1776:212.

Clythra plicata Fabricius, 1798:111-112.

Chlamys plicata (Fabricius); Olivier, 1808:876.

Chlamys gibbosa (Fabricius); Gemminger and Harold, 1874:3303.

Boloschesis gibbosus (Fabricius); Schaeffer, 1926:187.

Arthrochlamys plicata (Fabricius); Brown, 1943:127-128.

Chlamisus gibbosa (Fabricius); Proctor, 1946:186.

Chlamisus plicata (Fabricius); Fattig, 1948:8.

Arthrochlamys gibbosa (Fabricius); Jaques, 1951:275.

Diagnosis: The large size, moderately and regularly striate posterolateral
slopes of the pronotum, large elytral tubercles, truncate aedeagus, strongly
serrate prosternal process in males, two unequal spines on the front tibiae

A Revision of the Subfamily Chlamisinae 953

of males and absence of spines from front and middle tibiae in females
characterize this species.

Description: Male: 4.5-4.8 mm long; 2.6-3.7 mm wide. Female: 4.2-4.9
mm long; 3.1-3.8 mm wide. Color dark brownish to bright reddish or
yellowish bronze. Antennae yellowish brown; frons with large yellow spot
occupying about 1/2 of emargination of eye.

Head finely and faintly punctate, large fovea on lower frons near ventral
edge of antennal socket; lower frons and vertex usually with strongly de-
pressed central area. Antennal segment 4 subtriangular, 5 subrectangular,
6-10 rectangular. Pronotum regularly moderately striate, tubercles faint or
absent, impunctate; gibbosity deeply divided; posterior lobes transversely
and finely striate. Mesoscutellum granulate, posterior margin flat to faintly
depressed, posterolateral corners reflexed slightly. Elytra (Fig. 143) finely
but deeply punctate, with large tubercles and carinae; su-3 large, longer
than broad, ma-b obliquely carinate, ma longitudinally carinate along lateral
margin, anterior tubercles round, su-la faintly carinate. Pygidium finely and
faintly punctate, with smooth medial carina; large shallow foveae mostly
confluent. Prosternum (Fig. 144) more than twice as long as wide, pos-
terior half of process explanate, grooved and serrate in male; sides of pro-
sternal process parallel in female, not serrate and only slightly grooved
(Fig. 145). First abdominal segment impunctate, obscurely foveolate, non-
striate; last segment obscurely punctate with shallow confluent foveae, its
mid-ventral area flat with two groups of dense lateral hairs in male; female
with large, deep, mid-ventral fovea with fewer hairs than in male. One
large and one small spine on front tibia and single well developed spine
on middle tibia in male; female tibiae without spines.

Aedeagus (Fig. 26) robust, truncate; ejaculatory guide as in Figure 55.
Spermatheca (Fig. 82) with base of cornua slightly curved.

Types: Zimsen (1964) listed a single specimen of gibbosus, possibly
the type, in the Fabrician collection at Kiel (now on loan to the Zoological
Museum, Copenhagen). Dr. George W. Byers examined this specimen
and a specimen in the Paris Museum, believed to be the type of plicata.
Both were compared with a specimen in my collection and considered
to be conspecific with it. On the basis of these comparisons, I consider
plicata a junior synonym of gibbosus. The type locality for gibbosus was
given as North America. The type locality for plicata is probably South
Carolina, since it was collected by Bosc.

Discussion: This is the largest species of the genus in North America.
It has been confused with several others, including the highly tuberculate
insularis and bimaculatus. It also has a close affinity with platani, from
which it can be distinguished by its larger size, the large, rounded elytral
tubercles, regular and coarser striations on the pronotum, and especially

954

The University Science Bulletin

£

3

a

o
&

£

m

£

3
C
u,
1J

'Si

O

O

o

-Si

bo

o

ITS

1"

a

A Revision of the Subfamily Chlamisinae 955

the unequal tibial spines. Also, the truncate aedeagus is diagnostic for
gibbosus. Another closely related species, bebbianae, can be distinguished
from gibbosus by its smaller size, punctate pygidium, the presence of tibial
spines in the females, and the pointed shape of the aedeagus. Variation
creates problems in identification, since some individuals could be placed
in any one of these four species, if only one or two characters are used.
Each character must therefore be considered in combination with the others
mentioned.

Parasites: This species is heavily parasitized by Tetrastich us chlamytis
Ashmead and species of Ichneumonidae. Of 30 larvae collected at one
locality in Texas, 25 were parasitized.

Plant records: Salix sp., Rubus sp., Eubatus spp. (series with larvae), Qtiercus sp., wheat
and Phleum pratense.

Specimens examined (95 males and 186 females): (Map 6)

UNITED STATES: ALABAMA: Barbour Co., 18 June; Conecuh Co., 25 July; Houston
Co., 16 May; Marshall Co., 25 May; Mobile Co., 20 May, 1-8 June, July; Shelby Co., 16 Apr.,
17 July; Tallapoosa Co., Aug.; Tuscaloosa Co., 7 May. ARKANSAS: state record, 6 Oct.
DISTRICT OF COLUMBIA: Brightwood, 20 July; Washington, 7 May, 2 July. FLORIDA:
Alachua Co., 27 Feb., 14 Mar., 4-7 May, 12-24 June, 21 Oct.; Baker Co., 27 Apr., 26 May;
Brevard Co., 10 Oct.; Clay Co., 12 June; Columbia Co., 14 June; Dade Co., 22 July; Duval Co.,
no date; Gadsden Co., 10 July; Hamilton Co., no date; Hernando Co., June; Jackson Co.,

9 June; Nassau Co., 20 Aug.; Okaloosa Co., 30 July; Orange Co., Mar., 24 Apr.; Osceola Co.,
7 Mar.; Pinellas Co., 14 Feb.; Polk Co., 6 May; Putnam Co., 24 May; Seminole Co., 7 May,
29 June; Volusia Co., 29 July; Walton Co., 8-10 July. GEORGIA: Chatham Co., June, 5 July;
Clarke Co., 1 June; Cook Co., June; Dougherty Co., 23 July; Evans Co., 26 June; Fulton Co.,
3 Mar.; Paulding Co., 22 Oct., Stewart Co., 18 Aug. ILLINOIS: Franklin Co., 26 May;
Jackson Co., 27 May; Jefferson Co., 5 Aug.; Jo Daviess Co., no date; Marion Co., 31 May,
16 Aug.; Massac Co., 22 Apr.; Montgomery Co., 3 June; Perry Co., 6 Apr.; Pope Co., 13 May;
Pulaski Co., 7 June; Saint Clair Co., 23 Apr.; Washington Co., 5 Apr., 10 Aug.; Wayne Co.,
16 June. INDIANA: Boone Co., 2 May; Brown Co., 11 June; Owen Co., 25 May; Spencer Co.,
11 Aug. LOUISIANA: DeSoto Par., 21 Apr.; Jefferson Par., 11 Apr., 17 June; Orleans Par.,
28 Feb., 21 Apr.; St. Landry Par., 26 Apr.; St. Tammany Par., 4 June. MARYLAND: Prince
Georges Co., 29 June. MICHIGAN: Montcalm Co., 14 May. MISSISSIPPI: Claiborne Co.,
3 May; Granada Co., 11 June; Hancock Co., 21 May; Jackson Co., no date; Lowndes Co.,
6 May; Wilkinson Co., 1 Aug. MISSOURI: Boone Co., 3 June, 14 July, 12 Aug.; Callaway Co.,

10 July; Ozark Co., no date; St. Louis Co., 4 July. NEW MEXICO: Grant Co., 7 May. NORTH
CAROLINA: Buncombe Co., 12 May, 20-23 June; Johnston Co., 22 May; Moore Co., 4 May,
14 June; New Hanover Co., 15 July; Wake Co., 2 June. OHIO: Highland Co., 8 June;
Hocking Co., 20 June; Jackson Co., 6 Sept.; Scioto Co., 10 June; Washington Co., 15 June.
OKLAHOMA: Muskogee Co., 24 May-6 June, 19 Sept. SOUTH CAROLINA: Aiken Co.,
no date; Beaufort Co., 15 July; Chester Co., 13 June; Dorchester Co., 17 Aug.; Horry Co.,

11 Apr., 5 May; Orangeburg Co., 27 June. TEXAS: Anderson Co., 15 Apr.; Bastrop Co.,
13 July; Bexar Co., 21 May; Blanco Co., no date; Bowie Co., 15 June; Brazos Co., 11 Apr.-ll
May, 22-31 May, 13-14 June, 19 July, 29 Sept.; Cherokee Co., 2 Apr.; Colorado Co., 10 May;
Comal Co., 8 July; Galveston Co., 26 May, 10-13 Aug.; Gillespie Co., 14 June; Gonzales Co.,

19 Apr.; Lavaca Co., 27 May; Llano Co., 10 June; Kerr Co., 2 Apr., 18 Apr.; Matagorda Co.,

20 July; Panola Co., 8 May; Sabine Co., 25 Mar.; Tyler Co., Sept.; Uvalde Co., 10 June;
Val Verde Co., 7 June. VIRGINIA: Fairfax Co., 25 June, 8 July; Henrico Co., 26 June, 2 July;
James City Co., 25 May.

MEXICO: No locality, no date.

956 The University Science Bulletin

Neochlamisus bimaculatus new species

Diagnosis: This species has small velvety spots on the elytra, similar to
those found in the velutinus species group. It can also be identified by the
lack of tibial spines in the females and two spines on each front tibia in
males.

Description: Male holotype: 3.4 mm long; 2.6 mm wide. Color dark
brown with metallic reflections, antennae yellowish brown.

Head faintly punctate, center of clypeus depressed and finely vermiculate,
vertex finely striate. Antennal segment 4 subtriangular, 5 subrectangular,
6-10 rectangular. Pronotum rough, faintly punctate, moderately and irregu-
larly striate; tubercles faintly developed; posterolateral slopes with evenly
spaced striae; posterior lobes transversely, finely striate and impunctate.
Mesoscutellum granulate, posterior margin slightly depressed, posterolateral
corners reflexed slightly. Elytra (Fig. 146) with large tubercles and carinae,
moderately punctate; each discal velvety spot small, consisting of 4-6 deep
punctures separated by sharp ridges; su-2 slightly crescent-shaped and
carinate with su-la; anterior 1/5 of elytral suture closed, posterior 4/5
serrate; anterior tubercles carinate or acute. Pygidium with smooth medial
carina; large, confluent, shallow, circular foveae faintly vermiculate. Pro-
sternum more than twice as long as wide. First ventral abdominal segment
impunctate, granulate, with large obscure foveae; last abdominal segment
faintly punctate, finely striate ventrally and granulate laterally. Two well
developed spines on front tibia, single spine on middle tibia.

Aedeagus (Fig. 27) robust, truncate, with transparent apical edge;
ejaculatory guide (Fig. 56) long and slim, length five times width.

Female allotype: 3.7 mm long; 2.8 mm wide. Color and sculpturing
as in male. Apical spines of tibiae lacking; antennal segments as in male;
mid-ventral fovea of last abdominal segment large and deep, lateral hairs
long and more scattered than in males. Spermatheca (Fig. 83) conforms
in general shape to those of other species in the genus.

