S44 Pan
D7 4WHE
Lat,

es /
With Cuitnk, ay

ZOOLOGICA

SCIENTIFIC CONTRIBUTIONS OF THE
NEW YORK ZOOLOGICAL SOCIETY

FROM THE TROPICAL RESEARCH
STATION IN BRITISH GUIANA

VOLUME III. NUMBER 3

(Tropical Research Station Contribution Number 97)

A STUDY OF SOME SOCIAL BEETLES IN BRITISH
GUIANA AND OF THEIR RELATIONS TO fa
THE ANT-PLANT TACHIGALIA

By WILLIAM MorRTON WHEELER

rt ie

Eavebet Ss Ho ED BY THE rote 0 Sal Ot: Gi ee We
toe s2OOL0GICAL PARK, NEW YORK

DECEMBER 24, 1921

HSON
SEP 12 1958
LIBRARY

fA

If organic evolution had stopped with insects it would
still have been a succession of achievements that angels
might desire to look into. The entomologist watches by
the most copious fountain of wonder in the world—a
well of surprises for eye and intellect—J. ARTHUR
Tuomson, “The System of Animate Nature,” p. 545.

Volume III, Number 3

A STUDY OF SOME SOCIAL BEETLES IN BRITISH
GUIANA AND OF THEIR RELATIONS TO
THE ANT-PLANT TACHIGALIA*

By WILLIAM MorTON WHEELER

INTRODUCTION.

The materials for the following paper were obtained during
July, August and September, 1920, while I was working with
Prof. I. W. Bailey on the myrmecophytes of British Guiana. We
had gone to the Tropical Laboratory of the New York Zoological
Society at Kartabo for the purpose of making an intensive study
of the relations of the ants to such plants as Cecropia, Triplaris,
Cordia and Tococa; Prof. Bailey’s investigations being primarily
concerned with the anatomical peculiarities of the plants, my own
with the identification and habits of their various ants. During
these investigations I encountered two species of beetles of such
unusual social habits that I was led to devote considerable atten-
tion to their behavior. They live in the fusiform enlargements
of the petioles of a singular tree, Tachigalia, which is also inhab-
ited by numerous species of ants. Both the beetles and the ants
cultivate coccids in the hollow petioles, and with the beetles and
ecoccids there are, moreover, several associated, parasitic or
syneeketic insects. And since there are also other insects, besides
those already mentioned, associated with the plant, we may
regard the latter as the focus of a very interesting and compli-
cated biocoenose.

The elements of this biocoenose are so numerous and hetero-
geneous that I have had to appeal to several specialists for assist-
ance in making identifications and writing descriptions. Dr. EH.
A. Schwarz and Mr. H. S. Barber kindly studied and described
the social beetles and a little Coccinellid which feeds on their
coccids. Dr. Adam Boving made a fine study of the beetle larve;

*Contributions from the Eaton erer Laboratory of the Bussey In-
stitution, Harvard University. No.

36 Introduction

Mr. Harold Morrison identified the coccids; the Hymenopterous
and Dipterous parasites of the latter were studied by Prof. C. T.
Brues and Dr. E. P. Felt. Dr. J. Bequaert, Prof. Herbert Osborn,
Prof. J. W. Folsom and Dr. R. V. Chamberlin identified a wasp,
two Membracids, a Collembolan and a Myriopod found living in
Tachigalia petioles; and Prof. Roland Thaxter found and iden-
tified a fungus growing on the surfaces of one of the species of
social beetles. I have added the reports of several of these inves-
tigators as appendices to the present paper, and wish to express
my great indebtedness to all of them for their generous assist-
ance and to Mr. Wm. Beebe for his many kindnesses to Prof.
Bailey and myself while we were at Kartabo and the pains he
has taken, since our return to Boston, to ascertain further facts
concerning the Tachigalia and other myrmecophytes and to
collect additional species of Formicidz. So many of the ants in-
habiting the Tachigalia prove to be new to science that I have had
to provide a special paper for their taxonomic descriptions
(Zoologica III, No. 4).

Although I have endeavored to acquire a comprehensive
knowledge of the various insects associated with the Tachigalia,
I am aware that my account must be very fragmentary. The
territory covered was limited and in other parts of British Guiana
the same tree undoubtedly harbors other insects. This is indi-
cated by the few published observations on the trees of the genus
Tachigalia and their insects in Brazil and other parts of South
America. Moreover, only the leaves and petioles and their
inhabitants were studied and there are probably many peculiar
insects that live only in the wood and seeds or merely visit the
flowers, which in some species, at least, are conspicuous and
sweet-scented. The seeds were found only after they had germi-.
nated, and the trees showed no indications of flowering during
our stay in British Guiana. Our inference that they might bloom
during the winter months has not been confirmed, since up to the
time of this writing (March, 1921), Mr. Beebe has seen no
flowers on any of the trees which Prof. Bailey marked for obser-
vation. I trust, nevertheless, that my account of the Tachigalia.
biocoenose will give such a picture of the astonishing complexity
and exuberance of the insect fauna of the Neotropical Region,
and of the fierce competition among these organisms on the one

Introduction 330/

hand and of their intimate co-operation on the other, as to
stimulate some future investigator to complete my observations.
At any rate, the following pages may serve to direct the atten-
tion of our younger entomologists to one of the many wonderful,
almost untouched fields for investigation in tropical America.

BRITISH
GUIANA

5 MILES

LOCATION OF THE TROPICAL RESEARCH STATION OF THE
NEW YORK ZOOLOGICAL SOCIETY

The circle represents a radius of six miles.

THE TACHIGALIA.

The Leguminous myrmecophytes of the genus Tachigalia
comprise more than a dozen known species of trees and shrubs
belonging to the forest formation (hylea) of the Guianas and
the Amazon basin. The first species, paniculata, was described
by Aublet as long ago as 1775, and re-described by Tulasne in
1844. The former derived the generic name from the Carib
“tachigali,” “tachi” being the term employed by the natives of
the Guianas and Brazil for the stinging ants of the genus Pseu-
domyrma, which regularly inhabit the swollen petioles of the
species of Tachigalia and the hollow branches and trunk of the
various species of Triplaris. Spruce (1869) was also familiar
with several species of Tachigalia and their ant-inhabited petioles.
Owing to our not finding the trees at Kartabo in bloom, Prof.
Bailey and I have been unable as yet to ascertain their specific
name. They closely resemble, however, at least in their younger
stages, the species described by Harms (1906) as T. formicarum
and figured by Ule (1907), who discovered it in Brazil (Plate I).

I first found the Tachigalia on the Puruni Trail near the
Kartabo laboratory on July 22. The specimens were small and
slender, from a foot and a half to six or seven feet tall, with only
two or three to about a dozen leaves, and were growing in the
shade. The petioles were inhabited either by beetles or by ants
of the genera Azteca and Pseudomyrma. On the following day
Prof. Bailey found what we at first took to be a different Tachi-
galia on the left bank of the Cuyuni River. It was a small tree, 30
to 40 feet high, with denser foliage and very different leaves. Its
petioles were all inhabited by parts of a single huge colony of a
yellow Pseudomyrma. Later, on finding more material, we
became convinced that both trees were merely the juvenile and
adult forms of the same plant, the former growing in the shade,
the latter in the sun. Prof. Bailey found the seedlings in various
stages and was able to ascertain that the seed is a peculiar

40 Zoologica: N. Y. Zoological Society [III ;3

samara, not a pod as in most Leguminose, and that the plants
grow in loose colonies, the seedlings springing up in the shade
about the base of the parent tree, which rises to a height of
40 to 60 feet, with a crown of foliage at the summit of a very
slender trunk.

Further search revealed the fact that the Tachigalia is not
uncommon in many localities within a mile of the laboratory,
especially along the Puruni and Cuyuni Trails. It was also found
scattered through the beautiful primeval forest at Kalacoon and
Baracara on the right bank of the Mazaruni and in the jungle
behind the Penal Settlement on the opposite banks of the same
stream. Though the tree was often found in all the localities
mentioned, it is, nevertheless, rather sporadic compared with
many other components of the hylea. It may also be very local
in British Guiana and the adjacent countries. This is indicated
by the fact that there seemed to be no specimens of it in the
herbarium of the Botanical Garden at Georgetown nor in that
of the Botanical Garden at Port of Spain, Trinidad, and neither
the botanists of those institutions nor the chief of the forestry
department at Georgetown had ever seen the plant. Moreover,
the halfbreed Indian caretaker of the Kartabo laboratory, though
familiar with it and its ant-inhabitants, did not know its native
name, notwithstanding his remarkably accurate memory for the
aboriginal names of most other trees of the jungle. Thus we
were unable to ascertain even the generic name of the Tachigalia
till we had gone over the works on the South American flora in
the library of the Arnold Arboretum.

As Prof. Bailey will publish a detailed account of the
anatomical peculiarities of the tree, I may here confine myself
to a brief sketch of its appearance. For the purposes of this
study it will be advantageous to distinguish rather sharply
between the young, shade and large, sun forms, since their insect
inhabitants are, as a rule, very different. In the shade form
(Plate I) the Tachigalia has an extremely slender, straight
trunk, from about a foot to eight or twelve feet in height, with
long alternate, pinnate leaves coming off of it at right angles and
at such long intervals that even the smallest of the young plants

1921] Wheeler: Some Social Beetles 41

have only two or three, the largest hardly more than a dozen
leaves. These are dark green, smooth and somewhat shining, a
except at the base, where it forms a fusiform swelling two to
four inches in length and a third to half an inch in diameter. The
longer, more distal portion of the petiole bears six to eight pairs
of broadly lanceolate leaflets, which are not drooping or pendant.
The stipules at the base of the petiole are small and inconspicu-
ous. The sun form (Fig. 4) is much more vigorous and has
numerous branches at or near the summit of the long, slender,
foot or less in length. The petiole (Fig. 5a) is very slender
smooth, gray-barked trunk. The leaflets are pendant, more
crowded, much coarser, brighter green, with more rugose surfaces
and the petiole is thicker, with a large, three-cornered basal swell-
ing more gradually continued into the leaflet-bearing portion
(Fig. 5b). The cavity of the petiole also extends nearly through-
out its length instead of being confined to the basal, swollen
portion, and there are also cavities in the branches near each
leaf. The stipules are large and conspicuous and palmately
multifid. Forms of the plant intermediate between the shade
and sun forms were rarely seen. The specific identity of the two
forms was proved by finding young shoots of the typical shade
form growing from the trunk of a large sun tree in an abandoned
cassava patch at points where the wood had been cut back by
the natives.

THE TACHIGALIA BIOCOENOSE.

The interrelationships of the various organisms constituting
that portion of the Tachigalia biocoenose with which this paper
deals, 7. e., the numerous insects with their parasites and satel-
lites, that infest the young shoots and leaves, are represented in
the accompanying diagram (Plate II). It will be seen that the
portion of the plant which forms what may be called the center
of the biocoenose is the leaf-petiole, and it is, of course, the
peculiar structure of this organ that determines the specific
relations of the various insects to the host-plant and is, there-
fore, the key to an analysis and understanding of the whole living
complex. As Prof. Bailey will publish a detailed account of the

42 Zoologica: N. Y. Zoological Society [ll ;3

histological composition of the petiole, I here confine myself to a
general description, dwelling only on the more salient features.

The petioles of the shade-plants are very long and slender
and their enlargement is fusiform and restricted to the base,
proximal to the leaflets. In cross-section (Pl. IV) the enlarge-
ment is nearly circular, with one surface, corresponding to the
dorsal surface of the petiole, flattened. In the sun-plants the
basal thickening passes gradually into the leaflet-bearing portion
of the petiole and the cross-section (PI. III, figs. 3 and 4) is
distinctly triangular, one of the sides being dorsal, the two others
forming together the inverted roof-shaped ventral portion, so
that the petiole might be described as flattened dorsally and
carinate ventrally. The dorsal surface is distinctly winged, or
alate on each side. In the very young leaves the petiolar enlarge-
ment is solid, its interior being filled with juicy pith (PI. III,
fig. 3). This soon dries up, however, and is converted into a
flocculent or fibrillar, cinnamon brown substance, lying loosely
in a large cavity, the walls of which are lined by a thin layer
of the same reddish tissue. The same kind of tissue, the cells
of which are filled with a homogeneous amber-colored substance,
is also continually forming in four longitudinal strands or rays
in the dorsal wall of the petiole (Pl. IV, fig. 1). Later a few
similar strands make their appearance also in the ventrolateral
walls. Since the amber-colored substance which characterizes
this tissue evidently has a high nutritive value, I shall call it the
nutritive parenchyma. In a few petioles in this stage I found
from four to six small Curculionid larve feeding on the loose
material and reducing it to a red, powdery frass. Unfortu-
nately these larve could not be reared, so that their specific iden-
tity is unknown. They are not, however, necessary agents in the
preparation of the petioles for their future occupants.

As soon as the petiole has reached the stage just described
it is evident that any small insect, sufficiently enterprisng to bore
a hole in its wall, would find comfortable lodgings, which could
easily be rendered even more comfortable by tossing the dried
remnants of the pith out of the entrance or by compacting them
in the narrow ends of the spindle-shaped cavity. And if the
insect were a vegetable feeder it might find also an abundant
food-supply in the remains of the pith still forming the lining

FIG. 4. BRANCH OF AN ADULT (SUN) TACHIGALIA

The leaflets have been removed from two of the petioles.
Photograph by John Tee-Van.

FIG. 5. BASES OF LEAF-PETIOLE OF TACHIGALIA SP.

a, of young, shade tree; b, of large, sun tree, both nearly 44 natural size. Pieces of
the older petiole and adjacent trunk have been cut out to show the cavity
inhabited by Pseudomyrma.

1921] Wheeler: Some Social Beetles 45

of the cavity and especially in the rays of nutritive parenchyma.
That the very exuberant neotropical fauna comprises a consid-
erable number of such enterprising insects, the following obser-
vations will show.

It will conduce to clearness if we divide the insects which
take possession of the petiolar enlargements into two series
(Plate II). The first series comprises a small number of ants,
social beetles and coccids which utilize all the advantages afforded
by the petioles, 7. e., use them not only as dwellings for them-
selves and their young but also as sources of food. Among these
three groups of insects the coccids occupy a peculiar and impor-
tant position. They are present in considerable numbers and of
all sizes in all petioles inhabited by established colonies of the
beetles and ants. Specimens of the coccids from petioles of both
young and adult trees were submitted to Mr. Harold Morrison,
who pronounces them all, without exception, to belong to a single
species of mealy-bug, Pseudococcus bromelix Bouché (Fig. 9),
a well-known form which has been taken in widely scattered
tropical and subtropical localities (India, Zanzibar, Brazil,
Florida and even in the hot-houses of Massachusetts) and on a
variety of plants (mulberry, Canna, Hibiscus, pineapple). As
these insects are, of course, quite unable to make openings in the
walls of the petioles, they either wander into the cavities from
the surface of the plant, after the ants and beetles have entered,
or they are carried in by the former insects. Once inside the
petioles the coccids attach themselves to the longitudinal strips
of nutritive parenchyma in the walls, insert their beaks and find
an abundant supply of sap. The ants can then absorb the saccha-
rine excrement (“honey dew”’) of the coccids and thus vicariously
feed on the plant. The beetles, as we shall see, not only utilize
this same source of liquid nutriment but also feed on the solid
nutritive parenchyma of the petiolar walls.

The other series, which comprises many more forms, merely
use the petioles as nesting, dwelling or hiding places. They enter
petioles that have been previously occupied and abandoned, and
behave in all respects like the insects which live in old oak-galls
in our northern woods. They are in fact merely tenants, or
inquilines, or what German zoologists call ‘“Raumparasiten.”’ No
coccids are found in the petioles inhabited by these insects, which

46 Zoologica: N. Y. Zoological Society [IEE 3

may be divided into two groups, miscellaneous arthropods and
ants. The former are probably very numerous but I have notes
only on the following five:

1. In one abandoned petiole I found that a female katydid
had enlarged the opening with her jaws, had thrust her ovi-
positor through it and had deposited her flattened eggs in the
cavity.

2. On another occasion I found a petiole of a young plant
inhabited by termites (Nasutitermes sp.) They had built an
earthen gallery along the stem from the ground to an opening
in the petiole and were living in the cavity.

38. At the Penal Settlement, near Bartica, I found in one
old petiole the mature and pigmented pupa of a solitary wasp,
which Dr. J. Bequaert has identified as Podium ruficrus Fabr.
The mother wasp had stored the petiole with spiders, laid an egg
on them and sealed up the entrance.

4. Another petiole, found near the laboratory, contained
a small, red female centipede, identified by Dr. R. V. Chamber-
lin as Otostigmus limbatus Meinert, previously known from
Brazil and Paraguay. She was coiled about her small, white,
recently hatched young.

5. Some of the old and abandoned petioles on young trees
were occasionally seen to be inhabited by spiders, especially by
Attids.

Inasmuch as the ants comprise the more numerous and more
important insects associated with the Tachigalia they may be
considered in somewhat greater detail, and since their relations
to the plant are of five different kinds (Plate II) they may be
most conveniently grouped under as many captions.

1. Defoliators. Only one species, Atta cephalotes L., the com-
mon leaf-cutting ant of Central America and northern South
America, was actually proved to defoliate young Tachigalias at
a time when many of their petioles were not inhabited by other
ants or inhabited only by recently fecundated queens of Pseu-
domyrma that had not yet produced their colonies of belligerent,
stinging workers. In the jungle behind the Penal Settlement
I observed a few such trees which had been either completely

1921] Wheeler: Some Social Beetles AT

defoliated or had had large semicircular pieces bitten out of their
leaflets by Atta workers. I have no doubt that these ants would
carefully refrain from thus injuring larger Tachigalias in the
possession of well-developed Pseudomyrma colonies. In all prob-
ability the smaller leaf-cutters of the genus Acromyrmex, not
uncommon in the same locality, may also occasionally visit and
defoliate the young trees, but this was not actually observed.

2. Attendants of Homoptera. The terminal shoots of young
plants are often infested with a small brown Membracid in all
stages (Endoastus (?) productus Osborn) and peculiar flat Mem-
bracid nymphs (probably belonging to the genus Microcentrus,
according to Osborn) the piercing mouthparts of which leave on
the surfaces of the petioles permanent scars that may perhaps
serve later as convenient points for the beetles and queen ants
to bore into the enlargements. At least four species of ants
were taken in attendance on these Membracids: Camponotus
femoratus Fabr., Crematogaster limata parabiotica Forel, Ecta-
tomma tuberculatum Oliv. and Dolichoderus attelaboides Fabr.
The first two were the most frequently encountered and belong
to another interesting biocoenose, that of the ‘“‘ant-gardens,”’
which I have described in a recent paper (1921), the last two
were of more sporadic occurrence. A few notes on their habits
are recorded in Zoologica, Vol. III, No. 4.

3. Inquilines. These comprise no less than sixteen different
forms, representing thirteen species, belonging to the genera
Neoponera, Leptothorax, Crematogaster, Allomerus, Solenopsis,
Pheidole, Camponotus and Brachymyrmex. Most of them are
small and all are timid and nonaggressive species which never
keep coccids, and are more frequently found in other situations,
especially in the dead twigs and branches of various shrubs and
trees. And most or all of them merely take possession of peti-
oles that have been previously perforated, inhabited and aban-
doned by other ants or by the social beetles. The inquiline ants,
moreover, are confined to the young Tachigalias growing in the
shade. Although the number of them I have collected is con-
siderable and although they represent very diverse genera and
even subfamilies, it is certain that further search, especially in
other localities, will greatly increase the number, for the petiolar
cavities of Tachigalia are sufficiently commodious to accomodate

48 Zoologica: N. Y. Zoological Society [II1;3

at least the young colonies of nearly all of the twig-inhabiting
ants of British Guiana and the list of these is a long one, com-
prising many small species of Cryptocerus, Procryptocerus, Pseu-
domyrma, Leptothorax, Crematogaster, Azteca, Tapinoma,
Myrmelachista, Camponotus, ete.

4. Thief-ants. The only species belonging to this category
is the tiny Solenopsis altinodis Forel, (Fig. 15). which is closely
related to a series of “lestobiotic” Solenopsis species that nest in
the walls of ant-or termite-nests and prey on their brood. It
does not live in the Tachigalia petioles but enters those inhabited
by the social beetles, when their entrances happen to be un-
guarded, and destroys their larve. In all probability it also
attacks small, defenceless colonies of the inquiline ants and
devours their brood. I infer that it nests in the ground from
the fact that it often appeared suddenly in considerable numbers
during the night and exterminated the colonies of the Pseudo-
myrmas in vigorous Tachigalia and Triplavis branches that had
been left on a table in the yard of the laboratory. I have also
seen it wandering about on the laboratory tables indoors, over
the foliage of the undergrowth in the dark jungle or feeding
on the pulp of injured fruits of the water cocoa (Pachira aqua-
tica Aubl.) growing along the river banks.

5. Obligatory Ants. I would thus designate the ants that
are definitively attached to the Tachigalia as their host-tree.
There are only four species: Pseudomyrma damnosa sp. nov.,
(Fig. 13), Ps. maligna sp. nov. (Fig. 14), with its two varieties,
cholerica and crucians, Azteca foveiceps sp. nov. (Fig. 16), and
A. traili Emery. The last is doubtfully included for reasons to
be given below. The recently fertilized queens of these ants
perforate and enter the petioles of young Tachigalias growing in
the shade, close the openings behind them with particles gnawed
from the walls and eventually produce their broods in the cavi-
ties. Occasionally queens of the two species of Pseuwdomyrma
or of these and A. foveiceps may be found, each in a petiole of
the same plant and even of one with only a few leaves. That
the founding of the colonies may be frought with many dangers
is shown by the fact that dead queens are often found shut up
in the petioles precisely as in the case of the Azteca queens in
the internodes of young Cecropias. Some of the unfortunate in-

1921] Wheeler: Some Social Beetles 49

sects undoubtedly succumb to hunger or the attacks of fungi,
but in many cases they seem to be killed by alien ants or by the
beetles boring into the same petioles. If the queens survive,
however, and produce their broods of workers, the latter open
from the inside the entrances made by the queens and eventually
take possession of the whole plant. Since all the petioles of
larger trees are invariably inhabited by a single flourishing
colony of a single species or variety, we must suppose that the
offspring of different queens on opening their respective petioles
either fight for the possession of the plant or, if they belong
to the same species or variety, unite to form a single polycladic
colony. The fact that the petioles of large trees contain several
dedlated fertile females of Ps. damnosa or maligna would seem
to indicate that the eventual climax colony, as it may be called,
is established by alliance of several broods rather than by the
survival of the offspring of a single queen. Furthermore, the
climax colony seems to be far too populous to represent the off-
spring of a single mother.

