“Leol
Insecta

THE EMBRYOLOGY OF THE
_ HONEY BEE

BY

~ JAMES ALLEN NELSON, Ph.D.,

Expert, Bee Culture Investigations, Bureau of Entomology
U. S. Department of Agriculture

4 ) v6
Vv. [ 7
PRINCETON UNIVERSITY PRESS | K | Ku
y|

PRINCETON

LONDON: HUMPHREY MILFORD
OXFORD UNIVERSITY PRESS

1915

PREFACE

The good bee keeper is he who is interested not only in those
things which have to do directly with the production of honey |
but to whom everything pertaining to honey bees has a deep
interest. This is shown by the fact that the anatomy of the adult
bee has been much studied by practical bee keepers. Aside from
the fact that from the egg there emerges in three days a small
white larva, no knowledge of the wonderful changes which occur
in that small compass is available, while in comparison with the
changes which occur, the rather fixed structure of the adult insect
seems simple. In this paper is presented to the beekeeping pub-
lic, as well as to those whose interests are more scientific, the
most thorough account of the complex development of the bee
egg yet published and to those interested in bees no apology for
investigations of this kind is needed. It is of interest to the bee
enthusiast, for, while possibly he may not fully appreciate all
the details discussed, he will assuredly want to take such facts
as his training will permit.

From another standpoint work of this kind is needed. All
practical work with bees rests on a foundation of bee activities,
to an extent usually not recognized. This is because with few
other animals with which man deals is it so imperative that the
normal behavior be followed. A study of bee behavior can
progress no farther than our knowledge of the structure of the
bee has gone and anatomy therefore becomes indirectly a vital
thing to the bee keeper. Adult structure cannot be adequately
understood without a study of comparative anatomy and espe-
cially of development. While the developmental studies are
therefore several stages removed from the practices of the
apiary, they are nevertheless of importance and the bee keeper
will welcome this addition to the foundations of his chosen in-
dustry.

E. F. PHI.ips,
In Charge, Bee Culture Investigations.

yi
‘ fh, ;
4
t
te’
{ i
‘
i i
; ay z
| ett
_ 228 Geer 1 at nye * 7 a} “ - ae
> 7 ‘Fa i e eyed ' Ae
B eevee ne . sah Sob! (2 x ¥ ah i ,

http://www.archive.org/details/embryologyofhone00 A

TABLE OF CONTENTS

PAGE

PREFACE ili
Be I IROVIEW 6 oso oe 58s bs ce ds bo alew etnies I
Di; WOORGAEIZATION OF THE EGG... oe ce eee 4
III. CLEAVAGE BEES Re Pa ira ao Ghar etarbrs aye ath ated opp ees 16

IV. FoRMATION AND COMPLETION OF THE BLASTODERM... 27
se MEME SAVERS cai ocig cose eles ode ees bea oeeas 43

I. Formation of the Mesoderm................ 43

2. Formation of the Rudiments of the Mid-intestine 56

VI. Tue AMNION AND THE CEPHALO-DORSAL Bopy...... 82
SMMC TTATAIOES cig. 8 oS SL eva Moen ek bas eemahe 82
@. the Cephalo-dorsal Body............5..-08- 88

VII. GENERAL ACCOUNT OF THE DEVELOPMENT OF THE
EMBRYO, WITH ESPECIAL REFERENCE TO THE Ex-
EMOTE 03.5.) ti gis Liat h aig Medial dio ip 08-9 9 alaee 4 95

1. Changes in the Form of the Egg up to Stage

Reg Oa ee OURS CANIS Sis Ao eaten 95

2. The Development of the Embryo........... 97
EIMORTACION <5 5) 600 vic ss le has ok bo ble «aie curate 105

ee UeEMRRVOUS SYSTEM o.oo. 65 eect ene els owe eaialei 113
Prrrmemiyenatched Larval. .is.c6s.sise sie e osebs 113
ao. cons od Be eis Ghia d Gk SiMte we aiel¥ 124

Bo be Ventral ord ir a 63h ws 5 00 8 ace 127

TX.

B: The Brags aoc ty we tate het eet ee 142

C. The Stomatogastric System..... Benn 157

D... Neurilemimay 020.0 sale eerie sak 160

EL. Cotpata (ARG sie eae awa aca 161

F.. Degeneratingr ) Cells: 35)... eck eas 164.
TRACHEAL SYSTEM, ENDOSKELETON AND Hypo-

DERMIS: 60:5 (cc¢sScsrec 0S co +. ¢)el arse Pier eeinie xe Wee 167

1. Tracheal Systems (av arodis ve vies a eed eas 167

2. Tentorium and Mandibular Apodemes...... 175

@ Hypoderimias 6903 64 wld iia eae es 179

THE OENOCYTES «sy <5. so eeeeeak, + ke ee eee 181

Muscies, Fat Bopy AND CircuLATORY SySTEM.. 189

SEX , ORGANS—OVARIES ¢:0!5 64/6 5,sr0 6 does yee esas ou 213
ATAMBNTARY (CANAD wisi, 2 ds ation 0 bnlenie idee ae ates 220
Your AND Youn CEELS:. 4 ne nees Haieitcan es wee 230
DURATION AND RATE OF DEVELOPMENT........... 245
TECHNIQUE « a9 ey Eds Vindin-w cian Wi aeess SiMe tele 251
DUMMARY 6 | 5 iiy's die da soca eteen She aie w MIR ee meets eee 253
Last...oF ABBREVIATIONS 5 <5. « :\<s Gh Glan a Sean hae 262
BIPLIOGRAP MY [yh cee hel SENS Roles eae ee 265
ENDER ooo Sa 8 dive aon Wace soar bhe bid A Ee ie a 273

I

HISTORICAL REVIEW

The first recorded observations on the embryology of the
honey bee are those of the late Professor Weismann, transmitted
by him in a letter to the famous Russian investigator, Prof. Elias
Mecznikow (Metschnikoff), who published these in his Em-
bryologishe Studien an Insekten (1866, pp. 489-490). Weis-
mann’s statement is as follows: “In the bee a germ layer is
formed, which however does not become part of the embryo,
but which soon separates from the yolk and becomes an amnion-
like envelope. This at first remains in connection with the yolk
at the poles, and becomes entirely free only at a later period,
when the yolk has become transformed to the true embryo. It
is clear that this amnion-like envelope is the embryo, from which
there then arises by metagenesis that which we term the bee
embryo.” These brief and remarkable observations were soon
followed by two important papers. The first of these was that
of Dr. Otto Biitschli, published in 1870, entitled “Zur Entwick-
lungsgeschichte der Biene.” This paper comprises 45 (8vo)
pages of text and four double plates. Biitschli studied only
the living eggs. The features of the development visible from
the exterior were quite accurately described and figured, but the
observations of this noted zoologist were naturally limited by
his method, and while he succeeded in discerning correctly the
origin and development of certain internal parts, such as the
tracheal system, a systematic and detailed account of the origin
of the germ layers and of the details of the organogeny was in
the nature of the case impossible.

In the next year (1871) appeared an extended memoir by
the Russian embryologist Kowalevski, entitled “Embryologische
Studien iber Witirmern und Arthropoden.” It comprises seventy
pages of text (4to) illustrated by twelve plates. This paper is
a veritable landmark in the domain of insect embryology. Kow-
alevski made use of a method new at that time, namely that of
cutting sections of the tissue, previously fixed and embedded in

I

2 THE EMBRYOLOGY OF THE HONEY BEE

paraffin. By the application of this method Kowalevski was able
greatly to extend the boundaries of the knowledge relating to the
development of insects, and his memoir may be said to inaugu-
rate the period of modern research in this field. In the part
relating to the arthropods are included nine pages devoted to the
development of the honey bee, illustrated by thirty figures cover-
ing plate XI and a half of plate XII. Although that part of the
section devoted to the honey bee, which records observations on
the living egg, covers much the same ground as that covered by
Biitschli in the previous year, nevertheless by the study of sec-
tions Kowalevski was able to add many important facts regard-
ing the formation of the blastoderm, the origin and fate of the
germ layers, and the development of the nervous system. The
figures are excellent, and although small, are sharp and clear,
while the text is written in a condensed but lucid style.