Types: Male holotype, Baton Rouge (East Baton Rouge Par.), Louisiana, 27 Apr. 1934,
F. E. Lyman (UK). Female allotype, Opelousas (St. Landry Par.), Louisiana, no date, G. R.
Pilate (USNM). Sixteen paratypes: \$, Langdale, Chambers Co., Alabama, H. H. Smith
(USNM); 1 $, Louisiana, A. Salle (USNM); 1 9, Harahan (Jefferson Par.), Louisiana, 17 June
1944, F. G. Werner (MCZ); 2 2, Lafayette (Lafayette Par.), Louisiana, 19 Mar. 1908, on
Rubus, R. A. Cushman (USNM); 2$, Alexandria (Rapides Par.), Louisiana, 17 Mar. 1908,
on Rubus, R. A. Cushman (USNM); 1$, Mobile (Mobile Co.), Alabama, 2 June, 1906,
H. P. Loding (UM); 1 2, Mobile Co., Alabama, June 1927, Darlington (MCZ); 2$, Grenada
Co., Mississippi, 21 April 1944, 11 June 1944, J. A. Wilcox (JAW); 2$, Morton (Scott Co.),
Mississippi, 10 Aug. 1959, S. D. and H. R. Burke (TAMU); 1 2, Montgomery Co., Mississippi,
7 May 1944, J. A. Wilcox (JAW); 12,6 mi. N. Loachapoka, Lee Co., Alabama, 4 Apr. 1964,
E. U. Balsbaugh, Jr., taken on Rubus (SDSU); 12, Ansby, Hancock Co., Mississippi, 8 Mar.
1966, H.R.Hepburn (JBK).

Discussion: This species is less variable than most of the others in the
species group. It resembles in many respects gtbbosas and insiduns, except

A Revision of the Subfamily Chlamisinae

957

Fig. 146. Neochlamisits bimacttlattts , dorsal and lateral aspects. Fig. 147. Neochlamisits
insitlaris, dorsal and lateral aspects.

for the small velvety spots on each elytron. A few specimens of the latter
two species have slight darkenings on the elytra that suggest faint velvety
spots, but the deep punctures and sharp ridges are lacking. Some specimens
have slightly convergent striations on the posterolateral slopes of the pro-

958 The University Science Bulletin

notum but they are never as rough and punctate as those found in the
velutinus species group. The striate pronotum suggests a relationship to the
gibbosus species group.

The finely vermiculate pygidium is similar to that of eubati, which also
has faint indications of velvety spots. Vermiculation is so fine in many
specimens that it is only clearly evident near the lateral and posterior borders
of the pygidium.

riant records: Riibus sp. (series), Quercus sp.

Specimens examined (12 males and 92 females): (Map 7)

UNITED STATES: ALABAMA: Chambers Co., no date; Clarke Co., no date; Lee Co.,
no date; Mobile Co., 2 June. ARKANSAS: White Co., 25 Apr. DISTRICT OF COLUMBIA:
Brookland, 17 June. FLORIDA: Alachua Co., 12 Nov.; Escambia Co., 5 Aug.; Jackson Co.,
23 Apr.; Leon Co., 1 Apr.; Nassau Co., 6 Aug. GEORGIA: Bibb Co., 4 May; Clarke Co.,

15 May; Crawford Co., no date; Fulton Co., 6-18 May; Glynn Co., 21 July; Talbot Co., no date.
LOUISIANA: East Baton Rouge Par., 27 Apr.; Jefferson Par., 17 June; Lafayette Par., 19
Mar.; Rapides Par., 17 Mar.; Sabine Par., 2 Apr.; St. Landry Par., no date. MARYLAND:
Montgomery Co., 23 May. MISSISSIPPI: Grenada Co., 21 Apr., 11 June; Hancock Co., 8 Mar.;
Montgomery Co., 7 May: Scott Co., 10 Aug. MISSOURI: St. Louis Co., no date; Vernon Co.,

16 Apr. NORTH CAROLINA: Buncombe Co., no date; Johnston Co., 16-22 May; Wake Co.,
29 May. SOUTH CAROLINA: Beaufort Co., 11-23 July; Charleston Co., 18 Aug.; Florence
Co., 5 Mar.; Lee Co., 20 Apr.: Oconee Co., 31 May; Richland Co., 8 June. TENNESSEE:
Shelby Co., 29 May. TEXAS: Anderson Co., 15 Apr.: Angelina Co., 6 Jan.; Bowie Co.,
26-28 Mar.; El Paso Co., 7 June; Harris Co., 12 Apr.; Jasper Co., 21-25 Apr.; Panola Co.,
8 May; Rusk Co., 8 June. VIRGINIA: Arlington Co., 18 May, 9 Sept.; Fairfax Co., 2 Aug.;
Fauquier Co., 5 May; Hampton City, 18 Apr.; Montgomery Co., 23 May; Nansemond Co.,
25 May: Northampton Co., 19-23 June; Nottoway Co., 7 May; Spotsylvania Co., May.

MEXICO: HIDALGO: 10 mi. NE Jacala, 1-2 Aug.

Neochlamisus insularis (Schaeffer)

Boloschesis insularis Schaeffer 1926:186.

Chlamisus jacobyi Blackwelder, 1946:648 (new synonymy; preoccupied by Chlamisus jacobyi
(Achard), 1913).

Diagnosis: This species is easily recognized by the numerous deeply
impressed punctures and many tubercles on the elytra, in addition to those
normally found in related species. Each front tibia of males is armed with
two subequal spines. The females usually have one spine, but there may be
another partially developed.

Description: Male: 3.8-4.6 mm long; 2.8-3.6 mm wide. Female: 4.1-4.8
mm long; 3.1-3.7 mm wide. Color dark brownish to purplish or reddish
bronze. Antenna yellowish brown; frons with large yellow spot in emargi-
nation of eye.

Head finely and faintly punctate, clypeus and vertex moderately punc-
tate, center of clypeus and upper frons depressed, posterolateral corners not
reflexed. Elytra (Fig. 147) moderately punctate, with large tubercles and
carinae; many small tubercles present in addition to normal complement;
general area of su-2a and ma-b represented by four well defined tubercles;
punctures large and depressed; no discal velvety spots; su-2 round at apex,

A Revision of the Subfamily Chlamisinae 959

broad at base; elytral suture closed and straight for anterior 1/5, remaining
suture serrate, anterior tubercles rounded. Pygidium with smooth medial
carina; foveae mostly confluent, granulate, slightly depressed, finely punc-
tate except coarsely punctate at base; small shiny areas between punctures
and foveae. Prosternum less than 1 1/2 times as long as wide, posterior
process longitudinally grooved, explanate and finely serrate. First and
last abdominal segments with large, granulate, punctate foveae, many foveae
confluent and obscure; mid-ventral area of last segment flat, with scattered
hairs in male; female with large, deep fovea, sides tumid, with long hairs.
Two subequal, well developed spines on front tibia separated by a U-shaped
space, single similar spine on middle tibia of male; female with single spine
(rarely two) on front and middle tibiae.

Aedeagus (Fig. 28) robust, blunt pointed, with clear apical edge; ejacula-
tory guide as in Figure 57. Spermatheca as in Figure 84.

Type: Male holotype of Boloschesis ins uteris Schaeffer; Wilson City,
Abaco, Bahamas, June, Coll. G. P. Engelhardt, USNM type 29465, labeled
Chtemys insularis Schaeffer. The name insularis has been used for two
other, different species of Chtemys. By placement of insularis Schaeffer in
Neochlamisus, it is no longer a homonym. The new name, Chtemisus
jacobyi Black welder (1946) proposed for insularis Schaeffer is no longer
necessary and was itself a homonym when proposed.

Discussion: Most members are the most roughly sculptured of the
North American species of Neochlamisus. This is because of the many
tubercles on the elytra in addition to those normally present. The usual
tubercles are small and divided so that the whole elytral surface is highly
tuberculate. In most cases the foveae on the abdomen and pygidium are
deep and convergent, imparting a rough appearance to these parts of the
insect.

Many specimens from the southern states have been misidentified as
gibbosus. In my paper on the genus Exema (Karren, 1966), it was noted
that southern specimens of certain species were usually darker and rougher
than those from the northern states. The rougher and more tuberculate
sculpturing is accompanied by differences in the form of the genitalia and
tibial spines. Some specimens of insularis are difficult to distinguish from
the more tuberculate specimens of gibbosus except by these last named
characters.

Predation: A series of 14 specimens was recovered by K. V. Krombein
from nests of Cerceris flavojaciata excavated at Kill Devil Hills, Dare Co.,
North Carolina, 9 August 1958.

Plant records: Barletia cristata, Quercus sp. (series), Myrica cerifera (series).

Specimens examined (53 males and 270 females): (Map 8)

UNITED STATES: ALABAMA: Dallas Co., 8-9 Apr. FLORIDA: Alachua Co., 13 Feb.,
Mar.-26 June, 15 July, 26 Sept.-2 Oct., 19 Nov.; Brevard Co., 18 May, 8 Nov.; Broward Co.,

960 The University Science Bulletin

26 July; Charlotte Co., 16 May; Clay Co., 12 June; Collier Co., 18 Mar.; Dade Co., 3 Mar.,
2-12 Apr., 12-26 May, 11 June, 25 June, 5 Nov.; DeSoto Co., 31 Mar., 30 Oct.; Duval Co.,

9 May, 27 June, July, 3 Aug., 16 Aug.; Escambia Co., 4 Aug.; Hamilton Co., 6 June; Hendry
Co., 11 Apr., 18-20 Apr.; Highlands Co., 4-6 Feb., 19 Apr.-l June, 16-24 June, 13 July;
Hillsborough Co., 3-5 Mar., 30 Mar.; Holmes Co., 12 June, 26 Aug.; Lake Co., 4 Mar.; Lee Co.,

10 Mar., 28 Apr.; Levy Co., 14 Mar., 7 June-19 July; Liberty Co., 24 Apr.; Marion Co., 24 Oct.;
Monroe Co., 28 Apr.; Nassau Co., 19 Aug., 31 Aug., 5 Oct.; Okaloosa Co., 15-16 Oct.;
Orange Co., 15 Feb., 29-30 Apr., 10 Aug.; Osceola Co., 19 Feb.; Palm Beach Co., 21 Feb.;
Pasco Co., 13 July; Pinellas Co., 21-30 Mar., 11 Apr., 28 Apr., 30 Oct.; Polk Co., 13 Feb.,
10 Mar., 28 Mar., 5-8 May, 8-10 Nov.; Putnam Co., Apr.; Seminole Co., 28 Apr., 3-7 May, 1
June, 1 Sept.; Suwanee Co., 15 June; Volusia Co., 17 Mar., 15 Apr., 26 Apr., 16 May. GEORGIA:
Chatham Co., 20-24 June; Fulton Co., 8 May; Glynn Co., 22 July; Mcintosh Co., 26 Apr.;
Richmond Co., 3 May; Ware Co., 26 June; Whitfield Co., 5 Oct. LOUISIANA: Jefferson Par.,
7 June; St. Tammany Par., 2 June; Vernon Par., 25 Feb., 14 Mar. MISSISSIPPI: Harrison Co.,
June; Jackson Co., 10 July. NORTH CAROLINA: Jones Co., 15 Sept. SOUTH CAROLINA:
Bamberg Co., 15 Aug.; Beaufort Co., 11-23 July. TEXAS: Blanco Co., no date; Uvalde Co.,
25 Aug.; Walker Co., 1 May, 9 July: Webb Co., no date.

N eochlamisus platani (Brown)
Chlamisus platani Brown, 1952:338-339.

Diagnosis: The bright yellowish bronze color when present is the
most obvious and easily recognized characteristic of this species. In addi-
tion, it differs from related species by the finely striate pronotum with
striae usually convergent on posterolateral slopes, foveolate pygidium and
ventral abdominal segments, and two equal spines on each anterior tibia
of the male.

Description: Male: 3.2-4.1 mm long; 2.4-2.9 mm wide. Female: 3.7-4.6
mm long; 2.8-3.3 mm wide. Color bright reddish to yellowish bronze,
occasionally dark brownish bronze; antennae light brown.