No sooner are the petioles opened by the young broods of
workers than the coccids enter or are carried in by the ants and
attach themselves to the areas of nutritive parenchyma which
furnish optimum conditions for feeding and growth. Here they
can multiply and be cared for by the ants which undoubtedly
find in them a most welcome source of food. As the tree grows
and puts forth new leaves, each young petiole as soon as it has
reached the proper size is perforated by a detachment of ants and
coccids. The later leaves, as I have stated, have cavities extend-
ing nearly the entire length of the petioles and adjacent to their
insertions there are also cavities in the branches, which are like-
wise entered and occupied by the insects. The largest leaves have
petioles with such broad cavities at their bases that the ants
find it necessary to make chambers in them by building carton
partitions. The materials for the carton are gnawed from the
walls. Azteca foveiceps builds a much more elaborate system of
partitions than the Pseudomyrmas. This is not surprising, be-
cause the Aztecas are nearly all wonderful experts in carton
construction.

When the climax stage of the Tachigalia biocoenose has been
reached, it has been reduced to only three organisms, the plant,

50 Zoologica: N. Y. Zoological Society {III ;3

the coccids and Azteca foveiceps or more frequently one of the
two species of Pseudomyrma or one of the varieties of Ps.
maligna. Merely rapping on the tree then reveals the situation,
for the angry workers rush out of the petioles, cover the leaflets,
trunk and branches and can be readily identified by their size and
color, Ps. damnosa being yellow, A. foveiceps and Ps. maligna var.
crucians black; maligna s. str. and its var. cholerica both black
with yellow markings, but differing in size. Of A. traili only a
single very young colony was seen, and this was inhabiting a
petiole of a small Tachigalia. There were coccids in the cavity
of the petiole, however, and both for this reason and because
the ant is known to be associated with other myrmecophytes, I
have included it among the obligates in the diagram. It is, in
fact, conceivable that in some localities in British Guiana traili
may be the dominant species and replace foveiceps in the climax
stage of the biocoenose.

I have noticed unmistakable indications of two other insects
on young Tachigalias, both attacking the foliage. One was a
caterpillar of some kind, which had gnawed the leaflets of a few
plants and had disappeared, so that I was unable to secure a
specimen for identification. The other was an Itonidid gall which
occurred in great numbers on a single plant on both the upper
and lower surfaces of the leaflets. This gall was 2-4 mm. long,
erect and cylindrical, with a conical tip, and somewhat resembled
the galls produced by Cecidomyia caryecola Osten Sacken on the
leaves of our northern hickories. The smaller galls were unde-
veloped and contained no larve but each of the larger ones had
a round hole at the tip and contained in the base of its tubular
cavity a cylindrical cocoon enclosing the pupa of a small parasitic
Hymenopteron, evidently a Eulophid. Some of the specimens
were fully pigmented and nearly ready to hatch, others had been
killed by an Hntomophthora-like fungus. Probably the mother
of the parasite had made the round hole in the tip of the gail
in order to enter and oviposit on its maker.

It will thus be seen that the Tachigalia biocoenose comprises
at least 51 different organisms, the host plant and 50 organisms
associated primarily with its leaves and terminal shoots or secon-
darily with the organisms thus associated.

1921] Wheeler: Some Social Beetles 51

THE SOCIAL BEETLES.
A. COCCIDOTROPHUS SOCIALIS SCHWARZ AND BARBER.

After a careful taxonomic study of the beetles which I found
in the Tachigalia petioles, Messrs. Schwarz and Barber conclude
that they belong not only to perfectly distinct species but to
different genera of the family Silvanide. While still at Kartabo
I convinced myself of their specific difference, but unfortunately
not till it was too late to make a detailed study of the differences
in their behavior. While the species described in Zoologica III,
No. 5 as Coccidotrophus socialis was so abundant that I obtained
nearly four hundred of its colonies, Hunausibius wheeleri was so
scarce that I encountered it in less than a dozen petioles. For some
time I took it to be merely a depauperate variety or aberration
of the Coccidotrophus and therefore gave it little attention.
Finally, when I had come to appreciate its distinct specific status,
my stay at the laboratory was drawing to a close and I could
secure only a few of its colonies. My account, therefore, relates
almost entirely to Coccidotrophus and my notes on Hunausibius
yield only a few rather summary remarks which may be most
conveniently relegated to a brief separate caption (p. 88).

The Coccidotrophus (Plate III, fig. 2; Plate VI, figs. 1-5)
was taken only near Kartabo and only in the petioles of young
Tachigalias growing in the shade along the Puruni, Hacka and
Cuyuni trails, within a mile of the Tropical Laboratory. Although
the tree was not uncommon in the jungle behind H. M. Penal
Settlement and along the trails in the primeval forest near
Kalacoon and Baracara, I found no traces of the beetle in these
localities. It would seem, therefore, to be rather local. Its body
is flattened-cylindrical, long, slender and parallel-sided, with
short, stout legs, the femora, especially the fore pair, being con-
spicuously thickened. It varies in size from 3.5 to 4.5 mm. and
when fully mature is deep castaneous brown to almost black, with
more reddish appendages. Specimens recently emerged from the
pupa, however, are pale reddish testaceous, with more yellowish
elytra. The surface of the body is shining and glabrous, the
head and thorax punctate, the elytra punctate-striate. The head
bears a singular and significant resemblance to that of certain
small ants of the genus Cryptocerus. The antenne are short,

52 Zoologica: N. Y. Zoological Society pains 33

compactly jointed and distinctly clubbed at the tip. The female |
is, perhaps, somewhat smaller than the male, but I have been
unable to detect any other external differences between the sexes,
and Schwarz and Barber in their more careful study of the struc-
tural peculiarities of the species, have been no more successful.
For further details the reader is referred to their taxonomic
description in Zoologica III, No. 6.

The eggs (Fig. 11) are pure white, regularly elliptical, 0.55
long and 0.25 mm. broad, with rather thick, leathery, smooth
and shining chorion. The larva is described and figured in great
detail in Dr. Boving’s article (Zoologica III, No. 7, Plates VII
and VIII). In life it is whitish and beautifully translucent so that
the straight alimentary tract and its contents can be distinctly
seen, but as maturity approaches the body becomes more opaque
and milk-white, owing to a considerable increase in the fat-body.
The pupa is even more opaque, but just before eclosion becomes
yellowish or testaceous, with pigmented eyes. Both beetle and
larva are very active and alert, the latter at all times, the former
when hungry or disturbed. At other times the beetle is rather
sluggish or quiescent, but shows no tendency to “feign death”
when handled. Though its wings are well-developed (Plate VI,
fig. 5) I have never seen it attempt to fly. The pupa (Plate
IX, figs. 19-21) is also very active when stimulated, wriggling
its abdomen from side to side, but is unable to move about.

The beetles enter petioles which have either not been pre-
viously perforated by other insects or those which have been
occupied for some time and have then been abandoned by ants
or other beetles or have been occupied by queens of Pseudomyrma
or Azteca that have died before they could produce their broods
of workers. The beetles either enter as a single pair or one
beetle enters and is very soon joined by an individual of the
opposite sex. I have been unable to decide which of these
methods is followed or whether the opening in the petiole is
made by the male or the female or by either indifferently. Cer-
tainly the great majority of colonies in their first, or incipient
stage consist only of a male and a female beetle. On the rare
occasions when only a single beetle was found in a petiole, the
other may have escaped or eluded my attention while I was
cutting into the cavity. Twice I have actually seen both beetles

FIG. 6. SIX SUCCESSIVE STAGES IN THE CONSTRUCTION OF THE COCOON
BY THE FULL-GROWN LARVA OF COCCIDOTROPHUS SOCIALIS.

See text for explanation.

1921] Wheeler: Some Social Beetles 55

in a cavity diligently cleaning it up for occupancy. Such house-
cleaning is necessary both in previously unoccupied and in pre-
viously occupied petioles since in the former the particles of fibril-
lar or powdery pith, which partially fill the cavity, and in the
latter the dead bodies of ants, beetles, ete. must be removed. The
entrance opening gnawed in the wall is transversely elliptical
and just large enough to admit the slender bodies of the beetles.
It is most frequently made in the lateral wall some distance
from the narrow ends of the petiolar swelling. The petiole
occasionally has several openings, sometimes as many as five to
seven. I believe that each of these must be the work of the
founders of one of the colonies which have successively occupied
the same cavity. In other words, the pair of beetles seems not to
utilize the openings of previous occupants for the purpose of
entering the petiole but insists on making an opening of its
own, so that the considerable number of orifices occasionally
noticed in an old petiole represent the number of colonies of
beetles or potential colonies of ants that have from time to time
taken up their abode in its cavity.

The beetles accomplish the removal of the loose pith or the
remains of previous tenants by pushing this refuse into the
pointed ends of the cavity with their flattened heads, much as a
slovenly servant might tidy a room by sweeping things under
the furniture or into closets. Smaller particles of pith are
sometimes thrown out of the entrance but the decomposed and
more or less disarticulated:bodies of queen ants and beetles are
too voluminous to be disposed of in this manner so that they can
only be packed away compactly into the ends of the cavity. This
behavior brings the insects into contact with the outermost layer
of pith still adhering to the ligneous walls of the cavity and the
strips of nutritive parenchyma laden with amber-colored sub-
stance. (Plate IV, fig. 1, Plate III, fig. 5). That this tissue
actually constitutes the food of the beetles is proved not only
by finding it in their intestines but also by actually observing
their feeding activities. Very soon, however, young coccids
begin to enter the petiole through the opening made by the
beetles and take up their positions on the walls of the cavity and
preferably along these very strips of nutritive tissue which, as
the beetles feed, become gradually deepened into grooves. The

56 Zoologica: N. Y. Zoological Society [III ;3

coccids station themselves in a row in each groove, with the long
axes of their bodies parallel with the long axis of the petiole.
Since both beetles and coccids center their feeding activities
on the tissue forming the floor of these grooves it is important
that they shall be kept clean so that the parenchyma, which is
continually proliferating, can be easily reached by both species
of insects. (Plate III, fig. 5). Hence the beetles carefully
deposit their feces, or frass, which has a chocolate brown color,
on the areas between the grooves. As time goes on the accumu-
lations of frass acquire the form of more or less longitudinal
ridges projecting into the petiolar cavity (Plate IV, figs. 2-3 and
Plate V). In many petioles these ridges are strikingly regular,
in others more vermiculate, interrupted or anastomosing. In
old petioles inhabited by old colonies of Coccidotrophus the inter-
ior of the petiolar cavity presents the appearance of the figures
on Plate VY. In addition to these frass ridges the beetles also
build a more or less circular wall of the same substance around
the entrance, so that the latter is converted into a short tube, in
which one of the beetles often stations itself on guard for
hours at a time, with the long axis of its body at right angles
to the long axis of the petiole and its flattened head exactly
filling the elliptical orifice. (Fig. 11). From time to time the
beetle may project its antenneze out into the air and wave them
about, so .that the petiole from the outside suggests the nest
of one of the smaller species of Cryptocerus, with a soldier or
worker ant on guard at the orifice.

The female beetle begins to lay her eggs either before or
after the entrance of the coccids. They are deposited singly
along the edges of the frass ridges and evidently at intervals of
several hours or even days, for dissection of the beetles shows
that only a few eggs mature at a time in the ovaries. They
are glued to the wall of the petiole rather firmly and always
with their long axis parallel with its long axis and that of the
food-grooves and frass ridges (Fig. 11). As I have not wit-
nessed oviposition I have been unable to determine the length
of the embryonic period. The eggs hateh, of course, at intervals,
so that the larve vary greatly in size, some being very small
and evidently just hatched, others a third or half-grown or
actually full-grown and ready to pupate. Like the beetles, the

1921] Wheeler: Some Social Beetles 57

FIG. 7. ORYZAEPHILUS SURINAMENSIS lL.

Beetle, pupa and larva, showing the teeth on the sides of the pronotum
of the beetle and tubercles of the pupa in which they are formed.
Courtesy of the Federal Bureau of Entomology.

larve feed on the nutritive parenchyma. Its amber colored
cells in the intestines may be distinctly seen through the clear
integument and body cavity of the larve. The colony, which is
now in its second stage consists of the pair of parent beetles,
about one or two dozen larve, mostly immature and in most
cases of about the same number of young or half-grown coccids.

When mature the larve make brown cocoons and pupate
in them, as will be described in detail below. These are formed
singly and the beetles emerging from them remain in the petiole
with their parents and larval brothers and sisters, mate and
produce eggs and larve in turn, thus leading to the third or
climax stage of the colony, which may eventually consist of
several dozen beetles of both sexes and numerous larve and
pup in all stages of development. The coccids also increase
in number, so that the cavity of the petiole sometimes becomes
so crowded that its inmates must find their movements greatly
impeded. In the meantime the old and exhausted beetles gradu-
ally die off and their bodies are consigned to the refuse accumu-

58 Zoologica: N. Y. Zoological Society {III;3

FIG. 8.

a, pronotum of pupa of Eunausibius wheeleri showing lateral tubercles; b, to f, pronota
of five pupz of Coccidotrophus socialis, showing vestigial lateral tubercles and
their variation.

lations, or kitchen-middens in the pointed ends of the cavity.
When this crowded condition is reached, beetles begin to leave
the colony either singly or in pairs, seek and enter other petioles
of the same or other Tachigalias and thus establish new colonies.

In order to study the stages of colony development just
described as well as those that supervene, I found it necessary to
split each petiole longitudinally and to place its two halves, with
their cavities turned outward in a slender vial or test-tube plug-
ged with cotton. The cotton could be pushed in till it held the
two halves firmly against the inner surface of the glass. Through
the latter the behavior of the beetles could then be studied with
the pocket-lens (magnifying 10-20 diameters) or the binocular
dissecting microscope. Splitting the petiole, of course, so greatly
disturbs the insects that many of them at once escape into the
tube. Moreover, a certain number are killed or injured by the

1921] Wheeler: Some Social Beetles 59

knife-blade. But all the uninjured soon return to the two half-
cavities and remain in them. At first I carefully kept the tubes
in the dark, but I soon found that they could be left in the
diffuse day-light on my table without disturbing the activities
of the insects. It was necessary, however, to keep them in a
horizontal position, like that of the petiole on the living, plant,
for when they were placed upright, gravity seriously interfered
with the activities of the beetles and especially of the larve,
causing them to drop and accumulate at the lower ends of the
cavities or of the tubes. Of course, this position did not inter-
fere with the coccids which remained attached by their sucking
mouthparts to the nutritive parenchyma. Colonies of Coccido-
trophus can be kept in tubes and under close observation for
a week to ten days but by the fifth or sixth day the petioles are
apt to become so dry even during the rainy season that the
beetles, larvee and coccids become demoralized. The modifica-
tions of behavior thus induced will be considered in the sequel.

There is, perhaps, nothing very remarkable in the fact that
both the beetles and their larve feed on the nutritive paren-
chyma of the Tachigalia, since other Silvanide, e.g., certain spe-
cies of Oryzephilus, Silvanus, Cathartus and Nausibius are
known to be vegetarian, but the fact that both the imaginal and
larval Coccidotrophus actually solicit and imbibe the saccharine
excrement (“honey dew’’) of the coccids, is so unusual and start-
ling that it will be advisable to give a more detailed account of
these insects and of their treatment by the beetles.

The adult female Pseudococcus bromelix (Fig. 9) measures
nearly 3 mm. in length and is broadly and regularly elliptical,
evenly convex dorsally, flattened ventrally and of a pinkish flesh-
color or pale dull red, but the body is so completely covered
with snow-white wax as to be scarcely visible in healthy speci-
mens. The wax is secreted in a thin, even, mealy layer over
the dorsal surface but around the periphery of the body as a
regular fringe of stiff, blunt pencils which are longest on the
posterior segments, somewhat shorter on the anterior border
of the head and still shorter along the sides. Large specimens
of the insect are less numerous in the petioles than the smaller
or recently hatched individuals, many of which scarcely exceed
-5 mm. in length. These are reddish because they have not yet

60 Zoologica: N. Y. Zoological Society {IlI;3

FIG. 9. PSEUDOCOCCUS BROMELIAE BOUCHE.

Sketch of an adult living female with intact covering and
peripheral pencils of wax.

secreted an appreciable quantity of wax from their dorsal in-
tegument, and lack the peripheral fringe of snow-white pen-
cils. I have not seen the males. Young individuals are rather
active, not infrequently moving about in the petiole, but the
older ones remain stationary in the food grooves till the petivle
begins to dry up, when they take to wandering about aimlessly
in search of more succulent pastures.

In specimens of the Psewdococcus treated with caustic potash
and stained according to the method recommended by Mac-
Gillivray (1921, Chapter II), the details of the integument are
clearly visible (Fig. 10). Only a few of the structures are of
interest in connection with the following behavioristic account,
such as the anus, which is on the ventral surface of the flat-

1921] Wheeler: Some Social Beetles 61

tened posterior border of the body, and, on the dorsal surface,
two pairs of peculiar organs which have the form of transverse,
mouth-like slits with thick lips. One pair of these organs is
situated near the posterior corners of the head, the other between
the sixth and seventh abdominal segments. Coccidologists have
long been familiar with these organs in certain genera of mealy-
bugs of the subfamily Eriococcine, and have called them “‘eye-
like glands” “cicatrices,” “osteoliform or labiate fove,’’ or
“dorsal ostioles.” Sule called them ‘“adipopugnatorische Organe”’
and MacGillivray has recently dubbed the two pairs ‘“‘cephalabiz”
and “caudalabiz” respectively, terms so barbarous that they
make one’s flesh creep. I shall call them anterior and posterior
ostioles. Berlese regarded them as apodemes, or invaginations
of the integument for the insertion of muscles. Comstock, New-
stead, Sule and MacGillivray regard them as glands. In 1882
Comstock stated that he had ‘‘observed in Dactylopius a pair of
openings on the dorsal side of the sixth abdominal segment,
which are evidently homologous with the honey tubes of Aphi-
dide. A female mealy-bug was gently rubbed near the caudal
end of the body, when suddenly there appeared two drops of a
clear fluid, resembling in appearance the honey-dew of plant-
lice. This experiment was repeated many times and with many
specimens. Mr. Pergande assures me that he has-.observed a
similar excretion from a pair of openings on the cephalic mar-
gin of the first thoracic segment.” Comstock was, of course,
under the erroneous impression that the honey-dew of aphids
is a secretion of the cornicles instead of being the excrement of
the insects and therefore extruded from the anus. According
to MacGillivray:-“There can frequently be observed on living
specimens a small globule of a clear fluid over the mouth of each
labia, more frequently the caudalabize than the cephalabiz, so
that they are probably also glandular in structure as suggested
by Comstock. For, as he suggested, when the specimens are
stroked with a pencil or dissecting needle, the insect will hump
up its back and extrude a globule of liquid. The insect is unable
to repeat this operation until the pocket is again filled with the
clear fluid. Specimens have been observed to extrude globules
from all four labize at the same time. The labize undoubtedly
have a glandular function which is probably of later origin than
their earlier function, a parademe for the attachment of muscles.”’

62 Zoologica: N. Y. Zoological Society [IlI;3

FIG. 10. PSEUDOCOCCUS BROMELIAE BOUCHE.

Adult female treated with caustic potash to show the open-
ings of the anterior and posterior ostioles, anal orifice, etc.
The specimen is mounted dorsal surface uppermost.

Sule (1909) made a histological and physiological study of
the anterior and posterior ostioles in Pseudococcus farinosus De
Geer. When the female of this insect was stroked with a brush,
each of the ostioles suddenly emitted a droplet of orange-yellow
liquid, which partly adhered to the brush and partly rolled off
from the wax-powdered dorsal surface. The liquid was found to
consist of cells and a few blood-corpuscles. Sule concludes that
the organs are repugnatorial, and that their secretion is employed
like that of the cornicles of aphids for gluing up the appendages
of insect enemies. His account of the function of the secretion ©
is by no means convincing, since it might also be regarded as an
exudate, derived directly from the fat-body, like the exudates
produced by various termitophiles and myrmecophiles (Cf.
Wheeler 1918), and hence employed for allurement instead of
repulsion.

Returning to a consideration of the Coccidotrophus in the
Tachigalia petioles, we find that the beetle often remains motion-
less for hours at a time, in a food-groove, which just fits its long
slender body. If at other times, when it is moving about, it

1921] Wheeler: Some Social Beetles 63

chances upon a coccid, it stops suddenly and seems at once
to become more alert or excited, for as soon as its clubbed anten-
nz touch the dorsal surface of the insect, their beat, hitherto
leisurely and exploratory, becomes greatly accelerated. With
each beat, each antenna rapidly describes a minute transverse
ellipse on the surface of the coccid, and the beats of the two
appendages seem not to be quite synchronous. At the same
time the beetle, with a much slower rhythm, rocks its body for-
ward and backward by bending its legs, while the mobile arti-
culations between the head and prothorax and between the pro-
thorax and mesothorax enable it to cover more of the coccid’s
dorsum and to keep the antennal clubs in contact with its
rounded surface. While engaged in this perfomance the beetle
resembles an expert pianist moving his hands from side to side
over the key-board, or a masseur with his hands in soft gloves,
massaging a patient. The beetle undoubtedly distinguishes a
large coccid’s posterior from its anterior end, since it lavishes
most attention on the former. It seems, however, to be quite
as interested in the medium-sized or smallest coccids and will
spend just as much time in stroking them. The time devoted
to the performance in any particular case seems to vary directly
with the beetle’s appetite or thirst. A beetle may thus spend
ten, twenty or even forty or more minutes massaging a single
coccid, with occasional short pauses. After a coccid in the
proper condition has been stroked in this manner for a few
moments it may slowly turn up its wax-penciled posterior seg-
ments and discharge from the anal orifice a perfectly limpid
droplet of sweet excrement, which the beetle at once greedily
swallows. The coccid then flattens down its posterior segments
and the beetle resumes its massage. The coccid may thus con-
tribute a droplet every few minutes or it may remain inert and
unresponsive. An ant confronted with such a situation would
take the hint and at once look up another coccid, but the beetle
stubbornly keeps on and may work for an hour or more with-
out receiving another drink. Usually, however, some of the
larve or one of the other beetles of the colony intervene and the
scene may change, as described in a later paragraph.

That the antenne of the beetle are beautifully adapted for
stroking the coccids is apparent at a glance (Plate III, fig. 2,

64 Zoologica: N. Y. Zoological Society [II1;3

Plate VI, fig. 1). Their compact structure and clubbed extremi-
ties recall the antennze of many myrmecophilous beetles or of
certain ants, for many of these insects, of course, use their
antenne in soliciting liquid food from one another. The basal
joint of the Coccidotrophus antenna is even elongated to form a
crude scape, although the remaining joints do not form an angle
with it as in ants. The relations of the beetle to the coccids,
moreover, are physiologically similar to those of symphilic beetles
to the host ants that feed them with regurgitated liquids, and
the Coccidotrophus like the symphiles has a short, broad tongue
and short labial palpi. Wasmann (1896) and Escherich (1902)
have dealt with these antennal and labial adaptations in detail,
pointing out that the tongue in symphilic beetles becomes short,
broad and spoon-like and that the palpi, especially those of the
labium, become shorter and have a reduced number of joints.
Precisely this condition is seen in the labium of Coccidotrophus
as shown in the figure of Schwarz and Barber (PI. VI, fig. 1).
The greater development of the maxillary palpi indicates that
they may occasionally function like the antenne in soliciting
honey-dew.