Owing probably to the excellence of the observations just de-
scribed, over ten years elapsed before the egg of the honey bee
was again made the subject of scientific investigation. In 1884
a paper by the Italian zodlogist Battista Grassi appeared in a
relatively obscure journal—Atti dell’ Academie Gioenia di scienzi
naturali in Catania—, and according to Carriére (1897) re-
mained for a time almost unnoticed. This paper covers seventy-
seven pages of text (4to), and is illustrated by 252 figures cover-
ing ten plates. This is the fullest, and in fact the only complete
account of the embryology of the honey bee ever published. Like
his predecessors Grassi studied the living bee, but he also made
free use of microtome sections. The text consists of an intro-
duction including a brief review of literature, twelve numbered
chapters, each descriptive of the development of a tissue or sys-
tem of organs, and in addition a chapter discussing the signifi-
cance of the facts recorded. This most excellent paper is scarcely
open to criticism, when judged by the standard of contempo-
raneous papers, and allowing for the relative crudity of the
histological technique of that time. Indeed the correctness of
most of the facts recorded are beyond question. Nevertheless, —
judged by modern standards, the text seems lacking in complete-
ness, while most of the figures appear crude and diagrammatic,
many of them losing much of their value by their small size.
In addition the correctness of some of Grassi’s statements are
open to question.

THE EMBRYOLOGY OF THE HONEY BEE 3

In this connection it will be necessary merely to mention
Blochmann’s paper (1889), since it relates only to the matura-
tion of the egg. This may also be said of Petrunkewitsch’s paper
(1901). A second paper by this investigator (1903), “Das
Schicksal der Richtungskorper im Drohnenei” however contains
some data and figures relating to the early development of the
egg of the honey bee. These will be mentioned later. The recent
paper by Nachtsheim (1913) should also be mentioned here.
Although this is also concerned principally with the fertilization
the maturation of the egg of the honey bee, nevertheless it
contains a number of excellent figures of cleavage cells together
with a considerable amount of data regarding their cytological
features.

The only other paper of recent date devoted to the embryology
of the honey bee is that of Otto Dickel (1903) entitled “Ent-
wicklungsgeschichtliche Studien am Bienenei,” comprising forty-
six pages (8vo) illustrated by forty-six text figures and two
double plates. This paper is extremely limited in its scope,
describing only the early development up to but not including
the formation of the germ layers. It was submitted as a thesis
for the degree of Doctor of Philosophy in the University of
Munich, and was produced under the supervision of Prof. Oscar
Hertwig of that institution. It was apparently written with one
end in view, namely to demonstrate that in the honey bee the
mid intestine arises from the yolk cells (“entoderm’’), and loses
much of its scientific value by its ill-concealed attempt to arrive
at a predetermined conclusion.

This closes the list of papers descriptive of the embryology of
the honey bee, but in this connection one paper should be noticed
which is of the highest value to all students of insect embryology,
especially that of the Hymenoptera. This is the beautiful memoir
of Carriere and Birger, “Die Entwicklungsgeschichte der Mauer-
biene (Chalicodoma muraria Fabr.) im Ei.” This consists of
165 pages (4fo) accompanied by one single plate and eleven
double plates, nine of which are colored. This work covers the
entire development of the egg of the mason bee, from the com-
mencement of cleavage to the hatching of the larva, and it stands
alone as the most complete account of the embryology of a
single insect.

IT

ORGANIZATION OF THE EGG

In form the egg of the bee approximates a long cylinder; one
end, however, being slightly larger than the other, and both
having a smoothly rounded hemispherical contour. The egg is
gently curved in its long axis, so that in profile one side is seen
to be decidedly convex, the other slightly concave (Fig. I). By
reflected light the egg appears pearly white, by transmitted light
it is seen to be translucent, with a saffron tinge.

In length the eggs vary from about 1.53 mm. to 1.63 mm.
(0.059-0.063 inch). The larger end at its broadest point has an
average diameter of about 0.317 mm. (0.0122 inch), or approxi-
mately one-fifth of the total length.

Both the differing size of the two ends and the marked bi-
lateral symmetry of the egg, expressed by the curvature of its
long axis are common among the eggs of insects in general and
always bear a direct and constant relation to the position of the
future embryo. In the egg of the honey bee, with reference to
the parts of the embryo, the larger of the two ends is cephalic
(anterior), the opposite caudal (posterior); the convex side,
ventral, the concave, dorsal. The position of the embryo is
therefore predetermined. Hence the terms cephalic pole, caudal
pole, dorsal and ventral sides may be applied directly to the egg
itself. A similar form and similar relations to the embryo are
possessed also by the ova of other Hymenoptera, for example
Formica (Ganin, 1869), Chalicodoma (Carriere and Burger,
1897), and Polistes (Marshall and Dernhehl, 1905). In many
other insects it is known, moreover, that the egg before deposi-
tion lies in the ovary of the mother in such a position that the
parts of the future embryo are directed or oriented coincidently
with those of the mother. That is, the cephalic pole of the egg
is turned toward the head of the mother. This is known as the
law of orientation of Hallez, named from its discoverer (1886).

4

THE EMBRYOLOGY OF THE HONEY BEE 5

Whether this relation obtains in the bee has so far apparently
not been demonstrated. It is at least certain, however, that the
egg is generally deposited with the cephalic pole directed out-
ward, or toward the mouth of the cell. If, as is in all prob-
ability true, this is the position in which it comes from the
ovipositor of the queen, then the future caudal pole of the embryo
would correspond with the caudal end of the queen, and, since
in the ovary of the queen the eggs are disposed parallel to her
long axis, the law of Hallez doubtless applies to the honey bee.

As is known to every bee keeper, the eggs are commonly
placed, one in each cell, at or near the center of its floor. The
egg is attached by its smaller or caudal end by means of a minute
quantity of an adhesive substance secreted by the queen, and it
is thus enabled to stand at right angles to the bottom of the cell.
While this is the usual manner in which the eggs are deposited,
many deviations from it are frequently observed. These take
the form of variations in number and in position. Frequently
two or more eggs may be laid in a single cell; Grassi records
finding as many as six; but this number is frequently exceeded,
bee keepers sometimes finding as many as two dozen. Such eggs
may or may not be in the same stage of development; they may
be laid separately or adhering to one another. In examples of
the case last mentioned it is of interest to note that the adhesive
area of the egg is not confined to the caudal pole but extends
well up toward the middle of the egg. Variations in position are
very common; for example, the egg may be placed either on its
side, or on its posterior end; against the bottom or the walls
of the cell. As far as known such variations as these do not
necessarily imply abnormality either in the queen or in the egg,
but it is of course impossible for two or more larvae to long
continue their development together in a single cell. Probably in
most cases, the workers see to it that superfluous eggs are soon
removed, although two larvae are occasionally found in a single
cell. There is a general belief among bee keepers that such devi-
ations inthe position or the number of eggs laid in a cell indicate
abnormality in the queen, since they most frequently occur when
the queen is unfertilised or senile, or in the case of workers be-
coming fertile in the absence of a queen and of worker brood.
It is of interest to note that, according to Marshall and Dernehl

6 THE EMBRYOLOGY OF THE HONEY BEE

(1905), the egg of Polistes is usually attached to that wall of
the cell nearest the center of the nest and about two-thirds of
the depth of the cell from its mouth.

°
aoa
*o

20 6. 2vo6
®

TO 0.

Fic. 1. A, Surface view of anterior end of egg, showing the reticulated
chorion, and the micropylar area (M); x 100. B, Surface view of a small
portion of the chorion showing reticulum and tubercules, x 713.