Head rough, clypeus and lower frons vermiculate, vertex striopunctate
between eyes, central area of upper frons depressed. Antennal segment 4
subtriangular, 5 subrectangular, 6-10 rectangular, segments 4-6 slightly
smaller in female than in male. Pronotum rough, finely and irregularly
striate, mp and la tubercles slightly developed; gibbosity deeply divided,
posterolateral slopes striate, striae converging on area at middle of base
of posterolateral slopes; posterior lobes finely and transversely striate. Meso-
scutellum granulate, with anterior medial carina, posterior margin slightly
depressed, posterolateral corners reflexed. Elytra (Fig. 148) with large
tubercles and carinae, moderately punctate, punctures depressed, anterior
tubercles rounded or sharply carinate; elytral suture closed anteriorly,
posterior 7/8 serrate. Pygidium with smooth medial carina, many deep to
shallow foveae, foveae usually shallow and confluent. Presternum nearly
1 1/2 times as long as wide, posterior process slightly grooved, explanate
and serrate. First abdominal segment with large, granulate, shallow foveae,
last segment with small mid-ventral foveae in male; mid-ventral fovea of
female large and deep with long hairs on anterolateral border. Two well

A Revision of the Subfamily Chlamisinae

961

Fig. 148. Neochlamisus platani, dorsal and lateral aspects. Fig. 149. Neochlamisits
bebbianae, dorsal and lateral aspects.

developed spines on front tibia of male; middle tibia of male and front
and middle tibiae of female with similar single spine.

Aedeagus (Fig. 29) acute at apex, robust; ejaculatory guide (Fig. 58)
long and slim, more than five times as long as wide, sheath hinged near
middle of asymmetrical paired structures. Spermatheca (Fig. 85) nearly
U-shaped with slight apical swelling.

Types: The male holotype and female allotype are both in the personal

962 The University Science Bulletin

collection of John A. Wilcox, Albany, New York. The above description
was taken from these specimens. They bear the following label data :
Lorain Co., Ohio, 7 Sept. 1947, Platanus, J. A. Wilcox, Collr. Eleven para-
types, from the type locality, collected on various dates were also examined.

Discussion: Most specimens are brightly colored, but samples from
apparently the same population may contain specimens of the dark form.
This color variation is not correlated with locality, host plant or season.

Considerable variation is also found in the sculpturing of this species.
Generally the posterolateral slopes of the pronotum are diagnostic and
have fine convergent striae, but in some specimens the striation is more
coarse and follows the contours of the gibbosity rather than converging on
the base. Others have only a few convergent and many sinuous striae. In
these specimens the other diagnostic characters of the species must be used
in order to make a correct identification.

The host specificity originally described for the species does not exist.
At least three samples have been taken from hosts other than Platanus in
each of which larval or adult feeding has been observed. These specimens
vary but are as similar to the type of platani as are the paratypes.

Females of this species exhibit considerable variation in the development
of the apical spines of the front and middle tibiae. This variation appears
to be geographical in that many of the eastern examples have a small second
spine or a raised edge on the front tibiae. A short series of females from
Estero, Florida, has two equal tibial spines similar to those found in males
from elsewhere.

The pronotum may be moderately and irregularly striate in all speci-
mens in certain samples. Host data are lacking for all such specimens,
but I have placed them in platani because of their agreement in other diag-
nostic characters, and because they come from within the geographical
range of the species. They will most likely be confused with bebbianae,
except for their bright reddish bronze color, moderately punctate elytra
and many carinate tubercles.

Plant records: Platanus occidentalis (series with larvae), Corylus americana (series),
Betttla nigra (series), Ulnius sp., Tri folium pratense.

Specimens examined (M> males and 148 females): (Map 9)

UNITED STATES: ARKANSAS: Pulaski Co., 28 Apr.: Washington Co., 6-27 |ul\.
COLORADO: El Paso Co., no date; Logan Co., 22 May. DISTRICT OF COLUMBIA: Wash-
ington, 14 June, 1-13 Aug. FLORIDA: Holmes Co., 6-12 May: Polk Co., 20 Mar.; Volusia
Co., no date. GEORGIA:"" Clarke Co., 21 Oct.: Glynn Co., 22 July; Whitfield Co., 25 May,
24 July. ILLINOIS: Alexander Co., 2 May; Champaign Co., ') May; Cook Co., 13 May;
Fayette Co., 19 Apr.; Gallatin Co., 27 May; Jefferson Co., 5 Aug.: Johnson Co., 4-7 June,

16 July; Logan Co., 11 June; McHenry Co., no date; Paulding Co., 6 Apr.; Pope Co., 10 May,

17 Oct.; Vermilion Co., 6 Apr.; Washington Co., 24 May. INDIANA: Harrison Co., 27 June,
28 Aug. IOWA: state record, 6 Sept. KANSAS: Douglas Co., 30 Apr.- 18 June, 25 Sept.,
15 Oct.; Reno Co., May; Shawnee Co., 24 Aug.; Wilson Co., 7 May. KENTUCKY: Butler
Co., 16 June. LOUISIANA: Claiborne Par., 14 July. MARYLAND: Montgomery Co., 6 Mar.,

A Revision of the Subfamily Chlamisinae 963

26 Apr., 23 May-23 June, 4 July, 25 July. MASSACHUSETTS: Hampden Co., no date.
MISSISSIPPI: Forrest Co., 24 Apr.; Grenada Co., 18 May. MISSOURI: Boone Co., 19 May,
10 June, 12 Aug.; Jackson Co., 5 July. NEW JERSEY: Cape May Co., 13 July. NEW
YORK: Suffolk Co., 14 May. NORTH CAROLINA: Moore Co., no date. OHIO: Delaware
Co., 21 June, 2 July; Franklin Co., 9 June; Hocking Co., 8 May, 20 May-6 June; Lorain Co.,
27-29 July, 17-24 Aug., 7 Sept. OKLAHOMA: Murray Co., 14 May. PENNSYLVANIA:
Montgomery Co., 4 May. SOUTH CAROLINA: Beaufort Co., 11 Oct. TENNESSEE: Lake
Co., 2 June; Montgomery Co., 16 May. TEXAS: Blanco Co., no date; Cherokee Co., 2 Apr.;
Cottle Co., 15 Mar.; Crosby Co., 2-17 Mar.; Gillespie Co., 5 Apr.; Harris Co., 3 Aug.; Jeff
Davis Co., 25 June; Kerr Co., 23 Apr.; Leon Co., 23 Aug.; Rusk Co., 8 June; Travis Co.,
13 Feb. VIRGINIA: Fairfax Co., 27 June, 8 Aug.; Loudon Co., 8 Oct.; Montgomery Co.,
5 May, 5 Oct., 19 Oct. WEST VIRGINIA: Greenbrier Co., 18 June.

Neochlamisus bebbianae (Brown)

Arthrochlamys bebbianae Brown, 1943:128-130.
Arthrochlamys alni Brown, 1943:128 (new synonymy).
Arthrochlamys tecta Brown, 1946:48-49 (new synonymy).
Chlamistts alni (Brown); Brown, 1961:971.

Diagnosis: One can identify this species by its moderately to coarsely
punctate pygidium, obscure foveae on the abdominal segments, moderately
punctate elytra, and irregular and moderately striate posterolateral slopes
of the pronotum.

Description: Male: 3.2-4.1 mm long; 2.6-3.0 mm wide. Female: 3.8-4.1
mm long; 2.6-3.1 mm wide. Color dark brownish bronze, occasionally
bright reddish bronze; antennae yellowish to light brown, gradually be-
coming darker toward apex.

Head moderately punctate, clypeus and genae vermiculate, vertex faintly
striopunctate, upper frons with a medial depression. Antennal segment 4
subcylindrical, 5 subtriangular, 6 subrectangular, 7-10 rectangular. Pronotum
rough, moderately and irregularly striate, mp tubercle faintly developed;
gibbosity deeply divided, posterolateral slopes irregularly striate, some but
never all striae convergent ventrally; posterior lobes finely and trans-
versely striate, impunctate. Mesoscutellum granulate, posterior margin flat
to slightly depressed, posterolateral corners reflexed slightly. Elytra (Fig.
149) with large tubercles and carinae, moderately punctate; anterior elytral
suture closed, posterior 4/5 serrate; anterior tubercles round, acute or
carinate. Pygidium moderately to coarsely punctate, with irregular medial
carina, depressed areas confluent. Presternum less than 1 1/2 times as long
as broad, posterior process faintly grooved and obscurely serrate. Last ab-
dominal segment foveolate, mid-ventral area flat with numerous long hairs
in male; female with large mid-ventral fovea with long anterolateral hairs;
first segment obscurely foveolate. Two well developed spines on front
tibia, single spine on middle tibia in male; female with single spines on
front and middle tibiae.

Aedeagus (Fig. 30) acute with transparent apical area; ejaculatory guide
(Fig. 59) long and narrow. Spermatheca as in Figure 86.

964 The University Science Bulletin

Types: Arthrochlamys bebbianae: male holotype, Arnprior, Ontario, reared from larva
taken on Salix bebbiana in August 1942, W. J. Brown, no. 5405 in the Canadian National Col-
lection. Arthrochlamys alni: female holotype, Norway Bay, Quebec, on Alntis incana ameri-
cana, 30 August 1942, W. J. Brown, no. 5404 in the Canadian National Collection. Arthro-
chlamys tecta: male holotype, Walsingham, Ontario, 29 May 1944, on Coryltts americana,
W. J. Brown, no. 5614 in the Canadian National Collection. Paratypes have been examined
from each of the type series.

Discussion: Brown first segregated the species in the gibbosus complex
in 1943. At that time it appeared that many, if not all, of the species were
host specific. Brown studied principally the Canadian species, more spe-
cifically those of eastern Canada. On the basis of samples collected from
one host plant, Brown described several "sibling species" in this complex.
He admitted that there was no morphological means of differentiating some
of these species. Others were so similar that only specimens in a series
associated with a particular host could be correctly identified.

By studying a large number of specimens from eastern United States,
including many series associated with host plants, I have found that the
species of this complex are less host specific than was at first suspected.
Some of the forms recognized by Brown as sibling species are only variants
of more widespread species. Further study of this group by Brown (1952)
suggested that some species may prefer a particular host, but will move to
other acceptable plants when the preferred host becomes unavailable. It
also seems that when the usual food plant is not present in a given locality,
different plants become the preferred hosts.

I have accordingly synonymized tecta and alni under bebbianae. This
species has a wide distribution with considerable morphological variation,
but no more than the other North American species of this genus. Speci-
mens are generally of medium size, dark in color and moderately tubercu-
late, striate and punctate. The posterolateral slopes of the pronotum are
usually somewhat irregularly striate, but in some specimens the striations
are not broken or convergent. In a few others the striations are irregularly
convergent. Individual tubercules and carinae of the elytra vary in every
species so that one must use these characters in combination with others to
identify species.

There is little variation in the form of the apical tibial spines. They are
sometimes difficult to see because of the position of the legs and may be
concealed by apical hairs.

Predation: A series of 55 specimens was recovered by J. C. Bridwell at
Vienna, Virginia, from nests of Cerceris fumipennis (Scullen, 1965).

Plant records: Salix bebbiana (series with larvae), S. hnmilis (series with larvae), Coryltts
americana (series), Alntts rugosa (series with larvae), A. serralata (series, literature record),
Ulmtis sp. (series), Acer rubrum (series with larvae), Vcrbascum sp., Iris versicolor, Juglans
rupestris (series), /. nigra, /. cinerea, Qiiercus sp. (series), Hamamelis rirgimana, Bettila nigra
(series with larvae).