Coccidotrophus larve of all stages, from those just hatched
and less than a millimeter in length, to those almost four milli-
meters long and nearly ready to pupate, likewise solicit and
obtain food from the coccids by stroking them with the antenne.
The small beetle larve show no preference for small coccids
since just hatched larve are often seen on the backs of adult
female coccids, feverishly stroking their waxen surfaces and
full grown larve may often devote themselves to coccids smaller
than their heads. The movements of the larva’s antenne, though
similar to the antennal strokes of the beetles, cover a smaller
portion of the coccid but the larve reinforce the titillation by
a simultaneous use of the maxille and maxillary palpi. The
larva is almost or quite as persistent as the beetle and drinks
up the periodic globules of honey-dew with quite as much gusto.
Both beetles and larve, however, stroke the dorsal surfaces of
the coccids so gently that their waxen bloom is neither removed
nor diminished even after the most prolonged solicitations.

In connection with the behavior of the larva, Dr. Boving’s
figures of its mouthparts and antenne are very interesting and

i — 2 eee , a aaa. ea =e
I 2t eis 7s = Foe ee pan ss —_ f - .
x 6 i ‘i, ay J a ° ve es . : : 7,
4 ay i = : cm : . -) t
. al eal = 1
j- : 7
- mY oS i
1 if
5 Oo)

FIG. 11. ENLARGED DRAWING OF A PART OF THE WALL OF A TACHI-
GALIA PETIOLE INHABITED BY COCCIDOTROPHUS SOCIALIS.
Showing the food grooves and frass ridges, the entrance with its wall, the eggs,

an intact and broken cocoon of the Coccidotrophus and two cocoons of Blepyrus
tachigaliae, one of them after the eclosion of the parasite.

1921] Wheeler: Some Social Beetles 67

suggestive. It will be noticed that his Figs. 1, 2, 5, 8, 16, Plates
VII and VIII etc. show that the tongue is large and flat like a
spatula or ladle and well-adapted for receiving the globules of
coccid excrement. The labial palpi are very small and 2-jointed,
but the maxilla is large, with extensive stipital articulation and
a large lacinia tipped with a claw-like tooth and fringed with
stiff hairs along the medial border. The maxillary palpi are
long and 3-jointed and the large articular membranes of the
separate joints suggest great mobility. The antenne are un-
usually interesting (Plate VII, figs. 1, 2, 5, 7). Though they
consist of only three joints, the second is greatly elongated and
distinctly drum-stick-shaped, the apical joint being much reduced
to form merely a sensory cap for the second joint. Now this
drum-stick-type of antenna is precisely the one found in a long
series of symphilic ant-guests of the Coleopterous family Clavi-
geride (Fustiger, Rhynchoclaviger, Adranes, etc.), which use
their antennz for soliciting liquid food from their hosts. Dr.
Boving’s figures of other larval Silvanids and of genera belong-
ing to closely allied genera of the Cucujid complex, namely
Cathartus (Plate VIII, fig. 12), Nauwsibius (Plate VIII, fig. 18),
Dryocera (Plate IX, fig. 33), Telephanus (Plate X, figs. 34, 37)
and Scalidia (Plate X, fig. 39), show very different conditions.
Thus we may say that the antenne and maxille of the Coccido-
trophus larva are specially adapted to their active role of solicit-
ing and the labium to its passive, spoon-like role of receiving the
liquid excreta of the Pseudococcus.

The question naturally arises as to the function of the
anterior and posterior ostioles, which, as I have shown, are
highly developed in Ps. bromelix, and the probability of their
secreting substances that may be ingested by the beetles. Un-
fortunately I did not know of the existence of the ostioles while I
was at Kartabo. My attention was called to them by Prof. Mac
Gillivray several months after my return. I feel very confident,
nevertheless, that these organs in Ps. bromelix cannot have the
function ascribed to them by Sule in Ps. farinosus. I have so care-
fully watched the coccids of all ages that I could not have over-
looked the emission of orange-yellow droplets from the ostioles
or of any sticky liquid that would glue up the appendages of
small insects. The smallest beetle larve are so delicate that

68 Zoologica: N. Y. Zoological Society [III ;3

they are at once immobilized or impeded in their movements
when they happen to run or fall into a minute drop of water, and
if the Ps. bromelix were at all hostile to the beetles or their
larve in the manner described by Sulc, their presence would
not only be a nuisance but a serious menace to the colonies. The
further fact that the coccids in a petiole are frequently decimated
or even exterminated by several small predatory insects (vide
infra p. 78) is also unfavorable to Sule’s contention.

I am, of course, willing to admit that the ostioles may be
glands or exudate organs which secrete substances that may be
ingested by the beetles and their larve, but the closest observa-
tion of which I was capable showed that the oniy liquid visibly
imbibed by the Coccidotrophus was the saccharine excrement, or
honey-dew. If the secretion of the ostioles is a liquid, it must
be emitted in droplets too minute to be visible under a Zeiss lens
magnifying 20 diameters, or it must be a volatile substance like
that secreted by the peculiar tubular organs which occur on
the eighth abdominal segment of many ant-attended Lycaenid
caterpillars and have been described by de Niceville (1890),
Thomann (1901) and others.t That the ostioles of Ps. bromelix
may actually emit some substance attractive to the beetles and
their larve is indicated by their often very prolonged stroking
of the dorsal, lateral or anterior portion of a coccid. That they
prefer to stroke its terminal abdominal segments may be due
to the fact that that region bears both the anus and the poster-
ior pair of ostioles.

The attraction of the mealy-bugs, whether due solely to their
ability to excrete honey-dew or because they can also secrete some
delicious exudate or fascinating aroma, is so great, that in popu-
lous colonies a single coccid may often become the center of a
circle of actively competing beetles and larve of various sizes.
This is not so apparent in colonies that have just been collected
and placed in glass tubes, because the petiole still contains a cer-
tain amount of sap and the coccids are able to excrete normally,
but after the colonies have been kept for several days or a

1 Conf. also the paper of Newcomer (1912), who figures a section of one
of these organs in Lycaena piasus (Pl. 2, Fig. 3). Although he believes
that the tubular organs are not glandular, the structural details certainly
seem to support Thomann’s contention.

1921] Wheeler: Some Social Beetles 69

week and the supply of water in the plant tissues has consider-
ably diminished, the excretions of the coccids are less frequent
and copious, so that the beetles and larve become more and
more thirsty and therefore more desperate and exacting. Thus
by the very simple device of keeping the colonies for some time
in tubes, it is possible greatly to exaggerate the attentions of
the beetles and their larve to the coccids and to witness certain
peculiarities of behavior which are less obvious in recently col-
lected colonies.

When a number of thirsty beetles and larve surround ea
large coccid, all stroking different parts of its dorsal surface
and periphery, only the individual that happens to be stationed
at its posterior end is able to secure any honey-dew. The beetles
and larve all keep at work, however, till the antenne of two of
them happen to meet. Then the larger individual stops stroking
for a moment and butts its competitor with the side of its head. If
the group is formed by a single beetle and several larve, the beetle
being the stronger, soon pushes any larva with which it may come
in contact away, but the latter usually at once returns and re-
sumes its stroking till contact with the beetle again occurs and
the butting is repeated. When several larve of different sizes
have preempted a coccid, the largest treats the others in the same
manner. This behavior is so suggestive of that of a number of
pigs eating out of the same trough that one can hardly doubt that
something more than a mere reflex is involved in the butting. The
larger beetle’s or larva’s indefatigable perseverance in butting
is only equalled by the pertinacity with which the butted indivi-
dual keeps returning and resuming its stroking movements. I
can illustrate this best by transcribing a few observations from
my note-book.

August 10, a beetle was seen standing over the posterior end
of a large coccid and stroking it busily. At short intervals the
coccid raised its anal segments and discharged a minute limpid
globule of honey-dew, which was at once avidly seized and
swallowed by the beetle. During 11 minutes the coccid raised
its tail 25 times and the globule could be distinctly seen on
most occasions, as the beetle paused suddenly in its manipula-
tions and moved its labium and palpi each time a globule was
imbibed. Sometimes the beetle would pause for a’ moment,

70 Zoologica: N. Y. Zoological Society [IlI;3

before proceeding with its titillation and remain with its head
flattened down, even when the coccid failed to move its anal seg-
ments. Once a large larva came up and endeavored to get some
of the excretion but was promptly butted away, and once another
beetle was treated in the same manner. After the 25 feedings
the beetle moved away and another beetle came up and received
from the same coccid four globules in less than two minutes.

August 12, I observed a beetle (No. 1), which was red and
therefore immature, soliciting from a nearly full-grown coccid
in a petiole collected a few days previously and already beginning
to become dry. The beetle stroked the coccid for 15 minutes,
during which time the latter produced only five droplets of
honey-dew at intervals of two to five minutes. Then a mature,
dark colored beetle (No. 2) came up and began to stroke the
anterior end of the coccid, gradually moving back over it. When-
ever the beetles met they butted each other with their heads or
even locked mandibles for an instant and then returned to their
former position and occupation. Beetle No. 1 worked for an-
other 15 minutes without a reward. The coccid then rotated
180 deg. on its dorsoventral axis so that its anal end was now
_ presented to beetle No. 2, and inserted its beak into another part
of the nutritive parenchyma. The beetle at once became more
alert and accelerated the beat of its antenne. During the suc-
ceeding eight minutes it received seven globules of honey-dew
in quick succession, probably as a result of the coccid’s change
of pasture. Throughout this period beetle No. 1 kept titillating
the coccid’s side, pausing now and then for a few seconds, and
after 40 minutes from the time I began the observation, moved
away. Beetle No. 2 continued to stroke the coccid for some
time, but I did not follow its behavior further.

In another colony at 8 P. M. on the same day I noticed a
nearly mature beetle (No 3) vigorously stroking the hind end
of a small coccid, while its sides were being stroked simultane-
ously by two just-hatched larve (A and B). From time to time
other beetles and older larvee came up and joined the party.
Beetle No. 3 continually butted the newcomers away and they
at last rather reluctantly departed, leaving the original trio in
possession of the coccid. Every few seconds the beetle gave

1921] Wheeler: Some Social Beetles 71

one of the larvze a shove with its head, but the tiny creature
instantly returned and went on with its stroking. At 8.14 the
beetle gave A such a vigorous knock that it stayed away from
the coccid for some time. B, however, kept returning so pertin-
aceously that the beetle twice seized it in its mandibles for an
instant and then dropped it. The larva was uninjured, however,
and at once returned and went to work. Then the periodic but-
ting continued till 8.35 when larva A returned and went to work
with B on the side of the coccid. One or the other was butted
away by the beetle every few seconds till 9 P. M. During the
entire hour the coccid remained stationary and unresponsive,
never once raising its caudal segments nor emitting a droplet.
All this time the beetle had remained in the same spot and had
butted every beetle or larva with which its antenne had come
in contact. The beetles soon left after receiving a few knocks
but the little larve A and B, which seemed to be famished,
persisted for a whole hour side by side, except for the 20
minutes during which A was absent. Larva B must have been
struck by the beetle more than a hundred times. Finally the lat-
ter’s patience seemed to be exhausted; it seized first A and then
B in its mandibles, carried the latter three millimeters away
from the coccid and hurled it to one side. Larva A returned, but
B had fallen out of the petiolar cavity onto the moist wall of
the glass tube, adhered and was unable to leave the surface. The
beetle now left the coccid and another very mature beetle (No.
4) took its place. It permitted larva A to stroke the posterior
end of the coccid without molestation, but beetle No. 3 soon
bustled up from the opposite direction, locked mandibles with
beetle No. 4 and pushed it away. During the scrimmage the
coccid suddenly raised its caudal end and discharged a droplet
which was eagerly inbibed by the larva, at length rewarded for
its incredible pertinacity and the innumerable knocks it had
received. Then beetle 3 and larva A, now in undisturbed posses-
sion of the coccid but in the reversed position, the former being
at the anterior, the latter at the caudal end of the coccid, con-
tinued their stroking, interrupted every few seconds by the
butting of the beetle and the temporary withdrawal of the
larva. This went on till 9.20. By that time my eyes which had
been following the performance under the lens for an hour and
twenty minutes were so fatigued that I had to desist from further

72 Zoologica: N. Y. Zoological Society [IlI;3

observation, just after the beetle had tossed the larva to a dis-
tance of about four millimeters from the coccid by an unusually
well-aimed blow.

Scenes of this description were so frequently enacted that
they could be readily observed in almost any of the colonies after
they had been kept for several days and the beetles, larve and
coccids had all grown very thirsty. In such colonies the bodies
of the coccids and larve become visibly attenuated and some-
what shrivelled as a result of the loss of water from the tis-
sues of the Tachigalia petioles. All the insects now become rest-
less. The beetles leave the petioles, wander about on the walls
of the tubes and finally collect about the plugs of cotton in an
endeavor to escape to the outside. The coccids, too, withdraw
their beaks from the parenchyma in the floors of the food-
grooves and wander aimlessly about, vainly seeking more favor-
able pastures. But before this stage of demoralization is reached,
both the beetles and the larvee become cannibalistic and one
may often see them, singly or in groups voraciously devouring
partly dismembered larve or immature beetles. Within a few
days all the larve and immature beetles are consumed, but the
coccids, immune from attack, still wander about till they die
of starvation.

I believe that Coccidotrophus is rarely or never cannibal-
istic under normal conditions. It is, as already stated, almost
impossible to split a freshly gathered Tachigalia petiole con-
taining one of the beetle colonies, without cutting some of the
insects iu two, and such disabled individuals are soon devoured
by their fellows, but both in such cases and in the cannibalism
that supervenes in dried petioles, I believe that thirst or the need
of water and not a veritable carnivorous instinct, such as seems
to be manifested by some species of Cucujid beetles and their
iarve, must be regarded as the true explanation. I am con-
firmed in this view by Heins’ recent investigations (1920) on
meal-worms (Tenebrio molitor). He found that when the larve
of this beetle are reared in dry meal as many as 24.2% of
them may be devoured by their fellows, but that if wet slices
of rusk or of vegetables are placed in the breeding jars the
mortality from cannibalism is reduced to 7.5%. In this connec-
tion Bodine’s observations (1921) on grasshoppers are also of

1921] Wheeler: Some Social Beetles 73

interest. He finds that during starvation, the loss of water in
these insects is always greater than that in body weight or in
the solids. “This shows that starvation in the grasshopper
results in a rapid loss in water which has a decidedly quick and
fatal effect.”

I have been unable to ascertain the length of the larval per-
iod or the number of larval moults of Coccidotrophus. As no
exuvize were found in the petioles it would seem that they
must be devoured either by the beetles or by the larve them-
selves. The food of the larve, as we have seen, consists of the
amber-colored nutritive parenchyma and of the sweet excreta of
the coccids, the former evidently supplying the proteids, the
latter the sugar and most of the water. So concentrated a diet
should be very favorable for growth and probably the whole lar-
val period at tropical temperatures occupies only two or three
weeks. The fat-body, however, does not seem to become very
voluminous till the last larval instar when the segments of
the body become more convex and puffed out with the accumula-
tions of adipocytes. Yet this condition, which immediately pre-
cedes pupation, does not tend greatly to inhibit the activities of
the larva.

When a petiole containing a colony in what I have called the
third stage is opened, one or more cocoons are invariably found
in the cavity (Plate V). They are oblong-elliptical structures,
5-6 mm. long and 2-3 mm. broad and seem to consist of the
same chocolate brown substance as the frass-ridges. Their
walls are of uneven thickness, with smooth inner and rough-
ened outer surfaces, and are easily fractured. These cocoons
do not lie loosely in the cavity but are attached to some flattened
surface of the wall where the lumen of the petiole is rather
broad, i.e., away from the pointed ends, and always have their
long axes parallel with the long axis of the cavity. They are
sometimes single but more frequently occur in pairs or in
groups of three or four. When in pairs, the two cocoons lie
abreast of each other, when in threes or fours, the third and
fourth cocoon are often built on top of a basal pair. Such groups
of cocoons are so voluminous that they obstruct the lumen of
the petiole and leave only a narrow passage for the beetles to
move between the more roomy spaces at either end of them.

74 Zoologica: N. Y. Zoological Society [lis

One naturally infers that either the larve must make the
cocoons of frass or the beetles must envelop the pupz with this
material, but observation shows that both inferences are incor-
rect. The larva does, indeed, build the cocoon, but utilizes nei-
ther the frass nor the materials of old, abandoned cocoons in its
construction. I have not seen the earliest stages in the process
but it is evident that the larva selects a flat surface and begins
to build a wall around an elliptical area, which thus becomes
the floor of the cocoon. Little material is added to the wall
at the end of the ellipse compared with the sides, where the
material is built up as a pair of folds like those shown in Fig. 16a.
I have seen several cocoons that had been abandoned in this or
a somewhat more advanced stage, but on two occasions I was
able to observe the completion of the structure from a stage like
the one figured. Since in both cases the insects behaved in
essentially the same manner I shall describe only one of them.

The larva was first seen working inside the cocoon in the
stage of Fig. 6a, but it soon came out, wandered away to a
distance of a few millimeters and, after careful search, bit off a
minute particle of the living tissue of the petiolar wall, avoiding
any frass-covered surface, returned, entered the cocoon at one
end (left hand side of Fig. 6a), carefully masticated the particle
with its maxille, while mixing it with saliva, applied it to the
border of one of the folds, pressed it into place, crept out of
the other end of the cocoon and went in search of another parti-
cle. Then it returned, entered the cocoon as before and repeated
the building process. Excursions were made every few minutes
and within a radius of 8 to 10 millimeters from the cocoon. The
particles, which were selected with the greatest care and often
after what seemed like some hesitation, were very minute and
greenish when first bitten off but had become brown (by a pro-
cess of oxydation?) by the time they had been incorporated in
the walls of the cocoon. The particles were applied now to
one fold of the wall, now to the other so that the edges became
rather irregular (Fig. 6b), but as most of the particles were
added to the middle of the folds, they began to approach each
other. Still, their growth was very slow, owing to the minute
size of the particles and the time consumed in their selection.
The larva labored incessantly, making trip after trip and choos-

1921] Wheeler: Some Social Beetles 75

ing every particle with the same diligence and avoiding the re-
mains of empty cocoons in the immediate vicinity, although
their materials, one would suppose, might have been easily ap-
propriated and quickly built into the cocoon under construction.
The two folds or side-walls slowly approached as the work pro-
gressed and eventually fused with each other, the larva always
entering at the same end, applying the particle to one of the
edges from the inside and leaving by the opening at the opposite
end. Then it set itself to building the walls around this latter
opening, which grew smaller and smaller (Fig. 6c), till the larva
could no longer squeeze through it and was compelled when
about to leave the cocoon to turn back on itself, bending its
body in a loop with the two limbs in contact, and crawl out of
the opening by which it had entered. This feat seemed to be
accomplished with considerable effort. but had to be performed
after each particle had been built into the wall of the cocoon.
Eventually the small opening was closed and the cocoon had
only a single large elliptical orifice at one end (Fig. 6d). The
larva now began to contract this orifice, but after a time, as it
grew smaller, the insect on returning, no longer entered the
cocoon and reversed its body in order to apply the particles to
the edge of the orifice, but merely thrust its head and a few of
its anterior segments into the cocoon and left the remainder
of its body outside. At such times it used as a support or ful-
crum a structure which J had not seen used at any previous
stage of larval life, namely, the proleg which terminates the
conical tenth abdominal segment. This structure is described by
Dr. Boving (Zoologica III, No. 7) and clearly shown in his figures
(Pl. VII, figs 1 and 2). When the size of the orifice had been
reduced till the larva could only just squeeze through it, the
insect entered the cocoon, reversed its position and continued
building along the edges of the orifice with particles scraped
from the inner surface of the structure. The orifice thus soon
grew too small to permit the egress of the builder. (Fig. 6e).
Then the imprisoned creature slowly closed the opening and the
cocoon was completed (Fig. 6f).

A few days after the larva has thus immured itself, it sheds
its cuticle and becomes a pupa which lies loosely in the cavity of
the cocoon and has the appearance of Dr. Boving’s Plate IX, figs.

“ tol.

76 Zoologica: N. Y. Zoological Society [IIl;3

19-21. Owing to the minuteness of the particles used in building
the cocoon, the care with which they are chosen and the many
trips necessary to secure them, the time consumed in completing
the structure is considerable. The earliest stage figured (Fig. 6a)
was first seen at 8 P. M., July 25. By 6 A. M. the following
morning the cocoon was in the stage shown in b. By 12.30 P.
M. the small opening had been closed and the large opening was
being contracted (c). At 6:30 P. M. a small opening remained
(e), and the cocoon was completed an hour later (f). As the
first stage must have been the work of the greater part of a day,
the structure was probably begun not later than 7 A. M. on
July 25th. At least 36 hours of continuous labor, requiring
hundreds of trips back and forth between the cocoon and ex-
posed patches of living parenchyma on the petiolar wall, were
therefore consumed in completing the cocoon. The second larva
observed in the act of building its cocoon was even slower, since
the latter was first seen on the evening of July 26 in a stage
corresponding to Fig. 6a and was not entirely completed till after
10 A. M., July 29th. There was nothing to indicate that the
first larva rested during the whole period of cocoon construc-
tion. While it was working in the manner described, it was
occasionally annoyed by some young or half-grown larva enter-
ing the cocoon and using it as a hiding place while its architect
was away gathering building materials. It was interesting to
see the latter on its return oust the intruder, which scampered
away with comical alacrity. When the cocoon is completed it
is rather smooth externally but may later become rough through
the beetles’ plastering their frass over its surface. This cer-
tainly strengthens the walls of the structure.

I endeavored to keep the two pup enclosed in the cocoons
which I had seen built, in order to determine the length of the
pupal period, but both died when the petioles dried out. The
pupal period as inferred from other cases covers, at least, seven
days. The callow emerging beetle under normal conditions
gnaws a round or elliptical opening at one end of the cocoon
and joins the other members of the colony. At first it is yellow-
ish white and etiolated, with the legs, dorsal surface of head
and prothorax and a transverse band on each segment of the
venter pale red. It runs about very actively, nevertheless, and

1921] Wheeler: Some Social Beetles ra

in the course of several days gradually takes on the deep chest-
nut brown color of the mature insect. But before this stage is
reached, and while the male and female are still of a bright red
or pale chestnut red color, they mate. On two of the three occa-
sions on which I witnessed copulation both the male and the
female were immature. The third couple, observed, August
10, consisted of an immature female and a mature male. It
would seem, therefore, that mating occurs not only among im-
mature beetles of the same generation soon after they leave
the cocoons, but that old males of a preceding generation at
least occasionally fecundate the recently emerged females.