Two membranes cover the external surface of the egg. The
outer of these, the chorion, (Fig. 1, A and B), is a very thin and
transparent but tough and dense sheet about half a micron in

thickness, completely surrounding the egg, and protecting it from
direct contact with the atmosphere. The composition of this

THE EMBRYOLOGY OF THE HONEY BEE 7

membrane has been studied by Tichomiroff (1885) in the silk-
worm, more recently by Lecaillon (1897), and by Lecaillon and
Henneguy conjointly (1903) in the eggs of the representatives
of several orders of insects. These investigators have all found
that the chorion is not composed of chitin, as is the integument
of the imago, but of a peculiar substance differing in composition
from both chitin and horn, which Tichomiroff has termed chorio-
nin. The tests employed by these investigators have not been
applied to the chorion of the bee’s egg, but it is altogether prob-
able that it also is composed of chorionin.

Like other insect eggs the chorion of the bee’s egg is not
smooth but sculptured. This sculpturing takes the form of deli-
cate ridges forming a meshwork over nearly the entire egg
(Fig. 1A). The meshes have the forms principally of pentagons
and hexagons elongated in the direction of the long axis of the
egg. Examination under high power shows that the ridges
forming the polygons (Fig. 1B) are made up of minute circular
papillae fused together at their bases, and that similar papillae are
scattered about over the area within the polygons. Patterns of
this general character are commonly found among insect eggs,
the polygons representing merely the imprint of the ovarian folli-
cle cells which secreted the chorion. At the posterior end of
the egg the ridges fade out and disappear. At the extreme an-
terior end the ridges converge toward a small area over which
the chorion is conspicuously thickened. This area in general
appearance and position corresponds to the micropylar area in
other insect eggs and will accordingly be termed such (Fig. 2).
It consists of a plate-like thickening of the chorion, approximately
circular in outline, whose margins are continuous with the ends
of the ridges mentioned above. This thickening exhibits a num-
ber of what appear to be perforations or fenestra; these are very
irregular in outline and at the edges of the micropylar area merge
with the spaces between the ridges.

In other insect eggs the micropyle is a perforation or system
of perforations of the chorion permitting the entrance of sperma-
tozoa into the substance of the egg. Frequently the micropyle is
very complex in structure, and sometimes closed by a gelatinous
plug. It is very frequently, although not always, situated at the
cephalic pole of the egg. The writer has not been able to demon-

8 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 2. Micropylar area, as seen in face view, x 713.

strate actual perforations in the micropylar area of the bee’s egg.
All of the sections of the eggs have been unsatisfactory in this
regard, This is in part due to the density and toughness of the
chorion, these qualities being intensified by the process of dehy-
dration preliminary to embedding in paraffin, so that clear and
sharp transverse sections of this portion of the chorion are rare.
The indirect evidence that this is the micropylar region is how-
ever very strong. In the first place, as has already been said,
the micropyle is in very many insect eggs situated at the an-
terior pole. In the second place no other portion of the chorion
displays any differentiation which could be interpreted as a
micropylar apparatus; and lastly Blochmann (1889), Petrunke-
witsch (1901) and Nachtsheim (1913) have found that the sper-
matozoa do actually enter the egg at the anterior pole. In spite
of the lack of direct evidence of the actual perforation of the
chorion at the anterior pole it is fairly safe to assume that this is
in fact the micropylar area.

The vitelline membrane, the second of the two membranes en-
closing the insect egg, is in the bee somewhat thinner than the

THE EMBRYOLOGY OF THE HONEY BEE 9

chorion and apparently structureless. It is sometimes found ad-
hering to the chorion, sometimes to the egg.

The contents of the egg of insects, as well as of all other ani-
mals, is made up, broadly speaking, of two portions, namely:
protoplasm—the so-called formative yolk,—the material basis of
all life and the part of the ovum immediately concerned in de-
velopment; and deutoplasm, a store of food material destined
to be consumed by the developing embryo. In the vast majority
of insect eggs the deutoplasm greatly exceeds the protoplasm
in amount.

If an egg of the bee, freshly taken from the comb during the
early stages of development, is placed in normal salt solution and
crushed slightly so as to rupture the chorion and allow some of
the contents of the egg to flow out, examination under the mi-
croscope (Fig. 3) will show that these consist of the following

Fic. 3. Yolk from a living egg, showing vitelline spheres, vitelline
bodies, and the minute refringent bodies supposed to represent Blochmann’s
corpuscles, x 600.

elements: (1) Large transparent spheres of a colorless fluid,
10-25 microns in diameter, the vitelline spheres. (2) Small re-
fringent bodies, apparently solid, lying within these spheres, the
vitelline bodies. ‘Their size is variable, but is about one-sixth the
diameter of the spheres; their form is also variable, rounded to
long ovate, best expressed by comparing them to the water worn
pebbles of glacial gravel. Their number is approximately the
same as that of the spheres. (3) A viscid interstitial substance,
the ovarian protoplasm, pale greenish in color, apparently cement-
ing the spheres together. (4) Multitudes of very tiny greenish
refringent bodies lying within the interstitial substance. These
may possibly correspond to the so-called Blochmann’s corpuscles
found in certain other insects. They are not distributed uni-

10 THE EMBRYOLOGY OF THE HONEY BEE

formly, but are gathered into small groups. These bodies, in
contact with the salt solution often display a lively dancing
(Brownian?) movement, recalling that of some of the flagellate
infusoria. Examination of a similar uninjured egg shows it to
be densely packed with the transparent vitelline spheres (1),
while surrounding them everywhere and filling the interstices be-
tween them is the interstitial substance (3), which also forms a
cortical layer around the periphery of the egg. The smaller
clear bodies (2) are uniformly distributed throughout the egg,
while the tiny greenish Blochmann’s corpuscles (4) although
present within the interstitial substance throughout the egg, are
especially abundant in the cortical layer. Since of these four
components of the egg, one, namely the viscid interstitial sub-
stance, is protoplasm; the remainder, therefore, with the pos-
sible exception of the Blochmann’s corpuscles, constitutes the
deutoplasm, and this makes up by far the greater volume of the
substance of the ovum. In its physical make-up the egg contents
closely approximate that of an emulsion.

Turning to sections of fixed and stained ova in the earlier
stages of development (Figs. 4A and B, Fig. 5) the egg is seen to
be filled with a network of deeply stained protoplasm enclosing
irregularly circular spaces. These are largest near the center
of the egg and diminish in size near the periphery. Around the
latter is a layer of protoplasm continuous with that forming the
meshwork. This is the cortical layer or ““Keimhautblastem” of
Weismann and other German investigators (Figs. 4A and 5, CL).
It has a thickness of about 60 microns near the anterior pole of
the egg and diminishes gradually toward the posterior pole to
about one-half of this thickness. Near the anterior pole, on the
ventral surface, this layer sends out a conical projection into
the interior (Figs. | and 5, PP). This is the polar protoplasm,
the “Richtungsplasma” of Petrunkewitsch. At this stage it con-
tains the polar bodies, or their remains. During the formation
of the blastoderm it disappears. Along the central longitudinal
axis of the ovum, particularly in its anterior half the strands of
the protoplasmic meshwork are much thicker than those nearer
the periphery of the egg (cf. Figs. 4A and B). Within the spaces
of the meshwork are scattered rounded bodies, spherical to long
ovoid in form, always more or less densely stained, but showing

Se = Ye ne

THE EMBRYOLOGY OF THE HONEY BEE II

an especial affinity for iron haematoxylin. Referring to the
description of the contents of the ovum as seen in the fresh con-

esse ;
4 &,

-
2

S4,.. cam
FF ‘ganar
oe

Fic. 4. Protoplasmic network, from sections of unsegmented eggs
fixed in picro-formol. A is taken from the periphery of the egg and shows
the cortical layer (CL); B is taken from near the center of the egg. The
vitelline bodies are conspicuous in both figures, x 1107.