Specimens examined (148 males and 905 females): (Map 10)

A Revision of the Subfamily Chlamisinae 965

UNITED STATES: ALABAMA: Baldwin Co., 15 May; Barbour Co., 24 Feb.; Hale Co.,
13 May; Mobile Co., 12 Apr., 17 June, July. ARKANSAS: Boone Co., no date. ARIZONA:
Pinal Co., no date; Santa Cruz Co., no date. CONNECTICUT: Fairfield Co., June, 5-18 July,
25 Aug.; Litchfield Co., 30-31 May, 23 June; New Haven Co., 14-20 June; New London Co.,

23 May, 4 July; Windham Co., 25 June. DISTRICT OF COLUMBIA: Washington, May, 5 June,
5 July, 31 July. FLORIDA: Bay Co., 20 May; Dade Co., 11 May; Highlands Co., 5 May,

16 June; Jackson Co., 7 Mar.; Lee Co., 6-12 May; Liberty Co., 16 Apr.; Manatee Co., 25 Mar.;
Nassau Co., 19 Aug.; Okaloosa Co., 9 July; Suwannee Co., 4 Aug. GEORGIA: Baldwin Co., 6
May; Bibb Co., 13 May; Chatham Co., 2-3 July; Clarke Co., 24 Apr.-2 May, 18 May; Crawford
Co., 12 May; Mcintosh Co., 28 Apr.; Meriwether Co., 9 May; Pike Co., no date; Rabun Co.,
10 May. ILLINOIS: Calhoun Co., 10 Oct.; Cook Co., 21-28 June, 5 Sept.; Lake Co., 24
Aug.; McHenry Co., no date; Pope Co., 30 Apr.; Woodford Co., 6 June. INDIANA: Lake
Co., 18 June. IOWA: Boone Co., 2 Oct.; Pottawatomie Co., 26 June. KANSAS: Douglas Co.,

24 Sept.; Leavenworth Co., 1 July; Montgomery Co., 26 Aug.; Riley Co., May; Shawnee Co.,
no date. KENTUCKY: state record, no date. LOUISIANA: Calcasieu Par., 10 Aug.; St.
Tammany Par., 29 Mar.; Vernon Par., 25 Feb. MAINE: Penobscot Co., 24 May; Piscataquis
Co., 29 June. MARYLAND: Baltimore Co., 6 June, 11 July; Montgomery Co., 12 Apr., 5 May,

17 May, 8-13 June, 5-10 July, 5 Aug., 17 Aug., 30 Aug.; Prince Georges Co., 17 May, 4 June,
4-5 July, 17 Sept.; St. Marys Co., 12 July, 3 Sept. MASSACHUSETTS: Barnstable Co., 4 July,
27 Aug; Bristol Co., 19 June, 6 Sept., 22 Sept.; Essex Co., 26 May, 12 June; Hampden Co.,
10 June, 28 June, 7 July, 16 Aug., 20-22 Sept.; Hampshire Co., 17 Aug.; Middlesex Co., 21 May-
10 June, 8 July, 22 July, 10-19 Aug., 30 Aug.; 7 Sept.-22 Oct.; Norfolk Co., 9 June;
Worcester Co., 19 May, 31 May, 21 Aug. MICHIGAN: Cheboygan Co., 26 June, 8 July,

I Aug.; Chippewa Co., 2 July; Grand Traverse Co., 27 May; Livingston Co., 12 June, 12
July; Osceola Co., 24 May; Washtenaw Co., 8 May; Wayne Co., 5 Sept., 5 Oct. MIN-
NESOTA: Anoka Co., 17 Sept.; Hubbard Co., 27 May, 4 June, 15-28 June, 11 Aug.; Millc
Lacs Co., 3 June; Olmstead Co., 4 July, Polk Co., 11 Aug.; Red Lake Co., 10 June; Stone Co., 15
Aug. MISSOURI: Boone Co., 20-23 Apr., 16 May; Howell Co., 8 June; Jackson Co., 25 May;
Lincoln Co., 26 Aug.; St. Louis Co., 7 June; Vernon Co., 4 Oct. NEBRASKA: Cherry Co.,
10 June; Douglas Co., 30 June; Holt Co., 9 Sept.; Keyapaha Co., 10 June; Nance Co., 1 July;
Thomas Co., 4-5 July. NEW HAMPSHIRE: Carroll Co., 24 June; Rockingham Co., 7 May.
NEW JERSEY: Atlantic Co., 5 June, 10-17 July; Bergen Co., 15 May; Burlington Co., 25 May,

4 July; Camden Co., 23 May-13 June; Cape May Co., 31 May, 13 July; Essex Co., 26 May, 8
June; Mercer Co., 23 Apr.; Morris Co., 26 May; Ocean Co., 13 May, 30 May, 24-26 Sept.;
Union Co., 9 Sept. NEW YORK: Albany Co., 11 May, 24 May, 11 June, 25 Aug.; Cortland Co.,

7 June; Dutchess Co., 2-11 June, 10 Aug., 30 Aug.; Erie Co., 28 June; Herkimer Co., 22 May;
Monroe Co., Oct.; Orleans Co., 15 June; Putnam Co., 4 May; Queens Co., June; Richmond Co.,
24 June-2 July; Rockland Co., 28 June; St. Lawrence Co., 30 June; Suffolk Co., 28 May; 11-17
June, Aug.; Sullivan Co., 3 Sept.; Tioga Co., 24 Apr.; Tompkins Co., 14-22 May, 4 June;
Westchester Co., 17 Mar., 26 May, 2 June, 30 Oct.; Wyoming Co., 2 June. NORTH CAROLINA:
Buncombe Co., no date; McDowell Co., 27 May; Moore Co., 4-10 May; New Hanover Co., 26
July. OHIO: Champaign Co., 22 May, 8 Oct.; Cuyahoga Co., 19 May; Delaware Co., 2 July:
Erie Co., 1 Sept.; Fairfield Co., 29 June, 15 Aug.; Fulton Co., 2 Sept.; Greene Co., 30 May-

5 June, 13 Sept.; Hocking Co., 13 May-2 June, 20 June, 2-10 July, 15-24 Sept.; Medina Co.,

II June, 28 Aug.; Ross Co., 21 Sept.; Scioto Co., 30 May-17 June; Wayne Co., 11 May.
OKLAHOMA: Comanche Co., 30 May. PENNSYLVANIA: Allegheny Co., 4 July; Dauphin
Co., 25-30 May, 4-5 July, 2 Sept.; Delaware Co., June; Fayette Co., 23 June; McKean Co.,
16 May, 9 June, 21-30 June; Monroe Co., no date; Northampton Co., 7-14 June, 28 June, 11 July;
Philadelphia Co., 15 June, 29 July; Pike Co., 9 June; Wyoming Co., 8 May. SOUTH
CAROLINA: Charleston Co., 2 July; Pickens Co., 6 July. SOUTH DAKOTA: Bennett Co.,
11-12 Sept.: Corson Co., 8 Sept.; Yankton Co., 15-20 June. TENNESSEE: Campbell Co.,
27 Aug.; Davidson Co., 23 June; Sevier Co., 14 June. TEXAS: Aransas Co., 31 Mar.; Brazos
Co., 4 Mar., 19-24 Apr., 3 May, 15 May, 9 Oct.; Cameron Co., 30 Mar., 22-26 Apr., 25 June,

8 Dec; Colorado Co., 11-20 Apr.; Cottle Co., 15 Mar.; Dallas Co., no date; Eastland Co.,
4 June: Gillespie Co., 7 May, 6 June, 23 June: Harris Co., 4 Aug.; Hidalgo Co., 24 Mar., 10 May;
Houston Co., 22 Apr.; Jeff Davis Co., 19 June, 4 July; Kendall Co., 11 July; Kerr Co., 31 Mar.-
14 Apr., 22-27 Apr.; Leon Co., 23 Aug.; Montague Co., 26 Apr. -2 May; Montgomery Co.,

966 The University Science Bulletin

23 Aug.; Pecos Co., 24 Apr.: Real Co., 29 July: Travis Co., 28 June; Uvalde Co., 23 May,
13-15 June, 25 Aug.: Val Verde Co., 9 Mar., 3 May. VERMONT: Orange Co., 23-27 May;
Windham Co., spring. VIRGINIA: Arlington Co., 1-19 May, 8 Aug., 20-30 Sept.; Caroline
Co., 30 Sept., 12 Oct.; Fairfax Co., 30 Apr.-5 May, 18 May, 22 June: Prince William Co.,
28 Sept.: Spotsylvania Co., 21 Mar., May. WISCONSIN: Polk Co.. 13 Aug.; Shawano Co.,
13-16 Sept.; Washburn Co., 22 June; Wood Co., 13 Sept.

CANADA: NOVA SCOTIA: Hants Co., no date. ONTARIO: Addington and Lennox Co.,
16 June; Carleton Co., 3-13 May, 30 May, 19 Aug., 31 Aug., 3 Sept., 14 Sept.; Oxford Co.,
30 May: Parry Sound Co., 31 July; Prince Edward Co., 19 June; Renfrew Co., 2 May, 24 May,
15 Aug., 7-8 Sept.; Russell Co., 20 Aug., 5 Sept. QUEBEC: Pontiac Dist., 24 May, 30 Aug.,
3-10 Sept.

Neochlamisus eubati (Brown)

Arthrochlamys eubati Brown. 1943:126-127.
Arthrochlamys coryli Brown, 1946:48 (new synonymy).
Chlamisus eubati (Brown): Wilcox, 1954:396.

Diagnosis: This species is recognized by the deeply divided pronotal
gibbosity, vermiculate pygidium, vermiculate and striate last abdominal
segment and the two equal spines on the front tibia of the males.

Description: Male: 33-3.7 mm long; 2.5-2.7 mm wide. Female: 3.7-4.5
mm long; 2.8-3.4 mm wide. Color black to dark reddish brown with
metallic reflections; antennae yellowish brown at base, becoming darker
brown at apex, mandibles light brown to black.

Head moderately punctate, center of clypeus depressed and vermiculate,
vertex coarsely punctate. Antennal segment 4 subtriangular, 5 subrectangu-
lar, 6-10 rectangular. Pronotum moderately striate, impunctate, tubercles
faintly developed; posterolateral slopes regularly striate, gibbosity deeply
divided; posterior lobes transversely and finely striate, impunctate. Meso-
scutellum granulate, posterior margin depressed, posterolateral corners re-
flexed slightly. Elytra (Fig. 150) with large tubercles and carinae, moder-
ately punctate; punctures with radiating striae; often faint discal velvety
spots consisting of 3-4 deep punctures separated by sharp ridges; su-2 broad
with low-angled anterior slope; anterior part of elytral suture closed, pos-
terior 4/5 serrate, anterior tubercles rounded. Pygidium vermiculate, mod-
erately punctate, foveae shallow, confluent and obscure. Presternum more
than 1 1/2 times as long as wide. First abdominal segment obscurely foveo-
late and striate; last segment striate and vermiculate, mid-ventral area with
two groups of lateral hairs in male, female mid-ventral fovea large but
shallow, with hairs more scattered than in male. Two well developed spines
on anterior tibia, single similar spine on middle tibia of male; female
without tibial spines.

Aedeagus (Fig. 31) with bluntly pointed, scarcely deflexed apex; ejacula-
tory guide (Fig. 60) long and slim, length five times width. Spermatheca
as in Figure <S7.

Types: Arthrochlamys eubati: male holotype, Kirk's Ferry, Quebec, on Rtibits (Eubatus)
sp., 27 May 1942, W. J. Brown, no. 5403 in the Canadian National Collection. Arthrochlamys

A Revision of the Subfamily Chlamisinae

967

Fig. 150. Neochlamisus eubati, dorsal and lateral aspects. Figs. 151-153. Neochtamtsus
comptoniae. 151, dorsal and lateral aspects; 152, cephalic aspect; 153, presternum.

968 The University Science Bulletin

coryli: female holotype, Walsingham, Ontario, on Coryltts americana, 30 May 1944, no. 5613
in the Canadian National Collection. I have examined paratypes from both type series.

Discussion: The type series (all females) of coryli was collected on hazel.
Brown considered coryli distinct from eubati because of different host plants,
sex ratios, and because of slight morphological distinctions. I can find
no consistent diversity between these two named forms, either in the two
series of paratypes or in numerous other examples. Differences pointed out
in the original description are well within the range of variation of eubati,
especially when one considers the species from its entire geographic range.

Since several species in this group have been taken on a variety of host
plants and exhibit widely differing sex ratios from one collection to another,
it seems unlikely that the host plant and sex ratio ascribed to coryli are
valid specific characters. There is less variation in this species than in several
others of the gibbosus group. Some is found in the striate and vermiculate
surface of the last abdominal segment. In a few specimens the lateral areas
of the segments are granulate and faintly punctate. There is sometimes
a little irregularity in the striation of the posterolateral slopes of the pro-
notum.