The observation of August 10 is here transcribed from
my note-book. The female was distinctly immature but uni-
formly red, i.e. no longer in the white, callow stage, and dis-
tinectly smaller and more slender than the male. The latter was
certainly very nearly or quite mature. When first noticed the
female was eating the parenchyma of the petiolar wall. The
male mounted her back and remained for 18 minutes, clasping
her sides with his legs and occasionally attempting to insert his
aedoeagus. Now and then he rubbed her occiput from side to
side with his mandibles and antennz and sometimes shifted his
position very much to one side. The antennal movements were
precisely like those employed in titillating the coccids. The
female kept on feeding and pressed the tip of her abdomen
against the wall of the petiole so that the male was unable to
introduce his aedoeagus. He then dismounted and ran away
but soon returned and attempted to mount and grasp the female
again, but she was unwilling, slipt out from under him and
escaped. He permitted her to go only a very short distance,
however, before he again seized her just as she had stumbled on
a coccid and had begun to stroke its posterior segments. While
the male was strenuously endeavoring to copulate she continued
to stroke the coccid and on this occasion kept the tip of her abdo-
men tightly pressed upward against the tips of her elytra. Then
the male again dismounted and left her and she and another
immature beetle turned their attention to a partly eaten larva
which they proceeded to devour. In a moment the male, appar-
ently in a high state of excitement, returned, mounted the
female and this time succeeded in introducing his sdoeagus by

78 Zoologica: N. Y. Zoological Society (iS

forcing the tip of the female’s abdomen downward and away
from the tips of her elytra. During coitus, which lasted a little
more than six minutes, the female continued to partake of the
larva, but the male remained motionless.

B. THE ENEMIES OF THE BEETLES AND COCCIDS AND THE DECAY
OF THE COLONIES.

Attention has been called to the decay of the Coccidotro-
phus colonies when the flow of sap to the petioles is artificially
cut off by severing the latter from the plant and both coccids
and beetles are deprived of their nourishment. But even under
the natural conditions of the jungle the colonies are doomed to
decay, though from very different causes. As has been shown,
Coccidotrophus lives only in the petioles of young Tachigalias
growing in the shade and in these plants inhabits the petioles
only till they are taken over by the obligate ants of the genera
Pseudomyrma and Azteca. The beetles are not permitted, so to
speak, to occupy their apartments after the rightful owners of
the plant have become sufficiently numerous and aggressive to
eject them. Sometimes this may occur even in rather young
plants four to six feet in height. Still the period during which
the beetles may be allowed to inhabit the young Tachigalias
must cover several months. I infer this from the fact that dur-
ing the more than two months of my stay at Kartabo I saw little
change in the plants, which grew very slowly notwithstanding
the almost daily, drenching rains, and their growth is probably
almost nil during the dry season. Since the leaves are persist-
ent, at least during the rainy months, there is ample time for
the development of the beetle colonies, even if the growth of the
larval broods and the coccids requires more than the three weeks
above suggested. Throughout the latter part of July, August
and the first half of September colonies were found in all stages,
from those represented by a single pair of beetles to those com-
prising numerous beetles, larve, pupe and herds of coccids. It
would seem, therefore, that although the individual colony may
live for only a few months, new colonies must be formed contin-
ually, at least during the rainy season, by emigration of pairs of
young beetles from old colonies to other plants which have

1921] Wheeler: Some Social Beetles 79

attained a sufficient size, i.e. have produced at least three or four
fully developed leaves.

Still neither the beetles nor the coccids are permitted to
live in perfect security till the obligate ants take possession of
their quarters. Where competition among insects is so very
keen as it is in the Neotropical jungle it is not surprising to find
that several predators and parasites are continually gaining
access to the petioles and decimating or even completely destroy-
ing their occupants. The greatest enemy of the beetles is the
small thief-ant Solenopsis altinodis Forel (See p. 48 and Zoologica
III, No. 4, p. 154), and the coccids have at least three formidable
enemies. All of these insects enter the petiole through the open-
ings made by the beetles and must therefore elude their watch-
fulness. We should expect the beetles to keep one of their num-
ber constantly on guard at the entrance, but they are neither suffi-
ciently constant in this behavior nor sufficiently discriminating
to keep out all intruders. When the petioles are taken into the
laboratory the beetles are often seen to remain for hours with
their heads in, the entrances and their bodies at right angles to
the longitudinal axis of the cavity, and even after the petioles
have been split longitudinally and placed in vials the insects still
exhibit this behavior, though it is now absurdly futile, since their
domicile is wide open. But not infrequently even the single
opening of a petiole may remain unguarded for long periods,
and when the petiole has several openings some of them are
apt to have no sentinels, so that predators and parasites small
enough to pass the narrow orifices, have no great difficulty in
gaining access to the colony. Moreover, a beetle that is guard-
ing an opening may fight off certain intruders but back away
and allow others to enter. On several occasions I held a beetle
with its head to a guarded entrance. The sentinel at once
grasped the stranger’s head with its mandibles and pushed it
away. But when I placed a worker Solenopsis altinodis in the
same position, the beetle beat a hasty retreat and the ant climbed
into the petiole. From these experiments we may infer that
the beetles are more intent on keeping alien beetles of their own
species than dangerous pests like the Solenopsis out of their
nests. More probably some peculiar odor of the ant induces the
beetle to withdraw. Thus while it seems to be probable that alien

80 Zoologica: N. Y. Zoological Society LS 33

beetles are often kept out of the colony, the fact that the beetles
ot two or more colonies occupying different petioles, will, when
the latter are split open and placed in the same tube, mingle
without the slightest signs of hostility, would seem to show that
even strange beeties may occasionaily enter an unguarded colony
and become members of it in good standing. It has _ since
occurred to me that female beetles, at least, might be permitted
to pass the sentinels unchallenged. Unfortunately I failed to
dissect and determine the sex of the beetles with which I
experimented.

The laxity of the beetles in guarding the entrances is, indeed,
amply proved by the presence in their nests of several species of
insects, some of which are harmless or indifferent while others
are injurious either to the beetles and their larve or to the
coccids. From analogy to the guests of ants, those of the former
category may be called “synoeketes,” or indifferently tolerated
- guests, the latter “‘synechthrans” (predators) and parasites. To
the synoeketes belong a Collembolan, a mite and a small Phorid
fly. The Collembolan is most frequently seen, especially in large
petioles containing small colonies of beetles and therefore allow-
ing ample space for its movements. It is a minute silver gray
species, which Prof. Folsom has described and figured as E’nto-
mobrya wheeleri (Zoologica III, No. 11), and occurs in droves of
individuals of all stages, running hither and thither over the walls
of the petiole, like certain species of the same genus (E. myrmeco-
phila Reut. and dissimilis Mon.) and Cyphodeirus (C. albinos
Nicol), which are often abundant in the nests of ants. Like the
ants the Coccidotrophus pay no attention to these diminutive in-
sects and are probably not even aware of their existence. The
mites (Hypoaspis sp.) and Phorid flies (Aphiochexta scalaris)
were more rarely seen. They probably breed in the accumulations
of refuse at the ends of the petiolar cavity and may therefore
be regarded as scavengers, like the mites and Phorids which
occur in many ant-nests.

Careful examination of the alimentary tract of the Coccido-
trophus would probably show that the beetle harbors a number
of entoparasites, at least certain bacteria, but I could not find the
time to make such an examination when fresh material was avail-
able and my alcoholic specimens are worthless for the purpose.

1921] Wheeler: Some Social Beetles 81

FIG. 12. HYPHOMYCETOUS PARASITE ON COCCI-
DOTROPHUS SOCIALIS.

From a drawing by Prof. Roland Thaxter.

It seemed probable, nevertheless, that the beetle, living as it
does in dark, moist cavities, might be infested with ectoparasitic
fungi, especially of the group Laboulbeniales. I therefore re-
quested Prof. R. Thaxter, the leading specialist in this group, to
examine a large number of the beetles. After carefully scrutin-
izing their external surfaces he reports that he found no Laboul-
beniales, but only a sterile Hyphomycete, growing on the elytron
of one of the specimens. Referring to his work on the similar
fungi of other insects he writes me as follows: “The fungus on
the Coccidotrophus probably belongs to the group spoken of at
the bottom of p. 237 in my first paper (1914), the most striking
form of which (Aposporella elegans Thaxter), found on the
wings of a small fly from the Kamerun, is figured on Plate III,
Fig. 30 of the second paper (1920). These fungi seem to produce
no spores and to reproduce by a kind of fragmentation; pieces
breaking off with little or no differentiation, and starting to grow
where they adhere. I have seen a considerable number of them
on a variety of tropical insects, and ran across one of them a
few days ago growing on a Laboulbenia from Kamerun. It
has seemed hardly desirable to give names to such nondescript

82 Zoologica: N. Y. Zoological Society puss

forms till it is quite certain that they have no differentiated type
of reproduction.”” Prof. Thaxter kindly enclosed a drawing of

the Coccidotrophus Hyphomycete which is here reproduced as
Fig. 12.

The various organisms just described can, of course, have
little or no effect on the health of the beetle colonies. Their
decay and eventual extinction is due to ants destroying the
beetles and their larve or to predators or parasites destroying
the coccids on which they depend for an important part of their
diet. Colonies in all stages may be invaded by the tiny Solenop-
sis altinodis, which I believe to be the most persistent and deadly
enemy of the beetles. I have opened petioles containing both
beetles and Solenopsis workers, but no beetle larve or only frag-
ments of them, showing that the ants begin their depredations by
slaughtering the offspring of the beetles. In such petioles the
coccids remain uninjured, and the same is true of the pupe im-
mured in their cocoons. Many other petioles reveal conditions
from which the last stages in the history of the colony may be
inferred. In such petioles only dead beetles and a number of
dead Solenopsis are found, indicating that the ants, after destroy-
ing all the larvee, attack the beetles and that they, in the ensuing
bitter conflict, often defend themselves with their powerful jaws
to such good purpose that they succeed in killing many of the
ants. But as more Solenopsis are probably continually entering
the petioles as auxiliaries the beetles finally succumb and the
colony is exterminated, with the exception of such pupz as may
be present in their protective cocoons. The beetles emerging
from these may, conceivably, after the Solenopsis invasion is
over, start a new colony in the same petiole or emigrate to other
petioles. If no pup are present the petiole is sooner or later
entered by a pair of young migrant beetles, which pack the dead
bodies of the previous occupants, together with the ants they
have slain, into the ends of the cavity and establish a new colony.
In some of the colonies in the last stages of devastation above
described, I failed to find any coccids. They may have been either
eaten or carried away by the Solenopsis. Sometimes they remain
undisturbed, however, and may, perhaps, be taken over by the
beetles emerging from the cocoons or by any pair of young
beetles entering the petiole and establishing a new colony.

1921] Wheeler: Some Social Beetles 83

We can picture to ourselves the fierce battles which rage
in the petioles between the beetles and the Solenopsis workers,
probably mostly at night, for the Solenopsis is a nocturnal spec-
ies, and the precarious life of the beetles in parts of the jungle
where the ant is abundant. The beetles must live, in fact, like
the ancient Greeks, always in danger of invasion from the war-
like hords of barbarians. Yet even in quiet recesses of the
jungle, where the Solenopsis may happen to be rare or absent,
the attachment of the Coccidotrophus colonies to the Tachigalia
is sure to be severed as the plant grows and the workers of the
colonies started in one or more of the petioles by dedlated queens
of Azteca or Pseudomyrma have become sufficiently numerous to
take possession of every petiolar cavity and patrol the whole sur-
face of the plant. Perhaps some of the inquiline ants may occa-
sionally kill or oust the beetles, but as these ants merely occupy
a petiole here or there on the young trees, they cannot be regarded
as very serious enemies. Many of them, too, are small, timid
ants, which probably have their own battles to fight with the
insidious Solenopsis and are destined to be supplanted, like
the beetles, by the obligate Pseudomyrmas and Aztecas.

During the struggles between the beetles on the one hand and
the Solenopsis, Pseudomyrmas, and Aztecas on the other,
the poor coccids evidently play somewhat the same defenceless
role as the cattle in a country overrun by contending armies—
they merely change masters and are either eaten, or carried off
or permitted to remain and produce honey-dew for the victors.
But before any such change of masters occurs they are often
decimated or even exterminated by three enemies of their own,
which may be briefly described seriatim.

In a few of the beetle colonies I have seen a number of
larve of a very small Coccinellid beetle, described by Schwarz
and Barber as Scymnus xantholeucus (Zoologica III, No. 6).
These larve when full-grown resemble the larger coccids so
closely in size, form and color and are covered with such a similar
layer of snow-white wax, that I frequently overlooked them in my
living colonies and detected their presence only in the preserved
material after my return to the United States. They move slowly
about among the beetles and their larve and devour the coccids.
I am inclined to believe that by the time they are ready to pupate,

84 Zoologica: N. Y. Zoological Society ality

they have also devoured all the beetle larve since I found two
petioles each containing nothing but a pair of Scymnus pupe, a
few shrivelled remains of coccids and the kitchen-middens at the
ends of the cavity. The pupz were attached to the wall by
their caudal ends and with their longitudinal axis parallel with
that of the petiole. They were of a waxy yellow color, with their
surface studded with short, blunt hairs. Several days later the
imaginal beetles emerged. They measured 1.7 mm. in length
and 1.2 mm. in width and were pale yellow, with the basal two-
thirds of the elytra, the meso- and metasternum and the median
third of the first and second abdominal segments fuscous.

A much more abundant enemy of the the coccids is a peculiar
predacious Itonidid (Cecidomyid) fly, which Dr. Felt has de-
scribed as Diadiplosis pseudococci (Zoologica III, No. 8). The
larve of this insect are orange red and are often found in clusters
of as many as eight to a dozen around groups of coccids. The
whole mass is covered by a tough, dense, white web, or tent, which
is spun by the larve in such a way as to shut them and the coccids
off from the cavity of the petiole and therefore from contact
with the beetles or their larve. Thus secure from interference
the Diadiplosis larve devour the coccids at their leisure, attack-
ing them from the ventral side where their integument is thin-
est and free from wax. The coccids are eventually reduced to
their dorsal integument. When mature the Diadiplosis larva
pupates where it has been feeding, often in the midst of a group
of young or full-grown larve, and without orienting itself with
respect to the longitudinal axis of the petiole. Just before eclo-
sion the pupa forces its body, head foremost, through the silken
tent and projects into the cavity. The fly then emerges and prob-
ably either lays its eggs among any surviving coccids or emigrates
from the petiole and enters other beetle colonies. The adult fly
is a very delicate little midge measuring only 1.25 to 1.5 mm.
and of a reddish-brown color, with the abdomen red internally
and the sclerites somewhat infuscated.

Considerable interest attaches to this insect, because, unlike
the great majority of Itonidids and the species noticed above (p.50),
it does not make plant galls but is predaceous. The genus Diadi-
plosis was originally established by Felt (1911a), for D. cocci Felt,
a species reared in the island of St. Vincent from larve preying

1921] Wheeler: Some Social Beetles 85

on the eggs of a coccid, Saissetia nigra Nietn., on stems of Sea
Island cotton. In another paper published during the same year
(1911b) he gives a list of 19 species of known zoophagous Amer-
ican Itonidids. The list includes species of Endaphis, Arthro-
cnodax and Mycodiplosis feeding on mites, an unidentified species
of Cecidomyia feeding on the eggs of Cicada septemdecim, sev-
eral species of Aphidoletes and Lestodiplosis preying on aphids
and of Lestodiplosis, Dentifibula, Diadiplosis, Coccidomyia, Ceci-
domyia, Lobodiplosis, Mycodiplosis and Dichrodiplosis preying
on various coccids. The Diadiplosis from British Guiana seems
to be closely related to the type of its genus. According to Kiis-
ter (1911), certain European Itonidid larve have been described
by Riibsamen (1899) and Kieffer (1902) as preying on the larve
of gall-makers of the same family.

Aimost as abundant as the Diadiplosis in the beetle colonies
is a Hymenopterous parasite of the Psewdococcus, namely Blep-
yrus tachigalie Brues (Zoologica III, No. 9), a small Encyrtid of
the family Chalcididze. The white larva of this insect lives in the
coccid and grows with it, eventually becoming so voluminous that
the coccid’s body is very convex both dorsally and ventrally. The
coccid grows increasingly sluggish and inert and its wax-glands
cease to function so that its integument takes on a dull brown-
ish color and the wax-pencils disappear from its periphery. The
beetles and their larve are, of course, quite unaware of these pro-
found changes in their parasitized cattle and still continue to
stroke them, often for long periods, although there is no honey-
dew forthcoming as a reward for their efforts.

When full-grown the larval Blepyrus does not escape from
the coccid but remains within it and forms an amber-colored,
regularly elliptical cocoon about 2 mm. long and therefore very
nearly as large as the coccid, which is now reduced to a mere
skin enveloping the huge parasite. The cocoon seems to consist
of a hard, glassy substance, possibly a modified silk, and is cov-
ered except on its ventral side with small circular spots which
represent thinner, depressed areoles in its wall. Where these
areoles are lacking on the ventral side the wall is homogeneous
and distinctly thinner than elsewhere, but has a number of small
pointed projections which seem to pierce the ventral integument
of the coccid and to attach the cocoon rather firmly to the wall of

86 Zoologica: N. Y. Zoological Society [Tes Ss

the petiole. Since the parasitized coccid remains in the food-
groove with its body in the usual position the cocoon necessar-
ily has the same position and orientation, i.e. with its long axis
parallel with the long axis of the petiole. The cocoon gradually
grows darker, passing from amber-yellow to dark brown and by
the time it has reached this stage, the dead tissues of the coccid
enveloping it, except those on the ventral side, between the cocoon
and the wall of the petiole, disappear, leaving the lateral and dor-
sal surfaces of the cocoon fully exposed. I am inclined to believe
that the dead tissues of the coccid are eaten away by the beetles
or their larve, but as they are very soft and disintegrate easily,
they may perhaps be rubbed off merely by the attrition of the
insects as they move back and forth in the petiole. Two of the
denuded cocoons in the stage and with the orientation just des-
cribed are shown in Fig. 11.

When the completed Blepyrus cocoon is cleared in carboi-
xylol, mounted in balsam and examined as a transparent object,
the larva is found to have pupated within it, after extruding a
number of large meconial pellets, the undigested remains of all
the food it swallowed while it was living on the tissues of its
host. In most cases, at least, the head of the pupa is at the caudal
end of the coecid. The imago, when mature, cuts a large round
hole in the end of the cocoon (see upper part of Fig. 11) and
emerges as a short, thickset, broad-headed fly, only 1.5 mm. long,
with a metallic green face, a black, more or less bronzed body,
black and yellow antenne and legs and basally infuscated wings.
It is very active, and like other small Encyrtids skips about by
using the long saltatory spurs on its middle tibie. After mating
the female undoubtedly oviposits in the young coccids either in
the same or in some other beetle-inhabited petiole.

This parasite seems not to be nearly so serious a menace to
the Coccidotrophus colonies as the Scymnus and Diadiplosis, since
the infested coccids are probably able to supply the beetles and
their larve with honey-dew till both host and parasite are nearly
mature. Hence one often finds several infested coccids and
Blepyrus cocoons in petioles inhabited by flourishing beetle col-
onies. In one such colony I counted more than fifty cocoons and
a dozen large coccids swollen with parasites that were still in
the larval stage.

1921] ( Wheeler: Some Social Beetles 87

I have failed to find more than one of the three species of
coccid enemies in a single petiole. Their combined action, if
they actually ever occur together, would, of course, not only
greatly hasten the extermination of the coccids, but would seri-
ously interfere with their own development. It may be noted
incidentally that none of these enemies occurs in the petioles of
the large Tachigalias inhabited by the obligate Pseudomyrmas
and Aztecas. In such plants the coccids are free from all preda-
tors and parasites and are not only more numerous but attain
a larger size than in the petioles of the small shade trees tenanted
by the Coccidotrophus. The ants are undoubtedly much more
efficient than the beetles in keeping small miscellaneous guests
and synoeketes out of their nests. This is particularly true of
the Pseudomyrmas. Although I have collected the entire colonies
of many of these ants on several different trips to the American
tropics, the only synoekete I have ever seen associated with them
was a Microdon larva described many years ago (1901) from
the nest of Pseudomyrma mexicana Roger. Even coccids are kept
and attended by only a few species of Pseudomyrma.

Before concluding my account of Coccidotrophus I may intro-
duce a few statistical data, which are probably valid only for
the particular time and locality of my observations. While at
the Tropical Laboratory I noted roughly the condition of the
contents of each of the Tachigalia petioles I opened on a particu-
lar day. On some days only a few petioles were opened and
the results are not worth transcribing. The following collection,
however, gives:a more interesting picture owing to the number
of petioles examined:

August 9. Collected 253 petioles from young Tachigalias
11% to 7 ft. high growing along the Cuyuni Trail. Of these 37
or about 14% were either too young to have inhabitants or con-
tained solitary Pseudomyrma queens founding colonies or smal!
colonies of inquiline ants; 203 or about 86% either contained or
had contained beetle colonies. Of the latter number, 50 contained
incipient colonies, i.e., a single pair of beetles or more rarely sin-
gle beetles which had just entered the petioles and were busy
“cleaning house.’ In one petiole one of the beetles of a pair was
guarding the entrance while the other was shovelling frass and
the remains of previous occupants with the top of its head into

88 Zoologica: N. Y. Zoological Society [1Il;3

the pointed ends of the cavity. Sixty-four of the beetle colonies
had larve and were in what I have called the second and third
stages. In nearly every case coccids were seen. Highty-nine of
the colonies were either moribund or extinct. Solenopsis altinodis
workers either living or dead, were present, sometimes in con-
siderable numbers, in 35 of these petioles, and 10 of them still con-
tained large coccids that had been shut off by webs and were being
devoured or had been already devoured by Diadiplosis larve. In
one petiole two of the flies had emerged. In 6 of these petioles
the Solenopsis workers had destroyed the beetles and their larve
and were still running about. When I tore away the webs
covering the Diadiplosis larve the ants at once seized and killed
them. The webs serve, therefore, not only to protect the Itoni-
dids from the beetles and their larvez, but also from the Solenop-
sis.

