Skip to main content

Full text of "The anatomy of the honey bee"

See other formats

,;^^'' W^'- 


^in^ - 



LIBRARY 'vi t .. %■ 

■--, '. New York State Colleges 

OF ,. * 

Agriculture and Home Economics 


Cornell University 





Date Due 

rnell University Library 
The anatomy of the honey bee / 

3 1924 003 168 865 

Cornell University 

The original of tiiis book is in 
tine Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 

Technical Series, No. 18. 


L.'O, HOWARD, Entomologist and Chief of Bureau. 



Agent anii Eipert. 

Issued May 28, ,19i0. 



Technical Series, No. 18. 


L. 0. HOWARD, Entomologist and Chief of Bureau. 



Agent and Expert. 

Issued May 28, 1910. 







L. O. Howard, Entomologist and Chief of Bureau. 

C. L. Maelatt, Assistant Entomologist and Acting Chief in. Absence of Chief. 

K. S. Clifton, Executive Assistant. 
W. F. Tastet, Chief Clerk. 

F. I-I. Chittenden, in charge of truck crop and stored product insect investigations. 

A. D. Hopkins, in charge of forest insect investigations. 

W. D. Hunter, in charge of southern field crop insect investigations. 

F. M. Webster, in charge of cereal and forage insect investigations. 
A. L. Quaintance, in charge of deciduous fruit insect investigations. 
E. F. Phillips, in charge of bee culture. 

D. M. Rogers, in charge of preventing spread of moths, field work. 
RoLLA P. CuRRiE, in charge of editorial work. 

Mabel Colcord, librarian. 

Investigations in Bee Culture. 
E. F. Phillips, m charge. 

G. F. White, J. A. Nelson, B. N. Gates, R. E. Snodgeass, A. H. McCray, agents 
and experts. 

Ellen Dashiell, preparator. 
Jessie E. Marks, c/c/7c-. 
T. B. Symons, collaborator for Maryland. 
H. A. Surface, collaborator for Pennsylvania. 
J. C. C. Price, collaborator for Virginia. 


U. S. Department of Agriculture, 

Bureau of Entomology, 
Washington, D. C, October 19, 1909. 
Sir: I have the honor to transmit herewith a manuscript entitled 
" The Anatomy of the Honey Bee," by Mr. R. E. Snodgrass, agent 
and expert, of this Bureau. It embodies the results of detailed 
studies made by Mr. Snodgrass and should prove of value as bring- 
ing to the bee keeper reliable information concerning an insect of 
such great economic importance, and also as furnishing a sound 
basis in devising new -and improved practical manipulations. I 
recommend its publication as Technical Series, No. 18, of the Bureau 
of Entomology. 

Respectfully, L. O. Howard, 

Entomologist and Chief of Bureau. 
Hon. James Wilson, 

Secretary of Agriculture. 



I. Introduction 9 

II. General external structure of insects 10 

III. The head of the bee and its appendages . 26 

1. The structure of the head 26 

2. The antennae and their sense organs 32 

3. The mandibles and their glands 39 

4. The proboscis 43 

5. The epipharynx 51 

IV. The thorax and its appendages 53 

1. The structure of the thorax 53 

2. The wings and their articulation j-. -. 59 

3. The l^s 66 

V. The abdomen, wax glands, and sting 69 

VI. The alimentary catial and its glands 84 

1. The general physiology of digestion, assimilation, and excretion. 84 

2. The salivary glands 87 

3. The alimentary canal 90 

VII. The circulatory system - - 107 

VIII. The respiratory system 112 

IX. The fat body and the cenocytes 119 

X. The nervous system and the eyes 122 

XI. The reproductive system. 130 

1. The male organs 132 

2. The female organs 134 

Explanation of the symbols and letters used on the illustrations 139 

Bibliography 148 

Index 151 




Fig. 1. Median longitudinal section of body of worker 8 

2. Diagram of generalized insect embryo 12 

3. Example of generalized insect mouth parts 17 

4. Diagram of generalized thoracic segment 19 

5. Typical insect leg -. 2] 

6. Diagram of generalized insect wing and its articulation 22 

7. Diagram of terminal abdominal segments of a female insect and early 

stage in development of gonapophyses 25 

8. Example of a swordlike ovipositor 25 

9. Head of worker bee 27 

10. Heads of worker, queen, and drone 29 

11. Median longitudinal sections of heads of worker and drone 30 

12. Antennal hairs and sense organs 36 

13. Mandibles of worker and drone 40 

14. Internal mandibular gland of worker 42 

15. Mouth parts of worker 43 

16. Median section through distal half of mentum and base of ligula of 

worker 50 

17. Epipharynx and labrum of worker 51 

18. Sense organs of epipharynx 52 

19. Median longitudinal section of head of worker 52 

20. Dorsal view of ventral walls of body of worker 53 

21. Thorax of worker 54 

22. Lateral view of mesotergum of worker 56 

23. Thoracic terga of worker 57 

24. Upper part of left mesopleiu-um of worker 58 

25. Wings of Hymenoptera 60 

26. Basal elements of wings of Hymenoptera 61 

27. Median section through thorax of drone 64 

28. Internal view of right pleurum of mesothorax of drone 65 

29. Legs of worker, queen, and drone 67 

30. Claws and empodium of foot of worker 68 

31. Tarsal clawa of worker, queen, and drone 69 

32. Lateral view of abdomen of worker 70 

83. Ventral view of abdomen of worker 70 

34. Dorsal view of abdominal sterna of drone 70 

35. Sixth abdominal sternum of worker, queen, and drone , 72 

36. Semidiagrammatic view of left side of sting of worker 75 

37. Ventral view of sting of worker 76 

38. Section of small piece of wall of poison sac 79 

39. Sections of alkaline gland of sting 79 

40. Details of sting of worker 81 

41. Tip of abdomen of worker with left side removed 82 


tfclATJligkTIOKS. ^ 


Fig. 42. Alimentary canal of worker 85 

43. Details of pharyngeal and salivary glands 88 

44. Honey stomach of worker, queen, and drone 94 

45. Longitudinal section of honey stomach and proven triculus of queen . 97 

46. Histological details of alimentary canal of worker 103 

47. Dorsal diaphragm of drone, from one segment 108 

48. Small part of dorsal diaphragm of drone HO 

49. Pericardial chamber of one segment in worker Ill 

50. Tracheal system of worker 113 

51. Tracheal system of worker 117 

52. Nervous system of worker 123 

53. Brain and sub oesophageal ganglion of worker 125 

54. Horizontal section of compound eye and optic lobe of worker 127 

55. Histological details of compound eye of worker 128 

56. Reproductive organs of drone 133 

57. Reproductive organ and sting of queen ". 135 



The anatomy of the honey bee has been for years a subject of much 
interest to those engaged in bee keeping both for pleasure and for 
profit. This interest is due not only to a laudable curiosity to know 
more of the bee, but to the necessity of such information in order 
to understand fully what takes place in the colony. All practical 
manipulations of bees must depend on an understanding of the be- 
havior and physiology of bees under normal and abnormal circum- 
stances, and those bee keepers who have advanced bee keeping most 
by devising better manipulations are those, in general, who know 
most of bee activity. In turn, a knowledge of bee activity must rest 
largely on a knowledge of the structure of the adult bee. 

Studies on the anatomy of the bee have not been lacking, for 
many good workers have taken up this subject for investigation. 
The popular demand for such information, however, has induced 
untrained men to write on the subject, and most accounts of bee 
anatomy contain numerous errors. This is probably to a greater 
extent true of the anatomy of the bee than of that of any other 
insect. Frequently the illustrations used by men not trained in 
anatomical work are more artistic than those usually found in papers 
on insect anatomy, and they consequently bear the superficial marks 
of careful work, but too often it is found that the details are in- 
accurate. It has therefore seemed the right time for a new presenta- 
tion of this subject based on careful work. 

The drawings given in the present paper are original, with the 
e'kteption of figures 12, 54, and 55, and have been prepared with 
a thorough realization of the need of more accurate illustrations of 
the organs of the bee, especially of the internal organs. Mistakes 
will possibly be found, but the reader may be assured that all the 
parts drawn were seen. Most of the dissections, moreover, were 
verified by Dr. E. F. Phillips and Dr. J. A. Nelson, of this Bureau, 
before the drawings were made from them. An explanation of the 
abbreviations and lettering is given on pages 139-147. 

It is hoped that the work will furnish the interested bee keeper 
with better information on the anatomy of the bee than has hereto- 
fore been offered to him, that it may provide a foundation for more 
detailed work in anatomy and histology, and, finally, that it will be 


of service to future students of the embryology and physiology of 
the bee. With this last object in view the writer has tried to sum 
up under each heading the little that is at present known of insect 
physiology in order to bring out more clearly what needs to be done 
in this subject. 


When we think of an animal, whether a bee, fish, or dog, we uncon- 
sciously assume that it possesses organs which perform the same vital 
functions that we are acquainted with in ourselves. We know, for 
example, that an insect eats and that it dies when starved ; we realize 
therefore that it eats to maintain life, and we assume that this involves 
the possession of organs of digestion. We know that most insects see, 
smell, and perform coordinated actions, and we recognize, therefore, 
that they must have a nervous system. Their movements indicate to 
us that they possess muscles. These assumptions, moreover, are en- 
tirely correct, for it seems that nature has only one way of producing 
and maintaining living beings. No matter how dissimilar two 
animals may be in shape or even in fundamental constitution, their 
life processes, nevertheless, are essentially identical. Corresponding 
organs may not be the same in appearance or action but they accom- 
plish the same ends. The jaws may work up and down or they may 
work sidewise, but in either case they tear, crush, or chew the food 
before it is swallowed. The stomach may be of very different shape 
in two animals, but in each it changes the raw food into a soluble and 
an assimilable condition. The blood may be red or colorless, con- 
tained in tubes or not, but it always serves to distribute the prepared 
food which diffuses into it from the alimentary canal. The situa- 
tion of the central nervous system and the arrangement of its parts 
may be absolutely unlike in two organisms, but it regulates the func- 
tions of the organs and coordinates the actions of the muscles just 
the same. 

Hence, in studying the honey bee we shall find, as we naturally 
expect to find, that it possesses mouth organs for taking up raw food, 
an alimentary canal to digest it, salivary glands to furnish a digestive 
liquid, a contractile heart to keep the blood in circulation, a respira- 
tory system to furnish fresh oxygen and carry off waste gases, ex- 
cretory organs for eliminating waste substances from the blood, a 
nervous system to regulate and control all the other parts, and, finally, 
organs to produce the reproductive elements from which new indi- 
viduals are formed to take the places of those that die. 

The study of anatomy or the structure of the organs themselves 

is inseparably connected with a study of johysiology or the life 

. functions of the animal. While physiology is a most interesting 

and important subject, and, indeed, in one sense might be said to be 


the object of all anatomical research, yet the mere study of the 
structure of the organs alone, their wonderful mechanical adapta- 
tions, and their modifications in different animals forms a most fasci- 
nating field in itself, and besides this it gives us an insight into the 
blood relationships and degrees of kinship existing between the 
multitudes of animal forms found in nature. In the study of com- 
parative anatomy we are constantly surprised to find that structures 
in different animals which at first sight appear to be entirely differ- 
ent are really the same organs which have been simply changed in 
a superficial way to serve some new purpose. For example, the 
front wing of a bee and the hard shell-like wing cover of a beetle are 
fundamentally the same thing, both being front wings — that of the 
beetle being hardened to serve as a protection to the hind wing. 
Again, the ovipositor of a katydid and the sting of a bee are identical 
in their fundamental structure, differing in details simply because 
they are used for different purposes. Hence, in the study of anat- 
omy we must always be alert to discover what any special part cor- 
responds with in related species. In order to do this, however, it 
is often necessary to know the development of an organ in the 
embryo or in the young after birth or after hatching, for many 
complex parts in the adult have very simple beginnings in an imma- 
ture stage. 

Thus it becomes evident that the structural study of even one 
organism soon involves us in the subjects of anatomy, physiology, 
and embryology, and, if we add to this a studj^ of its senses, its 
behavior, and its place in nature, the field enlarges without limit. 
The student of the honey bee realizes that a lifetime might be spent 
in exploiting this one small insect. 

The differences between animals are much greater on the outside 
than on the inside. In the descriptions of the organs of the honey bee 
anyone will know what is meant by the " alimentary canal," the 
" nervous system," or the '' respiratory system," but the external 
parts are so different from those of animals with which we are more 
familiarly acquainted that no general reader could be expected to 
know what is meant by the names applied. Moreover, the bee and its 
allies are so modified externally in many ways that, at first sight, 
their parts look very different even from those of other insects. 
Hence, we shall give a preliminary account of the external structure 
of insects in general, for it is hoped that the reader will then more 
easily understand the special structure of the honey bee, and that the 
application of the terms used will appear more reasonable to him. 

Since all animals originate in an egg, the change into the adult 
involves two different processes: One ,is growth, which implies 
merely an increase in size, the addition of material to material; the 
other is development, which means change in shape and the produc- 



tion of a form with complex organs from the simple protoplasmic 
mass of the egg. The part of development that takes place in the 
eggshell is known as embryonic development; that which takes place 
subsequent to hatching is known as postemhryonic development. In 
insects there are often two stages in the postemhryonic development, 
an active one called the larval stage and an inactive one called the 
pupal stage. During the first of these the young insect is termed a 
larva; during the second, a pupa. When there is no resting stage the 
immature creature is often called a nymph. The final and fully de- 
veloped form is an adult, or imago. 

Since this paper is to deal only with the anatomy of the adult, the 
attractive fields of embryonic and postemhryonic development must 
be passed over, except for a few statements on 
fundamental embryonic structure, a knowledge 
of which is necessary to a proper understanding 
of the adult anatomy. 

When the embryo, in its course of development, 
first takes on a form suggestive of the definitive 
insect, it consists of a series of segments called 
metameres, or somites, and shows no differentia- 
tion into head, thoracic, and abdominal regions. 
Typically, each segment but the first is provided 
with a pair of latero-ventral appendages, hav- 
ing the form of small rounded protuberances. 
These appendages are of different sizes and take 
on different shapes in different parts of the 
body, for some of them are destined to form the 
antenn£E, some the mouth parts, others the legs 
and perhaps the cerci, while the rest of them 
remain very small and finally disappear. What 
we know of the embryology of insects is based 
on the observations of a number of men who 
have worked mostly on the development of dif- 
ferent species. Their observations are not all 
alike, but this is probably due in large part to the fact that the 
embryos of different insects are not all alike. Embryos have a very 
provoking habit of skipping over or omitting little and yet im- 
portant things m their development, but fortunately they do not 
all omit the same things. Therefore, by putting together all the 
reliable information we possess, we can make up an ideal embryo 
which would be typical of all insects. Such a generalized embryo is 
represented diagrammatically by figure 2. 

The first six or seven metameres very early begin to unite with 
one another and continue to fuse until their borders are lost. These 
consolidated embryonic segments form the head of the adult insect. 



2. — Diagram of a 
generaUzed insect em- 
bryo, showing the seg- 
mentation of the head, 
thoracic, and abdom- 
inal regions, and the 
segmental appendages. 


Observers differ concerning the fate of the seventh segment, but it 
is most probable that a part of it fuses with the sixth segment, thus 
taking part in the formation of the head, and that a part of it forms 
the neck or some of the neck plates of the adult. 

The appendages of these first seven segments form the antennse 
and mouth parts, except one or two pairs that disapjaear early in 
embryonic life. It is not certain that the first segment ever possesses 
appendages, but from it arise the large compound eyes and appar- 
ently also the upper lip, or labrum (Lm). The appendages of the 
second segment form the feelers, or antennae (lAnt) of the adult, 
those of the third (^Ant) disappear in insects, but they correspond 
with the second antennae of shrimps and lobsters. The appendages 
of the fourth segment form the mandibles (Md). Those of the 
fifth segment (Slin), when present, fuse with a median tonguelike 
lobe (Lin) of the following segment, and the three constitute the 
hypopharynx, or lingua of the adult. The next pair {IMx) form the 
maxillae, while the last {£Mx), or those of 'the seventh segment, 
coalesce with each other and constitute the adult labium, or lower lip. 

The bodies of the head metameres fuse so completely that it is 
impossible to say positively what parts of the adult head are formed 
from each. The last, as already stated, possibly takes part in the 
formation of both the head and the neck. Some embryologists at- 
tribute the plates which usually occur in this region to the last em- 
bryonic head segment, while others believe they come from the next 
segment following. Sometimes these plates are so well developed 
that they appear to constitute a separate segment in the adult, and 
this has been called the mAcrothorax. If this name, however, is 
given to the embryonic segment from which these plates are said to 
be derived, it must be remembered that it is not " thoracic " at all 
and belongs partly to the head. The name cervicum has been ap- 
plied to the neck region with greater appropriateness since it does 
not imply any doubtful affiliation with adjoining regions. What 
we really need, however, is not so much a name as more information 
concerning the development of the rear part of the head and the 
neck plates in different insects. 

The next three segments remain distinct throughout life in nearly 
all insects, but, since they bear the legs and the wings, they become 
highly specialized and together constitute the thorax. The indi- 
vidual segments are designated the prothorax, the mesothorax, and 
the metathorax. The legs are formed from the embryonic ap- 
pendages (fig. 2, IL, ^Z, 3L) of these segments, but the wings are 
secondary outgrowths from the mesothorax and metathorax and 
are, hence, not appendages in the strict embryological sense. 

The remaining segments, nearly always 10 in number, constitute 
the abdomen. The appendages of these segments, except possibly 


those of the tenth, disappear early in embryonic life in all insects, 
except some of the very lowest species, in which they are said to form 
certain small appendages of the abdominal segments in the adults. 

An adult insect is often described as being " divided " into a head, a 
thorax, and an abdomen, but this is not true in most cases. -AVhile all 
insects consist of these parts, the divisions of the body are usually 
not coincident with them. The prothorax in the adult is separated 
from the head by the neck and is very commonly separated from the 
mesothorax by a flexible membranous area. On the other hand, the 
mesothorax and metathorax are almost always much more solidly at- 
tached to each other, while, in most insects, the metathorax is solidly 
and widely joined to the first abdominal segment, though in the flies 
these latter two segments are usually separated by a constriction. In 
such insects as ants, wasps, and bees a slender, necklike peduncle 
occurs between the first and second segments of the abdomen, the 
first being fused into the metathorax so that it appears to be a part 
of the thorax. This' is the most distinctive character of the order 
Hymenoptera, to which these insects belong. 

The body wall of insects is hard on account of the thick layer of 
chitin which exists on the outer side of the true skin. Chitin is a sub- 
stance similar to horn, being brittle, though tough and elastic. It 
gives form and rigidity k) the body and affords a solid attachment for 
the muscles within, since insects have no internal framework of bones 
such as vertebrate animals have. The skin between the segments is 
soft and unchitinized and thus forms a flexible intersegmental mem- 
hrane which is often very ample and, in the abdomen, allows each seg- 
ment to telescope into the one in front of it. 

The chitin of each segment is not continuous, but is divided into 
plates called sclerites. The most important of these are a tergum 
above and a sternum below, but, in the case of the thorax, these two 
plates are separated on each side by another called the pleurum, which 
lies between the base of the wing and the base of the leg. Pleural 
plates are sometimes present also on the abdominal segments. These 
principal segmental plates are usually separated by membranous 
lines or spaces, which permit of more or less motion between them. 
Such lines are called sutures in entomology, though strictly this term 
should be applied only to the lines of fusion between adjoining parts. 

The terga, pleura, and sterna of each segment are furthermore 
subdivided into smaller sclerites, which may be termed tergites, pleu- 
rites, and stemifes, respectively. The sutures between them are 
sometimes membranous also, but most frequently have the form of 
impressed lines or narrow grooves. In such cases they are generally 
nothing more than the external marks of ridges developed on the 
inside of the body wall to strengthen the parts or to give attachment 
to muscles, Since these sutures are conspicuous marks on the outside 


of an insect, they are usually regarded as morphologically impor- 
tant things in themselves, representing a tendency of the tergum, pleu- 
rum, or sternum to separate into smaller plates for some reason. The 
truth about them would appear to be just the opposite in most cases — 
they are the unavoidable external marks of an internal thickening 
and strengthening of the plates. In a few cases they may be the 
confluent edges of separate centers of chitinization. Hence, most of 
the sutural lines in insects apjDear to signify a bracing or solidifying 
of the body wall rather than a division of it. 

Since the body wall of insects is continuous over all the surface it 
contains no articulations of the sort that occur between the bones in 
th'e skeleton of a vertebrate. Although insects and their allies be- 
long to the class of animals known as the Articulata, j^et an articu- 
late articulation is simply a flexibility — two chitinous parts of the 
exoskeleton are movable upon each other simply bj^ the intervention 
of a nonchitinized, flexible, membranous part. While there are often 
special ball-and-socket joints developed, these are always produced 
on the outside of the membranous hinge and simply control or limit 
the movement of the articulation. 

The head of an adult insect is a thin- walled capsule containing the 
brain, the ventral head ganglion of the nervous system, the pharynx 
and anterior part of the cesophagus, the tracheal tubes, and the 
muscles that move the antennae and the mouth parts. Its shape varies 
a great deal in different insects, being oval, globular, elongate, or 
triangular. In some it is flattened dorso-ventrally so that, the face is 
directed upward and the mouth forward, but in most, including 
the bee, it is flattened antero-jDosteriorly so that the face looks for- 
ward and the mouth is directed ventrally. In a few it is turned so 
that the face is ventral. The walls of the head are usually divided 
by sutures into a number of sclerites, which in general are located 
and named as follows: The movable transverse flap forming the 
upper lip is the labrum. Above it is a sclerite called the clypeus, 
which is a part of the solid wall of the head and carries the anterior 
articulations of the mandibles. The clypeus is sometimes divided 
transversely into an anteclypeus ("clypeus anterior," "epistoma") 
and into a post-clypeus ("clypeus posterior"). Above the clypeus 
is the front, a plate usually occupying the upper half of the face 
between the compound eyes and carrying the antennae. The top of 
the head is called the vertex, but does not constitute a separate scle- 
rite. The sides of the head below the compound eyes are often sepa- 
rated by sutures from the anterior and posterior surfaces and are 
known as the gence. The back of the head is formed by the occiput, 
which surrounds the large opening or foramen magnum that leads 
from the cavity of the head into that of the neck. The parts pos- 
terior to the gense, carrying the posterior mandibular articulations. 


are sometimes separated from both the occiput and the gense and are 
known as the postgena'. In a few insects, especially beetles, one or 
two median plates occur in the ventral wall of the head posterior to 
the base of the labium. These are the gular sclerites. Finally, small 
plates are sometimes found about the bases of the antennae and be- 
tween the bases of the mandibles and the genaj. The latter have 
been termed the trochantins of the mandibles. The term epicranium 
is often used to include all the immovable parts of the head, but is 
frequently applied only to the dorsal parts. Most of these sclerites 
preserve a pretty definite arrangement in the different orders, and 
they are probably homologous throughout the entire insect series, 
though they are in some cases very much distorted by special modi- 
fications and are often in part or wholly obliterated by the disap- 
pearance of the sutures. Embryologists are coming to the conclu- 
sion that the sclerites of the head have no relation to the primitive 
segments. The latter very earlj' consolidate into a head with a con- 
tinuous wall, while the sutures defining the sclerites are formed 
later. Some of the older entomologists were led, from a study of 
the sclerites, to suppose that the head consisted of a number of seg- 
ments, but it has been shown that these anatomical segments do not 
correspond with the embryonic ones. 

The appendages growing from the front of the face are the 
antennae (fig. 9A, Ant) or " feelers " and consist of a series of joints 
or segments. 

At the lower edge of the face is the front lip or Idbrum (fig. 9A, 
Lm) , behind which are the median epipharynx, the paired mandibles 
(Md) and maxilla}, the median hypopharynx, and the labium or under 
lip. All these organs together constitute what are known as the 
mouth parts or trophi. They vary greatly in shape and appearance 
in different insects according to the nature of the food, but their 
typical form is usually taken to be that shown by the lower insects 
which feed on solid food and have biting mouth parts. Figure 3, 
representing the jaws and lips of the common black cricket, is given 
as an example of generalized insect mouth parts. 

The labium (fig. 9A, Lm) is usually a simple transverse flap in front 
of the mouth, being developed, as already shown, from a similarly 
situated lobe on the first segment of the embryo (fig. 2, Lm). 

The epipharynx (fig. 19, Ephy) is a sort of dorsal tongue, and is 
situated on the membrane leading into the mouth from behind the 

The mandibles (figs. 3A; 9A, Md) are typically formed for 
biting, being heavy organs situated immediately behind the labrum 
and working sidewise on a hinge articulation with the head. Their 
cutting edges are usually notched and toothed, though smooth m the 
worker bee. 




The maxillse (fig. 3 B and B) are complicated appendages in their 
typical form. Each consists of a principal piece called the stipes {St) , 
which is hinged to the head by means of a smaller basal piece, the 
cardo (Cd). Terminally the stipes bears an outer lobe, the galea 
{Ga), and an inner lobe, the lacinia (Lc). On the outer side, at the 
base of the galea, it carries a jointed appendage called the maxillary 
palpus (Pip). 

The hypopharynx (fig. 3 C" and D, Hphy) is a median, ventral, 
tonguelike organ, called also the lingua, situated either on the upper 
surface of the labium or on the membrane between this organ and the 
mouth. It is de- 
veloped principally 
from a median lobe 
of the head of the 
embryo behind the 
mouth (fig. 2, Lin), 
but some entomol- 
ogists claim that it 
is compounded of 
this lobe and two 
smaller lateral ones 
developed from the 
appendages of the 
fifth embryonic 
head segment (fig. 
2, Slin) , the super- 

The labium (fig. 
3 C and D) consti- 
tutes the under lip 
of the adult, but it 
is formed from the 
two appendages of 
the seventh segment in the embryo, which fuse with each other. For 
this reason it is often called the second maxilla. It consists of a basal 
suhTnentum (Smt) bearing the mentum {Mt), which in turn carries 
three parts, a median ligula (Lg) and two lateral palpigers (Pig). 
The latter support the labial palpi (Pip), while the ligula bears four 
terminal lobes, of which the median ones are called the glossce {Gls) 
and the lateral ones the paraglossai {Pgl). If we should cut the 
labium into two parts along its midline we should see that even in 
the adult stage each half is very similar to one maxilla. The only 
discrepancy to be noticed in the example given (fig. 3) is that, there 
22181— No. IS— 10 2 

Fig. 3. — Example of generalized insect mouth parts, from 
common black cricket {Gryllus pcnnsylvanicus) : A, man- 
dibles ; B, B, maxllte, ventral view ; C, labium or second 
maxillse, ventral view ; D, labium, lateral view. 


is no maxillary palpiger, but many insects possess a corresponding 
part in the maxilla, frequently distinguished as the palpifer. 

The neck or cervicum is usually a short membranous cylinder which 
allows the head great freedom of motion upon the thorax. In nearly 
all insects its lateral walls contain several small plates, the cervical 
sclerites, while, in many of the lower species, dorsal, ventral, and 
lateral sclerites are present and highly developed. As already stated, 
the origin of these plates is doubtful. Some entomologists would 
derive them from the prothorax, others think they come from the 
last head segment, while still others think that they represent a 
separate segment. Only pure anatomists, however, entertain this 
last view and call this supposed segment the " microthorax," for 
embryologists have not yet reported a metamere between the labial 
segment and the prothoracic segment. Most embryologists who have 
studied the subject admit that some of the cervical sclerites may be 
formed from the last embryonic head somite which carries the labium 
and probably forms a part of the back of the head. Therefore, if 
it is desirable to retain the word microthorax as a name for a true 
segment, it can be applied only to this labial metamere."^ 

The thorax, as has already been stated, is a distinct anatomical 
region of the body rather than a " division " of the body, since it car- 
ries both the legs and the wings and contains the large muscles for 
each. Since the prothorax does not possess wings, it is not so highly 
developed otherwise as the two wing-bearing segments, and is, indeed, 
generally reduced in some ways, some of its parts being frequently 
rudimentary. Therefore we shall base the following description of 
a typical segment on the structure of the wing-bearing segments. 

A typical thoracic segment, then, presents four surfaces, as does also 
the entire body. These are a dorsum above, a venter below, and a 
latus * on each side. From these names we have the terms " dorsal," 

" In a former paper on the thorax of insects (Proc. TJ. S. Nat. Mus., XXXVI, 
1909, pp. 511-595) the writer probably drew a too dednite conclusion on the 
subject of the " microthorax." The origin of the neck sclerites has probably 
never yet been actually observed. Comstock and Kochi (Amer. Nat, XXXVI, 
1902, pp. 13^5), in summarizing the segmentation of the head, accredited 
the gular and cervical sclerites to the labial segment, but did not recognize the 
latter as taking part in the formation of the true head capsule. Riley, how- 
ever, in his study of the development of the head of a cockroach (Amer. Nat, 
XXXVIII, 1904, pp. 777-810), states that in Blatta the labial segment does 
form a part of the back of the head and that the posterior arms of the 
tentorium are derived from it. B'orner (Zool. Anz., XXVI, 1903, pp. 290-315) 
and Crampton (Proc. Acad. Nat. Scl. Phlla., 1909, pp. 3-54) believe that the 
cervical sclerites are. derived principally from the prothoracic segment. The 
notion that they constitute a separate segment, the " microthorax," equivalent 
to the maxilliped segment of the centipedes, has been elaborated principally 
by VerhoefE in his numerous writings on the Chilopoda and Dermaptera. 

'' The writer Introduces this word here because he knows of no other term 
applied to the side of the segment in this sense. 



"ventral," and "lateral." The chitinous parts of the dorsum con- 
stitute the tergum; of the venter, the sternum; and of the latus, the 

The tergum of the -^A-ing-bearing segments usually consists of 
two plates — a front one or true notum (fig. 4, N) carrying 
the wings, and a posterior one, which the writer has termed the 
fostnotum or pseudonotum (PJV), having no connection with the 
wings. The first is often more or less distinctly marked into three 
transverse parts called the prescutum (Psc), scutum (Set), and scu- 
tellum (Scl). In such cases the exposed part of the postnotum is 
called the postscutellum [Pscl). From either the anterior or the pos- 
terior margin of the tergum, or from 
both, a thin transverse plate projects 
downward into the interior of the 
thorax for the attachment of muscles. 
These plates are the phragmas {Aph 
and Pph). The notum supports the 
wing on each side by two small lobes, 
the anterior and posterior notal tcing 
processes {ANP and PNP). Behind 
the latter is the attachment of the 
axillary cord (AxC) or basal ligament 
of the wing. A large V-shaped ridge 
on the under surface of the notum hav- 
ing its apex forward is the " entodor- 
sum." (A better name would be 

The pleurum consists principally of 
two plates, the epistemum (fig. 4, Eps) 
and the epimerum {Epm)' lying before 
and behind a vertical groove, the pleural suture (PS) , which extends 
from the pleural coxal process (CxP) below to the pleural wing 
process (WP) above. The pleural suture marks the position of a 
heavy internal ridge, the pleural ridge or entopleurum. The epi- 
merum is connected with the postnotum (PJV) behind the base of the 
wing. These parts occur in almost all insects. In some of the lower 
ones another plate is present in front of the episternum which may 
be called the preepisternum {Peps)."- Lying along the upper edge of 

" Objection may be made to the use of the term " preepisternum " on the 
ground that it combines a Latin prefix with a word compounded of Greek ele- 
ments. The same may be urged against " prephragma," " postphragma," "pre- 
paraptera," and " postparaptera," words introduced by the present writer in a 
former paper on the thorax (Proc. TJ. S. Nat. Mus., XXXVI, 1909, pp. 511-595). 
However, we are barred from making up equivalent terms with the Greek pre- 
fixes pro and meta because these are used to designate the first and the third 


Fig. 4. — Diagram of generalized 

thoracic segment, left side. 


the pleurum and associated with the under surface of the wing base 
are several small plates known as the paraptera (P) ." Two lie above 
the episternum in front of the pleural wing process and are the 
episternal paraptera or preparaptera {IP and 2P), while one or 
occasionally two are similarly situated behind the wing processes 
and are the epimeral paraptera or postparaptera {3P and ^P). The 
preparaptera afford insertion for the muscle concerned in the exten- 
sion and pronation of the wing. 

The coxa (Cx), or basal segment of the leg, is hinged to the seg- 
ment by a dorsal articulation with the pleural coxal process (CxP), 
and by a ventral articulation (TnC) with a plate called the trochan- 
tin (Tth) lying in front of it and connected above with the lower 
end of the episternum {Eps) . Hence, while the leg is of course con- 
tinuous all around its base, by means of membrane, with the body- 
wall, its movement is limited to a hinge motion by these two special 
articulations of the chitin. 

The sternum or ventral plate of the segment is not so complicated as 
are the tergum and pleurum. It is often divided transversely into 
three parts, however, and some authors say typically into four. These 
parts have been named the presternum, (Ps), sternum proper (S), 

segments of the thorax or their respective parts. Entomologists have already 
established the system of referring a part to the front or back of any individual 
segment by the Latin prefixes pre (or prce) and post as used in " prescutum," 
" presternum," " postscutellum," and " poststernellum." Furthermore, pre and 
post are so indiscriminately used in English combined with Latin, Greek, and 
even Anglo-Saxon words that they may be regarded as general property. 
Hence, in order not to sacrifice an anatomical system, which certainly needs 
to be fostered In every way, the writer has preferred to sacrifice strict gram- 
matical rules by applying pre and post, regardless of the origin of the noun 
in the case, to designate anterior and posterior parts of the same segment. We 
already use such hybrid terms as " presternum," " mesotergum," and " meta- 

.The name " preepisternum " has been applied by Hopkins (Bui. 17, Pt. I, 
technical series, Bur. Ent., TJ. S. Dept. Agr., 1909) to a part of the mesepister- 
num of Dendroctonus — a plate apparently not homologous with the preepisternal 
element of the thorax in primitive insects. 

"The name "parapterum" is taken from Audouin's term paraptdre (Ann. 
des Sci. Nat., I, 1824, pp. 97-135, 416-432), and its application, as used by the 
present writer, is based on Audouin's definition given in his Chapter III, 
" Consider ationes generates svr le Thorax," where he says (p. 122) : "Finally 
there exists a piece but little developed and seldom observed, connected with 
both the episternum and the wing. It is always supported by the episternum 
and is sometimes prolonged ventrally along its anterior margin, or again, 
becoming free, passes In front of the wing and may even come to lie above 
the base of the latter. At first we designated this sclerite by the name of 
Eypoptire but on account of its change of position relative to the wing base 
we now prefer the name of PAEAPTi;EE." The first part of his description leaves 
no doubt that Audouin referred to the little pleural plate beneath the front 
of the wing which is usually very inconspicuous except in carefully dissected 


stemellum {SI), and poststemellum (Psl). In some of the lower 
insects a plate (x) occurs at each side of the presternum or of the 
sternum which seems to fall in line with the preerpisternum of the 
pleurum. This has been variously called a part of the presternum, 
the coxosterniim, an accessory sternal plate, and the sternal laterale. 
The inner surface of the 
sternum carries a large 
two-pronged process 
called the furca or ento- 

This plan of structure 
for the mesothorax and 
the metathorax prevails 
throughout all insects. 
The honey bee probably 
presents the greatest de- £"¥■' 

i Jt -i u i- Fig. 5. — Typical insect leg. 

parture from it, but even 

here the modification consists principally of a suppression of the 

sutures of the pleurum resulting from a condensation of the parts. 

The leg (fig. 5) of an adult insect consists of a number of joints 
or segments. It is attached to the body, as just described, by a thick 

specimens. In such preparations, however, one finds that there are in most 
cases two sclerites here instead of one, and, furthermore, that one or occa- 
sionally two others are similarly situated beneath the rear part of the wing 
base behind the pleural wing process. The present writer has, therefore, 
made the term " paraptera " cover this whole row of little plates, distinguish- 
ing those before and those behind the pleural wing process by the designations 
given above. 

In the latter part of Audouin's definition it would seem that he may have 
confused the rudimentary tegula as it exists in some insects with the parapte- 
rum, but even this is not probable since he says it is always connected with 
the episternum, which is never true of the tegula. In his description of the 
thorax of beetles, Dytiscus, Caraius, Buprestis, and Giirculio, it is evident 
that he regards the anterior upper part of the episternum as the parapterum 
fused with the latter plate. In fact, in each case he definitely states that such 
is the case and, in describing Dytiscus circumflesous, he says (p. 420) : "The 
episternum, the parapterum, and the epimerum all fuse dorsally and constitute 
a support for the wings and tergum." While Audouln is undoubtedly mis- 
taken in this homology, especially in the mesothorax, he at least shows that 
his " paraptfire " is a part of the pleurum. Hence modern writers such as 
Packard and Folsom who make the term " paraptera " synonymous with 
" tegulas " are certainly wrong. The tegula is a dorsal scale or its rudiment 
at the humeral angle of the wing, while the parapterum is a co-existent scl«- 
rite below this part of the wing base. The present writer agrees with Comstock 
and Kellogg, who, in their Elements of Insect Anatomy (first edition), define 
the little sclerite in front of the base of the wing in the locust, articulated to 
the dorsal extremity of the episternum, as the " parapteron," though in this 
Insect there are here really two of these parapteral plates instead of one. 



basal joint called the coxa {Ox). Beyond this is a smaller joint 
called the trochanter {Tr), this is followed by a long and strong 
segment, the fem/tir (i?"), which extends outward from the body, while 
bending downward from its distal end is the long and slender tibia 
(Tb), followed finally by the foot, or tarsus (Tar). The tarsus itself 
consists typically of five small segments of which the last bears a pair 
of cluws (Cla). The under surfaces of the tarsal joints are often 
provided with small cushions or pads called ptilviUi. Those between 
the claws are generally specially prominent and are called the 
empodia {Emp). The leg varies greatly in shape in different in- 
sects but usually preserves all of these parts. The segments of the 
tarsus, however, are frequently reduced in number. 

The adult wing is a thin expanse of memhranc sujjported by hollow 
branching rods called i^eins. It originates as a hollow outgrowth of 
the body-wall, but soon becomes flattened out dorso-ventrally and the 

Fig. 6. — Diagram of generalized insect wing and its articulation to first plate (A') of 

the tergum, 

contained trachess or air tubes mark out the courses of the veins. 
These veins form various j)atterns in different insects, but they can all 
be derived by modification from one fundamental plan. This plan is 
shown diagrammatically by figure 6. The first vein, which usually 
forms the anterior margin of the adult wing, is the casta ( G) . The 
next vein is the subcosta {Sc), which in typical cases divides into 
two branches {Sc^ and 8c^). The third and usually the principal 
vein is the radius {R). It divides dichotomously into five branches 
{R^ to R^), the anterior branch of the first fork remaining single. 
The next vein is the media (M), which forms four branches {M^ to 
yl/4). The fifth is the cubitus (Cu), which again is two-branched. 
The remaining veins are called the anaJs and are designated indi- 
vidually as the first anal (lA), second anal {^A), etc. 

Several cross-veins of common recurrence should be noted. The 
first is situated near the base of the wing between the costal and 
subcostal veins and is known as the humeral cross-vein. A second 


occurs between the radius and the media near the center of the wing 
and is called the radio-medial cross-vein. Another one, the inedio- 
cuhital, is similarly located between the media and the cubitus, 
while a fourth, called the median, occurs between the second and 
third branches of the media. The areas of the wing surface inclosed 
by the veins, the cross-veins, and the margins of the wing are known 
as the cells. 

A great many different names are applied by different entomolo- 
gists to the veins of the wings, both of the same and of different 
insects. The nomenclature here given is the one first consistently 
applied by Comstock and Needham and now used by a large number 
of entomologists working in different orders of insects. 

The wing is articulated at its base (except in mayflies and dragon- 
flies) to the anterior and posterior wing processes of the notum 
(fig. 6, ANP and PNP) and to the wing process of the pleuruni (fig, 
4, WP) by several small articular sclerites called axillaries. Two 
of these, the -first {lAx) and the fourth {Ii.Ax), form a hinge with the 
anterior and the posterior notal wing processes, respectively, while 
the second {2 Ax) articulates below with the wing process of the 
pleurum, constituting thus a sort of pivotal element. The third axil- 
lary (SAx) intermediates between the bases of the anal veins and the 
fourth axillary — except when the latter is absent (as it is in nearly 
all insects except Orthoptera and Hymenoptera) , in which case it 
articulates directly with the posterior notal process. The thin mem- 
brane of the wing base may be called the axillary membrane {AxM) . 
On its anterior edge is a hairy pad, the tegula (Tg), which is some- 
times a large scale overlapping the humeral angle of the wing. The 
posterior margin of the axillarj' membrane is thickened and may be 
called the axillary cord (AxC) or hasal ligament of the wing. 

The base of the costa is not directly associated with any of the 
axillaries, but is specially connected by tough membrane below with 
the episternal paraptera. The subcosta abuts against the end of 
the curved neck of the first axillarj^ The radius is either attached 
to or touches upon the anterior end of the second. The media and 
cubitus are usually associated with each other at their bases and also 
more or less closely with one or two median plates (m) in the wing 
base. These plates, however, are not of constant shape and occur- 
rence as are the articulating axillaries. The anals are generally 
attached to the outer end of the third axillary, which acts as a lever 
in the folding of the wing. 

A few insects have a generalized wing almost identical with the 
diagram (fig. 6), but most of them depart from it in varying degrees. 
Few go so far, however, as the honey bee, whose venation is very 
different, but yet the fundamental basal structure is the same even 


here, as Avill be shown in the special description of the wing of the 

The abdomen consists almost always of 10 segments. There are 
never any more than this number well developed in adult insects, and 
if there are fewer the reduction is due to a modification of the ter- 
minal segments to accommodate the external organs of reproduction. 
The posterior opening of the alimentary canal is at the end of the 
tenth segment, which carries also two small appendages at the sides of 
the anus. These are called the cerci (fig. 8, Cer) . In some insects they 
are short, styletlike processes, in others they are long and many 
jointed, while in many they are absent. The cerci are supposed to 
be developed from the embryonic appendages of the tenth segment, 
although, on the other segments, these appendages disappear before 
the embryo hatches, except in some members of the lowest wingless 
order of insects, which have a pair of cercuslike appendages on each 
segment of the abdomen. 

Each abdominal segment presents a tergum above and a sternum 
below; the former usually also reaches far down on the sides and 
overlaps the edges of the sternum. In some insects one or more small 
pleural plates intervene between the tergum and the sternum, but 
the abdominal pleura are never developed in any way suggestive of 
a thoracic pleurum. Very frequently there is present an upper 
pleural plate, or epipleurite, adjoining the edge of the tergum and a 
lower, or hypopleurite, adjoining the edge of the sternum. The line 
separating these two sclerites, however, is horizontal and can not 
correspond with the vertical suture of a thoracic pleurum between the 
episternum and the epimerum extending from the base of the leg 
to the base of the wing. 

The most complicated structures on the abdomen are the external 
organs of reproduction. In the male these serve as clasping organs 
and take on a great variety of forms in different species. The organs 
in the female form an ooipositor and are of much more definite and 
constant structure. 

The ovipositor (fig. 8), in its most perfect development, consists of 
three pairs of long, closely appressed bladelike processes called 
gonapophyses {IG, 2G, 3G). These six pieces fit neatly together and 
form an organ by means of which the female makes a hole in the 
ground or in the bark of a tree, or punctures some other insect, and 
then places her eggs in the cavity thus produced. An interesting fact 
in this connection is that the sting of a wasp or bee is simply a modi- 
fied ovipositor. This can be proved by a comparison of the organs 
themselves or by a study of their development. Each is formed from 
six little peglike processes that grow out from the sterna of the eighth 
and ninth abdominal segments of the larva or young soon after hatch- 



ing (fig. 7, -?(r, ^(x, and 3G). At first there is only one pair of these 
processes on each of the two segments, but those on the ninth soon 
split each into two, thus producing two pairs on this segment. The 
opening of the aviduct {OvO) is on the 
eighth segment between the bases of the 
first gonapophyses. 

The ovipositor of the longhorned grass- 
hopper, shown by figure 8, may be taken as 
a typical example of this organ. The 

median pair of gonapophyses on the ninth 
segment {^G) remain slender and fuse at 
their bases into a small bulblike swelling 
open below {ShB). The pair from the 
eighth segment {IG) form two long blade- 
like pieces, which fit by sliding articula- 
tions upon the lower edges of the corre- 
sponding second gonapophyses {2G). The 
first can therefore be worked back and 
forth while they are braced and held in 
position by the second pair. The third 
gonapophyses {3G)\ or the outer pair of 
the ninth segment (the left one in figure 8 is shown as if cut off near 
its base), form two long flat blades which are closely appressed 
against the outer surfaces of the others. In the detailed study of 
the bee it will be shown how closely the structure of the sting corre- 
sponds in every way with that of this ovipositor. 

Fig. 7. — Diagram of terminal 
abdominal segments of a fe- 
male insect and early stage in 
development o£ gonapophyses 
(10, 20, and 30), from 
which is formed the ovi- 
positor of most Insects and 
the sting of wasps and bees. 

Sp jG 'G 

Fig. 8. — Example of a swordlike ovipositor, from a longhorned grasshopper (Cono- 
cephalus sp.), illustrating the fundamental similarity of structure with the sting of the 
bee, flg. 36. 

Some entomologists have supposed that the original two pairs of 
gonapophyses represent the embryonic appendages of the eighth and 
ninth segments, and they would thus establish a homology between 
the ovipositor or sting and the legs and mouth parts. It has been 
shown, however, that the true appendages of the abdominal segments 
disappear in embryonic life while the gonapophyses appear much 
later, during early nymphal or larval life. Furthermore, each pair 


of gonapophyses arises in a median depression on the ventral side of 
the segment while the true appendages are latero-ventral. Hence, 
the evidence is very much against this theory and the gonapophyses 
appear to be special secondary processes of the body. wall. 

All insects do not have ovipositors of the sort described above. 
Flies, beetles, moths, and butterflies do not. Such insects simply 
drop their eggs from the orifice of the oviduct or deposit them in 
masses upon ■the external surfaces of various objects. In some of 
the flies, however, the terminal segments are long and tubular and 
entirely telescoped into one another. They are hence capable of 
being protruded in the form of a long tapering tube having the open- 
ing of the oviduct near the tip. This enables the insect to deposit its 
eggs in deep crevices, but the structure is not a true ovipositor — it is 
simply the abdomen itself stretched out. 

Insects breathe through a series of small holes situated along each 
side of the body. These breathing apertures are called spiracles and 
they lead into a system of internal air tubes called trachew. There 
are nearly always 10 spiracles present on each side of the body. Two 
are located on the thorax, the first between the prothorax and the 
mesothorax, the second between the mesothorax and the metathorax, 
while the other eight are situated on the first eight abdominal seg- 
ments. Some embryologists believe that the spiracles of the pro- 
thorax move forward in early embryonic life and unite with each 
other in front of the hypopharynx to form the salivary opening, their 
tracheae constituting the salivary ducts. 

After this review of the general external structure of insects we 
may proceed to a more detailed account of the parts and organs of 
the honey bee. 


The head of an insect, as already explained, is a composite organ 
formed of six or seven primitive segments, each of which, except the 
first, typically bears a pair of appendages (fig. 2). The antennae are 
developed from the embryonic appendages of the second segment, 
the mandibles from the fourth, the maxillae from the sixth, and the 
second maxillae, or labium, from the seventh. The appendages of 
the third segment disappear in early embryonic life while those of 
the fifth segment, when the latter is present, fuse with a median 
tonguelike lobe of the next segment to form the hypopharynx of 
the adult. 


The general appearance and outline of the head of a worker bee 
are shown from before and behind by figure 9, A and B. In facial 
view the head is triangular, with the apex below. The side angles 




are rounded and capped by the large compound eyes {E). In the 
opposite direction the head is very much flattened, the greatest diame- 
ter being crosswise through the middle of the eyes. The face is con- 
vex, while the posterior surface is somewhat hollowed out and fits 
snugly upon the anterior end of the thorax. 

The large lateral eyes (fig. 9 A, E) are called the compound eyes, 
because each is composed of a large number of separate eye elements 
forming the little hexagonal facets visible on the surface. All of 
these facets together constitute the cornea, or the transparent outer 
surface of the eye, which in the bee is densely clothed with long hairs. 
The dark color of the eye is located in the deeper parts, but these will 
be described in the section dealing with the nervous system. On the 

Vx ten 


Gk'k' Pgl 

Fig. 9. — A, front view of head of worker bee with mouth parts (Pr6) cut off a short 
distance from their bases ; B, corresponding view of posterior surface of head. 

top of the head between the compound eyes are the three simple eyes, 
or ocelli {0), arranged in a triangle with the median ocellus in front. 

Between the lower halves of the large eyes and near the center of the 
face arise the antennae [Ant) , each of which is inserted into a small, 
circular, membranous socket of the head wall, and consists of a long, 
basal, 1-segmented stalk carrying a terminal H-jointed arm movably 
articulated to the stalk and generally hanging downward from it. 
(In the drone the terminal arm consists of 12 joints.) 

The mouth parts are attached at the lower part of the head, and 
consist of the mandibles {Md) laterally and the maxillm {Mx) 
and labium (Lb) mesially. The latter two include the set of elongate 
bladelike organs surrounding the protrusible " tongue," which to- 
gether constitute what is commonly known as the proboscis {Prb). 


When not in use the parts of the proboscis are bent back beneath 
the head. By referring to figure 9B, giving a posterior view of the 
head, it will be seen that the basal parts of both the maxillae (St) 
and the labium (Mt) are suspended in a large hollow on the back of 
the cranium. This may be called the cavity or fossa of the proboscis 
{PrhFs). Between the mandibles on the front of the head (fig. 
9A) is a transverse movable flap, the Idbrum (Lm), attached to the 
lower edge of the front wall of the head and constituting the upper 
lip. The mouth {Mth) lies behind the labrum and the mandibles 
close beneath it. 

Below the antennal sockets is a transverse, slightly arched suture 
(a) which turns downward on each side and extends to the inner 
angles of the bases of the mandibles. The area bounded by this 
suture is the clypeus (Clp) and the suture itself may be called the 
clypeal suture. 

On the posterior surface of the head (fig. 9B) is seen the pen- 
tagonal foramen magnum {For) by means of which the cavity of 
the head communicates with that of the thorax and through which 
pass the nerves, cESophagus, blood vessel, and tracheal tubes. A 
small rod {ten) inside the head arches transversely over the fora- 
men magnuiH^ cutting it into a dorsal and a ventral half. At each 
side of the foramen is a large pit (e) which marks the base of an 
internal chitinous beam of the head known as the mesocephalic pillar. 
The opposite end of this pillar unites with the front wall of the 
head on the clypeal suture below the antennas, where it produces 
another smaller pit (&). 

Below the foramen magnum and separated from it by a wide trans- 
verse bridge of the cranial wall is seen the large fossa of the proboscis 
(fig. 9B, PrhFs) having the shape of an inverted U. The side walls 
of this cavity are chitinous and from their upper edges are suspended 
the maxillae, while the base of the labium is contained in the mem- 
branous floor of the fossa. The base of the labium projects from the 
head beneath or behind the mouth opening and its dorsal surface 
forms the floor of a preoral cavity surrounded by the bases of the 
mouth parts and labrum. 