The mid-ventral fovea of the last abdominal segment of the female is
not as deep as it is in most other species of the gibbosus group.

Plant records: Genm album (series). Aster sp., Rtibus frondostis, R. I Eitbatus) spp. (series
with larvae), .thins sp. (series), Phletim pratcnse (series with larvae), Barbarea vulgaris (series),
Potentilla jrttticosa (series, literature record), Ulmits sp., Corylns americana, Sohdago sp.

Specimens examined (66 males and 429 females): (Map 11)

UNITED STATES: ARKANSAS: White Co., 25 Apr. ARIZONA: Navajo Co., 20 July.
COLORADO: Boulder Co., 9 Apr., 16 June, 8-10 Aug. CONNECTICUT: Fairfield Co., 9
May, 30 May-10 June: Litchfield Co., 14 May, 23 June: New Haven Co., 9-22 June: New
London Co., 5-9 Aug. DISTRICT OF COLUMBIA: Washington, 23-25 July. FLORIDA:
Pinellas Co., 27 Feb. GEORGIA: Murray Co., 9 Aug. ILLINOIS: Alexander Co., 7 Oct.; Clay
Co., 31 May; Crawford Co., 8 June; Jackson Co., 26 June: Johnson Co., 23 May; Mason Co.,

I Aug.; Piatt Co., 24 May. INDIANA: Brown Co., 2 Oct.; Crawford Co., 1 June; Lake Co.,
24 May, 22 Sept.: Marion Co., 17 May; Perry Co., 16-23 May; Spencer Co., 24-25 May; Starke
Co., 18 June. IOWA: Dickinson Co.. 1 July, 22 Aug.; Henry Co., 13 June. KANSAS:
Shawnee Co., no date. KENTUCKY: Bullitt Co., 9 May. LOUISIANA: state record, no
date. MARYLAND: Baltimore Co., Apr., 6 June; Frederick Co., 18 June; Montgomery
Co., 30 Apr., 10-18 May, 8-15 June; Prince Georges Co., 6 Mar., 14-15 May, 9 Aug.,
16 Sept.; Washington Co., 8 June. MASSACHUSETTS: Bristol Co., 19 May; Essex Co.,
3 June; Hampshire Co., June; Middlesex Co., 17-27 May, 23-27 June, 1 Aug., 22-28 Aug.,
28 Sept.; Norfolk Co., 17 June: Suffolk Co., 26 June, 21-25 July. MICHIGAN: Allegan Co.,

II Sept.; Berrien Co., Aug.; Oscoda Co., 19 Aug. MINNESOTA: Anoka Co., 3 June; Big
Stone Co., no date; Hubbard Co., 17 June; Olmstead Co., 13 June, Sept.; Ottertail Co., no date;
Red Lake Co., 3 Aug.; St. Louis Co., 1 June, 5 Sept. MISSOURI: Boone Co., 25 May; Camden
Co., May: Iron Co., 19 May; Jefferson Co., no date; Pike Co., 20 Apr., 3 Oct.; St. Louis Co.,
5 Sept.; Stoddard Co., 17 June. MONTANA: Carter Co., 5 Sept.; Rosebud Co., 20 June.
NEBRASKA: Cuming Co., June; Thomas Co., 9 June, 4 July, 4 Aug. NEW HAMPSHIRE:
Rockingham Co., 6 June. NEW JERSEY: Bergen Co., 25 Apr., 16 June: Burlington Co.. 4
June; Camden Co., 6 Mar., 27-30 May; Essex Co., 30 May; Hudson Co., 21 Apr.; Mercer Co.,
1 Aug.: Middlesex Co., 22 Apr.; Morris Co., 15 Apr., 29 Apr., 13-29 June: Ocean Co., 9 May;
Passaic Co., 25 July, 7 Aug.; Sussex Co., 24 May, 6 June; Union Co., 3 June; Warren Co.,

A Revision of the Subfamily Chlamisinae 969

21 May. NEW YORK: Albany Co., 15-31 May, 3-11 June; Dutchess Co., 27 May, 16 June;
Kings Co., 27 May, 18 July; Monroe Co., 30 May; Nassau Co., 6 Apr., 24 June; New York Co.,
11 Apr., 24 May; Niagara Co., 4 Aug.; Orange Co., 26 May; Orleans Co., 15 June; Rensselaer
Co., 17 Aug.; Rockland Co., 27 May; Saratoga Co., 2 June; Suffolk Co., 14 May, 23 July;
Tompkins Co., 17-21 May, 4-8 June; Ulster Co., 18 June; Westchester Co., Sept. NORTH
CAROLINA: Buncombe Co., 12 May, 23 May; Haywood Co., May-June; Macon Co., no date.
NORTH DAKOTA: Burleigh Co., 19 Sept.; Cass Co., 16 June; Traill Co., 4 Aug. OHIO:
Clinton Co., 12 May, 29 May, 21 June; Greene Co., 2 June, 23 June; Hocking Co., 10-23 May,
2 June, 14-20 June, 2 July, 5 Sept.; Jackson Co., 19 Aug.; Lorain Co., 6 June; Perry Co., 9
June; Scioto Co., 5 May, 9 June; Summit Co., 15-16 June. PENNSYLVANIA: Allegheny Co.,
May; Bucks Co., no date; Centre Co., 15 June; Delaware Co., 7 May, 27 May, 2-16 June, 30
Aug.; Franklin Co., 13 June; Montgomery Co., 13-17 May, 29 May; Northampton Co., 31 May-
7 June; York Co., 15 June. SOUTH DAKOTA: Brookings Co., no date; Brule Co., 5 June;
Custer Co., 19 June; Fall River Co., 11 Sept.; Jackson Co., 9 June; Lawrence Co., 26 July,
10-11 Sept.; Marshall Co., 18 Sept. UTAH: Uintah Co., 13 June. VIRGINIA: Alexandria
Co., 4 May, 18 June, July-Aug.; Arlington Co., 9-18 May, 1 June, 28 June, 8 July; Fairfax Co.,
17 Apr., 16 June, 14 Aug., 27 Aug. WEST VIRGINIA: Greenbrier Co., no date.

CANADA: ALBERTA: Edmonton, 5-9 May, 1 Nov. ONTARIO: Carleton Co., 9 May,
7 June, 22-30 July, Aug.-Sept.; Renfrew Co., 26 Aug. QUEBEC: Wright Dist., 27 May, 18
Sept.

MEXICO: No locality, no date.

Neochlamisus jragariae (Brown)
ChlamUus jragariae Brown, 1952:337-338.

Diagnosis: The faintly divided pronotal gibbosity, vermiculate pygidium
and single tibial spines in the males together distinguish this species.

Description: Male: 3.2-3.5 mm long; 2.3-2.5 mm wide. Female: 3.5-4.2
mm long; 2.6-2.9 mm wide. Color black to dark reddish bronze, antennae
yellowish brown at base becoming darker brown at apex, mandibles dark
brown.

Head moderately punctate, center of clypeus slightly depressed, lower
frons and clypeus vermiculate, vertex faintly punctate and striate. Antennal
segment 4 subtriangular, 5 subrectangular, 6-10 rectangular. Pronotum
moderately striate, impunctate, tubercles faintly developed; posterolateral
slopes regularly striate, gibbosity faintly divided; posterior lobes transversely
and finely striate, impunctate. Mesoscutellum granulate, posterior margin
slightly depressed, posterolateral corners reflexed slightly. Elytra with mod-
erate tubercles and few carinae, finely punctate, punctures with radiating
striae; discal velvety spots faint, consisting of several deep punctures; su-2
broad, with slowly rising anterior slope; anterior elytral suture closed, pos-
terior 4/5 serrate; anterior tubercles small and rounded. Pygidium ver-
miculate, moderately to coarsely punctate, most of circular areas confluent
and obscure. Prosternum about twice as long as wide. First abdominal
segment obscurely foveolate and striate; last segment striate and vermicu-
late, mid-ventral area with two groups of lateral hairs; female mid-ventral
fovea shallow and large, with hairs more scattered than in male. Single
well developed spine on anterior and middle tibiae of male; female without
tibial spines.

970 The University Science Bulletin

Aedeagus (Fig. 52) as in eubati; ejaculatory guide (Fig. 61) long and
slim. Spermatheca as in Figure 88.

Type: Male holotype, Hatfield Point (Kings Co.), New Brunswick, 17 Sept. 1951, G. T.
Morgan, on Fragaria ananassa; no. 5984 in the Canadian National Collection. Allotype, same
data. I have examined one male and two female paratypes.

Discussion: A very short series was available for study so that informa-
tion about variability is limited. This species is very similar to eubati, except
for the single spine on the front tibia of the males, faintly divided pronotal
gibbosity and other slight morphological differences. Specimens of jragariae
seem to be less tuberculate; the velvety spots are less distinct and the elytral
punctures more widely separated than in specimens of eubati. These last
mentioned characters may prove to be insignificant when more specimens
are examined.

Plant record: Fragaria ananassa (scries).

Specimens examined (5 males and 1 (> females): (Map 13;

UNITED STATES: MAINE: Hancock Co., 8-9 Aug.: Washington Co., 13-20 June.
MICHIGAN: Presque Isle Co.. 22 Aug. MINNESOTA: Hubbard Co., 3 June, 15 June;
Kanabec Co., 27 June: St. Louis Co., 31 May. NEW JERSEY: Middlesex Co., 4 July (?). NEW
YORK: Orange Co., 20 July. VERMONT: Windham Co., spring.

CANADA: NEW BRUNSWICK: Kings Co., 17 Sept.; Northumberland Co., 9 July.

Neochlamisus comptoniae (Brown)

Arthrochlamys comptoniae Brown, 1943:125-126.

Atrochlamys charaedapnes Brown; Proctor, 1946:186 (misidentification and misspelling).

Diagnosis: This species is very similar to cribnpennis except that it is
a little larger, punctation on the pygidium and abdominal segments is more
uneven, the elytral sculpturing is rougher and the tibial spines in the males
are more widely separated. It differs from assimilis by the shorter ejacula-
tory guide and the absence of front tibial spines of the male.

Description: Male: 3.2-3.3 mm long; 2.2-2.4 mm wide. Female: 3.7-4.1
mm long; 2.6-3.0 mm wide. Color dark brownish to reddish bronze; an-
tennae yellowish brown at base, becoming darker brown at apex; irons
with small yellow spot in emargination of eye.

Head coarsely punctate on vertex and on lower frons and clypeus; large
central depression on upper frons moderately punctate. Antennal segment 4
subtriangular, 5 subrectangular, 6-10 rectangular. Pronotum (Fig. 152)
coarsely and irregularly striate, with scattered faint punctures, tubercles
faint or absent; gibbosity divided, posterolateral slopes regularly striate;
posterior lobes transversely, finely striate, impunctate. Mesoscutellum rough,
faintly striate, posterior margin flat to slightly depressed, posterolateral
corners not reflexed. Elytra (Fig. 151) with small tubercles and carinae,
coarsely punctate, punctures usually deep with radiating striae; su-2 broad
and rounded, anterior tubercles small and rounded. Pygidium shiny with

A Revision of the Subfamily Chlamisinae 971

smooth medial carina, unevenly punctate, granulate foveae small and faint.
Presternum (Fig. 153) over twice as long as wide. First abdominal segment
vermiculate and obscurely foveolate, last segment punctate laterally, mid-
ventral surface flat, with two groups of long to short hairs separated by
narrow longitudinal smooth area in male; female fovea large, shallow to
deep, with scattered lateral hairs. Two well developed spines separated by
U-shaped space on front tibia, single similar spine on middle tibia of male;
female without tibial spines.

Aedeagus (Fig. 35) broadly pointed, with apex strongly deflexed, large,
transparent apical area; ejaculatory guide (Fig. 62) long and robust, length
five times width. Spermatheca as in Figure 89.