12 THE EMBRYOLOGY OF THE.HONEY BEE

:
ue

° lo >s

cc

me

'o 2

@,.6 of
pages
*-¢@

26°
Le oe5 :
A

°

‘Fic. 5. Anterior half of a sagittal section of an egg, Stage I, showing
the polar protoplasm (PP), the cortical layer (CL) and portions of sev-
eral cleavage cells (CC), x 243.
dition, the deep staining meshwork is readily identified as proto-
plasm (3), the cavities between the meshes representing the
spaces formerly filled by the vitelline spheres. These cavities
were of course originally spherical, but have become more or
less distorted by the action of the reagents with which the eggs
were treated. These spaces are however not invariably empty.
In many preparations, more particularly those treated with acetic
alcohol or Gilson’s fluid, a pale granular precipitate partly or en-
tirely fills these spaces. The small rounded deeply staining bodies
are evidently to be identified with the vitelline bodies (2).
They are, however, not always present, or rather are present in
varying number in different preparations. The natural inference
is that they are more or less soluble in the reagents with which
the eggs are treated preliminary to sectioning, since they are al-
ways visible in fresh material. These bodies are always found
in sections of ova fixed with picro-formol—although much more
abundant in some preparations than in others—but are sparse

THE EMBRYOLOGY OF THE HONEY BEE 13

and scattered in sections of ova fixed in acetic alcohol, and absent
in those fixed in Petrunkewitsch’s fluid. So far all the elements
seen ‘in the fresh egg have been accounted for except the tiny
greenish bodies embedded in the protoplasm (4). These seem
to be represented by deeply stained minute granules, which indeed
appear to lend to the protoplasmic network its dark appearance.

In addition to the structural elements above mentioned, there
are visible in the anterior end of the egg one or more irregular
island-like masses composed of the same granular protoplasm as
the network, the cleavage cells (Figs. I and 5 CC). Each of
these possesses a clear spherical nucleus. At the close of the
process of fertilization there is of course but one nucleated cell,
whose nucleus (first segmentation nucleus) arose by the union—
in the fertilized egg,—of the nuclei of the egg and the sperm
(male and female pronuclei).t This cell soon gives rise by
mitotic division to a group of daughter cells, four of which are
more or less plainly visible in the figure. The form of the cleav-
age cells is highly irregular, or amoeboid, their outlines indented
by concavities which represent the imprint of the vitelline spheres.
Between the indentations arise processes of irregular shapes
which stretch out into the surrounding cytoplasmic network and
are continuous with it. These cells will be considered at greater
length in the next section.

An extended review of the various accounts of the organiza-
tion of the eggs of other insects is not possible here; moreover a
survey of these accounts, as well as of the statements in the text-
books, suffices to show that they quite uniformly agree in de-
scribing the contents of the insect egg as composed of the
following morphological elements: (1) A protoplasmic mesh-
work, which in many insects is extended over the periphery of the
egg to form a cortical layer; in others this layer is absent. Inthe
protoplasm are also included the segmentation nucleus or its
products. (2) Yolk bodies, generally in the form of balls or
spheres, enclosed in the meshes of the protoplasm. (3) Fat
globules. (4) Minute rod-like or rounded bodies embedded in the
substance of the protoplasm, and present only in certain insect
eggs. These bodies were discovered and described by Bloch-

*For an account of this process in the honey bee, see Blochmann (1889),
Petrunkewitsch (1¢0i, 1902), and Nachtskeim (1913).

14 THE EMBRYOLOGY OF THE HONEY BEE

mann (1884, 1886, 1887, 1892) for the eggs of Phyllodromia
(Blatta), Periplaneta, Blabera, Lasius, Pieris, Musca, and
Vespa. In the first four genera named these bodies were rod-
shaped, and strikingly resembled certain bacilli. Moreover, this
resemblance was enhanced by the fact that these bodies multi-
plied by transverse fission. They were found to be especially
abundant in the cortical layer, and also present in the fat body
of the imago. In the three genera last named, these bodies were
rounded or granular in shape and only tentatively identified with
the bacillar form. Wheeler (1889) has also found and described
these bodies in Blatta, and gave them the name “Blochmann’s
corpuscles.” Mercier (1906) has demonstrated that these are
independent organisms, probably bacteria, and capable of culti-
vation in artificial media. Friederichs (1906) has recently de-
scribed similar bodies, which he terms “Blochmann’s corpuscles”
(““Blochmannische Korperschen”), embedded in the protoplasmic
meshwork and cortical layer of certain chrysomelid beetles
(Chrysomela, Rhagonycha), and Tanquary (1913) finds similar
bodies in the eggs of an ant, Camponotus. In the case of the
honey bee the minute bodies to which has been given the term
“Blochmann’s corpuscles” are of course comparable only to the
rounded form, and only provisionally identified with these, since
in the honey bee so little is known concerning these bodies that
final identification would be premature.

Turning to the accounts of the hymenopterous egg in particu-
lar, there are available the accounts of Henking (1892) for
Lasius, Carriére and Burger (1897) for the mason bee (Chali-
codoma) and Anthophora, Marshall and Dernhehl (1905) for
Polistes and Tanquary (1913) for Camponotus; for the honey
bee Butschli (1870), and Grassi (1884). Henking’s observa-
tions contain little of interest here, except that the egg of Lastus
possesses a well developed cortical layer. In the case of Chali-
codoma the account is brief, essentially contained in the state-
ment that the contents of the egg “very fluid, consists of an
emulsion of a considerable amount of deutoplasm (food yolk)
in a small amount of protoplasm.” The cortical protoplasmic
layer is wanting, but present in Anthophora. The data given by
Marshall and Dernhehl regarding Polistes are brief and relate
principally to the cortical layer. This is very similar to that of

THE EMBRYOLOGY OF THE HONEY BEE 15

the honey bee but is thicker on the ventral than on the dorsal
side. In Camponotus Tanquary mentions a well developed corti-
cal layer, containing a large number of Blochmann’s corpuscles,
and having also embedded in it numerous yolk granules. Vacu-
oles, so called, are present at the center of the egg, more espe-
cially near its anterior end. Possibly these correspond to the
yolk spheres of the honey bee.

In the descriptions of both Biitschli and Gras are noted the
protoplasmic network, the vitelline spheres, the vitelline bodies
and Blochmann’s corpuscles. The latter, it is true are not
specifically mentioned, but the protoplasm is spoken of in both
instances as “granular,” an appearance referable to the presence
of these bodies. These investigators also observed the cortical
layer, describing it also as “granular.”

As regards the chemical make-up of the deutoplasmic portion
of the insect egg, very little is known. It is commonly described
as consisting of yolk and oil globules. This paucity of informa-

tion is no doubt due to the fact that embryologists usually have

not the training requisite to enter the difficult field of micro-
chemistry. It is true that Tichomiroff (1885) made an elabo-
rate chemical analysis of the silkworm egg, but this work has
apparently stood alone. Friederichs (1906) subjected the con-
tents of the egg of Rhagonycha and Chrysomela to a few simple
reagents: water, normal salt solution, an aqueous solution of
osmic acid and mercuric chloride, and diluted hydrochloric and
acetic acids combined. The results convey but little information
concerning the real nature of the deutoplasm except to demon-
strate the presence of oil droplets. The presence of these ap-
pears to be quite general in insect eggs. Grassi assumed that
the vitelline spheres in the egg of the honey bee are of an oily
nature; a natural supposition, suggested by their appearance.
Repeated tests with osmic acid, however, failed to yield the
well known blackening reaction characteristic of fat, nor did
other elements in the egg show it. The egg of the bee therefore
is presumably totally devoid of fat or oil, as such. A few tests
with other reagents were applied to the vitelline spheres, but with
inconclusive results, except that they appear to indicate that the
vitelline spheres in the bee are similar to those in the eggs ex-
amined by Friederichs.

Kit

CLEAVAGE

The group of cleavage cells? formed by the division of the
segmentation nucleus is at first nearly spherical in outline, and
situated in the middle of the anterior end of the egg about equi-
distant from the cephalic pole and the ventral, dorsal and lateral
surfaces. By rapid multiplication and migration of its compo-
nent cells, this group elongates in a caudad direction until at
Stage II (4-6 hours) it extends some two-thirds of the length of
the egg toward the caudal pole. The cleavage cells are now
arranged in the form of a one-layered meshwork forming a
hollow figure whose outline is long conical or pyriform, the
slender pointed end directed caudad, the larger rounded end
close to the surface at the anterior pole (Fig. Il). A cross section
through the anterior portion of an egg at this stage is shown in
figure 6. The cleavage cells are here seen arranged in a circle
whose position is slightly eccentric with respect to the median
axis of the egg, being displaced a trifle toward the ventral
(lower) side. The cells are still situated at some distance from
one another, as well as from the periphery of the egg.