It will be seen from the above description that the head wall of the 
bee contains no suture except that bounding the clypeus and the one 
which separates the labrum from the latter. Many of the higher 
insects have the head wall completely continuous, showing no division 
at all into sclerites, but, in such forms as a grasshopper or cockroach, 
and, in fact, most of the lower insects, the head as well as the other 
parts of the body is made up of a number of plates. Hence this may 
be regarded as the primitive condition, and it is presumed that the 
head of the bee has been produced from one whose wall was divided 
by sutures into a number of distinct parts. Therefore the different 





regions of the bee's head may be named according to the sclerites with 
which they correspond in other insects. Thus, the part of the face 
above the clypeus and between the compound eyes may be called the 
front (fig. 9A, Ft), i\i& parts below the compound eyes the genm (Ge), 
and the top of the head the vertex ^ 

(Vx). The area on the back of the 
head around the foramen magnum 
may likewise be termed the occipital 
region (fig. 9B, Oc) and the parts be- 
hind the gense and the lower halves 
of the compound eyes the postgenm 


The worker, queen, and drone differ 
conspicuously in the shape and size of 
the head, as will be seen by comparing 
A, B, and C of figure 10. In these 
drawings the front has been removed 
in order to show various internal 
parts, which will be described later. 
While the head of the worker (A) is 
triangular in facial view, that of the 
queen (B) is more rounded and wider 
in proportion to its length. The head' 
of the drone (C) is much larger than 
that of the female and is nearly cir- 
cular in outline. In shape the head 
of the queen is intermediate between 
that of the worker and that of the 
drone, but in size it is somewhat 
smaller than the head of the worker. 
The eyes {E) of the worker and queen 
are about equal, but those of the drone 
are enormously enlarged and are 
broadly contiguous on the vertex and 
the upper part of the front. On this 
account the ocelli ( C ) of the drone are 
crowded down on the front nearer the 
bases of the antennae and the front 
itself is very much narrowed above. 
The antennae of the drone consist of 
13 segments, while those of the females 
have but 12 segments. The mandibles are largest proportionately in 
the queen and are very small in the drone. Those of the worker have 
a smooth terminal edge, while this edge is notched in the queen and 
the drone. The parts of the proboscis are much longer in the worker 


Fig. 10. — A, anterior view of head of 
worker, with front, antenna;, and 
proboscis removed ; B, correspond- 
ing view of head of queen ; C, same 
of drone. 



and capable of much more action than in the queen and drone, which 
are almost entirely dependent upon the workers for their food. 

The internal structure of the cranium may be studied best in a longi- 
tudinal section of the head (fig. 11). In order to prepare a section 
for this purpose imbed the head in paraffin and then carefully slice 
off one side with a sharp knife or razor just outside of the bases of 
the mandible and antenna. Holding the remainder in the block of 
paraffin or fastening the whole in a dish of water or alcohol, care- 
fully dissect away the soft parts from the head cavity so as to expose 

Fig. 11. — A, longitudinal section through head o£ worker between the median plane and 
outer edges o£ mandibles (Md) and antennae (Ant) of left side, all internal soft parts 
removed ; B, corresponding section through head of drone, except that the pharynx 
(Phy) and (Esophagus (CB) are not removed. 

the internal chitinous parts shown in figure 11 A and B. These 
figures, however, represent a slice of the head taken from between the 
median plane and the outer edges of the antennal and mandibular 
bases of the left side. Thus only the parts on one side of the mid- 
line are shown. Figure A is from a worker and Figure B from a 
drone. In the latter the pharynx and oesophagus are retained and 
the neck is not removed. Figure 20 shows the head cut open from 
above and the mouth parts removed. A specimen so ciit and boiled 
a short time in caustic soda or potash to remove the soft parts will 
be found a valuable adjunct to this study. 


The principal parts of the internal skeleton of the head, or ento- 
cranium, consist of two large, -oblique, strongly chitinous bars form- 
ing a brace between the anterior and the posterior walls of the head 
(fig. 11 A and B, Ten, showing the parts on the left side only, and 
fig. 19, Ten). These bars have been named by Macloskie (1881) the 
mesocephalic pillars. As already pointed out the base of each is 
marked externally by a conspicuous pit (fig. 9 B, c) laterad of the 
foramen magnum, and its facial end by a smaller pit (fig. 9 A, &) 
in the clypeal suture near the upper end of each side of the latter. 
The bases of these pillars are connected by the slender bar (fig. 11 A, 
ten), already noticed, arching over the foramen magnum (fig. 9 B, 
ten). This bar and the two pillars represent what is called in other 
insects the tentorium. In the embryo the tentorium is formed from 
tubular ingrowths of the head wall which unite internally and 
assume different shapes in different insects. Since the air tubes of 
the body also first appear as tubular ingrowths of the body wall, 
some entomologists have supposed that the hollow tentorial in- 
growths of the head represent the spiracular tubes of the head 
which are, otherwise, lacking. However, there is not sufficient evi- 
dence to support such a view as this, and there is no reason why the 
tentorium should not have been originally designed simply to give 
greater rigidity to the walls of the head where the latter support the 

The usual form of the tentorium in the lower insects is that of an 
X, with a large central body, situated like a brace across the lower 
part of the head, having two of the arms directed anteriorly and 
laterally and two directed posteriorly and laterally, and while the 
former are said to be ingrowths from the mandibular segment, there 
is some difference of opinion concerning the segment to which the 
latter belong. Eiley states that they are formed in the labial seg- 
ment of the cockroach and Carriere and Burger describe the same 
thing for the mason bee. Other authors have ascribed them to the 
maxillary segment, but they may, in later stages, lie in this segment 
and thus appear to belong to it, while they originated in the one 
following, having moved forward on account of the condensation 
of the back part of the head. The tentorium of the honey bee, 
consisting as it does of the two great mesocephalic pillars (fig. 11 
A and B, Ten) and the small arched bar ( ten) is so highly modified 
that it is hard to see just how its parts are to be homologized with 
the parts of an X-shaped tentorium. Probably the two pillars repre- 
sent the separated halves of the X, while the slender arch is an addi- 
tional structure. In any case we have not enough evidence to war- 
rant us in regarding the tentorial invaginations as modified tracheae, 
or their external pits as rudimentary spiracles. Similar processes 
extend inward from the walls of the thorax to strengthen it or to 
give attachment of muscles. Such processes in general form the 


entoskeleton and are individually called apodemes. Those of the 
head constitute the entocranium, those of the thorax the entothorax. 

The side walls of the fossa of the proboscis form two high, thin, 
vertical plates, as seen from the interior of the head (fig. 11), in 
front of the mesocephalic pillars. The posterior edge {d) of each 
of these plates is so much thicker than the rest of it in the worker 
that it appears at first sight to be a separate rod. Its upper end 
projects above the body of the plate as a free arm (e) to which is 
articulated the basal piece of the maxilla {Cd). It thus constitutes 
the maxillary suspensorium. (Macloskie includes under this term 
both the arm of the cranial wall and the cardo of the maxilla.) 

The head of the drone (fig. 11 B) presents, besides the parts de- 
scribed, a thin plate (/) depending from the vertex of the cranium 
along the line between the compound eyes. 

Besides these apodemes of the cranial wall itself there are others 
which project into the head cavity from the bases of the appendages 
to afford points of insertion for their muscles. These are specially 
developed in connection with the mandibles and will be described in 
the discussion of these organs. Still other internal chitinizations are 
developed in the walls of the pharynx, but these likewise will be 
described later. 


The antennae of the bee are the two slender, jointed appendages- 
movably attached to the center of the face, where each is inserted 
into a circular membranous area or socket just above the upper part 
of the clypeal suture. Their general shape and position are shown 
by figures 9 A, 11 A, and 19, Ant. Each is seen to consist of two 
parts, forming a prominent elbow with each other, and usually so 
held that the first or proximal part extends outward and upward 
from its frontal attachment and carries the other in a pendent posi- 
tion from its distal end. The first part thus forms a basal stalk, 
called the scape (figs. 9 A; 19, Scp)^ consisting of a single joint 
inserted into the antennal socket of the front by a prominent basal 
condyle bent toward the face. This articular knob is attached to 
the rim of the socket by a circle of membrane, but it is also pivoted 
on a slender peglike process projecting upward from the lower edge 
of the socket. Hence, while the flexible membrane allows each 
antenna to revolve freely in any direction, the latter is at the same 
time held firmly in position by the pivot. The antennae are moved 
by special sets of muscles inserted upon their bases within the head. 
The second or distal division of the antenna is cylindrical and longer 
than the first, forming a flexible flagellum, (fig. 9 A; 19, FT) hanging 
downward from the distal end of the scape. It is composed of 11 


small joints in the worker and queen and of 12 in the drone. The 
male antenna thus consists of 13 joints in all, while that of the female 
has but 12. The first joint of the flagellum is freely articulated to 
the scape, but the others do not have much play upon one another, 
though they give flexibility to the flagellum as a whole. 

Each antenna is a hollow tube containing the large antennal nerve, 
minute extensions of the tracheal system, and the small muscles which 
move the segments upon one another. 

Popularly the antennae of insects are known as the " feelers," be- 
cause they are constantly moved about in all directions with a nervous 
kind of motion as if the creature were feeling its way along by means 
of them. In fact " feelers " is a better name for these appendages 
than the scientific term, for there can be no doubt that the sense of 
touch is very highly developed in them and that by means of them 
insects acquire a great deal of information concerning their surround- 
ings and their companions. Moreover, a large mass of evidence 
derived from experiments shows unquestionably that the organs of 
smell also are located upon the antennse in a great many if not all 
insects, while some investigators believe that in some species they 
carry in addition the organs of hearing. 

The study of the senses of insects is a most elusive subject, and 
becomes more so the more we ponder on the results of experiments. 
In the first place, it is manifestly impossible for us to acquire any 
real knowledge of an insect's sensations, for what is to us an odor, 
11 taste, a color, or a sound may be something quite different to such a 
differently organized creature. We can, however, by experiments 
determine that some things which give us the sensation of an odor 
are perceived also by insects when placed near them. Also it can be 
shown that some of them distinguish substances of different taste in 
their food, and likewise that they perceive movement and distinguish 
the colors and in a vague way the outlines of objects. Furthermore, 
it is known that some of their perceptions are more delicate than ours, 
and that some insects at least see color where we see none. They may 
even possess senses of which we have no conception. 

Hence, while it can be positively stated that insects perceive differ- 
ences of touch, taste, smell, sound, and light, and act accordingly, we 
can not say what the sensations they acquire are like. In fact we 
do not know that they have conscious sensations at all. What looks 
like an action due to intelligent perception may be purely a reflex one, 
unaccompanied by any sensation. This of course involves the ques- 
tion as to whether such creatures or insects are possessed of conscious- 
ness or not — a question which can not be answered one way or the 

Understanding, then, that our knowledge of insect senses amounts 
only to this, that what gives us the sensation of light, sound, taste, 
22181— No. 18—10 3 


touch, or smell makes also some sort of an impression on the insect 
and varies in degree and kind much as it does in us, we may go on to 
a study of the senses located on the antennae. 

Here, again, however, we are confronted by a difficulty, for while, 
at first thought, it seems very easy to hold some strong-smelling sub- 
stance near the antennae of a beetle, ant, or bee and observe the evident 
displeasure with which the creature turns away, yet we may be en- 
tirely wrong if we conclude that the insect " smells " the substance 
that repels it. Strong-smelling, volatile liquids may simply produce 
pain in some of the delicate nerve endings of the antennae. Some 
other kind of a being, experimenting on our senses, might close up 
pur nose and mouth and prove that we smell by means of our eyes 
on observing the blinking we should perform when strong formalin 
or ammonia was held close to the face. Furthermore, irritant gases 
and volatile liquids affect the mucous membranes of our noses and 
throats in a way quite indejDendent from the odor that we perceive, 
and there is no reason why the same, may not be true of insects. As 
pointed out by Forel, experiments on the sense of smell should be 
made with odorous substances that the insect meets with in a state of 
nature, which would be princijjally the materials it feeds on. In- 
sects are indifferent to almost every mildly odorous substance not 
used as food, which, however, does not prove that they do not sniell 

Again, in many cases, it would be difficult to decide whether the re- 
sults of an experiment should be accredited to smell or sight. For 
example, every bee keeper knows that hungry bees are attracted to 
honey a long distance from their hives, and it would seem almost self- 
evident that they are guided by a sense of smell. Yet one might con- 
tend that they find the honey by sight, as, indeed, is claimed by a 
number of entomologists Avho have made experiments on the olfactory 
powers of bees. This question has been decided in some other insects 
by painting the eyes with some opaque substance or by removing the 
antennae, but the evidence is not conclusive on either side in the case of 

Experiments made by a large number of competent investigators, 
including Lubbock, Schiemenz, and Forel, have proved conclusively 
that the organs of the sense of smell in insects are located principally 
on the antennae. The most interesting of these experiments are per- 
haps those which Forel ( 1903 ) made on carrion-feeding beetles. He 
found the dead and putrid bodies of a hedgehog and a rat infested by 
a swarm of these beetles belonging to several genera. He collected 
more than 40 specimens from the carcasses and removed their an- 
tennae. Then he placed them all at one place in the grass and moved 
the dead bodies to a distance of 28 paces from the beetles where he 
concealed them in a tangle of weeds. Examination the next day 


revealed the fact that not one of the mutilated beetles had found the 
carcasses. Repeated experiments gave the same results — no beetle 
without its antennae was ever found on the dead animals, although at 
each examination new individuals of the several species were present. 
It might be supposed that the mutilation itself distracted the beetles 
to such an extent that they did not care to eat. In order to test this 
point Forel next cut off all the feet on one side of the body from a 
dozen intact beetles and changed the location of the dead bodies again. 
The next day five of this lot were found on the carcasses. 

The same results have been obtained from experiments on other 
insects. Ants distinguish between their comrades and enemies by 
means of their antennal sense organs. Males of the silkworm moth 
and many other moths and butterflies perceive the presence of 
the females and are guided to them by an evident sense of smell 
located on the antennse, for they fail completely to find them when 
these appendages are removed, although one immediately recognizes 
a female when placed in contact with her. 

Similar experiments have been made on the bee, testing the ability 
of the workers to find honey hidden from their sight. The results, 
according to Forel, seem, curiously enough, to indicate that bees can 
perceive odors but a very short distance from their heads. Forel 
found that hungry bees in a cage would pass and repass hundreds of 
times within a few millimeters of some honey concealed from their 
sight by a lattice without discovering it. They ate it greedily, how- 
ever, when the lattice was removed, though it had been perfectly 
accessible to them all the time. Forel believes that " bees guide them- 
selves almost exclusively by vision," and Lubbock holds the same 
opinion. At the same time it would probably be a very difficult mat- 
ter to convince many practical bee keepers that bees do not " smell " 
from long distances. It is a well-known fact that at times when nec- 
tar is scarce bees are attracted in large numbers to the houses of an 
apiary where honey is stored, though, when the natural flow is suf- 
ficient, they pay no attention to it. Tests of the olfactory sense should 
undoubtedly be made under natural conditions. Bees inclosed in a 
box with some honey concealed from their sight might not be able to 
locate it in such close quarters though they might be smelling it all 
the time. An odor in a room may so fill the air that it does not seem 
to come from any particular direction and we ourselves would have 
to exert our intelligence to discover its source. 

While, then, it does not seem probable that bees have such limited 
olfactory powers as some investigators claim their experiments indi- 
cate, it may be accepted as proved that the organs of smell are located 
principally on the antennae. It has already been stated that the sense 
of touch also is very highly developed on these organs, although in a 
less sensitive degree it is distributed over most of the other parts of 



the body. It is again specially developed on the palpuslike append- 
ages of the sting. (See figs. 36 and 37, StiiPlp.) Sections of a bee's 

antenna show that there are 
Hn. ,,^*^ on its surface a great number 

of minute structures of sev- 
eral different kinds, though 
all apparently are to be re- 
garded as modified hairs, 
which are undoubtedly the 
sense organs. Now the diffi- 
culty arises of deciding which 
of these to assign to the sense 
of touch and which to the 
sense of smell. Different au- 
thors have made such differ- 
ent interpretations of the 
sense organs of insects that 
the student attempting to get 
informatioji on the subject 
from books must soon be dis- 
couraged by their conflicting 
statements. But it must be 
realized that only intelligent 
guessing is possible where 
several senses are located on 
the same part. In the case of 
the bee some authors have 
ascribed even a third sense, 
that of hearing, to the an- 
tennae, but there is little evi- 
dence that bees possess the 
power of hearing. The senses 
of taste and touch are pos- 
sessed by the mouth parts, 
and some entomologists think 
that they contain organs of 
smell also. Thus, the organs 
of sight are apparently the 
only ones that can not be con- 
fused with some other sense. 
The best account of the 
antennal sense organs of the 
bee is that of Schiemenz (1883) , whose drawings are here reproduced 
(fig. 12) and whose text is the basis of the following descriptions. 
The organs consist, as before stated, of modified hairs and their basal 

Fig. 12. — Antennal hairs and sense organs 
(after Schiemenz). A, example of antennal 
hairs (Hr) imbedded in cuticle (Ctl) but 
having no nerve connection ; B, hollow hair 
containing prolongation of special cell (CZ) ; 
C, D, straight and curved tactile hairs con- 
nected with basal cells (CI) and nerve fibers 
(Nv) ; E, conical hair (Hr) sunljen in a pit 
(Ft) of the cuticle, probably an olfactory 
organ ; F, closed sac shut in by thin disc 
(/if) on surface of antenna and containing a 
delicately poised cell (CI) with nerve con- 
nection (Sv). 


insertions which are connected with the ends of nerve fibers. Some 
of them stand expbsed on the surface of the cuticle while others 
are sunken into, or entirely concealed within, pits of the integument. 
In addition to these, there are two other kinds of special hairs on 
the antennae which have no nerve connections, while, finally, the ordi- 
nary hairs, such as are found on all parts of the body, occur also on 
them, especially on the scape. 

The special hairs not provided with nerve endings are of two 
sorts. One is a solid curved or hooked hair (fig. 12 A, Hr) which 
is simply articulated into a socket of the cuticle {Gtl), while the 
other (B) is hollow and is situated over a channel through the cuticle, 
and contains a prolongation of a specially enlarged epithelial cell 
{Gl) lying beneath it. These hairs can not be regarded as sensory, 
since they have no communication with the central nervous system, 
and it is not clear just what purpose they do serve. 

The simplest sensory organ is a short, hollow, conical hair (C, 
Hr) arising directly from the surface of the cuticle, over a wide 
oprning through the latter, and containing the end of a sensory cell 
{CI) connected with a nerve fiber {Nv), which goes into the main 
trunk of the axial antennal nerve. A modified form of this organ 
consists of a curved hair (D, Hr) set into a small depression over 
the cuticular channel. Such hairs are probably tactile in function; 
that is to say, by means of them the bee can perceive that its antennae 
are in contact with some surface. The general integument is too 
thick and dense to allow of any sort of delicate touch sensation being 
communicated through it, but if one of these movable hairs brushes 
against an object the nerve within it must be at once stimulated. 
Tactile or touch hairs are distributed especially over the outer sur- 
face of the antenna* and at its apex, but occur also scattered over 
the other parts of the body and on the mouth parts. 

Microscopic sections of the antennae reveal still other organs 
which are not so apparent on the surface as the hairs just described. 
One of these is shown at E of figure 12. It consists of a small pit 
{Pt) in the integument, widened basally, and having a small papilla 
on its floor, in whose summit is the opening of a still deeper cavity 
which also expands toward its deeper end. This inner cavity is 
almost filled up by a conical plug {H7') which arises from its floor 
and ends just below the aperture into the outer pit. The plug con- 
tains a thick nerve ending which arises from a ganglion cell con- 
nected with the antennal nerve by a nerve' fiber. Ten or more of 
these sense organs occur on the terminal and the first three segments 
of the flagellum. It is evident that each is simply a sensory hair 
which has been doubly sunken into a cavity of the integument. 


As before stated, it has been conclusively proved by several investi- 
gators that bees perceive odors, and it is said that if the antennae 
are covered with shellac, bees can distinguish between distasteful 
substances only by means of the proboscis. Schiemenz and most 
other writers on the subject therefore conclude that the sunken cones 
are the organs of smell, since, being below the surface, they could not 
be organs of touch. Some other authors, however, among whom are 
Cheshire, regard these inclosed cones as hearing organs. They sup- 
pose that the sound waves of the air enter the pit, as into an ear 
cavity, and these set up a vibration in the cone which stimulates the 
attached nerve ending. However, the appearance of one of these 
cones would suggest that it is too stable a structure to be affected 
by sound waves, so the olfactory theory seems much more probable. 

Finally, Schiemenz describes the most specialized of all the anten- 
nal sense organs as a closed cavity {Pt) in the cuticle {Ctl) extend- 
ing into the hollow of the antenna as a long, curved, tapering sac. 
This is shown at F of figure 12. A nerve {Nv) enters the lower 
extremity of the pouch, expands slightly into a nucleated ganglion 
cell {CI), and then extends toward the top as a delicate spindle 
drawn out into a fine tapering point. The surface covering of the 
pit is a thin layer of chitin presenting several concentric light and 
dark rings surrounding a central disc {hr). Sections show that this 
appearance of rings is due to circular thickenings of the membrane, 
and Schiemenz points out that the central disc is probably a modi- 
fied hair, while the whole structure is to be regarded simply as a 
modification of a tactile organ such as that shown at D with the 
nerve- ending and its ganglion inclosed in a sac. These organs are 
most abundant on the antennae of the drones, where they are situ- 
ated, especially on the under surface, so close together that but little 
space is left between them for the tactile hairs, while in the workers 
and queens they are farther apart and are interspaced with many 
tactile hairs. Hence, whatever sense they accommodate must be 
much more highly developed in the males than in the females. 
Schiemenz described these organs, as well as the sunken cones, as 
organs of smell. He ascribed only the senses of touch and smell to 
the antennae, and both Cheshire and Cowan concur in his view of the 
closed pits. Arnhart (1906), however, argues that an organ of smell 
must be open to the air in order to permit the ingress of odor par- 
ticles. Such an organ is constituted by the sunken cones, but the 
closed pits have nothing to recommend them for an olfactory func- 
tion. Arnhart then further points out that the buried sacs, inclosing 
a delicately poised nerve-ending and covered by an external tym- 
panum, have all the mechanical elements of an organ of hearing. 
He finally argues that bees must hear, since they produce special 
sounds such as the piping of the queens, and that, since no possible 


organs of hearing have been discovered on any other part of the body, 
some of the antennal sense organs must be auditory in function. His 
conclusion from tliese premises is, of course, inevitable that the 
closed sacs on the antenna; are tlie hearing organs of the bee. What 
invalidates the argument, however, is the fact that no one has yet 
produced any actual evidence that bees joerceive sound. 

The following, then, may be stated as a general summary of the 
evidence concerning the antennal senses and their sense organs in 
the bee: (1) The antenna; are highly sensitive to touch and are the 
seat of the sense of smell. (2) They are covered by several kinds 
of minute structures which are modified hairs containing special 
nerve-endings. (3) By inference, it would seem certain that these 
are the sense organs, but we can only form an opinion, based upon 
their structure, as to which are tactile and which olfactory. (4) One 
set of organs does not appear to belong to either of these categories 
and their structure suggests an auditory function, but, in the absence 
of evidence that bees hear, the purpose of these organs must be re- 
garded as problematical. 


The mandibles (fig. 9 A, Md) are the dark, strongly chitinous 
appendages of the head, commonly called the jaws, situated at each 
side of the mouth, anterior to the base of the proboscis. In all in- 
sects with biting mouth parts the jaws work sidewise, each being- 
attached to the head by an anterior and a posterior articulation. 
They can thus swing in and out on a longitudinal axis in such insects, 
as the bee, that cai'iy the head with the mouth directed downward, 
or in the same way on a vertical axis in those that carry the head 
with ,the mouth forward. 

Both mandibular articulations are of the ball-and-socket type, 
although in the bee the socket is a very shallow one, the anterior 
consisting of a condyle on the outer angle of the clypeus fitting 
against a facet on the mandible, and the position of a facet on the 
lower edge of the postgena receiving a condyle from the mandible. 
The motion of the mandible is thus reduced to a hinge-joint move- 
ment, and, on this account, insects can only bite and crush their 
food; they can not truly chew it, since their jaws are incapable of 
a grinding motion. Each mandible is, of course, as pointed out in 
the introduction, really suspended from the head by a continuous 
membrane between its base and the cranium, being simply a modified 
saclike outgrowth of the head wall. The two articulations are pro- 
ductions of the chitin on the outside of this membrane. 

Figure 9 A shows the location and shape of the mandibles {31 d) 
of the worker as seen in a facial view of the head. Figure 11 A 




shows tlie ajipearance of the left mandible in side view, while the 
right one is shoAvn detached from the head in figure 13 A. The 
mandibles differ consjiicuously in size and shape in the three forms 
of the bee as already described and as shown in figure 10 A, B, and C. 
That of the worker is hollowed out somewhat on the distal half of 
its inner face (fig. 13 A, Md) forming a spoon-shaped organ, the 
edge of which is smooth and rounded. The mandibles of both the 
queen (fig. 10 B) and the drone (C), however, are pointed at the 
apex and have a conspicuous subapical notch. Those of the drone 

are smaller than those of 
either form of the female, 
but appear to be especially 
small on account of the 
great size of the drone's 
head. The mandible of the 
worker is undoubtedly to 
be regarded as the special- 
ized form, since the notched 
mandible of the drone and 
queen is of the ordinary 
Hymenopteran type. Both 
the drone and the queen 
are, under normal circum- 
stances, fed almost entirely 
by the workers, and they 
probably never have any 
use for their jaws as feed- 
ing organs. The queen 
needs her large, sharp- 
pointed mandibles for bit- 
ing her way out of the 
thick wax cell in which 
she is reared, but the 
drone, on the other hand, 
being reared in an ordinary cell resembling that of a worker, except 
in size, is easily able to cut through the thin cell cap with his com- 
paratively weak jaws. The workers, however, have numerous uses 
for their mandibles, such as biting through the cell caps, eating 
pollen, and modeling wax. The last is the especial function of 
the worker mandible, and probably it is to accommodate this jjur- 
pose that it has acquired its specialized spoonlike shape. 

Each mandible is jnoved by two sets of muscles within the head. 
The outer one constitutes the extensor muscle (fig. 13 A, EMd) and 
the inner the flexor muscle (RMcl). The latter is the stronger of 

Fig. 13. — A, right mandible of worljer, anterior 
view, witli extensor and flexor muscles (EAIrl 
and Rilcl) and mandibular glands (IMdOl) at- 
tached ; B, corresponding view of mandible of 
drone, with muscles cut off a short distance from 
their bases. 


the two, since all the work of the mandible falls upon it, the extensor 
being used simply to open the jiiw. While these muscles their 
origins on the walls of the head, they are not inserted directly upon 
the .mandibles, but on large apodemes (fig. 13 A, EAp and BAp) 
attached to the edges of the mandible. 

A gland opens at the inner margin of each mandible between the 
anterior articulation and the base of' the apodeme of the flexor 
muscle (fig. 13 A and B, IMdGl). In the worker it consists of a 
large sac covered with secreting cells lying within the front part of 
the head between the clypeus and the compound eye (fig. 10 A, 
IMdGl). These mandibular glands may be most easily studied by 
removing the front as shown in figure 10 A, B, and C. In order to 
do this, pull the head from the thorax and allow the prothoracic legs, 
which will usually come off with the head, to remain attached to it. 
Next melt a small hole in the bottom of a paraffin dish with a heated 
needle and fasten the head face upward into this, the attached legs 
helping to anchor the head in the paraffin. Cover the specimen with 
weak alcohol and by means of sharp needles remove the part of the 
front on either side between the clypeus and the lower half of the 
compound eye in the worker and drone and the entire front of the 
queen. In figure 10 the whole front is removed in all three forms in 
order to expose other internal parts of the head. 

The mandibular gland {IMdGl) is of greatest size in the queen 
(fig. 10 B) , though it is large in the worker (fig. 10 A and fig. 13 A) , 
but it is reduced in the drone (fig. 13 B) to a very small oval sac, 
Avhich is hidden by another gland {SGI) in front (fig. 10 C) . It was 
first described by Wolff (1875) as an olfactory mucous gland {Riech- 
SGhleimdrlisse) and was supposed by him to secrete a liquid which 
was poured upon the roof of the mouth in order to keep this surface, 
on which Wolff thought the olfactory organs were located, in a moist 
condition capable of absorbing odor particles. There is absolutely 
no evidence, however, of the presence of organs of smell in the mouth, 
and furthermore, as pointed out by Schiemenz (1883), the gland 
varies in the three forms of the honey bee according to the size of the 
mandible, which is proportionately largest in the queen and smallest 
in the drone. Of the three, we should expect the drone or the worker 
to have the sense of smell most highly developed, and hence, even if 
we did not know that the sense of smell is located in the antenna3, 
it would seem more reasonable to suppose that the glands of the 
mandibles are connected in some way with the functions of these 
organs themselves. 

The mandibles, as already stated, are used for eating pollen and as 
tools for manipulating and modeling wax. Therefore, according to 
Arnhart (1906), since the queen does not eat raw pollen, the product 



of the mandibular glands must be intended for softening the wax 
when it is worked in the jaws. The secretion of the glands is said 
to be very volatile and strong smelling and to have an acid reaction. 
It is probably entirely possible that it may have a solvent effect upon 
the wax, or even, when mixed with it, change somewhat the chemical 
com2DOsition of this substance; in fact, some investigators claim that 
the wax of the comb differs chemically from that freshly taken from 
the wax plates. Even this explanation, however, does not seem en- 
tirely satisfactory, for the only occasions on which the queen has any- 
thing to do with wax is when she gnaws her way out of her cell after 
hatching or bites her way into the cells of young queens in order 
to sting them. However, these occasional uses by the queen of her 
mandibles appear to be important enough to maintain the large size 
of these organs in the queen, and it may be reasonable to assume that 
the demand upon their glands is likewise a large one when it does 

occur. Yet the mandibles of the 
queen are toothed and sharp 
pointed, which should provide her 
with sufficient cutting power both 
to emerge from her own cell and to 
enter the cells of other queens, and 
so, on the whole, the opinion of 
Schiemenz that the secretion of the 
mandibular glands is merely sali- 
vary in function would seem to be 
the simplest explanation and the 
most logical one. However, an 
actual test should certainly be made 
to determine whether the worker's 
manipulation of the wax with her mandibles produces any change in 
it, and to discover whether the queen simply bites her way mechan- 
ically through the wall of the cell or at the same time softens the wax 
by a secretion from her mouth. The male in any case has little use 
for his mandibles, and the glands are so small that they must certainly 
be functionless. 

A second mandibular gland (fig. 14, 2MdGT) is present in the 
worker. It consists of a delicate, flattened, racemose mass lying 
against the internal face of the wall of the fossa of the proboscis, 
whose duct opens into the mouth cavity at the posterior inner edge 
of the mandible. This gland was first described by Bordas (1895) as 
the internal mandibular glared. According to him, it corresponds 
with a similar gland in the Bombidse (bumblebees) and in the Ves- 
pidas (yellow jackets) and to the maxillary glands of other Hy- 
menoptera. Nothing is known of its secretion. 


Fig. 14. — Internal mandibular gland 
(2MdGl) of worker, lying against inner 
wall of postgena iPge) and opening 
••(Bet) at inner edge of base of man- 



The conspicuous group of mouth appendages in the honey bee, 
forming what is commonly known as the yroboscis (fig. 9 A, Prb), 

Fig. 15. — Mouth parts of the worker: A, tip of glossa, showing labellum (TJbV), guard 
hairs (S"r), and ventral groove (fc) ; B, same, from above; C, small piece of glossal 
rod (>•) with adjoining parts of walls (or) of glossal canal attached, showing ventral 
channel (!) guarded by rows of hairs. D, parts forming the proboscis, labium in middle 
and maxillas at sides, flattened out, ventral view ; B, cross section of glossa showing its 
invaginated channel (Lum) and position of rod (r) along its dorsal wall, and likewise 
position of channel (I) of rod along median line within the glossal channel ; F, end of 
mentum (.Mt) and bases of ligula (Liji and labial palpi (LliPlp), showing opening 
of salivary duct {SaWO), dorsal view; G, lateral view of proboscis showing parts on 
left side; H, lateral view of glossa (GJs) with its rod (r) torn away at base showing 
attachment of retractor muscles {2BMcl). 

by means of which the bee takes up liquid food, consists of what cor- 
respond with the maxillse and the labium of insects that feed on solid 


food alone. By sejjarating the parts of the proboscis a little (fig. 
!) B) it will be seen that, while there are five terminal pieces present, 
three of them arise from one median basal sclcrite {Mt), the two 
wider lateral appendages {Mx) being carried each by a separate lat- 
eral basal piece {St). The median group constitutes the labium and 
the separate lateral parts the maxillm. 

If the reader will now turn again to figure 3 C (p. 17), which may 
rejDresent any generalized insect labium, and compare with it the 
drawing of the bee labium, forming the median series of parts in fig. 
15 D, he will at once be able to identify the parts of the latter. The 
princijDal elongate median basal plate is the mentum {Mt), the small 
triangular plate at its base is the submentum {Smt), and the two 
jointed lateral appendages of the mentum are the labial palpi 
{LbPlp) , each carried by a basal palpiger {Pig) ■ It is only the parts 
of the bee's labium that lie between the paljii which are actually 
different from those in the generalized diagiam where they consist 
of the four lobes of the liijula {Gls and Pgl). But even here it will 
be seen that the two small lobes {Pgl) in the bee's labium, partly con- 
cealed within the bases of the palpi, correspond with the paraglossa. 
Hence we have only the long median apjDendage to account for and it 
is unquestionably the representative of the gloxficv {Gls) which are 
here fused together and drawn out into this flexible tonguelike organ. 
In fact, a comparison with the mouth parts of other Hymenoptera in 
which the elements are much less modified leaves no doubt of this 
being the true interpretation of the bee's labium. It is simjDly an 
example of how nature constantly prefers to modify an already exist- 
ing part to serve some new jDurpose rather than to create a new organ. 

If, then, we bear in mind that the slender median appendage of 
the bee's labium represents the glossas of other insects, we may for 
convenience call it the " tongue," as it is popularly termed, or, since 
it is a single organ, there is probably no grammatical objection to 
calling it the glossa. The word " tongue," however, to use it prop- 
erly, should be applied to the true lingua or hypopharynx (fig. 3 C 
and D, Hpliy) which arises fi'om the upper surface of the labium. 
Many of the older entomologists, adopting the notion from Kirby 
and Spence, who defined the term in 1826, regarded the glossa of 
the bee as the homologue of the lingua in other orders. Even Pack- 
ard in his Text-book of Entomology calls the glossa the " hypo- 
pharynx." Cheshire named it the " ligula,'' and his mistake has been 
perpetuated by several other writers on bee anatomy, including Cook 
and Cowan. The term ligvla properly includes both the glossa and 
the paraglossa^ or should signify the basal piece from which these 
four lobes arise (fig. 3 C, Lg), so that it can not be applied to the 
o-lossa alone. 


The derivation of anatomical names counts for nothing in their 
application — this must be determined by scientific usage and priority. 
Thus, glossa is the Greek word for " tongue," but it was first applied 
in entomology to the median lobes of the labium ; Viiujua is its equiva- 
lent in Latin and was first given to the true tongue or hypopharynx 
in insects; Ihjula is a diminutive derivative from "lingua"' and has 
come to be applied collectively to the terminal parts of the labium 
beyond the mentum but not including the palpi. Hence, all these 
words mean the same thing by their origins, but their anatomical 
applications should be carefully distinguished. In this paper there- 
fore the slender median appendage (6-7 a) of the labium will be 
called the glossa, or, for convenience, the tongue, but with the strict 
understanding that the organ in question is not the true tongue. 
This latter should be called the '' hypopharynx," but, as will be shown 
later, it is absent in the bee. 

The glossa of the bee (figs. 9 B; 11 A and B, and 15 D, F. and G, 
Gls) is covered with long hairs which increase in length toward the 
end. The tip is formed of a small sjDoon-shaped lobe, the labellum or 
houton (LM), which is covered by short delicate processes branched 
at their ends (fig. 15 A and B, LM). The long hairs of the glossa 
are arranged in circles and the transverse rows of hair bases give 
the tongue a multiarticulate appearance. Surrounding the dorsal 
side of the base of the labella and forming two short subterminal 
rows on the ventral side of the glossa are a number of stiff, out- 
wardly curved, spinelike hairs (Hr). These hairs have been de- 
scribed as taste organs, but their appearance would suggest that they 
are simply protective spines guarding the delicate tip of the tongue. 
Between the two ventral rows of these spines is the termination of 
a groove (A, 7i,) which extends along the midline of the under sur- 
face of the glossa (D, k) to its base (fig. 9 B, /<;). The cleft of this 
groove is covered by two fringes of converging hairs whose tips are 
inclined also toward the tip of the tongue. 

Let us now return to a study of figure 15 D. The series of lateral 
pieces as already explained are the maxillae. A comparison with 
figure 3 B representing a generalized maxilla will show that these 
organs in the bee have suffered a greater modification than has the 
labium, but the parts can yet be quite easily made out. The main 
basal plate (St) is the combined stipes, suigalcci, and palpifer, the 
basal stalk is the cardo (Cd), and the little peglike process {MxPlp) 
at the outer end of the stipes is the greatly reduced maxillary palpus. 
Hence, we have left only the terminal bladelike lobe (.1/^') to account 
for, and it is evident that it must be either the galea or the lacinia 
(see fig. 3 B, Ga and Lc) or these two lobes combined. Hei'e again 
a comparative knowledge of the mouth parts of Ilymenoptera comes 


to our aid and shows clearly that the part in question is the outer 
lobe or galea, for the inner one becomes smaller and smaller in the 
higher members of the order and finally disappears. 

The base of the submentum is connected in the bee with the upper 
ends of the cardines by a -flexible, widelj' V-shaped band, the lorum, 
(Lr). The posterior angle of the submentum rests in the apex of the 
lorum, while the tips of the loral arms are movably articulated with 
the distal ends of the cardines. The name " lora " was given to this 
structure by Kirby and Spence, bvit " lorum " is more correct, since 
this is the Latin form of the word (meaning a thong or lash). Some 
recent entomologists have spoken of the structure as consisting of 
two rods, thus making the word do duty as a plural, but the thing 
itself is all one piece. Cheshire and some others have incorrectly 
applied the name to the submentum. 

The lorum is peculiar to the Hymenoptera, and the reason for it 
is clear when we examine the attachments of the parts of the proboscis 
to the head. As already stated, the maxillse and labium are sus- 
pended in a large cavity on the back of the head which may be called 
the fossa of the proboscis (fig. 9 B, PrhFs). The maxillse are articu- 
lated by their cardines {Cd) to the maxillary suspensoria (fig. 11 
A, e) at the upper edges of the side walls of the fossa. The labium, 
on the other hand, is not attached to the solid walls of the cranium 
but is suspended in the membranous floor of the fossa. This is to 
afford it freedom of movement during feeding, but, in order to 
give it more substantial support and to make the regulation of its 
motions possible, the submentum is slung to the ends of the cardines 
by the lorum. 

The terminal lobes of the labium and maxillae when not in use 
are ordinarily folded down beneath the head against the mentum 
and stipites (fig. 19). When, however, the bee wishes to imbibe a 
thick liquid such as honey or sirup in large quantity, these parts are 
straightened out and held close together so as to form a tube between 
them leading into the mouth, the terminal joints of the labial palpi 
alone diverging from the rest (fig. 11 A). 

The action of the mouth parts while feeding may be observed quite 
easily if some bees are given a small amount of honey and then 
watched through a lens while they are eating. A most convenient 
method is to put a few workers in a small screen-covered cage, such 
as are used for queen nurseries, spread a small drop of honey on the 
wire, and then place the cage under a simple microscope. It will be 
seen that the maxillse are held almost stationary but that the base 
of the labium slides back and forth between the maxillary bases 
with a very regular to-and-fro movement as if the honey were being 
either pumped or sucked up into the mouth. It is probable that there 
is a sucking force exerted by the pharynx (fig. 11 B, Phy) but not 


by the honey stomach (fig. 44, HS), which latter, as Cheshire re- 
marks, could no more suck honey through the ajsophagus than a 
balloon could suck gas from a pipe. The liquid undoubtedly runs 
up the temporary tube between the blades of the mouth parts first 
by capillary attraction, but it must be greatly assisted along its way 
to the mouth by the retraction of the labium. The load brought up 
when this is pulled back is then sucked into the mouth by the 
pharynx while the labium immediately goes out again after more. 
It acts thus as a sort of mechanical feeder and this function is prob- 
ably derived from the lapping motion of the under lip in wasps and 

The mentum (fig. 15 D and G, Jit) is hinged freely upon the 
submentum (Smt), the latter, as already described, is set into the 
socketlike angle of the lorum, while, finally, the arms of the lorum 
are articulated to the distal ends of the cardines of the maxillae. 
Now, when the labium is retracted by means of muscles attached to 
the mentum, the submentum turns in the loral socket and assumes a 
position at right angles to the mentum while the lorum itself turns 
somewhat on its articulations with the cardines. This great freedom 
of motion is permitted by the loose membrane of the fossa in which 
both the maxillaj and the labium are suspended. 

The observer, however, can not fail to note that beside this motion 
of the entire labium the tongue itself, or glossa (6-7.y), performs a 
conspicuous independent movement of its own. It is by far the most 
active member of the mouth parts during feeding, being activeh' 
thrust out and retracted while its tip is constantly moved about in 
a way suggestive of its being delicately perceptive of taste or touch 
or perhaps to both of these senses. So great is the retractile power 
of the tongue that its tip, which normally extends far beyond the end 
segments of the labial palpi, can be drawn back entirely within the 
latter. This contractile activity appears at first sight to be due to 
elasticity, but a closer examination will show that the entire ligula, 
i. e., the paragiossae (Pgl) as well as the glossa (Gls), moves back 
and forth and that the action is due to a retraction of the base of the 
ligula (fig. 15 F, Lg) into the anterior end of the mentum (Jit). 
The ligula is supported on a membranous cone at the end of the 
mentum whose walls are strengthened by three thin chitinous plates, 
two above (F, p) and one below (D, g). By the contraction of 
muscles situated within the mentum (fig. 16, IRMd) and inserted 
upon the base of the ligula the latter is pulled into the end of this 
cone whose walls, including the chitinous plates, simply turn inward. 

But the tongue does possess also a contractile power of its own by 
means of which it actually shortens its length. A flexible rod arising 
from the median ventral supporting plate (fig. 15 D, o) of the ligula 
extends throughout its length. The base of this rod is curved down- 


ward and has two muscles attached to it. This is shown by figure 
15 H, where the rod (r) is torn from the glossa {Gls) basally so as 
to show the muscles {2RMcl) inserted upon it and its connection 
with the plate (o). By the contraction of the muscles the rod bends 
at its base and is drawn back into the mentum. The glossa thus 
shortens and becomes bushy just as does a squirrel's tail when one 
attempts to pull the bonp out of its base. 

The protrusion of the parts is due to the pressure of blood driven 
into the ligula from the mentum, Avhile probably the glossa extends 
also by the straightening of its rod as the muscles relax. Wolff 
described a protractor muscle at the base of the ligula. The rod of 
the tongue is certainly not in itself contractile, as supposed by 
Cheshire, who looked for evidence of muscular striation in it. It has 
mostly a transparent and cartilaginous appearance, but is presumably 

The mouth parts, their action in feeding, and the muscular mech- 
anism by which they are moved have been elaborately described 
and illustrated by Wolff (1875) in his monograph on the organs of 
smell in bees. Most unfortunately, however, Wolff's paper was 
written to show that the seat of the sense of smell is in the mouth, 
a most erroneous notion, and the title of his paper based on this 
notion has caused little attention to be paid to this work on the mouth 
parts of the bee, which is one of the best anatomical treatises evei' 
produced on the mouth parts of any insect. 

It, still remains for us to describe the details of the glossa and its 
particular function in feeding. The tongue is not a solid appendage 
nor yet is it truly tubular. A. compromise is effected by the longi- 
tudinal groove (fig. 15 A and D, k) on its ventral surface which 
expands within the tongue into a large cavity occupying half of its 
interior (E, Lum). The glossal rod (r), which has already been 
mentioned, lies in the dorsal wall of this channel and is, hence, 
really not an internal but an external structure. The rod is itself 
grooved along its entire ventral length (E, I) and this groove agaiii ■ 
is converted into a tube by two rows of short hairs which converge 
from its margins. The lips of the ventral groove of the glossa are 
so deeply infolded that its cavity is almost divided along the midline. 
Hence, the glossa might be described as containing three channels. — 
a small median dorsal one (?) and two large latero- ventral ones 

The glossal rod (fig. 15 C, r) is very flexible but not contractile, as 
already stated, and is mostly clear and cartilaginous in appearance, 
its ventral groove {I) alone being lined by a deposit of dark chitin 
(fig. 15 C and E). Its shape in section is sufficiently shown by the 
figures. The walls of the large channels of the proboscis consist of 
a delicate membrane (C and E, q) covered with very small hairs. 


The entire ventral cavity (Lum) with the rod (r) can be evaginated 
through the ventral cleft (k) by blood pressure from within. At 
Cheshire points out, this permits of the channels being cleaned in 
case of clogging by pollen or any foreign matter. 

It is supposed that these glossal tubes are of especial service to the 
bee by enabling it to take up the smallest drops of nectar- -quantities 
that would be lost in the clumsy tube formed between the parts of 
the Jabium and the maxillae. The suction must be in large part 
capillary attraction, but here again the shortening of the glossa by 
the retraction of its rod must squeeze the contained nectar out of the 
upper ends of the channels where it is received upon the ventral flaps 
of the paraglossa3 (fig. 15 F, Pffl), from which it runs around the 
base of the tongue (Gls) within the paraglossaj to the dorsal side of 
the mentum (Mt) and so on to the mouth. 

The maxillae and labium of both the queen and the drone (fig. 11 
B) are smaller and weaker than those of the worker, and neither of 
these two forms is capable of feeding itself to any extent. If a 
hungry queen be given some honey she attempts to eat it and does 
imbibe a small quantity, but at the same time she gets it very much 
smeared over her head and thorax. 

The mouth is hard to define in insects; practically it is the space 
surrounded by the bases of the mouth parts, but strictly speaking it 
is the anterior opening of the alimentary canal situated behind the 
bases of the mouth parts (fig. 19, Mth). Yet the enlargement of the 
alimentary canal (Phy) immediately following this opening is never 
spoken of as the mouth cavity but is called the pharynx. On the 
other hand the so-called epipharynx {Ephy) and hypopharynx 
(absent in the bee) are located in front of this opening and are con- 
sequently not in the pharynx at all, the former being attached to the 
under surface of the labrum and clypeus, while the latter is situated 
on the upper surface of the base of the labium. These and numerous 
other inconsistencies in the nomenclature of insect morphology have 
to be endured because the parts were originally named for descrip- 
tive purposes by entomologists who were not familiar with scientific 
anatomy. In this paper the term mouth will be applied to the true 
oral opening (fig. 19, Mth). The space in front of it between the 
bases of the mouth parts may be called the freoral cavity. 

The duct of the salivary glands of insects in general opens upon the 
base of the labium in front of the hypopharynx. In the honey bee 
the salivary opening is on the dorsal side of the base of the ligula 
between the paraglossse (fig. 15 F, SalDO). This alone would show 
that the glossa is not the hypopharynx of the bee, as many authors 
have supposed, for otherwise the opening of the salivary duct should 
be ventrad to the base of the glossa. In fact, this makes it clear that 
22181— No. 18—10 4 


the bee does not possess a hypopharynx. There is, however, a con- 
spicuous chitinous plate located on the anterior part of the floor of the 
pharynx (fig. 19, s) having two terminal points hanging downward 
over the lower lip of the oral aperture, but, although this plate is truly 
hypopharyngeal in position, it is not the homologue of the organ 
called the hypopharynx in other insects. It is variously developed 
in all Hymenoptera, being simply a chitinization of the floor of the 
pharynx, and should be called the pharyngeal plate {Schlundbein of 
Wolff). It will be more fully described in connection with the ali- 
mentary canal. If a hypopharynx were present it should be situated 
on the upper side of the labium (see fig. 3 D, Hphy) but there is here 
present only a plain arched membranous surface in the honey bee 
and other typical Hymenoptera. 

The external location of the salivary opening enables the saliva 
to be mixed with the food before the latter enters the mouth. This 
is necessary in insects since the jaws are also on the outside of the 


\ Ls .. TMcl 



Fig. 16. — Median section tlirougli distal half of mentum (Mt) and base of ligula (Lg) 
of worlser, sliowing opening of salivary duct (SaWO), and muscles connected with 
ligula and the "salivary syringe" (t). 

mouth, and whatever chewing or crushing the food receives from 
them is consequently done in the preoral cavity. 

In some insects the saliva is used for other purposes than diges- 
tion. For example, the saliva of some predaceous insects with pierc- 
ing mouth parts belonging to the order Hemiptera is poisonous, and 
when one of these insects " bites," the saliva is injected into the 
wound by a special pump. The bite of the mosquito is made painful 
likewise by an irritant secretion from a part of the salivary glands. 
Bees appear to have the power of letting their saliva run down the 
tongue when necessary to dissolve a hard substance like sugar and 
render it capable of being taken up in solution, for they do not eat 
sugar with their mandibles. Moreover, there is even a sort of pump 
or so-called " salivary syringe " at the termination of the salivary 
duct in the ligula, by means of which the secretion can be forcibly 
ejected from the opening. 

The salivary opening on the base of the ligula (fig. 15 F, SaWO) 
leads into a deep transverse pit with collapsible cartilage-like walls 
having its deepest part turned horizontally toward the base of the 



labium (fig. 16, t). The salivary duct (SalD) bends downward in 
the anterior part of the mentum (3It) and opens into the posterior 
end of the pit (t). When the retractor muscles {IRMcl) of the 
ligula pull the latter back into the mentum the lips of the salivary 
pit must necessarily be closed. The simultaneous contraction of the 
elevator muscle (m) attached to the roof of the horizontal part of the 
pit must expand the latter and suck the saliva from the salivary duct. 
When, finally, these muscles relax and the ligula is driven out by 
blood pressure in the mentum, probably produced in part by the 
contraction of its dorsal transverse muscles {TMcl), the saliva in 
the temporarily formed bulb must be squirted out upon the base of 
the tongue. Wolff (1875) calls each dorsal longitudinal muscle of 
the mentum {IRMcl) — the two inserted upon the basal hooks (w) of 
the glossa (fig. 15 H and fig. 16) — the retractor lingum longus. The 
large ventral retractor muscle of each side {'BR Mel) he calls the 
retractor linguw biceps since its anterior end divides into two parts, 
one of which is inserted by a tendonous prolongation upon the base 
of the glossal rod (fig. 15 H and fig. 16, r) and the other upon the 
base of the ligula. The use of the word " lingua " in these names is 
objectionable because, as already explained (page 45), the lingua is 
properly the true tongue or hypopharynx. " Ligulse " should be sub- 
stituted for " linguae." The dilator muscle (fig. 16, u) of the salivary 
pit (t) is termed the protractor lingual by Wolff because, as he sup- 
poses, when the ligula is pulled back into the mentum the position 
of this muscle is reversed, so that a contraction of its fibers would 
help to evert the ligula. 

The glands that furnish the saliva lie within the head and the 
thorax and will be described later in connection with the alimentary 
canal and the process of digestion. 

Fig. 1Y. — Epipharynx (Ephij) and labrum (Lm) of worker: A, ventral Tiew ; B, 

anterior view. 


The epipharynx of insects in general may be described as a dorsal 
tongue, it being a median lobe developed on the roof of the preoral 
cavity from the under surface of the clypeus or labrum and situated 
opposite the hypopharynx. 