Type: Male holotype, Natick (Middlesex Co.), Massachusetts, 8 Aug. 1942, on sweet fern,
C. A. Frost, no. 5407 in the Canadian National Collection. Male and female paratypes of this
species have been examined.

Discussion: This species is closely related to chamaedaphnes but differs
mainly by having two spines instead of one on the front tibia of the males
and more evenly striate posterolateral slopes on the pronotum. Females are
extremely difficult to identify because of subjective interpretation of this
latter character. It is much easier to identify a series containing both males
and females, although males are less frequent in both species. The host
plant will be some help in identification, although neither species is re-
stricted to a single host. There is very little difference between comptomae
and assimilis except for those characters listed in the key. The host plants
are different, and assimilis has a limited and more southern distribution.

The species referred to by Proctor (1946) as Atrochlamys charaedapnes
is a misspelling and a misidentification of specimens of comptomae. There
are a few specimens of comptoniae , including a large series from Cass Co.,
Minnesota, that have a moderately divided pronotal gibbosity and unevenly
striate posterolateral pronotal slopes. In these two respects they resemble
chamaedaphnes, but in all others, including the two apical spines of the
tibia in the males, they are typical comptoniae.

A series of 18 specimens labeled "Farmingdale, Florida," are probably
mislabeled. The locality does not appear in gazeteers of Florida, and
Florida is outside the normal range of the species. There are other speci-
mens collected at Farmingdale, New York, by the same collector at approxi-
mately the same date. Four specimens labeled Las Mercedes, California, are
considered incorrectly labeled since the locality is so far removed from the
known distribution of the species. These localities are not included on the
distribution map.

Plant records: Myrica asplenijolia ( = Co<nptonia percgnna ) (series with larvae), Solidago
sp. (series), Kalmia sp. (series), and Corylus sp. (series).

Specimens examined (66 males and 460 females): (Map 12)

972 The University Science Bulletin

UNITED STATES: CONNECTICUT: Litchfield Co., 23 Mar., 8-14 May, 14-23 June;
New Haven Co., 2 Sept.; New London Co., 20 May, 26-28 July, 5-6 Aug.; Tolland Co., 15 May,
6 June. MAINE: Cumberland Co., 1-13 Aug.; Hancock. Co., 3 June, July, 12 Sept.; Lincoln
Co., 15 June, 5 Aug.; Oxford Co., 8 Sept.; Washington Co., 21) June, 5 July, 16 Aug. MARY-
LAND: Prince Georges Co., 15 May, 21 July. MASSACHUSETTS: Barnstable Co., 18 Aug.,
4 Sept.; Bristol Co., 11 May, 9-14 June, 24 Aug., 6 Sept.; Essex Co., 10 June; Franklin Co.,
20 Aug.; Hampden Co., 14 July, 1 Oct.; Middlesex Co., 23 Apr., 6-14 May, 27 May-24 June,
4 July, 22-26 July, 8 Aug.-29 Sept., 10 Oct.; Nantucket Co., Sept.; Norfolk Co., 27 May, June,
1-2 Sept.; Suffolk Co., 20 June, 25 July; Worcester Co., 27 Aug., 14 Sept. MICHIGAN: Glad-
win Co., 4 Sept.; Presque Isle Co., 31 July. MINNESOTA: Cass Co., 21-23 Aug.; Hubbard
Co., 4 June. NEW HAMPSHIRE: Belknap Co., 3 Aug., 22 Aug.; Carroll Co., 24 June, 19 Aug.;
Coos Co., July; Grafton Co., 29 June, 1 Aug.; Hillsboro Co., Sept.; Rockingham Co., no date;
Strafford Co., 1 June, 14 Aug. NEW JERSEY: Atlantic Co., 26 May, 12 June, 5 July; Camden
Co., 4 May, 30 May, 9 Aug.; Cape May Co., 28 May, 10 June; Monmouth Co., 13-16 June;
Ocean Co., 4-9 May, 24-27 May, 17 July, 29 Aug. NEW YORK: Albany Co., 25 Aug.;
Clinton Co., 27 June, Essex Co., 1 Sept., 28 Sept., 5 Oct.: Nassau Co., 12 Sept.; Orange Co.,
6 July; St. Lawrence Co., 26-27 Aug.; Saratoga Co., 25 Aug.; Suffolk Co., 26 Apr., 28 May,
20 June, 31 July, 25-27 Aug.; Warren Co., 28 Aug.; Wyoming Co., no date. PENNSYLVANIA:
Dauphin Co., 6-11 June; Monroe Co., no date; Pike Co., 30 May-June, 17 Sept.; Tioga Co.,
no date.

Neochlamisus assimilis (Klug)

Chlamys assimilis Klug, 1824:239.

Chlamys affinis Klug, 1824:115-116.

Chlamistts azalae Brown, 1961:972-973 (new synonymy).

Diagnosis: This species is recognized by the presence of a single spine
on the front tibia of the female, and the elongate form of the ejaculatory
guide.

Description: Male: 3.2 mm long; 2.3 mm wide. Female: 3.5-4.0 mm
long; 2.6-3.8 mm wide. Color dark brownish to yellowish bronze; antennae
yellowish brown at base, becoming darker brown at apex; frons with small
yellow spot in emargination of eye.

Head coarsely punctate on vertex, lower frons and clypeus; large central
depression on upper frons faintly punctate. Antennal segment 4 subtri-
angular, 5 subrectangular, 6-10 rectangular. Pronotum coarsely and irregu-
larly striate, finely punctate, tubercles faint or absent; gibbosity slightly
divided, posterolateral slopes regularly striate; posterior lobes transversely,
finely, carinate; posterior margin depressed, posterolateral corners slightly
reflexed. Elytra with small tubercles and carinae, moderately punctate,
punctures usually deep with faint radiating striae, su-2 broad and rounded,
anterior tubercles small and rounded. Pygidium shiny, with smooth medial
carina, coarsely punctate at base, apex becoming finely punctate near middle.
Presternum less than twice as long as wide. First abdominal segment
faintly vermiculate and obscurely foveolate; last segment punctate laterally
to foveolate near center; mid-ventral surface in male flat, with large group
of long and short hairs, female with large shallow fovea and fewer scat-
tered lateral hairs than in male. Two large spines separated by U-shaped

A Revision of the Subfamily Chlamisinae 973

space on front tibia, single similar spine on middle tibia of males; female
with single similar spine on anterior and middle tibiae.

Aedeagus (Fig. 34) broadly pointed with apex strongly reflexed; apical
area large, transparent; ejaculatory guide (Fig. 63) long and slim, length
eight times width, length of basal orifice 1/3 of total length. Spermatheca
as in Figure 90.

Type: The Humboldt University Museum, Berlin, has many of Klug's
types. There are two female specimens in the collection that are considered
by F. Hieke, the curator of Coleoptera, to be the type series. I here desig-
nate one of them as lectotype. The type locality is stated as North America.
Chlamisus azaleae: male holotype, Highlands (Macon Co.), North Caro-
lina, 12 June 1957, W. J. Brown, swept from Azalea, no. 6917 in the Cana-
dian National Collection.

Discussion: Female paratypes of azaleae agree with the lectotype of
assimihs, which was borrowed from the Berlin Museum. The only male
specimen examined of this species was a male paratype of azaleae. Since
few specimens were available for study, little is known of variation in the
species.

The one specimen from Indiana seems to be out of the range of this
species but it is morphologically the same as the other specimens. Addi-
tional collecting may indicate a much wider range than is now known for
this species.

Plant records: Azalea sp. (series).

Specimens examined (1 male and 8 females): (Map 12)

UNITED STATES: FLORIDA: Jackson Co., 7 Mar.; Osceola Co., no date. INDIANA:
Porter Co., 30 July. NORTH CAROLINA: Macon Co., 26-29 June. SOUTH CAROLINA:
Oconee Co., no date. VIRGINIA: Fairfax Co., 18 Apr.

Neochlamisus chamaedaphnes (Brown)

Arthrochlamys chamaedaphnes Brown, 1943:130.

Diagnosis: The coarsely striate posterolateral slopes of the pronotum
are curved and uneven, the gibbosity is slightly divided, and the elytral
tubercles are large. Males have a single spine on the front tibia.

Description: Male: 3.2-3.7 mm long; 2.5-2.7 mm wide. Female: 3.6-4.2
mm long; 2.5-3.0 mm wide. Color dark brown to reddish bronze. Antennae
yellowish brown at base, becoming darker brown at apex.

Head finely and faintly punctate, moderately punctate on clypeus and
vertex. Antennal segment 4 subtriangular, 5 subrectangular, 6-10 rectangu-
lar. Pronotum finely punctate, moderately and irregularly striate, tubercles
faintly developed; gibbosity moderately divided, posterolateral slopes un-
evenly striate; posterior lobes transversely, finely striate, impunctate. Meso-
scutellum subquadrate, granulate, posterior margin slightly depressed, pos-
terolateral corners reflexed. Elytra (Fig. 154) with large tubercles and

974

The University Science Bulletin

Fig. 154. Neochlamisus chamaedaphnes, dorsal and lateral aspects. Figs. 155-157. Neo-
chlamisus cribripennis. 155, dorsal and lateral aspects; 156, cephalic aspect; 157, prosternum.

carinae, moderately to coarsely punctate, su-2 broad and rounded; anterior
elytral suture closed, posterior 4/5 serrate; anterior tubercles rounded.
Pygidium shiny, unevenly moderately to coarsely punctate; foveae shallow
and obscure. Prosternum more than 1 1/2 times as long as wide. Abdominal
segment partially foveolate, mid-ventral surface of last segment flat with
two slightly separated groups of long hairs in male; female with large and
deep mid-ventral fovea, lateral hairs more scattered than in male. Single
well developed apical spine on front and middle tibiae of male; smaller
spines in female.

Aedeagus (Fig. 35) robust, bluntly pointed, with transparent apical

A Revision of the Subfamily Chlamisinae 975

area; ejaculatory guide (Fig. 64) long and robust; length five times width.
Spermatheca as in Figure 91.

Types: Male holotype, South March, Ontario, on Chamaedaphne caly-
culata, 19 Sept. 1942, W. J. Brown, no. 5406 in the Canadian National Col-
lection. Allotype and paratypes, same data. I have examined a number
of male and female paratypes of this species.

Discussion: The females are sometimes difficult to separate from those
of bebbianae because of similarities in the pronotum and tibial spines. The
posterolateral slopes of the pronotum are more coarsely and irregularly
striate in bebbianae, and the pronotal gibbosity is more deeply divided.
The striations of the pronotum are fairly constant in chamaedaphnes and
useful in identifying females not associated with males.

Except for Chamaedaphne calyculata, all the host plants of this species
are shared by many species of the gibbosus group. Corylus americana, for
example, is a host of platani, bebbianae, eubati, chamaedaphnes, and prob-
ably comptoniae. Specimens of chamaedaphnes have been taken on Rubus
sp., Alnus rugosa, Potentilla fritticosa and V accinium sp., all host plants
of one or more other species of Neochlamisus. It appears that the preferred
host of chamaedaphnes, as in several other species of this group, may differ
in widely separated localities, and that other acceptable hosts occur in each
locality.

One specimen labeled Portland, Oregon, is probably mislabeled. It has
not been included on the map or in the list of specimens examined.

The reference of Proctor (1946) to Atrochlamys charaedapnes in Maine
is a misspelling and misidentification. The species referred to is N. comp-
toniae.

Plant records: Potentilla jruticosa (literature record), V accinium sp. (series), Alnus rugosus
(series), Rubus occidentalis (series), Chamaedaphne calyculata (series with larvae), and
Corylus americana (series).