C. EUNAUSIBIUS WHEELERI SCHWARZ AND BARBER.

This beetle, though superficially very similar to Coccido-
trophus socialis, can be easily distinguished in all its postembry-
onic stages. The adult beetle (Plate III, fig. 1, Plate VI, figs. 6
to 10) is distinctly smaller, measuring only 3-3.5 mm., perman-
ently of a red color like the immature Coccidotrophus and there-
fore never deepening into the dark chestnut color of the latter.
The antennal clubs are larger and broader and much more dis-
tinctly marked off from the more proximal joints, the eyes are
much larger, the anterior border of the front is much less
deeply emarginate, the femora are less incrassated and the pos-
terior pair has a small tooth on the flexor side. The surface of
the body is smoother, the punctation being less pronounced. The
larva is more slender, with the head and dorsal surface distinctly
gray, owing to a deposition of fine pigment granules in the integu-
ment. The pupa can be at once recognized by the presence of
four large, equidistant tubercles on each of the parallel lateral
borders of the pronotum (Fig. 8a, Plate IX, fig. 23). For many
of the less obvious differences between the various instars of
the two beetles the reader may be referred to the excellent des-
criptions and figures of Schwarz and Barber (Zoologica III, No.
6) and Boving (Zoologica III, No. 7).

1921] Wheeler: Some Social Beetles 89

The prominent tubercles on the sides of the pupal pronotum
of Hunausibius merit somewhat fuller consideration, because
they present a striking instance of the retention in an earlier
ontogenetic stage of a character which may be completely lost
in the adult. An examination of the common saw-toothed grain-
beetle, Oryzxphilus (formerly Silvanus) surinamensis L., repre-
sented in Fig. 7, and other species of the same genus, shows
that the adult beetle has six acute teeth on each of the lateral
borders of the pronotum, corresponding to and arising within
as many large, blunt tubercles of the pupa. These structures
were long ago noticed by Coquerel (1849) and Perris
(1852) and by the former erroneously supposed to be por-
tions of some tracheal system peculiar to the pupa. In
Nausibius (N. clavicornis) the pronotum of the beetle bears
six obtuse teeth on each side. In other Silvanid genera, such
as Silvanus and Cathartus as well as in Hunausibius and
Coccidotrophus these teeth are either altogether absent in the
imago, or reduced to the first pair, which form the anterior
corners of the thorax. Hunausibius has well-developed teeth in
this position but in Coccidotrophus the anterior corners of the
pronotum are merely rectangular. It is therefore interesting to
find that the pupa of Hunausibius has four well-developed pairs
and a fifth vestigial pair of tubercles, that these tubercles
decrease in size anteroposteriorly, and that only the first pair
gives rise to teeth that persist in the adult. In Coccidotrophus
the reduction of the pupal tubercles is carried much further since
there are only small vestiges of the three anterior parts of Huna-
usibius, none of which gives rise to teeth in the imago. More-
over, the second and third pairs of pupal tubercles may be repre-
sented by only a single tubercle on one side of the pronotum, as
Béving observed (Zoologica III, No. 7). The tubercles are, in
fact, so evanescent that they have become very variable, like ves-
tigial organs in general. This is seen in Fig. 8—b to f, represent-
ing the prothoraces of five Coccidotrophus pupe selected from a
series of fifty specimens. We may safely conclude, therefore, first,
that Hunausibius and Coccidotrophus are derived from ancestors
which had a 12-toothed pronotum like the species of Oryzae-
philus; second that this condition disappeared first in the imago
and still tends to linger on in the pupa, and third, that the tuber-
cles have a tendency to disappear in sequence in a posteroanter-

90 Zoologica: N. Y. Zoological Society [III ;3

ior direction. There can be little doubt that the dentation of
the sides of the thorax is a very ancient character not only in
the Silvanids but also in the Cucujids (as restricted by Boving),
since vestiges of the teeth can also be clearly seen in the imagines
of certain genera of the latter family (Cucujus, Brontes).

I have already alluded to the fact that Ewnausibius colonies
are much rarer at Kartabo than those of Coccidotrophus though
both species may occur in the same localities and even in differ-
ent petioles of the same young Tachigalia. And not only are
all the instars of Hunausibius smaller than those of Coccidotro-
phus but the colonies are also much less populous. The largest
I have seen consisted of less than a dozen beetles and not more
than two dozen larve. The habits, so far as I have been able
to observe them, are much like those of Coccidotrophus. The
Eunausibius also feed on the nutritive parenchyma in the walls
of the petiole but they do not dig long grooves in the tissue but
only narrow elongate pits, nor do they build up their frass in
parallel or vermiculate ridges but plaster it in a thin layer over
the walls of the petiole, so that the latter are smooth and even.
The elongate entrance to the petiole seems not to be provided
with a wall of frass. In one petiole I found that the pair of
parental beetles had entered through a large hole about 2.5 mm.
in diameter which had evidently been made by some larger insect.
The beetles had plugged the opening with frass, leaving a small
elliptical opening in the center just large enough to fit the head
of the beetle when acting as a sentinel. Coccids are found in
the elongated pits in the nutritive tissue but are few in number
and of small size, though the Hunausibius solicit and drink their
saccharine excretions in the same manner as Coccidotrophus. The
cocoons of Hunausibius, apart from their smaller size and some-
what more delicate walls, are very much like those of Coccido-
trophus and are, in all probability, constructed in the same man-
ner.

I have seen so few colonies of Hunausibius that I can give
no account of its enemies nor of those of its coccids. In all prob-
ability it is even less able than the more vigorous and more proli-
fic Coccidotrophus to withstand the insidious attacks of Solen-
opsis altinodis. The whole appearance of the beetle and its colon-

1921] Wheeler: Some Social Beetles 91

ies is that of a feeble, anemic and harried species on the verge
of extinction.

GENERAL CONSIDERATIONS

The behavior of the social Silvanids described in the pre-
ceding pages and the conditions under which they live are suf-
ficiently startling to stimulate reflection and a comparison with
other species of the same and allied families. Such comparison,
as an ethological method, has so often thrown light on what ap-
peared at first sight to be unique and incomprehensible instincts
and their settings that we may hope by resorting to it to trace
the peculiar conditions in Coccidotrophus and Ewunausibius to
simpler and more general phenomena. Since both the setting,
or environment and the responses, or behavior of the beetles are
rather complicated it will be best to consider them separately and
to begin with the setting, i.e. with the Tachigalia biocoenose.

The general ethological concept of the ‘biocoenose’ was, of
course, more or less clearly recognized by many of the early zoolo-
gists. Although the term seems to have been first used by Mo6-
bius (1877), even Réaumur had an inkling of the value of study-
ing insects in association with their host plants. He says: “I
would that the observers who busy themselves with the history
of insects gave catalogues of those that feed on every plant.”
In the middle of the last century Perris (1852-1862) devoted
many years to the study of the insects associated with the mari-
time pine and the chestnut in France, and Kaltenbach (1874)
attempted to list all the phytophagous insects of Germany accord-
ing to their host plants. In the United States Packard’s volume
(1881) on the forest and shade-tree insects and Mrs. Dimmock’s
paper (1885) on the insects of the birch represent more modest
studies of the same kind. Perhaps none of our entomologists
has been more thoroughly convinced of the advantages of study-
ing insects and other animals as components of biocoenotic com-
plexes than Forbes. Forty years ago he expressed his general
convictions on this subject in his paper on the food of fishes and
insects (1880) and he has returned to the subject in a recent
address (1915). His fine papers on the strawberry and maize
plants and their associated organisms (1884, 1894-1905) also
clearly illustrate the great value of biocoenotic investigations.

92 Zoologica: N. Y. Zoological Society Iss

Picard (1919) has recently published an interesting paper on the
insects of the fig in Southern France.

As a mere record of the insects associated with a tropical
plant my study of the Tachigalia biocoenose is necessarily frag-
mentary, owing to the few weeks I could devote to it, but it ac-
quires considerable interest from the fact that the Tachigalia is a
myrmecophyte, or one of those plants which are supposed to be
peculiarly adapted structurally to association with battalions of
protecting ants. The only organs which can be cited, however, as
such an adaptation aré the fusiform enlargements of the petioles,
which undoubtedly furnish excellent lodgings for all the various
ants, both inquiline and obligate. The plant is utilized also as a
source of food by the obligate species through the instrumental-
ity of the coccids, which are kept in the petioles and draw their
food by preference from the strands of nutritive parenchyma.
The beetles also use the petioles as lodgings and not only utilize
the species of coccid as a copious source of sugar and water but
also feed directly on the tissues of the plant. The plant is there-
fore more completely exploited by the beetles than by the ants and
might be said to be more perfectly adapted to the former than to
the latter. But the question as to whether the peculiar structure
of the petiole is really an adaptation to either of these groups of
insects is one which I shall leave to the botanist. Prof. Bailey
will no doubt deal with it in connection with the same probler
in the other South American myrmecophytes which he has inves-
tigated. That both the ants and the beetles have adapted them-
selves to the plant cannot be doubted and this adaptation, as I
have shown, is exhibited in three degrees, the inquiline ants
merely using the petiolar enlargements as lodgings, the obligate
ants as lodgings and through their herds of coccids as indirect
sources of food, and the beetles as lodgings and as both direct
and indirect sources of nutriment.

Of course, a particular biocoenose is not an isolated, per-
fectly self-contained association of organisms but shares some
of its components with other biocoenoses. Thus the Tachigalia
is part of a large association, or biocoenose of jungle trees grow-
ing under certain conditions of soil, humidity, light, tempera-
ture, ete. The Atta cephalotes, which occasionally defoliates the
young tree is the center of an elaborate biocoenose of its own,

1921] Wheeler: Some Social Beetles 93

comprising all the trees it habitually defoliates, its fungus gar-
dens, its myrmecophiles of the Blattid genus Atiaphila, the toads,
lizards and ant-eaters which feed on the foraging workers, the
Amphisbaenians which live in the penetralia of the huge nests,
etc. The two ants, Camponotus femoratus and Crematogaster
parabiotica, which attend Membracids on the young shoots of
the Tachigalia, are really characteristic members of the very
peculiar ‘“‘ant-garden” biocoenose, which I have described in an-
other paper (1921), and the Dolichoderus attelaboides belongs
to still another biocoenose of which many Melastomaceous plants
and their Membracid parasites are important components.

A particular biocoenose must also, of course, have a phylo-
genetic history, i.e. we must conceive it to have been gradually
built up, integrated and organized in time from components
which detached themselves from other biocoenoses and attached
themselves as satellites to an organism which furnished more
congenial conditions of life. Owing to the basic nutritive inter-
dependance of animals and plants, a particularly favorable plant
usually constitutes the primary focus of a biocoenose. The vari-
ous parasites, scavengers and synoeketes, which live with the
insects that immediately depend on this plant merely use the
former as so many secondary or tertiary foci. Thus in the
Tachigalia biocoenose the primary focus is the young plant
and the center of the focus the leaf-petiole, the secondary focus
is represented by the coccids and the tertiary foci by the ants
and beetles to the extent that they attract predators, parasites
and scavengers.

It is permissible, perhaps, to reconstruct the phylogenetic
sequences of the various organisms that have become associated
to form the Tachigalia biocoenose. Not improbably the tree, like
many other trees of the Neotropical jungle, was originally peo-
pled throughout its life by a certain number of miscellaneous,
inquiline ants. Among these were several species of Pseudo-
myrma and Azteca, both large genera comprising numerous forms
which still habitually inhabit any available hollow twigs or peti-
oles of the most diverse trees and shrubs. Later the number
of these ants was reduced, through the advent of the coccids and
their definitive association with the Tachigalia, to a very few
species, the putative ancestors of the present Ps. damnosa and

94 Zoologica: N. Y. Zoological Society [III;3

maligna and A. foveiceps, because the coccids enabled them to
acquire very intimate trophic relations to the plant. The coc-
cids present an unsolved problem in this connection. It would
seem that the Tachigalia must be their true host-plant, and that
the various other plants on which they are known to live, are
subsequent, or secondary hosts, possibly acquired when the
natives of British Guiana and of the surrounding countries took
to making clearings in the jungle and growing in them various
introduced plants such as pine-apples, Hibiscus, etc. The truth of
this statement can, of course, be established only by further in-
vestigation of Pseudococcus bromelix throughout its range.
When the obligatory ants had thus acquired their definitive at-
tachment to the tree, the miscellaneous inquilines necessarily
became restricted to its youngest stages since they were no
longer able to compete with the obligates for the possession of
nesting sites on the adult plant. The Silvanid beetles were prob-
ably relatively late intruders which found that they could inha-
bit the young tree for a considerable period before the queens of
the obligate ants had succeeded in maturing their broods of
belligerent workers. At first the beetles merely used the petiolar
cavities as lodgings and fed on the nutritive parenchyma in their
walls, but later they discovered the coccids and learned how to
obtain their honey dew and came to depend more and more on
this saccharine nutriment. The various parasites, scavengers,
etc., which infest the beetle colonies and their droves of coccids
obviously represent still more recent accessions to the biocoenose.
The other insects, such as Atta cephalotes, the Membracids and
their attendant ants, the caterpillars and gall-flies of the leaves,
etc., may belong to the ancient miscellaneous fauna which origin-
ally attacked or frequented the Tachigalia in all its stages, when
it was quite as “unprotected” as the great majority of jungle
plants.

Turning now to a consideration of the beetles themselves,
it would seem to be desirable to review their activities in the
light of what is known concerning the other members of the
natural family to which they belong. Here, however, we en-
counter difficulties, for the family Cucujide (sensu lato) has been
more neglected by taxonomists and students of insect behavior
alike than any other family of equal size in the order Coleoptera.

1921] Wheeler: Some Social Beetles 95

As understood by Coleopterists the family Cucujide belongs to
the huge and very inadequately analyzed Clavicorn complex of
families, but its characters are so striking that they have arrested
the attention of some of the specialists. Thus Leconte and Horn
(1883) long ago remarked: “This family is evidently an antique
and synthetic type, which exhibits alliances with both Hetero-
mera and Rhynchophora more than any other Clavicorn fam-
ily.” And Handlirsch (1908) says: “The family Cucujide,
which Ganglbaur places in the midst of typically Clavicorn
forms, exhibits many primitive characters and at the same time
high specialization. I do not believe that their antenne can be
derived from those of the Clavicorn type, although the Cucujids
agree with this group in the number of their Malpighian tubules
(six). Perhaps the Cucujids branched off very near the base
from the Cantharid stem, but possibly, and I regard this as
more probable, they form an independent series.” In his phyle-
tic tree (opposite p. 1278), Handlirsch therefore depicts the
family as arising from the Protopolyphaga as far back as the
beginning of the Coenozoic. Several species of Silvanus and one
of Passandra are, in fact, known from the Baltic Amber (Lower
Oligocene Tertiary), and Wickham (1920) cites Laemophloeus
vestitus Scudder from the Green River Eocene and three species
of Lithocoryne and a Pediacus from the Miocene of Florissant.
That the family must be an old one is indicated also by the

fact that New Zealand possesses some 20 indigenous species of
' Cucujide, distributed over 12 genera, mostly peculiar to the
islands, which are said to have been separated from Australia
during the Jurassic.

Kolbe (1910) is also of the opinion that the Cucujids are a
primitive group. He says: “The very lowly organized Cucu-
jids are not only in part characterized by a prothorax of very
primitive structure (as in the Adephaga) but primitively in-
serted (inframarginal) and primitively constructed (filiform or
moniliform) antenne.”’ Leng (1920) places the Cucujids in the
lower portion of the series of Clavicorn families, near the Rhizo-
phagide and Erotylide. In a brief study of the larve of Cucu-
jids, de Peyerimhoff (1902-’03) calls attention to their great
diversity and their resemblance on the one hand to the larve of
Cryptophagus among the Clavicorns and on the other to Pyro-

96 Zoologica: N. Y. Zoological Society [III ;3

chroa among the Heteromera. Hamilton (1886) long ago noticed
the resemblance of the larval Cucujus clavipes to the Pyrochroid
Dendroides canadensis larva. As shown in his very valuable
paper on the larve of Coccidotrophus and other genera (Zoo-
logica III, No. 7) Béving divides the Cucujide auctorwm into
four families, the Silvanide, Cucujide (sens. str.), Leemophleide
and Sealariide. The last of these he relegates to the group
Cleroidea, and states that they are closely connected with the
family Bothioderidz of Craighead (1920).

Apart from several species of considerable economic im-
portance, the little that is known concerning the habits of these
four families of beetles is scattered through the literature. Such
data as I have been able to glean in regard to the European,
North American and cosmopolitan species have been brought to-
gether in condensed form in Zoologica III, No. 5. From these data
it will be seen that the Cucujids, taken as a whole, exhibit certain
tendencies which are not without significance in connection with
the peculiar behavior of Coccidotrophus and Hunausibius. If we
exclude the Scalariide, which Fiske (1905) has shown to be
parasitic in their larval stages—resembling in this respect the
Bothrioderide—we notice that the various genera and many of
the species of the remaining families show an extraordinary
diversity, one might say versatility of behavior. They occur in
a great variety of habitats such as stored human foods of vege-
table origin, under bark, in decaying wood, in the burrows of
bark-beetles, under dead leaves and rubbish, and feed on all
sorts of substances mainly of a vegetable and especially of a
concentrated or highly nutritious character. Many of the spe-
cies are scavengers, others are undoubtedly predaceous and prey
on the larve of other insects. The adult beetles seem to be
rather long-lived and usually, if not always, live gregariously
with their larve, all the active stages feeding on the same sub-
stances. The developmental period is certainly very brief in
some species, as e.g. in Cathartus advena, the whole life-cycle
of which, from the egg to the imago may require only three weeks
and in the saw-toothed grain beetle (Oryzephilus surinamensis)
less than a month. As a rule both the beetles and the larve of
the vegetarian and detritivorous species are very tolerant of the
presence of other insects and actually seem to seek their compan-

1921] Wheeler: Some Social Beetles 97

ionship, especially when the food supply is abundant. Thus O.
surinamensis is often found living with the rice weevil (Cal-
andra oryz%), Cathartus advena with the Indian meal moth
(Plodia interpunctella) and O. mercator with a Tenebrionid
grain-beetle, Palorus subdepressus. Many of the species of
Lemophloeus constantly live in the burrows of Scolytid beetles
and feed on their dejecta. Owing to these peculiarities and espe-
cially to their very diverse and plastic feeding habits many of the
Silvanids and Lzemophloeids have become cosmopolitan house-
hold pests capable of doing considerable damage to many of the
staple stored foods of our own species.

If with this general complex of behavioristic tendencies
exhibited by the European and North American Cucujids (sens.
lat.) we compare the activities of Coccidotrophus and Hunausi-
bius, we find that the latter while retaining many of the ancient
and primitive family traits nevertheless exhibit several of them
in a peculiar and highly specialized form. Thus the merely gre-
garious habits of the adults and larve of the northern and cos-
mopolitan Cucujids have become more definitely social in the
Tachigalia beetles, and their toleration of alien insects has
increased; the feeding of the ancient Cucujids on various vege-
table substances has become specialized to the point of concen-
tration on a particular tissue of a particular plant and both
adult beetles and larvee have become coccidophilous. The con-
struction of the cocoon, too, exhibits peculiarities not found in
any other Cucujids. Owing to the unusual interest of these
various specializations they may be discussed at greater length
under separate captions.

1. SocIAL LIFE AMONG THE COLEOPTERA

If we regard as truly social only those insects in which the
parent or parents live with their offspring, protect them and
either feed them directly or prepare materials for their susten-
ance, there seem to be only three groups of beetles that meet these
requirements, namely, the Platypodidx, the Scolytide (Ipide)
and the Passalide. The Platypodide and that portion of the
family Seolytide, comprising, according to Hagedorn (1910) the
tribes Corythaline, Xyleborine and Spongiocerine#, with some
400 described species, mostly tropical, are commonly known as

98 Zoologica: N. Y. Zoological Society [IIl;3

ambrosia beetles, because like the Attiine ants of the Neotropical
Region and many Old World termites they cultivate fungi as
food for themselves and their larve. The remarkable social
organization and the food-fungi of these beetles have been stud-
ied by Hichhoff (1881), Hubbard (1897a, 1897b), Hopkins
(1898), Neger (1908a, 1908b, 1909, 1911), Schneider-Orelli
(191la, 1911b, 1912, 1913) and others, Hubbard’s account of
the habits of Platypus compositus of our Southern States is so
interesting that I quote it at length:

“These social instincts reach their highest development, ap-
parently in the genus Platypus. The species of this genus are
readily known by their very long cylindrical bodies, their promin-
ent head, flattened in front, the flattened and spur-tipped joint of
the front legs, and in the males the spine-like projections of the
wing cases behind. They are powerful excavators, generally
selecting the trunks of large trees and driving their galleries
deep into the heartwood. The female is frequently accompanied
by several males, and as they are savage fighters fierce sexual
contests take place, as a result of which the galleries are often
strewn with the fragments of the vanquished. The projecting
spines at the end of the wing cases are very effective weapons in
these fights. With their aid a beetle attacked in the rear can
make a good defence and frequently by a lucky stroke is able to
dislocate the outstretched neck of his enemy. The females pro-
duce from one hundred to two hundred elongate-oval pearly
white eggs, which they deposit in clusters of ten or twelve in the
galleries. The young require five or six weeks for their develop-
ment. They wander freely about in the passages and feed
in company upon the ambrosia which grows here and there upon
the walls. The chitinous ridges upon the thoracic segments, to-
gether with the row of tubercles upon the other segments, enable
the larva to move as rapidly through the galleries as if it were
possessed of well-formed legs. The mouthparts of the larva are
also provided with strong cutting mandibles, but the inner jaws
are not adapted to masticating hard food, such as particles of
wood. The older larve assist in excavating the galleries, but they
do not eat or swallow the wood. The larve of all ages are sur-
prisingly alert, active and intelligent. They exhibit curiosity
equally with the adults, and show evident regard for the eggs

1921] Wheeler: Some Social Beetles 99

and very tender young, which are scattered at random through
the passages, and might easily be destroyed by them in their
movements. If thrown into a panic, the young larve scurry
away with an undulating movement of their bodies, but the older
larve will frequently stop at the nearest intersecting passageway
to let the small fry pass, and show fight to cover their retreat.
When full grown the larva excavates a cell, or chamber, into
which it retires to undergo its transformations. The pupa cells
are cut parallel with the grain of the wood and generally occur
in groups of eight to twelve along some of the deeper passages.
The older portions of the galleries are blackened by the long-
continued formation of the food fungus. In the ambrosia of
Platypus compositus the terminal cells are hemispherical, and
are borne in clusters upon branching stems.”