The epipharynx of the bee is a large three-lobed appendage de- 
pending from the roof of the preoral cavity just in front of the mouth 
(fig. 19, Ephy). Seen from below it is triangular (fig. 17 A) with 
the apex forward. Its median lobe has the form of a 
fcJl fC^ I'^igh) vertical, keel-like plate, while the lateral lobes 
are rounded but have prominent elevated edges con- 
^'e^ging toward the front of the keel. The appearance 
in anterior view is shown by figure 17 B. Situated 
on the posterior parts of the lateral lobes are a num- 
ber of sense organs, each consisting of a small cone 
with a pit in the summit bearing a small hair (fig. 18). 
These are regarded as organs of taste. 
Wolflf (1875) made a most thorough study of the epipharynx, 
which he called the "palate sail" {Gaumensegel) on account of the 
high median crest. His drawing is the standard illustration of the 
organ found in nearly all books on the anatomy of the honey bee 


V '-5., 

Fig. 18. — Sense 
organs, prob- 
ably of taste, 
from e p i - 

Fig. 19.- 

-Median longitudinal section of head of worker, but with entire labium attached, 
showing internal organs except muscles and brain. 

and in most works on general insect anatomy and the sense organs. 
Wolff, however, regarded the sensory cones as having an olfactory 
function, and this led him to erroneous conclusions regarding the 
functions of several other organs. For example, he thought that 
the mandibular glands poured a liquid upon the surface of the 





epipharynx which kept it moist and capable of absorbing odor 
particles, while he ex- 
plained the inhalation 
of the latter into the 
preoral cavity as 
brought about through 
the contraction of the 
air sacs situated about 
the mouth. Wolff's 
anatomical researches 
are without doubt 
some of the best ever 
made on the bee, and 
it is due to his mis- 
taken idea of the loca- 
tion of the sense of 
smell, which, as al- 
ready explained, is on 
the antenna', that we 
have received from 
him a most excellent 
account and detailed 
drawings not only of 
the epipharynx but of 
the mandibular glands, 
the mouth parts, the 
salivary " pump," and 
the respirator}'' organs. 



The apparent thorax 
of the bee (fig. 20, 
T^-IT, and fig. 21) 
and of most other 
Hymenoptera is not 
exactly the -equivalent 
of the thorax in other 
insects. The middle 
division of the body, 
so conspicuous in this 
order, consists not only of the three leg-bearing segments, which alone 

Fig. 20.- 

-Dorsal view of ventral walls and internal skele- 
ton of body ot worker. 



constitute the thorax of all other insects, but also of the first ab- 
dominal segment. The conspicuous necklike constriction posterior to 
the base of the hind legs (fig. 21, Pd) is, therefore, between the first 
and the second abdominal segments (fig. 1, IT and I IT). 

The thorax of the honey bee at first sight looks entirely different 
in structure from that of all other insects except related Hymenoptera, 
in the higher families of which group it is more highly modified than 
in any other order of the whole series of insects. When, however, we 
examine the thorax of one of the lowest members of the Hymenop- 
tera, such as a sawfly, we are surprised to find that, in each segment, 
the structure agrees very closely with our ideal diagram of a general- 

ized thoracic 
la segment (fig. 4). 

I :^ 1 ^Jg The three seg- 

■2 ,, — ,c„r -' ments are per- 

fectly distinct, 
and the first 
abdominal seg- 
ment, while it 
may be clearly 
separated from 
the rest of the 
abdomen, is not 
fused into the 
thorax so as to 
appear to be a 
part of it. If, 
now, we exam- 
ine representa- 
tives of several 
families inter- 
mediate between 
the sawflies and 

the bees, the line of specialization that has produced the bee thorax 
becomes perfectly evident. The principal features in these modifi- 
cations are the following: 

(1) The lateral and ventral parts of the prothorax (figs. 20 and 21, 
EpSi and S-^) are suspended loosely in a large membranous area 
which is continuous anteriorly as the neck. They thus form a sort 
of suspensorium for the front legs, which appears detached from the 
rest of the thorax. (2) The protergum {T^) is solidly attached to 
the anterior edge of the mesothorax and its lateral parts extend 
downward till they meet on the venter behind the prosternum (figs. 
20 and 21). (3) The postnotum (postscutellum) of the mesothorax 
(figs. 22, PN ; 23 A, PN^) is entirely invaginated into the cavity of 
the thorax and is reduced to the form of two lateral arms of the large 

Fig. 21. — Thorax of worker, left side, with intersegmental lines 
somewhat exaggerated, showing prothorax (Ti, EpSi, Cxi), 
mesothorax (T2, Eps^, Epnis, S^, Cx-z) , metathorax (T^, Ph, 
ph, Oxs) and propodeum or first abdominal segment (IT). 


internal postphragma {Pph) which has no median tergal connection 
at all. (4) The metatergum (figs. 21 and 23 A, T^) consists of a 
single narrow plate. (5) The metapleurum (fig. 21, Pl^ and fl^) 
shows no trace of a division into episternum and epimerum, but is 
divided into an upper {Pl^) and a lower {pl^) pleural plate. (6) 
The first abdominal tergum (fig. 21, IT) is solidly attached to the 
metathorax and forms an intimate part of the thoracic mass. 

We shall now proceed with a more detailed account of the thorax, 
and the reader should occasionally turn back to figure 4 (p. 19) in 
order to keep clearly in mind the parts that make up a generalized 
thoracic segment. 

The parts of the prothorax are so separated from each other that 
they appear to belong to different segments. The protergum (fig. 21, 
y'l) forms a collar completely encircling the front of the mesothorax. 
On each side a large lobe (w) projects posteriorly as far as the base 
of the front wing and constitutes a protective shield over the first 
thoracic spiracle. The tergum presents a median transverse groove, 
dividing it into an anterior and a posterior part, which parts may 
be called the scutum (fig. 23 A, T^, Set) and scutellum {Scl). The 
propleurum (figs. 20, 21, Eps^) consists of a large plate presenting 
both a lateral surface (fig. 21) and a ventral surface (fig. 20). On 
account of the position of the coxal articulation (fig. 21) this plate 
would seem to be the anterior pleural plate alone (see fig. 4), which 
is the episternum. In some Hymenoptera the epimerum is repre- 
sented by a very small plate on the rear edge of the episternum. 
The anterior ends of the two episterna form knobs which loosely 
articulate with the occipital region of the head (figs. 11 B, 20, and 
21). Lying just ventrad of each is a slender cervical sclerite (fig. 21, 
mi). The prosternum (S^) is shown by figure 20. It carries a large 
entosternura {Fu^), forming a bridge over the nervous system behind 
the prothoracic ganglion (fig. 52). 

The mesotergum, as seen in its natural position (fig. 21, T^), consists 
of a large anterior scutum (Sot,) and of a smaller but very prominent 
posterior scutellum (Scl^), separated by a very distinct suture (c). 
The scutellum has two latero-anterior areas partially separated from 
the median area by sutures. When the mesotergum is detached from 
the rest of the thorax (fig. 22) it is discovered that there is attached 
laterally to the scutellum a large posterior internal part, which does 
not show on the sutface at all. This is the representative of the 
postscutellum (Pscl) and its phragma (Pph) constituting the post- 
notum (PJV) of our diagrammatic segment (fig. 4). The proof of 
this, again, is to be derived from a study of the lower Hymenopteran 
families. In some of the horntails (Siricidee) the postnotum or 
postscutellum is a prominent plate on the surface of the dorsum be- 
hind the scutellum. In /S/rex (Siricidae) this plate is sunken below 



the general surface and mostly concealed between the naesothorax 
and the metathorax. In higher families such as the Pompilidse the 
postnotum of the mesotergum is entirely concealed by invagination, 
but it still carries a very large, phragma. When, now, we come to 
the highest members of the order we find that the median part of the 
postnotum in the mesothorax is gone entirely and that it is repre- 
sented only by the lateral arms (figs. 22, PN ; 23 A, PNo) carrying 
the large, purely internal postphragma (Pph). 

The mesopleurum is large and consists principally of the episternum 
(fig. 21, EpSo), which, however, is continuously fused with the meso- 
sternum (figs. 20 and 21, 82)- The pleural suture (fig. 21, PS^) is 
short and sinuous and does not reach more than half way from the 
wing process to the base of the middle leg. The epimerum is reduced 
to a small double plate lying above the episternum and posterior to 
the wing process (figs. 21, Epm., and 24 A, Epm and Epm). The 

pleural ridge (fig. 24 B, PR) 
is weak, but the wing process 
(WP) is well braced by a num- 
ber of accessory internal ridges. 
One preparapterum {2P) and 
one postparapterum {3P) are 
present. Lying behind the 
postparapterum is another 
larger sclerite (fig. 24 A and 
B, pn), whose anterior end is 
articulated to the edge of the 
epimerum and whose posterior 
tapering end is loosely asso- 
ciated with the terminal arms 
of the postnotum (fig. 22, PN and pn). This sclerite might be 
regarded as the fourth parapterum, but it is much more probably 
the representative of a small terminal bar of the postnotum present in 
other Hymenoptera, such as Pepsis, which connects this tergal plate 
with the epimerum, though in this genus it is not detached from the 
main postnotal sclerite. 

Both the mesostermim (fig. 20, S.^) and the metasternum (S^) con- 
tribute to the formation of a large entosternum (T^Wj+g), which forms 
a protecting bridge over the combined mesothoracic and metathoracic 
ganglia (fig. 52) and affords attachment for the»ventral longitudinal 
muscles of the thorax (fig. 27, Imcl). 

The metathorax consists of a very narrow series of plates (fig. 21, 
7*3, PZg, and pl^) compressed between the mesothorax and the first 
abdominal tergum {IT) . Its back plate is a single, narrow, transverse 
sclerite (figs. 21 and 23 A, T^) widening on the sides, where it carries 
the wings by the two wing processes' (fig. 23 A, ANP and PNP) . The 

Fig. 22. — Lateral view of mesotergum of 
worker, removed from the rest of thorax to 
show large internal postscutellum (post- 
notum, PX) and its phragma {Pph) not 
visible normally in the bee from exterior. 



ordinary tergal divisions seem to be entirely obliterated. The meta- 
pleurum consists of a dorsal plate (fig. 21, Pl^) supporting the hind 
wing and of a ventral plate {pl^) carrying the hind leg. These two 
functions certainly identify these two plates as constituting together 
the metapleurum, but there is absolutely no trace of a division into an 
episternum and an epimerum. Once more, therefore, Ave have to go 
back to the generalized Hymenoptera to find out what has happened. 


Fig. 23. — ^A, thoracic terga of worker separated from one another, showing protergum 
[T-l), mesotergum (Ta) and its internal postscutellum (postnotum PN2) and phragma 
(Pp7»2), metatergum (Ts) and propodeum or first abdominal tergum (IT) ; B, ventral 
¥iew of principal or notal plate of mesotergum. 

The answer is simple. Sirex has a typical metapleurum consisting of 
an episternum and epimerum separated by a complete pleural suture. 
In the higher forms this suture simply disappears, and consequently 
the pleurum shows no traces of its original component plates. The 
division into a wing-bearing and a leg-bearing plate is, therefore, a 
purely secondary one. 

None of the Hymenoptera has separate trochantinal sclerites (see 
fig. 4, Tn), but, since the coxae are articulated ventrally to knobs 



(figs. 20 and 21, z) apparently belonging to the sterna, it might be 
supposed that the trochantins have fused with the latter plates. 

The posterior part of the thoracic mass (fig. 21) consists of the 
first abdominal tergum {!T), which fits into the deeply concave pos- 
terior edges of the metathorax and forms the j^eduncle {Pd) that 
carries the rast of the abdomen (fig. 32). It consists of a single large, 
strongly convex sclerite (figs. 21 and 23 A, IT) bearing the first 
abdominal spiracles laterally (ISp) and having its surface divided 
into several areas by incomplete sutures. 

Many entomologists find it difficult to believe that this plate, which 
so apparently belongs to the thorax, is really derived from the abdo- 
men. But the proof is forthcoming from a number of sources. In 
the first place, the thorax is complete without it and the abdomen is 
incomplete without it, the latter having otherwise only nine seg- 
ments. Again, if the plate is reckoned aH a part of the thorax we 




Fig. 24. — A, upper part of left mesopleurum of worker, external ; B, inner view of same. 

should have the anomalj^ of a thorax with three pairs of spiracles — 
there being the normal two on each side situated, as they always are, 
between the true thoracic segments. Furthermore, comparative anat- 
omy shows us that in some of the sawflies (Tenthredinidae) the first 
abdominal tergum, while separated by a wide membranous space 
from the second, is not at all incorporated into the thorax. In a horn- 
tail such as Sirex (Siricidse) the entire first abdominal segment is 
fused to the posterior edge of the metathorax and is only loosely 
joined to the next abdominal segment by membrane. This insect 
affords, therefore, a most complete demonstration of the transference 
of this segment from the rest of the abdomen to the thorax. Finally, 
we have absolute jDroof of its abdominal origin based on a knowledge 
of development, for it has been shown by Packard from a study of the 
bumblebee that the first abdominal segment of the larva • is trans- 
ferred during the pupal metamorphosis to the thorax and forms the 


part under discussion. We hence see that not only the first abdomi- 
nal tergum but the entire segment has undergone transposition, 
though the ventral part has disappeared in all the higher families. 
This transferred part has been named both the median segment and 
the propodeum, by writers who recognize it as belonging to the abdo- 
men and not to the thorax. 

The names current among systematists for the back plates of 
Hymenoptera afford an excellent example of the errors that ento- 
mologists may be led into through an ignorance of the comparative 
anatomy of insects. They recognize the protergum as such and then, 
loiowing that there are yet two segments to be accounted for, they 
call the mesoscutum the " mesonotum," the mesoscutellum the 
" scutellum," the metatergum the " postscutellum " (being unaware 
that the true postscutellum is deeply concealed within the thorax), 
while the first abdominal tergum is called the metathorax. Such 
a nomenclature assigns both pairs of wings to the mesothorax. Too 
many systematists working in only one order of insects do not care 
whether their names are applied with anatomical consistency or not. 


In the study of insects the wings always form a most interesting 
subject because by them insects are endowed with that most coveted 
function — the power of flight. It has already been stated that the 
wings are not primary embryonic appendages, but are secondary out- 
growths of the body wall from the second and third thoracic seg- 
ments. Therefore it is most probable that the early progenitors of 
insects were wingless, yet for millions of years back in geological time 
they have possessed these organs in a pretty well developed condition. 
. Nearly all of the insect orders have some characteristic modifica- 
tion of the wingf-veins and their branches. None of them, however, 
departs nearly so far from the normal type as do the Hymenoptera, 
even the lowest members of this group possessing a highly specialized 
venation. Before beginning a study of the Hymenopteran series 
which leads up to the bee the student should first turn back to figure 
C (p. 22) and again familiarize himself with the generalized condi- 
tion of the veins and the articular elements of the wing. By com- 
paring, now, with this diagram the basal parts of the wing of a 
sawfly {Itycorsia discolor, fig. 26 A) it will be easy to identify the 
parts of the latter. Vein C has two little nodules {C, C) cut off from 
its basal end which lie free in the axillary membrane. Yein Sc articu- 
lates by an enlarged and contorted base {8c) with the first axillary 
{lAx), while vein R is continuous with the second [2 Ax). The next 
two veins that come to the base and unite with each other are appar- 
ently not the media and cubitus but the first and third anals {lA and 



SA), since they are associated with the third axillary {J Ax). In this 
species the subcosta (Sc) is entirely normal, but in the related horntail 
{Sirex -flavicornis, fig. 26 B) the enlarged basal part of the subcosta is 
almost separated from the shaft of the vein, while the latter (fig. 25 A, 
Sc) is short and weak. A study of the venation of this wing leads 
us to believe that the vein which arises from the radius a short dis- 
tance from its base is the cubitus {Cu). Therefore the basal part 

Pig. 25. — Wings of Hymenoptera and their basal articular sclerites (lAx—iAx) : A, Sirex 
flavicomis, front wing ; B, Pepsis sp., front wing ; C, honey bee, front wing ; D, honey 
bee, hind wing. 

of the media is either gone or is fused with the radius. Since we dis- 
cover its branches in the distal field of the wing, arising from the 
trunk of the radius, we conclude that the latter is the case. By this 
sort of reasoning we may arrive at the Comstock and Needham inter- 
pretation of the wing illustrated at A, fig. 25. From this it is evident 
that the branches of both the radius and the media have been bent 
back toward the posterior margin of the wing. 



Fig 26. — Basal elements ot wings of Hymenoptera : A, base of front wing of a sawfly 
(itycorsia discalor) showin!;' comparatiyely generalized arrangement of veins and 
axillaries ; B, bases of anterior veins of front wing of a liorntail (Sirex flavicornis) , 
showing detachment of base of subcostal vein (Sc) from its shaft ; C, corresponding 
view of anterior veins in front wing of a tarantula-killer (Pepsia sp.), showing com- 
plete absence of shaft of subcosta, but presence of basal part ISc) fused with base of 
radius (R) ; D, axillaries of anterior wing of honey bee worker; B, tegula of worker; 
F base of anterior wing of worker showing absence of shaft of subcosta buf presence 
of scale (Sc) derived from its base ; G, axillaries of hind wing of worker, the fourth ab- 
sent in bee ; H, base of hind wing of worker, showing absence of costal and subcostal veins 
and fusion of bases of subcosta (Sc) and radius (R) into large humeral mass; I, attach- 
ment of front wing to scutum (Sctz) and scutellum (Sch) of mesotergum ; .T, under view 
of end of mesoscutellum (8cU) showing attachment of both first (lAx) and fourth 
axillaries (iiAx) to posterior wing process (PUP), an unusual connection for first axillary. 


Taking this wing of Sirex as a foundation let us proceed a little 
higher and examine the wing of a Pompilid, such as Pepsis (figs. 
26 C and 25 B). We observed that in Sirex (fig. 26 B) the basal 
part of vein Sc is almost separated from the distal shaft. In Pepsis 
(fig. 26 C) it is entirely a separate piece, to which is fused also the 
base of vein E. Moreover, the shaft of Sc has disappeared entirely 
(fig. 25, B). Thus there is at the humeral angle of the wing a large 
chitinous mass (fig. 26 C, Sc and R) representing the fused bases 
of both the subcosta and the radius, which is associated with 
both the first axillary (lAx) and the second axillary {2 Ax). 

If now we proceed to a study of the front wing of the bee we 
find that its basal characters (fig. 26 F) are more similar to those of 
Sirex (B), while its venation (fig. 25 C) resembles more closely that 
of Pepsis (B). The subcostal scale at its base (fig. 26 F, Sc) is 
not fused with the base of the radius, but the distal part of the 
subcosta is gone (fig. 25 C), as in Pepsis. In the hind wing of the 
bee (fig. 26 H) the bases of the subcosta and radius are fused into 
one large humeral mass articulating with the first two axillaries 
{lAx and £Ax). The third axillary {3 Ax) is well developed but 
the fourth is absent. The venation (fig. 25 D) is reduced to a very 
simple condition, but to one just the opposite from primitive. 

The details of the axillaries in the two wings are shown by figure 
26 D and G. The fourth {4^ix) is well developed in the front wing 
(D) and has a large accessory sclerite (y) connected with it, upon 
which is inserted a long slender muscle (fig. 28, cc). A very small 
accessory sclerite (ax) occurs close to the muscle plate of the third 
axillary {3Ax). These are called "accessory" sclerites because 
they are of irregular occurrence in the wing bases of insects generally 
and are developed in connection with the muscle attachments. Simi- 
lar ones occur in the hind wing (G, ax) in connection with the 
second {2 Ax) and third axillaries {3 Ax). 

The front wing is attached to the posterior half of the side of 
the mesonotum. The anterior notal wing process is bilobed (figs. 
22, 23 A, ^2, ANP) and is carried by the scutum, while the pos- 
terior process {PNP) is carried by the scutellum and is mostly 
hidden beneath the anterior wing process. The two wing processes, 
in fact, are so close together that the first axillary articulates not 
only with the first but also with the second (fig. 26 J). The axillary 
cord (fig. 26 F, AxG) arises from a lobe of the scutellum overlapped 
by the lateral margin (I and J, AxC). In the hind wing, where the 
fourth axillary is absent, the third articulates directly with the 
posterior notal wing process of the metatergum (fig. 23 A, T'j, PNP). 

The base of the front wing is overlapped by a large scale (fig. 26, 
E and I, Tg) called the tegula. It is carried by the axillary mem- 


brane, to which it is attached between the humeral angle of the wing 
base and the edge of the notum. The tegulaj are present in most in- 
sects, generally on the base of each wing, but they usually have the 
form of small inconspicuous hairy pads, as hhown in the diagram 
(fig. 6, Tg). In the flies, moths, butterflies, and Hymenoptera, 
however, the tegulae of the front wings develop into large conspicu- 
ous scales overlapping the humeral angles of the bases of these 

The motion of the wing in flight consists of both an up-and-down 
movement and a forward-and-backward movement, which two com- 
bined cause the tip of the wing to describe a figure-eight course if 
the insect is held stationary. Corresponding with these four move- 
ments are four sets of muscles. In the dragonflies nearly all of the 
wing muscles are inserted directly upon the base of the wing itself, 
but in other insects, excepting possibly the mayflies, the principal 
muscles are inserted upon the thoracic walls and move the wing 
secondarily. In the lower insects, such as the grasshoppers, crickets, 
stoneflies, net-winged flies, etc., the two wing-bearing segments are 
about equal in their development and each is provided with a full 
equipment of muscles. In these insects the wings work together by 
coordination of their muscles, although each pair constitutes a sepa- 
rate mechanism. In such insects, however, as the true flies and the 
wasps and bees the metathorax, as we have seen in the case of the 
bee, is greatly reduced, and what is left of it is solidly attached to 
the mesothorax. In the flies the hind wings are reduced to a ])air 
of knobbed stalks having no function as organs of flight, while in 
the bees the hind wings, which are very small, are attached to the 
front wings by a series of booklets on their anterior margins (fig. 
25 D, mc) which grasp a posterior marginal thickening of the 
front wings. Moreover, when we examine the interior of the bee's 
thorax we find that the muscles of the metathorax are greatly 
reduced or partly obliterated and that the great mesothoracic mus- 
cles serve for the movement of both wings, thus assuring a perfect 
synchrony in their action. Hence, it is clear that the union and 
consolidation of the thoracic segments in the higher insects is for 
the purpose of unifying the action of the wings. 

The muscles of flight in the bee may be very easily studied by cutting 
the thorax of a drone into lateral halves. The cavity of the thorax 
is occupied almost entirely by three great masses of muscles. One 
of these is longitudinal, median, and dorsal (fig. 27, LMcl^), extend- 
ing from the mesoscutum {Sctr.) and the small prephragma {Aph) 
to the large mesothoracic postphragma (Pph^). A small set of 
muscles (LMcl^) then connects the posterior surface of this phragma 
with the lower edge of the propodeum (IT). On each side of the 



anterior end of this great longitudinal muscle is a thick mass of 
dorso- ventral fibers (VAIcl) extending from the lateral areas of the 
mesoscutum (Set.,) to the lateral parts of the mesosternum {S.^). A 
contraction of the vertical muscles must depress the tergal parts, 
at the same time expanding the entire thorax in a longitudinal direc- 
tion and stretching the longitudinal muscles. A contraction, then, 
of the latter muscles (LMcl) restores the shape of the thorax and 
elevates the tergal parts. Remembering, now, that the wings are 

supported from be- 
low upon the 
pleural wing proc- 
esses and that each 
is hinged to the 
back by the notal 
wing processes, it 
is clear that a de- 
pression of the 


dorsum of 
thorax must 


vate the wings and 
that an elevation 
of the dorsum de- 
presses them — the 
pleural wing proc- 
esses acting as the 
fulcra. Hence, the 
chief up-and-down 
movements of the 
wings are pro- 
duced by these 
great thoracic mus- 
cles acting upon 
the shape of the 
thorax as a whole 
and not directly 
upon the wings 
themselves. The vertical muscles are the elevators and the longi- 
tudinal the depressors. 

But besides being moved up and down the wings can also, as before 
stated, be extended and flexed, i. e., turned forward and backward in 
a horizontal plane upon the pleural wing process. The muscles 
which accomplish these movements lie against the inner face of the 
pleurum (fig. 28), and each wing is provided with a separate set. 
The extensor muscle {PMcl) is the most anterior and is inserted by 
a long neck upon the preparapterum {^P). The latter is closely 

Fig. 27. — Median section through thorax of drone, showing 
longitudinal muscles (LMch) of mesothorax going from 
mesotergal scutum (Scfe) and small anterior phragma 
(Aph) to posterior phragma (Pph^) of Internal postscutel- 
lum (postnotum) of same segment, also showing vertical 
mesothoracic muscles (FJfcO, and ventral longitudinal mus- 
cles (Imcl), and longitudinal muscles of metathorax 
(LMch) going from postphragma of mesothorax (Ppha) to 
posterior edge of propodeum or first abdominal tergum (IT). 
By alternate contraction of dorsal longitudinal muscles and 
vertical muscles, roof of thorax is elevated and depressed, 
causing wings to heat downward and upward respectively, 
being supported on fulcra formed by pleural wing processes 
(fig. 28, WPi) of side walls of thorax. 



WP, 2Ax sAx 

connected with the anterior part of the base of the wing so that a 
contraction of the muscle turns the wing forward and at the same 
time depresses its anterior margin. For this reason the parapterum 
and the extensor muscle have been called the pronator apparatus, and 
the muscle is known also as the pronator muscle. In some insects 
which fold the wings back against the body this muscle is a great 
deal larger than in the bee. The -flexor muscle {RMcl) consists of 
three parts situated upon the anterior half of the pleurum and in- 
serted upon the third axillary {3Ax) by long tendonlike necks. 
These muscles are antagonistic 
to the extensor and by their 
contraction pull the wing 
back toward the body. 

The mechanism which pro- 
duces the wing motion thus 
seems to be a very simple one 
and may be summarized as 
follows: Each wing rests and 
turns upon the wing process 
of the pleurum (figs. 24 arid 
28, WP) by means of the 
pivotal sclerite or second axil- 
lary in its base (figs. 26 F and 
'2,^, 2 Ax). It is hinged to the 
back by the first and fourth 
axillaries (fig. 26 F, lAx and 
Ji-Ax) which articulate with 
the anterior and posterior 
notal wing processes (fig. 23 
A, r^, ANP and PNP), re- 
spectively. The large vertical 
muscles (fig. 27, VMcl) of 
the thorax depress the ter- 
gum, which pulls down with 
it the base of the wing and 
hence elevates the distal part — 
the fulcrum being the pleural wing process. The dorsal longitudinal 
muscle {LMcl) restores the shape of the thorax, elevates the tergum, 
and consequently depresses the wing. Extension and flexion of the 
wing are produced by special muscles (fig. 28, PMcl and RMcl) acting 
upon its base before and behind the pleural wing process, respectively. 
Besides these muscles there are several others (fig. 28) associated 
with the wing whose functions are less evident. Most conspicuous 
of these is a muscle occupying the posterior half of the mesopleurum 
{aa) and inserted upon the outer end of the scutellum. This may 
22181— No. 18—10 5 


Fig. 28. — luternal view ot right yleurum of 
mesothorax of drone, showing muscles in- 
serted upon parapteral plates (2P and 3/') 
and upon third axillary (SAx). The wing 
rests upon wing process of pleurum {'WP2) 
by second axillary (gAi') ; it is turned for- 
ward and downward by the pronator muscle 
{PMcl), inserted upon anterior parapterum 
(iP) which is attached to costal head of 
wing, and is turned back toward body by 
flexor muscle (RUcl) inserted upon third 
axillary (SAx). 


be simply accessory to the large vertical sterno-scutal muscle (fig. 27, 
VMcl). Another is a long slender muscle (bh) attached to the upper 
end of the mesocoxa and inserted upon the postparapterum {3P). 
This is sometimes termed the coxo-axillary muscle. A third (cc) is 
inserted upon the tip of the accessory sclerite (y) of the fourth 
axillary and is attached -to the lateral arm of the large entosternum 
of the mesothorax and metathorax. 


The legs of the honey bee are highly modified for several special 
purposes besides that of walking, but they are so well known and 
have been so often jiescribed that it will not be necessary to devote 
much space to them here. 

The front legs (fig. 29 A) have a structure formed by the adjoining 
ends of the tibia and the first tars'al joint, which is called, on account . 
of its use, the antenna cleaner. It consists (fig. 29 C) of a semi- 
circular notch {dd) in the base of the first tarsal joint provided 
with a comblike row of bristles. A specially modified, flat, movable 
spur (ee), shown in ventral view at B, is so situated on the end of 
the tibia {Th) that it closes over the notch when the tarsus is bent 
toward the tibia. By grasping an antenna between the notch and 
the spur and drawing it through the inclosure the bee is able to re- 
move from this sensitive appendage any pollen or particles of dirt 
that may be adhering to it. 

The middle legs (fig. 29 D) present no special modifications of any 
importance. It will be observed, however, that they, as well as the 
other legs (A and F), have the first joint of the tarsus (ITar) very 
greatly enlarged. 

The hind legs of all three forms, the worker (F) , the queen (E) , and 
the drone (H), have both the tibia and the large basal segment of 
the tarsus very much flattened. In the queen and drone there seems 
to be no special use made of these parts, but in the worker each of 
these two segments is modified into a very important organ. The 
outer surface of the tibia iF,Th) is fringed on each edge by a row of 
long curved hairs. These constitute a sort of basket (CI) in which 
the pollen collected from flowers is carried to the hive. The struc- 
tures are known as the pollen hasJcets, or corhicula. The inner sur- 
face of the large, flat, basal segment of the tarsus (ITar) is pro- 
vided with severalrows of short stiff spines (G) forming a brush by 
means of which the bee gathers the pollen from its body, since it 
often becomes covered with this dust from the flowers it visits for 
the purpose of getting nectar. When a sufficient amount is accumu- 
lated on the brushes it is scraped off from each over the edge of the 
tibia of the opposite hind leg and is thus stored in the pollen baskets. 
Hence the worker often flies back to the hive with a great mass of 



Fig. 29. — A, left front leg of worker, anterior view, showing position of noteli (dd) of 
antenna cleaner on base of first tarsal joint (ITar) and of closing spine (ee) on end 
of tibia (T&} ; B, spine of antenna cleaner (ee) in flat view; C, details of antenna 
cleaner ; D, left middle leg of worlcer, anterior view ; E, left bind leg of queen, anterior 
or outer view ; F, left bind leg of worker, anterior or outer view, showing the pollen 
basket (Cb) on outer surface of tibia (T&) ; G, inner view of first tarsal joint of hind 
leg of worker showing rows of pollen-gathering hairs and the so-called " wax shears " 
{ff) ; H, left hind leg of drone, anterior or outer view. 



pollen adhering to each of its hind legs. The pollen baskets are 
also made use of for carrying propolis. 

Between the ends of the hind tibia {Th) and the first tarsal joint 
{ITar) is a sort of pincerlike cleft (F and G, ff) guarded by a row 
of short spines on the tibial edge. This is popularly known as the 
" wax shears " and it is supposed to be used for picking the plates 
of wax out of the wax pockets of the abdominal segments. The 
writer, however, has watched bees take the wax from their abdomen 
and in these observations they always poked the wax plates loose 

Fig. 30. — A, dorsal view of end of last tarsal joint of first foot (Tar), the claws (Cla), 
and empodium {Emp) of worker; B, ventral view of same; C, lateral i view of same, 
showing empodium in ordinary position when not in use. 

with the ordinary hairs or spines of the tibise or tarsi and then by 
means of the feet passed them forward beneath the body to the 

The last tarsal joint of each leg bears a pair of claws (E, Cla^ and 
a single median empodium {Em,p). Each one of the claws is bi- 
lobed, consisting of a long tapering outer point and a smaller inner 
one (figs. 30 and 31). The claws of the worker (fig. 31 A) and the 
queen (B) are only slightly different in details of outline, although 
the claws of the queen are much greater in size than those of the 



worker, but the drone's claws (C) are large and very strikingly 
different in shape from those of either the worker or the queen. 

The empodium (fig. 30 A, B, and C,_£m£) consists of a terminal 
lobe bent upward between the^cla^S- (C) and deeply cleft on its 
dorsal surface (A) , and of a thick basal stalk 
whose walls contain a number of chitinous 
plates. One pi these plates is dorsal (A and 
C, hh) and bears five very long, thick, curved 
hairs projecting posteriorly over the terminal 
lobe, while a ventral plate (B and C, ii) is 
provided with numerous short thick spines. 
A third plate (A, B, and C, ffg) almost 
encircles the front of the terminal lobe, its 
upper ends reaching to the lips of the cleft. 

When the bee walks on any ordinary sur- 
face it uses only its claws for maintaining a 
foothold, but when it finds itself on a smooth, 
slippery surface like glass the claws are of no 
avail and the empodia are provided for such 
emergencies as this. The terminal lobe is 
pressed down against the smooth surface and 
its lateral halves are flattened out and adhere 
by a sticky liquid excreted upon them by 
glands said to be situated in front of them, 
the muscle that flattens the empodial lobes the latter spring back 
into their original position by the elasticity :of the chitinous band 
(gg) in their walls. 

C ' 

Pig. 31. — A, outer view of 
hind claw of worker ; B. 
same of queen ; C, same 
of drone. 

On the relaxation of 


The abdomen of the worker and queen appears to consist of six seg- 
ments (figs. 1, 32, 33, II-VII), but it must be remembered that, as 
has already been explained, the thoracic division of the body in the 
Hymenoptera includes one segment, the propodeum or median seg- 
ment, which really belongs to the abdomen and is its true first seg- 
ment according to the arrangement in all other insects. Hence, 
counting the propodeum (figs. 21 and 32, IT) as the first, we find 
seven exposed abdominal segments in the worker and queen and 
nine in the drone. Each one except the first consists of a tergum 
(T) and a sternum (S), the former reaching far down on the side 
of the segment, where it carries the spiracle (Sp) and overlaps the 
edge of the sternum. The two plates of the last or seventh segment 
in the worker and queen are separated by a cleft on each side, and 
if they are spread . apart it is seen that the tip of the abdomen 



incloses a cavity which lodges the sting and its accessory parts. The 
end of the abdomen of the male (fig. 56 D) is quite different from 
that of the female, while in it parts at least of nine segments are 


Sp Sm 

Pig. 32. — Lateral view o£ abdomen of workei', showing tlie propodeum (IT) as a part 
of the ahdomen, of which it is the true first segment. 

visible, the last is very much modified and is exposed only on the 
sides and below. 

An internal view of the ventral plates and the lateral parts of the- 

j v' ^ 


- :.. Mb 



PiG. 33. — Ventral view of abdomen of 
worker, showing tip of sting (Stn) and 
palpuslike appendages (StnPlp) pro- 
jecting from sting chamber within 
seventh segment (Til). 


'^sj ^ 

\A ^ 


J Vffl 

lClsp'''^0\ '-Jl'.y^ 



Pig. 34. — Dorsal view of ab- 

dominal sterna 

of drone, 

showing clasping 


UClsp and SClsp) of ninth 


terga in the worker is shown by figure 20, while a corresponding 
view of the male sterna is shown by figure 34. It will be seen that 
each sternum is very widely underlapped (viewed from above) by the 


one next in front of it and that the intersegmental membrane (Mb) 
is reflected from the middle of the dorsal surface of each to the 
anterior edge of the following sternum. By removing an individual 
plate (fig. 35 A) this is more easily shown. It is also clearly seen 
that the transverse line of attachment of the membrane (Mb) divides 
the sternum into a posterior part (Rd), which is merely a prolonged 
reduplication underlapping the following sternum, and into an an- 
terior part underlapped by the preceding sternum. The posterior 
half is, hence, purely external while the anterior half forms the true 
ventral wall of the segment, its dorsal face being internal and its 
ventral face external. The anterior part is also very smooth and 
shiny and somewhat bilobed and for this reason it is sometimes called 
the " mirrors." Its edge is bounded by a thickened ridge giving off a 
short apodeme (Ap) on each side. The mirrors of the last four 
sterna are also, and more appropriately, called the wax plates because 
the wax is formed by a layer of cells lying over them. It accumu- 
lates on the ventral side in the pocket between the wax plates and the 
jjosterior underlapping prolongation of the preceding sternum. Wax 
is formed only on the last four visible segments, i. e., on segments 
IV-VII, inclusive. 

In studying any part of the body wall of an insect it must always 
be borne in mind that the chitin is originally simply an external cutic- 
ular layer of a true cellular skin or epidermis (erroneously called 
"hypodermis" in insects), but that in the adult stage the latter 
almost everywhere disappears as a distinct epithelium. Thus the 
chitin comes to be itself practically the entire body wall, the cell layer 
being reduced to a verj- inconspicuous membrane. However, in cer- 
tain places the epithelium may be developed for special purposes. 
This is the case with that over the Avax plates which forms a thick 
layer of cells that secrete the wax and constitute the so-called wax 
glands. The wax is first secreted in a liquid condition and is ex- 
truded through minute pores in the wax plates of the sterna, harden- 
ing on their under surfaces into the little plates of solid wax with 
which every bee keeper is acquainted. 

The secretion of the wax has been studied by Dreyling (1903), who 
made histological sections through the glands at different times in 
the life of the bee. He found that in young, freshly emerged workers 
the epidermis of the wax plates consists of a simple layer of ordinary 
epithelial cells. As the activities of the bee increase, however, these 
cells elongate while clear spaces appear between them and, when the 
highest development is reached, the epithelium consists of a thick 
layer of very long cells with liquid wax stored in the spaces between 
them. In old age most of the cells become small again and in those 
bees that live over the winter the epithelium degenerates to a simple 
sheet of nucleated plasma showing no cell boundaries. It is thus 
evident that the secretion of wax is best performed during the prime 



of life, which in bees is at about 17 days of age or before, and that 
old bees can only gather honey and pollen. Bees do not normally 
secrete wax while performing the other more ordinary duties of their 
life. When comb is needed a large number of young bees or bees 
that have not passed their prime hang together in vertical sheets 
or festoons within the hive and are fed an abundance of honey. After 
about twenty-four hours they begin to construct comb. During this 
time the. wax is excreted through the wax plates and accumulates in 

the external wax pockets below. 
It is poked out of these pockets by 
means of the spines on the feet 
and is passed forward beneath the 
body to the mandibles. By means 
of these organs it is manipulated 
into little pellets and modeled 
into the comb. Dreyling describes 
the pores of the wax plates as ex- 
cessively fine, vertical, parallel 
canals only visible in very thin 
sections and under the highest 
power of the microscope. 

Corresponding abdominal sterna 
present quite different shapes in 
the three forms of the bee (fig. 35 
A, B, and C). In the queen (B) 
the sterna are much longer than in 
the worker (A), while in the 
drone (C) they are shorter and 
have very long lateral apodemes 

The last three abdominal seg- 
ments — the eighth, ninth, and 
tenth — are very different in the 
two sexes on account of their 
modification in each to accom- 
modate the external organs of re- 
production and egg laying. In the female these segments are entirely 
concealed within the seventh, but, in the male, parts of both the 
eighth and ninth segments are visible externally. 

The seventh segment of the drone (counting the propodeum as 
the first) is the last normal segment, i. e., the last one having a com- 
plete tergum and sternum resembling those of the anterior part of 
the abdomen (fig. 56 D, VIIT and VI IS). Behind the seventh ter- 
gum and partly concealed within it is the eighth tergum {VI I IT) 
carrying the last abdominal spiracles {Sp). The eighth sternum is 

Fig. 35. — Dorsal surface of sixth abdominal 
sternum : A, worlter ; B, queen ; C, drone ; 
showing division of plate by line of at- 
tachment of intersegmental membrane 
(Jfft) into anterior part with polished 
internal surface, in worker bearing wax 
glands, and into large posterior external 
part (Kd) underlapping anterior half of 
succeeding sternum. 


almost entirely concealed within the seventh. It is very narrow 
below, but is expanded at the upper parts of its sides (VIIIS) , where 
it is partly visible below the eighth tergum and behind the seventh 
sternum. The dorsal part of the ninth segment is membranous except 
for a small apodeme-bearing plate on each side hidden within the 
eighth tergum. The ninth sternum, on the other hand, is a well- 
developed semicircular band (IXS) forming the ventral and ventro- 
lateral parts of the ninth segment. It bears on each side two con- 
spicuous lobes — one a small, darkly chitinized, dorsal plate (IClsp) 
carrying a large bunch of long hairs, the other a large, thin, ventral 
plate {2Clsp). Between these four appendicular lobes is ordinarily a 
deep cavity, which is the invaginated penis (fig. 56 E), but in 
figure D this organ is shown partly evaginated (Pen). While the 
penis is really an external organ, the details of its structure will be 
described later in connection with the internal organs of reproduction. 
The tenth segment is entirely lacking in segmental form. The anal 
opening is situated in a transverse membrane beneath the eighth ter- 
gum {VI I IT), and below it is a thin chitinous plate, which may 
belong to the tenth segment. 

In many insects the modification of the terminal segments of the 
males in cbnnection with the function of copulation is much greater 
than in the bee. The ninth segment often forms a conspicuous 
enlargement called the hypopygium, which is usually provided with 
variously developed clasping organs in the form of appendicular 
plates and hooks. 

The development of the external genital parts of the drone has been 
described by both Michaelis (1900) and Zander (1900). A small 
depression first appears on the under surface of the ninth segment of 
the larva shortly after hatching. Soon two little processes grow 
backward from the anterior wall of this pouch and divide each into 
two. The part of the larval sternum in front of the pouch becomes 
the ninth sternum of the adult, while the two processes on each side 
form the upper and lower appendicular lobes (the valva externa and 
the valva interna of Zander) . The penis at first consists of two little 
processes which arise between the valvae internae, but is eventually 
formed mostly from a deep invagination that grows forward between 
them. These four processes arising on the ventral side of the ninth 
segment of the male larva are certainly very suggestive of the similar 
ones that are formed in the same way on the same segment of the 
female and which develop into the second and third gonapophyses 
of the sting. If they are the same morphologically we must homol- 
ogize the two clasping lobes of the ninth sternum in the male with 
the two gonapophyses of this segment in the female. Zander (1900) 
argues against such a conclusion on the ground that the genital pouch 
is situated near the anterior edge of the segment in the female and 


posteriorly in the male, while the parts in the two sexes develop 
later in an absolutely different manner. These arguments, how- 
ever, do not seem very forcible — in the earliest stages the processes 
certainly look alike in the two sexes. 

The sting of the bee is situated in the sting cavity at the end of the 
abdomen, from which it can be quickly»protruded when occasioii de- 
mands. This sting chamber contains also the reduced and modified 
sclerites of the eighth, ninth, and tenth abdominal segments. In 
fact, the sting chamber is formed by an infolding of these three seg- 
ments into the seventh. It is consequently not a part of the true in- 
terior of the body or body cavity which contains the viscera, but is 
simply a sunken and inclosed part of the exterior, in the saiiie sense 
that the oven of a stove is not a part of the real inside of the stove; 
Consequently the parts of the sting, though normally hiddeif from 
view, are really external structures. 

A very gentle pull on the tip of the sting is sufficient to reftiove it 
from its chamber, but a sting thus extracted brings along with it the 
ninth and tenth segments, most of the eighth segment, the poison 
glands, and the terminal part of the alimentary canal. This is due 
to the fact that the inclosed segments are attached to the surround- 
ing parts by very delicate membranes. For the same reason they so 
easily tear from the living bee as the latter hurriedly leaves' its victim 
after stinging. The worker thus inflicts a temporary wound and 
pain at the cost of its own life. Undoubtedly, however, nature re- 
gards the damage to the enemy as of more importance to the bee 
community as a whole than the loss of one or a dozen of its members. 
The entire stinging apparatus with a bag of poison attached is thus 
left sticking in the wound while the muscles, which keep on working 
automatically, continue to drive the sting in deeper and deeper and 
at the same time pump in more poison. Such a provision certainly 
produces much more effective results than would a bee giving a thrust 
here and another there with its sting and then rapidly flying away 
to escape from danger. 

The sting itself, when extracted from its chamber, is seen to con- 
sist of a straight tapering shaft with its tip directed posteriorly and 
its base swollen into a bulblike enlargement. In superficial appear- 
ance the shaft appears to be solid, although we shall presently show 
that it is not, but the bulb is clearly hollow and is open below by 
a distinct median cleft. Several plates of definite shape and arrange- 
ment always remain attached to the sting and overlap its base. The 
entire apparatus, including the base of the large poison sac, is shown 
somewhat diagrammatically in side view by figure 36. The bulb of 
the sting (ShB) is connected with the laterai plates by two arms 
which curve outward and upward from its base. (Only the left side 
is shown in the figure.) Between these arms the two poison glands 



{PsnSc and BGl) open into the anterior end of the bulb. From the 
posterior ends of the plates two whitish fingerlike processes (StnPlp) 
project backward. When the sting is retracted these lie at the sides 
of the shaft (figs. 33 and 37), but in figure 36 the sting is shown in a 
partly protracted position. These appendages, often called the sting 
palpi, undoubtedly contain gAse organs of some sort by means of 
which the bee can tell wherf her abdomen is in contact with the object 
upon which she desir^ to use her sting. 

A close examination of the sting shows that it is a much more com- 
plicated structure than it at first sight appears to be. The shaft, for 
example, is not a ^mple, solid, tapering, spearlike rod, but is a hollow 
organ made of ithree pieces which surround a central canal. One of 
these pieces is dorsal (fig. 36, ShS) and is the true prolongation of 
the bulb (ShB), while the other two (Let) are ventral and slide 
lengthwise on tracklike ridges of the dorsal piece. Moreover, each 
basal arm of the 
sting is double, con- 
sisting of a dorsal 
or posterior piece 
(ShA); which is like- 
wise a prolongation 
of the bulb, and a 
ventral or anterior 
piece (Let) , which is 
continuous with the 
ventral rod of the 
shaft on the same 
side. Hence the sting 
may be analyzed into 
three elements, which 
are characterized as follows: The dorsal piece, known as the sheath, 
consists of a prominent basal swelling or iulb (ShB) containing a 
large cavity, of a terminal tapering sho.ft{ShS) , and of two curved 
basal arms (ShA). The ventral part consists of two long slender 
rods, called the lancets or darts (Let), which slide freely upon two 
tracks on the ventral edges of the sheath and diverge upon continua- 
tions of these tracks along the basal arms of the latter (ShA). The 
bulb is hollow, containing a large cavity formed by invagination 
from below, where it is open to the exterior by a lengthwise cleft. 
This cavity continues also through the entire length of the shaft of 
the sting as a channel inclosed between the dorsal sheath and the 
latero-ventral lancets. This channel, as will be explained later, is 
the poison canal of the sting. 

Each arm of the sheath (ShA) is supported at its end farthest 
from the bulb by an oblong plate (fig. 36, Ob), which normally over- 

FiG. 36. — Semidiagrammatic view of left side of sting of 
worlier, accessory plates {Trij Oh, Qd), sting palpus 
iStnPlp), alkaline poison gland (BOl), and base of large 
poison sac (Psn^c) of acid gland. 



laps the side of the bulb, and which carries distally the palpi of the 
sting {StnPlp). Each lancet is attached at its base to a triangular 
plate (Tri) which lies latero-dorsad to the base of the oblong plate 
and articulates with a knob on the dorsal edge of the latter by its 
ventral posterior angle. By its dorsal posterior angle the triangular 
plate is articulated to a much lav gej> quadrate plate {Qd) which 
overlaps the distal half of the oblong plate. A thick membranous 
lobe {IXS) , concave below, where it is thickly set with long hairs, 
overlaps the bulb of the sting and is attached on each side to the 
edges of the oblong plates. All of these parts are shown flattened out 

in ventral view by 
,.Bib figure 37. 

The presence of 
the two basal arms 
of the sheath might 
suggest that this 
part is to be re- 
garded as made up 
of fused lateral 
halves. In this case 
we should have six 
appendicular ele- 
ments, viz, the two 
lancets, the two 
halves of the sheath, 
and the two pal- 
puslike organs. If 
now we turn back to 
figure 8, showing 
the component parts 
of the ovipositor of 
a longhorned grass- 
hopper, we can not 
fail to be struck at 
once by the great similarity between this organ and the sting of 
the bee (fig. 36) . The first gonapophyses {!(?) of the ovipositor are 
identical with the lancets (Let) of the sting, and their sliding connec- 
tion, by means of longitudinal tracks, with the second gonapophyses 
{^G) suggests at once that the latter represent the sheath of the 
sting (ShS). The identity is still more strongly suggested when we 
observe the small bulb (ShB) formed by the fused bases of these 
gonapophyses. The third gonapophyses {3 G),' -which inclose between 
them the other parts of the ovipositor, represent the palpi of the 
sting (StnPlp). If, finally, we study the development of the parts of 
the sting we are convinced that this similarity between the sting and 
an ovipositor means something more than an accidental resemblance 

Fig. 37. — Ventral view of sting of worker and accessory parts, 
flattened out. , j p^ 


between two different organs — in fact we can not doubt that the sting 
is simply an ovipositor which, being no longer needed for egg-laying 
purposes, has been modified into a poison-injecting apparatus. Zan- 
der (1899, 1900) and others have shown that the sting of the bee 
arises from six little abdominal processes of the larva, two of which 
arise on the eighth segment and four on the ninth. Those of the first 
pair develop into the lancets, those of the middle pair on the ninth 
segment fuse to form the sheath, while those of the outer pair, be- 
come the palpi. The ovipositor, it will be remembered, develops in 
the lower insects from two pairs of processes arising on the eighth 
and ninth abdominal sterna, the second pair of which very soon 
splits into four processes. The simultaneous appearance of six on 
the bee larva is simply an example of the hurrying process or accelera- 
tion that the embryos and young of most higher forms exhibit in 
their development. 

It is only the higher members of the Hymenoptera, such as the 
wasps and the bees and their close relatives, that possess a true sting. 
The females of the lower members have ovipositors which closely re- 
semble those of such insects as the katydids, crickets, and cicadas, but 
which, at the same time, are unquestionably the same as the sting of 
the stinging Hymenoptera. It is said that the queen bee makes use 
of her sting in placing her eggs in the cells, but both the wasps and 
the bees deposit their eggs in cells or cavities that are large enough to 
admit the entire abdomen, and so they have but little use for an egg- 
placing instrument. But the females of the katydids and related 
forms like Gonocephalus (fig. 8) use their ovipositors for making a 
slit in the bark of a twig and for pushing their eggs into this cavity. 
The cicada and the sawfly do the same thing, while the parasitic 
Hymenoptera often have extremely long and slender piercing oviposi- 
tors for inserting their eggs into the living bodies of other insects. 

An examination of the sting in place within the sting chamber, as 
shown by figure 41, will suggest what the accessory plates represent in 
other less modified insects. It has already been explained that the last 
external segment of the female abdomen (fig. 32, VII) is the seventh. 
Within the dorsal part of the sting chamber is a slight suggestion of 
the eighth tergum (fig. 41, VIIIT), which laterally is chitinized as a 
conspicuous plate bearing the last or eighth abdominal spiracle {Sp) . 
The triangular plate (Tri), as Zander has shown by a study of its 
development, is a remnant of the eighth sternum, and the fact that it 
carries the lancet (Let) shows that even in the adult this appendage 
belongs to the eighth segment. The quadrate plate (Qd), since it is 
overlapped by the spiracle plates of the eighth tergum, might appear 
to belong to the eighth sternum, but Zander has shown that, by its 
development, it is a part of the ninth tergum. In many other adult 
Hymenoptera, moreover, the quadrate plates are undoubtedly tergal, 


for they are sometimes connected by a bridge behind the eighth 
tergum. The oblong plate {Oh) and its stalk represent the ninth 
sternum, and since it carries both the arm of the sheath {ShA) and 
the palpus (Pip) it still maintains its original relationships to the 
gonapophyses. The membranous lobe arising from between the 
oblong plates and overlapping the bulb of the sting (figs. 36 and 37, 
IXS) must belong to the median part of the ninth sternum. 