Specimens examined (59 males and 179 females): (Map 8)

UNITED STATES: CONNECTICUT: Fairfield Co., no date; New Haven Co., 4 Oct.
DISTRICT OF COLUMBIA: Washington, 29 July, 30 Sept. KENTUCKY: state record, no
date. MARYLAND: Prince Georges Co., 17 Sept. MASSACHUSETTS: Middlesex Co., 18
Apr., 10-13 June, 7-13 Sept., 26-29 Sept., 2 Oct.; Norfolk Co., 15 Aug.; Plymouth Co., 29 May.
MICHIGAN: Allegan Co., 17 June; Cheboygen Co., 21 June-3 July, 22 July; Gladwin Co.,
4 Sept.; Grand Traverse Co., 27 May; Kent Co., 9 Sept.; Luce Co., 8 Sept.; Marquette Co.,
15 June; Schoolcraft Co., 25 Aug.; Van Buren Co., 1 Aug.; Washtenaw Co., 7 July; Wayne Co.,
Aug. MINNESOTA: Crow Wing Co., 3 July; Hubbard Co., 29 May, 11 Aug.; Lake of the
Woods Co., 7 Aug.; Mille Lacs Co., 13-21 May, 3 June; Pine Co., 7 Aug. NEW JERSEY:
Burlington Co., 29 June, 12 July; Gloucester Co., 28 June; Hunterdon Co., 30 Apr.; Ocean
Co., 25 May, 3 July, 23 Aug. NEW YORK: Chenango Co., 2 Sept.; Genesee Co., 1 July;
Herkimer Co., no date; Monroe Co., Oct.; Nassau Co., 12 Sept.; Richmond Co., 1 Oct.;
Steuben Co., 20 June; Suffolk Co., 8 May. OHIO: state record, 7 May. VIRGINIA: Arlington
Co., 3-5 May, 18 May, 26 June, 20 Sept.; Nelson Co., 26 June; Spotsylvania Co., 21 Mar.
WISCONSIN: Oneida Co., 7-8 Sept.

CANADA: ONTARIO: Carleton Co., 20 May, 10 June, 15 Aug., 17-19 Sept.

976 The University Science Bulletin

Neochlamisus cribripennis (LeConte)

Chlamys cribripennis LeConte, 1878:614.

Arthrochlamys cribripennis (LeConte); Brown, 1943:125.

Chlamisus cribripennis (LeConte); Proctor, 1946:186.

Diagnosis: The strongly and evenly punctate pygidium and abdominal
segments, together with the subtriangular shape of the prosternum, finely
punctate lateral pronotum, and faintly divided pronotal gibbosity, serve to
differentiate this species.

Description: Male: 2.8-2.9 mm long; 2.1-2.3 mm wide. Female: 3.3-3.4
mm long; 2.4-2.6 mm wide. Color brownish to reddish bronze; antennae
yellowish brown; frons with small yellow spot in emargination of eye.

Head moderately punctate on vertex to coarsely punctate on clypeus.
Antennal segment 4 subtriangular, 5 subquadrate, 6-10 rectangular. Prono-
tum (Fig. 156) coarsely and irregularly striate, lateral areas finely punctate,
tubercles faint or absent; gibbosity slightly divided, posterolateral slopes
regularly striate, posterior lobes transversely and faintly striate, impunctate.
Mesoscutellum granulate, posterior margin flat to slightly depressed, pos-
terolateral corners not reflexed. Elytra (Fig. 155) with rounded, faintly
developed tubercles and carinae. Coarsely to moderately punctate, su-2 small
but broad; anterior elytral suture closed, posterior 4/5 serrate. Pygidium
shiny and moderately to coarsely punctate; no large circular granulated
areas. Prosternum (Fig. 157) about twice as long as wide, sides not or only
slightly emarginate; posterior process grooved longitudinally. Abdominal
segments moderately punctate, almost nonfoveolate; mid-ventral fovea of
last segment nearly flat and faintly punctate with scattered hairs, hairs
more numerous in male than in female. Two well developed spines on
front tibia separated by a V-shaped space, single spine on middle tibia of
male; female usually without tibial spines.

Aedeagus (Fig. 36) robust, pointed, with narrow truncate apex; ejacu-
latory guide (Fig. 65) long and slim, length five times width. Spermatheca
as in Figure 92.

Type: Male holotype, collected near Detroit (Wayne Co.), Michigan, in the LeConte collec-
tion (MCZ).

Discussion: My specimens of cribripennis agree with the holotype ex-
cept for minor differences. The foveae of the first abdominal segment are
less distinct and more confluent in the holotype. Specimens are among
the smallest of the gibbosus group and also of North American Neo-
chlamisus.

There is little variation in the diagnostic characters given, but this may
be the result of the availability of too few specimens and the limited number
of localities represented. Some specimens of comptoniae have the pro-
sternum and pronotal gibbosity similar to those found in cribripennis.

A Revision of the Subfamily Chlamisinae 977

although the gibbosity is usually more deeply divided in comptomae. The
abdominal segments are also partially punctate in comptoniae but never as
distinctly, deeply, and uniformly as in cribripennis.

Plant records: Vaccinium angnsti folium. This plant may not be the only host since only
a few specimens were available for study, but it is a host in which large series of both larvae
and adults have been observed feeding.

Specimens examined (9 males and 26 females): (Map 13)

UNITED STATES: INDIANA: Porter Co., 22 June. MASSACHUSETTS: Middlesex Co.,
19 Sept. MICHIGAN: Cheboygan Co., 11-22 July; Gladwin Co., 4 Sept.; Wayne Co., no date.
MINNESOTA: Chisago Co., 16 Aug. WISCONSIN: Douglas Co., 16 Aug.

CANADA: ONTARIO: Carleton Co., 9 June, 10 July, 22-30 July, 6 Sept. QUEBEC:
Wright Dist., 28 May; "Duparquet," 2 Aug.; "Peribonca," 24 Aug.

Neochlamisus tuberculatus (Klug)

Chlamys tuberculata Klug, 1 824 : 1 1 7- 1 1 8.
Chlamisus tuber culata (Klug); Fattig, 1948:8.

Diagnosis: This small, brightly colored species can be distinguished by
the obscurely striate front half of the pronotum, impunctate first abdominal
segment, obscurely foveolate pygidium and granulate texture of the elytra.

Description: Male: 2.8-3.0 mm long; 2.1-2.3 mm wide. Female: 2.9-3.3
mm long; 2.2-2.7 mm wide. Color bright brownish to reddish bronze,
underside sometimes partially greenish bronze, antennae yellow to light
brown.

Head finely punctate, except more coarsely punctate on clypeus and
vertex. Antennal segment 4 subtriangular, 5 subrectangular, 6-10 rectangu-
lar. Pronotum (Fig. 159) faintly, finely, and irregularly striate, tubercles
faintly developed; gibbosity moderately or slightly divided, posterolateral
slopes moderately and regularly striate; posterior lobes transversely, finely
striate, impunctate. Mesoscutellum subquadrate, granulate, posterior margin
slightly depressed. Elytra (Fig. 15S) with small tubercles and carinae, finely
punctate; several deep punctures on disc in some specimens; anterior
elytral suture closed, posterior 4/5 serrate. Pygidium with faint medial
carina, centrally foveolate; small shallow fovea becoming convergent and
obscure laterally. Prosternum (Fig. 160) subtriangular, about twice as long
as wide; sides only slightly emarginate. First abdominal segment obscurely
foveolate, impunctate; last abdominal segment with many small foveae
laterally, mid-ventral area flat with two groups of hairs in male; female
with slightly depressed large fovea with scattered hairs laterally. Two
well developed tibial spines, separated by U-shaped space, on front tibia,
single similar spine on middle tibia; female tibiae without spines.

Aedeagus (Fig. 37) robust with acute apex; ejaculatory guide (Fig. 66)
long and slim. Spermatheca as in Figure 93.

Types: The type of tuberculatus is supposed to be in the Humboldt
University Museum, Berlin, but was not located there. I assume, from the

978

The University Science Bulletin

Figs. 158-160. Ncoc/ilamisus tuberculatus. 158, dorsal and lateral aspects; 159, cephalic
aspect; 160, presternum.

descriptions by Klug and Lacordaire, as well as the type locality, that I
have correctly identified the species. Its small size and sculpturing make
this one of the more easily recognized species of the gibbosus group.

Discussion: This species has many similarities to cribnpennis but the
two are easily distinguished by the characters given in the key. They also
are quite distinct geographically and can be separated by means of locality.
Both are somewhat restricted in distribution as compared to most of the
other species in this group. Most of the specimens examined show little
variation in the usually variable characters mentioned for the group. A few
specimens, however, are more tuberculate and striate than usual and have
moderately deep punctures in the region of the elytra where some species
have a small velvety spot. The other elytral punctures are also larger and
more depressed. Other characters in these individuals agree with those of
tuberculatus so I am placing them in this species.

Plant records: Vacanium sp. and Qttercus sp.

Specimens examined (42 males and 48 females): (Map 13)

UNITED STATES: ALABAMA: Tallapoosa Co., Aug. ARKANSAS: Benton Co., 6 June.
FLORIDA: Alachua Co., 13 Mar., 19 Apr.-6 May, 21-31 May; Brevard Co., 8 Nov.; Duval Co.,
no date; Hamilton Co., 6 June; Hendry Co., 26 Feb., 20 Apr.; Highlands Co., 5 May, 16 June;
Hillsborough Co., 13 Apr. -7 May; Lee Co., 23 Apr., 3-5 May; Levy Co., 4 Apr., 17 July;

A Revision of the Subfamily Chlamisinae 979

Okeechobee Co., 28 Feb.; Osceola Co., 19 Feb.; Pasco Co., 13 July; Pinellas Co., 20 Jan., 9-11
Feb., 21 Mar., 2 Apr., 29-30 Apr., 14 Nov., 26 Nov., 15 Dec, 30 Dec; Seminole Co., 5 Apr.,
May; Suwannee Co., 3 July; Volusia Co., 10-15 Apr., 25 May, 5 Oct. GEORGIA: Charlton Co.,
June-25 July; Clarke Co., 2 May; Union Co., 25 May. INDIANA: Harrison Co., 26 Sept.
MARYLAND: Anne Arundel Co., 21 May. NORTH CAROLINA: Harnett Co., no date;
New Hanover Co., no date. SOUTH CAROLINA: Kershaw Co., 29 May.

LITERATURE CITED

Achard, J. 1913. Note de synonymie (Col. Phytophaga). Bull. Soc Ent. France 1913:160.
. 1914. Coleoptera Phytophaga, fam. Chrysomelidae, subfam. Chlamydinae & Sphaero-

charinae. In Genera Insectorum (P. Wytsman, ed.), fasc. 160, 25 pp.
Baly, J. S. 1865-1867. Phytophaga Malayana. Trans. Ent. Soc. London. 3rd ser., 4:1-300.
. 1878. Descriptions of new species of phytophagous Coleoptera. J. Linn. Soc. Zool.

14:336-356.
Berthold, A. A. 1827. Latreille's Naturliche Familien des Theirreichs. Aus dem Franzosischen.

Mit Anmerkungen und Zusatzen. Weimar. 606 pp.
Blackwelder, R. E. 1946. Checklist of the coleopterous insects of Mexico, Central America,

the West Indies, and South America. Part 4. Bull. U.S. Nat. Mus. 185:551-763.
Boving, A. G., and F. Craighead. 1930-1931. An illustrated synopsis of the principal larval

forms of the order Coleoptera. Ent. Am. (n.s.) 11:1-351.
Bowditch, F. C. 1913. Notes on Chlamydae with descriptions of a few new species. Trans.

Am. Ent. Soc 39:1-21.
Brown, W. J. 1943. The Canadian species of Exema and Arthrochlamys (Coleoptera,

Chrysomelidae). Canadian Ent. 75:119-131.
. 1944. The dimorphism in the male copulatory organ of the chrysomelid Arthro-
chlamys bebbianae Brown. Canadian Ent. 76:70-72.
. 1946. Some new Chrysomelidae, with notes on other species (Coleoptera). Canadian

Ent. 78:47-54.

. 1952. Some species of Phytophaga (Coleoptera). Canadian Ent. 84:335-342.