With the change in the form of the group of cleavage cells
from spherical to elongate there arises a difference between the
cells composing its anterior and posterior portions. The differ-
ence consists in this: that the cells in the anterior portion of the
cleavage figure are more numerous, more compact in form and

*Blochmann (1887a) has called attention to the fact that the complex
made up of the cleavage nuclei, together the protoplasm immediately
surrounding them, and the protoplasmic meshwork, including the cortical
layer, is a syncitium. It is therefore not strictly correct to use the term
“cleavage cells” as applied to the cleavage nuclei and the protoplasm im-
mediately surrounding them. The employment of this term may however
be justified both on the ground of convenience, and because the cell terri-
tory appertaining to each nucleus is always distinguishable, except for a
very brief period during the formation of the blastoderm. This position
is held by Heider (1889), Heymons (1895), Lecaillon (1897),. ete.

16

en) fata —

THE EMBRYOLOGY OF THE HONEY BEE 17

Fic. 6. Tranverse section of anterior end of egg of Stage II, showing
cleavage cells (CC) x 243.
joined more closely together than those situated further caudad ;
the latter present by comparison a scattered and somewhat at-
tenuated appearance. This difference is plainly evident in figures
IT and III (cf. also Figs. 7A and 7B). Moreover, although all
the cells in the egg are of the same age those in the posterior
portion are retarded in development in proportion to their dis-
tance from the anterior end, and since they reach their destina-
tion in the cortical layer later than those situated anterior to
them, a handicap, so to speak is placed on the posterior region
of the egg, from which it is slow to recover. This condition has
also been noted by Carriére and Birger for Chalicodoma, but is
entirely ignored in Dickel’s account (1904). Marshall and
Dernhehl also make no mention of it in the case of Polistes, nor
does it appear in Lasius (Henking).

As the number of cells continues to increase the pyriform

figure grows both in length and diameter until it finally touches

the cortical layer of the egg at one or more points. The point
where the cleavage cells first actually come into contact with the
periphery of the egg is usually situated on the ventral side, one-

Fic. 7. Transverse sections of egg of Stage III. A, from anterior
half, showing cleavage cells (CC) near the periphery, and yolk cells (YC)
near the center. B, from posterior end. showing cleavage cells (CC), x
243.

THE EMBRYOLOGY OF THE HONEY BEE 19

fourth to one-third of the length of the egg from the cephalic
pole. In some cases the cells appear as though concentrated on
this point, but its location is probably subject to considerable
individual variation. Soon after the first cells reach the surface
the anterior half of the egg becomes rapidly invested with a
covering of cells, yet the ventral surface is, as figures III and 7A
show, generally attained a little before the dorsal. The attain-
ment of the periphery of the egg by the cleavage cells now pro-
gresses slowly caudad, the caudal pole itself E hente the last to
receive a cellular investment.

By far the greater number of cells advance toward the peri-
phery in a single row, eventually attaining the cortical layer and
then forming blastoderm, nevertheless, a few linger behind
within the yolk and become the yolk cells. Three of these are
clearly seen in figure 7A, YC. At this period they are not dis-
tinguishable from the other cells by any visible characteristics.
Their origin has already been correctly described by Dickel
(1904) for the honey bee, and conforms to that of the higher
insects in general. The further history of these cells will be
considered later.

The varied and highly irregular amoeboid contour of the in-
dividual cleavage cells is illustrated by figures II, III, 6, 7A and
B, as well as by the series represented in figures 8A-F. As the
figures show, the outline of the cell is indented by concavities ;
these concavities represent the spaces occupied by the vitelline
spheres. Between these concavities irregular slender processes
extend out from the cell; these either merge with the protoplas-
mic network or unite with corresponding processes of adjacent
cells, so that the cleavage cells, together with the other proto-
plasm of the egg—protoplasmic meshwork, cortical layer—form a
single continuous whole, and in fact constitute a true syncitium.®
As the cells move toward the surface they become more numer-
ous by repeated divisions, and more closely crowded together,
their connecting processes become correspondingly more numer-
ous, shorter, stouter and more regular in outline. Figure 9 il-
lustrates a tangential section through the layer of cleavage cells
near the anterior end of an egg between Stages II and III show-
ing the manner in which the cells are linked together at this time.

“See footnote p. 16.

20 THE EMBRYOLOGY OF THE HONEY BEE

:
2

Fic. 8. Cleavage cells, showing stages of cell division. A and B,
prophases; C and D, metaphases; E and F, telephases; G, multipolar
spindle, x 1107.

THE EMBRYOLOGY OF THE HONEY BEE 21

The nuclei of the cleavage cells, during the resting stage, are
circular to broad elliptical in outline. Each contains a network
of achromatic material bearing on it a number of small rounded
granules of chromatin. No nucleoli—karyosomes or plasmo-
somes—were observed. The size of the nuclei is, in a given egg,
quite uniform, from the beginning to the end of the period under
consideration, but varies considerably in different eggs, ranging
from 9-14 microns.

The position of the nucleus within the cell is very variable,
but it can be stated that in general the nuclei during the earlier
Stages (I-I1) occupy a position near the center of the cell body
(Fig. 7B). As cleavage progresses, however, the nuclei evi-
dence a tendency to take up a position near the margin, and this
tendency becomes more and more marked as the cells approach
the periphery of the egg (Figs. 6 and 7A).

Nachtsheim has recently (1913) published a detailed account
of mitosis in the cleavage cells of both the fertilized and un-
fertilized eggs of the honey bee. Prior to the appearance of
this account the writer had also studied this subject and pre-
pared figures 8A-F, illustrating the different phases of mitosis.
The writer’s observations agree with those of Nachtsheim ex-
cept in one or two minor details. A clearly defined spireme was
never observed, although in the prophases (Figs. 8A and B) the
chromosomes are seen to be united in short strings, which in
many preparations, as in that from which figure 8B was drawn,
give the effect of a relatively small number of curved rod-shaped
chromosomes, and it is only by the use of the highest magnifica-
tion available that these resolve themselves into the minute spher-
oidal chromosomes of which they are composed. No special
effort was made to determine the number of these. It seems

probable that Nachtsheim is right in estimating the number as

thirty-two. There are certainly more than sixteen. The centro-
somes,* densely staining minute spherical bodies, are readily seen.
at all stages. In figure 8A they have taken positions at opposite
sides of the nucleus preparatory to division. Soon after this the
nuclear membrane breaks down, usually disappearing first at the

*Nachtsheim, following Boveri, prefers to call these the centrioles, and
the surrounding zone the centrosome. The latter the writer regards as
the attraction-sphere.

Bo THE EMBRYOLOGY OF THE HONEY BEE

Fic. 9. Tangential section through cone of cleavage cells, Stage II-III,
showing arrangement of cells and their relation to one another, x 534.

points opposite the centrosomes, when the astral rays may fre-
quently be seen attached to the chromosomes and traversing the
nucleus. These fibres are plainly evident in preparations fixed
in Petrunkewitsch’s fluid, but less so in those fixed in picro-
formol, as were those from which the series of drawings were
made. In figure 8B an attraction-sphere surrounds each centro-
some. Figures 8C and D show the metaphases of division. One
of the centrosomes is seen to have divided, and this division be-
comes even more evident in the anaphases and telophases (Figs.
SE and F). Nachtsheim describes a division of the centrosomes
at both poles, which would naturally be expected, but the numer-
ous preparations studied by the writer for some reason showed
2 division at only one pole. During the anaphases (Fig. 8E) the
centrosomes increase in size and the nuclei frequently have a bi-
lobed appearance. This division of the nuclei into two lobes is
not uncommon in the early stages of development of other ani-
mal forms, as for example Cyclops (Riickert 1895) and Crepi-
dula (Conklin 1897). One of the two lobes in such cases is
supposed to represent the maternal half of the nucleus, the other

THE EMBRYOLOGY OF THE HONEY BEE 23

the paternal, corresponding to the two pronuclei which unite to
form the first cleavage nucleus.