The tenth segment (fig. 41, X) consists of a short, thick tube having 
the anus {An) at its tip. It takes no part in the formation of the 
sting, but is entirely inclosed in the dorsal part of the sting chamber 
beneath the seventh tergum. 

In the accessory plates of the bee's sting we have a most excellent 
illustration of how the parts of a segment may become modified to 
meet the requirements of a special function, and also an example 
of how nature is ever reluctant to create any new organ, preferring 
rather to make over some already existing structure into something 
that will serve a new purpose. 

There are four glands connected with the sting, two of which 
are known to secrete the poison, which is forced through the canal 
between the sheath and the lancets and ejected into the wound made 
by the latter. It is this poison that causes the pain and inflammation 
in the wound from a bee's sting, which would never result from a 
mere puncture. The other two glands have been described as " lubri- 
cating glands," being supposed to secrete a liquid which keeps the 
parts of the sting mechanism free from friction. They lie within 
the body cavity, one on each side against the upper . edge of the 
quadrate plate, where thej'^ are easily seen in an extracted sting, each 
being a small oblong or ovate whitish cellular mass. Transverse 
microtome sections through this region show that each of these 
glands opens into a pouch of the membrane between the quadrate 
plate and the spiracle-bearing plate of the eighth tergum. Each 
gland cell communicates with this pouch by a delicate individual 
duct. The secretion of the glands is thus poured upon the outer sur- 
faces of the quadrate plates and might easily run down upon the 
bases of the lancets and the arms of the sheath, but, for all that, the 
notion that it is lubricative in function is probably entirely conjectural. 

The large, conspicuous poison sac (figs. 36, 37, 41, and 57, PsnSc) 
that opens by a narrow neck into the anterior end of the bulb of the 
sting is well known to everyone at all acquainted with bees. The 
poison which it contains comes from the delicate branched thread 
attached to its anterior end (fig. 57), a minute tube which, if traced 
forward a short distance from the sac, will be seen to divide into two 
branches, which are long and much coiled and convoluted, each ter- 
minating finally in a small oval enlargement {ACfl). These terminal 
swellings are generally regarded as the true glands and the tubes 



Fig. 38. — Section o£ 
small piece of wall of 
poison sac of sting. 


{AGID) as their ducts, but the epithelium of the tubes appears to be 

of a secretory nature also, and, if it is not, it is hard to see any reason 

for their great length. It also does not look 

probable that the two little end bodies could 

form all the poison that fills the comparatively 

enormous sac. 
The walls of the poison sac (fig. 38) are lined 

by a thick coat of laminated chitin {Int) thrown 

into numerous high folds. In the neck i5art of 

the sac the folds are arranged very regularly in 

a transverse direction and form interrupted 

chitinous rings, holding the neck rigidly open. 

The epithelium {Epth) contains nuclei {Nu), 

but the cell boundaries are very slightly marked. 

There is a distinct basement membrane (BM) , 

forming a tunica propria externallj', but there 

are no muscle fibers of any sort present except 

a few which are inserted upon the sac from some of the surrounding 

organs and which apparently act as suspensoria. 

The poison found in the sac has an 
acid reaction and is supposed to consist 
princij)ally of formic acid. Hence its 
gland is known as the acid gland {AGl) 
of the sting. 

The other sting gland is a short, very 
inconspicuous, and slightly convoluted 
whitish tube (figs. 36, 37, 41, and 57, 
BGl) opening directly into the base of 
the bulb ventrad to the opening of the 
poison sac. Its walls consist of a thick 
epithelium of distinct cells (fig. 39, 
Epth) lined with a thin chitinous in- 
tima (Int) and surrounded by a distinct 
basement membrane {BM), but, as in 
the other gland, there are ho muscles 
present. The secretion of this gland is 
said to be alkaline and the gland is 
therefore known as the alkaline gland 
{BGl) of the sting. 

Experiments made by Carlet (1890) 
show that it is only the mixture of the 
products from the two poison glands 

that is fully effective in stinging properties. Carlet's experiments were 

made upon houseflies and blowflies. He shows (1) that flies stung by a 

bee die almost instantly, (2) flies artificially inoculated with the secre- 



riG. 39.- 

-Sections of alkaline gland 
of sting. 


tion of either gland alone do not die for a long time even in spite of 
the necessary mutilation, while (3) successive inoculations of the 
same fly first from one gland and then from the other "produce death 
in a much shorter time than when inoculated from one gland alone — 
presumably as soon as the two liquids mix within the body. 

The two secretions, one acid and the other alkaline, are poured 
together into the base of the sting bulb and mix within the cavity 
of the latter. The resulting poison is then driven through the chan- 
nel in the shaft to near the tip of the latter, where it makes its exit 
into the wound. Since the large poison sac is not muscular, the poison 
is not forced through the sting by it, as is often supposed. A glance 
at figure 57 (see p. 135) will show that the accessory plates of the sting 
support several very compact sets of muscles on their inner faces. 
These muscles so act during the process of stinging that the triangular 
plates (figs. 36 and 37, Tri) turn upon their hinge-joint articulations 
with the oblong plates {Oh). By this motion of the triangular 
jjlates the attached lancets {Let) are moved back and forth along 
the tracks on the lower edges of the sheath and its arms {ShA). 
Each of these tracks consists of a ridge with a constricted base which 
dovetails into a correspondingly shaped groove on the dorsal surface 
of the lancet. This structure, as seen in cross sections through the 
shaft and bulb of the sting, is shown by fig. 40 A, B, and C. The 
lancets are thus held firmly in place, while at the same time they may 
slide back and forth with perfect freedom. The figures show also 
that all three parts of the sting are hollow, each containing a pro- 
longation {he) of the body cavity. Between them, however, is in- 
closed another cavity through which the poison flows. This is the 
'poison canal {PsnC). In the bulb (fig. 40 C) the body cavity is 
reduced to a narrow cleft {he) by the great size of the invaginated 
poison canal {PsnC). 

It will now be most convenient to describe the apparatus by means 
of which the poison is ejected from the sting. As before pointed out, 
the large poison sac can have no functions in this connection because 
its walls are entirely devoid of muscle fibers. On the other hand, 
there is an actual pumping apparatus situated within the bulb. This 
consists of two pouchlike lobes, having their concavities directed 
posteriorly, attached to the upper edges of the lancets (fig. 40 D and 
G, Vlv) on the anterior ends of the parts of the latter which slide 
within the lower edges of the bulb chamber. The lobes lie side by 
side within the bulb (fig. 40 C, Vlv) , when the lancets are in the same 
position,. and each has an accessory lamina against its own inner wall. 
When the lancets are pushed backward the walls of the lobes flare 
apart against the poison contained in the bulb and drive this liquid 
before them into the channel of the shaft, while at the same time they 
suck more poison into the front of the bulb from the glands. When, 




on the other hand, the lancets are retracted the pouches collapse so 
that they may be drawn back through the poison-filled bulb without 
resistance, but they are ready for action again as soon as the move- 
ment of the lancets is reversed. The whole apparatus thus consti- 
tutes an actual force pump in which the lobes on the lancets alter- 
nately act as a piston and as valves. The lancets need not work 
together ; in fact, 
they more often "^ -^^s 

perhaps work al- 
ternately, the lobes 
being of such a 
size as to be ef- 
fective either when 
acting together or 

The reader ac- 
quainted with 
other works on 
the anatomy of 
the bee, such as 
those of Cheshire 
(1886), Cook 
(1904), Cowan 
(1904), and Ani- 
hart (1906), will 
see often repeated 
the statement that 
the poison leaves 
the sting both > by 
a ventral opening 
between the lan- 
cets near their tips 
and by several lat- 
eral pores near the 
ends of the lancets 
opening from the 
poison canal upon 
the bases of the barbs. The writer, however, has never been able 
to observe the exit of the poison from any such lateral pores, while, 
on the other hand, it is very easy to watch it exude from between 
the lancets on the ventral side of the sting near the tip. If an 
excited bee is held beneath a microscope and the tip of the sting 
observed, the poison will be seen to accumulate in little drops near 
the tip on the ventral side. If, also, the bulb of an extracted sting 
22181— No. 18—10^ 6 

Fig. 40. — Details of sting of worker ; A, section througti tip of 
sting showing lancets (Lct\ and shaft of sheath (SAS) sur- 
I'ounding central poison canal (PsnC), and each containing 
a prolongation of the body-cavity (be) ; B, section of same 
near base of bulb ; C, section of sting through basal bulb, 
showing poison canal as large invaginated cavity (PsnC) 
in bulb of sheath (aiiB) containing the two valves (Ylv) 
of lancets (Let) ; D, part of left lancet carrying valve {Tlv), 
dorsal view ; E, tip of lancet showing pores opening on 
bases of barbs (oo) coming from body-cavity (be) of lancet — 
not from poison canal ; F, dorsal view of shaft of sheath 
showing lateral series of pores (oo) from prolongation of 
body-cavity (Be) ; G, lateral view of left valve and part of 



be squeezed gently between a pair of forceps the poison will be seen 
to emerge in the same way. In fact, it can be actually squirted out 
by a sudden compression when the bulb is well filled with poison, but 
there is never any evidence of its escape through the sides. 

An examination of the end of each lancet does reveal a number of 
oblique pores (fig. 40 E, oo) which have been figured by other writ- 
ers, and they certainly open on the bases of the barbs as described, 
but their inner ends apparently communicate wj^th the body cavity 
(be) of the lancet instead of passing clear through the lancet and 
opening into the poison canal. Furthermore, a paired series of 
exactly similar pores extends the entire length of the shaft of the 
sheath (fig. 40 F, oo), opening on its dorsal surface from the body 
cavity (ic). No one could possibly claim that the poison emerges 

Fig. 41. — Tip of abdomen of worker with left side removed, showing right halves of sev- 
enth tergum (VIIT) and sternum (VIIS), containing the sting chamber (fcfc) cut open 
along the line bio, exposing the eighth tergum (VII IT), the rudimentary tenth segment 
(Z) carrying the anus (An), and the sting and accessory parts shown by fig. 36. 

also through these pores, which, very curiously, do not appear to 
have been described before, although they are even more conspicuous 
as well as more numerous than those of the lancets. The writer has 
not been successful in preparing histological sections of the sting 
which show these pores, but they probably constitute the ducts of 
some kind of subcuticular glands. 

A cross-section through the sting a short distance in front of its 
tip shows that the lancets are here separated by a narrow cleft (fig. 
40 A), while elsewhere (B and C) they are contiguous. This cleft 
between the ends of the lancets forms the exit for the poison from the 

The sting of the queen is much longer than that of the worker 
and is more solidly attached within the sting chamber. Its shaft is 


strongly decurved beyond the bulb. The lancets have fewer and 
smaller barbs than those of the worker, but the two poison glands 
are well developed (fig. 57, AGl and BGl), while the poison sac 
(PsnSc) is especially large. 

A number of minute unicellular glands open upon the interseg- 
mental membrane between the seventh and eighth terga of the ab- 
domen. These are sometimes called the glands of Nassanoflf, after 
their discoverer. Nassanoff suggested that they are sweat glands, 
while Zoubareff thought that they form small drops of liquid said 
to be excreted by bees during flight derived from the excess of water 
in the newly collected nectar. Their function, however, has been 
much more carefully investigated by Sladen (1902), who found that 
they are scent organs producing a strong odor* even when the part 
of the back to which they are attached is removed from the rest 
of the abdomen. He furthermore identified this smell as the same 
that bees give off when a lot of them are shaken from a frame on 
the ground close to the front of the hive. Under such circumstances 
also, as in natural swarming or during the first flights in the spring 
or after a period of bad weather, bees are well known to produce a 
peculiar sound called the " joyful hum." Sladen observed that this 
was produced, in the case of bees shaken before the hive, by those 
individuals who first found the hive entrance, then by those next to 
them, until very soon all the others were informed of the location 
of the entrance and proceeded to make their way in. Also, when a 
swarm loses sight of its queen, those Hhat find her first set up this 
" joyful hum " and immediately the rest of the swarm is attracted 
to the spot. In the springtime the young bees seem to be guided 
in their flights by this same hum of the old ones. Sladen, however, 
observing the odor emitted at the same time, thinks that this and 
not the sound is the real means of information, the sound being 
simply incidental to the special movement of the wings produced 
for the purpose of blowing the odor away from the body. He argues 
that we have no evidence of an acute sense of hearing in bees, while 
it is well known that they possess a delicate sense of smell located on 
the antennae. This argument certainly seems reasonable, and we 
may at least accept Sladen's theory as the best explanation of the 
function of the glands of Nassanoff. 





It is no exaggeration to say that eating is the most important thing 
that any animal does and that its alimentary canal is the most im- 
portant organ it possesses. The entire system suffers when there is a 
deficiency in the food supply or an impairment in the digestive appa- 
ratus. Every other function is either subservient to or dependent 
upon that which furnishes nourishment to the cells. The senses of 
sight, smell, and taste are all more or less concerned in the acquisition 
of food. The muscular system enables the animal to hunt for it, to 
dig for it, to climb for it, or to chase living prey either on the ground, 
in the water, or in the air, and to kill, tear, and chew it when ob- 
tained. The blood is the servant of the stomach, for its entire func- 
tion in insects is to carry the products of digestion to the body cells. 
The heart furnishes the motor power of the blood. The respiratory 
function is accessory to that of digestion, inasmuch as it furnishes the 
oxygen which unites with the waste materials ejected from the cells 
and renders them capable of being removed fcem the blood. This 
removal is accomplished partly by the respiratory system itself and 
partly by special excretory organs. Thus %e see that the sense organs 
and the muscular system are the agents that cooperate in obtaining 
the raw food, the digestive tract is the kitchen of the body in which 
the food is prepared for use, the blood is the waiter that distributes 
it, while the respiratory and excretory systems are the refuse gath- 
erers that remove waste products. The nervous system holds the con- 
trolling power over all these organs. It regulates them in the per- 
formance of their duties and coordinates their actions so that they 
all work together. It makes a unified organism out of what would 
otherwise be simply a complex mass of" variously specialized cells. 

The reproductive function alone contributes nothing to the indi- 
vidual. In fact, the production of spermatozoa by the male and of 
eggs by the female and the nourishing of the embryo and the young 
create a demand upon all the other organs for material whjch is 
separated from the individual that produces it. But this is what the 
organism exists for; this is its reason for being. At least this is 
what it amounts to in the case of the individual, though from a wider 
philosophical standpoint the real truth is probably just the reverse, 
viz, any species exists because its individuals reproduce themselves. 

The writer has already made frequent use of the word " cell," 
assuming that the reader is familiar with the meaning of this word 
as used in anatomy and physiology. The entire body of an animal 
or plant is made up of cells or their products. The word, however, is 
misleading, for a cell is not a small sac or empty space, as was at 






Fig. 42. — Alimentary canal of worker {Phy-Bect), together with pharyngeal glands UGl), 
and salivary glands of head (20!) and of thorax (SGI), as seen by cutting body open 
from above and pulling the ventricnlus (Vent) out to left. 


first supposed from the study of plants, but is a little protoplasmic 
body or corpuscle, visible only under the microscope, surrounded by 
a membranous cell wall and containing a small internal body called 
the nucleus. The different cells of the body are specialized in groups 
to do some one particular thing— the salivary cells secrete saliva, the 
muscle cells contract, the excretory cells pick out waste substances 
from the blood, and 39- on. Btit this specialization does not signify 
that each cell does not perform its^own vital processes in addition to 
its specialty. The fact that it remains alive and works means that 
the complex chemical components of its body substance or protoplasm 
are constantly being reduced to simpler compounds which are ex- 
pelled', while new protoplasm is built up from the supply of food 
material brought by the blood. This double process of destruction 
and reconstruction is known as metaiolism, while its two phases, the 
breaking-down process and the building-up process, are known as 
katabolism and anaSoKsm, respectively. 

Now, while all the cells of the body must have nourishment, none 
of them, except those of the alimentary canal, is capable of utiliz- 
ing the raw food materials that an animal obtains in a state of nature. 
These materials must therefore be changed into some other form in 
order that they may be assimilated by the cells. This change is called 
digestion. ^ 

The single cell composing the body of a Protozoan, living free in 
nature, digests its own food and then assimilates the products of its 
own digestion. But, of the cells constituting the body of any mul- 
ticellular animal, only those of the alimentary canal are capable of 
digesting raw foodstuffs, and, moreover, as digestion is the specialty 
of these cells, they have also to digest the food for all the other ^pUs 
of the body. 

The two most important changes that must be brought about in 
the natural food by digestion are those which make it soluble in the 
blood and which render it capable of passing through animal tissues. 
In the first place, the food must diffuse through the walls of the 
alimentary canal as a liquid which mixes with the blood, for there 
are no pores or openings of any sort from the alimentary canal into 
the body cavity; and in the second place, it must pass through the 
walls of the cells themselves. The digestive changes result chiefly in 
a breaking down of the complex molecules of the raw food materials 
into more simple chemical substances. These are taken up by the 
cells and reconstructed into complex protoplasmic molecules which 
can not escape through the cell membrane until they are again broken 
down into simpler forms. 

The waste products of the cells consist principally of carbon, hy- 
drogen, and nitrogen. These are converted by the oxygen supplied 
by the respiratory system into carbon dioxid, water, and compounds of 



urea. The first, being a gas, mixes with the air in the tracheal tubes 
and so reaches the exterior during exhalation. Much of the water is 
also given off through the tracheal system in the form of vapor which 
exhales from the spiracles, but, since insects are covered by their 
hard chitinous shell, it is probable that they do not " sweat." The 
compounds of urea, and probably also some water, are separated 
from the blood by the excretory glands, called Malpighian tubules 
in insects, which empty their products back into the alimentary 
canal, whence they are discharged with the faeces from the intestine. 

Digestion is brought about by substances called enzymes which are 
contained in the various liquids mixed with the food in the alimentary 
canal. These liquids are secreted by the salivary glands and by the 
cellular walls of the stomach. 


The opening of the salivary duct on the base of the proboscis has 
already been described (see pp. 49-51). The true salivary glands, or 
those corresponding with the salivary glands of other insects, are 
arranged in two pairs, one situated within the head (figs. 19 and 42, 
2GI) and the other within the thorax (fig. 42, 3GI) . The four ducts 
unite into one median tube, which enters the base of the labium (fig. 
19, SalD) and opens upon the upper surface of the ligula (fig. 15 F, 
and fig. 16, SalDO). The large and conspicuous glands lying within 
the anterior and upper parts of the head and opening into the 
pharynx will be described later in connection with this organ. They 
are special pharyngeal glands in no way homologous with the salivary 
glands of other insects, and are by many supposed to secrete the 
brood food instead of a digestive liquid like saliva. 

The salivary glands of the head {System No. '2 of Cheshire, fost- 
cerehral glands of Bordas) lie against the posterior walls of the 
cranium. In the worker each consists of a loosely arranged mass of 
pear-shaped follicles or acini whose individual ducts unite irregu- 
larly with one another and eventually form a common duct on each 
side (figs. 19, 42, and 43 F, 2GI) . Their two ducts unite with the 
median duct from the thoracic glands just before the bases of the 
mesocephalic pillars (fig. 19). In the drone these glands have a 
quite different appearance from those of the female, each consisting 
of a compact mass of very small follicles connected by minute ducts 
and flattened against the posterior walls of the head (fig. 43 B and C, 
2GI). A large lobe of this gland in the drone extends forward on 
each side against the face, between the compound eye and the clypeus 
(fig. 10 C, 2GI) , thus occupying the position of the large mandibular 
gland in the worker (A, IMdGl) and in the queen (B, IMdGl). 
There is also a prominent triangular mass of glandular cells in the 
drone situated just above the ocelli (fig. 10 C, 2GI) which has been 



described, by Bordas (1895) as a separate gland opening by two ducts 
into the oesophagus just behind the pharynx. The writer, however, 
has been utterly unable to discover any such ducts, though two sus- 
pensorial ligaments of the anterior end of the oesophagus are at- 
tached to the wall of the head at the posterior ends of these glands 
(fig. 11 B, g) and might easily be mistaken for ducts. These " post- 
ocellar glands " of Bordas, moreover, appear to be simply detached 
lobes of the postcerebral glands. They are prominent also in the 
queen (fig. 10 B, 2GI) and are represented by a few follicles in the 

Fig. 43. — A, small piece of large lateral pharyngeal glands in head of worker ; B, piece of 
postcerebral salivary glands in head of drone ; C, postcerebral glands ($GZ) in normal 
position against posterior wall of head in drone ; D, pharyngeal plate (s) of worker, 
ventral view, showing bases of lateral pharyngeal glands (,1GI) and their receptacula 
(mm), and median ventral pharyngeal gland (iGl) ; E, corresponding view of pharyngeal 
plate of drone, showing entire absence of lateral pharyngeal glands, and greater devel- 
opment of small median glands i.'iGl) ; F, part of postcerebral gland of worker. 

Bordas describes the follicles of the postcerebral glands in the 
worker as hollow sacs, each having a large lumen lined with a chiti- 
nous intima. Their secretion, he says, is a thin viscid liquid, pale 
yellow in color and having a slightly alkaline reaction. According to 
Schiemenz (1883) each gland is developed as an outgrowth from the 
common duct of the thoracic glands. 

The salivary glands of the thorax in the bee {System No. 3 of 
Cheshire, thoracic salivary glands of Bordas) are the ones that cor- 
respond with the ordinary salivary glands of other insects. They 
are described by Schiemenz (1883) as being formed inside of the 


outer covering (tunica propria) of the first part of the larval silk 
glands. But it is of common occurrence in insects that the salivary 
glands are temporarily specialized as silk-producing organs in the 
larva. In the adult worker these glands lie in the ventral part of 
the anterior half of the thorax (fig. 42, 3GI). The two are widely 
separated anteriorly, but their posterior ends are contiguous. Each 
consists of a mass of small, many-branched, glandular tubes opening 
into several collecting ducts which empty into a sac near the ante- 
rior end of the gland (II) . From each of these reservoirs, then, a duct 
{Dct) runs forward and fuses with the one from the opposite side 
just within the foramen magnum of the head. The common duct 
thus formed turns downward within the head, receiving the two ducts 
of the postcerebral salivary glands and then enters the base of the 
mentum (figs. 19 and 43 C, SalD) , to open as already described on the 
upper side of the ligula at the root of the glossa and between the 
bases of the two paraglossse (fig. 15 F and 16, SalDO). The 
secretion of the thoracic glands is said also to be weakly alkaline. 
Therefore the entire salivary fluid poured out upon the labium is 
alkaline, and it must be designed to act especially upon the food 
taken through the proboscis. This action, furthermore, on account of 
the location of the salivary opening, may take place before the food 
enters the mouth. 

The food of the bee consists normally of pollen, nectar, and honey. 
The first is eaten entirely with the mandibles, while the other two are 
taken through the proboscis. The pollen is to the diet of the bee what 
meat is to ours; that is to say, it contains the proteid or nitrogen- 
containing ingredient of the food which is necessary to the sup- 
port of any animal, and also substances comparable with fat, called 
in general hydrocarbons. The nectar and honey consist principally 
of grape sugar, fruit sugar, and cane sugar, which belong to the class 
of chemical substances known as carbohydrates. Now, all of these 
foodstuffs, except the grape and fruit sugars, have to be changed 
chemically by the digestive process before they can be absorbed into 
the blood. The pollen, which contains the proteids and hydrocarbons 
of the food, is taken directly into the mouth by means of the man- 
dibles and apparently is not digested until it reaches the small in- 
testine, and therefore it would seem that it is the cane sugar which 
must be affected by the saliva. The change, or inversion, as it is 
called, of cane sugar, which has a very large molecule (CioHj.Oii), 
consists of its reduction to grape and fruit sugars which have smaller 
molecules (CeHi^Oe). Starch (CeHioOj) must also be reduced to 
simpler and more soluble compounds before it is capable of absorp- 
tion. Its inversion is effected in us partly by the saliva, but starch 
appears to form a very inconsiderable element in the bee's diet. 



The alimentary canal is a tube which extends through the entire 
length of the body and, on account of being more or less coiled, it is 
generally considerably longer than the length of the body in insects. 
It has no openings of any sort into the body cavity. The internal 
organs are packed closely about it, and the interstices are filled with 
the blood, there being no special arteries or veins in insects. The 
amount of space occupied by the alimentary canal varies according to 
the amount of food it contains, and for this reason it seldom looks 
exactly alike in any two individuals examined. 

The part of the canal immediately following the mouth forms an 
enlargement (fig. 42, Phy) called the pharynx. Succeeding this is 
a slender tube which leaves the head by the foramen magnum above 
the small transverse tentorial bar and traverses the entire length 
of the thorax. This is the (esophagus {(E)- In the -anterior part of 
the abdomen the oesophagus expands into a large thin-walled sac 
which is ordinarily called the crop or ingluvies, but which, in the 
bee, is known as the honey stomach (US). Behind this is a short, 
narrow, necklike division, with rigid walls constituting the pro- 
ventricuhis (Pnent) . Then comes a large U-shaped part, with thick, 
spongy-looking walls containing numerous annular constrictions. 
This is the ventriculus {Vent), or stomach, of the bee, frequently re- 
ferred to as the " chyle stomach." Following the ventriculus is a 
short, narrow, coiled small intestine {SInt) having a circle of about 
one hundred long, greatly coiled, blind, threadlike tubes opening into 
its anterior end. These latter are called the Malpighian tubules 
{Mai). Functionally they^do not belong to the digestive tract, since 
they are excretory organs, corresponding with the nephridia of other 
invertebrates and with the kidneys of vertebrates. Following the 
small intestine is the largv intestine, or rectum {Rect), which is often 
distended by its contents into a great sac occupying a large part of 
the abdominal cavity. Six whitish bands on its anterior end are 
called the rectal glands {RGl). The rectum opens to the exterior 
through the anus, which is situated, as already described, at the end of 
the rudimentary tenth or last segment of the abdomen (fig. 41, An). 

After this brief general survey of the parts of the alimentary 
canal, we shall proceed with the description of each in detail, and at 
the same time give what is known of the role each plays in the 
process of digestion. What is known, however, about digestion in 
the bee, or in any insect, for that matter, really amounts to nothing, 
but the views of various writers on the subject must be discussed 
briefly, in order to show how little has actually been demonstrated. 

The pharynx (figs. 11 B, 19, and 42, Phy) lies in the anterior part 
of the head close behind the clypeus, extending from the mouth 


dorsally to above the antennae, where it turns posteriorly and con- 
tracts into the much narrower oesophagus (CE). Attached to its 
walls are numerous suspensorial muscles, whose contraction must 
expand the pharyngeal cavity, while the latter may be contracted 
by the sheet of muscles surrounding its walls. In this way the 
pharynx is undoubtedly able to perform a sucking action, by means 
of which the liquid foods are taken into the mouth. Its lateral 
walls are strengthened by two long, chitinous rods (figs. 11 B and 
19, A) , which arise from a median anterior plate in its floor (fig. 19, s) . 
The anterior end of this plate is prolonged into two free, tapering 
lobes which hang down over the lower rim of the mouth. The plate, 
in the worker, and the bases of the rods are shown in ventral view, 
removed from the pharyngeal wall, in figure 43 D. Near where the 
rods join the plate are two long, chitinous pockets (mm), opening 
above, which receive the ducts of the two large glands (JGl) lying 
within the anterior part of the head. Between these two pockets is a 
transverse row of cells (4(^1), which have been described by Bordas 
(1895) as the " sublingual glands," but this name is not appropriate 
in insects, for, while the gland in question may be suggestive of the 
sublingual salivary gland of vertebrates, it does not lie beneath the 
tongue or lingua of the bee. Although the pharyngeal plate lies 
upon the floor of the true mouth, it is not, as already explained (p. 
44) , the equivalent of what is properly called the tongue, lingua, or 
hypopharynx in other insects — this organ being absent in most 
Hymenoptera. The only suggestion the writer can make, however, 
is to call this group of cells the ventral or median ventral pharyngeal 
gland in distinction to the large lateral glands. A comparative view 
of the pharyngeal plate and its accessory parts in the drone is given 
in figure 43 E. The plate itself (s) is shorter than in the worker, 
and its anterior lobes are smaller. The lateral glands and their 
receptacula are entirely absent, but the median glands (4<?0~.are 
much larger than those of the worker. Bordas says that each acinus 
of the latter glands in both the worker and the drone is provided 
with a fine, sinuous canaliculus, and that these tiny ducts open 
separately in two bundles on the lateral parts of the pharyngeal 
plate. The' lateral glands are present in the queen, but are very small 
and rudimentary. 

Especial interest attaches to the large lateral pharyngeal glands of 
the worker {System No. 1 of Cheshire, the sufracerehral glands 
of Bordas) , because they are regarded by many as the source of the 
brood food and the so-called " royal jelly," which is fed to the larvae 
and to the adult queens and drones by the workers. Each consists 
of a long coiled string of small ovate follicles attached to one median 
duct (fig. 43 A) and the two are intricately packed into the anterior 
and upper parts of the head (figs. 10 A, 19, and 42, IGl). Each 


acinus consists of a solid mass of several small cells, which are pene- 
trated by a large number of fine, chitinous ducts, arising in the neck 
of the acinus from the common duct of the gland. These follicular 
ducts can be very clearly shown by treating a part of the gland with 
weak caustic potash, which dissolves the protoplasm of the cells 
and brings out the bunch of ductules very clearly. 

The fact that these glands are entirely absent in the drone and at 
best rudimentary in the queen shows that they must in some way be 
connected with the special functions of the worker. Schiernenz (1883 ) 
and Cheshire (1886) have shown that their development in the dif- 
ferent species of bees is in proportion to the social specialization. 
They vary from a group of cells opening by separate ducts upon the 
pharyngeal plate to the highly developed condition they present in 
the honey bee. The writer questions, however, whether these authors 
did not mistake the median pharyngeal glands of these lower genera 
of bees for rudimentary representatives of the lateral glands. Bordas 
states that the former occur in all Hymenoptera, but Schiernenz and 
Cheshire did not seem to recognize them. The bumblebees (Bombus) 
have them almost as well developed as the honey bee (Apis), espe- 
cially the large females. In the genus Psythirus they are similar to 
those of Bombus but are smaller, while in such genera as Andrena 
and Anthophora they are rudimentary or consist of a few scattered 
cells. Both Schiemenz and Cheshire have thus argued strongly that 
these glands of the pharynx are the organs that produce the brood 
food. On the other hand, Schonfeld (1886) has made an equally 
strong plea in favor of the ventriculus as the producer of this impor- 
tant material. He believes that the brood food, especially royal 
jelly, is regurgitated chyle. Both Schonfeld and Cook (1904) fed 
bees in a hive some honey containing powdered charcoal and later 
found this in the brood food in the comb cells, thus apparently con- 
firming its ventricular origin. However, the charcoal that got into 
the cells might have come from the mouth, the oesophagus, or the 
honey stomach. It, of course, could not have gone through the 
stomach walls and entered the pharyngeal glands, as proved by Dr. 
J. A. Nelson, of this Bureau, from microtome sections of bees fed on 
lampblack. The arguments, then, in favor of the stomach and the 
pharyngeal glands seem equally strong, and perhaps the truth is, as 
occurs in so many such cases, that both sides are right — ^that the brood 
food is a mixture of chyle from the stomach and of secretion from 
the pharyngeal glands. 

Arnhart (1906) seems to adopt the position that the brood food 
is chyle which has been acidified by the addition of an acid from the 
glands. He states that the acid reaction of the royal jelly is due to 
the presence of three- fourths of 1 per cent of tartaric acid. The 
contents of the ventriculus, on the other hand, and for that matter 


of all the parts of the alimentary canal, are alkaline. Hence, it 
seems very logical to suppose that if the brood food comes from the 
stomach, its acid constituent is furnished by the glands in the head. 
But the difference between the brood food found in the cells and the 
contents of the ventriculus is so great that it would seem as if a very 
substantial addition of something more than a mere preservative acid 
must be made to the latter. 

The brood food given to the queen larvae, known as royal jelly, is a 
gummy paste of a milky-white color when fresh, but when taken out 
of the cell it soon acquires a darker tone with a yellowish tint. Under 
the microscope it appears to be a homogeneous, very minutely granu- 
lar mass. It is very acrid and jDungent to the taste, and must be 
strongly acid. Samples examined by the writer taken from cells 
containing queen larvae two and four days old contained a number of 
fresh undigested pollen grains but no bits of hairs such as occur in 
the stomach. 

The possible ventricular origin of a part of the brood food and its 
regurgitation will be further discussed when we treat of the stomach 
(page 98). The writer does not advocate any personal view regard- 
ing the origin of this larval food — the fact is, there is not enough 
known about it to enable one to formulate any opinion worth while. 
We know only that the whitish paste comes out of the mouths of the 
workers, but we know nothing of where it is made or of how it is 
made. Hence we can but await the evidence of further investigation. 

The brood food is fed to the larvae by the workers and is produced 
in greatest abundance by the younger individuals. The larvae of the 
queens are said to receive nothing but pure royal jelly throughout 
their entire developmental period, while the larvae of the drones and 
the workers are given the pure product only during the first three 
days of their life. From the beginning of the fourth day on, honey 
is said to be mixed with the diet of the drones and workers and, in 
the case of the former, undigested pollen also. Moreover, the adult 
queens and the drones receive a certain amount of prepared food 
throughout their lives ; if they do not get it they become weak. While 
they can feed themselves with honey they apparently can not eat 
pollen, and consequently are not able to obtain the proteid element of 
diet unless fed this in a predigested condition by the workers. Dur- 
ing egg-laying activity the queen especially demands this food, and 
by furnishing or withholding it the workers probably have the power 
of stimulating or inhibiting her production of eggs. Arnhart (1906) 
says that the workers feed it to weak or starved members of their own 
class, the material being accumulated upon the upper surface of the 
mentum of one bee whence it is sucked up through the proboscis by 
the other. All of these statements, however, concerning the feeding 
of the brood and the differences in the diet need to be verified. They 




are based chiefly on the work of Planta, published in 1888. Cheshire 
(1886) states that the stomachs of queens contain a substance which 
is " microscopically indistinguishable from the so-called royal jelly," 
scarcely a pollen grain being discoverable in it. If this is so, it would 
seem to prove that the queen is fed this substance by the worker, for 
the stomach of the latter is invariably filled with a dark-brown slime 

containing a vary- 
/-c- Ul --rj ing amount of pol- 

len and in no way 
resembling royal 
jelly. Cheshire 
further says that 
before impregna- 
tion the stomachs 
of the queens al- 
ways contain pol- 
len, the royal jelly 
being found in 
them two or three 
days after impreg- 
nation, when all 
traces of pollen 
have disappeared. 

The narrow 
oesophagus (fig. 42, 
(E)isa. simple tube 
with a thick chiti- 
nous lining and 
muscular walls. 
The epithelium (fig. 
45) is very rudi- 
mentary, its cell 
boundaries being 
lost and its nuclei 
(Nu) appearing as 
if imbedded in the 
lower layers of the 
thick transparent 
intima (Int). The muscles are disposed in an outer layer*' of trans- 
verse fibers {TMcl) and an inner layer of longitudinal ones (LMcl). 
The honey stomach (fig. 42, HS) is simply an enlargement of the 
posterior end of the oesophagus lying within the anterior part of 
the abdominal cavity. It is best developed in the worker (fig. 44 A) , 
but is present also in the queen (B) and in the drone (D). The 
organ should perhaps have been named the nectar stomach, for its 

Fig. 44. — A, honey stomach (HS) of worker with posterior end 
of oesophagus (CE), proventriculus (Pvent), and anterior 
end of ventriculus (Vent) ; B, same of queen; C, honey- 
stomach {.HS) of worker mostly cut away exposing the 
stomach-mouth (mi) of proventriculus (Pvent) leading into 
ventriculus (Vent) ; D, honey stomach of drone. 


principal function in the bee is to hold the nectar as it is collected 
from the flowers and to allow the worker to accumulate a consider- 
able quantity of this liquid before going back to the hive. Hence, 
since the honey stomach is a sac with very distensible walls, its 
apparent size varies greatly. When empty it is a small flabby pouch, 
but when full it is ah enormous balloon-shaped bag with thin tense 
walls. The histological structure of the honey stomach (fig. 4.5, HS) 
is exactly the same as that of the oesophagus. The numerous high 
folds into which its epithelium (Epth) is thrown permit the enor- 
mous expansion of which the sac is capable. When a worker with 
its honey stomach filled with nectar reaches the hive, the nectar is 
either stored directly in a cell or is given up first to some other 
worker, who places it in a cell. 

It would appear that all the food swallowed by a bee must go first 
into the honey stomach, and since the bee's diet consists of pollen and 
honey as well as nectar, one would suppose that in regurgitating the 
latter the bee would also disgorge the pollen it might have recently 
eaten. Honey which is made from the regurgitated nectar does 
indeed contain some pollen, but most of the pollen eaten by the bee 
is undoubtedly retained in the stomach as food. The apparatus by 
means of which the pollen is supposed to be separated from the nec- 
tar belongs to the following division of the alimentary canal, but it 
is not known that the worker takes nectar, and pollen for food, into 
its honey stomach at the same time. 

The proventriculus (figs. -±2 and 44, Pvent) forms the necklike stalk 
between the honey stomach [HS) and the true stomach or ventricu- 
lus {Vent), but a very important part of it also projects up into the 
honey stomach (fig. 44 C). If the honey stomach be slit open, a 
short, thick, cylindrical object will be seen invaginated into its pos- 
terior end and having an X-shaped opening at its summit (fig. 44 C, 
nn). This opening is the mouth of the proventriculus, and its four 
triangular lips, which are thick and strong, mark four longitudinal 
ridges of the proventricular tube. This structure is commonly known 
as the " stomach-mouth " and is supposed to be an apparatus de- 
signed especially to enable the worker to pick out pollen grains from 
the honey stomach and swallow them on down into the true stomach 
or ventriculus, while the -nectar is left to be stored in the hive. 
Cheshire says : " While the little gatherer is flying from flower to 
flower her stomach-mouth is busy separating pollen from nectar." 
This notion is so prevalent among bee writers in general that it 
passes for a known truth. Yet it has really never been shown that 
the worker eats pollen while she is gathering nectar. Probably no 
more pollen is ever mixed with the nectar in the honey stomach than 
is found in the honey itself. Furthermore, under normal conditions 
pollen never accumulates in the honey stomach, even when the bee 


is not collecting nectar — or, at least, the writer has not observed it — 
while, finally, both the proventriculus and its mouth are just as well 
developed in the queens and drones as in the workers, though neither 
of the former are known to eat pollen, and they certainly do not 
gather nectar. 

If the honey stomach be cut open in a freshly killed bee, the 
pro ventricular mouth may be seen still in action.. The four lips 
spasmodically open wide apart with a quivering motion and then 
tightly roll together and sink into the end of the proventricular 
lumen. This, of course, suggests their picking pollen out of the 
nectar, but it is probably simply the ordinary process by means of 
which the proventriculus passes any of the food in the honey stomach 
on to the ventriculus. Nearly all insects have some such proventricu- 
lar apparatus, which simply takes the stored food from the crop as 
it is needed by the stomach. In some insects it forms apparently a 
straining apparatus, which prevents coarse, indigestible fragments 
from entering the stomach, while in some the proventriculus may be 
a triturating organ comparable with a bird's gizzard. Bees, how- 
ever, do not crush the pollen either in their mandibles or in the 
proventriculus, for it occurs in perfect condition in the ventriculus. 

Hence, before the current notion that the " stomach-mouth " is 
for the special purpose of taking pollen out of the nectar in the 
honey stomach can be accepted it must be first demonstrated that 
the workers eat pollen while the honey stomach contains nectar to 
be stored in the cells, i. e., any more than is disgorged along with 
the nectar; and, secondly, a reason must be shown why the queens 
and drones should have a " stomach-mouth " as well developed as 
that of the worker. In the meantime it appears most logical to 
regard the proventricular mouth as simply the ordinary apparatus, 
"possessed by insects in general, by means of which all of the food is 
passed from the crop to the stomach. 

A longitudinal section through the honey stomach, the proventric- 
ulus, and the anterior end of the ventriculus is shown in figure 45, 
which is made from a queen. The proventriculus does not differ from 
that of a worker, but the honey stomach is smaller and not so much 
turned to one side (cf. fig. 44 A and B). The two muscle layers of 
the oesophagus continue down over the walls of the honey stomach 
(TMcl and LMcl). The outer layer of transverse fibers, however, 
ceases at the posterior end of this organ, while the longitudinal fibers 
continue posteriorly over the proventriculus and the ventriculus as 
an external layer {LMcl) . A new layer of internal transverse fibers 
begins on the proventricular walls and extends backward on the 
ventriculus {TMcl) beneath the longitudinals. Hence the muscles 
on the oesophagus and crop are in reverse order from those of the 
proventriculus and ventriculus. The proventriculus is deeply in- 



vagina ted into the posterior end of the honey stomach. Each lobe 
of its mouth forms a thick triangular ridge on the walls of its 
lumen, in which lies a special mass of longitudinal muscle fibers 
(LMcl) . The epithelium of the lumen is lined by a thick, smooth, 
chitinous intima (Int), while the lobes of the mouth (nn) are pro- 
vided with bristles point- 
ing inward and backward 
into the mouth opening. 

The posterior opening 
of the proventriculus into 
the ventriculus is guarded 
by a long tubular fold 
of its epithelium (fig. 45, 
PventVlv), the pro ventric- 
ular valve. This would 
appear to constitute an 
effective check against the 
escape of any food back 
into the proventriculus. It 
looks like one of those traps 
which induces an animal to 
enter by a tapering funnel 
but whose exit is so small 
that the cajDtive can not 
find it from the other side. 
Yet Schonfeld has elab- 
orately described experi- 
ments by means of which 
he induced the ventriculus 
to discharge its contents 
through the proventriculus 
into the honey stomach and 
even into the end of the 
oesophagus. He says that 
he did this by gently tap- 
ping on the honey stomach 
and the ventriculus at the 
same time. The experiment 
was repeated manj' times with unvarying results and Schonfeld de- 
scribes so minutely what happened that we can not disbelieve his 
statements. From these experiments he argues that the larval food- 
stuff is prepared in the stomach and regurgitated through the proven- 
triculus directly into the oesophagus by a contraction of the honey 
stomach which brings the stomach-mouth against the base of the oesoph- 
22181— No. 18—10 7 




Fig. 45. — Longitudinal median section of base of 
cesopliagus ((B),lioney stomaeli (7/S), proventricu- 
lus (.Pvent) and ventriculus {Vent) of a queen. 


agus. We shall have to postpone a further discussion of this subject 
to page 99, after the ventriculus and its contents have been described. 

The ventriculus (fig. 42, Ve7it) is the largest part of the alimentary 
canal in the bee and is bent into a U-shaped loop of which the pos- 
terior arm is dorsal. It is cylindrical and does not vary so much in 
shape and diameter according to its contents as do the other parts of 
the canal, although the numerous transverse constrictions which give 
it a segmented appearance are not at all constant. When examined 
under alcohol the ventriculus has an opaque whitish appearance, but 
in the natural condition — that is, as seen when examined in a freshly 
killed or asphyxiated bee — it is of a dark -brown color with lighter 
rings corresponding to the constrictions. The latter represent in- 
ternal folds where the walls are really thicker than elsewhere, the 
color being due to the contents which naturally show more plainly 
through the thin parts. 

The contents of the ventriculus invariably consist of a dark brown 
mucilaginous slime and generally also of a varying amount of pollen. 
The latter is most abundant in the posterior arm of the ventricular 
loop and is often densely packed in its rear extremity, while the an- 
terior arm may be almost entirely free from it. The pollen in the 
ventriculus is always fresh-looking, the native color showing dis- 
tinctly through the enveloping slime while most of the grains yet re- 
tain all of their contents. The writer has examined many samples 
of pollen from the stomachs of workers and, in all, the great mass of 
it showed no evidence of digestion, the color being fresh and the 
contents perfect — only a few had the latter shrunken and seldom was 
an empty shell observed. On the other hand, the pollen contained 
in the small intestine has invariably lost its bright color, the contents 
of the majority of the grains are more or less shrunken, while a num- 
ber of empty shells are to be found. That in the rectum, finally, con- 
sists in large part of empty shells or of grains having the contents 
greatly shrunken and apparently mostly dissolved out, although a 
few perfect and bright-colored. grains are always present, looking as 
if entirely unaflFected by the digestive liquids. From these observa- 
tions the writer would conclude that the digestion of pollen takes 
place principally in the intestine. In all parts of the alimentary 
tract there occur numerous bits of feathered bee-hairs, but these seem 
to be especially numerous in the ventriculus. 

We are now in a position to discuss the possibility of the production 
of the brood food in the stomach. ScEonfeld (1886), as has already 
been stated, argues that this substance is regurgitated " chyle " from 
the ventriculus. Arnhart (1906) adopts this view and elaborates 
considerably upon the chemical process by means of which the trans- 
formation of " chyle " into this larval food is effected through the 
addition of tartaric acid from the pharyngeal glands of the head. 


The ventricular contents do become slightly milky when treated with 
a solution of tartaric acid, but they are not changed into anything 
at all resembling royal jelly. Moreover, a transformation of the 
brown slimy contents of the ventriculus into the white gummy paste 
on which the larvae are fed does not seem possible without the addi- 
tion of much other material. In fact the added material must make 
up the conspicuous part of the larval foodstuff and, from a purely 
argumentative standpoint, it would not seem necessary to assume that 
it contains any " chyle " at all. Again, if it were not for Schonfeld's 
experiments one could not easily believe that the food could be dis- 
gorged through the proventricular valve. The conspicuous action of 
the proventricular mouth is a swallowing motion, and the writer has 
not been able to induce the ventriculus to disgorge its contents 
through it in the way that Schonfeld describes, although perhaps 
suiEcient care was not observed in exposing the organs. Cheshire 
states that. the proventricular tube (fig. 45, PventVlv) in the ventricu- 
lus " rather makes regurgitation improbable than impossible," while 
he argues that the down-pointing bristles of the stomach-mouth would 
further interfere with this process. Cowan adopts the view of 
Dufour and Schonfeld that the brood food is of ventricular origin, 
and says in this connection: ''Although saliva from the glands 
(especially System I) is probably added to the food, this can not, 
from its great variability, be entirely a secretion, as stated by 
Schiemenz. The work of Doctor Planta, we think, conclusively proves 
that the food is not a secretion, and that the nurses have the power 
of altering its constituents as may be required for the different bees." 
If the variation of the food is under the control of the workers pro- 
ducing it, it does indeed look impossible that it should be produced ' 
entirely by glands. Cov/an illustrates by a diagram how regurgita- 
tion through the proventriculus may be possible in spite of the pro- 
ventricular tube projecting into the ventriculus. Since this tube is 
simply a cylindrical fold its walls, as shown in figure 45, PventVlv, 
consist of two layers, and Cowan says that " when the bee wishes to 
drive the chyle food from the chyle-stomach (Vent) into the cells 
it forces the stomach-mouth {nn) up to the oesophagus {(E) and the 
prolongation {PventVlv) unfolds, extending the chyle-stomach to the 
oesophagus, making a direct communication through which the food 
is forced by compression of the chyle-stomach by its muscles." The 
honey-stomach of the worker is much larger than that of the queen, 
shown by figure 45, in which there is not enough space for the unfold- 
ing of the proventricular tube. This mechanism suggested by Cowan 
looks simple and conclusive in a diagram, but when one attempts to 
unfold the proventricular tube by grasping the stomach-mouth in a 
pair of fine forceps and pulling the top of the proventriculus upward 
it is found that, while the tube can be entirely straightened out, doing 


SO involves the tearing of all the fine muscle fibers and tracheal 
branches uniting the honey-stomach to the upper end of the ventricu- 
lus (fig. 45). If, then, the organ itself can not be made to work 
according to this scheme, it might be supposed that the inner wall of 
the proventriculus and the tube are evaginated through the stomach- 
mouth (nn), but the walls of the former certainly appear to be en- 
tirely too rigid to permit of any such performance as this. Finally, 
it is not clear how any eversion of the tube could be produced by the 
proventricular muscles as they exist. 

The various facts and arguments bearing on the origin of the 
brood food may be summarized as follows: 

1. The brood food itself is a milky-white, finely granular, and 
gummy paste having a strong acid reaction said to be due to the 
presence of tartaric acid. 

2. The pharyngeal glands of the head are developed in proportion 
to the social specialization of the various species of bees; they are 
always largest in those individuals that feed the brood, and they 
reach their highest development in the workers of the honey bee. 
From this it would seem that they are accessory to some special 
function of the worker. 

3. The contents of the stomach in the workers consist of a dark 
brown, slimy, or mucilaginous substance in no way resembling the 
brood food, even when acidulated with tartaric acid. Pollen is 
present in varying quantity, mostly in the posterior end of the 
stomach, and shows little or no evidence of digestion. Since the 
brood food is highly nutritious, it must contain an abundance of 
nitrogenous food material which is derived only from pollen in the 
•bee's diet. Therefore it is not clear how the stomach contents can 
alone form brood food. 

4. The constituents of the food given to the different larvae, at 
different stages in their growth, and to the adult queens and drones 
show a constant variation apparently regulated by the workers pro- 
ducing it. A variation of this sort can not be explained if it is 
assumed that the brood food is produced by the glands alone. 

5. Powdered charcoal fed to a hive of bees appears after a short 
time in the brood food in the cells, and this has been urged as proof 
that the latter is regurgitated " chyle." But it is certainly entirely 
possible that the charcoal found in the food might have come only 
from the honey stomach or even from the oesophagus or mouth. 

6. We have Schonfeld's word for the statement that a regurgita- 
tion of the stomach contents may be artificially induced by irritation 
of the honey stomach and ventriculus in a freshly dissected bee, but 
all explanations offered to show how this is mechanically possible 
in spite of the proventricular valve are unsatisfactory when the 
actual anatomical structure is taken into consideration. 


The only conclusion, then, that we are really warranted in draw- 
ing concerning the origin of the royal jelly or of any of the larval 
food paste is that we do not know anything about it. Cheshire is 
probably responsible for the widespread opinion that it is formed 
by the pharyngeal glands, though Schiemenz (1883) published a 
large paper containing this theory three years before Cheshire's 
book was printed. The " chyle " theory, which also has many advo- 
cates, originated with Dufour but was principally elaborated by 
Schonfeld. Arnhart would derive the brood food from both the 
stomach and the glands. But we are still absolutely in the dark, 
since we lack definite and conclusive information. A satisfactory 
study of the subject would involve the chemical investigation of 
vei-y minute quantities of substances, and it may be a long time before 
any interested person is found capable of undertaking a work of this 
sort. The writer of the present paper is professedly preparing an 
account only of the structure of the organs, but is doing this with 
the hope that it may furnish a basis for some future investigator who 
shall have time to devote himself to a study of the chemistry and 
physiology of the digestive organs and their glands. 

In vertebrate animals the digestive secretion of the stomach is acid 
and its enzymes bring about especially the digestion of proteids. The 
resulting acid mixture is called chyme. In the intestine the contents 
are flooded with various alkaline liquids whose enzymes then take up 
the digestion of the other food elements. The final prepared product, 
which is absorbed by the lacteals, is called chyle. These names have 
been applied to the contents of the alimentary canal in insects — espe- 
cially by Arnhart (1906), who speaks of the material undergoing 
digestion as " chyme " and the completed products as " chyle." But 
absolutely nothing is known of the digestive process in insects beyond 
the fact established by Plateau (1874) that the contents of all parts of 
the alimentary tract are alkaline during digestive activity and either 
neutral or weakly alkaline at other times. Hence, if we make use of 
these words in insect physiology, it must be with the understanding 
that no chemical significance is implied. The ventriculus is very 
commonly called the " chyle stomach " but there is probably no reason 
for calling it a " chyle stomach " any more than a " chyme stomach," 
and likewise there is no reason for supposing that the intestine does 
not contain chyle — in fact, it almost certainly does. The word 
" chyle " may be used with entire propriety in insect physiology to 
signify the completed products of digestion, but to designate a part 
of the alimentary tract as the " chyle stomach " is applying the term 
without an adequate basis of facts. 