The habits of several genera of ambrosia beetles of the
family Scolytide were investigated by Hubbard, and one of our
species, Xyleborus xylographus, which has an extensive circum-
polar distribution and is common in the wood of fruit-trees, has
also been studied by Eichhoff (1881) and Hopkins (1898). I
quote the latter’s account of this insect which may serve as a
paradigm of the whole group:

“The fertilized females pass the winter in their brood cham-
bers and emerge in the spring (April and May, near Morgan-
town, W. Va.). They are then attracted to sickly, dying or
felled trees, in the living or moist dead wood of which they
prefer to excavate their brood galleries. A crevice or opening in
the bark, such as may be made by other insects, or, as I have
observed, those made by the yellow-bellied woodpecker, but more
commonly the edge of a wound, in a dead place on a living tree,
is selected as a favorite point of attack. Here a female will
commence the excavation of a mine, and after she has penetrated
the wood a short distance, another female (as I have observed)
will come to her assistance, one working at the excavation, while
the other guards the entrance and assists in expelling the borings.
The primary or main gallery is usually extended into the heart-
wood before eggs are deposited. When the primary gallery is
completed (according to Hubbard) a bed is provided on the
sides of the gallery for the propagation of the special species or
variety of ambrosia fungus which is to furnish food for the

100 Zoologica: N. Y. Zoological Society Pontes

future broods. The first set of eggs are few in number (five to
ten) and are placed without any protection on the sides near
the end of the main gallery, or in cavities or short branching
galleries, one-half to one inch from the end, where upon hatch-
ing, the young larve find a supply of ambrosial food. After the
first set of larve have attained considerable size, another set of
eggs are deposited, and so on at intervals until a large family
is reared, in which eggs, larve of all stages of development,
pup, and young and old adults are found crowded promiscu-
ously in leaf-like brood-chambers which are continually broad-
ened or extended by the adults and possibly by the larve, to make
room for the increase. It appears that the brood-chambers are
broadened and extended by the adults, and that the borings,
mixed with the fungus, are softened and furnish additional food
for the larve and young beetles.” At this point in his account
Hopkins introduces the following note: “In a brood-chamber
before me just cut from a nearby apple tree, I find a pupa minus
an abdomen. No predaceous enemies can be found, but two or
three half-grown larve are in such a position as to make the
circumstantial evidence quite plain that they are to blame for
the multilation. The remaining portion of the pupa is in a
normal condition, which would indicate that the attack had been
recent and when the victim was alive. This would also indicate
that the helpless pupze may furnish food for the larva in case
of a scarcity of ambrosia, or that they may be thus disposed
of to prevent an overcrowded brood-chamber.”

The account of Xyleborus continues: ‘Mr. Hubbard records
the discovery of a death chamber, or a kind of catacomb, in
which the dead mother beetles and other dead friends or foes
of a large colony are consigned by the survivors. In some fresh
specimens of galleries before me I find the same thing, but it
appears that in addition to a resting place for the dead, it is also
utilized for the disposal of all objectional and refuse matter,
which owing to the crowded condition of the chamber, cannot
be conveniently expelled from the entrance. One of the males
found in this set of chambers was excavating a burrow in the
mass of material in the death or garbage chamber. Whether he
was excavating his own tomb, or simply providing bachelor
quarters, I cannot say. The proportion of males in this, as in

1921] Wheeler: Some Social Beetles 101

all other species of the genus Xyleborus, is remarkably small.
There are usually not more than three males in the largest
colonies, or groups of brood-chambers. It would appear from
observations made by Swiner and Hichhoff in Germany, and
the numerous colonies I have examined in this country that there
is, on an average, about one male to twenty females. The males
have no wings, therefore probably do not leave the brood-cham-
bers, but remain with the over-wintering colony until all have
emerged in the spring. They are then left to be smothered in
overabundant ambrosial food, or to the tender mercies of preda-
tory insect enemies which had previously been prevented from
entering the brood-chambers by one or more female sentinels
at the entrance. A few females may emerge from time to time
during the summer to start new colonies, but from the excessively
crowded condition of the brood-chambers during the fall and
winter months, it would appear that the older adults of the

broods excavate branching chambers in which new broods are
- developed, and that in these old and new chambers they pass
the winter.”

The third group of social beetles comprise the large
Lamellicorns of the family Passalide, abundant in the tropics
of both hemispheres but represented in the United States by
only a single species, Passalus cornutus Fabr., which ranges as
far north as Massachusetts and Illinois. None of our Coleop-
terists seem to have taken the trouble to study the habits of this
common and conspicuous insect, so that it was left to Ohaus
(1899-1900, 1909) to discover the social behavior in certain
Brazilian species. He found that they live in rotten logs in
colonies, each consisting of an adult male and female with their
larve. The beetles excavate spacious galleries, comminuting
the wood and probably treating the particles with some digestive
enzyme, so that they can be eaten by the larve, which slowly
follow along the galleries just behind their tunneling parents.
Owing to the structure of their mouthparts the larve are quite
unable to break down the wood, and when removed from their
parents soon die. The beetles not only guard their greenish eggs
and diligently provide food for their larvee, but also protect the
pupz and feed the imaginal young till their chitinous integument
is completely hardened. In a former paper (Wheeler and Bailey,

102 Zoologica: N. Y. Zoological Society [1II;3

1920) I have published an account of the stridulatory organs of
the larval and adult Passalus and have given reasons for believing
that all the members of colony are kept together by the shrill
sounds they are able to emit.

During the summer of 1920 while in Trinidad and British
Guiana my son Ralph and I made a few observations on severai
of the species of Passalus which are very common in rotten logs
throughout the jungle. Just under the bark the beetles make
large, flat cavities, which are later very often occupied by the
fungus-growing ants of the genera Apterostigma, Myrmicocrypta
and Cyphomyrmex, and evidently furnish just the right places
for their more or less globular gardens. In each of the Passalus
colonies examined during July and August there were only two
adult beetles, usually accompanied by a troop of larve varying
little in size and evidently belonging to a single brood. In one
log, however, we found a pair of the beetles guarding a batch of
about 40 large, olive-green, broadly elliptical eggs, some of which
had just hatched. The young larve closely resembled the older
individuals in the structure of the peculiarly modified paw-like
metathoracic legs, which are rubbed over the finely ridged middle
coxe during stridulation, but the hairs on the body were conspic-
uously longer and coarser. Our observations on the beetles and
their larve both in the field and in the laboratory, confirm the
statements of Ohaus.

The preceding account of the Platypodids, Scolytids and
Passalids will suffice to show that they have reached a more
advanced stage of social development than Coccidotrophus and
Eunausibius, though the latter exhibit certain interesting resem-
blances to such ambrosia beetles as Xyloterus. The two Silvanids
really represent a stage in social development intermediate
between that of the families mentioned and the merely gregarious
Silvanus, Oryzxephilus, Nausibius ete. Although the colonies of
Coccidotrophus and Eunausibius are founded by pairs of parent
beetles and in the climax stage of their development may comprise
a considerable number of offspring in all stages of development, yet
the latter do not seem to be the recipients of any special care on
the part of the parents, unless we interpret as such the guarding
of the petiolar cavity and the deepening of the food-grooves which
would seem to render the nutritive parenchyma more accessible

1921] Wheeler: Some Social Beetles 103

to the larve. And although all the members of the colony, beetles
and larve alike, seem to be very indifferent to one another, except
when they are competing for the honey-dew of the same coccid
or when larve occupy cocoons in process of construction by other
larve, yet under normal conditions there are no signs of hostility
on the part of the beetles and larve even when other individuals
are very annoying. Moreover, the use of the petiolar cavity as
a common domicile, with its kitchenmiddens and more or less
definite arrangement of the frass-ridges and wall about the
entrance, the droves of coccids and the definite orientation of the
eggs and cocoons, all show a much more socialized condition than
anything that has been hitherto observed in other Cucujids. I
believe, therefore, that I am justified in regarding the two
Tachigalia Silvanids as representing a fourth group of social
beetles, of a more primitive type than any of the three families
above considered and differing in the absence of any definite
preparation of larval food by the parents. No such preparation
is necessary, in fact, owing to the peculiar conditions under which
the Coccidotrophus and Eunausibius live, since. both the young
and the adults feed on the same substances and these are
furnished by the plant and the coccids, which in turn feed on the
same specialized parenchyma as the beetles.

No doubt the toleration by the beetles and larvze of such
different insects as the coccids, the Scymnus larve, the larval and
adult Diadiplosis, the Entomobrya, Aphiochaeta and probably
also of the adult Blepyrus, is due to the same causes as the tolera-
tion by so many ants and termites of numerous myrmecophiles
and termitophiles. Such guests, parasites and synoeketes can, of
course, manage to live only among insects which through long
association with individuals of their own species have come to
tolerate or even to seek the presence of insects belonging to alien,
or unrelated species.

2.—THE DEVELOPMENT OF THE FEEDING-HABITS OF
THE SOCIAL SILVANIDS.

There would seem to be little doubt that the primitive food
of Coccidotrophus and Eunausibius is the nutritive parenchyma
of the Tachigalia petioles. But this is a very specialized diet

104 Zoologica: N. Y. Zoological Society [aut

compared with that of other Silvanide. Although such genera as
Silvanus, Oryzephilus, Cathartus and Nausibius eat by prefer-
ence vegetable substances with high protein, starch or sugar con-
tent, none of these forms is known to devour the tissue of grow-
ing plants, and it is even doubtful whether White’s statement
* (1872) that the Cucujid Dendrophagus crenatus feeds on the in-
ner bark of conifers, is correct (see Zoologica III, No. 5). The
Tachigalia beetles are primarily attracted by the tree and there
is every reason to regard this peculiar Leguminous plant as
their only host, so that in this respect, also, they are highly
specialized, for while some of the Silvanids, Lemophloeids and
Cucujids (sens. stv.) prefer particular trees, they seem never-
theless to thrive equally well in trees of different species,
probably because they do not eat the living plant tissues
but merely require special moisture conditions or the presence
of certain other insects. We must .suppose, furthermore,
that the social Silvanids are primarily attracted to the Tachi-
galia by certain chemical substances in the petioles, a suppo-
sition which seems to offer the only satisfactory explanation, as
Picard (1919) has shown, for the selection of particular host
plants by particular insects. We should have to suppose also,
that the beetles can discriminate between the petiolar substances
of young shade—and older sun-trees, since they confine their
attentions to the former. This is not surprising when we consider
that some insects, e.g., certain Cynipid gall-flies exhibit even
more delicate powers of discrimination since, when ovipositing,
they seem to be able to distinguish between the viability of
different buds or leaves on the same branch.

An even more interesting problem is presented by the
coccidophily of Coccidotrophus and Hunausibius, for nothing like
it has been observed in any other beetles, and apart from the
ants, few insects are known to have developed the ability to
solicit honey-dew from any of the Homoptera. I find only the
two following cases in the literature, the first an observation by
Belt in his “Naturalist in Nicaragua” (1884, p. 228), on wasps
attending Membracids: “Similarly as, on the savannahs, I had
observed a wasp attending the honey-glands of the bull’s horn
acacia along with the ants, so at Santo Domingo another wasp,
belonging to quite a different genus (Nectarinia), attended some

1921] Wheeler: Some Social Beetles 105

of the clusters of frog-hoppers, and for the possession of others
a constant skirmishing was going on. The wasp stroked the
young hoppers, and sipped up the honey when it was exuded,
just like the ants. When an ant came up to a cluster of leaf-
hoppers attended by a wasp, the latter would not attempt to
grapple with its rival on the leaf, but would fly off and hover
over the ant; then when its little foe was well exposed, it would
dart at it and strike it to the ground. The action was so quick
that I could not determine whether it struck with its fore-feet or
its Jaws, but I think it was with the feet. I often saw a wasp
trying to clear a leaf from ants that were already in full posses-
sion of a cluster of leaf-hoppers. It would sometimes have to
strike three or four times at an ant before it made it quit its
hold and fall. At other times one ant after the other would be
struck off with great celerity and ease, and I fancied that some
wasps were cleverer than others. In those cases where it
succeeded in clearing the leaf, it was never left long in peace.
Fresh relays of ants were continually arriving, and generally
tired the wasps out. It would never wait for an ant to get near
it, doubtless knowing well that if its little rival once fastened
on its leg, it would be a difficult matter to get rid of it again.
If a wasp first obtained possession, it was able to keep it; for the
first ants that came up were only pioneers, and by knocking
these off, it prevented them from returning and scenting the
trail to communicate the intelligence to others.”

The second case is more remarkable and refers to a Gerydine
Lycenid butterfly of India, described by Bingham (1907, p. 287) :
“A remarkable habit in one member of the subfamily, viz.,
Allotinus horsfieldi, has been communicated to me by Colonel
H. J. Barrow, R. A. M.C. He writes: ‘I don’t know whether you
have observed the habits of a small plain butterfly which I caught
in Maymyo. I watched it often in the jungle, sometimes for an
hour at a time. It puzzled me at first to know why it took such
an immense time to settle. It would keep within one yard of a
spot and almost settle, twenty times perhaps, before it actually
did. Its legs are immensely long and I discovered why. It settles
over a mass of Aphides and then tickles them with its proboscis,
just as ants do with their antennz and seems to feed on their
exudations.’ The butterfly would settle over rather large ants

106 Zoologica: N. Y. Zoological Society [Ill ;3

that were attending the aphids ‘and did not mind one or two
actually standing up and examining its legs to see who was
there. The ants did not attack it in any way.’ ”

A comparison of the behavior of the insects considered in
the foregoing paragraphs is very instructive. The predilection
of ants for various Homoptera (aphids, coccids, membracids,
cercopids and psyllids) is well known. Though never observed
among the predatory Doryline and Cerapachyine and rare among
the Ponerine and Pseudomyrmine this predilection is, neverthe-
less, so prevalent among the higher subfamilies (Myrmicine,
Dolichoderine and Formicine) that it has not escaped the most
casual observer. When this type of behavior is highly developed,
as in our species of Lasius, the ants display not only an exquisite
deftness in stroking their trophobionts but also a decidedly pro-
prietary interest in them, most clearly evinced by building
peculiar carton or earthen shelters over them, aggressively
defending them from their foes, or even collecting them and
their eggs in the nests, distributing them over the surfaces of
suitable plants and conveying them to places of safety when the
colonies are disturbed. The whole performance is so elabo-
rately adaptive as to suggest on the part of the ants an intimate
acquaintance with the requirements and habits of their wards.
This is also indicated when the latter fail to respond to stroking,
for the ants do not wear themselves out by prolonged solicitation
after their cattle have discharged their honey-dew, but stand
around as if waiting for more of the saccharine liquid to
accumulate.

The wasp described by Belt and the butterfly described by
Bingham are really robbers, the former having learned to dispos-
sess the ants of their wards, at least temporarily, the latter to
overreach the ants and obtain the honey-dew by stealth. Though
very different, the relations of the beetles to their coccids are no
less extraordinary. The case is, indeed, so far as known, unique
among the Coleoptera. Unlike many species of ants, the beetles
have not yet learned to pick up the coccids and carry them about,
but merely accept them as an integral part of the normal environ-
ment or as members of the colony. The fact that the beetles
clearly recognize the signal of the prospective emission of the
honey-dew by the coccids, when they raise their caudal segments,

1921] Wheeler: Some Social Beetles 107

and the fact that they devote more attention to the posterior than
to the anterior end of the larger coccids, implies a delicate
discrimination, because both ends of the coccid’s body are so very
much alike. Furthermore, the beetle’s antenne and mouthparts
seem to be so clearly adapted to dealing with the coccids as to
indicate that the trophobiotic relations between the two species
have been in existence for a very long time. The beetle’s extraor-
dinary perseverence in stroking individual coccids after they
have been exhausted by repeated emissions of honey-dew might
be interpreted either as a very thorough and hence highly
adaptive method of exploiting the coccids, or as due to a very
imperfect discernment of their physiological peculiarities.

Obviously the most remarkable item of behavior in
Coccidotrophus and Eunausibius is the stroking of the coccids
by the larve of all stages, as well as by the adult beetles. No
one has even considered the possibility of a similar performance
by the larve of ants, wasps or butterflies, since it is difficult
to imagine creatures more unfitted for such behavior. Neverthe-
less, Mr. W. F. Fiske informs me that while he was investigating
certain injurious insects in British East Africa, he saw small
worker ants climbing a tree with their larve and holding them
to the posterior ends of aphids, so that they could feed on the
honey-dew voided in response to the antennal solicitations of their
nurses! Unfortunately no specimens of the ants were preserved,
and from Mr. Fiske’s description I am unable to determine even
the subfamily to which they belonged, but I have no reason to
doubt the statement of an entomologist so competent and so
keenly observant. I surmise that the ant must have been some
Myrmicine which is unable to feed its larve by regurgitation, as
otherwise such behavior would be superfluous.

So specialized a habit as the coccidophily of the two genera
of social Silvanids calls for some consideration of its possible
phylogenetic origin. Under existing conditions, the beetles either
find the coccids already established in petioles that have been
previously inhabited by other beetle colonies or by other insects,
or the coecids enter the young petioles just after they have been
perforated by the beetles, for insects with sucking mouthparts
cannot, of course, gain access to the cavities in any other way.
That they migrate into the petioles as very young individuals is

108 Zoologica: N. Y. Zoological Society (it ;3

certain. Later, after completing their growth, they are often
too bulky to escape through the entrances made by the beetles,
and as such imprisoned coccids contain eggs, they might be
supposed to breed in the petioles. I have never been able to find
either the males or the deposited eggs in the petioles, and as the
total number of coccids in a petiole is too small to indicate the
survival of many of the young, I suspect that the beetles, though
averse to devouring the young or mature coccids, nevertheless
consume many of the eggs. How the coccids manage in the first
place to reach the individual Tachigalia plant is a problem which
presents itself also in the case of any of the other often widely
distributed species of the family. That ants have much to do
with carrying certain species of coccids to their host-plants is
very probable. The only other active agents in such distribution
would seem to be birds or the wind.

Soon after they enter the petioles the coccids seek out the
strands of nutritive parenchyma, and sink their slender beaks
into the tissue. And as the beetles keep gnawing at the same
strands, grooves or narrow depressions are soon made in which
the coccids settle, one behind the other, in rows, with the long
axes of their bodies parallel with the long axis of the petiole.
Thus the coccids naturally and inevitably come to lie in the paths
of the feeding beetles, so that these can hardly avoid continual
contact with the waxy creatures and their excrement. Probably
at first the coccids simply voided their excrement in the grooves,
thus drenching the surface of the nutritive parenchyma, so that
the beetles found their bread spread with syrup. But this could
hardly be an unalloyed blessing, because a sticky liquid spread
on the walls of, the petiole would almost certainly be injurious
or fatal to the eggs, pups and younger larve of the beetles. We
may therefore conjecture that the latter soon learned to stroke
the coccids and to swallow the honey-dew at its very source, and
that they have even acquired so keen an appetite for the liquid
that it has now become a very important if not an essential
constituent of their diet. The exploitation of such a constant
and energizing supply of syrup, moreover, would surely tend not
only to lengthen the original life-span of the adult beetles but
also to increase the number of their progeny and hence the size
and vigor of their colonies. This is suggested by the conditions

1921] Wheeler: Some Social Beetles 109

in certain ants, e. g., our yellow, hypogeic species of Lasius,
which are able to develop populous colonies mainly, if not exclu-
sively, on a diet of honey-dew derived from root-aphids and root-
coccids.

Inasmuch as the larve, from their very youngest stages,
no less than the adult beetles, continually stroke the coccids, the
question arises as to whether the habit was first acquired by the
larve, or by the beetles, or whether both instars developed it
simultaneously. An answer to this question might, perhaps, be
forthcoming if we could determine whether the larva or the
adult beetle shows the greater structural adaptation of the
antennze and mouthparts to dealing with the coccids. The larve
as I have shown, use both their antennze and maxille in the
process, the adults only the antenne, but although the beetles
are able to cover a greater area of the coccid’s surface with their
antennal clubs, the larger larve at least, by combining both pairs
of organs, can probably produce a stimulus no less intense and
effective. The larve are certainly more alert; restless and
inquisitive than the beetles and the mandibles in the youngest
stages seem to be very poorly adapted to feeding on the nutri-
tive parenchyma. It is therefore quite as probable that the habit
of stroking the coccids was first acquired by the young larve
and later continued in the adult as that it originally appeared in
the latter and was inherited in earlier and earlier ontogenetic
stages till it came to be manifested by the just-hatched larva
less than a millimeter in length. As shown by the observations
recorded on p. 70 such larve show an even greater avidity for
the honey-dew than the adult beetles. Since the stroking of the
female beetle by the male during the courtship is precisely like
the stroking of the coccids, we might be tempted to conjecture
that the habit had arisen first in the adult as a modification of
the sexual appetite, but this is, perhaps, rather far-fetched.

3.—THE BUILDING OF THE COCOON.

The few data I have been able to gather concerning the
processes accompanying pupation in the beetles allied to Cocci-
dotrophus and Hunausibius are reproduced in Zoologica III, No. 5.
The pupating larvae of some Lemophloeids, Silvanids and Cucu-

110 Zoologica: N. Y. Zoological Society uu

jids are described as attaching themselves by the last segment to
the substratum, and pushing back the larval skin to the tip of
the abdomen which remains fixed to the exuvium. In OryZze-
philus surinamensis the larva before pupating makes a rude
cocoon by agglutinating particles of food or detritus with an
oral secretion, but I have seen no circumstantial account of the
process which may be of considerable interest in connection with
the cocoon-building of Coccidotrophus. The Cucujus larva also
makes a rude cocoon (see p. 177).

The construction of a substantial cocoon by the larve of
the social Silvanids would seem to be necessary, because a nude
pupa, even if attached to the petiolar wall by its anal end, would
be exposed to injury by the numerous beetles and larve moving
about in the narrow cavity. But the way in which the cocoon is
constructed is, to say the least, very unusual. So far as known,
the larve of Coleoptera and other insects, when engaged in
making such structures, remain in situ and build the cocoon as
an envelope around the body, using for the purpose extraneous
particles of earth, detritus, wood or frass, or threads of silk
spun from the sericteries or more rarely a secretion of the
Malpighian tubules, as in ant-lions and certain weevils, as
described by Knab (1915a, 1915b) and others, or several secre-
tions as described by Boving for Donacia (1910). The Coccido-
trophus larva, however, laboriously collects minute particles of ~
living plant-tissue, mixes them with saliva and builds them up
in a very definite manner, repeatedly leaving the structure to
go afield for the purpose of collecting the necessary materials.
When the cocoon is all but completed the larva enters, and
becoming a voluntary prisoner, closes the aperture at the end with
materials scraped from the inner surfaces of the walls. I have
failed to find in the entomological literature any account of such
a method of cocoon-building, which in many particulars resembles
the nest-building of certain birds and rodents.

The only suggestion I can make in regard to the possible
origin of this behavior is that it may be derived in some way
from the beetle’s habit of building up its frass in more or less
regular ridges, or welts between the food-grooves or immediately
around the entrances to the petioles. I have been unable actually

1921] Wheeler: Some Social Beetles 111

to witness this performance as the beetles never exhibited it
after the petioles were cut open. I am inclined to believe that
the feces are not simply voided in the spaces between the food-
grooves but actually built up with the aid of the mouthparts.
This seems to be clearly indicated by the circular wall around
the entrances (Fig. 11). Perhaps the larve have quite as much
to do with the construction of the ridges as the beetles.