. 1961. Notes of North American Chrvsomelidae (Coleoptera). Canadian Ent. 93:967-

977.
Brues, C. T., A. L. Melander and F. M. Carpenter. 1954. Classification of insects. Bull.

Mus. Comp. Zool. vol. 108, 917 pp.
Chapuis, M. F. 1874. Histoire naturelle des insects. Genera des Coleopteres. Famille des

phytophages. vol. 10, 455 pp.
Chen, S. H. 1940. Attempt at a new classification of the leaf beetles. Sinensia 11:451-479.
Chevrolat, L. A. A. 1835. Coleopteres du Mexique. Strasbourg, fasc. 5, 50 pp.
Chujo, M. 1940. Fulcidacinae of Formosa (Col. Chrvsomelidae). Trans. Nat. Hist. Soc.

Formosa 30:265-293.
. 1942. Fulcidacinae of Japan proper and Loo-Choo (Coleoptera-Chrysomelidae). Trans.

Nat. Hist. Soc. Formosa 32:77-100.
. 1955. A taxonomic studv on the Chrysomelidae (Insecta-Coleoptera) from Formosa.

IX Subfamily Chlamisinae (1). Tech. Bull. Kagawa Agric Coll. 7:51-59.
. 1956. A taxonomic studv on the Chrysomelidae (Insecta-Coleoptera) from Formosa.

IX Subfamily Chlamisinae (2). Tech. Bull. Kagawa Agric. Coll. 7:153-167.
Costa, O. G. 1835. Fauna del Regno di Napoli, Famiglia de'Coccinigliseri, o de'Gallinsetti.

Emitteri Napoli. 23 pp.
Emden, F. van. 1932. Die Larven von Discoloma cassideum Reitt. (Col. Colyd.) and Sf{warraia

paradoxa Lac. (Col. Chrysom.). Zool. Anz. 101:1-17.
Fabricius, J. C. 1776. Genera Insectorum. Chilonii. 310 pp.

. 1798. Supplementum entomologiae systematicae. Hafniae. 522 pp.

Fall, H. C. 1927. Expedition of the California Academy of Sciences to the Gulf of California

in 1921. The Chrysomelidae (Coleoptera). Proc. California Acad. Sci. ser. 4,

16:381-395.
Fattig, P. W. 1948. The Chrysomelidae or leaf beetles of Georgia. Emory Univ. Mus. Bull.

no. 6, 47 pp.
Gemminger, M., and E. von Harold. 1874. Catalogus coleopterorum hucusque descriptorum

synonymicus et systematicus, Monachii. 11:3233-3479.
Gressitt, J. L. 1946. Chinese chrvsomelid beetles of the subfamily Chlamisinae. Ann. Ent.

Soc. America 39:84-100.
Ihering, H. von. 1907. Les Mollusques fossiles du Tertiaire et du Cretace superieur de l'Argen-

tine. Anal. Mus. Nac. Hist. Nat. Buenos Aires, ser. 3, 7:1-611.

980 The University Science Bulletin

Iherixg, H. von, and R. von Ihering. 1905. Bibliographia 1902-1904 historia natural e

anthropologia do Brazil. Revista do Museu Paulista (1904) 6:584-659.
Jacobson, G. G. 1924. Annotationes synonymicae et sustematicae de coleopteris. Rev. Russe

d'Entomologie 18:237-243.
Jacoby, M. 1881. Biologia Centrali-Americana, Insecta, Coleoptera, Cryptocephalidae, Chlamydi-

dae, Lamprosomidae, Eumolpidae. 6(1):73-144.
. 1889. Biologia Centrali-Americana, Insecta, Coleoptera, supplement to Phvtophaga.

6(1) suppl.:81-168.
Jaques, H. E. 1951. How to Know the Beetles. W. C. Brown Co., Iowa. 372 pp.
Karren, J. B. 1964. Protective coloration and form in the North American genus Exema

(Chrysomelidae, Coleoptera). Proc. North Central Branch, Ent. Soc. Amer. 19:77-79.
. 1966. A revision of the genus Exema of America, North of Mexico (Chrysomelidae,

Coleoptera). Univ. Kansas Sci. Bull. 46:647-695.
Klug, J. C. F. 1824. Entomologische Monographien. Berlin. 242 pp.
Knoch, A. W. 1801. Neue Beytrage zur Insectenkunde. Leipzig. Vol. 1, 208 pp.
Lacordaire, J. T. 1848. Monographic des coleopteres subpentameres de la famille des phyto-

phages, vol. 2. Mem. Soc. Roy. Sci. Liege vol. 5, 890 pp.
LeConte, J. L. 1878. Descriptions of new species. In The Coleoptera of Michigan (H. G.

Hubbard and E. A. Schwarz, eds.). Proc. Am. Philos. Soc. 17:593-626.
LeConte, J. L., and G. H. Horn. 1883. Classification of the Coleoptera of North America.

Smithsonian Misc. Coll. 26:1-567.
Leng, C. W., and A. J. Mutchler. 1927. Supplement 1919 to 1924 (Inclusive) to Catalogue

of the Coleoptera of America, North of Mexico. Sherman, New York. 78 pp.
. 1933. Second and Third Supplements 1925 to 1932 (Inclusive) to Catalogue of the

Coleoptera of America, North of Mexico. Sherman, New York. 112 pp.
Leonard, M. D. 1928. A list of the insects of New York. Cornell Univ. Agri. Expt. Sta. Mem.

101, 1121 pp.
Linell, M. L. 1898. New species of Coleoptera of the family Chrysomelidae, with a short

review of the tribe Chlamydini. Proc. U.S. Nat. Mus. 20:473-485.
Monros, F. 1951. Diez nuevas especies de Chlamisus neotropicales (Col., Chrysomelidae) (6a

contribucion al conocimiento de Chlamisinae). Dusenia 2:253-272.
. 1952. Revision de las especies argentinas de Chlamisinae (Col., Chrysomelidae) (5a

contribucion al conocimiento de Chlamisinae). Acta Zoologica Lilloana 10:489-672.

. 1954. Notes and synonyms in Chrysomelidae. Proc. Ent. Soc. Washington 56:23-26.

.1959. Notas sobre Chrysomelidae, (Coleoptera). Acta Zoologica Lilloana 17:1-24.

Monte, T. de. 1957. Prima nota sull'apparato plasmatore degli escrementi, caratteristica

modificazione dell'intestino retto femminile in alcune tribu' della sottofamiglia

Cvltrinae (Col., Chrvsomelidae). Estratto Dalle Memorie Della Societa Entomologica

Itahana 36:143-159. '
Navatas, E. 1944a. Fulcidacinae (Co.) de Monte Alegre com a descricao de uma especie nova.

Papeis Avulsos 6:75-80.
. 1944b. Algumas notas sobre a nomenclatura dos Fulcidacideos (Col. Chrysomeloidea).

Papeis Avulsos 4:213-220.
— . 1946. Os genotipos da fam. Fulcidacidac (Col. Chrvsomeloidea). Papeis Avulsos

7:243-245.
Neave, S. A. 1939. Nomenclator Zoologicus. London. Vol. 1, 957 pp.
Olivier, A. G. 1808. Entomologie, ou histoire naturelle des insectes, avec leurs caracteres

generiques et specifiques, leur description, leur synonymie, et Ieur figure enluminee.

Coleopteres. Paris. Vols. 6 and 8 (plates).
Pallister, J. C. 1953. The leaf beetles of North Central Mexico collected on the David Rocke-
feller Mexican Expedition (Coleoptera, Chrvsomelidae). Am. Mus. Novitates no. 1623,

pp. 1-95.
Powell, E. F. 1941. Relationships within the family Chrysomelidae (Coleoptera) as in-
dicated by the male genitalia of certain species. Am. Mid. Nat. 25:148-195.
Proctor, W. 1946. Biological Survey of the Mount Desert Region, Part VII, (Revision of

Parts I & VI), the Insect Fauna. Wistar Institute, Philadelphia. 566 pp.

Rafinesque, C. S. 1815. Analyse de la nature ou tableau de l'univers et des corps organises.

Palermo. 224 pp.
Roding, P. F. 1798. Museum Boltenianum. Pars secunda. Hamburg. 199 pp.
Schaeefer, C. F. A. 1906. On new and known genera and species of the family Chrysomelidae.

Sci. Bull., Mus. Brooklyn Inst. Arts Sci. 1:221-253.
. 1919. Synonymical and other notes on some species of the family Chrysomelidae and

descriptions of new species. J. New York Ent. Soc. 27:307-340.

A Revision of the Subfamily Chlamisinae 981

. 1926. New species of Boloschesis (=zChlamys) with notes on known species

(Coleoptera; Chrysomelidae; Fulcidacinae). Proc. Ent. Soc. Washington 28:181-187.
Scullen, H. A. 1965. Review of the genus Cerceris in America north of Mexico (Hymen-

optera: Sphecidae). Proc. U.S. Nat. Mus. 116:333-548.
Spett, G., and M. Lewitt. 1926. Versuch einer Verwertung des Receptaculum seminis als

systematisches Merkmal bei den Chrysomeliden. Arch. Naturgesch. Abt. A, 92:96-142.
Torre-Bueno, J. R. de la. 1937. A Glossary of Entomology. Brooklyn Ent. Soc, New York.

336 pp.
Varma, B. K. 1954. Taxonomic value of spermathecal capsules as subfamily characters

among the Chrysomelidae (Coleoptera). Indian J. Ent. 17:189-192.
. 1955. Phylogenetic study of the family Chrysomelidae, Coleoptera. Current Science

21:18-19.
Wilcox, J. A. 1954. Leaf beetles of Ohio (Chrysomelidae: Coleoptera). Ohio Biological

Survey Bull. 43:353-506.
Zia, Y. 1936. Comparative studies of the male genital tube in Coleoptera, Phytophaga. Sinensia

7:319-352.
Zimsen, E. 1964. The type material of I. C. Fabricius. Munksgaard, Copenhagen. 656 pp.

982

The University Science Bulletin

Map 1. Range of Pseitdochlamys semirufescens, Chlamisus flayidiis, C. arizonensis, C.
tiigromactilatus, C. tcxamis, and C. quadrilobatus . Map 2. Range of Chlamisus joveolatus, C.
huachucae, and C. maculipes. Each spot represents one or more collections within a county or
at a locality. S=state record. R=indefinite locality within a country.

A Revision of the Subfamily Chlamisinae

983

Map 3. Range of Diplacaspis prosternalis. Map 4. Range of Keochlamisus velutinus and N.
memnonius. Each spot represents one or more collections within a county or at a locality.

984

The University Science Bulletin

Map 5. Range of Neoc/ilamisus moestificus. Map 6. Range of Neochlamisus scabripennis,
N. subelatus, and N. gibbosus. Each spot represents one or more collections within a county or
at a locality. S=state record.

A Revision of the Subfamily Chlamisinae

985

Map 7. Range of Neoc/ilaniisus bimaculatus. Map 8. Range of Neochlamistis chamaedaph-
ncs, and N. insularis. Each spot represents one or more collections within a county or at a
locality. S=state record.

986

The University Science Bulletin

Map 9. Range of Neochlamisits platani. Map 10. Range of Neochlamisus bebbianae. Each
spot represents one or more collections within a county or at a locality. S=state record.
R = inderinite locality record within a country.

A Revision of the Subfamily Chlamisinae

987

Map 11. Range of Neoc/tlamistts ettbati. Map 12. Range of Neochlamisits comptoniae and
N. assimilis. Each spot represents one or more collections within a county or at a locality.
S=state record. R=indefinite locality within a country.

988

The University Science Bulletin

13

* N. frogor
A N. cribrip
■ N. tuberc

Map 13. Range of Neoc/ilamisits fragariae, N. cribripennis , and N. tubercidatits. Each spot
represents one or more collections within a county or at a locality.

3 2044 093 362 291

Date Due

f>ni«»'n

•A**