The contents of the ventriculus are surrounded by several concen- 
tric layers of thin filmy membrane which form an interior tube ex- 
tending the entire length of the stomach and reaching down into the 


anterior end of the intestine. This tube can be very easily seen by 
carefully cutting open the outer walls of the ventriculus, but it is 
best demonstrated by transverse microtome sections of a specimen 
prepared for histological purposes. Such a section is shown by figure 
46 A. A small amount of solid food matter {qq) is seen in the cen- 
ter of the specimen, surrounding which are numerous irregular con- 
centric rings of membrane {Pmb), some adhering to each other in 
places, others entirely free, most of them structureless, but others 
partly cellular. These are known as the peritropMo membranes 
{Pmb). They keep the solid contents of the stomach away from the 
epithelial walls, from which, as will be presently explained, they are 
given off from time to time. 

The walls of the ventriculus (fig. 46 A) are thick and consist of 
numerous cells {Epth) apparently very irregularly arranged. On 
their inner surfaces is a thin intima [Int) and on their outer surfaces 
a still finer basement membrane (BAI) . Outside of the last are two 
layers of muscles, the external layer consisting of longitudinal fibers 
(LMcl) and the inner of transverse ones {TMcl). Numerous an- 
nular depressions of the walls form internal folds (fig. 45), but any 
part of the ventricular wall can be stretched out into a flat sheet, 
which is then seen to be full of little pits, giving the whole a screenlike 
appearance. Sections show that the pits result from circular invagi- 
nations of the basement membrane (fig. 46 B, BM), and that at the 
bottom of these pockets the cells are very small and convergent, while 
those on their lips are very large. Figure 46 B is a very perfect 
example of this structure of the epithelium, which is usually more 
or less obscured, as in figure 46 A, by a great proliferation of small 
cells from the lips of the cups — and then a large section seldom gives 
a symmetrical view of all the parts. The cups are all filled to over- 
flowing by a gelatinous mass (pp) which fuses over their edges into 
a continuous coating beneath the intima over the entire inner surface 
of the epithelium. This mass appears to be formed mostly by the 
cells at the bottoms of the cups, for the outermost of these (fig. 46 B, 
7t) often insensibly fade into it. 

Figure 46 E shows an opposite condition of the epithelial cells. 
Here the lip cells of the cups appear to be very actively dividing, 
and proliferating a great number of small cells (Em) which float 
off into the gelatinous covering. These discharged cellules are all 
nucleated, but their protoplasm does not stain in preparations and 
consequently they appear clear and transparent as compared with the 
cells they apparently come from. The writer has not been able to find 
any of these cells actually in the process of division, but a comparison 
of figures B and E (which are camera lucida drawings and not dia- 
grams) would certainly suggest that the condition of the cells in E 
has resulted from a very active division of the cells of the walls and 



lips of the cups, which are quiescent in B. Comparing this with 
what is known to take place in other insects during digestion, there 
is every reason for believing that the proliferated cellules are filled 
with the digestive secretion, and that E represents a stage immedi- 


Fig. 46. — Histological details of alimentary canal of worker ; A, cross section of ventriculus 
showing peritrophic me'mbranes (Pmli) ; B, section of wall of ventriculus showing 
epithelial cups with cells in resting condition and covered by gelatinous mass (pp) ; 
C, section of Malpighian tubule ; D, cross section of small intestine ; E, section of 
ventricular epithelium after formation of numerous small digestive or enzyme cells 
(Ens) given off into gelatinous matrix (pp) ; F, section of anterior end of rectum 
through reqjal glands (iSGJ) ; G, part of slightly oblique section through posterior end 
of ventriculus and anterior end of small intestine, showing openings of Malpighian 
tubules (Mai) into the latter. 

ately subsequent to one of greatest secretive activity, in which there 
is a large number of little cells (Em) highly charged with the 
enzyme-containing digestive juices imbedded in a gelatinous matrix 
covering the inner surface of the epithelium. This matrix next 


separates itself from the ends of the remaining epithelial cells, which 
at the same time secrete a new intima over their inner surfaces. The 
lower part of figure 46 A shows this indisputably. ^The whole thing, 
then, finally contracts about the food and, as the digestive cellules 
give up their contents, shrivels and shrinks and becomes a peritrophic 
membrane. In figure A the outermost peritrophic layer is still in 
both conditions — its dorsal part is shrunken to a thin membranous 
form, while its lower part is gelatinous and filled with secretion 
cellules, though it is separated from the epithelium by a new intima 
and is detached at intervals from the latter. Beneath the new intima, 
furthermore, is seen at places the formation of a new gelatinous mass. 
Some of the inner peritrophic layers shown in A also retain remnants 
of dells. 

Figure 46 A is drawn from a specimen which is typical of all in 
several series of sections through the ventriculus. The peritrophic 
layer partly adhering to the epithelium is no artifact, because the 
same condition may often be directly observed in dissections of fresh 
specimens. In the opposite end of the series from which the specimen 
was selected this layer is entirely free from the epithelium. 

The peritrophic membrane has been described in some insects as 
being a prolongation from the intima of the proventriculus, the ven- 
triculus itself being supposed never to secrete an intima. It is per- 
fectly conceivable that the anterior end of the membranes might be 
generated by the outer cellular layer of the proventricular funnel and 
remain attached to it after the rest of it had become free from the 
ventricular wall, and thus give the appearance of belonging to the 
proventriculus. The writer, however, has several sets of longitudinal 
sections through these part^ in the bee, but none of them nor any dis- 
sections made show such a condition. 

Absorption is commonly supposed to take place largely in the ven- 
triculus. If so, the food must pass through the several peritrophic 
membranes and then through the thick epithelium. It is entirely 
possible that it may do so, but the pollen contained in the ventriculus, 
as already stated, shows little or no evidence yet of digestion and does 
not begin to do so until it reaches the small intestine. On the other 
hand, the dark mucilaginous slime of the ventriculus does not appear 
in any quantity in the much drier contents of the small intestine. 
Therefore it may be supposed that this slime contains the sugar ele- 
ments of the food and that the latter are principally digested in, and 
absorbed from, the ventriculus. The absorption of the proteids and 
hydrocarbons must take place in the intestine and rectum since these 
food elements in the bee's diet are derived only from the pollen. 
However, these conclusions are purely tentative, being based on the 
writer's observation of the contents of the different parts of the ali- 
mentary tract, which, while fairly extensive and continued through 


most of a year, are confessedly not nearly adequate to serve as a 
basis for conclusive statements on the digestive process. They are 
sufficient, however, to show the utter lack of a basis in facts for many 
other opinions on this subject. 

Cheshire (1886) describes two kinds of cells in the ventricular 
epithelium, " one secreting a digestive fluid (gastric juice) from the 
surrounding blood into the stomach, so that the pollen grains may be 
made fit for assimilation by a transformation not unlike that lique- 
fying gluten in our own case; the other absorbing the nutrition as 
prepared and giving it up to the blood." Though Cheshire refers 
to his figures to show these two kinds of cells, he does not point out 
which are which — in fact, he does not even designate two different 
kinds in his drawings nor even represent two kinds. 

The small intestine (fig. 42, SInt) forms a loop upon itself and con- 
stitutes a narrow tube connecting the stomach {Vent) with the large 
intestine or rectum {Rect)= Its anterior end is somewhat enlarged 
and carries the circle of malpighian tubules {Mai). Its epithelium 
(fig. 46 D, Epth) is very simple and is thrown into six longitudinal 
folds that project into its lumen. On the outside is a thick sheath 
of transverse muscle fibers {TMcl) with distinct nuclei {Nu). The 
latter are designated by Cheshire (1886) as "longitudinal muscles" 
(see his figure 14 D), but this is a very evident mistake — ^the small 
intestine has no longitudinal muscles at all. It is evident that the 
folds of the epithelium permit the ordinarily narrow tube to expand 
very considerably when necessary to allow the passage of a large 
amount of food. The contents of the small intestine are usually 
drier than those of the ventriculus, consisting principally of masses 
of partly digested pollen, that is to say, the contents of the grains are 
partly dissolved out — presumably signifying that they are under- 
going digestion. There is usually only a small amount of the brown 
slime present such as fills the ventriculus. 

The Malpighian tubules (fig. 42, Mai) are wrapped and coiled about 
one another and about the viscera of the abdominal cavity. There 
are about 100 of them in the honey bee and they all open separately 
into the anterior end of the intestine. Each is a very long thread- 
like tube consisting of a single layer of epithelial cells provided with. 
a very delicate basement membrane and intima (fig. 46 C) . The ends 
of many of the cells are clear and bulge into the lumen. Figure 
46 G shows a section through the junction of the ventriculus and the 
intestine where the tubules open by narrow necks penetrating the 
epithelium. The wall of the ventriculus forms a short double-layered 
fold {VentVlv) projecting backward into the anterior end of the 
intestine, behind which are the orifices of the Malpighian tubules. 
The section from which figure G was drawn is cut somewhat obliquely 
and takes in this fold only on one side. 


The Malpighian tubules are regarded as excretory in function and 
are supposed to remove from the 'blood the nitrogenous waste prod- 
ucts resulting from metabolism. Minute crystals of urates are often 
to be found in them and they probably perform the work of the 
kidneys in vertebrate animals. 

The large intestine (fig. 42, Red), called the rectum in insects, is 
an enormous sac which may lie limp and flabby in the rear part 
of the body or it may be so immensely distended by the amount of 
its solid and liquid contents as to occupy a large part of the abdomi- 
nal cavity. The recognizable elements of the material within it 
consist mostly of the empty shells of pollen grains or of grains hav- 
ing their contents greatly shrunken and distorted — presumably as 
a result of the absorption of the protoplasm, although a considerable 
number are usually present which are only slightly digested, while 
there are always to be observed a few perfect and fresh-looking 
grains showing no evidence at all of digestion. The rest of the in- 
definite mass of solid rectal material consists of some unrecognizable, 
finely triturated substance, probably derived in part from fragments 
of the peritrophic membranes. There are always present a few bits 
of feathered bee hairs. 

The epithelium of the rectum is, like that of the CESophagus, rudi- 
mentary, being distinguishable only by the nuclei (fig. 46 F, Nu) 
remaining in the outer layer of the thick transparent intima {Int). 
Outside of this is an external layer of longitudinal muscle fibers 
{LMcl) and an inner layer of transverse fibers {TMcT). The intima 
(/wi) is thrown into numerous folds whose edges qonverge, forming 
pocketlike grooves between them in which are lodged small masses 
of the rectal contents. This is very suggestive that absorption takes 
place in this part of the alimentary tract, although it is not com- 
monly supposed to do so, but if the pollen is not fully digested until 
it reaches the rectum, how can it be absorbed by the anterior part 
of the alimentary canal? 

The so-called rectal glands (fig. 42, RGT) consist of six hollow 
epithelial tubes (fig. 46 F, RGl) and are the only parts of the rectal 
epithelium in which the cells are well developed. The cells on the 
outside of each " gland " are small, but the inner ones are very large 
and are covered by a thick layer of dark chitin {Int). The lumen 
is intercellular and does not communicate with that of the rectum. 
"WTien the rectum is distended the " glands " bulge out on the surface 
as six short opaque ridges (fig. 42, RGl) , but when it is empty they 
sink into the walls as in figure 46 F. Nothing is known of the 
function of these organs, and their glandular nature is entirely con- 
jectural. If they are glands, it is not clear why the intima should 
be so especially dense on their inner faces. 



The liquid medium that distributes the digested food from the 
alimentary canal to the cells of the body tissues is called the hlood, 
and the contractile organ that keeps the blood in motion is the heart. 
In vertebrate animals the blood is contained entirely within tubes 
called arteries and veins, but in insects and most other invertebrate 
animals the blood simply fills the empty spaces between the viscera 
of the body-cavity, which spaces may, however, constitute definite 
channels or sinuses, and may even be shut in by special membranes. 
Besides carrying and distributing the digested food that is absorbed 
into it in solution, the blood of animals generally has also to dis- 
tribute oxygen to the tissue cells and carry off their waste products. 
Oxygen is obtained from the air and, like any other gas, is soluble 
in liquids. Hence it is present in the blood not in the form of small 
bubbles of gas but in solution, just as it is in all water exposed to the 
air. The respiratory system (see page 116) is simply a special con- 
trivance for bringing air into close proximity to the blood so that 
its gases may diffuse into the latter, but many soft-bodied animals 
like earthworms absorb air directly through the skin. Vertebrate 
animals have a substance in their blood called hemoglobin which is 
contained in the red corpuscles and has a great capacity for absorb- 
ing oxygen. It, therefore, enables the blood to carry much more of 
this gas than could be dissolved simply in its plasma. Invertebrate 
animals do not need so much oxygen as vertebrates, and, therefore, 
most of them can get along with that which dissolves in the color- 
less blood plasma without the special aid of hemoglobin. Most 
insects, however, being excessively active creatures, must have a 
rapid metabolism in their cell tissues, and consequently they need 
much oxygen to consume the product of this metabolism, but they 
belong to the class of animals without red blood and, hence, nature 
has provided them with another means of obtaining a special supply 
of air, namely, a set of air-tubes branching minutely over nearly all 
the internal organs, the tissues, and even most of the cells in the 
body. (See " The Respiratory System," page 112, for discussion of 
oxidation and removal of waste products.) 

The blood of insects is usually a colorless liquid containing opaque 
granular cells or corpuscles floating in it. There are no special blood 
vessels, but there are very definite channels between the muscles and 
viscera through which the blood flows, while conspicuous membranes 
stretched across the dorsal and ventral walls of the abdomen (fig. 1, 
DDph and VDph) inclose special dorsal and ventral sinuses which 
play an important part in the circulation. These membranes, called 
diaphragms, are rhythmically contractile, and contribute much to 



maintaining the circulation of the blood. " The heart (fig. 1, Ht) is 
located in the dorsal sinus, which latter is therefore often called the 
pericardial chamber. The pulsations of the diaphragms are produced 
by fine muscle fibers lying in their walls. These are usually ar- 
ranged in a number of fan-shaped bunches on each side radiating 
from the edges of the diaphragm (fig. 47, DphMcl) toward the mid- 
dle, where most of them are continuous with the fibers from the oppo- 
site side. It used to be supposed that those of the dorsal diaphragm 
produced the expansion of the heart, and they were for this reason 
called the " wing muscles of the heart," but the latter organ is now 
known to be a muscular tube and to contract and expand by its own 

FiQ. 47. — Dorsal diaphragm of drone, from one segment and adjoining parts of two 
neighboring segments, showing median heart {Ht) as seen through transparent dia- 
phragm (DDph), fan-shaped bunches of diaphragm muscles (DphMcl), and lateral 
tracheal sac (TraSc) giving off sac-bearing trunks into pericardial chamber above 

power. In some insects the muscles of the dorsal diaphragm form a 
meshwork of fine fibers surrounding numerous large and small holes 
in the membrane, which probably permit the entrance of blood into 
the sinus above, but in most species the diaphragm is imperforate 
and the blood enters the pericardial chamber above its scalloped edges 
(figs. 1 and 47) . 

The heart of insects in general is a long narrow tube (fig. 1, Ht) 
situated in the dorsal sinus or pericardial chamber of the abdomen 
along the midline of the body. It is swollen toward the middle of 
each segment into a heart chamber (ht) which presents a vertical 
slitlike opening or ostium (Ost) on each side. Theoretically^ in 


generalized insects, there should be a chamber to each segment, but 
the heart is variously shortened from both ends so that the chambers 
are always fewer than the segments. The posterior end of the heart 
is closed, but its anterior end is produced into a long narrow tube 
called the aorta (fig. 1, Ao) which extends through the thorax and 
opens by a few simple branches into the cavity of the head. 
. The heart of the bee (fig. 1, Ht) consists of only four chambers 
(Iht-liM) lying in the third, fourth, fifth, and sixth segments of the 
abdomen. In the front of this part of the body it bends downward 
and forms a large convoluted loop (i) of about 18 folds where it 
passes through the abdominal constriction. All of this convoluted 
part really belongs to the abdomen, since it lies in the propodeal part 
of the apparent thorax, which is the true first abdominal segment. 
The aorta {Ao) extends forward from here as a very fine tube making 
a large arch between the muscles of the thorax and then enters the 
back of the head. According to Pissarew (1898) the convolutions 
of the anterior end of the heart are peculiar to the honey bee, .being 
absent in its nearest relatives such as Bomhus and MegacMlei. The 
heart walls, as before stated, are muscular and produce a rhythmical 
contraction of the tube whose pulsations follow each other from be- 
hind forward. Thus the contained blood is driven out of the anterior 
end of the aorta into the head, where it bathes the brain and the other 
organs of this region, and then flows backward, percolating through 
the spaces between, the organs of the thorax. 

From the thorax it enters the cavity of the ventral sinus — not the 
general abdominal cavity, at least in the bee — and is pumped back- 
ward by the pulsations of the ventral diaphragm and dorsally over 
the inner walls of the thorax and through definite channels about all 
the viscera, finally collecting in the dorsal sinus where it again enters 
the heart through the lateral ostia. The lips of the ostia are pro- 
vided with small membranous lobes which project inward and con- 
stitute valves that prevent the expulsion of the blood. A similar 
valve is placed at the anterior end of each chamber of the heart to 
prevent a possible backward flow. 

In the bee, both the dorsal and the ventral diaphragms are well 
developed, the former (fig. 1, DDph) extending from the third 
abdominal segment to the seventh, inclusive, while the latter iVDph) 
extends from the abdominal constriction to the eighth segment. 
The ventral diaphragm is much more muscular than the dorsal and 
its pulsations, which are very strong, follow each other from before 
backward. They may easily be observed by removing the top of 
the abdomen from an asphyxiated bee. The ventral sinus is very 
ample, inclosing the nerve cord of the abdomen, and receives into 
its anterior end the blood channels of the thorax, so that the latter 



communicate with the general cavity of the abdomen only through 
the ventral sinus. 

The dorsal diaphragm (fig. 1, DDph) ends by a free transverse 
edge near the front of the third abdominal segment. A part of it is 
shown by figure 47 extending across one segment and the adjoining 
parts of two others. The fan-shaped bunches of muscle fibers 
{DphMcl) are seen radiating from the anterior edges of the terga 
toward the midline, where they are mostly continuous with those 
from the opposite side, only a few of the anterior and posterior ones 
ending free in the membrane of the diaphragm. The latter is imper- 
forate, but its edges are deeply scalloped between the points where 
the muscles are attached, allowing free entrance to the blood from 
the intervisceral channels of the abdomen. The dorsal surface of 
the diaphragm is covered by a network of cells (figs. 47 and 48, 
DphCls) arranged in flat branching and fusing bands. These cells 


Fig. 48. — Small part of dorsal diaphragm of drone, showing bands of flat diaphragm cells 
(DphCls), the diaphragm membrane itself (DpftmS), and the muscle fibers (DphMcl). 

may be called the diaphragm, cells to distinguish them from the 
pericardial cells to be described later. 

The abdominal circulation is very easy to observe in a living 
bee. The best way to demonstrate it is to pin an asphyxiated bee 
to a block of cork or paraiEn and remove the top of the abdomen 
by making an incision with a small pair of scissors clear around it. 
Gently pull the alimentary canal to one side so as to expose the 
ventral diaphragm, which will be observed pulsating strongly back- 
ward. Next cut a small hole in the top of the thorax and insert 
into it a drop of some stain in a water solution (the writer used 
carmalum). Almost immediately this will appear in the ventral 
sinus of the abdomen, in which it is pumped backward by the dia- 
phragm, and from which it goes upward through invisible channels 
between the air sacs and the alimentary canal and especially up 



wide channels against the lateral walls of each segment. It does not 
run out free into the abdominal cavity, however, unless through a 
rent, nor does it enter the latter from the thorax except by way of 
the ventral sinus. The dorsal circulation of course can not be 
observed in this specimen because the back is removed. Therefore, 
take another bee and fasten it in the same manner, but make simply 
a shallow median slit along the back, thus exposing the dorsal sinus 
and the heart from above. Now insert a drop of stain into the 
thorax as before. After about two minutes this will appear in jper- 
ceptible amount in the dorsal sinus, very much diluted, to be sure, 
with the blood, but there will be enough to give white blotting paper 
a distinct red tint. In a short time the heart becomes filled with 
the stained blood and appears as a red tube along the median line. 

The dorsal sinus contains not only the heart but also two pairs of 
pericardial air sacs in each segment. These are seen entering the 

Fig. 49. — Pericardial chamber of one segment in worljer, seen from below looking through 
transparent dorsal diaphragm {DDph), showing median heart (Ht), lateral pericardial 
air sacs (BtTraSc) given off from large lateral sacs (TraSc), and the padding of 
pericardial cells (HtCla) against inner surface of tergum (T) . 

sinus from the large lateral air sacs of the abdomen (TraSc) in 
figure 1 and also in figure 47. In the latter the heart (Ht) is seen 
along the median line as it shows through the diaphragm. Figure 
49 gives a view of the pericardial sinus as seen from below, in one 
segment, by focusing through the transparent diaphragm (DDph). 
In the middle lies a chamber of the heart (Ht) with the slitlike 
ostium on each side. Laterally are the four pericardial air sacs 
(HtTraSc) giving off branches that ramify profusely upon the 
heart. Above the heart and the air sacs is a thick bed of large 
granular cells {HtCls) which make a soft padding between the hard 
tergal wall and the delicate organs of the sinus. These are called 
the pericardial cells. They may have some physiological function, 
as has often been supposed, but if so no one has decided what it is. 



The lives of all animals depend upon a constant distribution of 
free oxygen gas throughout their bodies. This oxygen, continually 
inhaled and exhaled, is not used in the formation of tissues, it does 
not become a part of the living protoplasm of the animal — it is the 
physiological scavenger that eats up certain waste products of me- 
tabolism which are deadly to the system unless constantly removed or 
changed into less harmful compounds. The action of oxygen upon 
these waste substances within the body is comparable with ordinary 
combustion in that it results in the formation of carbon dioxid gas 
and water and in the generation of heat. Since the air, which is com- 
posed of both oxygen and nitrogen, is the source of the oxygen supply, 
the ordinary breathing processes involve an inhalation also of nitro- 
gen gas, and the tissues become permeated with it as well as with 
oxygen. The nitrogen of the air, however, is not known to serve any 
physiological purpose in the body, its presence being simply unavoid- 
ably incidental to the inhalation of oxygen. While oxygen and nitro- 
gen are two most important food elements, the tissues -of animals can 
not make use of either in the gaseous condition, but must be supplied 
with substances containing these elements in combination with others 
in the form of solid and liquid food stuflfs taken into the alimentary 
canal. Hence, air is not a food, and the respiratory system is to be 
regarded as chiefly excretory in function. 

The means by which different animals receive oxygen into 
their systems are various. All aquatic breathers of course use 
that which is naturally dissolved in water. Many of the lower ani- 
mals absorb air directly through their skins and into their tissues, 
while the carbon dioxid escapes the same way. Others that live in 
the water and whose bodies are covered by an impervious skin or 
shell have thin-walled, hollow, branching appendages, called gills, 
through which the blood circulates freely and through whose walls 
the necessary exchange of gases takes place. Land animals very 
commonly have some sort of an invagination from the exterior which 
allows the air to enter thin-walled tubes or cavities and be absorbed 
into the blood. Land vertebrates have a tube opening from the back 
of the mouth whose inner end branches profusely and forms a pair of 
organs called the lungs, through which the blood circulates freely in 
delicate tubes that allow the transfer of gases. Insects, finally, have 
a system of internal air tubes, called trachece, opening to the exterior 
by a number of small orifices, called spiracles, situated along the sides 
of the thorax and abdomen, which give off branches that ramify 
minutely to all parts of the organism, thus virtually making a lung 
of the entire body. The tracheae are thin tubes made of flat epithelial 





Fig. 50. — Tracheal system of worker as seen from above, one anterior pair of abdominal 
sacs (fig. 1, 9) removed and transverse ventral commissures of abdomen not shown. 

221S1— No. 18—10 8 


cells lined with a delicate layer of chitin. The latter, however, is 
strengthened by circular thickenings which give the appearance of 
an internal spiral thread, but a closer examination shows that each 
thickening makes only a few turns and that several lie in parallel 
bands. This structure is for the purpose of maintaining an open 
passageway for the air through the very thin-walled tubes. The 
tracheae branch into fine capillaries and these terminate in excessively 
delicate end-tubes. In some cases it is easy to see that a great 
number of capillary branches surround the cells of a tissue, if 
they do not actually enter the cell walls, but in others it can not be 
shown that the tracheae really penetrate below the surface of a mass 
of cells. 

Gases in solution, like solids, pass freely back and forth through 
moist animal membranes, going in the direction of the least pressure 
of each particular gas. By this simple method the gases go back and 
forth through the walls of the gills, lungs, or air tubes and permeate 
the tissues themselves. Vertebrate animals, as already explained, 
have a red substance in the blood called hemoglobin which has a 
very great oxygen-absorbing power and which greatly increases the 
oxygen-carrying power of the blood, but still a certain amount of 
oxygen is carried in solution by the liquid or plasma of the blood. 
Now, tl\,e blood of insects has none of this hemoglobin and all the 
oxygen it can carry is that which dissolves in its plasma, but, on 
account of the extensive ramification of the air tubes, it is not neces- 
sary for the blood to distribute the oxygen to the organs. It is usually 
stated that the blood in insects does not carry oxygen at all, ^cept 
for its own use, but it would seem physically impossible that the gases 
should not diffuse out of the fine terminal air-tubes into the blood 
when they do so in all other cases. If the blood of a crab or crayfish 
is capable of carrying enough oxygen in solution to supply the wants 
of the body, there is no reason why that of an insect, which has much 
better facilities for obtaining air, should not do the same. Further- 
more, we can not suppose that the products of katabolism have to 
accumulate about the end tracheae in order to be oxidized. They are 
produced wherever metabolism is going on, which is everywhere in 
the living cell substance, and, hence, the latter must be permeated 
with oxygen in solution, which must also be in the blood along with 
the carbon dioxid formed. The carbon dioxid diffuses back into the 
end tracheae from the blood. Therefore, while the great extent of the 
tracheal system in insects relieves the blood of the work of distribut- 
ing the oxygen, the blood must nevertheless serve as an intermediary 
medium for both the oxygen and the carbon dioxid between the fine 
terminal tracheal branches and the cells. 

It has sometimes been suggested that certain large cells called 
cenocytes, found especially in connection with the tracheal system, 


function as intermediaries between the trachea and the cells, but 
Koschevnikov (1900) has shown that these cells appear to be tem- 
porary storehouses for waste products from the tissues— presumably 
uric acid compounds which have been already oxidized. Even the 
fat-body has been regarded as a sort of lung in which oxidation takes 
place, but there is no evidence to support this theory, although, for 
that matter, there is little evidence in favor of any theory in insect 

The process of metabolism, or the vital activity of the cells them- 
selves, results in a breaking down of the complex and highly unstable 
protoplasmic molecules into chemical substances of much simpler con- 
struction, and it is these by-products of metabolism that are attacked 
by the oxygen in the blood furnished by the respiratory system. Pro- 
toplasm consists principally of the elements carbon, oxygen, hydro- 
gen, and nitrogen, and the oxidation process results, as before stated, 
in the formation of carbon dioxid (COj) and water (H2O), while 
the residuary products are mostly organic comjiounds of nitrogen 
related to uric acid (C5H4N1O3) and urea (CON^Hj). The carbon 
dioxid is a soluble gas which diffuses into the end tubes of the trachea; 
and is exhaled. A part of the water at least is given off with the 
" breath " in the form of water vapor, for drops of it can be collected 
by inclosing bees or any insects in a tube for a short time. The nitro- 
gen compounds and probably a part of the water are dissolved in the 
blood and removed by the Malpighian tubules, which are the kidneys 
of insects. 

Besides this oxidation of waste products, which allows the process 
of metabolism to go on unhindered, the inhaled oxygen serves also 
another purpose, namely, that of maintaining the body heat. Al- 
though insects are usually classed as " cold-blooded " animals, they 
nevertheless maintain a temperature which is always higher than 
that of the surrounding air and is often a number of degrees above 
it. It is well known that the temperature of a beehive during the 
brood-rearing season is almost as high as that of the human body, 
and that even during winter it remains at nearly 80° F. ; but this is, 
of course, due to the accumulation and condensation of the warmth 
from the bodies of a great many bees, and is much higher than the 
temperature of any bee outside of the hive. In our own bodies 
certain substances are consumed by oxidation in the blood simply 
to produce the necessary heat energy for maintaining metabolism, 
and hence it seems reasonable to suppose that the same thing takes 
place in insects, although of course to a much less degree. 

There are generally ten pairs of spiracles or breathing apertures 
in insects, two being situated on the sides of the thorax between 
the segments, but probably belonging to the mesothorax and the 


metathorax (although the first is often regarded as prothoracic) , 
while the other eight are situated on the sides of the first eight 
abdominal segments — in the bee on the lateral parts of the terga 
(figs. 32 and 33, 8p). The breathing apertures are usually pro- 
vided with a closing apparatus of some sort consisting of the swollen 
lips of slitlike spiracles, of a small lid, or of a flexible and collapsible 
chitinous ring, each with special occlusor muscles attached.. In the 
bee a chitinous band surrounds the tracheal tube opening at each 
spiracle, a short distance from the aperture, and has two opposite 
loops projecting on the same side, connected by a muscle whose con- 
traction approximates the two halves of the band so as to close the 
lumen of the trachea." It is supposed that after an inhalation the 
spiracles are closed momentarily, so that the first force of the ex- 
piratory contraction of the abdomen is exerted against the air shut 
in the tracheae, with the result of driving it into the extreme tips 
of the latter — the spiracles then opening, the rest of the contraction 
is expended in exhalation. 

The internal tracheal system consists, among insects generally, 
of a large tracheal trunk lying along each side of the body, connected 
by short tubes with the spiracles and by transverse commissures with 
each other, while -they give off segmental branches into the body 
cavity which ramify minutely upon the organs and tissues. In the 
thorax specially large tubes are given off on each side to the legs 
and to the bases of the wings, while in the head others go to the 
eyes, antennae, and mouth parts. The whole body is thus virtually 
a lung with ten pairs of openings along the sides. 

The tracheal system of the bee (figs. 1, 50, and 51) is best developed 
in the abdomen, where the longitudinal trunks are enlarged into two 
enormous lateral air sacs (TraSc), which are of greatest diameter in 
the anterior end of the abdomen. They are segmentally connected by 
large transverse ventral commissures (fig. 51, TraCom), most of which 
are themselves distended into small air sacs. Dorsally the lateral 
sacs give off in each segment a large tube which divides into two sac- 
culated branches (figs. 49 and 50, UtTraSc) that enter the pericardial 
chamber and supply the heart and pericardial cells with tracheae. In 
the thorax a large sac lies on each side of the propodeum (figs. 1 and 
50, 7), which bears a short tube opening to the first abdominal 
spiracle (figs. 21 and 50, ISp). Above these sacs is a narrow trans- 
verse median one (figs. 1 and 50, 8) occupying the large cavity of 
the turgid mesoscutellum (fig. 21, Scl^). In the ventral part of the 
thorax there is a large median posterior sac (figs. 1, 50, and 51, 5) 

"For a detailed description of ttie spiracles in the bee and their occlusor 
apparatus see Djathchenko (1906). 





51. — Tracheal system of worker showing lateral and ventral parts as seen from 
above, with dorsal sacs and trunks removed in both thorax and abdomen. 


which gives off trunks to the middle and hind legs and a large sac 
on each side (fig. 51, 6 and 6) to the ventro-lateral walls of the thorax. 
Two large strong tubes (figs. 1, 50, and 51, Tra) — the only tracheae 
in the bee's body well developed as tubes — extend backward from the 
head through the neck and prothorax to the first thoracic spiracles 
(figs. 50 and 51, ISp) . Each of these gives off a branch which divides 
into the trachea for the first leg and into another that connects with 
the posterior ventral thoracic sac (5). An anterior median thoracic 
sac {Ji) is connected with the two large anterior tubes near where these 
enter the neck. In the head are a number of large sacs which are 
situated above the brain (see figs. 1, 50, and 51, 1), about the bases 
of the eyes and optic lobes (see figs. 1 and 50, ^), and above the bases 
of the mandibles j(see fig. 1, 3). 

Nearly all of the tracheae in the bee's body are excessively delicate 
and their walls mostly lack the spiral thickening that ordinarily holds 
a tracheal tube open. They are consequently very distensible and, 
when inflated, they show as opaque glistening white vessels, which, 
however, when empty, are extremely difficult or actually impossible to 
see. The smaller branches are so numerous and flabby in the thorax 
and the legs (fig. 1, LTra) that they appear to form everywhere 
meshworks or sheets of tiny glistening air-cavities imbedded between 
the muscle fibers. Only the large trunks in the anterior part of the 
thorax have the normal tracheal appearance. 

The body of the bee is thus most abundantly aerated, probably 
more so than that of any other insect. The numerous large and 
small sacs form great storehouses of air — tanks containing reserve 
supplies of oxygen. They are not present for the purpose of lighten- 
ing the weight., of the body, because inflation with air does not 
decrease the weight of any object surrounded by air. 

The respiratory movements are limited to the abdomen in the bee 
on account of the solidity of the thorax. They vary a great deal 
according to the activity of the individual. While sitting quietly 
at the entrance of the hive or walking slowly about, bees usually 
exhibit almost no respiratory motion, only a very slight vibratory 
trembling of the abdomen being noticeable. Others that are walk- 
ing hurriedly about lengthen and shorten the abdomen very percepti- 
bly, the motion being specially pronounced at the tip. A bee that 
has just alighted after flying exhibits still more pronounced abdomi- 
nal movements, not only a contraction and expansion but an up- 
and-down motion as well. When a bee is becoming asphyxiated in 
a killing bottle the extension and contraction of the abdomen is most 
pronounced, although much slower than in the ordinary breathing 

The muscles of the abdomen that produce respiration have been 
described by Carlet (1884), who distinguishes seven different sets of 


them as follows: There are two dorsal sets: (1) The internal dorsal^ 
going from the anterior edge of one tergum to the anterior edge of 
the next following tergum, and (2) the external dorsal, going from 
the lateral edge of one tergum to the corresponding edge of the fol- 
lowing tergum. Both of these are expiratory, since their contrac- 
tions bring the two segments together. On the sides are three sets: 
(3) The external oblique, going from the anterior edge of each tergum 
to the side of the corresponding sternum; (4) the internal oblique, 
crossing under the last from the anterior edge of each tergum to the 
side of the preceding sternum. These two sets are likewise expira- 
tory, because their contractions approximate the terga and sterna. 
The third set of lateral muscles is (5) the transverse, lying between 
the overlapping surfaces of each tergum and its corresponding 
sternum and being, therefore, inspiratory, because the contraction 
separates the terga and sterna. Finally, there are two sets of ventral 
muscles: (6) The external ventrals and the internal ventrals, forming 
a letter M between the anterior edge of each sternum and the one 
following, and (7) the interventral, situated between the overlap- 
ping surfaces of consecutive sterna and causing their separation by 
contraction. These last are therefore also inspiratory. 

It would thus seem that the abdomen is much better equipped with 
expiratory than with inspiratory muscles. Perhaps the expansion 
is partly due to elasticity, and perhaps, also, it is true that the abdo- 
men contracts upon the full tracheae and air sacs, before the spiracles 
open to allow exhalation, in order to drive the air into the farthest 
recesses and terminal tubes of the tracheal system, which necessitates 
an extra contractive force. 


The fat tissue of insects is not miscellaneously distributed through 
the tissues, imbedded beneath the skin and packed between the 
muscles, but is disposed in sheets and strands within the body 
cavity, especially in the abdomen, or forms a definite mass, the 
fat body. The fat cells are large and extensively vacuoled with 
clear globules of fatty oils. In some insects the fat bodies have a 
brilliant yellow, golden, or orange color. Associated with the fat 
cells are other much larger and often gigantic cells, called wnocytes, 
attaining the largest size of all the cells in the body except the eggs. 
They were first discovered in segmental clusters attached to the 
tracheae near the spiracles, but they are now known also to be scat- 
tered through the depths of the body cavity, where they occur im- 
bedded especially between the fat cells. The term " cenocyte " signi- 


fies merely that those cells first observed by Wielowiejski, who gave 
them this name, were slightly wine-colored. 

Both the fat cells and the cenocytes of the honey bee have been 
specially studied by Koschevnikov (1900), who gives the history of 
the fat body as follows : In the larva it consists of gigantic lobes, the 
cells of which are in general all alike and so closely packed in 30 
or more layers that, in the younger stages, most of them assume 
angular forms. Many of them are binucleate, and the protoplasm 
is strongly vacuolated except for a small area about the nuclei. In 
the full-grown larvse the fat cells become globular and filled with a 
number of round granules, which, during the early part of the pupal 
stage, are set free by a dissolution of the cell walls and float free in 
the body cavity. In pupae a little older, having even but a very 
thin chitinous covering, the adult fat body is fully formed, and yet 
neither the disappearance of the larval granules and nuclei nor the 
formation of new adult fat cells is to be observed. It seems that 
the granules of the larval fat cells, set free at the beginning of 
histolysis, are reassembled about the nuclei to form the fat cells of 
the adult. In the very young imago the cells of the fat body are 
very distinct, and each possesses a considerable amount of protoplasm, 
with enormous vacuoles which press upon all sides of the nucleus. 
In old bees the vacuolation is much reduced and may even be entirely 
lacking, while the cells become filled with a solid granular plasma. 
Old workers examined in the fall show the fat cells united into 
syncytia or masses in which the cell boundaries are lost, although 
the nuclei remain distinct. A queen does not appear to form these 
syncytia in old age. 

The function of the fat body is still unsettled, but we do not 
know of any reason why it should not be comparable physiologically 
with the fat of vertebrate animals and constitute a reserve supply 
of materials which can be used both as food and as a source of heat 
oxidation. It has already been stated (p. 115) that insects main- 
tain several degrees of body temperature. Some entomologists have 
supposed that the fat body gives rise to the corpuscles of the blood, 
others have believed it to be an excretory organ because concretions 
of -uric-acid salts are often found associated with its cells, while 
still others have regarded it as the seat of the combustion of waste 
products by the tracheal oxygen. 

The cenocytes of the bee are described by Koschevnikov (1900) as 
enormous cells imbedded in the fat bodies. He says that those of 
the larva persist into the pupal stage where they undergo dissolu- 
tion and disappear, while new imaginal cenocytes are formed from 
proliferations of the ectodermal epithelium. The new ones are at 


first small and increase about five times in diameter before reaching 
their adult proportions. The fat cells and the cenocytes, although 
closely associated with each other, are easily diKtinguishable by their 
size and by their reaction in life to staining solutions. Koschevnikov 
fed some bees honey or sugar siiiip containing sesquichlorate of iron 
and then, after a few hours, removed the fat body, washed it in ferro- 
cyanide of potassium, and placed it in alcohol acidulated with hydro- 
chloric acid. He found a precipitate of Berlin blue in the fat cells 
while the cenocytes remained perfectly colorless. Thus he showed 
conclusively that the two classes of cells are physiologically different 
in life, although, he says, if a piece of dissected fat body be placed 
in the staining solution the color diffuses alike throughout all the 

The cenocytes have a golden brown pigmentation but no differen- 
tiated contents in young workers and queens. In old workers, to- 
ward the end of the summer, yellow granules begin to appear in 
them. During winter and especially in early spring the cenocytes 
of the workers contain a great number of these granules, but they are 
present in greatest quantity in queens several years old, while in the 
latter the fat cells also contain similar granules. Although 
Koschevnikov admits that the chemistry of these granules is entirely 
unknown, he thinks that they are undoubtedly excretory substances, 
that the waste products of metabolism are first taken up by the ceno; 
cytes and then delivered to the blood, and that the accumulation of 
the granules in the cells during old age means the loss of power 
to discharge them, which brings on the decline in the life activity of 
the bee. If this is so, then the cenocytes are, as he states, excretory 
organs without ducts — cells which serve as depositories for waste 

According to this theory of Koschevnikov, the cenocytes might be 
likened in function to the liver of vertebrate animals, which, accord- 
ing to the present views of physiologists, is the seat of the splitting 
up of the immediate nitrogenous products of katabolism, discharged 
into the blood from the tissues, into those final compounds of nitro- 
gen excreted by the kidneys., 

Wheeler'' also describes the fat cells and cenocytes of insects 
as perfectly distinct in their origins, the fat cells arising from the 
mesoderm, which is the embryonic cell layer between that which 
forms the outer body wall and that which forms the embryonic ali- 
mentary canal, while the cenocytes are derived from internal pro- 
liferations of ectodermal cells. 

"Concerning the Blood Tissue of Insects. Psyelie, VI, 1892, pp. 216-220, 
233-236, 253-258, PI. VII. 



We have learned so far that the bee is a complex animal made up 
of a large number of tissues and organs all definitely interrelated, and 
we speak of these tissues and organs as performing their own special 
functions. Yet, in itself, a mass of cells, even though a living mass, 
is incapable of doing anything — it is inert unless stimulated into ac- 
tion. The legs would not move, the heart would not beat, the glands 
would not secrete, the respiratory movements would not be produced, 
and the animal would cease to live were it not for a vital force that 
incites them all into activity. This force is generated by certain 
cells of the nervous system and is sent out to the other organs along 
the nerve cords, but we know nothing more about it than simply that 
it exists in living animals and is dependent upon the maintenance of 
the nerve cells. 

Now, in order that an animal may be " alive," it is not only neces- 
sary that the muscles should be made to contract, the glands to secrete, 
and all the other organs induced to perform their individual roles, 
but it is equally important that they should all work together and 
accomplish definite results. The muscles must perform harmonious 
movements to produce walking, flying, breathing, or swallowing, the 
heart must beat in proper rhythm, the glands must secrete their juices 
at the right time and in needed amounts. Hence, the functions both 
of stimulation and coordination devolve upon the nervous system. 
The nerve-cells generate a force which, delivered through the nerve- 
fibers to the various organs, irritates the tissues into activity, but, at 
the same time, the cells send out this force in such a methodical man- 
ner that the activities produced in the different organs are definitely 
correlated and cooperate in maintaining the necessary condition for 
the life of the cells. 

The nervous system, however, is more than simply the source of 
these physical and chemical processes that constitute the visible phe- 
nomena of life, for it is also the seat of all sense perceptions and, in 
the higher animals, of consciousness. We do not know, however, that 
insects possess consciousness — ^that they are actually aware of their 
own existence, and we therefore do not know that they have conscious 
sense perceptions. We do know that they are affected by external 
objects — ^by light, heat, taste, odor, pressure, and perhaps sound 
acting upon specially sensitized cells of the ectoderm called sense 
organs, but we do not know that the reaction of the individual is 
anything more than the exhibition of a very highly developed re- 
flex nervous system. It is most probable that bees do all that they 
do — make the comb, store up honey and pollen, feed the young, attend 
to the wants of the queen, and so on — ^without knowing why, and we 
have no evidence that they are even conscious of the fact that they do 



FlQ. 52. — Nervous system of worker, dorsal view. 


these things. Some authors have tried to prove that insects reason, 
but the burden of joroof is still with them. We can admit that in- 
sects iiiaij he possessed of very slight conscious intelligence, but we 
can not admit that any one has ever proved it. Of course, a great 
deal of very interesting insect literature owes its readableness to the 
fact that the author endows his subjects with human emotions and 
some intelligence, or makes it appear that they consciously do things 
from a blind sense of obligation. The bee of literature is often quite 
a different creature from the bee of science. 

If, then, we are forced to admit that we have no proof of intelli- 
gence or of conscious sensations in insects, we have, on the other 
hand, all the more evidence of a very high degree of nervous coordi- 
nation. The body of a bee can be very greatly mutilated and the 
creature will still remain " alive " as long as the nervous system is 
left intact. The segments can be cut apart and each will yet be able 
to move its appendages as long as its nerve center is not destroyed. 
This shows that there are numerous vigorous centers of nervous 
stimulation, but proper coordination results only when all the parts 
are together and intact. 

The nervous system of insects (figs. 1 and 52) is comparatively 
simple, consisting of a series of small nerve masses called ganglia 
{Gng) lying along the mid- ventral line of the body, each two con- 
secutive ganglia being connected by a pair of cords called the com- 
missures." The ganglia contain the nerve cells, which are the source 
of the stimuli sent out to the other tissues, while they also receive the 
stimuli from the ectodermal sense organs. Thus there are incoming 
or afferent stimuli and outgoing or efferent stimuli. The commis- 
sures and the nerve-trunks that branch to all parts of the body con- 
sist of fibers which are fine prolongations of the nerve cells. These 
fibers are the electric wires that convey the stimuli to and from the 
nerve centers and are of two kinds, afferent and efferent, according 
to the direction of the stimulus each transports. 

In a generalized embryo we should theoretically find a nerve gan- 
glion develoj)ed from the ventral wall of each segment, making seven 
head ganglia, three thoracic, and at least ten abdominal ones. In 
the adult, however, many of these fuse with one another. In the 
head, for example, in place of seven ganglia there are only two, one 
situated above the oesophagus, called the train, and one situated 
below it and called the submsophageal ganglion. The connecting 
cords are known as the circumoisophageal commissures. The 
brain is composed of three embryonic ganglia, and in the adults 
of many lower insects these are still evident as three well-marked 
cerebral divisions or swellings, called the protocerebrum, the deuto- 
cerehrum, and the tritocerehrum. The first carries the optic lohes 
and innervates the compound and simple eyes, the second bears 



two large antennal lobes, from which are given off the antennal 
nerves. The third innervates the lower part of the face and the 
labrum, while it gives off also a pair of nerves which unite in a small 
swelling, the frontal ganglion, that lies between the pharynx and the 
front of the head. A nerve runs posteriorly from this on the dorsal 
side of the pharynx or oesophagus to behind the brain, where it 
divides into several branches, some of which bear small ganglia while 
others extend backward on the cesophagus to the stomach. These 
nerves, originating in the frontal ganglion, constitute the stomato- 
gastric system, sometimes called also the " sympathetic system." 

— AntNv 

Fig. 53. — Brain and subcesophageal ganglion of worker and their principal nerves, 

anterior view. 

The subcesophageal ganglion consists of at most four ganglia which 
innervate the mandibles, the hypopharynx, the first maxillae, and the 
labium or second maxillas. In adult insects the body ganglia also 
very commonly fuse with one another in varying combinations, for 
the number present is always less than the number of segments, vary- 
ing from eleven to one. 

The brain of the bee (fig. 53, Br) is distinctly composed of two 
parts, the protocerebrum [IBr), carrying tlie large optic lobes {OfL), 
and the deutocerebrum {'BBr) , which consists principally of the con- 


spicuous antennal lobes {AntL) that give off the large antennal 
nerves (AntNv). The tritocerebrum is not present as a distinct brain 
division, and its nerves, the labral {LmNv) and the frontal {FtCom), 
appear to arise from the deutocerebrum at the base of the antennal 
lobes. The frontal ganglion {FtGng), formed at the union of the 
two frontal nerves, gives off a very small, anterior, median nerve and 
a much larger, posterior, stomatogastric trunk {StgNv, represented 
in the drawing as cut off a short distance behind the frontal ganglion) 
which goes backward on the dorsal side of the pharynx beneath the 
brain. Behind the latter, and just where the pharynx contracts to 
the tubular oesophagus, the stomatogastric nerve bears a pair of small 
ganglia which are connected by short nerves with the brain, and 
then it breaks up into branches that go posteriorly along the oesopha- 
gus but have not been traced. 

The circumoesophageal commissures are so short in the bee that 
the suboesophageal ganglion appears to be attached directly to the 
lower ends of the brain, while the oesophagus appears to penetrate 
the latter between the antennal lobes. The three principal pairs of 
nerves from the lower ganglion {3/dNv, MxNv, and LhNv) go to the 
mouth parts. 

A most thorough study of the internal structure of the brain of 
the bee has been made by Kenyon (1896), to whose paper the reader 
is referred if interested in this subject. Kenyon 's descriptions have 
never been verified, but his work has an appearance of thoroughness 
and carefulness. He applies the term " brain " to both of the nerve 
masses of the head, distinguishing the upper as the " dorsocerebrum " 
and the lower as the " ventrocerebrum," being led to do this from 
physiological considerations, the separation of the two being merely 
incidental to the passage of the oesophagus. 

In the thorax of the bee (figs. 1 and 52) there are two large ganglia 
{IGng and 2Gng). The first is prothoracic, being situated above 
the prosternum, in front of the entosternum (fig. 52, Fu.^), and it 
innervates the prothorax and the first pair of legs. The second, 
vrhich is situated in front of the middle legs and is protected above 
by the arch of the common entosternum of the mesothorax and meta- 
thorax (fig. 52, Fu^+s), is a combination of the mesothoracic and 
metathoracic ganglia and the first two abdominal ganglia. This 
composite structure is evident from the fact that it innervates both 
the middle and the hind legs, the bases of both pairs of wings, the 
mesothorax, the metathorax, the propodeum, and the first abdominal 
segment behind the constriction (the true second segment) . The first 
and second ganglia of the abdomen (fig. 52, SGng and ^Gng) lie in 
the first two segments (// and ///) behind the constriction, which 
are the true second and third segments. But since the nerve trunks 
of these ganglia go, in each case, to the segments behind them, we 



assume that they really belong to these latter segments, that is, to 
segments /// and IV. The next three ganglia {5Gng, 6Gng, and 
7Gng) lie in the segments they innervate (F, VI, and VII) and, 
hence, belong to the fifth, sixth, and seventh abdominal segments. 
The last, that is, the seventh ganglion, supplies all of the segments 
behind it with nerves and is therefore probably a compound of the 
ganglia originally belonging to the seventh, eighth, ninth, and tenth 

In connection with the nervous system it is most convenient to 
give a description of the simple and compound eyes. The other 

' ^ " X. 

Pig. 54. — Horizontal section of compound eye and optic lobe of worker (after Phillips) : 
BM, basement membrane ; Cor, cornea ; /mi, Jm2, fms, outer, middle, and inner librlllar 
bodies of optic lobe ; inner ch, inner chlasma ; Om, ommatldium ; OpL, optic lobe ; outer 
ch, outer chlasma. 

sense organs will be found already described along with the parts 
on which they are located (see pp. 36 and 52). All the sense organs, 
to be sure, are of ectodermal formation and are only secondarily 
connected with the nervous system, but the eyes are so intimately 
associated with the optic lobes of the brain that their description 
here seems most appropriate. 

The compound eye of the bee (figs. 9 A, 10, 52, and 53, E) has been 
specially studied by Phillips (1905) and figures 54 and 55 are re- 
produced from his drawings, while the following statements are 
based on his paper: The convex outer surface or cornea of the eye 



presents a honeycomb appearance under the microscope, and each 
little hexagonal facet is the outer end of an eye tube called an omma- 





[ -retn. 