4.—CONCLUDING REMARKS ON THE BEHAVIOR OF THE
SOCIAL SILVANIDS

All modern observers of insects have been deeply impressed
by the highly mechanized character of their behavior, but it is
equally true that those who have most closely studied these
organisms both under natural and experimental conditions have
failed to find that the behavior of any one of them can be com-
pletely reduced to a rigid system of automatic or stereotyped
reactions. While the behavior of certain forms such as the
larval ant-lion, according to Doflein (1916) or the larval worm-
lion (Vermileo), as shown by my unpublished studies, seems to
consist almost entirely of a small number of reflexes, the behavior
of other insects, such as the solitary wasps, termites and social
Hymenoptera, often exhibits considerable plasticity, modifiability
or adaptability. Between these extremes we find the majority
of insects with a certain modicum of the latter type of behavior.
To this group we may assign the social Silvanids. The interpre-
tation of their various activities necessarily involves some refer-
ence to the behavior of insects in general and the assumption
of a definite attitude towards certain intricate and much dis-
cussed questions. The limitations of space compel me, therefore,
either to leave the whole matter unconsidered or to treat it in a
very brief and sketchy manner. I prefer to adopt the latter
course.

The fashion which required one to explain as much as
possible of the behavior of an insect in terms of tropisms, or
taxes, and measured the value of one’s work by the success
achieved in the endeavor, seems to be rapidly passing. Thirty
years ago I followed the fashion with some enthusiasm, but
continued observation of the ways of insects has made me very

112 Zoologica: N. Y. Zoological Society [Ill s3

dubious in regard to the whole subject of the tropisms. My
present position concerning them is not essentially different
from that of Jennings (1904, 1906, 1909), von Buddenbrock
(1915), Claparede (1912, 1913), and others. I should, therefore,
interpret them as adaptive, secondarily developed reflexes and
not as unique, primitive elements in the genesis of instinctive
behavior.

There are, nevertheless, in the behavior of Coccidotrophus
certain phenomena, which some might be inclined to interpret
as tropisms, especially the reactions to contact, light and chemical
stimuli. The reader who. has followed my account of the beetle
will have noticed the peculiar orientation of some of its stages
and of some of the associated insects with respect to the walls
of the long fusiform petiolar cavity which they inhabit. Thus
the eggs, cocoons and pup of the beetle are always placed with
their long axes parallel with the long axis of the petiole, and
the coccids, while feeding, the cocoons of Blepyrus, which are
formed within the bodies of the coccids, and the nude pupe of
Scymnus assume the same orientation. The food grooves which
are excavated by the beetles and the frass-ridges which they
build, as well as the longer axis of the entrance are all longitu-
dinal. Moreover, the beetles spend much time lying in the food-
grooves with their narrow bodies longitudinally oriented and
with as much as possible of their surface in contact with the
floor and walls of the grooves.

At first sight this striking series of orientations would seem
to be best described as tropistic, perhaps as due to some form
of thigmotropism, but it is evident that the only behavior which
might be legitimately regarded as such is that of the adult beetie
when resting or moving in the food grooves. The Collembolans
and the larvee of Scymnus and Diadiplosis exhibit not the slightest
tendency to assume a similar orientation, and the Coccidotrophus
larve show no traces of it till they start their cocoons. Even
then, though they orient their cocoons, they assume a position
with their long axes parallel with the long axis of the petiole
only while actually adding particles to the walls and after they
have pupated. It is evident, furthermore, that nearly all the
orientations mentioned can be traced more or less directly to
peculiarities in the structure of the petiole, i.e., to the shape of

1921] Wheeler: Some Social Beetles 113

its cavity and the histological structure of its walls, and especialiy
to the singular arrangement of the nutritive parenchyma in long,
narrow strands. The beetles gnaw these out and thus form
grooves, which in turn orient the coccids and their internal para-
sites. The orientation of the frass ridges and beetle cocoons is
also determined by the grooves, and the position of the Coccido-
trophus eggs is very probably due to their being laid and attached
by the beetles while they are lying in the depressions between
the frass ridges. Thus the various orientations are merely so
many direct or indirect adaptations to the nutritive parenchyma
and the long narrow petiolar cavity. The latter clearly deter-
mines to some extent the longitudinal arrangement of the bulky
cocoons, just as a long Pullman car makes it advisable for us
to arrange the berths in a similar manner.

One orientation, that of the entrance, is not so easily
explained. In all the petioles I have examined, the long axis of
the entrance is parallel with the long axis of the petiole. It is
evident that it precisely fits the head of the beetle and that the
latter while gnawing it must stand on the outer surface of the
petiole at right angles to its long axis. When the surface is
longitudinally grooved, as is sometimes the case, the entrance can,
of course, have no other orientation, but often the surface is
quite smooth and it is difficult to see why it should be easier for
the insect to gnaw through the tissue lengthwise rather than
crosswise of the grain. It is also evident that while guarding
the entrance the beetle has its long axis at right angles to the
long axis of the petiole, but since a tubular wall is built around
the inside of the orifice (Fig.11) the insect’s reaction might be
regarded as thigmotropic, like its reactions to the walls of the
food grooves.

Nevertheless, I am convinced that the responses of the adult
beetle to contact stimuli may be more properly interpreted as
typical and highly adaptive reflexes. This is clearly indicated
by the structure of the insect. The long, parallel-sided, sub-
cylindrical form of the body and the shortness of the legs are
merely so many adaptations to living in narrow tubular cavities.
The same type of structure reappears as an independent develop-
ment in each of many different families of beetles, which live
in cylindrical cavities or burrows, e.g., the Scolytide, Platypodi-

114 Zoologica: N. Y. Zoological Society [Il1;3

dz, Brenthide, Bostrychide, Buprestide (Agrilus), Cleride,
Trogositide, Histeride (Teretrichus), Colydiide (Colydium),
Lyctide, Lucanide (Ceruchus), Elateride, Parandride, many
Cerambycide, Lymexylondide, etc. This type of body and one
more extremely flattened and adapted for living under bark
(Cucujus, Brontes), are very common among the Cucujide sens.
lat. The peculiar conformation of the front and of the mand-
ibles of Coccidotrophus and Hunausibius, so strikingly like that
of many ants (Cryptocerus, Cataulacus, Colobopsis, etc.) is, more-
over, a definite adaptation to guarding elliptical or circular en-
trances to solid-walled nesting cavities. It is interesting to note
also that the general shape of the body of Coccidotrophus and
Eunausibius reappears in several genera of ants which regularly
live in narrow plant cavities, e.g., Colobopsis, Simopone, Cylin-
dromyrmex, Metapone, Pachysima, Tetraponera and Pseudo-
myrma (see Plate III, figs. 1 and 2). Even the larve and pup of
these ants have assumed a similar form. (See Wheeler, 1918, and
Wheeler and Bailey, 1920). We may say, therefore, that the
whole general bodily structure of the various insects I have
mentioned has been adaptively modified during their phyloge-
netic history and that such a modification can hardly be any-
thing but an expression of a concomitant adaptation of their
nervous system and reflexes.

Equally unsatisfactory from an ethological point of view
is the reference of other behavioristic peculiarities of the social
Silvanids to simple tropisms. Let us take as an example the
attraction to the Tachigalia. I agree with Picard (1919) in his
contention that phytophagous insects are attracted to their
respective host-plants by particular chemical substances in the
latter. Entomologists have always believed this but have usually
described the phenomena as due to “odor” or “taste.” To desig-
nate them as chemotropism really adds nothing to our knowledge
but a technical term. The ethological question remains: Why
is a particular insect species or sex attracted to a particular part
of a particular species of host plant, or, in the case of the social
Silvanids, why do they fly to and bore into the swellings of the
petioles of young individuals of a certain species of Tachigalia?
Undoubtedly the exquisite sense-organs in their antennal clubs
enable the beetles to detect certain very delicate effluvie emanat-

1921] Wheeler: Some Social Beetles 115

ing from the young Tachigalia as a whole or perhaps even from
the nutritive parenchyma in its petioles, but the attraction of
these odors is probably due to their having acquired a “meaning”
for the beetles, because the latter throughout their larval and
early imaginal stages fed on these very substances, and had, in
fact, long been familiar with the petiolar cavities, the coccids, ete.
The latter part of this statement also applies to the young queens
of the obligate Tachigalia ants of the genera Pseudomyrma and
Azteca and would account for the rather unusual attachment of
these insects to a definite host-tree. Hence in these cases organic
memory, or “mneme,” or even individual memory yields a more
satisfactory explanation of the phenomena than a naked tropism.

A consideration of the responses to light leads to similar
results. The beetle colonies, as we have seen-are “‘photophobiec,”’
or live in the dark, and when the petioles are opened in the light
the insects are at first much agitated but soon settle down and
continue the regular routine of their existence even when the
pieces of petiole are kept exposed to artificial or diffuse day-light
in the laboratory. The adult beetles have well-developed eyes
and the larve have three pairs of small simple eyes on each side
of the head. (Plate VII, fig. 2, Plate VIII, fig. 10). And since a
certain amount of light enters the petiole through the entrance, at
least when it happens to be unguarded, it is probable that under
ordinary conditions the eyes are mainly useful in enabling the in-
sects and particularly the larve, to stay in the dark. There is noth-
ing to show that the young beetles, which leave the petioles to
establish new colonies, do so because they become positively
phototropic. The emigration may, perhaps, take place at night
and even if it occurs during the day the light in the parts of
the jungle where the young Tachigalias are growing, is very
subdued. I believe that the young beetles must emigrate either
because the space in the petiole has become too greatly reduced
by the growth of the colony or because the food-supply has for
the same reason become insufficient. In either case emigration
would be due to internal stimuli (‘‘physiological states”). These
may also be important factors in the swarming of ants, bees and
termites. We might, perhaps, even suppose that the guarding
of the entrance by the beetles is due to an abortive or inhibited
emigration impulse due to feebler or vaguer internal stimuli

116 Zoologica: N. Y. Zoological Society platings s;

comparable with those which sometimes impel a person who has
been sitting for hours in a dimly-lighted room, to stand at the
window or open door or to step out into the sunlight.

Thus while it is easy to interpret many of the activities of
the social Silvanids as adaptive reflexes it is difficult to assign to
such stimuli as light, contact, chemicals, etc., the leading réle
which they have in the theories of Loeb and Bohn. Much of the
behavior of the beetles, such as their treatment of the coccids, the
building of the cocoon, mating, the guarding of the entrance,
etc., so obviously depends on internal or physiological states that,
in so far it has not become completely mechanized, we may more
properly regard it as made up of cyclical activities like those
recognized by Herrick (1910), Craig (1918) and others in birds.
Craig especially has shown how much of the behavior of birds
can be interpreted as cycles of appetence, or appetite and aver-
sion, which he defines as follows: ‘An appetite (or appetence,
if this term may be used with purely behavioristic meaning),
so far as externally observable, is a state of agitation which
continues so long as a certain stimulus, which may be called the
appeted stimulus, is absent. When the appeted stimulus is at
length received it stimulates a consummatory reaction, after
which the appetitive behavior ceases and is succeeded by a state
of relative rest. An aversion is a state of agitation which con-
tinues so long as a certain stimulus, referred to as the disturbing
stimulus, is present; but which ceases, being replaced by a state
of relative rest, when that stimulus has ceased to act on the
sense-organs.”

Rignano (1920) gives this same conception a more general
physiological formulation in the following passage: “Every
organism is a physiological system in a stationary condition and
tends to preserve this condition or to restore it as soon as it is
disturbed by any variation occurring within or without the
organism. This property constitutes the foundation and essence
of all “needs,” of all “desires,” of all the most important organic
“appetites.” All movements of approach or withdrawal, of
attack or flight, of taking or rejecting which animals make
are only so many direct or indirect consequences of this perfectly
general tendency of every stationary physiological condition to
remain constant. We shall soon see that this tendency in its

1921] : Wheeler: Some Social Beetles ilAlef

turn is only the direct result of the mnemic faculty characteristic
of all living matter. This single physiological tendency of a
general kind, accordingly, is sufficient to give rise to a large num-
ber of the most diversified particular affective tendencies. Thus
every cause of disturbance will produce a corresponding ten-
dency to repulsion with special characteristics determined by
the kind of disturbance, by its strength, and by the measures
capable of avoiding the disturbing elements; and for every
incidental means of preserving or restoring the normal physio-
logical condition, there will be a quite definite corresponding
tendeney such as “longing,” “desire,” “attraction” and so forth.
Even the instinct of self-preservation—when understood in the
usual narrow sense of “preservation of one’s own life’”—is only
a particular derivative and direct consequence of this very gen-
eral tendency to preserve physiological invariability.”” This ten-
dency, however, as Rignano remarks, is supplemented by another,
“for as soon as the previous stationary condition cannot be
restored by any means, that is by any movements or change ot
location, the organism disposes itself in a new stationary con-
dition consistent with its new external and internal environment.
In this way there originate a large number of new phenomena

, 99

called ‘adaptations’.

2)” 66

Of course, the contention that the appetites are fundamen-
tally important in animal behavior is not new. It is merely
astonishing that they have been so consistently ignored by many
modern observers. The role of appetency, or appetite, was set
forth with great acumen by the philosopher Fouillée, especially
in the third book of his ‘‘Evolutionnisme des Idées Forces”
(1920, first edition 1890). The germ of the conception, however,
can be traced back to Reimarus (1798) and Leibnitz (Dwelshav-
ers, 1908, p. 181), to the ‘‘appetitus sensitivus” of the schoolmen
(see Maher, 1903, Chap. 10), and the docEic and td doextindv
of Aristotle (De Anim. 3, 10; Eth. Nic. 1, 13, 18). Recently the
behavioristic psychologists, psychopathologists and students of
the sympathetic nervous system and internal secretions, Mosso
(1899), Drever (1917), Smith and Guthrie (1921), Goddard
(1919), Kempf (1918), Crile (1915), Cannon (1915), and others,
have emphasized the importance of the appetites and others have
stressed their peculiarities in such terms as “‘tumescence” and

118 Zoologica: N. Y. Zoological Society [IlI;3

“detumescence” in sexual psychology and ‘enhancement,’
“relief” and ‘‘catharsis” in art. Kempf (1921) in his valuable
contribution to psychoanalysis has used the term “craving” in-
stead of appetite, avoiding the “libido” of the Freudians which
embodies the same notion. In a recent volume’ Bertrand
Russell takes essentially the same view of the phenomena
of appetite but uses the word “desire.” Most of the authors
cited deal, of course, with man, but one can hardly over-
estimate the value of their work for the animal behaviorist and
entomologist. It is certain that insects have well-developed sym-
pathetic and glandular systems and that their alimentary and
sexual behavior presents quite as definite a picture of appetites
as does the corresponding behavior of the higher animals.

Fouillée believes that every appetition involves a rudimentary
cognition and that automatic behavior like that of the habits
and reflexes is merely lapsed appetition. If it could be shown
that the latter really can have this derivation and that such
ontogenetic mechanisms as habits can acquire representation in
the germ-plasm and hereditary transmission, we might be in a
position to give a consistent account of all animal behavior, and
one which would lead us to regard the reflexes and the tropisms
as ultimate, highly specialized end-stages instead of primitive,
elemental components of behavior.

1“The Analysis of Mind,’ London and New York, Allen, Unwin &
Maemillan, 1921.

POSTSCRIPT.

Just as the final proof of this paper was being returned to
the printer Prof. I. W. Bailey received a letter from Col. David
Prain, Director of the Royal Botanical Gardens of Kew, with
the identification of the Tachigalia. It proves to be T. paniculata
Aublet. Col. Prain compared our specimens with Aublet’s type
in the British Museum Herbarium.

Among the various social beetles considered in the latter part
of my paper I should have included Phrenapates bennett: Kirby,
_ the habits of which were studied by Ohaus (1909) in Ecuador.
I shall have occasion to return to this insect in a future
publication.

BIBLIOGRAPHY

AUBLET.
1775. Histoire des Plantes de Ja Guiane Francoise. 4 vol., 1775. London
and Paris, Didot.
BELT, T.
1888. The Naturalist in Nicaragua. 2d Ed. London, Bumpus, 1888.

' BINGHAM, C. T.

1907. The Fauna of British India, including Ceylon and Burma.
Butterflies. Vol. II, 1907, London, Taylor and Francis.

BLATCHLEY, W. S.

1910. An Illustrated Descriptive Catalogue of the Coleoptera or Beetles
(Exclusive of the Rhynchophora) Known to Occur in
Indiana. Indianapolis, Nature Publ. Co., 1910.

Buisson, J. F. J.

1849. Description de la Larve et de la Nymphe du Silvanus sexden-
tatus Fabr. Ann. Soc. Ent. France (2) 7, 1849, pp. 163-172,
pl. 6, figs. 1-18.
BopINng, J. H.
1920. Factors Influencing the Water Content and the Rate of
Metabolism of Certain Orthoptera. Journ. Exper. Zool. 32,
1921, pp. 137-164, 6 figs., 6 tables.
BOvING, A.
1910. Natural History of the Larve of Donaciine. Internat. Rev.
gesamt. Hydrobiol. Hydrograph. 3, 1910, pp. 45-108.
Biol. Suppl., pp. 1-4, 7 pls. and 70 textfigs.

Von BUDDENBROCK, W.
1915. Die Tropismentheorie von Jacques Loeb. Biol. Centralbl. 36,
1915, pp. 481-506. Trans. in Journ. Anim. Behav. 6, 1916,
pp. 341-366.

CANNON, W. B.

1915. Bodily Changes in Pain, Hunger, Fear and Rage. N. Y. and
London, D. Appleton & Co., 1915.

CasEy, T. L.

1884. Revision of the Cucujide of America North of Mexico. Trans.
Amer. Ent. Soc. 11, 1884, pp. 69-112, 5 pls.

CHAMPION, G. C.
1896b. Silvanus surinamensis (Linn.) Ent. Month Mag. (2) 7, 1896,
pp. 268, 269.

CHITTENDEN, F. H.

1895. The More Important Insects Injurious to Stored Grain. Year-
book U. S. Dep. Agric. (1894) 1895, pp. 377-394. 9 figs.

1896. A New Grain Beetle. Canad. Ent. 28, 1896, pp., 197, 198.

1897a. Some Insects Injurious to Stored Grain. Farmers’ Bull. No. 45.
U. S. Dep. Agric., 1897, pp. 1-24, 18 figs.

1897b. Some Little-known Insects Affecting Stored Vegetable Products.
Bull. No. 8 (N. S.) U. S. Dep. Agric. Div. Ent., 1897, pp.
1-45, 10 figs.

1911. A List of Insects Affecting Stored Products. Bull. No. 96, Pt. 1,
U. S. Dep. Agric. Bur. Ent., 1911, pp. 1-7.

120 Zoologica: N. Y. Zoological Society [TII;3

CLAPAREDE, E.

1912. Point de Vue Physico-Chimique et Point de Vue Psychologique.
Scientia 11, 1912, pp. 251-258.

1913. Article “Thierpsychologie” in “Handworterbuch der Naturwis-
senchaften”. Bd. 9, Jena, Gustav Fischer, 1913.

Comstock, J. H.
1882. Note on the Structure of Mealy Bugs. Rep. U. S. Dep. Agric.
(1881), 1882, p. 214.
COQUEREL, C.

1849. Note sur les Métamorphoses du Silvanus sexdentalus. Ann
Sec. Ent. France (2) 7, 1849, p. 172.
Craic, W.

1918. Appetites and Aversions as Constituents of Instincts. Biol. Bull.

34, 1918, pp. 91-107.
CRAIGHEAD, F. C.

1920. Biology of Some Coleoptera of the Families Colydiide and
Bothrioderide. Proc. Ent. Soc. Wash. 22, 1920, pp. 1-13, 2
pls., 1 textfig.

CRILE, G. W.

1915. The Origin and Nature of the Emotions. Phila. and London,
W. B. Saunders Co., 1915.

DimmMock, Mrs. A. K.
1885. The Insects of Betula in North America. Psycne, 1885, pp.
239-243, 271-286.
DIMMOCK, G.

1884. Article “Beetles” in Standard Natural History. Vol. 2, 1884.
Boston, 8. C. Cassino & Co.

DOFLEIN, F.
1916. Der Ameisenloéwe, eine biologische, tierpsychologische und
reflexbiologische Untersuchung. Jena. Gustav Fischer, 1916,
138 pp., 43 figs., 10 pls.
DorAN, E. W.

1892. The Evolution of Heat by Silvanus cassie. Proc. Ent. Soc.
Wash. 2, 1892, pp. 325-327.

DREVER, J.
1917. Instinct in Man, a Contribution to the Psychology of Education.
Cambridge, Univ. Press, 1917.
DWELSHAUVERS, G.
1908. La Synthése Mentale. Paris, F. Alean, 1908.

EICHOFF, W.
1881. Die Europaischen Borkenkafer. Berlin, J. Springer, 1881, pp.
280-281.
EMERY, C.

1898. Studio Monografico sul Genere Azteca Forel. Mem. R. Accad.
Se. Ist. Bologna (5) 3, 1898, pp. 319-352, 2 pls.

ESCHERICH, K.
1902. Biologische Studien iiber algerische Myrmekophilen, zugleich
mit allgemeinen Bemerkungen tiber die Entwicklung und
Bedeutung der Symphilie. Biol. Centralbl. 22, 1902, pp.
638-663, 4 figs.

1921] Wheeler: Some Social Beetles 121

FELT, E. P.
191la. A Generic Synopsis of the Itonide. Journ. N. Y. Ent. Soc. 19,
1911, pp. 31-62.
1911b. Habits and Galls of American Gall Midges. Journ. Econ. Ent.
4, 1911, pp. 451-484.
FISKE, W. F.
1905. Catogenus rufus. A Coleopterous Parasite. Proc. Ent. Soe.
Wash. 7, 1905, pp. 90-92.

FORBES, 8. A.
1880. Studies of the Food of Birds, Insects and Fishes Made at the
Illinois State Laboratory of Natural History at Normal,
Illinois. Bull. Ill. State Lab. Nat. Hist. 1, 1880 (second
ed. 1903), 176 pp.
1884. Insects Injurious to the Strawberry. Rep. State Ent. Illinois
13, 1884, pp. 60-180.
1894— 1905. A Monograph of the Insect Injuries to Indian Corn. Part
I, 18th Rep. State Ent. Illinois, 1894, 171 pp., 15 pls. Part
II, ibid. 23rd Rep. 1905, 273 pp.
1915. The Ecological Foundations of Applied Entomology. Ann. Ent.
Soc. Amer. 8, 1915, pp. 1-19.
FOREL, A.
1904. In und mit Pflanzen lebende Ameisen aus dem Amazonas-
Gebiet und aus Peru. Zool. Jahrb. Abth. Syst. 20, 1904,
pp. 677-707.
1906. Fourmis Néotropiques Nouvelles ou peu Connues. Ann. Soc.
Ent. Belg. 50, 1906, pp. 225-249.
1912. Formicides Néotropiques. Part IV. Mém. Soc. Ent. Belg. 20,
1912, pp. 1-32.
FOuILLEE, A.
1920. L’Evolutionnisme des Idées-Forces. 6th Ed. Paris. F. Alean, 1920
(First ed. 1890).
FOWLER, GC
1889. The Coleoptera of the British Islands. Vol. III. Clavicornia
London, L. Reeve & Co., 1889.