A prominent and deeply stained mid-body is frequently ob-
servable between newly divided daughter nuclei (Fig. 8F).

A peculiarity of mitosis, confined to the cleavage cells, at once
evident in glancing over the series represented in figs. 8A-H, is
the small size of the spindle as compared with that of the resting
nucleus. This disparity ceases to appear after the cells have
attained the cortical layer and commenced to form blastoderm.

Multipolar spindles, as described by Lecaillon (18972) were
not observed during cleavage, but in certain preparations, irregu-
lar mitotic figures were visible of the type represented by figure
8G. These are so rare, however, that it would not be permissible
to conclude that they are normal phenomena.

In a number of insect eggs,’ it has been observed that the
division. of the cleavage cells take place in a definite direction,
that is, at right angles to the egg’s surface. Examination of
sections of the bee’s egg during Stages I and III, shows in di-
viding cells mitotic spindles turned in various directions and
apparently not conforming to any rule, except when the cells are
approaching the periphery of the egg. In this case the spindles
are more frequently parallel to the surface, indicating that the
resulting division plane will be normal to the surface. Never-
theless, a consideration of the relation of the cleavage cells to
one another will make it evident that the majority of the di-
visions must be in effect at right angles to the egg’s surface,
since the cleavage cells—with the exception of those few destined
to form yolk cells,—are arranged in a single layer, and this for-
mation is maintained throughout the period of cleavage. The
products of every division—with the exception just noted—must
therefore eventually come to lie in this layer, so that whatever
the directions of the mitotic spindle may be, the plane ultimately
separating the daughter cells will be normal to the egg’s surface.

In figures 5, 6, and 7B there is observable a marked contrast be-
tween the structure of the egg within the zone of cleavage cells
and that outside of it. This difference concerns the protoplasmic
meshwork. On the outside of the zone it is evident that the

*E.g., Aphis, Musca, Blochmann (1887); Blatta, Wheeler (1889) ;
Forficula, Heymons (1895) ; Clytra, Lecaillon (1897a).

24 THE EMBRYOLOGY OF THE HONEY BEE

strands composing the network present the same appearance as
in the unsegmented egg; within the zone they are scarcely visible.
On examination of the interior region with high powers it is
seen that the network has now all but disappeared, of it there
remains only a cobweb-like remnant, with tiny polyhedral thick-
enings marking the nodes (Fig. 10). Since the network periphe-

' a mex
o A L a mi @

Fic. 10, Protoplasmic reticulum from center of egg, Stage II, showing
diminution of reticulum subsequent to the peripherad migration of cleav-
age cells, x I107.

rad of the zone of cells is not altered, it is evident that what was
lost from the interior of the egg has been drawn into the
cleavage cells. This process of absorption continues as the cells
advance until they reach the cortical layer, when the condition
formerly seen only within the line of cleavage cells now prevails:
throughout the entire egg, except in the immediate neighborhood
of the yolk cells. By far the greater portion of the protoplasm
originally present in the ovum is drawn to its surface to furnish
material for the building up of the future embryo. A similar
process of absorption of protoplasm by the cleavage cells doubt-
less exists in most insect eggs, but is much more strikingly evi-
dent in certain of the Hymenoptera than elsewhere. Carriere
and Birger have described and figured it for both Chalicodoma
and Anthophora; in both of these forms it is identical with that
seen in the honey bee. Grassi and Dickel have also represented

THE EMBRYOLOGY OF THE HONEY BEE 25

in their figures the lack of protoplasm within the line of cleav-
age cells, without specifically mentioning it in the text.

Omitting an extended review of the results of other investi-
gators who have dealt with this stage of insect development? it
may be said that speaking broadly, the cleavage process in the
bee is fairly typical for the higher insects, and offers but few
peculiarities worthy of extended comment. Such differences as
exist relate principally to details. Taking up some of these, it
may be mentioned first, that while the formation by the cleavage
cells of a hollow one-layered figure is common to most of the
higher insects, yet this figure in most cases conforms more closely
to the outline of the egg than in the case of the bee. This is
illustrated by the Coleoptera (Hydrophilus, Clytra), Diptera
(Musca), Lepidoptera (Zygaena) and some Hymenoptera
(Lasius, Polistes). Second, that part on the egg’s surface first
reached by the cleavage cells may be located near the middle of
the egg, or near one of the two poles. <A table compiled by
Marshall and Dernhehl (1906)* showing the location of this
place in twenty-six insects belonging to seven different orders,
also shows that different insects within the same order may
differ widely in this respect. The Hymenoptera illustrate this
point quite well, since, in Formica (Ganin 1869), Lasius (Henk-
ing 1892), and Rhodites (Weismann 1882), the cleavage cells
first reach the periphery of the egg near the posterior end, in
Platygaster (Kulagin 1897), near the equator, in Chalicodoma
(Carriére and Birger, 1897), and Polistes (Marshall and Dern-
hehl, 1906), near the anterior end, on the ventral side, as in
Apis. It is therefore evident that this feature possesses but little
significance.

In the ant Camponotus, according to Tanquary (1913) the
conditions obtaining during cleavage are quite different from
that of the majority of other non-parasitic Hymenoptera, and
decidedly remarkable, recalling the conditions found by Weis-
mann (1882) in Rhodites and Biorhiza. According to Tan-
quary’s rather imperfect observations, a large nucleus, presum-

° For a complete review of this phase of insect embryology see Lecaillon
1897a, Henneguy 1904 and Marshall and Dernhehl 1906. The last named
is particularly complete on the historical side.

min, 6; P.. 140.

26 THE EMBRYOLOGY OF THE HONEY BEE

ably the first cleavage nucleus, is seen at the posterior end of the
egg. Later smaller cleavage cells are found in the anterior half.
These multiply, spread apart, and finally form the blastoderm,
which, as in the honey bee, first appears on the ventral side of
the egg, near the anterior end. The large nucleus at the posterior
end of the egg meanwhile disappears.

In concluding it should be said that the account of the cleavage
in the egg of the honey bee agrees essentially with those of
previous investigators, as far as their observations extend. The
accounts of Kowalevski, Butschli and Grassi are, as regards
the cleavage, rather fragmentary and incomplete; that of Dickel
is much more complete and in general correct, although his figures
are diagrammatic, and evidently not accurately drawn from
actual preparations. Petrunkewitsch (1901) also gives a brief
description of cleavage in the drone egg, illustrated by two fig-
ures. Among the other Hymenoptera investigated, those whose
early development conforms most closely to the bee are Chali-
codoma and Polistes.

IV
FORMATION AND COMPLETION OF THE BLASTODERM

In the preceding section the cleavage cells were followed to
the point where they reach the periphery. This stage in their
migration is illustrated by figures 7A and 11A. In the second
figure mentioned only a single cell is represented, much magni-
fied, together with a corresponding portion of the cortical layer.
A distinction is now apparent between the protoplasm of the
cortical layer and that of the cleavage cells. The former is
clear and transparent, the latter dark and granular. The cleav-
age cells rest upon the cortical layer as upon a base, supported
by short intervening processes which are separated from one
another by irregularly circular openings. On meeting the corti-
cal layer these processes spread out to unite under this layer to
form a continuous sheet. On comparing figure 7A with an ear-
lier stage, figure 6, it becomes evident that these processes ex-
tending peripherad from the cleavage cells are only the strands
of the protoplasmic meshwork underlying the cortical layer,
shortened and thickened, while the openings represent the spaces
occupied by the vitelline spheres. As the cells advance peri-
pherad the processes shorten, while the open spaces flatten and
ultimately disappear (Fig. 11B) ; the nucleus of each cell is thus
brought into contact with the cortical layer (Fig. 11C). The
nucleus continues to advance peripherad, followed by the deep
staining protoplasm of the cell body, while over it the cortical
layer begins to rise up on the exterior in a more or less rounded
swelling, each of these being marked off from its neighbors by a
wide and deep furrow (Figs. 11C, D and E). This stage, when
seen in the living egg, presents a very striking and characteristic
appearance, and has been well described and figured by Kowa-
levski (1871). A differentiation in the character of the proto-
plasm of the cell body now begins to appear. That part
accompanying the nucleus is seen to be less deeply stained and

27

28 THE EMBRYOLOGY OF THE HONEY BEE

Pe tery
=, ote
intsaieeeee

eres

Fic. 11. A series of stages to illustrate the entrance of a cleavage cell
into the cortical layer during the formation of the blastoderm. In A the
cleavage cell has just come into contact with the lighter tinted cortical
layer. In F the external portion of the cleavage cell has entered the
cortical layer and has produced a well marked elevation on the surface of
the egg. In Gand H the cleavage nucleus is dividing, x 1107.