Fig. 55. — Histological details of compound eye of worker (after Pliillips) : A, entire 
ommatidium (somewhat diagrammatic), adult; B, entire ommatldium, as if dissected 
out, without outer pigment cells (diagrammatic), adult; C, section of entire om- 
matidium, showing distribution of pigment, adult ; D, cross section just proximal to 
lens, slightly oblique ; B, cross section through extreme distal ends of retinulaj and 
proximal ends of cones, slightly oblique ; F, cross section through retinulee, showing 
relation of out^r pigment cells in this region ; G, cross section through retinulae in 
region of nuclei ; H, cross section through retinulae in region of proximal nucleus ; I, 
cross section of eye, cutting basement membrane parallel (the distinctness of nerve 
fibers of each ommatidium is shown) ; BU, basement membrane; VG. crystalline cone; 
CL, crystalline lens ; c.-p.c, corneal pigment cell ; li.c._, hair-cell ; l.ret.n., lower retinular 
nucleus ; n.f., nerve fiber ; A^e, nerve ; o.-p.c, outer pigment cell ; rei, retinula ; ret.n.^ 
retinular nucleus ; rhb, rhabdome. 

tidium, all of which converge toward the internal base of the eye, 
since each is vertical to the outer surface. Figure 54 is a horizontal 


section through the eye and the optic lobe of the brain. The omma- 
tidia (Om) are seen converging upon the basement membrane (BM) 
which is penetrated by the nerve fibers from the optic lobe (OpL). 
The outer ends of the ommatidia are transparent, forming the 
facets which together constitute the cornea (Cor) of the eye. The 
nerve fibers, by a complicated course through the optic lobe, reach 
the nerve cells of the brain, which are the true seat of sight percep- 
tion, as of all other sensations, whether conscious or otherwise. 

The ommatidia (Om), or eye tubes, are separated from one an- 
other by cells containing a dark coloring matter and known as the 
pigment cells. Each tube (fig. 55 A) consists of several parts, as fol- 
lows: First, on the outside, is a clear six-sided, prismatic structure, 
with convex outer and inner surfaces, called the crystalline lens (CL) , 
and which forms one of the facets of the cornea. Beneath the lens 
is a crystalline cone (CC) having its apex directed inward and 
followed by a crystalline rod or rhabdome {rhb) which extends to 
the basement membrane {BM) through the middle of the omma- 
tidium. (The rhabdome is represented black for the sake of distinct- 
ness in figure 55 A; its natural appearance is more as shown in B 
and C.) Surrounding the rod is a circle of eight or nine long re- 
tinulcB cells {ret), each containing a conspicuous nucleus {ret. n) 
above its middle and continuing basally into an optic nerve fiber {Nv) 
penetrating the basement membrane. The arrangement gf these 
cells about the rhabdome is shown in cross section at F and G. The 
inverted apex of the crystalline cone (A, B, and C, GC) is sur- 
rounded by the corneal pigment cells {c.-p. c.) , while the entire omma- 
tidium below the lens — the base of the cone, the corneal pigment 
cells, and the retinulse — is surrounded by the long outer pigment 
cells {o.-p. c), forming a pacldng between all the ommatidia, as 
shown in cross section at E. 

The entire compound eye is simply a modified part of the epidermis 
(so-called " hypodermis " of insect histologists) in which the cuticle 
is transformed into the lenses or cornea, the cones, and the rods, the 
epithelium into the pigment and retinulsB cells, and the basement 
membrane into the floor of the eye perforated by the optic nerve fibers. 
According to Phillips the ommatidia arise from the ectoderm of the 
bee larva as groups of epithelial cells whicli become arranged in the 
form of spindles surroimded by smaller cells. The cells of the 
spindles become the retinulse, while the surrounding small cells become 
the pigment cells and the cone cells. The cone cells come to occupy 
a position external to the retinulse by an invagination of the latter, 
and, through a transformation of most of their protoplasm into a 
crystalline substance, they form the crystalline cone of the eventual 
ommatidium. The approximated edges of the retinulse cells are 
22181— No. 18—10 9 


transformed into the crystalline rod. The cornea is secreted by the 
corneal pigment cells, which at first lie distal to the cone, and possibly 
also by the outer pigment cells. The nerve fibers are formed as 
differentiated parts of the retinulae which penetrate through the base- 
ment membrane (fig. 54, BM) and enter the retinular ganglion be- 
neath it at the outer end of the optic lobe of the brain. Hence the 
retinulae are simply sense end-organs of the skin comparable at an 
early stage of their development with other sensory epidermal cells, 
and we thus see how a simple layer of epithelium may be transformed 
into such an immensely complex organ as the compound eye. 

There has always been a great deal of discussion as to how insects 
see by means of the compound eyes. The weight of opinion now 
favors the theory that they see a part of the object or field of vision 
with each ommatidium. But it is most certain that, at best, most 
insects see very indistinctly, and, in fact, it is often questioned 
whether they really see the shape of objects at all or not. A few of 
them, however, such as dragonflies, appear to have a very acute 
vision. In the case of the honey bee there is yet much difference of 
opinion as to whether the workers discover nectar by the bright color 
of the flowers (i. e., by the sense of sight) or by the sense of smell. 
The sense of sight in bees and in insects generally, however, may be 
found elaborately discussed in many books dealing with the senses of 

The simple eyes or ocelli (figs. 9 A, 10, 52, and 53, 0) are con- 
structed on quite a different plan from that of the compound eyes, 
each consisting of a lenslike thickening of the cuticle back of which 
the epithelial cells are specially elongated, and sensitized by nerve 
connections. The ocelli of the bee, however, have never been care- 
fully studied. 


The reproductive organs are those that produce the cells from 
which new individuals are formed. All animals grow from at least 
one cell called the egg and almost all from a combination of the egg 
with another cell called a sfermatozoon. The uniting of these two 
cells is called the fertilization of the egg. In a few animals the two 
different kinds of reproductive cells are -formed in the same individ- 
ual, but in most of them, including all insects, the sperm and the eggs 
are produced in different individuals — the males and the females. 
In the honey bee the males are called drones, while the females are 
called queens or workers, according to their functions in the hive. 
The queens have the egg-producing organs or ovaries greatly devel- 
oped, while these organs are rudimentary in the workers. The single 
active queen in each hive, therefore, normally produces all the eggs 
of the colony, while the work of rearing and providing for the brood 


falls to the lot of the workers. Most other female insects lay their 
eggs at some place where the young will be able to find food when they 
hatch out, and the mother never in any way feeds or protects her 
offspring; in most cases she dies before her brood emerges from the 
eggs. But the wasps and bees are different in that nearly all of them 
make a nest of some sort for the protection of the young larvae when 
they hatch, in which also they store up food for them to eat. In many 
species of wild bees all the work of constructing the nest, laying the 
eggs, and collecting and storing food for the young devolves upon 
the single female, as it naturally should, since insects do not ordinarily 
have servants, and the males of most species are utterly irresponsible 
in such matters. In some of the higher wasps, such as the hornets 
and yellow jackets, however, the first females that hatcK out as adults 
in the spring help their mother provide for a still larger family by 
increasing the size of the house and collecting more provisions. 
Nature designed them for this purpose, moreover, by making them 
all sterile, allowing them to retain the maternal instincts, but de- 
priving them of organs capable of producing offspring of their own. 
Thus there is here a beginning of that division of labor which reaches 
its highest development in the honey bee, where one form of the fe- 
male is specialized entirely to produce the young and the other to 
rear the brood, keep the home in order, gather the food, and ward 
off enemies. The differences between the queens and the workers 
are supposed to result from the different diet on which larvae designed 
to be queens are brought up, but a more thorough investigation of the 
food given to the different larvae of the brood is yet needed before 
we can decide on the merits of this explanation. The work of numer- 
ous investigators seems to have demonstrated conclusively that the 
drone of the honey bee is always produced from an egg cell alone — 
that is, from an unfertilized egg — while the queens and workers are 
produced from fertilized eggs. The production of eggs that develop 
normally without the addition of the male element is called partheno- 
genesis. In a number of insects, such as some species of scales, a few 
beetles, and some of the gall-forming Hymenoptera, there are no males 
known, although the females are extremely abundant. Such cases 
are often regarded by entomologists as examples of parthenogenesis, 
and, if they are such, the result of the development of unfertilized 
eggs is here the formation of females only. A few other insects, such 
as some of the plant lice, produce eggs that develop without fertiliza- 
tion into females or into both males and females, but such cases nearly 
always occur in a cycle of alternating generations in which, at some 
stage, all the eggs are fertilized. As far as is known the production 
of males alone from parthenogenetic eggs is confined to the order 



The reproductive organs of the drone are shown by figure 56 A. 
They consist of the testes (Tes), the vasa deferentia (VDef), the 
vedculcB seminales (Ves), the accessory or mucous glands {AcGT), 
the ductus ejaculatorius {EjD), and the penis {Pen). 

The testes of the bee {Tes) are said to be best developed in the pupa, 
at which stage they form the spermatozoa. Each consists of a lairge 
number of small tubules opening into a collecting reservoir at the end 
of the vas deferens. The spermatozoa pass down through the coiled 
vas deferens {VDef) and collect in the saclike enlargement of this 
duct, which constitutes the vesicula seminalis {Ves). In the mature 
adult drone these elongate sacs are densely packed with the active 
spermatozoa, while the testes that produced them become rudimentary. 
The vesiculfe when freshly dissected appear to be alive, for they 
bend and twist themselves about like small worms. Each opens by a 
short duct into the base of the accessory mucous gland {AcGl) of the 
same side. These organs have the forna of two great sacs and are 
filled with a thick, white, homogeneous, finely granular liquid, which is 
supposed to mix with the spermatozoa as the latter are discharged. 
The two open at the bases into the single median ejaculatory duct 
{EjD) which opens into the anterior end of the penis {Pen) . This last 
organ, shown in lateral view by figure 56 E, is an unusually large 
structure in the bee and is deeply invaginated into the cavity of the 
abdomen from the end of the ninth segment (D, Pen) as already de- 
scribed (see page 73). While the penis is simply an ectodermal tube, 
its walls present a number of very curious diflFerentiations. The upper 
part is enlarged into a bulb (fig. 66 A and E, B and PenB) having 
two large irregular but symmetrical chitinous plates {tt) in its dorsal 
wall, beneath which is a large gelatinous thickening (B, ss). 
Near the base of the bulb is a double pinnate lobe (A and E, uu) 
projecting from the dorsal wall. Below this, on the ventral side, 
is a series of close-set, transverse plates (E, vv), followed again by 
large dorsal and ventral plates {ww and xx). The terminal part 
makes a thin-walled chamber (A and E, yy) , from which project 
backward two very large membranous pouches {zz) ending in blunt 
points. The whole tube of the penis is capable of being turned 
inside out, and it is said that copulation is effected by its eversion 
into the oviduct of the queen, the basal pouches of the penis (ss) 
being forced into corresponding pouches of the oviduct, and the 
spermatozoa in the bulb placed, near the opening of the spermatheca 
in the vagina. By their own activity probably the spermatozoa now 
make their way up into this receptacle of the female, the spermatheca, 
where they remain until ejected upon eggs passing down the oviduct. 
The spermatozoa received from one drone normally last the queen 



Pen AcGI 


ji-iQ 58. A, reproductive organs of drone, dorsal view, natural position ; B, inner surface 

of dorsal wall of bulb of penis (E, PenB), showing gelatinous tbiclsening (ss) ; C, 
group of spermatozoa and intermixed granules ; D, terminal segments of male abdomen, 
showing the seventh tergum (VIIT) removed from its sternum (¥IIS) and the penis 
(Pen) partly protruded ; B, lateral view o£ penis as invaginated within abdomen, and 
ejaculatory duct (EjD). 


throughout her life, so that after mating she goes into the hive never 
again to emerge except with a swarm, and her entire life is devoted 
to egg laying. The drone, on the other hand, dies immediately after 
mating, while those that do not mate are driven out of the hive in 
the fall and left to starve. 

The spermatozoa (fig. 56 C) are minute threadlike cells, capable of 
a vibratory motion. As found in the vesiculae, they are usually bent 
into closely compressed loops, although many are extended to their 
entire length. One end is blunt, but not noticeably enlarged, the 
other is tapering, while the half toward the tapering end seems to 
be the part chiefly endowed with the power of motion. The sperm 
threads are contained in a liquid within the vesiculae, in which float 
also a great number of minute granules. The vibrations of the 
spermatozoa keep these granules in constant motion. 


The organs of the female that produce the eggs are called the 
ovaries (fig. 57, Ov). In insects they consist of a varying number of 
egg tubules or ovarioles {ov) forming two lateral groups, in each of 
which the tubules converge at both ends, the anterior ends being 
drawn out into fine threads whose tips are connected, while the poste- 
rior ends are widened and open into the anterior end of the oviduct 
{OvD) on the same side of the body. An egg is simply a very large 
cell whose size is due to the great accumulation of yolk in its proto- 
plasm, which serves as food for the future embryo. The eggs are 
formed in the terminal threads of the ovarioles and are at first appar- 
ently ordinary undifferentiated cells, but as they pass downward' in 
the tubule they increase in size at the expense of some of the other 
ovarian cells. Hence the ovarioles usually have the form of a string 
of beads arranged in a graded series from very tiny ones at the upper 
end to others the size of the mature egg at the lower end. The two 
oviducts converge posteriorly and unite into the common median duct 
or vagina (Vag) which in most insects opens to the exterior upon 
the eighth sternum, as already described in the general account of the 
external anatomy of insects (see page 25), but in the bee and many 
other insects the eighth sternum is entirely lacking as a distinct 
sclerite, and the genital opening is therefore behind the seventh ster- 
num and below the base of the sting. The posterior part of the 
vagina is very large, forming a tursa copulatrix {BCpx). In addi- 
tion to these parts there is nearly always present in insects a special 
receptacle for the spermatozoa called the spermaiheca (Spm). This, 
in most insects, opens directly into the vagina as it does in the bee, but 
in some it opens into the roof of the genital chamber above the eighth 
sternum, when this is present, by a separate orifice behind that of the 



— StnPlp 

Fia. 57. — Reproductive 'organs, sting, and poison glands of queen, dorsal view. 


vagina. In the bee the two poison glands {AGl and BGl) do not 
open into the vagina but, as already described, into the base of the 
sting. They are, hence, probably special organs having no homo- 
logues in nonstinging insects. 

The ovaries of the queen bee form two large gourd-shaped masses 
(fig. 57, Ov) whose posterior or basal ends are enlarged and whose 
anterior ends are narrowed, curved, and attached to each other. 
Since the queen lays, eggs continuously during her entire life the 
ovaries always contain eggs in all stages of growth, and conse- 
quently do not vary so much in appearance as they do in those insects 
that ripen only one lot of eggs and deposit these all at. one time. 

The structure of the ovarioles and the formation of the eggs in the bee 
have been specially studied by Paulcke (1900) and the following is 
a resume of his paper : The terminal threads of the ovarioles are 
covered by a thin tunica propria and are filled with a protoplasmic 
mass containing transversely elongate nuclei in a single close series, 
but no cell outlines.. Farther down, in the upper end of the ovariole 
proper, the nuclei become arranged in two rows, while here also the 
cell boundaries begin to appear; still farther along, where the cells 
are clearly defined, the latter become differentiated into epithelial 
cells and germ cells. Next, the germ cells themselves divide into 
egg cells and food or nurse cells. AVlien first formed the egg cells 
occur in any part, of the diameter of the tube, but they soon become 
arranged in a row down the middle of the ovariole and are separated 
by groups of nurse cells. The epithelial cells at this time arrange 
themselves on the periphery just within the tunica propria, but 
farther down they form a capsule or follicle about the egg and, less 
definitely, about the group of nurse cells at its upper end. The upper 
end of the egg becomes narrowed by a constriction of the epithelial 
capsule, which, however, does not shut it off from the nurse cells, 
a connection being retained with the latter in the form of a neck 
from the egg abutting against them. There are 48 of these nurse 
cells to each egg, which fact is accounted for by supposing that each 
original germ cell divides into 4, one of which ceases further divi- 
sion and becomes the egg cell, while each of the other 3 divides into 
16 by four consecutive divisions. The latter are the nurse cells and 
their function is to nourish the egg cells. They persist down to 
the time when the egg is fully formed, when they suddenly disappear 
by being absorbed bodily into its yolk. Toward the end of the 
growth of the egg the follicle cells become thinner and thinner, so 
that when the egg is ready to go into the oviduct it has but a thin 
membrane to break through. 

The organs of most especial interest to the student of the bee are the 
spermatheea and the apparatus by means of which the queen is able 
to dole out the spermatozoa to the eggs as she deposits the latter. 


The spermatheca consists of a globular seminal sac (fig. 57, Spm), 
of a pair of tubular accessory glands (SpmGl) , and of a duct whose 
upper end is connected with the sac and receives also the duct of the 
glands,' and whose lower end opens into the anterior part of the 
dorsal wall of the vagina just caudad of the united bases of the 

The spermatozoa are discharged by the male into the upper end 
of the vagina, and in some manner they make their way up into the 
sperm sac through the duct. Cheshire (1885) described the latter 
as forking toward its lower end into an anterior branch which opens 
into the vagina and into a posterior branch which turns backward 
and becomes lost in the lower end of the vaginal wall. This second 
branch he believes is open in the young queen and is the one through 
which the spermatozoa enter the sac. Breslau (1906) has shown, how- 
ever, that Cheshire was entirely wrong in his supposed observation 
of the forking of the duct, that the latter is a single tube, and that 
consequently the spermatozoa must both enter and leave the sac by the 
same conduit. It used to be supposed that the sperm sac had muscular 
walls and that it forced the spermatozoa out as from a compressed 
bulb, but Breslau has shown that this also is a mistaken notion, that 
the walls of the sac are entirely devoid of muscular fibers, and that 
the spermatozoa are sucked out by a muscular apparatus in the wall of 
the duct, which structure he names the sperm pump. Cheshire (1885) 
had previously described this apparatus in a very imperfect manner 
without recognizing any pumping function, for he supposed that 
by the relaxation of certain muscles the spermatozoa simply passed out 
of the sac and went down the tube. Breslau says, however, that 
the spermatozoa have not enough energy of their own to come out of 
the sac and, hence, do not need to be kept in by a special sphincter 
muscle, as described by Leydig. 

The upper end of the spermathecal duct makes an S-shaped bend 
jugt beyond the opening of the sac, and a number of muscles dis- 
posed upon this part constitute Breslau's sperm-pump. Breslau shows 
that a contraction of certain of these muscles flattens the bend of 
the S and causes an enlargement of the lumen of the upper end of 
the loop. This, therefore, sucks into itself a small bundle of sperma- 
tozoa from the sac. The contraction then of other muscles forces 
the rest of the sperm-threads back into the mouth of the sac and 
drives the small bundle thus cut off down through the duct and into 
the vagina. Moreover, Breslau claims that this explanation is not 
theory only, for, by preparing histological sections from queens 
killed at different moments of eggrlaying, he procured specimens 
showing the various stages in the pumping process and in the passage 
of the sperm through the duct. Cheshire calculated that a normal 
queen lays 1,500,000 eggs in her lifetime and that the spermatheca 


holds about 4,000,000 spermatozoa, and therefore, allowing for drones, 
he concludes that there can not be more than four sperm-threads given 
to each female egg. But Breslau, figuring from the size of the sperm- 
bundle taken into the ducl^ for each egg, states that each egg is 
actually given 75 to 100 spermatozoa. We feel that the latter calcula- 
tion must be much more reliable than that of Cheshire because it 
is based on an actual observation of the size of the sperm mass de- 
livered to the egg. Moreover, the myriads and myriads of tiny 
spermatozoa contained in the spermathecal sac make any attempt 
at a calculation of the number look absurd, and we can not believe 
that it is possible to even approximate the number present. Fur- 
thermore, as Breslau states, 100 spermatozoa make such an excessively 
small bundle that it requires a most effective and perfect apparatus to 
deliver even this number with anything like exactness — it is incon- 
ceivable that a mechanism could be perfect enough to give out only 
4 or 5 or even 7 at a time. 

On the floor of the vagina, opposite the opening of the spermathecal 
duct, is a free flap provided with muscles, which is so situated 
that when elevated its end fits into the opening of the duct above. 
Leuckart (1858) explained this flap as a contrivance for holding the 
passing egg tight against the upper vaginal wall so that its aperture 
through which the spermatozoa is received, called the micropyle, would 
come against the opening of the duct and thus insure fertilization. 
Breslau, on the other hand, does not think the flap in question has 
any such function and he regards it as a valve which by fitting into 
the orifice of the spermathecal duct closes the latter and so prevents 
the pump from sucking up the contents of the vagina at the same time 
that it sucks a bundle of spermatozoa out of the sac. Since, however, 
the flap is on the floor of the vagina and is pressed down by the passing 
egg it is not clear how it can at such a time act as a valve to close the 
orifice of the duct in the dorsal wall, since the pump is supposed to 
work by reflex action as the egg is entering the vagina, though, of 
course, it may so function before the egg has gone far enough to 
intervene between it and the duct opening; but it would certainly 
seem that a valve to close the latter, if needed at all, would be de- 
veloped in the dorsal wall of the vagina in connection with the orifice 
itself. Furthermore, a collapsible tube like the spermathecal duct, 
even though lined with chitin, should automatically close at its 
lower end when a suction force is applied at the upper end. 

Finally, Breslau attributes to the sperm pump not only the func- 
tion of delivering a definite mass of spermatozoa to each egg, but also 
that of sucking the spermatozoa up from the vagina of a newly fer- 
tilized queen into the spermathecal sac. He does not seem now to see 
in the valve any obstacle to such an action. The spermatozoa are 
usually supposed to make their way up the duct by their own vibra- 
tory motion. 


The anatomy of the spermatheca and the muscular apparatus of its 
duct for delivering the spermatozoa to the egg does not, as Breslau 
points out, throw any light on the determination of sex in bees. It is a 
common notion that all eggs of an unfertilized female develop into 
drones, but this is by no means proved; in fact, there is just as good 
reason for believing that, while no females develop, there are also no 
more than the normal number of drones produced — the eggs that 
might otherwise have developed into females, if laid by a fertile queen, 
all dying in the cells of the comb, from which they are removed by the 
workers. Modern investigation of the determination of sex shows that 
there is probably just as much reason in many cases for supposing that 
sex is established in the egg of the ovary before fertilization, as there 
is for believing it to result from fertilization or from subsequent en- 
vironment of the egg or young embryo. Hence, it is not only very, 
doubtful that the queen determines the sex of her offspring by con- 
trolling the fertilization of the eggs, but it is also very uncertain that 
fertilization itself has anything to do with it. Parthenogenesis in 
the bee may amount simply to this, that the male eggs, predetermined 
as such in the ovary, are capable of developing without fertilization, 
while the female eggs are incapable of such a development and die if 
they are not fertilized. 

Each unlaid egg of insects in general has a small hole in the upper 
end of its shell, called the micropyle, which admits the spermatozoa to 
its interior. One or several spermatozoa may enter the egg through 
this aperture, but the nuclear part of only one unites with the egg 
nucleus, this constituting the fertilization of the egg. After this the 
micropyle closes and the egg is deposited in a cell of the comb by the 
queen. The nucleus and a part of the protoplasm of the egg then 
begin to split up into a number of small cells which — but this is 
taking us into the development of the next generation, which is 
beyond the limits of our subject, and so here we must stop. 



The writer has made an attempt to work out a set of convenient 
symbols for all the principal external and internal parts in the anat- 
omy of an insect. It has been found, however, that entire consistency 
is incompatible with practicability, especially in making compound 
abbreviations, and, therefore, the latter has been given first considera- 
tion in many cases. For example, the symbol Dct suggests the word 
" duct " when standing alone much better than simply the letter Z>, 
but such combinations as SalDct and OvDct are unnecessarily long 
and the shortened forms of SalD and OvD are sufficiently suggestive 
of " salivary duct " and " oviduct." The abbreviation So is used 
in such compound symbols as PsnSc for " poison sac " and TraSc 


for " tracheal sac," notwithstanding that Sc alone means " subcosta." 
The symbol T is used for " tergum," and T^, T^, etc., and IT, IIT, 
etc., indicate individual thoracic and abdominal terga, but TMcl is 
used to signify " transverse, muscle." And so, in several other cases, 
it has been found expedient to sacrifice strict uniformity to practical 

A combination of lower-case letters duplicating one entirely or 
jDartly of capitals signifies that the part so designated is a part or sub- 
division of the other. For example, Ten refers to the principal part 
of the tentorium and ten to a minor part ; PI and pi are subdivisions 
of the same pleurum ; Lmcl and Imcl are both longitudinal muscles. 

The most logical method of referring symbols to any particular 
segment of the body would be, perhaps, to prefix them with either a 
Roman or an Arabic numeral corresponding with the number of the 
segment. A common objection, however, to both would arise from 
the fact that entomologists are not at all agreed as to how many seg- 
ments there are in any region of an insect's body. Furthermore, 
Roman numerals prefixed to all the symbols necessarily used on a 
drawing of the thorax, for example, would occupy entirely too much 
space. Finally, it is very desirable to have a method of referring to 
repeated structures without implying any segmental connection, and 
prefixed Arabic numerals are certainly most convenient and sug- 
gestive for such a purpose. A system often adopted to indicate the 
segment to which a part belongs, especially in the thorax, .is the use 
of one, two, or three accents in connection with the abbreviation. 
But accented symbols lack artistic unity, and some of the accent 
marks are too easily lost in the engraving and printing. For these 
several reasons the writer has adopted the following system: 

Numerical order of any repeated structure is indicated by an 
Arabic numeral placed before the abbreviation, and has no segmental 
significance. Thus IP, 2P, etc., mean simply " first parapterum," 
" second parapterum," etc; IG-ng, 2Gng, etc., mean " first ganglion," 
" second ganglion," etc., without implying that the ganglion belongs 
to any particular segment. 

Symbols are referred to the prothorax, the mesothorax, or the meta- 
thorax, respectively, by the figures 1, 2, and 3 placed below and 
after them, except on the wings, where such numbers designate the 
branches of the veins according to the Comstock-Needham system. 

The abdominal segments, counting the propodeum as the first, are 
indicated by the Roman numerals / to X, and, when any one of these 
is placed before an abbreviation, it refers the symbol to that indi- 
vidual segment. 

The lower-case letters are used, singly and in pairs, to refer to 
miscellaneous parts having, in most cases, no individual or general 
anatomical names. 







































CI, Gls, 










anal vein; XA, first anal. 2.1, second anal, etc. 

accessory gland of male reproductive organs. 

acid gland of sting, opening into poison sac (PsnUo). 

duct of acid gland of sting. 


anterior wing process of notum. 

anterior marginal ridge of notuni. 


antennal lobe of brain. 

antennal nerve. 


apodeme, any internal cbitinous process of body-wall. 

anterior pbragma of any tergum, prephragma. 

the axlllaries or articular sclerites of the wing base, designated 

individually as lAx, 2AJ;, 3Ax, and 4A.i: 
accessory axillary sclerites of irregular occurrence in connection 

with the principal axillaries (Ax). 
axillary cord, or ligament-like thicliening of posterior edge uf basal 

membrane of wing, attached to posterior angle of scutellnm. 
axillary membrane, the thin membrane of wing base, containing 

the axillary sclerites and forming in some cases the lobes called 

bulb (bulb of penis or of sheath of sting). 
any particular part of body cavity such as that prolonged into the 

mouth parts, legs or pieces of the sting, 
bursa copulatrix. 
alkaline gland of sting, 
basement membrane, 



costa, first vein of wing. 

pollen basket or corblculum on hind tibia of worker, 
crystalline cone of compound eye. 

crystalline lens of compound eye. 
cell, cells, 
clasping lobes of ninth segment of male, ])erhaps equivalent to the 

four gonapophyses of ninth segment of female, 
upper or outer clasper. 
lower or inner clasper. 

commissure (of either nervous or tracheal .system). 

cuticle, the chltinous layer of the epidermis, 
cubitus, fifth vein of generalized wing. 








pleural coxal process. 




dorsal diaphragm. 




diaphragm cells. 


membrane of diaphragm. 


muscle fibers of diaphragm. 


compound eye. 


apodeme of extensor muscle. 


ejaculatory duct. 


lateral emargination of notum. 


extensor muscle. 




digestive vesicles formed by ventricular epithelium. 
















foramen magnum. 




frontal commissure. 


frontal ganglion. 


frontal nerve. 


furca or median entosternal apodeme of thoracic sterna. 










large pharyngeal gland in anterior part of head of worker. 


salivary gland in posterior part of head. 


thoracic salivary gland. 


small median gland below pharyngeal plate (s). 










hooks on front edge of hind wing. 






surface disk of " auditory " organ of antenna, probably modified 

base of sensory hair. 


honey stomach. 




individual chamber of heart. 


pericardial cells. 


pericardial, tracheal sac. 


intima, the chitinous lining of any internal organ. 


tergum of first abdominal segment, the median segment, or pro- 

podeum, Incorporated into thorax. 























































labial nerve. 

labial palpus. 


lancet of sting, equivalent to first gonapophysis UG). 

ligula. • 

"lubricating" gland of sting (not shown in figures). 

median lobe of lingua or hypopharyux. 


longitudinal muscles. 

ventral longitudinal muscles of thorax. 

labral nerve. 


trachea of leg. 

lumen, the cavity of any hollow organ, whether the glossa, sting, 

alimentary canal, or gland, 
media, fourth vein of wing. M^-M^, first to fourth branches of 

median plate or plates of wing base. 
Malplghian tubules, 
intersegmental membrane, 

medio-cubital cross-vein, 
disclike muscle apodeme. 

outer saclike mandibular gland, 
inner racemose mandibular gland, 
mandibular nerve, 
mesothorax, designated by figure 2 placed after and below any 

thoracic symbol, 
metathorax, designated by figure 3 placed after and below any 

thoracic symbol, 
the chitinous plates of the neck collectively, the " microthorax," 

Individually designated mi. 
cervical (microthoracic) sclerites. 
median cross-vein, 
mouth parts or trophi. 

maxillary palpus, 
maxillary nerve, 
oblong plate. 

circumoesophageal commissures, 
optic lobe. 









IP, 2P, 
3P, 4P, 




























ostium or lateral aperture of heart. 


ovarlole, individual ovarian tube. 


opening of vagina or median oviduct. 

paraptera, small pleural plates belowr base of wing, typically two 

episternal paraptera or preparaptera (IP and 2P) before pleural 

wing process {WP)^ and two epimeral paraptera or postparap- 

tera {3P and JfP) behind wing process, 
episternal paraptera, preparaptera. 
epimeral paraptera, postparaptera. 
arm of pleural ridge, 
muscle disc of episternal paraptera, giving insertion to pronator 

muscle (not present in the bee), 

bulb of penis, 

subdivision of pleurum. 
palpifer, palpus-carrying lobe of maxilla, 
palpiger, palpus-carrying lobe of labium, 

peritrophic membrane, 
pronator muscle, 
postnotum or pseudonotum, the second or postalar tergal plate of 

the wing-bearing segments of most insects, the " postscutellum " 

of higher orders, 
small rod connecting postscutellum (postnotum PN) with upper 

edge of epimerum, probably a detached piece of the former (see 

figs. 22 and 24). 
posterior notal wing process, 
posterior marginal ridge of notum. 
posterior phragma or postphragma of any tergiim, carried by the 

second notal plate or postnotum (PA), the "postscutellum" of 

higher forms, 
internal pleural ridge, the entopleurum, marked externally by 

pleural suture (PS). 
fossa of proboscis, 
pleural suture, external line separating episternum and epimerum, 

marking site of internal pleural ridge, 

postscutellum ( postnotum ) . 
poison canal of sting. 
poison sac of sting into which opens the acid gland (AGl). 




Pvent Ylv, 
Qd, , 





































sensory pit. 

peritreme, spiracle-bearing sclerite. 


IJroventricular tube or valve in ventriculus. 

quadrate plate of sting. 

radius, third vein of generalized wing. Ri-Ro, first to fifth branches 
of radiuB. Rg, radial sector. 

apodeme of flexor muscle. 

posterior extension or reduplication of any tergal or sternal plate 
overlapping plate following it. 

rectum, the large intestine of insects., 

rectal glands. 

radio-medial cross-vein. 

flexor muscle of mandible or wing. 

dorsal retractor muscle of ligula. 

ventral retractor muscle of ligula. 

radial sector, or second branch of radius at first forliing. 


salivary duct. 

external opening of salivary duct. 

subcosta, second vein of generalized wing. 





sheath of sting, equivalent to the second gonapophyses (2G) or 
middle pair on ninth abdominal segment. 

basal arm of sheath of sting. 

bulb of sheath of sting or ovipositor. 

shaft of sheath of sting. 

small intestine. 


superlingua, embryonic lateral lobes of hypopharynx, true append- 
ages of fifth head segment. 


suboesophageal ganglion. 



spermathecal gland. 


stomatogastric nerve. 


palpuslike appendages of the sting, equivalent to the third gona- 
pophyses (30) or the outer pair on ninth abdominal segment. 


first abdominal tergum, the propodeum, Incorporated into thorax. 

second abdominal tergum. 



large tentorial arms of head, the mesocephalic pillars. 

slender tentorial arch over foramen magnum. 



-No, 18—10 10 



































transverse muscle. 

trochantiu (not separated from sternum in bee). 

coxal condyle of trochantiu. 



transverse ventral tracheal commissures of abdomen. 

tracheal sac. 

triangular plate of sting. 


vas deferens. 

ventral diaphragm. 


ventricular fold or valve in small intestine. 

vesicula seminalis. 

valve of sting carried by lancet. 

large vertical muscles of thorax. 

internal, median V-shaped notal ridge, the " entodorsum." 



mesothoracic wing nerve. 

metathoracic wing nerve. 

wing process of pleurum. 


clypeal suture. 

anterior tentorial pit, in clypeal suture. 

posterior tentorial pit, in occiput beside foramen magnum. 

thickened posterior edge of lateral wall of fossa of proboscis. 

process at upper end of d articulating with cardo of maxilla and 

forming maxillary suspensorium. 
internal median keel of vertex in cranium of drone, 
suspensorial ligaments of anterior end of oesophagus, 
pharyngeal rod. 

convolutions of dorsal blood vessel, 
anterior articular knob of mandible, 
ventral groove of glossa. 
ventral groove of maxillary rod. 
median plates of wing base, 
basal hooks of glossa. 
median ventral plate of ligula. 

dorsal plates of anterior end of mentum, supporting ligula. 
inner wall of canal of glossa. 
chitinous rod of glossa. 

pharyngeal plate, on anterior part of floor of pharynx, 
salivary pouch opening on dorsal side of base of ligula, receiving 

common duct of salivary glands {SaW). 
oblique muscles Inserted upon dorsal side of salivary pouch of 

transverse or V-shaped suture on surface of mesonotum or metano- 

tum, formed by the internal V-shaped ridge or " entodorsum " 

lateral lobe of pronotum projecting posteriorly over the first 



a>, tMoracic plate lying laterad of anterior part of sternum, often 

regarded as a part of presternum. 

y, accessory sclerlte of fourth axillary (.iAx) of front wing, affording 

insertion for slender muscle (fig. 28, cc) attached below to 
common apodeme of mesostei-num and metasternum. 

z, coxal condyles of mesotlioracic and metatboraclc sterna, probably 

really the coxal condyles of trochantins (fig. 4, TnC) fused en- 
tirely with the sterna and episterna in each segment. 

aa, muscle arising from inner wall of mesothoracic pleurum and in- 

serted upon outer end of corresponding scutellum, probably ac- 
cessory in function to the great vertical muscles (fig. 27, TMcl) 
between the mesothoracic sternum and scutum. 

66, coxo-axillary muscle, extending from upper end of coxa to third 

parapterum (3P). 

cc, muscle inserted upon accessory sclerite (y) of fourth axillary 

(IfAa:) from common putosternum of mesothorax and meta- 

M, notch of antenna cleaner on first tarsal Joint (ITar) of front leg. 

ee, spine of antenna cleaner situated on distal end of tibia (Ti). 

jf, so-called " wax shears " or " wax pincers." 

gg, transverse chitinous band of empodium (Enip), which compresses 

its two lobes when not in use and spread out by muscular effort. 

lili, dorsal plate supporting empodium. 

ii, ventral plate supporting empodium. 

jj, dorsal groove of lancet interlocking with ventral ridge of sheath of 


kk, sting chamber within end of seventh abdominal segment, lodging 

sting whose accessory plates are derived from eighth and ninth 

II, reservoir of thoracic salivary gland. 

mm, receptacular chitinous pouches on ventral side of pharyngeal plate 

(s) receiving ducts of large lateral pharyngeal glands of head 

nn, " stomach-mouth " at summit of proventricular projection within 

honey stomach (HS). 

00, pores on lancets (fig. 40 E) and shaft of sting sheath (F) open- 

ing to exterior from prolongation of body-cavity (6c) contained 
in each. 

pp, gelatinous layer secreted upon inner surface of ventricular epi- 


gg, food contents of alimentary canal. 

rr, cells of ventricular epithelium apparently forming the internal 

gelatinous layer. 

ss, cartilaginous mass on inner surface of dorsal wall of bulb of 

penis (flg. 56 B, PenB). 

tt, dorsal plates of bulb of penis. 

MM, fimbriated dorsal lobes of penis at base of bulb. 

vv, ventral scalarlform row of plates on tube of penis. 

WW, dorsal basal plates of penis. 

xso, ventral basal plates of penis. 

yy, basal pouch of penis. 

zn, copulatory sacs of penis. 



Abonyi, Sandor. 

1903. Morphologische und physiologische Besclireibung des Darmkanals der 

Honigblene (Apis mellifloa). Ablatt Kozlem., II, pp. 137-168, pis. 
Abnhabt, Ltjdwig. 

1906. Anatomie und Physiologle der Honigbiene, 99 pp., 4 pis., 53 figs. Wien, 

1906. [Extract from " Allgemelnes Lehrbuch der Blenenzucht " 
by Alois Alfonsus.] 


1894. Anatomie du tube digestif des Hymenopt^res. Comptes Kendus de 

I'Acad. des Sci., Paris, CXVIII, pp. 1423-1425. 
1894. Glandes salivaires des Apinae (Apis melliflca $ et ? ). Compte- 

rendu de la See. Phil, de Paris, 1894, No. 4, pp. 12-14. 

1894. Glandes salivaires des Apinee (Apis melliflca ^ et S). Bui. de la 

See. Pbil. de Paris, 8 ser., Vll, 1894-1895, pp. 9-26, 9 figs. 

1895. Appareil ggnital male des Hymfnoptgres. Ann. Sci. Nat., XX, 7 ser., 

1895, pp. 103-181. 
Beeithadpt, Paul Feanz. 

1886. Ueber die Anatomie und die Functionen der Bienenzunge. Arch. f. 
Xaturg., LIII, pp. 47-112, pis. IV~\. [Brief summary in Journ. 
Roy. Micr. Soc, VII, Pt. I, 1887, pp. 224-225.] 
Bbeslau, Eenst. 

1905. Die Samenblasengang der Bienenkonigin. Zool. Anz., XXIX, 1905- 
1906, pp. 299-323, 7 figs. 
Bbiant, T. J. 

1884. On the anatomy and functions of the tongue of the honey bee 
(worker). Journ. Linn. Soc. London, Zool., pp. 408-417, Pis. 
Cablet, G. 

1884. Sur les muscles de I'abdomen de I'abeille. Comptes Rendus de I'Acad. 
des Sci. de Paris, XCVIII, pp. 758, 759. 

1884. Sur le venin des HymgnoptSres et ses organes sficreteurs. Comptes 

Rendus de I'Acad. des Sci. de Paris, XCVIII, pp. 1550, 1551. 
1890. Memoir sur le venin et I'aiguillon de I'abeille. Ann. des Sci. Nat, 

Zool., 7 ser., IX, pp. 1-17, pi. 1. 
1890. Sur les organes secreteurs et la secretion de la cire chez I'abeille. 

Comptes Rendus de I'Acad. des Sci. de Paris, CX, pp. 361-363. 
Cheshire, Prank R. 

1885. The apparatus for differentiating the sexes in bees and wasps. An 

anatomical investigation into the structure of the receptaculum 
seminis and adjacent parts. Journ. Roy. Micr. Soc, ser. 2, V, pp. 
1-15, Pis. I, II. 

1886. Bees and bee keeping, 2 vols., London. (Vol. I devoted mostly to 

anatomy of the bee.) 
Clerici, F. 

1875. L'Ape sua anatomia — suoi memici. Milan. [30 colored plates drawn 

from anatomical preparations by G. Bart6.] 
Cook, A. J. 

1904. The bee keeper's guide, 18th ed., Chicago. 
CoPELAND, Manton, and Mark, E. L. 

1907. Some stages in the spermatogenesis of the honey bee. Proc. Amer. 

Acad. Art. and Sci., XLII, pp. 103-111, 1 pi. 


Cowan, T. W. 

1904. The honey bee, 2d ed., London. • 

Djathchbnko, Sophie. 

1906. Zur Frage der Athumsorgane der Biene. Ann. de I'lnst. agron. de 

Moscou, XII, pp. 1-14, 9 flgs. 

Dretling, L. 

1903. Ueber die wachsbereitenden Organe der Honigbiene. Zool Anz., 
XXVI, pp. 710-715, 2 flgs. 


1907. The senses of insects. English ed., translated by Maclood Yearsley, 



1878. Les abeilles. Paris. 

GiEDWOYN, Michel. 

1876. Anatomie et physiologie de I'abeille. Mem. de la Soc. Polonaise des 
ScL Exac, VI, 39 pp., 12 pis., Paris. [Also published separately.] 


1904, 1905. Anatomie et physiologie de I'abeille domestique. Le Microg. 
Prep., XII, pp. 49-60, pis. XXVII, XXVIII; XIII, pp. 15-25, 
1906. Apiculture (Encyclopedie Agricole). Anatomie de I'abeille, pp. 20- 
103, Paris. 
I-IUBEE, Francois. 

1814. Nouvelles observations sur les abeilles, 2 vols. 2d ed., Paris and 
Kenyon, F. C. 

1896. The brain of the bee. Journ. Comp. Neurol., VI, pp. 133-210, pis. 


1897. The optic lobes of the bee's brain in the light of recent neurological 

methods. Amer. Nat., XXXI, pp. 369-876, pi. IX. 


1900. Ueber den Fettkorper und die CEnocyten der Honigbiene {Apis rnelli- 
fera L). Zool. Anz., XXIII, pp. 337-353. 
Keaepelin, K. 

1873. Untersuchungen iiber den Bau, Mechanismus und Entwickelung des 
Stachels der bienenartigen Tiere. Zeit. f. wiss. Zool., XXIII, pp. 
289-330, pis. XV, XVI. 
LuDwiG, August. 

1906. Unsere Bienen. Die Anatomie, pp. 34-112, Berlin. 
Macloskie, George. 

1881. The endocranium and maxillary suspensorium ot the bee. Amer. 
Nat., XV, pp. 353-362, 6 figs. 
Maek, E. L., and Copeland, Manton. 

1907. Some stages in the spermatogenesis of the honey bee. Proc. Amer. 

Acad. Art. and Sci., XLII, pp. 103-111, 1 pi. 
Meves, Feiedeich. 

1903. Ueber Richtungskbrperbildung im Hoden von Hymenopteren. Anat. 

Anz., XXIV, pp. 29-32, 8 figs. 
1907. Die Spermatocytenteilungen bei der Honigbiene {Apis nielli flea L.), 

nebst Bemerkungen uber Chromatinreduction. Arch. f. Microsk. 

Anat. und Entwick., LXX, pp. 414-491, 5 flgs. pis. XXII-XXVI. 



1900. Bau und Entwickelung des mannlichen Begattungsapparates der Ho- 
nigbiene. Zeit. f. wiss. Zool., LXVII, pp. 439-460, pi. XXVI. 
Patjlcke, Wilhelm. 

1900. Ueber die Differenzirung der Zellelemente im Ovarium der Bienen- 

konigin {Apis mellifica). Zool. Jahrb., Anat. und Ontog., XIV, 1900, 

1901, pp. 177-202, pis. 12-13a. 
Phillips, Everett Franklin. 

1905. Structure and development of the compound eye of the honey bee. 

Proc. Acad. Nat. Sci. Phila., LVII, pp. 123-157, 7 figs., pis. VI-VII. 
Pissarew, W. J. 

1898. Das Herz der Biene (Apis mellifica). Zool. Auz., XXI, pp. 282, 283, 

1 fig. 
Planta, a. von. 

1888 and 1889. Ueber den Futtersaft der Biene. Zeit. f. Physio. Chemie, 

XII, 1888, pp. 327-354; XIII, 1889, pp. 552-561. 
Plateatt, Felix. 

1874. Recherches sur les ph6nom&nes de la digestion chez les insectes. M6m. 

de I'Acad. Roy. des Sci., des Let. et des Beaux-arts de Belgique, 
XLI, pp. 3-124, pis. I-III. 
Root, A. I., and Root, E. R. 

1908. The ABC and X Y Z of Bee Culture, Medina, Ohio. 
^AMtrELSON, J., and Hicks, J. B. 

1860. The honey bee. London. 


1883. Ueber des Herkommen des Futtersaftes und die Speicheldriisen der 
Bienen, nebst einem Anhange iiber das Riechorgan. Zeit. f. wiss. 
Zool., XXXVIII, pp. 71-135, pis. ^'-VII. 
ScHMiD, A., and Kleinb, G. 

1861. Umrisse zur Anatomie und Physiologie der Bienen. Die Bienen- 

zeitun^, I, pp. 498-525, pis. I-VII. 


1886. Die physiologische Bedeutung des Magenmundes der Honigbiene. 
Arch. f. Anat. und Physiol., Physiol. Abth., pp. 451-458. 

Sladen, p. W. L. 

1901. A scent-producing organ in the abdomen of the bee. Gleanings in 

Bee Culture, XXIX, August, pp. 639, 640, 1 fig. 

1902. A scent-producing organ in the alsdomen of the worker of Apis 

mellifica. Ent. Mag. Lond., XXXVIII, pp. 208-211, 1 fig. 
Wolff, O. J. B. 

1875. Das Riechorgan der Biene. Nova Acta der Ksl. Leop.-Carol. Deut. 

Akad. der Naturf., XXXVIII, pp. 1-251, pis. I-VIII. 

Zander, Enoch. 

1899. Beitrage zur Morphologic des Stachelapparates der Hymenopteren. " 

Zeit. f. wiss. Zool., LXVI, pp. 288-333, pis. XVIII, XIX. 

1900. Beitrage zur Morphologic der mannlichen Geschlichtsorgane der 

Hymenopteren. Zeit. f. wiss. Zool., LXVII, pp. 461-489, 9 figs,, 
pi, XXVII. 



A-bdomen, defined 13-14 

general structure 24-26 

of honey bee 69-71 

muscles 118-119 

wax glands and sting 69-83 

Absorption 104-105 

Accessory sternal plate in generalized thorax 21 

Acid gland of sting 79 

Alimentary canal 90 

and its glands 84-106 

Alkaline gland of sting 79 

Anabolism : 86 

Anal veins of generalized wing 22 

Hymenoptera » 59 

Andrena, pharyngeal glands 92 

Anteclypeus 15 

Antenna cleaner 66 

Antennae and their sense organs 32-39 

defined 16 

of honey bee 27, 32-33 

sense organs 36-39 

Antennal lobes 125 

Anterior notal wing process, defined 19 

of mesothorax of honey bee 62 

metathorax of honey bee 56 

Anihophora, pharyngeal glands 92 

Anus 78 

Aorta 108-109 

Apodomes 32 

Appendages 12-13 

Assimilation 84-87 

Axillaries, defined 23 

of front wing of honey bee ■ 62 

hind wing of honey bee 62 

Hymenoptera 59-62 

Axillary cord, defined 19 

of generalized wing 23 

honeybee 62 

membrane of generalized wing 23 

honey bee 61 

Basal ligament of wing. (See Axillary cord.) 

Basement membrane 129 

Blatta, development of head 18 

Blood 207 

Body wall 14_2 5 




Bombiis, pharyngeal glands 92 

Bouton 45 

Brain, general deBcription 124 

of honey bee 125-126 

Brood food 92-94 

production of, and summary of facta known concerning it 98-101 

Bursa copulatrix .'. 134 

Carbohydrates 89 

Cardo, defined 17 

of honey bee 45 

Cells of body, defined 84-86 

wing, defined 23 

Cerci, defined 24 

Cervicum 13 

Chiasma, inner 127 

outer 127 

"Chyle" 101 

"stomach " 90 

' ' Chyme " 101 

Circulation of the blood 107-111 

Circulatory system 107-111 

Circumoesophageal commissures, defined 124 

in honey bee 126 

Clasping organs, defined 24 

of drone 73 

Claws of tarsus 22, 68-69 

Clypeal suture : 28 

Clypeus, defined 15 

of honey bee 28 

Cockroach, Blatta, development of head 18 

Commissures 124 

Compound eyes 27 

detailed structure 127-130 

Conocephalus, ovipositor 25 

Corbicula 66-68 

Cornea 27, 127-129 

Corneal pigment cells 129 

Costa of generalized wing '. 22 

Hymenoptera 59 

Coxa, defined 20, 22 

of honey bee 67 

Coxo-axillary muscle of wing of honey bee 66 

Coxosternum 21 

Cranium, internal structure 30-32 

Cricket, Gryllus pennsylvanicus, mouth parts 16-18 

Crop, defined 90 

of honey bee (honey stomach) 94-95 

Cross veins of generalized wing 22-23 

Crystalline cone 129 

lens 129 

rod 129 

Cubitus of generalized wing 22 

Hymenoptera 60 

INDEX. 153 

Darts. (See Lancets.) Page. 

Dendroctonus, " preepisternum '' 20 

Depressor muscles of wing of honey bee 64 

Determination of sex in honey bees 139 

Deutocerebrum, defined 124 

in honey bee 125-126 

Development, defined 11-12 

Diaphragm cells 110 

Diaphragms, defined 107-108 

dorsal 109-110 

ventral 109 

Digestion 86, 89 

assimilation, and excretion, general physiology 84-87 

Dorsal diaphragm 109-110 

sinus 107-108, 111 

" Dorsocerebrum " 126 

Dorsum, defined 18 

Drones, defined 130 

Ductus ejaculatorius 132 

Egg, defined 130 

fertilization 137, 138, 139 

formation 136 

Elevator muscles of wing of honey bee 64 

Embryo 12-14 

Embryonic development, defined 12 

Empodium, defined 22 

of honey bee 68, 69 

Entocranium 31, 32 

Entodorsum, defined 19 

Entopleurum, defined 19 

of mesothorax of honey bee. (See Pleural ridge.) 

Entostemum (furca), defined 21 

of prothorax of honey bee 55 

mesothorax and metathorax of honey bee 56 

Entotergum, defined I9 

Entothorax, defined 32 

Enzymes 87 

Epicranium, defined Ig 

■ Epimeral paraptera (postparaptera), defined 20 

of mesopleuium of honey bee 56 

Epimerum, defined I9 

of mesopleurum of honey bee 5g 

Epipharynx, defined 16 

of honey bee 51_53 

sense organs 52-53 

Epipleurites, defined 24 

Episternal paraptera (preparaptera), defined 20 

of mesopleurum of honey bee 56 

Episternum, defined j^g 

of mesopleurum of honey bee 5g 

Excretion 84-87 

Extensor muscle of mandible of honey bee 40 

External genital organs of drone honey bee 72-73 

development 73-74 



External mandibular glands 41 

Eye, compound 27, 127 

simple 27, 130 

Facets of compound eye. 27 

Fat body 119,120 

and cenocy tes 119-121 

Female organs of reproduction 134-139 

Femur, defined 22 

of honey bee 67 

Fertilization of egg, defined 130 

of honey bee 137-139 

First abdominal segment (propodeum) 58-59 

Flagellum 32-33 

Flexor muscle of mandible of honey bee 40 

wing of honey bee 65 

Food of adult honey bees 89 

larvae. {See Brood food and Royal jelly.) 