GANGLBAUR, L.
1899. Die Kafer von Mitteleuropa. Bd. 3. Vienna, C. Gerold’s
Sohn, 1899.
GaviT, J. E.
1849. Corn or Grain Weevil of Europe (Calandra granaria) and
Silvanus surinamensis, the Weevil most Common in America.
Trans. N. Y. State. Agric. Soc. for 1848. Vol. 8, 1849,
pp. 656-662, plate.

GIRAULT, A. A.
1912. Insects Injurious to Stored Grains and their Ground Products.
Bull. No. 156, Univ. Illinois Agric. Exper. Station, 1912,
pp. 67-92, 12 figs.

GLOVER, T.
1869. The Food and Habits of Beetles. Rep. U. S. Com. Agric. 1868
(1869), pp. 78-117, 189 figs. ;

GoppDarD, H. H.
1919. Psychology of the Normal and Subnormal. New York, Dodd,
Mead & Co., 1919.

122 Zoologica: N. Y. Zoological Society [III;3

GROUVELLE, A.

1896. Nitidulides, Colydiides, Cucujides et Parnides récoltés par M. E.
Grouvelle au Brésil et autres Clavicornes nouveaux
d’Amérique. Ann. Soc. Ent. France 65, 1896, pp. 177-216,
26 figures.

1913. Notes sur les Silvanini (Col. Cucujidz). Ann. Soc. Ent. France
81, 1913, pp. 313-386.
GUILLEBEAU, F.
1890. Note sur les Silvanus. Rev. d’Ent. 9, 1890, pp. 220-224.
HAGEDORN, M.

1907. Pilzziichtende Borkenkafer. Naturw. Wochenschr. N. F. 6,
1907, pp. 289-298, 12 figs.

1910. Coleoptera. Fam. Ipide. Wytsman’s Genera Insectorum, 1910,
pp. 1-178, 14 pls.
HAMILTON, J.
1886. Natural History Notes on Certain Coleoptera. Canad. Ent. 18,
1886, pp. 26-30.
HANDLIRSCH, A.

1908. Die Fossilen Insekten und die Phylogenie der Rezenten Formen.
Text and Atlas. Leipzig, Wilh. Engelmann, 1908.
Harms, H.

1906. Leguminose in II Beitrage zur Flora der Hylea nach den
Sammlungen von Ule’s Amazonas-Expedition. Verh. bot.
Ver. Prov. Brandenburg 48, 1906, p. 164.
HEIN, S. A. A.
1920. Technical Experiences in the Breeding of Tenebrio molitor.
K. Akad. Vetens. Amsterdam. Proc. 238, 1920, pp. 193-218.
HERRICK, T. H.
1910. Instinct and Intelligence in Birds. Pop. Sci. Month., 1910, pp.
532-556, 82-97, 122-141.
HopkKINS, A. D.

1898. On the History and Habits of the Wood Engraver Ambrosia
Beetle Xyleborus wylographus (Say), Xyleborus saxeseni
(Ratz), with Brief Descriptions of Different Stages.
Canad. Ent. 30, 1898, pp. 21-29, 2 pls.

Howarp, L. O. and C. L. MARLATT.

1896. The Principal Household Insects of the United States, with
a Chapter on Insects Affecting Dry Vegetable Foods by
F. H. Chittenden. Bull. No. 4, N. S., U. S. Dep. Agric. Div.
Ent. 1896, 130 pp., 63 figs.

HUBBARD, H. G.

1897. The Ambrosia Beetles of the United States. Bull. 7 (N. S.)
Div. Ent. U. S. Dep. Agric. 1897, pp. 9-30, 34 figs.

1906. Ambrosia Beetles. Yearbook, U. S. Dep. Agric. 1906, pp.
421-430, 7 figs.

HUBER, J.

1908. Materiales para a Flora Amazonica. VII. Plante Duckeane
austro-guyanenses. Bol. Mus. Goeldi 5, 1908, pp. 387, 388.

1921] Wheeler: Some Social Beetles 123

JENNINGS, H. S.

1904. Contributions to the Study of the Behavior of Lower Organisms.
Carnegie Inst. Wash. Publ. No. 16, 1904.

1906. Behavior of the Lower Organisms. New York, Columbia Univ.
Press, 1906.

1909. Tropisms. C. R. 6me, Congres. Internat. Psychologie Geneva,
1909, pp. 307-324.

KALTENBACH, J. H.
1874. Die Pflanzenfeinde aus der Klasse der Insekten. Stuttgart,
Julius Hoffmann, 1874.
KeEmpr, E.
1918. The Autonomie Functions and the Personality. Nervous and
Mental Disease Monogr. Series No. 28. N. Y. and Wash-
ington, 1918.
KeEmpPr, E. J.
1920. Psychopathology. St. Louis, C. V. Mosby Co., 1920.

KIEFFER, J. J.
1920. Déscription de quelques Cécidomyides Nouvelles. Marcellia 1,
1902, p. 115.
Knap, F.
1915a. The Secretions Employed by Rhynchophorous Larve in
Cocoon-making. Proc. Ent. Soc. Wash. 17, 1915, pp. 154-158.
1915b. Dung-bearing Weevil-Larva#. Proc. Ent. Soc. Wash. 17, 1915,
pp. 198-194.
KoLBE, H.
1911. Die Vergleichende Morphologie und Systematik der Coleopteren.
Mem. ime. Congr. Internat. d’Ent. Bruxelles, 1910. Vol. 2,
1911, pp. 41-68.
KUSTER, E.
1911. Die Gallen der Pflanzen. Leipzig, L. Hirzel, 1911.

Lea, A. M.

1904. Descriptions of New Species of Australian Coleoptera. Proc.
Linn. Soc. N. S. W. 29, 1904, pp. 60-107, 1 pl.

LEBARON, W.
1874. Fourth Annual Report of the Noxious and Beneficial Insects
of the State of Illinois. Springfield, 1874, pp. 18-199. 94 figs:
LECONTE, J. L. AND Horn, G. H.
1883. Classification of the Coleoptera of North America. Smiths.
Misc. Coll. 507, 1888, pp. XXVIII, 567.
LENG, C. W.

1920. Catalogue of the Coleoptera of America North of Mexico. Mt.
Vernon, N. Y., J. D. Sherman, 1920.

LuccER, O.
1899. Coleoptera, or Beetles. Fifth Ann. Rep. Ent. Minnesota, 1899,
248 pp.
MACGILLIVRAY, A. D.
1921. The Coccide. Urbana, IIl., Scarab Co., 1921.

124 Zoologica: N. Y. Zoological Society [III ;3

MaHe_R, M. (S. J.).
1903. Psychology, Empirical and Rational. London, Longmans, Green
& Co., 1903.

Mostus, K.
1877. Die Auster und die Austernwirthschaft. Berlin, 1877. Transl.
“The Oyster and Oyster Culture.” Rep. U. S. Fish Comm.
(1880), Part VIII, 1883, pp. 683-751.

Mosso, A.

1899. The Mechanism of the Emotions. Clark Univ. Decen. Cel.
Vol., 1899, pp. 396-407.

NEGER, F. W.

1908a. Ambrosiapilze. Ber. Deutsch. Bot. Ges. 26, 1908, pp. 735-754.

1908b. Die Pilzkulturen der Nutzholzborkenkafer. Centralbl. f. Bak-
teriol. 20, 1908, pp. 279-282.

1909. Ambrosiapilze II. Die Ambrosia der Holzbohrkafer. Ber.
Deutsch. Bot. Ges. 27, 1909, pp. 372-389.

1911. Zur Uebertragung des Ambrosiapilzes von Xyleborus dispar F.
Naturw. Zeitschr. f. Forst. u. Landwiss. 9, 1911, pp.
223-225.

NEWCOMER, EF. J.
1912. Some Observations on the Relations of Ants and Lyczenid
Caterpillars, and a Description of the Relational Organs
of the Latter. Journ. N. Y. Ent. Soc. 20, 1912, pp. 31-36,
2 pls.

DE NICEVILLE, L.
1890. The Butterflies of India, Burmah and Ceylon. 3. Calcutta, 1890.

OHAUS, F.
1899—1900. Bericht iiber eine entomologische Reise nach Zentral-
brasilien. Stett. Ent. Zeitg. 60, 1899, pp. 204-245; 1900,
pp. 164-191, 193-274.
1909. Bericht tiber eine entomologische Studienreise in Siidamerika.
Stett. Ent. Zeitg. 70, 1909, pp. 3-139.

PACKARD, A. S.
1881. Insects Injurious to Forest and Shade Trees. Bull. No. 7, Dep.
Inter. U. S. Ent. Com., Washington, 1881, 275 pp. 100 figs.
PERRIS, E.
1852— 1862. Histoire des Insects du Pin Maritime. Ann. Soc. Ent.
France (2) 10, 1852, pp. 491-522, pl. (3) 1, 1853, pp. 555-644;
2, 1854, pp. 85-160, 593-646; 4, 1856, pp. 173-257, 423-486;
5, 1857, pp. 341-395, pl. Suppl. (4) 2, 1862, pp. 173-243, pl.
1876. Larves de Coléoptéres. Ann. Soc. Linn. Lyon. 22, 1786, pp.
259-418.

DE PBYERIMHOFF, P.

1902. 1903. Descriptions des Larves de Trois Coléoptéres Exotiques.
Ann. Soc. Ent. France, 71, 1902-03, pp. 710-718, 6 figs.

PICARD, F.
1919. Contribution a l’Etude du Peuplement d’un Végétal. La Faune
Entomologique du Figuier. Ann. Serv. Epiphyt. 6, 1919, pp.
35-174, 36 figs.

1921] Wheeler: Some Social Beetles 125

PIERCE, W. D.

Oe AY Manual of Dangerous Insects Likely to be Introduced in the
United States through Importations. U. S. Dep. Agric.
1917, 256 pp., 48 pls., 107 text figs.

REpI, F.
1671. Experimenta circa generationem Insectorum. Amstelodami,
1671.
RIGNANO, E.

1920. Psychologie du Raisonnement. Paris, F. Alcan, 1920. (First
Chapter translated and entitled “On the Mnemonic Origin
and Nature of Affective Tendencies.” Monist 221, 1911,
pp. 321-356.)

REITTER, E.

1911. Fauna Germanica. Die Kafer des deutschen Reiches, III Bd.
Stuttgart, K. G. Lutz, 1911.

REIMARUS, H. S.

1798. Allgemeine Betrachtungen iiber die Triebe der Thiere, haupt-
sdchlich tiber ihre Kunsttreibe. Hamburg, C. E. Bohn, 1798.

RUBSAMEN, E. K.

1899. Ueber die Lebensweise der Cecidomyiden. Biol. Centralbl. 19,
1899, pp. 529, 561.

SCHNEIDER-ORELLI, F.

191la. Die Uebertragung and Keimung des Ambrosia-Pilzes von
Xyleborus (Anisandrus) dispar F. Naturw. Zeitschr. f.
Forst- und Landwiss. 9, 1911, pp. 186-192, 223-225.

1911b. Ueber die Symbiose eines einheimischen pilzztichtenden Borken-
Kafers (Xyleborus dispar F.) mit seinem Nahrpilz. Verh.
Schweiz. Naturf. Gesell. 94, 1911, pp. 279, 280.

1912. Der ungleiche Borkenkiafer (Xyleborus dispar F.) an Obstbiau-
men und sein Nahrpilz. Naturwirtsch Jahrb. Schweiz. 1912,
pp. 326-334.

1913. Untersuchungen iiber den Pilzziichtenden Obstbaumborkenkafer
Xyleborus (Anisandrus) dispar und seinen Na&hrpilz.
Centralbl. Bakter. Parasit. 38, 1913, pp. 25-110.

SMITH, J. B.
1909. The Insects of New Jersey. Ann. Rep. N. J. State Museum,
1909.

SMITH, S. AND GUTHRIE, E.
1921. Chapters in General Psychology. Seattle, University of Wash-
ington Press, 1921.

i_

SPRUCE, R.

1908. Notes of a Botanist on the Amazon and Andes. 2 vols. N. Y.,
Macmillan & Co., 1908. (Chapt. 24, “Ants and Plant Struc-
ture,” being a paper submitted and read to the Linn. Soc.,
Apr. 15, 1869, but not published.)

Stitz, H.
1913. Ameisen aus Brasilien, gesammelt von Ule. Deutsch. Ent.
Zeitschr. 1913, pp. 207-212, 2 figs.

126 Zoologica: N. Y. Zoological Society [III;3

Suc, K.
1909. Zur a netcmis der Cocciden. Zool. Anzeig. 34, 1909, pp. 164-172,
gs.
THAXTER, R.
1914. On Certain Peculiar Parasites of Living Insects. Bot. Gazette
58, 1914, pp. 235-253, 4 pls.
1920. Second Note on Certain Peculiar Fungus Parasites on Living
Insects. Bot. Gazette 69, 1920, pp. 1-27, 5 pls.

THOMANN, H.
1901. Schmetterlinge und Ameisen. Beobachtungen tiber eine Sym-
biose zwischen Lyccena argus L. und Formica cinerea Mayr.
Hnsug: Diss. Univ. Ziirich, Chur, J. Casanova, 1901, 44 pp.,
1 pl.
TUSLANE, L. R.
1844. Légumineuses Arborescentes de l’Amérique du Sud. Arch. Mus.
d’Hist. Nat. 4, 1844, p. 160.
ULE, E.
1907. Ameisenpflanzen des Amazonasgebietes. In Karsten and
Schenck’s Vegetationsbilder 4. Reiche. Jena, 1907, 14 pp.,
6 pls.
WASMANN, E.
1896. Die Myrmekophilen und Termitophilen. C. R. 3me Congr. Intern.
Zool. Leyden (1895) 1896, pp. 411-440.
WEISSE, J.

1897. Biologische Mitteilungen. Deutsch. Ent. Zeitschr, 1897, pp.
389-395.

WHEELER, W. M.

1901. Microdon Larve in Pseudomyrma Nests. Psyche 9, 1901, pp.
222-224, 1 fig. :

1918. A Study of Some Ant Larve, with a Consideration of the Origin
and Meaning of the Social Habit among Insects. Proc.
Amer. Philos. Soc. 57, 1918, pp. 293-348, 12 figs.

1921. A New Case of Parabiosis and the Ant Gardens of British
Guiana. Ecology, 2, 1921, pp. 89-103, 3 figs.

WHEELER, W. M. AND BAILey, I. W.
1920. The Feeding Habits of Pseudomyrmine and Other Ants. Trans.
Amer. Phil. Soc. N. S. 22, 1920, pp. 235-279, 5 pls., 6 text-figs.
Wuite, F. B.
1872. Contributions Towards a Knowledge of the Life-histories of
Certain Coleoptera. Ent. Month. Mag. 8, 1872, pp. 196-198.
WickKHaAM. H. F.

1920. A Catalogue of the North American Coleoptera Described as
Fossils. In Leng’s Catalogue (q. v.) pp. 349-365.

AS
ANT
b4
Ny)
hs
4

J

4 t

4
~—

PLATE I. YOUNG TACHIGALIA FORMICARUM HARMS, FROM LETICIA, PERU.

From a photograph taken by E. Ule and reproduced from Plate V of his article in Kartsen
and Schenck’s Vegetationsbilder 4, Reihe, 1907

a
te)
0
:
e 4
e <
® ®
)
< ig
» ie}
4 Zz
% Wl
iS) By
% 0
Ys v
S, 4
is)
4
a fm)
= 0
3
a =
o% Bu
ra oA
{YSIS3HM VAYSOWOLN] E Soi
Oo <u
Q ° Font
ee
3 fo) u
ew tw)
ae
Ve
fe)
)
4
a”
Sy
0]
3
of
Sa
w.o000NnaAsSdA SISO 1NdIOVWIA os
ao
ae
uo
2 1)
ae a
Res
par
6
»
ao
\
m2) j
a
ie
(oy) SO BRACK,
rel if oS MYR
i
] ie}
OF ae
AZTECA FOVEICEPS —— E 5 BRACHYMYR
CF
- 3
Zz \
or ~— on
Xe) N me
on™ S ch
wr r i:
ye
on™ & / z
sev? ¥ v te)
1S r 0 0 o
< e Z we
r yu < r
> ~ 7 c
ae fo) O
5
RS r z
oO 1¢)
? O a
x <F < @
@ é 2
i ia
ss >
O 2
fe) fo)
S) fa)
m7 5)
iY) wW
Qa 1)
a

SJSONAOSOOIS VIIVSOIHOVL

3% 3
ef sel
on on?
= 3 ee
va
Y
2) Zi AE
5 wee
~ ot a
) yon
fo) Pad
eae Bse Zo
X 2 -
Ss Foes
\ EL a)
or
Vi °o
/ . ie)
(é) S929nG0y, SLOOHS apni
3 é¢)Sn Sia 04v30 _-st
S30IOSWISitv SNYsqoHSIN[a +Svoans AWNIWYS4 ae oz r r
WOivINDeaeNnL vYWWoO1vi9os a uy 2
v v Dv
1 ¥ § is
VOILOlavavd HSLiSsvOO1WWaYyD 2) 8 (o} 9 =
a
SN1LVYOWSS SNLONOGWYD eSouoWw t 4 7 i
sonyuin 1 (e) rT 26 0
Ge See) Go a 2
/
s3700i113d / S, 2 > > 9
/ > > x x of 2)
WA ” a
ae c v 2 ~?
2 ES 7 io) y
7 tH) O
Z 2°
o
d a” ae°
r 2» ssh
1) a neo
2
S, vw
36 A Roane
7 of 6ne?
\
} v8
noe
ele

xKINLIN SISDONZ10S

~

Sp
ty
Sw Ys
Srwe
ee
So
+o
wauv7 (S3NITIN@NI)
SiNny

OINOIWNDYADS

SSNITINONI

S3dIsnH WnId0d
SNOWVvA

ASON3OOOIS VIIVSDIHOVL

LIS
oh AOS
ero?
4 a ae
»
yy VIO
4 Ze
2, 7
x yee
a oe Ae
~ » [e)
fe) A Ad
ay ao’
a ea
WS y aw 4
wr Aes ‘o”
b \o*
[o)
wo
st Ongoys SLOOHS Dee ae a
S3q01I0Sv13SitvY SNYSQOHSIN[CG (é) “"L8vogK5 IVNIWSS LL pat ows ae li Bs
WONLvINDeSsBNnL YWWO1vLOg 2 uy L
Z
=) iw
VOILOIavVavYd YSLSVYODOLYNSYNO asl 3) ix 8 3 (6) “Y
SNLVYOWSS SNLONOGWYD naooHol™ ch x 2 y
snyt t z fe) (o) ea
% n 2D a 2
s3n0I13d ~ = P) > > Q
re) ? x x 7, @
‘ ” »
. L S c v v ~”
s Zz ie ah ° 4
7 (0) o c er o”
Guo Se ag a > ae
Ma, ©:% 4 ny 2 2)
Sy ei ely pole, & By 22
42% ce 0 0 2» wv
7 4 a Cc 2 S oY
aS Y) 2 4 w?
y y) es ?
Sy : vw? _
Sn S, g x on
Yaw 9, fol yaw
igh es Ys ow a ei
OW ™~ le
Plaisg Ao =
Lo oat
s SANIMIN®NI >
SdisaNyM Wooded Se Sa Ree KIMLIAN SISGONZTOS
ee o Pn
—_ — a a Ss & i)
~ = x ° R S
va iv < »>
x > 5 9 4 *y
= © v Q
wFinawoua I 2 2 Se
sno0090a0n4Sd 0) 5 0 “
< De ie)
= 0
Q <
ig
o
i
: g
= > yy Se
‘s 2 . Pee
7) x 0 N
{e}
iy a) a
IW3ISSHM 5 ) SF
sniaisnynna z= v <
SIGONILAY SISdON370S 2 0 & (Ayo1voinso) SNVIONYSD YWYAWOONSSG
Ssintvi3d0s wy ol © SLNV
SNHdOwY LOGIDD09 0 fr) x \
¢ 2 \ La “2,
a 0 ~ Y~» Say
as 0 ony
2° ——{ =
rs (9) L &
at \ - | a Way
» ~ | \ iS) Ay
4 = 4 ~ | “ 4) fo)
“2 > m a Y rh D Ne [ny
— fo) 72 < c a 3 Vy Sy
oe on - ° fs W ah L
ao z a Q z 4,
: o 1S c u a 4
0 > a,
Q o LS fe)
0 2 *~ ix fo) “ °
< s u a 2)
N »
> x ba AS}
< x Y
z 0 z
aC Wl [e)
m F o
m N a
n < ny
a

PLATE III.

. 1. Eunausibius wheeleri Schwarz and Barber. X 3%.

Coccidotrophus socialis Schwarz and Barber. X 8.

Cross section of base of very young petiole of adult Tachigalia, showing pith
still in the eavity. Photograph by Prof. I. W. Bailey.

Cross section of base of large petiole of Tachigalia inhabited by Pseudomyrma
damnosa. The dark areas in the wall are nutritive parenchyma, which is not
disturbed by the ants but nourishes their coccids. A portion of one of the carton
partitions is shown on the left side. Photograph by Prof. I. W. Bailey.

Cross section through one of the strands of nutritive parenchyma showing the
cells with their homogeneous, amber-colored contents. Photograph by Prof.

I. W. Bailey.

Fig. 1.

Fig. 2.

Hig. 3.

PLATE IV.

Cross section of an as yet uninhabited petiolar swelling of a young Tachigalia,
showing the intact nutritive parenchyma, large pith cavity, and thin layer of
pith cells lining it. Photograph by Prof. I. W. Bailey.

Cross section of a petiolar swelling inhabited by Coccidotrophus socialis, show-
ing the gnawed nutritive parenchyma and the frass ridges. Photograph by
Prof. I. W. Bailey.

Section similar to that of Fig. 2 but showing three Coccidotrophus cocoons in
cross section. Photograph by Prof. I. W. Bailey.

PLATE V.

Halves of various Tachigalia petioles seen from the inside, showing the arrangements

of the cocoons, frass ridges, entrances, and in some cases also the beetles,
larvee and coccids end a pupa (second figure from below).

¥

Vid gle. ae A

u| ft,
Into tT
ne MAL

q

Ao
ROHAN dS

‘)

QL Wheeler, William Morton.
591 Study of some social
B74W48 beetles in British Guiana,,
Ent. ISP

wi ili

nhent QL591.B74W48
Study of some social beetles in British