THE EMBRYOLOGY OF THE HONEY BEE 29

less granular than the deeper lying portions, although it, in turn,
is still less transparent and darker than the protoplasm of which
thé cortical layer is composed, and sharply marked off from it.
In figure 11F the migration is nearly completed; each cleavage
nucleus is embedded in a rounded swelling of the cortical layer,
and flanked on each side by a mass of cytoplasm derived from
the cleavage cell; these are united behind (centrad to) the
nucleus and merge with a still darker mass of cytoplasm which
still retains the branching and amoeboid form characteristic of
the cleavage cells. These irregular masses of dark cytoplasm be-
longing to adjacent cells now form a continuous although irregu-
lar stratum underlying the nuclei with their accompanying
cytoplasm. Next the cell boundaries become distinguishable,
faint indications of cell walls extending inwards from the bot-
tom of the furrows; the deeper stained bases of the blastoderm
cells are however still linked to one another, so that cell terri-
tories are yet not completely demarcated (Fig. 12A). Soon after,
the walls separating the cells are completed; a delicate basement
membrane is formed near the bases of the cells, cutting off the
latter from the narrow inner zone of deeply staining protoplasm.
The cells are now completely delimited and the blastoderm, as
such, is established (Fig. 13A). It should, however, be always
borne in mind that these processes do not take place simultan-
eously over the entire surface of the egg, or even over any
extended part of it. They commence where the cleavage cells
first reach the cortical layer, namely, on the ventral side near
the anterior end, then extend rapidly over the entire anterior
third of the egg, thence progressing slowly caudad, so that all
phases of the process of blastoderm formation may frequently be
observed in a single egg.

A striking feature of these early phases in the formation of the
blastoderm is the role played by the nucleus. As mentioned in
the section preceding, when the cleavage cells approach the corti-
cal layer, the nuclei take up a peripheral position in the cells.
This is plainly seen in figures 7A and 11A; the nuclei are now
very close to the external margin of the cell body. From this
point on the nucleus leads the advance into the cortical layer,
and after coming into contact with it (Fig. 11B) has apparently
become denuded on its peripheral side of cytoplasm belonging

Rimaramamenyetia

petiaicins

Fic. 12. Later stages in the formation of the blastoderm. A, newly
formed blastoderm, from a sagittal section, cell boundaries just appearing.
B and C, blastoderm cells dividing obliquely. D, blastoderm cells so ar-
ranged that their nuclei lie at two levels. E, blastoderm cells from
ventral side of egg, 24-26 hours. F, same, 28-30 hours, x I107.

THE EMBRYOLOGY OF THE HONEY BEE 31

to the body of the cleavage cell. That this migration is accomp-
lished by an actual movement on the part of the nucleus itself is
contrary to the facts of biology; the source of the movement
must therefore be referred to the cytoplasm, as Marshall and
Dernhehl have pointed out. Moreover, nuclei are frequently
seen elongated at right angles to the line of movement, as though
somewhat flattened by pressure before and behind (Figs. 11D
and F). A flattening of the nuclei has also been observed to oc-
cur under similar circumstances by Lecaillon (1897a) in Clytra,
and by Noack (1901) in Musca, and is similarly explained by
these investigators. In these instances, however, the nuclei are
flattened only at their peripheral ends.

A second feature to be noted is the sharp separation of the
protoplasm derived from the cell body and that of the cortical
layer. This separateness is readily seen in figure 12A, when the
cell walls are beginning to appear. An ultimate fusion of the
material from these two sources occurs later, but seems to take
place only very slowly. The cytoplasm of the blastoderm cells
is therefore derived in part from the cleavage cells, but not all
of the cytoplasm from this source is thus employed, a consider-
able portion being temporarily cut off and left within the blasto-
derm in the form of a deeply stained granular layer, described
above. Such a layer of granular protoplasm was first observed
by Weismann (1863) in Musca and named by him the “innere
Keimhautblastem”; this layer may then accordingly be termed
the inner cortical layer, but it must be remembered that it has
no close genetic relationship with the cortical layer proper. This
inner layer has subsequently been found to exist in the eggs of
many insects, but appears to be wanting in the orders Orthoptera
and Dermaptera.

Figures 13A and 13B represent transverse sections through an
egg just after the blastoderm has become established. As is
readily apparent, the blastoderm is not of uniform thickness
throughout, but is thinner on the dorsal than on the ventral side.
Comparing figure 13B, from the posterior region of the egg,
with figure 13A, through its anterior region, it is evident that
the blastoderm-forming cells of the posterior region are much
less numerous than those of the anterior region. This difference
is directly referable to the preceding cleavage stages when a

THE EMBRYOLOGY OF THE HONEY BEE

32

ae

$

a

3

Fie get
,

‘vor etee ®

ions from eg
near anter

Tranverse sect

Fic, 13.
fermation of blastoderm.

g about 16 hours old, just after

ior end, x 243.

, B, near poster

10r

b

A

oe eo oe

THE EMBRYOLOGY OF THE HONEY BEE 33

similar numerical relation occurs (see p. 16 and Figs. II and III).
Figure 14 represents a sagittal section of a stage a trifle older

Fic. 14. Anterior end of a sagittal section of an egg about 16 hours
old. The point of junction of the future ventral blastoderm with the
thin dorsal blastoderm is designated by a star, x 243.

than those of the two figures preceding. Here the difference in
thickness between the dorsal and ventral blastoderm is more
marked. Moreover the thicker blastoderm of the ventral side
is continued over the anterior pole, being at this point even
thicker than the ventral blastoderm itself. At the point on the
dorsal surface where the thickened area joins the thinner and
somewhat flattened cells of the mid-dorsal area (marked in the
figure by a star) the cells of the latter have an irregular amoe-
boid form, and separated by short intervals, as though pulled
apart. The existence of this area of scattered cells was first
noticed in the bee by Dickel (1904) who has made much of it
in his endeavor to construct an original interpretation of the
origin of the germ layers in insects. A discussion of Dickel’s
view of the relation of this area to the germ layers will be given
later.

About the time that the nuclei enter the germ layers they un-
dergo mitotic division (Fig. 11G and H). This may occur either
during the migration (Fig. 11H) or immediately after (Fig.

34 THE EMBRYOLOGY OF THE HONEY BEE

11G); the division plane is always normal to the external
surface.