Foramen magnum, defined 15 

of head of honey bee. . ! 28 

Fossa of proboscis 28, 46 

Frontal ganglion 125 

Front, defined 15 

of head of honey bee 29 

Furca (entosternum), defined 21 

of prothorax of honey bee 55 

mesothorax and metathorax of honey bee 56 

Galea, defined 17 

of maxilla of honey bee 46 

Ganglia, defined 124 

" Gaumensegel " 52 

Genae, defined 15 

of head of honey bee 29 

General physiology of digestion, assimilation, and excretion 84-87 

Gills, defined.... 112 

Glands, external mandibular 41 

internal mandibular 42 

lateral pharyngeal 91, 92 

median pharyngeal 91 

mucous glands of male organs 132 

of Nassanoff ; . . 83 

sting. 78-80 

acid ,. ." 78-79 

alkaline 79 

"lubricating " 78 

postcerebral 87-88 

rectal 90, 106 

salivary, of head 87-88 

thorax 88-89 

sublingual 91 

supracerebral 91 

Glossa, defined ^ 45 

details, in worker of honey bee 48 

Glossse, defined 17 

of labium of generalized insect ^ 17, 44 

INDEX. 155 


Glossal rod ^^ 

Gonapophyses, defined 24 

of ovipositor of longhorned grasshopper 25 

sting of honey bee '^6 

Grasshopper, longhorned ( Conoceplialus), ovipositor 25 

Growth, defined 1^ 

Gryllus pennsylvanicus , mouth parts 16-18 

Gular sclerites, defined 1^ 

Head, defined 15 

of honey bee and its appendages 26-53 

external structure 26-30 

internal structure 30-32 

worker, queen and drone, compared 29-30 

Heart, chambers 108 

general description 108 

of honey bee 109 

Honey stomach .- 90i 94-95 

Horntail, Sirex flavicornis, first abdominal segment 58 

metapleurum 57 

wing veins 60-62 

Humeral cross-vein, defined 22 

Hydrocarbons 89 

Hypopharynx, confusion with glossa of honey bee 44 

defined..., • 16,17 

Hypopleurites, defined 24 

Hypopygium, defined - 73 

Imago, defined - 12 

Ingluvies, defined 90 

Insects, general external structure 10-26 

Internal mandibular glands 42 

Intersegmental membrane, defined - 14 

in abdomen of honey bee 70-71 

Intestine 90, 105, 106 

Itycorsia discolor, wing veins 59-61 

Joyful hum 83 

Katabolism 86 

Labella - 45 

Labial palpi, defined 17 

of honey bee 44 

Labium, defined 16 

of generalized insect 17 

honeybee 27,44 

Labrum, defined 15 

of generalized insect 16 

head of honey bee 28 

Lacinia, defined 17 

Lancets of sting 75 

Large intestine 90, 106 

Larva, defined 12 

Larval stage, defined 12 

Lateral pharyngeal glands 91-92 

Latus, defined 18 

Legs of generalized insect 21-22 

worker, queen, and drone 66-69 



Ligula, defined 17 

of honey bee .' 44^5 

Lingua, confusion with, other terms 44, 45 

defined 17 

Lorum 46 

" Lubricating " glands of sting 78 

Male organs of reproduction 132-134 

Malpighian tubules, defined 90 

of honey bee ' 105-106 

Mandibles, defined 16 

of generalized insect 16 

honey bee ^ 27, 39-11 

Mandibular glands, external 41 

internal 42 

Maxillae, defined 16 

of generalized insect 17 

honey bee 27, 44 

Maxillary palpus, defined 17 

of honey bee 45 

Maxillary suspensorium 32 

Median pharyngeal glands 91 

plates of generalized wing 23 

segment (propodeum) of Hymenoptera 59 

Media of generalized wing 22 

wings of Hymenoptera 60 

Medio-cubital cross-vein, defined 23 

Mentum, defined - 17 

of labium of honey bee 44 

Mesocephalic pillars 31 

Mesopleurum 56 

Mesosternum 56 

Mesotergum 55 

Mesothorax, defined 13 

of honey bee 55-56 

Metabolism, defined 86 

process 115 

Metameres, defined 12 

Metapleurum • 57 

Metastemum ; 56 

Metatergum 56 

Metathorax, defined 13 

of honey bee 56-58 

Micropyle of egg, defined 139 

Microthorax, defined 13 

of generalized insect - 18 

Motion of wings in flight 63 

Mouth, defined 49 

of honey bee 28 

parts, defined 16 

of generalized insect 16-18 

honey bee 27-28, 39-53 

action in feeding 46-49 

Mucous glands of male organs 132 

Muscles of flight 63-66 

INDEX. 157 


Neck or cervicum, defined 13 

of generalized insect 18 

Nervous system and the eyes 122-130 

general description 124-125 

physiology 122-124 

of abdomen of honey bee 126-127 

head of honey bee 125-126 

thorax of honey bee 126-127 

Notum, defined 19 

Nymph, defined 12 

Oblong plate of sting 75-76 

Occiput, defined 15 

of head of honey bee 29' 

Ocelli 27, 130 

OEnocytes 114-115,120-121 

ffisophagus, defined 90 

of honey bee 94 

Ommatidia of compound eye 128 

Optic lobes of brain 124 

Ostia of heart = 108 

Ovaries, defined 134 

structure in queen bee 136 

Ovarioles, defined 134 

structure in queen bee 136 

Oviducts, defined .• 134 

Ovipositor .' . 24-26 

Palpifer, defined 18 

of maxilla of honey bee 45 

Palpiger, defined 17 

of labium of honey bee 44 

Palpi, labial, defined 17 

of honey bee 44 

maxillary, defined 17 

of honey bee 45 

of sting of honey bee 76 

Paraglossse, defined 17 

of labium of honey bee 44, 47- 

Paraptera, defined , 20 

of mesopleurum of honey bee 56 

Parthenogenesis 131 

Penis 132 

Pepsis, wing veins : '. 60-62 

Pericardial air sacs Ill 

cells Ill 

chamber, defined 108 

of honey bee (dorsal sinus) 107, 111 

Peritrophic membranes 101-102, 104 

Pharyngeal glands, lateral _. 91-92 

median 91 

plate 50,91 

Pharynx, defined 90 

of honey bee 90-91 

Phragmas, defined ig 

of mesotergum of honey bee 55-56 



Pigment cells of compound eye 129 

Pleural coxal process, defined 19 

ridge (entopleurum), defined 19 

of mesopleurum of honey bee * 56 

suture, defined 19 

of mesopleurum of honey bee 56 

wing process, defined 19 

of mesopleurum of honey bee 56 

Pleurites, defined 14 

Pleurum, defined 14 

of generalized insect 19-20 

mesothorax of honey bee 56 

metathorax of honey bee 57 

prothorax of honey bee 55 

Poison glands of sting of honey bee 78-79 

sac of sting of honey bee , 78-79 

Pollen baskets of worker bee 66 

Postcerebral glands 87-88 

Postclypeus, defined 15 

Postembryonic development, defined 12 

Posterior notal wing process, defined 19 

of mesothorax of honey bee 62 

metathorax of honey bee 56 

Postgena, defined = 16 

of head of honey bee 29 

Postnotum (pseudonotum), defined 19 

of mesotergum of honey bee 55-56 

Postparaptera, defined 20 

of mesopleurum of honey bee 56 

Postscutellum, defined 19 

of mesotergum of honey bee 55 

Poststemellum, defined 21 

Preepistemum, defined 19-20 

Preoral cavity 49 

Preparaptera, defined 20 

of mesopleurum of honey bee 56 

Prescutum, defined 19 

Presternum, defined 20 

Proboscis 27, 43-51 

Pronator apparatus of wing 65 

Propodeum 58-59 

Proteids 89 

Prothorax, defined 13 

of honey bee ' 55 

Protocerebrum, defined 124 

in honey bee 125 

Protoplasm 86 

" Protractor linguse " 51 

Proventricular valve 97 

Proventriculus, defined 90 

of honey bee 95-98 

Pseudonotum, defined 19 

JPsithyrus, pharyngeal glands 92 

INDEX. 159 


Pulvilli, defined 22 

Pupa, defined 12 

Pupal stage, defined I - 

Quadrate plate of sting of honey bee 76 

Queens, defined 130 

function in hive - 130-131 

Radio-medial cross-vein, defined -'^ 

Radius of generalized wing 22 

wings of Hymenoptera 59-62 

Rectal glands, defined 90 

of honey bee 106 

Rectum, defined. . . i 90 

of honey bee 106 

Reproductive system 130-139 

of drone bee 132- 1 34 

queen bee 134-139 

Respiration, movements 118 

muscles 118-119 

physiology 112-114 

Respiratory system 112-119 

Retinulse cells 129 

"Retractor linguae biceps" 51 

"longus" 51 

Rhabdomes of compound eye 129 

" Riechschleimdrusse " 41 

Royal jelly 92-94 

Salivary glands 87-89 

of head 87-88 

thorax 88-89 

opening on labium 49 

pump So 

Sawfly, Itycorsia discolor, wiug veins 59-61 

Scape 32 

"Schlundbein" 50 

Sclerites, defined 14 

Scutellum, defined 19 

of mesotergum of honey bee 55 

Scutum, defined 19 

of mesotergum of honey bee 55 

Second maxillae, defined 17 

Sheath of sting 75 

basal arms 75 

bulb ' 75 

shaft 75 

Simple eyes 27, 130 

Sinuses, defined 107 

Sirex flavicornis, first abdominal segment 58 

metapleurum 57 

wing veins 60-62 

Small intestine, defined 90 

of honey bee 105 

Smell, sense 33-39 



Somites, defined 12 

Spermatheca, defined 134-135 

structure in queen 136-137 

Spermatozoa, defined 130 

of honey bee 134, 137-138 

Sperm pump of spermatheca 136-138 

Spiracles, defined 26, 112 

of honey bee 115-116 

Sternal laterale, defined 21 

Sternellum, defined 21 

Sternites, defined 14 

Sternum, defined 14 

of generalized insect 19 

mesothorax of honey bee 56 

metathorax of honey bee 56 

proper, defined 20 

Stimuli, afferent, defined 124 

efferent, defined 124 

Sting 74-83 

injection of poison 80-82 

morphology 77-78 

of queen bee 82-83 

Stipes, defined 17 

of maxilla of honey bee '. 45 

Stomach mouth (proventriculus) 95-96 

(ventriculus), defined 90 

of honey bee 98 

Stomatogastric nervous system, defined 125 

of honey bee 126 

Subcosta of generalized wing 22 

wings of Hymenoptera 59 

Subgalea 45 

Sublingual glands 91 

Submentum, defined 17 

of labium of honey bee 44 

Subcesophageal ganglion, defined 124 

of honey bee 126 

Superlinguse of embryo 17 

Supracerebral glands 91 

Sutures, defined 14 

Sympathetic nervous system, defined 125 

of honey bee 126 

Tarsus, defined '. 22 

of honey bee 67 

first joint 66 

last joint 68-69 

Taste organs 52 

Tegula, defined 23 

Temperature of honey bees 115 

Tenth segment of abdomen 78 

Tentorium 31 

Tergites, defined 14 




Tergum, defined 14 

in generalized thoracic segment ^^ 

of abdominal segments o£ honey bee 69-70, 72-73 

first abdominal segment of honey bee 58-59 

mesothorax of honey bee '^° 

metathorax of honey bee 56-57 

prothorax of honey bee ^^ 

Testes 132 

Thoracic salivary glands 88-89 

segment, typical 18-19 

Thorax, defined 13 

generalized segment -. 18-19 

of honey bee and its appendages 53-69 

special characters, in honey bee 54-55 

Tibia, defined 22 

of honey bee 67 

Tongue 27, 44-45 

Trachea, defined 26, 112 

of honey bee 116-118 

Triangular plate of sting of honey bee 76 

Tritocerebrum 124 

Trochanter, defined 22 

of honey bee 67 

Trochantin, defined 20 

of honey bee 57-58 

mandibles, defined 16 

Trophi, defined 16 

Vagina, defined 134 

valve 138 

"Valva externa" 73 

' ' interna " 73 

Vasa deferentia 132 

Veins of generalized wing 22-24 

wings of Hymenoptera 59-62 

Venter, defined 18 

Ventral diaphragm 109 

or median pharyngeal glands 91 

sinus 107-109 

Ventriculus, defined 90 

of honey bee 98 

contents 98 

histology 102-104 

" Ventrocerebrum " 126 

Vertex, defined 15 

of head of honey bee 29 

Vesiculse seminales 132 

Wax glands 71 

secretion 71-72 

shears 68 

Wing processes of notum, defined 19 

of mesonotum of honey bee 62 

metanotum of honey bee 56 

22181— No. 18—10 11 



Wing processes of pleura, defined 19 

of mesopleurum of honey bee 56 

Wings, articulation in generalized insect 23 

honey bee 62-63 

defined 13 

motion 63 

muscles 1 63-66 

of generalized insect 22-24 

honey bee 62-66 

Hymenoptera 59-62 

pronator apparatus 65 

veins 59-62 

Workers, defined 130 

function in hive 130-131 



L. O. HOWARD, Entomologut and Chief of Buieau. 


D. B. qASTEEL, Ph. D., 

Collaborator and Adjunct Professor of Zoology, 
University of Texas. 

Issued December 31, 1912. 





L. O. HowABD, Entomologist and Chief of Bureau. 

C. L. Maklatt, Entomologist and Acting Chief in Atsence of Chief. 

R. S. Clifton, Executive Assistant. 

W. F. Tastet, Chief Clerk. 

F. H. Chittenden, ijn charge of truck crop and stored product insect investigations. 

A. D. Hopkins, m charge of forest insect investigations. 

W. D. HuNTEB, in charge of southern field crop insect investigations. 

F. M. Webster, in charge of cereal and forage insect investigations. 
A. L. QtTAiNTANCE, in charge of deciduous fruit insect investigations. 
E. F. Phillips, in charge of See culture. 

D. M. RoGEBS, in charge of preventing spread of inoths, field work. 
ROLLA P. CuKBiE, TO Charge of editorial work. 
Mabel Colcoed, vn charge of library. 

Investigations in Bee Cultube. 
E. F. Phillips, in charge. 

G. F. White, J. A. Nelson, experts. 

G. S. Demttth, a. H. MoCbat, N. B. McIndoo, apicultural assistants. 
Peable H. Gabbison, preparator. 
D. B. Casteel, collaborator, 


U. S. Department oi" Agriculture, 

Bureau of Entomology, 
Washington, D. G., September £3, 1912. 
Sir : I have the honor to transmit herewith a manuscript entitled 
" The Behavior of the Honey Bee in Pollen Collecting," by Dr. Dana 
B. Casteel, of this bureau. The value of the honey bee in cross pol- 
linating the flowers of fruit trees makes it desirable that exact infor- 
mation be available concerning the actions of the bee when gathering 
and manipulating the pollen. The results recorded in this manu- 
script are also of value as studies in the behavior of the bee and will 
prove interesting and valuable to the bee keeper. The work here 
recorded was done by Dr. Casteel during the summers of 1911 and 
1912 at the apiary of this bureau. 

I recommend that this manuscript be published as Bulletin No. 121 
of the Bureau of Entomology. 

Respectfully, L. O. Howard, 

Entomologist and Chief of Bureau. 
Hon. James Wilson, 

Secretary of Agriculture. 




Infioduction 7 

The structures concerned in the manipulation of pollen 7 

The pollen supply 10 

General statement of the pollen-collecting process 11 

Action of the forelegs and mouthparts. 13 

Action of the middle legs 14 

Action of the hind legs 16 

Additional details of the basket-loading process 18 

Pollen moistening 22 

Storing pollen in the hive 29 

Summary 31 

Bibliography 33 

Index 35 




Fig. 1. Left foreleg of a worker bee 8 

2. Left middle leg of a worker bee 9 

3. Outer surface of the left hind leg of a worker bee 10 

4. Inner surface of the left hind leg of a worker bee : 11 

5. A flying bee, showing the manner in which the forelegs and middle legs 

manipulate pollen 14 

6. A bee upon the wing, showing the position of the middle legs when they 

touch and pat down the pollen masses 15 

7. A bee upon the wing, showing the maimer in which the hind legs are 

held during the basket-loading process 17 

8. The left hind legs of worker bees, showing the manner in which pollen 

enters the basket 19 

9. Inner surface of the right hind leg of a worker bee which bears a com- 

plete load of pollen 22 




While working upon the problem of wax-scale manipulation dur- 
ing the summer of 1911 the writer became convinced that the so- 
called wax shears or pinchers of the worker honey bee have nothing 
whatever to do with the extraction of the wax scales from their 
pockets, but rather that they are organs used in loading the pollen 
from the pollen combs of the hind legs into the corbiculae or pollen 
baskets (Casteel, 1912). Further observations made at that time dis- 
closed the exact method by which the hind legs are instrumental in 
the pollen-loading process and also the way in which the middle legs 
aid the hind legs in patting down the pollen masses. During the 
summer of 1912 additional information was secured, more particu- 
larly that relating to the manner in which pollen is collected upon 
the body and legs of the bee, how it is transferred to the hind legs, 
how it is moistened, and finally the method by which it is stored in 
the hiv0 for future use. In the present paper a complete account will 
be given of the history of the pollen from the time it leaves the flower 
until it rests within the cells of the hive. The points of more par- 
ticular interest in the description of pollen manipulation refer to 
(1) the movements concerned in gathering the pollen from the 
flowers upon the body and legs, (2) the method by which the baskets 
of the hind legs receive the loads which they carry to the hive, and 
(3) the manner in which the bee moistens pollen and renders it suf- 
ficiently cohesive for packing and transportation. 



The hairs which cover the body and appendages of the bee are of 
the utmost importance in the process of pollen gathering. For the 
purposes of this account these hairs may be classified roughly as 
(1) branched hairs and (2) unbranched hairs, the latter including 
both long, slender hairs and stiff, spinelike structures. 

Of these two classes the branched hairs are the more numerous. 
They make up the hairy coat of the head, thorax, and abdomen, with 
the exception of short sensory spines, as those found upon the an- 
tennae and perhaps elsewhere, and the stiff unbranched hairs which 






cover the surfaces of the compound eyes (Phillips, 1905). Branched 
hairs are also found upon the legs ; more particularly upon the more 
proximal segments. A typical branched hair is composed of a long 
slender main axis from which spring numerous short lateral barbs. 
Grains of pollen are caught and held in the angles between the axis 
and the barbs and between the barbs of contiguous hairs. The hairy 
covering of the body and legs thus serves as a collecting surface upon 
which pollen grains are temporarily retained and from which they 
are later removed by the combing action of the brushes of the legs. 
Although, as above noted, some unbranched hairs are located upon 
the body of the bee, they occur in greatest numbers upon the more 
distal segments of the appendages. They are quite diverse in form, 
some being extremely long and slender, such as those which curve 

over the pollen 
Jjoxa baskets, others 

being stout and 
stiff, as those 
which form the 
collecting brushes 
and the pecten 

The mouth- 
parts of the bee 
are also essential 
to the proper col- 
lection of pollen. 
The mandibles 
are used to scrape 
over the anthers 
of flowers, and 
considerable pollen adheres to them and is later removed. The same 
is true of the maxillae and tongue. From the mouth comes the fluid 
by which the pollen grains are moistened. 

The legs of the worker bee are especially adapted for pollen gath- 
ering. Each leg bears a collecting brush, composed of stiff, un- 
branched hairs set closely together. These brushes are located upon 
the first or most proximal tarsal segment of the legs, known techni- 
cally as the palmse of the forelegs and as the plantae of the middle 
and hind pair. The brush of the foreleg is elongated and of slight 
width (fig. 1), that of the middle leg broad and flat (fig. 2), while 
the brush upon the planta of the hind leg is the broadest of all, and 
is also the most highly specialized. In addition to these well-marked 
brushes, the distal ends of the tibiae of the fore and middle legs bear 
many stiff hairs, which function as pollen collectors, and the distal 
tarsal joints of all legs bear similar structures. 

Fig. 1. — Left foreleg of a worker bee. (Original.) 



The tibia and the planta of the hind leg of the worker bee are 
greatly flattened. (See figs. 3, 4.) The outer surface of the tibia is 
marked by an elongated depression, deepest at its distal end, and 
bounded laterally by elevated margins. From the lateral boundaries 
of this depression spring many long hairs, some of which arch over 
the concave outer surface of the tibia and thus form a kind of recep- 
tacle or basket to which the name corbicula or pollen-basket is given. 
The lower or distal end of the tibia articulates at its anterior edge with 
the planta. The remaining portion of this end of the tibia is flat- 
tened and slightly concave, its 




r-Brusk onPlania 

surface sloping upward from 
the inner to the outer surface 
of the limb. Along the inner 
edge of this surface runs a row 
of short, stifp, backwardly di- 
rected spines, from 15 to 21 in 
number, which form the pec- 
ten or comb of the tibia. The 
lateral edge of this area forms 
the lower boundary of the 
corbicular depression and is 
marked by a row of very fine 
hairs which branch at their 
free ends. Immediately above 
these hairs, springing from the 
floor of the corbicula, are found 
7 or 8 minute spines, and above 
them one long hair which 
reaches out over the lower edge 
of the basket. 

The broad, flat planta (meta- 
tarsus or proximal tarsal seg- 
ment of the hind leg) is marked 
on its inner surface by several 
rows of stiff, distally directed 
spines which form the pollen 
combs. About 12 of these transverse rows may be distinguished, 
although some of them are not complete. The most distal row, which 
projects beyond the edge of the planta, is composed of very strong, 
stiff spines which function in the removal of the wax scales (Casteel, 
1912) . The upper or proximal end of the planta is flattened and pro- 
jects in a posterior direction to form the auricle. The surface of the 
auricle is marked with short, blunt spines, pyramidal in form, and a 
fringe of fine hairs with branching ends extends along its lateral edge. 
This surface slopes upward and outward. 
61799°— Bull. 121—12 2 

Fig. 2. — Left middle leg of a worker bee. 
( Ouiginal. ) 






When bees collect pollen from flowers they may be engaged in this 
occupation alone or may combine it with nectar gathering. From 
some flowers the bees take only nectar, from others only pollen; a 

third class of flowers furnishes 
an available supply of both of 
these substances. But even 
where both pollen and nectar 
are obtainable a bee may 
gather nectar and disregard 
the pollen. This is well illus- 
trated by the case of white 
clover. If bees are watched 
while working upon clover 
flowers, the observer will soon 
perceive some which bear pol- 
len masses upon their hind 
legs, while others will continue 
to visit flower after flower, 
dipping into the blossoms and 
securing a plentiful supply of 
nectar, yet entirely neglecting 
the pollen. 

The supply of pollen which 
is available for the bees varies 
greatly among different flow- 
ers. Some furnish an abun- 
dant amount and present it to 
the bee in such a way that 
little difficulty is experienced 
in quickly securing an ample 
load, while others furnish but 
little. When flowers are small 
and when the bee approaches 
them from above, little, if any, 
pollen is scattered over the 
bee's body, all that it acquires 
being first collected upon the 
mouth and neighboring parts. 
Very different conditions are 
met with when bees visit such 
plants as corn and ragweed. The flowers of these plants are pendent 
and possess an abundant supply of pollen, which falls in showers over 
the bodies of the bees as they crawl beneath the blossoms. The 


Fig. 3. — Outer surface of the left hind leg. of a 
worker bee. (Original.) 




supply of pollen which lodges upon the body of the bee will thus 
differ considerably in amount, depending upon the type of flower 
from which the bee is collecting, and the same is true regarding the 
location upon the body of a bee of pollen grains which are available 
for storage in the baskets. 
Moreover, the movements 
concerned in the collection 
of the pollen from the va- 
rious body parts of the 
bee upon which it lodges 
will differ somewhat in 
the two cases, since a 
widely scattered supply 
requires for its collection 
additional movements, 
somewhat similar in na- 
ture to those which the 
bee employs in cleaning 
the hairs which cover its 







Polien (hmk 

A. very complete knowl- 
edge of the pollen-gather- 
ing behavior of the worker 
honey bee may be obtained 
by a study of the actions 
of bees which are work- 
ing upon a plant which 
yields pollen in ahun- 
dance. Sweet corn is an 
ideal plant for this pur- 
pose, and it will be used 
as a basis for the descrip- 
tion which follows. 

In attempting to out- 
line the method by which ^^-- '-'^'^'^'^^^Z:' ^J^^" ''' "' ^ 
pollen is manipulated the 

writer wishes it to be understood that he is recounting that which 
he has seen and that the description is not necessarily complete, 
although he is of the opinion that it is very nearly so. The move- 
ments of the legs and of the mouthparts are so rapid and so many 


members are in action at once that it is impossible for the eye to 
follow all at the same time. However, long-continued observation, 
assisted by the study of instantaneous photographs, gives confidence 
that the statements recorded are accurate, although some movements 
may have escaped notice. 

To obtain poUen from corn the bee must find a tassel in the right 
stage of ripeness, with flowers open and stamens hanging from them. 
The bee alights upon a spike and crawls along it, clinging to the 
pendent anthers. It cf'awls over the anthers, going from one flower 
to another along the spike, being all the while busily engaged in the 
task of obtaining pollen. This reaches its body in several ways. 

As the bee moves over the anthers it uses its mandibles and tongue, 
biting the anthers and licking them and securing a considerable 
amount of pollen upon these parts. This pollen becomes moist and 
sticky, since it is mingled with fluid from the mouth. A considerable 
amount of pollen is dislodged from the anthers as the bee moves over 
them. All of the legs receive a supply of this free pollen and much 
adheres to the hairs which cover the body, more particularly to those 
upon the ventral surface. This free pollen is dry and powdery and 
is very different in appearance from the moist pollen masses with 
which the bee returns to the hive. Before the return journey this 
pollen must be transferred to the baskets and securely packed in them. 

After the bee has traversed a few' flowers along the spike and has 
become well supplied with free pollen it begins to collect it from its 
body, head, and forward appendages and to transfer it to the pos- 
terior pair of legs. This may be accomplished while the bee is 
resting upon the flower or while it is hovering in the air before 
seeking additional pollen. It is probably more thoroughly and rap- 
idly accomplished while the bee is in the air, since all of the legs are 
then free to function in the gathering process. 

If the collecting bee is seized with forceps and examined after it 
has crawled over the stamens of a few flowers of the corn, its legs 
and the ventral surface of its body are found to be thickly powdered 
over with pollen.. If the bee hovers in the air for a few moments 
and is then examined very little pollen is found upon the body or 
upon the legs, except the masses within the pollen baskets. "While in 
the air it has accomplished the work of collecting some of the scat- 
tered grains and of storing them in the baskets, while others have 
been brushed from the body. 

In attempting to describe 'the movements by which this result is 
accomplished it will be best first to sketch briefly the roles of the 
three pairs of legs. They are as follows : 

(a) The first pair of legs remove scattered pollen from the head 
and the region of the neck, and the pollen that has been moistened 
by fluid substances from the mouth. 


(b) The second pair of legs remove scattered pollen from the 
thorax, more particularly from the ventral region, and they re- 
ceived the pollen that has been collected by the first pair of legs. 

(c) The third pair of legs collect a little of the scattered pollen 
from the abdomen and they receive pollen that has been collected 
by the second pair. Nearly all of this pollen is collected by the 
pollen combs of the hind legs, and is transferred from the combs to 
the pollen baskets or corbiculse in a manner to be described later. 

It will thus be seen that the manipulation of pollen is a succes- 
sive process, and that most of the pollen at least passes backward 
from the point where it happens to touch the bee until it finally 
reaches the corbiculse or is accidentally dislodged and falls from the 
rapidly moving limbs. 


Although the pollen of some plants appears to be somewhat sticky, 
it may be stated that as a general rule pollen can not be successfully 
manipulated and packed in the baskets without the addition of some 
fluid substance, preferably a fluid which will cause the grains to 
cohere. , This fluid, the nature of which will be considered laterj 
comes from the mouth of the bee, and is added to the pollen which 
is collected by the mouthparts and to that which is brought into con- 
tact with the protruding tongue and maxillssj and, as will appear, 
this fluid also becomes more generally distributed upon the legs and 
upon the ventral surface of the collecting bee. 

When a bee is collecting from the flowers of corn the mandibles are 
actively engaged in seizing, biting, and scraping the anthers as the 
bee crawls over the pendent stamens. Usually, but not always, the 
tongue is protruded and wipes over the stamens, collecting pollen 
• and moistening the grains thus secured. Some of the pollen may 
possibly be taken into the mouth. All of the pollen wliich comes in 
contact with the mouthparts is thoroughly moistened, receiving more 
fluid than is necessary for rendering the grains cohesive. This 
exceedingly wet pollen is removed from the mouthparts by the fore- 
legs (fig. 5), and probably the middle legs also secure a little of it 
directly, since they sometimes brush over the lower surface of the 
face and the mouth. In addition to removing the very moist pollen 
from the mouth the forelegs also execute cleansing movements over 
the sides of the head and neck and the anterior region of the thorax, 
thereby collecting upon their brushes a considerable amount of pollen 
which has fallen directly upon these regions, and this is added to the 
pollen moistened from the mouth, thereby becoming moist by contact. 
The brushes of the forelegs also come in contact with the anterior 
breast region, and the hairs which cover this area become nloist with 
the sticky exudation which the forelegs have acquired in the process 
of wiping pollen from the tongue, maxillae, and mandibles. 



The middle legs are used to collect the pollen gathered by the 
forelegs and mouthparts, to remove free pollen from the thoracic 
region, and to transport their load of pollen to the hind legs, placing 
most of it upon the pollen combs of these legs, although a slight 
amount is directly added to the pollen masses in the corbiculse. Most 
of the pollen of the middle legs is gathered upon the conspicuous 
brushes of the first tarsal segments or plantaj of these legs. 

In taking pollen from a foreleg the middle leg of the same side is ex- 
tended in a forward direction and is either grasped by the flexed fore- 
leg or rubbed over the foreleg as it is benf downward and backward. 
In the former movement the foreleg flexes sharply upon itself until 

Fig. 5. — A flying bee, showing the manner in which the forelegs and middle legs manipu- 
late pollen. The forelegs are removing wet pollen from the mouthparts and face. The 
middle leg of the right side is transferring the pollen upon Its brush to the pollen 
combs of the left hind planta. A small amount of pollen has already been placed in 
the baskets. (Original.) 

the tarsal brush and coxa nearly meet. The collecting brush of the 
middle leg is now thrust in between the tarsus and coxa of the fore- 
leg and wipes off some of the pollen from the foreleg brush. The 
middle leg brush is then raised and combs down over the flexed fore- 
leg, thus removing additional pollen from the outer surface of this 
leg. The middle leg also at times reaches far forward, stroking down 
over the foreleg before it is entirely flexed and apparently combing 
over with its tarsal brush the face and mouthparts themselves. 
"V\Tien the middle leg reaches forward to execute any of the above 
movements the direction of the stroke is outward, forward, and then 
back toward the body, the action ending with the brush of the leg in 
contact with the long hairs of the breast and with those which spring 


from the proximal segments of the forelegs (coxa, trochanter, 
femur). As a result of the oft-repeated contact of the brushes of 
the middle and forelegs with the breast, the long, branched hairs 
which cover this region become quite moist and sticky, since the 
brushes of these two pair of legs are wet and the pollen which they 
bear possesses a superabundance of the moistening fluid. Any dry 
pollen which passes over this region and touches these hairs receives 
moisture by contact with them. This is particularly true of the free 
dry pollen which the middle pair of legs collect by combing over the 
sides of the thorax. 

The pollen upon the middle legs is transferred to the hind legs in 
at least two ways. By far the larger amount is deposited upon the 
pollen combs which lie on the inner surfaces of the plantse of the 

Fig. 6. — A bee upon the wing, showing the position of the middle legs when they touch 
and pat down the pollen masses. A very slight amount of pollen reaches the corbiculce 
through this movement. (Original.) 

hind legs. To accomplish this a middle leg is placed between the 
plantse of the two hind legs, which are brought together so as to grasp 
the brush of the middle leg, pressing it closely between them, but 
allowing it to be drawn toward the body between the pollen combs 
of the two hind legs. (See fig. 5.) This action results in the trans- 
ference of the pollen from the middle-leg brush to the pollen combs 
of the hind leg of the opposite side, since the combs of that leg scrape 
over the pollen-laden brush of the middle leg. This action may take 
place while the bee is on the wing or before it leaves the flower. 

The middle legs place a relatively small amount of pollen directly 
upon the pollen masses in the corbiculse. This is accomplished when 
the brushes of the middle legs are used to pat down the pollen masses 
and to render them more compact. (See fig. 6.) The legs are used 


for this purpose quite often during the process of loading the baskets, 
and a small amount of pollen is incidentally added to the masses 
when the brushes come into contact with them. A misinterpretation 
of this action has led some observers into the erroneous belief that 
all or nearly all of the corbicular pollen is scraped from the middle- 
leg brushes by the hairs which fringe the sides of the baskets. The 
middle legs do not scrape across the baskets, but merely pat down- 
ward upon the pollen which is there accumulating. 

It is also possible that, in transferring pollen from the middle leg 
of one side to the planta of the opposite hind leg, the middle-leg 
brush may touch and rub over the pecten of the hind leg and thus 
directly place some of its pollen behind the pecten spines. Such a 
result is, however, very doubtful. 


The middle legs contribute the. major portion of the pollen which 
reaches the hind legs, and all of it in cases where all of the pollen 
first reaches the bee in the region of the mouth. However, when 
much pollen falls upon the body of the bee the hind legs coUect a 
little of it directly, for it falls upon their brushes and is collected 
upon them when these legs execute cleansing movements to remove 
it from the ventral surface and sides of the abdomen. All of the 
pollen which reaches the corbiculse, with the exception of the small 
amount placed there by the middle legs when they pat down the 
pollen masses, passes first to the pollen combs of the plantse. 

When in the act of loading pollen from the plantar brushes to the 
corbiculse the two hind legs hang beneath the abdomen with the tibio- 
femoral joints well drawn up toward the body. (See fig. 7.) The 
two plantse lie close together with their inner surfaces nearly parallel 
to each other, but not quite, since they diverge slightly at their distal 
ends. The pollen combs of one leg are in contact with the pecten 
comb of the opposite leg. If pollen is to be transferred from the 
right planta to the left basket, the right planta is drawn upward in 
such a manner that the pollen combs of the right leg scrape over 
the pecten spines of the. left. By this action some of the pollen is 
removed from the right plantar combs and is caught upon the outer 
surfaces of the pecten spines of the left leg. 

This pollen now lies against the pecten and upon the flattened 
distal end of the left tibia. At this moment the planta of the left 
leg is flexed slightly, thus elevating the auricle and bringing the auri- 
cular surface into contact with the pollen which the pecten has just 
received. By this action the pollen is squeezed between the end of the 
tibia and the surface of the auricle and is forced upward against the 
distal end of the tibia and on outward into contact with the pollen 
■ mass accumulating in the corbicula. As this act, by which the left 



basket receives a small contribution of pollen, is being completed, the 
right leg is lowered and the pecten of this leg is brought into contact 
with the pollen combs of the left planta, over which they scrape as 
the left leg is raised, thus depositing pollen upon the lateral surfaces 
of the pecten spines of the right leg. (See fig. 7.) 

Eight and left baskets thus receive alternately successive contribu- 
tions of pollen from the planta of the opposite leg. These loading 
movements are executed with great rapidity, the legs rising and fall- 
ing with a pump-like motion. A very small amount of pollen is 
loaded at each stroke and many strokes are required to load the 
baskets completely. ' 

If one attempts to obtain, from the literature of apiculture and 
zoology, a knowledge of the method by which the pollen baskets 

Fig. 7. — A bee upon the wing, showing the manner in which the hind legs are held during 
the basliet-loading process. Pollen is l)eing scraped by the pecten spines o£ the right 
leg from the pollen combs of the left hind planta. (Original.) 

themselves are loaded, he is immediately confused by the diversity of 
the accounts available. The average textbook of zoology follows 
closely Cheshire's (1886) description in which he says that " the legs 
are crossed, and the metatarsus naturally scrapes its comb face on the 
upper edge of the opposite tibia in the direction from the base of the 
combs toward their tips. These upper hairs * * * are nearly 
straight, and pass between the comb teeth. The pollen, as removed, 
is caught by the bent-over hairs, and secured. Each scrape adds to 
the mass, until the face of the joint is more than covered, and the 
hairs just embrace the pellet." Franz (1906) states that (translated) 
" the final loading of the baskets is accomplished by the crossing over 
of the hind-tarsal segments, which rub and press upon each other." 
Many other observers and textbook writers evidently believed that 
the hind legs were crossed in the loading process. 
61799°— Bull. 121—12 3 


On the other hand, it is believed by some that the middle legs are 
directly instrumental in filling the baskets. This method is indicated 
in the following quotation from Fleischmann and Zander (1910) 
(translated) : 

The second pair of legs transfer the pollen to the hind legs, where it is 
heaped up in the pollen masses. The tibia of each hind leg is depressed on its 
outer side, and upon the edges of this depression stand two rows of stiff hairs 
which are bent over the grooye. The brushes of the middle pair of legs rub 
over these hairs, liberating the pollen, which drops into the baskets. 

A suggestion of the true method is given by Hommell (1906), 
though his statements are somewhat indefinite. After describing 
the method by which pollen is collected, moistened, and passed to 
the niiddle legs he states that (translated) "the middle legs place 
their loads upon the pollen combs of the hind legs. There the sticky 
pollen is kneaded and is pushed across the pincher (a, traverse la 
pince), is broken up into little masses and accumulates within the 
corbicula. In accomplishing this, the legs cross and it is the tarsus 
of the right leg which pushes the' pollen across the pincher of the 
left, and reciprocally. The middle legs never function directly in 
loading the baskets, though from time to time their sensitive ex- 
tremities touch the accumulated mass, for the sake of giving assur- 
ance of its position and size." 

The recent valuable papers of Sladen (1911, 1912, a, h, c, d, and e), 
who was the first to present a true explanation of the function of 
the abdominal scent gland of the bee, give accounts of the process 
by which the pollen baskets are charged, which are in close accord 
with the writer's ideas on this subject. It is a pleasure to be able to 
confirm most of Sladen's observations and conclusions, and weight is 
added to the probable correctness of the two descriptions and in- 
terpretations of this process by the fact that the writer's studies and 
the conclusion based upon them were made prior to the appearance 
of Sladen's papers and quite independent of them. His description 
of the basket-loading process itself is so similar to the writer's own 
that a complete quotation from him is unnecessary. A few differences 
of opinion will, however, be noted while discussing some of the move- 
ments which the process involves. As will later be noted, our ideas 
regarding the question of pollen moistening, collecting, and transfer- 
ence are somewhat different. 


The point at which pollen enters the basket can best be deilermined 
by examining the corbiculse of a bee shortly after it has reached a 
flower and before much pollen has been collected. Within each 
pollen basket of such a bee is found a small mass of pollen, which lies 



along the lower or distal margin of the basket. (See fig 8, a.} It is 
in this position because it has been scraped from the planta of the 
opposite leg by the pecten comb and has been pushed upward past 
the entrance of the basket by the continued addition of more from 
below, propelled by the successive strokes of the auricle. Closer 

Fig. 8. — Camera drawings of the left hind legs of worker bees to show the manner in 
which pollen enters the basket, a, Shows a leg taken from a bee which is just begin- 
ning to collect. It had crawled (.vcr a few flowers and had flown in the air about five 
seconds at the time of capture. The pollen mass lies at the entrance of the basket, 
covering over the fine hairs which lie along this margin and the seven or eight short 
stiff spines which spring from the floor of the corbicula immediately above its lower 
edge. As yet the pollen has not come in contact with the one long hair which rises 
froDj the floor and arches over the entrance. The planta is extended, thus lowering 
the auricle ; i, represents a slightly later stage, showing the increase of pollen. The 
planta is flexed, raising the auricle. The hairs which extend outward and upward from 
the lateral edge of the auricle press upon the lower and outer surface of the small 
pollen mass, retaining it and guiding it upward into the basket ; <•, d, represent slightly 
later stages in the successive processes by which additional pollen enters the basket. 

examination of the region between the pecten and the floor of the 
basket itself shows more pollen, which is on its way to join that 
already squeezed into the basket. 

If the collecting bee is watched for a few moments the increase will 
readily be noted and the fact will be established that the accumulat- 
ing mass is gradually working upward or proximally from the lower 


or distal edge of the corbicula and is slowly covering the floor of this 
receptacle. (See fig. 8, 5, c, and d.) In many instances the suc- 
cessive contributions remain for, a time fairly separate, the whole 
mass being marked by furrows transverse to the long axis of the tibia. 

Sladen (1912, 6) notes the interesting fact that in those rather 
exceptional cases when a bee gathers pollen from more than one 
species of flowers the resulting mass within the corbicula will show 
a stratification parallel to the distal end, a condition which could 
result only from the method of loading here indicated. 

As the pollen within the basket increases in amount it bulges out- 
ward, and projects downward below the lower edge of the basket. 
It is held in position by the long hairs which fringe the lateral sides 
of the b£Dsket, and its shape is largely determined by the form of 
these hairs and the direction in which they extend. When the basket 
is fully loaded the mass of pollen extends laterally on both sides of 
the tibia, but projects much farther on the posterior side, for on this 
side the bounding row of hairs extends outward, while on the anterior 
edge the hairs are more curved, folding upward and over the basket. 
As the mass increases in thickness by additions from below it is held 
in position by these long hairs which edge the basket. They are 
pushed outward and many of them become partly embedded in the 
pollen as it is pushed up from below. When the pollen grains are 
small and the whole mass is well moistened the marks made by some 
of the hairs will be seen on the sides of the load. (See fig. 9, a.) 
These scratches are also transverse in direction and they show that 
the mass has been increased by additions of pollen pushed up from 

Even a superficial examination of a heavily laden basket shows 
the fallacy of the supposition that the long lateral fringing hairs are 
used to comb out the pollen from the brushes of either the hind or 
middle legs by the crossing of these legs over the laterail edges of the 
baskets. They are far from sufficiently stiff to serve this purpose, 
and their position with relation to the completed load shows con- 
clusively that they could not be used in the final stages of the loading 
process, for the pollen mass has completely covered many of them 
and its outer surface extends far beyond their ends. They serve 
merely to hold the pollen in place and to allow the load to project 
beyond the margins of the tibia. 

The auricle plays a very essential part in the process of loading 
the basket. This structure comprises the whole of the flattened 
proximal surface of the planta, except the joint of articulation itself, 
and it extends outward in a posterior direction a little beyond the 
remaining plantar edge. The surface of the auricle is covered over 
with many blunt, short spines and its lateral margin is bounded by 
a row of short rather pliable hairs, branched at their ends. When 


the planta is flexed the auricle is raised and its surface approaches 
the distal end of the tibia, its inner edge slipping up along the pecten 
spines and its outer hairy edge projecting into the opening which 
leads to the pollen basket. (See fig. 8, 5.) With each upward stroke 
of the auricle small masses of pollen which have been scraped from the 
plantar combs by the pecten are caught and compressed between the 
spiny surface of the auricle and the surface of the tibia above it. 
The pressure thus exerted forces the pasty pollen outward and up- 
ward, since it can not escape past the base of the pecten, and directs 
it into the entrance to the corbicula. The outward and upward slant 
of the auricular surface and the projecting hairs with which the outer 
edge of the auricle is supplied also aid in directing the pollen toward 
the basket. Sladen (1911) states that in this movement the weak 
wing of the auricle is forced backward, and thus allows the escape of 
pollen toward the basket entrance, but this appears both doubtful and 
unnecessary, since the angle of inclination of the auricular surface 
gives the pollen a natural outlet in the proper direction. 

If the corbicula already contains a considerable amount of pollen 
the contributions which are added to it at each stroke of the auricle 
come in contact 'with that already deposited and form a part of this 
mass, which increases in amount by continued additions from below. 
If, however, the corbicula is empty and the process of loading is just 
beginning, the first small bits of pollen which enter the basket must 
be retained upon the floor of the chamber until a sufficient amount 
has accumulated to allow the long overcurving hairs to offer it effec- 
tive support. The sticky consistency of the pollen renders it likely 
to retain contact with the basket, and certain structures near the 
entrance give additional support. Several small sharp spines, seven 
or eight in number, spring from the floor of the basket immediately 
within the entrance, and the entire lower edge of the corbicula is 
fringed with veiy small hairs which are branched at their ends. 
(See fig. 3.) One large hair also springs from the floor of the basket, 
somewhat back from the entrance, which may aid in holding the 
pollen, but it can not function in this manner until a considerable 
amount has been collected. 

As the pollen mass increases in size and hangs downward and back- 
ward over the pecten and auricle it shows upon its inner and lower 
surface a deep groove which runs outward from the entrance to the 
basket. (See fig. 9, b.) This groove results from the continued im- 
pact of the outer end of the auricle upon the pollen mass. At each 
upward stroke of the auricle its outer point comes in contact with 
the stored pollen as soon as the mass begins to bulge backward from 
the basket. 

Although the process is a rather delicate one, it is entirely pos- 
sible so to manipulate the hind legs of a recently killed bee that the 




corbiculse of the two legs receive loads of pollen in a manner similar 
to that above described. To accomplish this successfully the operator 
must keep the combs of the plantse well supplied with moistened 
pollen. If the foot of first one leg and then the other is grasped 
with forceps and so guided that the pollen combs of one leg rasp over 
the pecten spines of the other, the pollen from the combs will be 
transferred to the corbiculae. To continue the loading process in a 

proper manner, it is also nec- 
essary to flex the planta of 
each leg just after the pollen 
combs of the opposite leg 
have deposited pollen behind 
the pecten. By this action 
the auricle is raised, com- 
pressing the pollen which 
the pecten has secured, and 
forcing some upward into 
the corbicula. Bees' legs 
which have been loaded in 
this artificial manner show 
pollen masses in their cor- 
biculse which are entirely 
similar in appearance to 
those formed by the labors 
of the living bee. More- 
over, by the above method 
of manipulation the pollen 
appears first at the bottom 
of the basket, along its lower 
margin, gradually extends 
upward along the floor of 
the chamber, comes in con- 
tact with the overhanging 
hairs, and is shaped by them 
in a natural manner. All 
attempts to load the baskets 
by other movements, such as 
crossing the hind legs and 
scraping the plantar combs over the lateral edges of the baskets, 
give results which are entirely different from those achieved by the 
living bee. 


Many descriptions have been written by others of the method by 
which pollen is gathered and moistened. Some of these are indefi- 
nite, some are incorrect, while others are, in part, at least, similar 

Fig. 9. — Inner surface of the right hind leg of a 
worker bee which bears a complete load of 
pollen, a. Scratches in the pollen mass caused 
by the pressure of the long projecting hairs 
of the basket upon the pollen mass as it has 
been pushed up from below ; 6, groove in the 
pollen mass made by the strokes of the auricle 
as the mass projects outvv.ird and backward 
from the basket. (Original.) 


to my own interpretation of this process. A few citations will here 
be given : 

The bee first strokes the head and the proboscis with the brushes of the 
forelegs and moistens these brushes with a little honey from the proboscis, so 
that with later strokes all of the pollen from the head is collected upon these 
brushes. Then the middle-leg brushes remove this honey-moistened pollen from 
the forelegs and they also collect pollen from the breast and the sides of the 
thorax.— [Translation from Alefeld, 1861.] 

In his account of the basket-loading process Alefeld assigns to 
the middle-leg brushes the function of assembling all of the pollen, 
even that from the plantar combs, and of placing it on the corbiculae, 
this latter act being accomplished by combing over the hairy edge of 
«ach basket with the middle-leg brush of the same side. 

It appears probable that the bee removes the pollen from the head, breast, 
and abdomen by means of the hairy brushes which are located upon the medial 
sides of the tarsal segments of all of the legs, being most pronounced upon the 
hind legs. The pollen is thus brought together and is carried forward to the 
mouth, where it is moistened with saliva and a little honey. — [Translation from 
Franz, 1906.] 

Franz then says that this moistened pollen is passed backward and 

Since the pollen of many plants is sticky and moist it adheres to the surface 
of the basket. Dry pollen is moistened by saliva, so that it also sticks. — 
[Translation from Fleischmann and Zander, 1910.] 

Pollen is taken from flowers principally by means of the tongue, but at times, 
also,, by the mandibles, by the forelegs, and middle legs. The brushes of the 
hind legs also load themselves, collecting from the hairs of the body; The pollen 
dust thus gathered is always transmitted to the mouth, where it Is mixed with 
saliva. — [Translation from Hommell, 1906.] 

Sladen considers the question of how pollen is moistened by the 
honey bee, bumblebee (bumblebee) , and some other bees, but does not 
appear to reach definite conclusions. In one of his papers (1912, c) 
he states that the pollen of some plants may be found in the mouth 
cavity and in the region of the mouth, but he reaches the conclusion 
that this pollen is comparatively " dry," using the word in a " rela- 
tive sense." He asserts that " nowhere but on the corbicula and 
hind metatarsal brushes did I find the sticky pollen, except some- 
times on the tips of the long, branched haics on the back (upper) 
edges of the tibiae and femora of the middle legs, and then only 
in heavily laden bees, where it is reasonable to suppose it had 
collected accidentally as the result of contact with the hind metatarsal 

These and other considerations lead Sladen to think that, in the 
case of the bumblebee at least, the pollen " may be moistened on the 
hind metatarsus with the tongue." He states that the tongue of 
the bumblebee is of sufficient length to reach the hind metatarsus 


(planta) and that it might rub over the brushes of the metatarsi 
or be caught between them when they are approximated and thus 
moisten the two brushes simultaneously. However, he has never 
seen the tongue of the collecting honey bee brought near to the hind 
legs, and it appears probable to him that it can not easily reach them. 
" Possibly the middle or front legs are used as agents for conveying 
the honey" (in the case of the honey bee). "In the bumblebee the 
tongue is longer, and it could more easily moisten the hind legs in 
the way suggested." 

In an earlier paper Sladen (1912, a) gjves the following as his 
opinion of the " way in which pollen dust is moistened with nectar," 
although he states that this is one of the points " which still remains 
obscure " : 

The only satisfactory manner In which; it seems to me, this can be done is 
for the tongue to lick the tarsi or metatarsi of the forelegs, which are covered 
with stiff bristles, well suited for holding the nectar, the nectar being then 
transferred to the metatarsal brushes on the middle legs, and from these, again, 
to the metatarsal brushes on the hind legs. The latter being thus rendered 
sticky, the pollen dust would cling to them. The different pairs of legs were 
certainly brought together occasionally, but not after every scrape of the 
hind metatarsi, and their movements were so quick that it was Impossible 
to see what was done. Still, several pollen-collecting bees that I killed had the 
tarsi and metatarsi of the forelegs and the metatarsal brushes of the middle 
and hind legs moistened with nectar, and I think it probable that the moisten- 
ing process, as outlined, is performed, as a rule, during the flight from flower 
to flower. 

Sladen (1912, c) also considers the possibility that the fluid which 
moistens the pollen might be secreted through the comb at the end 
of the tibia, through the tibio-tarsal joint, or from the surface of the 
auricle, but finds no evidence of glandular openings in these regions. 
A suggestion of a similar nature, apparently unknown to Sladen, 
was made by Wolff (1873), who describes "sweat-glands" which, 
he claims, are located within the hind tibia and the planta, and 
which pour a secretion upon the surface of the corbicula and upon 
the upper end of the planta through many minute openings located 
at the bases of hairs, particularly those which arise from the lateral 
margins of the corbicula. Wolff is convinced that the fluid thus 
secreted is the essential, cohesive material by which the grains of 
pollen are bound together to form the solid mass which fills each 
fully loaded basket. He noticed that the mouthparts are used to 
collect pollen, and that some of it is moistened with " honey " or 
" nectar," but he does not consider that the fluid thus supplied is 
sufficient to explain adequately the facility with which the collecting 
bee brings together the scattered gi'ains of pollen and packs them 
away securely in the baskets. Wolff's description of the basket-load- 
ing process itself is strikingly similar to that advocated later by 


The writer is not prepared to deny the possibility that the surface 
of the chitin of the hind legs of worker bees may be moistened by 
the secretion of glands which lie beneath it, but he is convinced that 
any fluid thus secreted bears little or no relation to the cohesion of 
the pollen grains within the baskets. Sections and dissected prepa- 
rations of the hind legs of worker bees show certain large cells which 
lie within the cavity of the leg and which may function as secreting 
gland cells; but similar structures occur in even greater numbers 
within the hind legs of the drone and they are found within the hind 
legs of the queen. 

As has been noted, the extreme moisture of the plantar combs and 
of the tibio-tarsal articulation of the hind leg is readily understood 
when one recalls the manner in which moist pollen is compressed 
between the auricle and the tibial surface above it. 

From the account already given it is evident that, in the opinion 
of the writer, the mouth is the source from which the poUen-jnoisten- 
ing fluid is obtained. It is extremely difficult to determine with 
absolute accuracy the essential steps involved in the process of adding 
moisture to the pollen. In an endeavor to solve this problem the 
observer must of necessity consider a number of factors, among which 
may be noted (1) the location upon the body of the collecting bee 
of " moist " and of comparatively " dry " pollen, (2) the movements 
concerned in the pollen-gathering and pollen-transferring processes, 
(3) the relative moisture of those parts which handle pollen, (4) the 
chemical differences between the natural pollen of the flower and 
that of the corbiculse and of the cells of the hive, and (5) the observer 
must endeavor to distinguish between essential phenomena and those 
which are merely incidental or accidental. 

In the first place it should be noted that the relative dampness of 
pollen within the corbiculse depends very largely upon the character 
of the flower from which the pollen grains are gathered. When 
little poUen is obtained it is much more thoroughly moistened, and 
this is particularly true in cases when the pollen is all, or nearly all, 
collected in the region of the mouth, the forelegs, and head. When 
a bee takes pollen from white or sweet clover practically all of it 
first touches the bee in these regions. It immediately becomes moist, 
and in this condition is passed backward until it rests within the 
baskets. There is here no question of "dry" and "wet" pollen, 
or of collecting movements to secure dry pollen from other regions 
of the body, or of the ultimate method by which such free, dry pol- 
len becomes moist. 

The sticky fluid which causes pollen grains to cohere is f oimd upon 
all of the legs, in the region of their brushes, although the pollen 
combs and auricles of the hind legs are likely to show it in greatest 
abundance, since nearly all of the pollen within each basket has 


passed over the auricle, has been pressed upward and squeezed be- 
tween the auricle and the end of the tibia and the pollen mass above, 
and by this compression has lost some of its fluid, which runs down 
over the auricle and onto the combs of the planta. It is not necessary 
to invoke any special method by which these areas receive their 
moisture. The compressing action of the auricle squeezing heavily 
moistened pollen upward into the basket is entirely sufficient to 
account for the abundance of sticky fluid found in the neighborhood 
of each hind tibio-tarsal joint. As has been noted, the brushes of 
the forelegs acquire moisture directly by stroking over the proboscis 
and by handling extremely moist pollen taken from the mouthparts. 
The middle-leg brushes become moist by cojitact with the foreleg and 
hind-leg brushes, probably also by touching the mouthparts them- 
selves, and by passing moist pollen backward. The hairy surface of 
the breast is moistened by contact with the fore and mid leg brushes 
and with the moist pollen which they bear. 

The problem of the method of pollen moistening is somewhat more 
complicated in the case of flowers which furnish an excessive supply. 
Under such conditions the entire ventral surface of the collecting bee 
becomes liberally sprinkled with pollen grains which either will be 
removed and dropped or will be combed from the bristles and branch- 
ing hairs, kneaded into masses, transferred, and loaded. The ques- 
tion naturally arises whether the movements here are the same as 
when the plant yields but a small amount of pollen which is collected 
by the mouthparts and anterior legs. In the opinion of the writer 
they are essentially the same, except for the addition of cleansing 
movements, executed chiefly by the middle and hind legs for the col- 
lection of pollen which has fallen upon the thorax, upon the abdomen, 
and upon the legs themselves. Indeed it is questionable as to just 
how much of this plentiful supply of free pollen is really used_ in 
forming the corbicular masses. Without doubt much of it falls from 
the bee and is lost, and in cases where it is extremely abundant and 
the grains are very small in size an appreciable amount still remains 
entangled among the body-hairs when the bee returns to the hive. 
Yet it is also evident that some of the dry pollen is mingled with the 
moistened material which the mouthparts and forelegs acquire and 
together with this is transferred to the baskets. 

In all cases the pollen-gathering process starts with moist pollen 
from the mouth region. This pollen is passed backward, and in its 
passage it imparts additional moisture to those body regions which 
it touches, the brushes of the fore and middle legs, the plantse of the 
hind legs, and the hairs of the breast which are scraped over by the 
fore and middle leg brushes. This moist pollen, in its passage back- 
ward, may also pick up and add to itself grains of dry pollen with 
which it accidentally comes in contact. Some of the free, dry pollen 


which falls upon the moist brushes or upon the wet hairs of the 
thorax is also dampened. Some of the dry pollen which is cleaned 
from the body by the action of all of the legs meets with the wet 
brushes or with the little masses of wet pollen and itself becomes wet 
by contact. Pollen grains which reach the corbiculae either dry or 
but slightly moistened are soon rendered moist by contact with those 
already deposited. Little pollen gets by the sticky surfaces of the 
combs of the plantse or past the auricles without becoming thoroughly 

Sladen (1912, c) very aptly compares the mixture of dry pollen 
with wet to the kneading of wet dough with dry flour and suggests 
that the addition of dry pollen may be of considerable advantage, 
since otherwise the brushes, particularly those of the hind legs, 
would become sticky, " just as the board and rolling pin get sticky 
in working up a ball of dough if one does not add flour." The addi- 
tion of a considerable amount of dry pollen gives exactly this result, 
for the corbiculse then rapidly become loaded with pollen mixed 
with a minimum supply of moisture and the brushes reniain much 
dryer than would otherwise be the case. However, if too much dry 
pollen is added the resulting loads which the bees carry back to the 
hives are likely to be irregular, for the projecting edges of the masses 
may crumble through lack of a sufficient amount of the cohesive 
material by which the grains are bound together. 

On the other hand, it does not appear at all necessary to mix much 
dry pollen with the wet, nor do the brushes become sufficiently 
" sticky " from the presence of an abundance of the moistening fluid 
to endanger their normal functional activity. I have observed bees 
bringing in pollen masses which were fairly liquid with moisture, 
and the pollen combs also were covered with fluid, yet the baskets 
were fully and synunetrically loaded. 

Sladen's different interpretations of the pollen-moistening process 
are rather confusing, and it is difficult to distinguish between what 
he states as observed facts and what he puts forward as likely 
hypotheses. He agrees with me in his observation that all of the 
legs become moist in the region of their brushes and also in his sup- 
position that this moisture is transferred to them from the mouth. 
In this moistening process my observations show that the fluid con- 
cerned is passed backward by the contact of the middle-leg brushes 
with the wet foreleg brushes and that the middle-leg brushes in turn 
convey moisture to the plantae as they rub upon them. I am also 
convinced that the wet pollen grains furnish additional moisture to 
the brushes as they pass backward, and this is particularly true in 
the case of the extremely moist surfaces of the auricles and the pollen 
combs of the planta, since here moisture is pressed from the pollen 
upon these areas. The pollen upon the fore and middle leg brushes 
is not always " dry " even in " a relative sense." 



In describing pollen manipulation several writers state that dry 
pollen is picked up by the brushes of the legs and is carried forward 
to the mouth, there moistened (according to some, masticated), and 
is then carried backward by the middle legs for loading. Obviously 
such accounts do not apply to cases in which all of the pollen is col- 
lected by mouthparts and forelegs. Do they apply in cases where 
much pollen falls on the body and limbs ? Without doubt a certain 
amount of this free pollen is brought forward when the middle legs, 
bearing some of it, sweep forward and downward over the forelegs, 
mouthparts, and breast. However, it does not appear to the writer 
that this dry pollen is carried to the mouth for the specific purpose of 
moistening it, or that it is essential to its moistening that it be 
brought in contact with the mouth. Some of it touches the moist 
hairs on the forelegs and breast and is moistened by contact. All 
that remains on the brushes of the middle legs secures moisture from 
these brushes or from -wet pollen which the brushes collect from the 
mouthparts or forelegs. The supposed necessity of carrying forward 
pollen to the mouth for moistening is a delusion. Some is acci- 
dentally brought forward and into contact with the mouth and gets 
wet, but the process is not essential. 

If the pollen which bees transport to their hives has been moistened 
with some fluid substance which causes the grains to cohere, this 
addition should be indicated by differences in the results of an analy- 
sis of pollen from a plant as compared with that found in the cor- 
biculse of a bee which has been working on this plant. For the sake 
of determining this difference and in an endeavor to ascertain, if 
possible, the approximate nature of the added fluid, analyses were 
made of three kinds of pollen, as follows: (1) Pollen collected by 
hand from the corn plant itself; (2) pollen taken from the corbiculaa 
of bees which had secured their supply from corn; (3) pollen stored 
in the cells of the hive. In the first two cases pollen from the same 
species of plant (com) was used. The material from the cells of the 
hive was composed largely of corn pollen, but contained an admixture 
of some other pollens. 

The writer is indebted to Dr. P. B. Dunbar, of the Bureau of 
Chemistry, for the following analyses : 


Com pol- 
len from 


Total solids 

Moisture .' 

Reducing sugar before inversion 


Total reducing sugar after inversion 

Dry basis: 

Reducing sugar 










These analyses show conclusively that a very large amount of 
sugar has been added to the pollen by the time it reaches the cor- 
biculse. Calculated on a dry basis just about twice as much sugar is 
present in the basket pollen as in that from the corn plant. Not only 
is this so, but the additional fact is disclosed that over three times as 
much reducing sugar is present in the corbicular pollen as sucrose. 
This latter result indicates that honey (largely a reducing sugar) 
rather than nectar (containing more sucrose) is the chief sugar in- 
gredient of the corbicular pollen. The additional amount of sugar 
(here again a reducing sugar) in the stored pollen of the hive is 
what might be expected, since it is supposed that the workers add 
honey and possibly other ingredients to the pollen within the 
storage cells. 

The total solid percentages, corn 53.47, corbicula 66.94, stored 
pollen 79.66, also show that the fluid substance which is added is one 
highly charged with solids, a ;;ondition which honey amply fulfills. 

In the descriptions which have been cited of the pollen-gathering 
process in which the mouth is supposed to supply the requisite fluid 
three substances are mentioned: Nectar, honey, and saliva. The 
analyses herein given indicate that reducing sugar is mingled with 
the pollen, and in the case of corn it is indicated that honey is used 
in greater abundance. Without doubt a certain amount of saliva 
also finds its way to the pollen, but the proportion of this substance 
has not been determined. This salivary fluid may have adhesive 
qualities, but this is scarcely necessary, since honey alone is amply 
sufficient for this purpose. 

It appears probable that the fluid which a bee adds to the pollen 
which it is collecting varies somewhat in amount, since the pollen of 
different plants differs considerably in moisture content and that of 
the same plant will differ in this respect at different times. Pollen 
collected in the early morning before the dew has left the plant is 
much more moist than that found upon the same plant later in the 
day, and the grains, if taken when moist, have a natural tendency to 
become aggregated and form small masses. Moreover, this may ex- 
plain the fact that bees make their pollen-collecting trips during the 
morning hours, rather than in the afternoon, although some may be 
seen upon the flowers throughout the whole day. 


When the bee has fully loaded its baskets and before it returns to 
the hive it often spends a little time upon the plant from which it 
has been collecting, occupied with the task of cleaning scattered 
grains of pollen from its body and of patting down securely the loads 
which it has obtained. Upon its return to the hive it hurries within 
and seeks for a suitable place in which to deposit the pollen. Some 


returning bees walk leisurely over the combs and loiter among their 
sister workers, while others appear to be greatly agitated, shaking 
their bodies and moving their wings as though highly excited. 
Many pollen-bearing bees appear eager to receive food upon their 
return to the hive, and they will solicit it from other workers or 
take it from the honey-storage cells. The workers of the hive at 
times take a little of the fresh pollen from the baskets of the laden 
bee, nibbling it off with their mandibles or rasping off grains with 
their tongues. 

If the combs of a colony are examined, stored pollen will be found 
in various parts of the hive. In the brood frames the greatest amount 
is located above and at the sides of the brood and between this and 
the stored honey. Cells scattered through the brQod from which 
young bees have lately emerged may also contain pollen. In the 
outer frames of the hive, where brood is less likely to be found, 
nearly all of the cells may be packed with pollen, or honey-storage 
cells may be found interspersed with those filled with pollen. As a 
rule pollen is not stored in drone comb, although this occasionally 

As the pollen-bearing bee crawls over the combs it appears to be 
searching for a suitable cell in which to leave its load. It sticks 
the head into cell after cell until finally one is located which meets 
its requirements, although it is an open question as to why any one of 
a group should be chosen rather than another. This selected cell 
may already contain some pollen or it may be empty. If partly filled, 
the pollen which it contains is likely to be from the same species of 
plant as that which the bee carries, although different kinds of pollen 
are often stored in the same cell. 

In preparation for the act of unloading the bee grasps one edge 
of the cell with its forelegs and arches its abdomen so that the pos- 
terior end of the abdomen rests upon the opposite side of the cell. The 
body is thus held firmly and is braced by these two supports with the 
head and anterior thoracic region projecting over one of the neigh- 
boring cells. The hind legs are thrust down into the cell and hang 
freely within it, the pollen masses being held on a level with the outer 
edge of the cell, or slightly above it. The middle leg of each side 
is raised and its planta is brought into contact with the upper 
(proximal) end of the tibia of the same side and with the pollen mass. 
The middle leg now presses downward upon the pollen mass, work- 
ing in between it and the corbicular surface, so that the mass is 
shoved outward and downward and falls into the cell. As the pollen 
masses drop, the middle legs are raised and their claws find support 
upon the edge of the cell. The hind legs now execute cleansing move- 
ments to remove small bits of pollen which still cling to the corbicular 



surfaces and hairs. After this is accomplished the bee usually leaves 
the cell without paying further attention to the two pellets of pollen 
although some collecting bees will stick the head into the cell, possi- 
bly to assure themselves that the pollen is properly deposited. It has 
been stated by some (Cheshire, for example) that the spur upon the 
middle leg is used to help pry the pollen mass from the corbicula. 
This structure is in close proximity with the mass while the middle 
leg is pushing downward upon it, but its small size renders difficult 
an exact estimate of its value in this connection. It is certainly true 
that the entire planta of the middle leg' is thrust beneath the upper 
end of the pollen mass, but the spur may be used as an entering 

Pollen masses which have been dropped by the collecting bee may 
remain for some time within the cell without further treatment, but 
usually another worker attends to the packing of the pollen shortly 
after it has been deposited. To accomplish this the worker enters the 
cell head first, seizes the pollen pellets with its mandibles, breaks 
them up somewhat or flattens them out, probably mingles additional 
fluid with the pollen, and tamps down the mass securely in the bot- 
tom of the cell. As is shown by the analyses of corbicular pollen and 
of stored pollen, certain substances are added to the pollen after the 
collecting bee leaves it in the cell. Sugar is certainly added, and it is 
generally supposed that secretions from some of the salivary glands 
are mixed with the pollen after deposition. It appears probable that 
the stored pollen or " beebread " is changed somewhat in chemical 
composition through the action of the fluids which have been added 
to it, either during the process of collection, at the time of packing, 
or later. 


PoUen m^-y be collected by the worker bee upon its mouthparts, 
upon the brushes of its legs, and upon the hairy, surface of its body. 
When the bee collects from small flowers, or when the supply is not 
abundant, the mouthparts are chiefly instrumental in obtaining the 

The specialized leg brushes of the worker are used to assemble the 
pollen, collecting it from the body parts to which it first adheres and 
transporting it to the pollen baskets or corbiculse of the hind legs. In 
this manipulation the forelegs gather pollen from the mouthparts and 
head; the middle legs, from the forelegs and from the thorax; the 
hind legs, from the middle legs and from the abdomen. 

The pollen baskets are not loaded by the crossing over of one hind 
leg upon the other or to any great extent by the crossing of the middle 
legs over the corbiculse. The middle legs deposit their loads upon the 


pollen combs of the hind plantse, and the plantse, in turn, transfer the 
pollen of one leg to the pecten comb of the other, the pecten of one 
leg scraping downward over the pollen comb of the opposite leg. 
(See fig. 7.) A little pollen is loaded directly from the middle legs 
into the baskets when these legs are used to pat down the pollen 
masses. (See fig. 6.) 

Aside from the foregoing exception, all of the pollen which reaches 
the baskets enters them from below, since it is first secured by the 
pecten combs, and is then pushed upward by the impact of the 
rising auricles, which squeeze it against the distal ends of the tibiae 
and force it on into the baskets to meet that which has gone before. 

The long hairs which form the lateral boundaries of the baskets 
are not used to comb out pollen from the brushes of any of the legs. 
They serve to retain the accumulating masses within the baskets and 
to support the weight of the pollen, as it projects far beyond the 
surfaces of the tibiae. 

Pollen grains are moistened and rendered cohesive by the addition 
to them of fluid substances which come from the mouth. Analyses 
show that honey forms a large part of this moistening fluid, although 
nectar and secretions from the salivary glands are probably present 

In the process of pollen manipulation this fluid substance becomes 
well distributed over the brushes of all of the legs. The forelegs 
acquire moisture by brushing over the mouthparts, and they transfer 
this to the hairs of the breast and to the middle-leg brushes when 
they come in contact with them. The middle-leg brushes transmit 
their moisture to the pollen combs of the hind legs when they rub 
upon them. All of these brushes also transport wet pollen which 
has come from the mouthparts and thereby acquire additional mois- 
ture. The auricles and the plantae of the hind legs become particu- 
larly wet from this source, since fluid is squeezed from the wet pollen 
when it is compressed between the auricles and the distal ends of the 
tibiae. Dry pollen which falls upon the body hairs becomes moist 
when brought into contact with the wet brushes or with wet pollen. 

During the process of manipulation pollen passes backward from 
its point of contact with the bee toward its resting place within the 

Pollen which the collecting bee carries to the hive is deposited by 
this bee within one of the cells of the comb. As a rule, this pollen is 
securely packed in the cell by some other worker, which flattens out 
the rounded masses and adds more fluid to them. 



Alefeld, Dr. — Vol. 5, Nos. 15 and 16, Blchstadt Bienen Zeltung. Summarized 

in "Die Blenenzeitung In neuer, gescliicliteter und systematlsche geordneter 

Ausgabe." Herausgegeben vom Sclimid und Kleine: Erste Band, 

Tlieoretischer Theile. 1861. 

Casteel, D. B., 1912. — ^The manipulation of the wax scales of the honey bee, 

Circular 161, Bureau of Entomology, TJ. S. Dept. Agriculture, pp. 15. 
Cheshibe, p. R., 1886. — Bees and bee-keeping ; scientific and practical. Vol. I, 

scientific ; II, practical. London. 
Fleisohmann und Zander, 1910. — Beitrage zur Naturgeschichte der Honigbiene. 
Fbanz, a., 1906. — In "TJnsere Bienen," herausgegeben von Ludwig, A., Berlin. 

pp. [viii]+831. 
HoMMELL, R., 1906. — ^Apiculture, Encyclopgdie Agricola, Paris. 
Phillips, E. F., 1905. — Structure and development of the compound eye of the 

bee. Proc. Acad. Nat. Sci. Philadelphia, vol. 57, pp. 123-157. 
Sladen, F. W. L., 1911. — ^How pollen is collected by the social bees, and the 
part played in the process by the auricle. British Bee Journal, vol. 39, 
pp. 491-493, Dec. 14. 
Sladen, F. W. L., 1912. — (a) Hovr pollen is collected by the honey bee. Nature, 
vol. 88, pp. 586, 587, Feb. 29. 

1912. — (6) Further notes on how the corbicula is loaded with pollen. 

British Bee Journal, vol. 40, pp. 144, 145, Apr. 11. 
1912.— (c) Pollen collecting. British Bee Journal, vol. 40, pp. 164-166, 

Apr. 25. 
1912. — (d) How propolis is collected. Some further notes on pollen- 
collecting. Gleanings in Bee Culture, vol. 40, pp. 335, 336, June 1. 
1912. — (e) Hind legs of the worker honey bee. Canadian Bee Journal, 
vol. 20, p. 203. July. 
WotPF, O. J. B., 1873. — Das Pollen-Einsammeln der Biene. Eichstadt Bienen- 
Zeitung. 29 Jahrg. Nrs. 22 u. 23, pp. 258-270. 


Alefeld on pollen moistening by worker bee 23 

Antenna cleaner of worker bee, figure S 

Auricle of hind planta of worker beCj definition 

figure 1-1 

r61e and action in pollen collect- 
ing 16-17, 19, 20-22 

Basket, pollen. (See Corbicula.) . 

Brusb of foreleg of worker bee, action and rSle in pollen collecting 13 

figure S 

bind leg of worker bee, action and rSle in pollen collecting__.: — 16 
middle leg of worker bee, action and rSle in pollen collecting — 14-16 

figure 9 

Brushes of legs of worker bee, use in pollen collecting 8-9 

Bumblebee, moistening of pollen, views of Sladeu 23-24 

Cheshire on process of loading pollen baskets by worker bee 17 

Comb or pecten of hind tibia of worker bee, definition 9 

figure 77 

role and action in pollen col- 
lecting 16-19 

Corbicula of worker bee, definition 9 

figure 10 

process of loading 15-22 

Corn, sweet, pollen collecting therefrom by honey bee 11-13 

Coxse of worker bee, figures 8, 9 

DuNBAE, Dr. P. B., analyses of com pollen from plant, from corbiculae 

of bees, and from hive cells 28 

Femora of worker bee, figures 8, 9, 10, 11 

Fleisohmann and Zandeb cia process of loading pollen baskets by worker 

bee . IS 

Flowers, variable amounts of pollen from different plants 10-11 

Fbanz on pollen moistening of worker bee 23 

process of loading pollen baskets by worker bee 17 

Hairs, branched, of honey bee, use in pollen collecting 7-8 

fringing pollen basket, function _ 20 

unbranched, of honey bee, use in pollen collecting 7, 8 

HoMMELL on pollen moistening of worker bee 23 

process of loading pollen baskets by worker bee IS 

Honey, use by worker bee for moistening pollen 24,28-29 

Leg, hind, of worker bee, loaded with pollen, figure 22 

Legs, fore, of worker bee, action and r61e in pollen collecting 12, ]3 

hind, of worker bee, action and role in pollen collecting 13, 16-18 

stages in basket-loading process, figure 19 

middle, of worker bee, action and role in pollen collecting 13, 14r-16 

of worker bee, action in unloading pollen 30-31 

structures used in pollen collecting 7-9 




Mandibles of honey bee, action and rdle in pollen collecting 8, 13 

woi'ker bee, use in packing pollen in the cell ♦ 31 

Maxillse of honey bee, action and rSle in pollen collecting 8, 13 

Moistening of pollen by bumblebee, views of Sladen : 23-24 

honey bee 13,22-29 

Mouthparts of honey bee, action and rSle in pollen collecting 8, 13 

Nectar, supposed use by worker bee for moistening pollen 24-29 

Palma of foreleg of worker bee, definition 8 

Pecten of hind tibia of worker bee^ definition 9 

figure , 11 

r61e and action in pollen collecting 16-19 

Planta of hind leg of worker bee, definition 8 

figures 10,11 

structures concerned in pollen collect- 
ing 9 

middle leg of worker bee, definition 8 

Pollen, chemical composition 26 

collecting by worker bee, bibliography 33 

general statement regarding it 11-13 

summary of process 31-32 

corn, from plant, from oorblculae of bees, and from hive cells, 

analyses to determine nature of moistening fluid 28-29 

moistening by bumblebee, views of Sladen 23-24 

honey bee 22-29 

storage in the hive , 29-31 

structures of honey bee concerned in manipulation 7-9 

supply of honey bee 10-11 

unloading process by worker bee 30-31 

Saliva, supposed use by worker bee in moistening pollen 23,29 

Sladkn, observations on process of loading pollen baskets by worker 

bee __ 18, 20, 21 

views as to pollen moistening by worker bee 23-24, 27 

Spur of middle tibia of worker bee, figure 9 

Storing pollen in the hive 29-31 

Structures of honey bee concerned in manipulation of pollen 7-9 

" Sweat glands " of Wolff within hind tibia and t)lanta of worker bee, 

supposed function 24 

Tibia of hind leg of worker bee, modifications and structures for pollen 

collecting 9 

Tibiae of worker bee, figures , 8,9,10,11 

Tongue of worker bee, action and r51e in pollen collecting 8, 13 

Trochanters of worker bee, figures 8,9 

Wax shears or pinchers, so-called, use in loading pollen by worker bee 7 

Wolff on pollen moistening by worker bee 24 

Zandee, Fleischmann and. (See Fleischmann and Zander.) , 


Issued October 4, 1912. 



" L. O. HOWARD, Entomoloeist and Chief of Bureau. 


D. B. CASTEEL/Ph. D., 

Collaborator; Adjunct Professor of Zoology, 
University of Texas. 



L. O. HowAKD, Entomologist and Chief of Bureau. 

C. L. Maelatt, Entomologist and Acting Chief in Absence of Chief. 

E. S. Clifton, Executive Assistant. 
W. F. Tastet, Chief Clerk. 

F. H. Chittenden, in charge of truck crop and stored product insect investiga- 
A. D. Hopkins, in charge of forest insect investigations. 
W. D. HtTNTEB, in charge of southern field crop insect investigations. 
F. M. 'VKebsteb, in charge of cereal and forage insect investigations: 
A. L. Quaintance, in charge of deciduous fruit insect investigations. 
E. F. Phillips, m charge of Bee culture. 

D. M. EoGEES, in charge of preventing spread of moths, field work. 
EoLLA P. Cukeie, in charge of editorial work. 
Mabel Colcoed, in charge of library. 

Investigations in Bee Cultuee. 

B. F.' Phillips, in charge. 

G. F. White, J. A. Nelson, experts. 

G. S. Demhth, a. H. McCkat, N. B. McIndoo, apicultural assistants. 

D. B. Casteel, collaborator. 
Peaele H. Gaeeison, preparator. 

ADDITIONAL COPIES of this publication 
xi- may be procured Irom the Supebintend- 
EST OF DocnuEirTS, Govenunent Fiintiiig 
OfBce, Washington, D. C, at 5 cents per copy 

Circular No. 161. 

Issued October 4, 1912. 

United States Department of Agriculture, 

L. O. HOWARD, Entomologist and Chief of Bureau. 


By D. B. Casteel, Ph. D. 

Collahoratar; Adjunct Professor of Zoology, University of Texas. 


The particular form of bee activity with which this paper deals 
■ is that which results in the removal of the wax scales from the bodies 
of the worker bees and in the application to the comb of the wax 
thus obtained. A detailed presentation of the facts will be given 
and attention called to certain current conceptions of this process 
which are in error. 

Since the bee is a very lively insect it is not surprising that the 
bodily movements upon which some of its activities depend are 
extremely difficult to follow and may easily be misunderstood. All 
of its highly specialized legs may be used at once in the performance 
of some intricate process, and the observer is in need of keenness of 
^ight and patience if he would gain more than an approximate 
understanding' of the parts taken by the several members. 

In the m'ore permanent literature of apiculture and of zoology 
will be found well-written accounts of the habits of bees, accounts 
which are founded upon a large amount of careful observation and 
which represent the work of many students of bees from the time 
of Huber on. As the years have gone by, errors of sight and of 
judgment have gradually been eliminated, so that at the present time 
our knowledge of bee life, so far as it goes, rests upon a fairly 
satisfactory foundation of authenticated facts. Yet many puzzling 
questions are still unanswered, .and some supposed facts may still be 

An examination of a number of bees from an active colony will 
show great variation in the appearance of the wax scales of differ- 

54505°— Clr. 161—12 1 


ent members of. the colony. In some cases no scales can be observed, 
even upon dissection. In others the scales will be found to be ex- 
tremely thick and completely filling the wax pockets. Some bees 
will show scales in two or three pockets and none in others. Many 
of the workers will possess a complete supply of scales, either all 
of about the same thickness or varying considerably in thickness. 
These and other diverse conditions present themselves for explana- 

The present account is particularly concerned with the manipula- 
tion of the wax scales. By what process or series of processes are 
the scales of wax removed from their pockets and added to the 
comb? That the wax which forms the comb is produced by the bees 
themselves, being elaborated within their bodies and given out in the 
form of thin plates or scales, is a fact well known to all students 
of bees; but many differences of opinion have been expressed con- 
cerning the exact method of wax manipulation. It is also well 
known that the workers of the hive perform many duties — build- 
ing the comb, gathering the stores of pollen and honey, caring for 
the brood and the queen, repairing, cleaning, ventilating, and guard- 
ing the hive— and it has been fairly well established that in some 
cases, at least, these duties vary with the age of the individual 
worker, although more accurate information on this point is much- 
to be desired. Dreyling's ^ results, in particular, indicate that bees 
of certain ages are incapable of producing wax, since their glands 
are either undeveloped or atrophied. Do these bees use the wax 
secreted by others, taking it from them, manipulating it, and form- 
ing it into comb? By careful observation bees devoid of wax scales 
or with scales too thin for satisfactory removal may be discovered 
working with the wax. Do these bees procure their wax from other 
workers, or are they merely reworking the wax of the comb? Upon 
each hind leg of a worker bee is located a peculiar pincers-like 
structure long known as the wax shears. Do bees really use this 
instrument in extracting the scales from the pockets, and if so, does 
the owner of the scale perform this operation, or is the scale re- 
moved by another worker? Or may it not be that the wax scales 
drop from their pockets when they reach a suitable thickness, and 
are salvaged by other workers and added to the comb? All of the 
above interpretations of these processes have been advanced by 
various observers. It is the object of this paper to present a true 
account of the manner in which the scales of wax are transported 
from their pockets to the comb and to point out some of the causes 
which lead to diversity in scale number and scale form. 

1 Dreyling, L. 1903. TJeber die waehbereitenden Organe der Honigbiene. Zoologischer 
Anzeiger, Vol. 26. 

Same. — 1905. Die wachbereitenden Organe bei den gesellig lebenden Bienen. Zoolo- 
giscbe Jahrbiicher, Abtheilung fur Anatomie u. Ontogenie d. Thlere, Vol. 22. 



The way in which the wax scales are formed, as secretion products 
arising from the surfaces of the wax plates on the ventral side of 
Ihe abdomen of the workers, has been well described by others and 
Avith apparent accuracy. The accounts of Dreyling embody the 
results of a very considerable amount of work, and will, for the 
present, at least, be taken at their full value. The work of Snod- 
grass ^ upon the anatomy of the wax plates and wax glands may be 
relied upon. Only a brief statement will here be given of the struc- 
ture of these organs and of the manner in which the scales are 

As is well known, wax is produced by the worker bees only. The 
location of the wax-secreting sur- 
faces, or wax plates, may be readily 
determined by an examination of the 
ventral surface of a bee's abdomen. 
By stretching the abdomen somewhat 
it will be seen that each of the last 
four visible sternal or ventral plates 
is divided into two regions: A pos- 
terior projecting edge which is dis- 
tinctly hairy, and a smooth anterior 
half which is usually covered by the 
next preceding plate. This anterior 
region is divided by a median ridge 
into two distinct, irregularly oval 
areas, which thus lie on either side 
of the midventral line. These areas 
are the wax plates, and upon them 
the wax scales are formed. Each one 
of the last four sternal plates bears 
two wax plates, making eight in all. 
(See fig. 1.) 

The glands which secrete the wax 
lie on the floor of the abdomen im- 
mediately above and in contact with the wax plates, and their 
secretion is deposited upon the external surfaces of the plates, exud- 
ing through the many minute pores which perforate the plates. Upon- 
coming in contact with the air the fluid wax hardens, forming a cov- 
ering over the entire outer surface of the plate, which gradually in- 
creases in thickness with the continued addition of wax through the 
pores. In this way the wax scales are produced, and since they are 

1 Snodgrass, E. E., 1910. The Anatomy of the Honey Bee, Bur. Ent, Tech. Ser. 18, 
V. S. Dept. Agr. 


Fig. 1. — Ventral abdominal plates 
of a worker bee dissected to show 
the position of the wax plates. 


molded upon the surfaces of the eight wax plates they correspond to 
them in number and in form. 

In its natural position each wax scale lies between its wax plate 
and the overlapping edge of the next preceding sternal plate. The 
scale thus fits into a little crevice or wax pocket and is well protected 
from injury. If the bee extends its abdomen the rear edges of the 
scales can be seen protruding from their pockets, or if the scales be- 
come very thick they will push the covering sternal plates outward 
and will project from the pockets. 


The problem of wax secretion has been extensively studied by 
Dreyling, who shows that the wax glands differ markedly in struc- 
ture in bees of different ages. In the newly emerged bee the epider- 
mis which underlies the wax plate is composed of epithelial cells 
nearly cubical in form. As the bee grows older these cells become 
elongated and are separated by clear spaces, and when the bee has 
reached the height of its activity as a wax producer these gland cells 
are elongated and show liquid wax stored in the spaces between them. 
When the wax-secreting period is over these cells degenerate, so that 
in sections through the glands of old field bees, or of bees that have 
lived over winter, the layer of cells beneath the wax plates appears 
greatly shrunken, and individual cells can be distinguished by th^ir 
nuclei only. These histological data are given by Dreyling in sup- 
port of the conclusion that the secretion of wax in much more abun- 
dant at a certain period in the bee's life and that old bees and- very 
young bees are, as a rule, incapable of wax production. These con- 
clusions are in harmony with the practical experiences of bee keepers. 


In a study of the behavior during scale removal and wax building 
it is necessary to watch the bees while they are working naturally 
within the hive. To accomplish this, observatory hives are used in 
which glass is substituted for wood in part of the construction. 
Most of the work is done upon colonies in modified nucleus boxes 
(fig. 2). The two sides are removed from each hive and are replaced 
with glass in the form of sliding doors, two to a side, and glass plates 
are fitted to the top. It all cases wooden shades cover both sides and 
top when the bees are not under observation. Although bees are 
somewhat disturbed when light is first admitted to the hive, they ap- 
pear to become accustomed to it and work normally unless the hive is 
left open for too long a period. 

When a hive is well crowded with bees, and when the frames are 
widely spaced, the workers are apt to extend the comb above the level 
of the top bars of the frames until it comes in contact with the glass. 
This gives the observer an excellent opportunity to study the comb 


-ivorkers at close range, and it also obviates the necessity of placing 
glass ends in the hive against which comb might be built. 

Even with the best of arrangements it is difficult to follow some of 
the movements of the workers during the act of scale removal. As 
an aid to vision a Zeiss binocular microscope is used, the tubes being 
removed from the stand and held to the eye after the manner of a 
iield glass. By the use of this instrument a bee appears to acquire 

Pig. 2. — Observatory hive. The sides are fitted with sliding glass doors, and two pieces of 
glass cover the top. The sliding glass doors allow the observer to gain access to any 

. small area of the outer comb without removing the glass from the entire side of the 
hive. Screens of wood cover the glass of the sides and top when the bees are not under 
observation. (Original.) 

the dimensions of a large-sized rat, and the action of its legs and 
mandibles may be followed with great precision. 

For tke sake of later identification many of the bees ace marketl 
by painting different colors on their baclts, and some are numbered. 
Such distinctive marks make it possible to follow the actions of an 
individual bee from day to day. 

The observations here recorded were made during the summer of 
1911 at the apiary of the Bureau of Entomology. 



Fig. 3. — Ventral view of a worker bee In the act of 
romoTing a wax scale. The two middle legs and 
the right hind leg are used for support, while the 
left hind leg removes the scale. (Original.) 

The determination of the exact method by which tUe wax scales 
are removed either comes as the result of prolonged and patient 

watching or is the product 
of good fortune. Long be- 
fore the observer is able to 
decide upon all of the de- 
tails of the process he 
becomes convinced that 
usually the scales are re- 
moved by the bee which 
secretes them and by this 
bee are masticated and 
added to the comb. The 
workers never assist each 
other in the process of re- 
moval, although, as will 
be mentioned later, free 
scales may, in some cases, 
be handled by other 

As a rule, the scales are 
removed while the bee is 
standing on the comb or its support, and the wax thus obtained is 
applied to the comb near the place where it is removed. Since the 
whole process of removal takes place beneath the worker's body it 
can be observed most 
satisfactorily when 
the bee is seen from 
the side or when it is 
building comb against 
a glass plate. 

The posture of a 
bee in the act of re- 
moving a scale is 
rather characteristic 
and is at once recog- 
nized by one familiar 
with it. Immediately 
before the scale is to be 
removed the bee may 
be busily engaged upon the surface of the comb, plying with its man- 
dibles the wax of the scale last extracted or reshaping and polishing 
wax already deposited, its whole body somewhat agitated, moving 

Fig. 4.- 

-Side view of a worker in the same posture as 
that shown in figure 3. (Original.) 


backward and forward or from side to side as it adapts its position to 
the wox-k in hand. Suddenly its body becomes very quiet. The fore- 
legs and mandibles are raised from the comb, and the head is held 
with the face inclined tow- 
ard the comb. The hind 
leg of one side is now 
raised, and its flattened 
first tarsal segment or 
planta is slipped alpng the 
ventral surface of the ex- 
tended abdomen and comes 
in contact with the pro- 
truding wax scales of the 
corresponding side (figs. 
3 and 4). The weight of 
the bee is now supported 
upon three legs; upon the 
middle leg of the side 
from which the scale is to 
be removed and upon the 
middle and hind legs of 

Fig. 5. — Ventral view of a worker bee showing tlie 
position of the wax scale Just before it is grasped 
by the forelegs and mandibles. The scale is still 
adhering to the spines of the pollen combs. The 
bee is supported upon the two middle legs and 
upon the hind leg which is not removing the scale. 

the 'other side. The first 
tarsal segment of the leg 
which is to • remove the 
scale is now pressed firmly 
against the abdomen, and the edge of a protruding scale becomes 
engaged with it. Steady, continuous pressure is now exerted both 
against the abdomen and toward the rear, with the result that the 

scale is drawn out of its 
pocket but remains- at- 
tached to the leg which 
removed it. The hind 
leg bearing the scale is 
now quickly flexed tow- 
ard the thorax and 
head, thus carrying the 
scale forward under the 
body of the bee and 
placing it in a position 
where it may be readily 
grasped by the forelegs 
or the mandibles (figs. 5 and 6). Sometimes the scale is appar- 
ently removed from the hind leg by the mandibles alone, but usually 
the forelegs aid in this process and also manipulate the scale while 

Fig. 6. — Side view of a worker bee in the same posture 
as that shown in figure 5. (Original.) 



the mandibles are masticating it. After the scale has been thor- 
oughly masticated the wax is applied to the comb. 


A point of particular interest in the process of wax scale removal 
is that which deals with the manner in which the scale is grasped by 

the hind leg which removes it. As is 
well known, each hind leg of the worker 
bee bears a pincerslike structure — the 
so-called wax shears — located at the 
juncture of the tibia and the flattened 
first tarsal segment or planta (fig. 7). 
According to the statements of numer- 
ous writers, the wax scales are grasped 
between the edges of the supposed 
pincers formed by the pecten above and 
the auricle below, and are either snipped 
off or are held by the jaws of the 
pincers and thus drawn from the pock- 
ets. Cowan's^ account may be given 
as typical of others which are current 
in the literature of apiculture and of 

The articulation of the tibia and planta 
being at the anterior angle, and the absence 
of the spur on the tibia (which only the 
honey bee does not possess) give the pecten a 
freedom of action it would not otherwise have 
and enable it to be used together with the 
auricle on the planta, which is quite smooth, 
as a true pair of pincers, and as an instru- 
ment for laying hold of the thin flakes of 
wax, and for bringing them forward to be 
transferred by the' other legs to the jaws for 

As a matter of fact, the wax shears 
have nothing whatever to do with the 
removal of the wax scales. They per- 
form an entirely different function, be- 
ing concerned with the gathering of 
pollen in a manner to be described in a 
future paper. 

In coming to the above conclusions the writer was first convinced 
that the so-called wax shears are not -used in removing scales by 
noting that the position of the tibio-tarsal joint at the time of scale 



FiG. 7. — Inner surfape of the left 
hiifa leg of a worker bee, show- 
ing the position of a wax scale 
Immediately after it has been re- 
moved from the wax pocket. The 
scale has been pierced by seven 
of the spines of the pollen combs 
of the first tarsal segment or 
planta. The jaws of the so- 
called wax shears or pincers are 
formed by the pecten spines 
above and the surface of the 
auricle below. (Original.) 

1 Cowan, T. W., " The Honey Bee," 2d ed., London, 1904. 


removal is such as to make it impossible for the pincerslike crevice 
to grasp the scale. Moreover, the open jaws of the shears point lat- 
terally and away from the scales rather than toward them, nor, 
indeed, is it possible for the shears to grasp even the projecting edges 
of any of the ventral or lateral body plates and thus steady or guide 
the leg as it seeks contact with the scales. 

The transverse rows of spines upon the planta, called pollen combs, 
and not the wax shears are instrumental in the removal of scales, 
Snodgrass (1910), in discussing the anatomy of the hind leg and 
its functions, states that the wax is " poked out of " the " pockets 
by means of the spines on the feet " — " with the ordinary hairs or 
spines of the tibiae or tarsi," and the same general conclusions were 
reached independently by the writer, but with this exception; only 
the spines of the first tarsal segment (planta) function in this 
manner, and usually only certain large spines in the rows at the dis- 
tal end of this segment. 

It is exceedingly difficult to capture a bee at the very moment at 
which the scale is being drawn from its pocket and before it has been 
carried to the mouth, and even if this is accomplished the captive is 
very likely to drop the scale from the hind leg in its struggles to 
escape. If, however, one is successful, the scale-removing leg will 
show the little wax scale adhering to the distal end of the inner 
surface of the first tarsal segment, being pierced in several places 
by the strong spines which project from the lower rows of the pollen 
combs. (See fig. 7.). 

It can also be shown experimentally that this method of remov- 
ing the wax scales is entirely possible, for if the hind tarsus of a bee 
is mounted upon a small stick and is gently rubbed along the ventral 
side of a fully extended dead bee's abdomen, holding it in such a- 
position that the pollen combs brush over the projecting edges of the 
scales, one of the scales will probably be removed and will be seen 
adhering to the spines in the manner above described. 

In any hive where comb is being constructed rapidly many free 
scales will be found upon the bottom board and upon the lower bars 
of the frames. If these scales are examined microscopically some 
will be found without marks upon them, having, evidently been 
loosened from their pockets accidentally during the movements of 
the workers over the comb and around the hive. Others will show 
certain marks and scratches upon them, indicating that they were 
voluntarily removed from the pockets, and in some cases they may 
bear the marks of the mandibles, showing that they were dropped 
during the process of mastication. Most of the scales which are 
marked at all are indented with several small punctures showing the 
places where the spines of the pollen combs have pierced them. 
These scars are exactly similar in appearance to those on the scale 


shown in figure 7. Such free scales are not marked as they would be 
had they been extracted by such a structure as the so-called wax 

So far as can be determined there does not appear to be any regu- 
lar order for the removal of scales. One may be taken from the 
left side and then one from the right, or the bee may remove two or 
three from one side in succession. An attempt to remove a scale is 
by no means always successful, the worker often trying first one side 
and then the other, pressing the pollen combs against the more ante- 
rior scales and running them down to the most posterior, until at 
last a scale is impaled upon the spines or the bee discontinues its 


When a scale has become attached to the spines it is transferred to 
the mouth with great rapidity, so swiftly, in fact, that the eye can 
scarcely follow the action. This is not surprising, for it is necessary 
only to flex the leg toward the head to bring the scale in close con- 
tact with the forelegs and mandibles. The leg is rotated through 
the arc of a circle, downward, forward, and upward, while at the 
same time the head is slightly turned under to receive the scale. 
The process of mastication is more prolonged. It is usually sup- 
posed that the pure wax of the scale differs in chemical composition 
from the wax of the comb, this change being accomplished during 
mastication, by which process the wax is mixed with saliva, becomes 
translucent rather than transparent, changes somewhat in color, and 
becomes more pliable. 

The behavior of a bee upon receiving a wax scale at its mouth is 
subject to considerable variation. On some occasions the scales are 
apparently manipulated by the mandibles alone, while at other times 
the forelegs are brought into requisition and assist the mandibles. 
When a scale is thin and small and has been firmly grasped by the 
mandibles little assistance is needed from the legs. But if a 
scale of medium or extra thickness is presented, or if the mandibles 
do not hold it securely and it is in danger of falling from the mouth, 
the two forelegs are used to great advantage in readjusting the scale 
and in so holding it that the mandibles may be applied to it most 
advantageously. If a scale is small and thin, it may be masticated 
entirely before any wax is applied to the comb ; but if of considerable 
size a portion only may be prepared, this deposited upon the comb, 
and then the remainder treated in a similar manner. 

As a rule the wax which is deposited upon the comb by the pro- 
ducing bee is first subjected to the action of the mandibles and 
mixed with saliva. Such, however, is not always the case, for some 
bees appear to be " careless " and will mingle small unchewed por- 
tions of scales with the masticated wax. Indeed, it is not uncom- 


mon to find nearly perfect scales mixed with the wax of a newly 
made comb. The masticated wax itself is spongy and flaky when 
it is deposited by the producmg bee and will later be reworked, 
thereby gaining greatly in compactness and smoothness. 

The entire process of the removal of one scale, its mastication, and 
the application of the wax to the comb is completed in about four 
minutes, only a very small portion of this interval being consumed in 
the work of extracting the scale from its pocket and passing it to the 
mouth, except, in cases in which scales appear to be removed with 


When wax scales are voluntarily removed they are taken off by 
the bee which secretes them and in the manner above described. 
Many, however, are accidentally detached, being loosened from their 
pockets by movements of the abdomen, incidental cleansing move- 
ments of the legs, or by contact with objects both within and without 
the hive. Such scales, and also those which are dropped in the 
course of transference from the wax pocket to the mouth, may or 
may not be recovered later and added to the comb. Since old wax 
is used over and over again in the rebuilding of comb, it is but 
natural to expect that scattered scales would likewise be utilized by 
the colony and not be allowed to go to waste, and it is probably true 
that such is usually the case. Yet there appears to be no concerted 
action among the workers to salvage such particles of wax, ho class 
of comb workers whose duty it is to pick such material from the 
bottom board of the hive and carry it to the comb. Scales which 
drop are likely to remain for a long time, and some may even be 
carried out through the entrance with waste material. If, however, 
scales accidentally dislodged or voluntarily removed fall on the 
comb among the comb workers they are often noticed by them, picked 
up, masticated, and built into the comb. If a scale slips from the 
pollen combs or is fumbled by the bee before being grasped by the 
mandibles, it is seldom recovered by the worker to which it belongs 
unless it falls very near her or she stumbles upon it accidentally. 


Although a bee endeavors to remove an entire wax scale at one 
operation, the attempt is not always successful. A scale that has 
become very thick is difficult of removal, particularly so if the outer 
edge is broken or beveled. When the bee applies its pollen combs to 
such a scale the spines may fail to get a hold upon the wax, or they 
may not become sufficiently well fixed in it to make possible the re- 
moval of the entire scale. Instead of this, shreds and small pieces of 
wax are torn off and remain sticking to the bristles of the pollen 


combs. These may be entirely disregarded by the bee, or they may 
be cleaned off by scraping the combs together, the shreds of. wax 
dropping to the bottom of the hive. More usually, however, if a 
worker is actively engaged in the task of adding to the comb these 
bits of wax will be carried forward to the mouth, masticated, and 

In one case which came under observation a worker had removed 
all of its wax scales except a very large, thick one which was evi- 
dently sticking tightly in its pocket. Repeated efforts were made by 
the bee to accomplish the extraction of this scale, but with only 
partial success, since the main portion of the scale remained in the 
pocket. But as the result of its efforts the bee succeeded in beveling 
off the entire projecting edge of the scale, rasping it off bit by bit 
and carrying the small pieces forward to the mouth, masticating 
them, and dispositing the wax upon the comb. 


The presence of well-developed scales protruding from the pockets 
of a worker does not necessarily indicate that this individual will 
shortly add this wax to the comb, even though the colony may at the 
time be producing comb at a rapid rate. Such a bee may be working 
upon the comb as a molder of wax rather than as a producer. One 
who is intent upon a study of the process of scale removal will often 
be disappointed after following for a time the movements of a 
worker 'that is evidently manipulating wax and which shows the 
protruding edges of scales beneath its abdomen, for the wax with 
which it is working is being picked up, little by little, from the comb 
and comes not from its own body. This reworking of wax is one of 
the most characteristic features of comb construction, for it goes on 
continually while new comb is being produced, and it is, of course, a 
necessary process in the reconstruction of old comb. 

The claim has been made by several investigators and writers that 
the bees which sculpture the wax are not at the same time concerned 
with its secretion and deposition— that there are producing bees 
and building bees. In a sense this is true, but not entirely so. With- 
out doubt many active comb workers are, at the time, nonproductive, 
for the wax glands of many are not functionally active. The re- 
sults of Drey ling would indicate that the old bees, at. least, might be 
considered as falling in this class, and the direct observations of the 
writer lead to the conclusion that old bees devoid of wax scales per- 
form a considerable share of the labor of reworking newly deposited 
wax and of shaping and polishing the cells of the comb. 

However, as noted above, bees with well-developed wax scales 
often busy themselves with wax working rather than with produc- 
tion. Moreover, a bee that is removing its scales may discontinue 


this work and give its attention to the molding of wax laid down 
by others. This may occur immediately- after a worker has removed 
the last of its scales, or the bee may turn to sculpturing while several 
scales yet remain in the pockets. It is thus evident that the produc- 
ing bee may also be a worker of wax produced by others and that 
nonproductive bees do not monopolize the work of sculpturing and 
polishing the comb. 


As is well known, the wax produced by the worker bee occurs in 
the form of scales, eight in number, which appear upon the surfaces 
of the eight wax plates. These wax plates are located upon the 
last four visible ventral plates of a worker bee's abdomen. The wax 
is secreted by glands which lie upon the inner surface of each wax 
plate. The liquid wax exudes through pores which perforate the 
wax plates, and it hardens to form the scales as it comes in contact 
with the air. 

Unless accidentally dislodged the wax scales are always removed 
and manipulated by the bee which secretes them. 

In the process of removal the scale is not grasped by the so-called 
wax shears, but it is pierced by a few of the stiff spines en the distal 
end of the first tarsal segment of the hind leg and is then drawn 
from its pocket and remains adhering to these spines until removed 
for mastication. 

By flexing the hind leg the scale is brought forward beneath the 
bee's body and into proximity with the mouth. In the process of 
mastication the forelegs usually aid the mandibles by holding the 
scale in an advantageous position. 

No definite sequence is observed by the bee in the order in which 
it removes its scales. 

As a riile entire scales are removed at one operation, although it 
sometimes happens that a thin scale is broken in extracting it from 
its pocket or an extremely thick one is gradually beveled off by the 
continued rasping of the pollen combs. 

Scales which are removed accidentally or which are dropped 
during manipulation may be recovered later and built into the comb, 
but the recovery of free scales is usually not accomplished by the 
bee which secreted them. 

Bees which are producing wax may also rework the masticated 
wax laid down by others. Producing bees may turn to the work of 
building and sculpturing the comb either before all their scales 
are removed or immediately after this has been accomplished.