In many accounts of the insect development some details are
given of the manner in which the cleavage cells became trans-
formed into blastoderm, but probably the most circumstantial
accounts are those of Blochmann (1887a) and Noack (1901)
for Musca; of Heider (1889) for Hydrophilus, Carriere and
Biirger (1897) for the mason bee (Chalicodoma) and its para-
site, Anthophora, and Marshall and Dernhehl (1905) for Poli-
stes. The last named account is by far the richest in details. In
Musca and Hydrophilus the process (as described) is essentially
the same. The cleavage cells, on arriving at the cortical layer,
fuse with it to form a continuous sheet without any cell boun-
daries—a syncitium, in other words—in which the nuclei divide,
the spindles of the mitotic figures being parallel to the egg’s
surface. Cell boundaries now appear, first as furrows of the
external surface of the egg, thence extending inward, delimiting
the cells on their lateral surfaces. Meanwhile the darker stain-
ing bases of the cells form a continuous layer within, constituting
the “inner Keimhautblastem.” In Hydrophilus this is soon ab-
sorbed by the cells of the blastoderm. In Musca this layer is
much thicker and, in certain portions of the egg, separated from
the blastoderm by a layer of yolk. Moreover it is only slowly
absorbed by the blastoderm cells, but is ultimately taken up by
them. In Anthophora the process is as just described, except
that no “innere Keimhautblastem” is formed. It is also absent
in. Chalicodoma, as is the cortical layer (Keimhautblastem).
The account of Marshall and Dernhehl allows a much closer
comparison with the condition met with in the honey bee. In
Polistes the cleavage cells enter into the cortical layer separately,
with the nuclei situated at the peripheral margin of the cell body,
which has the appearance of trailing after the nucleus like the
tail of a comet, remaining meanwhile separate from the cortical
layer and distinguished by its darker shade, as in the honey bee.
A regular inner cortical layer was not observed, but subsequent
to the formation of the blastoderm irregular patches of a rather
granular mass were found at the bases of some of the blasto-
derm cells. The separation of the cell territories was accomp-
lished in the usual manner. There is then almost complete agree-

THE EMBRYOLOGY OF THE HONEY BEE 35

ment between Polistes and Apis in the particulars relating to the
entrance of the cleavage cells into the cortical layer and their
subsequent transformation into blastoderm. The chief points
of difference are: the narrow, almost pointed shape of the
cleavage cells of Polistes when entering the cortical layer, and
the relatively scanty amount of basal (centres) material left to
- form an inner cortical layer.

The further changes in the history of the blastoderm can best
be followed by taking them up in the order of their age. This
is in fact the only safe method of determining the order of
development during this period.

Seventeen to twenty hours. The formation of the blastoderm,
it is estimated, is completed in from 14 to 16 hours, since the
stage under consideration, identified by means of eggs whose age
is actually known, immediately follows and is directly connected
with the stages just described by changes which can occupy only
a brief interval.

While the cleavage cells are entering the cortical layer, or im-
mediately after this event, they divide mitotically, with the spin-
dles parallel to the egg’s surface, as already described. This may
possibly be followed by another similar division, but whether
this occurs is uncertain; at all events not more than two divisions
of this character intervene between the entrance of the cleavage
cells into the cortical layer and the stage of 17-20 hours. Sub-
sequent mitotic divisions however do occur, probably in every
blastoderm cell, but the spindles of the mitotic figures are no
longer uniformly parallel to the surface of the blastoderm; they
now more frequently are inclined at an angle to it, in other
words, they are oblique (Fig. 12B and C). As these illustrations
show, the angle formed by the spindles of different cells with
the surface of the blastoderm is not uniform; some spindles are
parallel to the surface, as in the left hand cell of figure 12B;
some form an angle of 45 degrees with it, as in the middle cell
of the same figure, while others are inclined at a lesser angle.
The effect of these oblique divisions is shown in figures 12D and
15. At first glance the blastoderm appears now to consist of
two layers of cells, superimposed one above the other, since two
rows of nuclei are seen, except along the dorsal mid-line. That
two layers of cells are actually present is rendered improbable

36 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 15. Transverse section through the anterior region of an egg
about 17-20 hours old, showing the nuclei of the blastoderm arranged in
two layers, x 243.

by the fact that very many of the cells situated next to the peri-
pheral surface are pyriform in outline, with their slender ends
reaching inwards to the basement membrane. These alternate
more or less regularly with shorter cells lying close to the base-
ment membrane, but which frequently possess a slender peri-
pheral process. This apparent two layered condition was
described and figured by Kowalevski (1871), who explained it
ingeniously by supposing it to be an artifact, brought about dur-
ing the formation of the action of the fixing agent. The latter
was assumed to have contracted the layer of cells first appearing
at the surface of the egg, closing up the intervals between them
and crowding them down on those cells arriving later. This
same stage has also been figured for the drone egg by Petrunke-
witsch (1901, Fig. 24), a dividing cell is also shown. Oblique
divisions of the blastoderm cells are also described for Polistes —
by Marshall and Dernhehl (1905).

Viewing the blastoderm as a whole, it appears more compact
and sharply outlined than it was directly after its inception; it
now constitutes a well developed epithelium. Both its outer and

THE EMBRYOLOGY OF THE HONEY BEE 37

inner surfaces are regular and smooth in contour. On the inner
surface the inner cortical layer is evident, but this is also more
uniform as well as thinner than before. Moreover the processes
on its inner surface are no longer so strikingly apparent, although
they still exist.

Twenty-four to twenty-six hours. This stage is represented
by a sagittal section, figure 16. The blastoderm at this stage

Fic. 16. Anterior portion of a sagittal section of an egg about 24-26
hours old. The contrast between the thick ventral blastoderm (ventral
plate) and the thin dorsal strip is very marked. The point of junction
between them is marked by a star, x 243.

differs from that of the preceding stage in several particulars.
The contrast between the blastoderm on the dorsal surface and
that over the remainder of the egg is very marked, since there is
now evident, along the dorsal mid-line, a fairly well developed
strip about one-third of the diameter of the egg in width, and ex-
tending from one pole to the other, composed of flattened cells.
On the other hand, the cells of the ventral and lateral surfaces
have become elongated and columnar in form, whereby, the thick-
ness of the blastoderm in these regions is much increased. In
addition the thickened blastoderm over the anterior pole has
disappeared, giving way to a thin sheet of cytoplasm containing

38 THE EMBRYOLOGY OF THE HONEY BEE

scattered nuclei; the boundary between the thickened blastoderm
of the ventral surface and the thin dorsal sheet is sharp, and is
situated precisely at the anterior pole (marked by a star), instead
of dorsad of it, as in earlier stages (cf. Figs. 14 and 16). This
change is apparently brought about by a shortening of the ventral
blastoderm in the longitudinal axis, since stages intermediate in
age (22-24 hours) show intermediate conditions in respect to the
position of the point of juncture of the thicker ventral blastoderm
and the thinner dorsal strip. In addition to these differentiations
of the blastoderm, its average thickness is much greater in the
anterior than in the posterior half of the egg. Beginning at the
cephalic pole and passing backward the blastoderm first rapidly
increases in thickness as far as the middle of its anterior half
(Fig. 16), from this point it grows thinner slowly but uniformly
to the posterior pole. A slight difference in thickness in favor
of the anterior half of the egg is observable at all stages, but at
the present one is especially noticeable.

The cells composing the blastoderm in the available prepara-
tions of this stage present on close examination a very peculiar
aspect. All—except those of the dorsal strip—have now elon-
gated to a somewhat irregularly prismatic form, with their bases
frequently constricted. The nuclei are situated close to their
peripheral ends (Fig. 12E). The cells have not however all
elongated to the same extent, so that the nuclei do not yet lie uni-
formly at the same level. The cytoplasm of the cells is pale,
and finely vacuolated in structure; centrad of each nucleus is a
large vacuole approximating the nucleus in size. The lateral
cell walls are seen to pass inward some two-thirds of the thick-
ness of the blastoderm, where they are lost to sight in a zone
of vacuolated cytoplasm formed by the bases of the cells them-
selves, which have become continuously fused with one another.
Centrad of this zone, but merging with it, is a darker one, recog-
nizable as the inner cortical layer (JCL). Referring to figure
12A, it will be readily apparent that the blastoderm cells, so far
as their relation to one another and to the inner cortical layer is
concerned, have resumed the condition existing at the time that
the blastoderm was first established, the basement membrane
formerly underlying the blastoderm cells and separating them
from the inner cortical layer having disappeared. A discussion

Oe a

ee ee

THE EMBRYOLOGY OF THE HONEY BEE 39

of the significance of this change will be deferred to the end of

the section.
Twenty-eight to thirty hours. At this stage (Fig. 17) a cross

Fic. 17. Transverse section through the anterior half of an egg about
28-30 hours old. All of the cells of the blastoderm, including those of the
dorsal side, have assumed a columnar form, x 243.

section through any portion of the egg shows the blastoderm of
nearly equal thickness over the entire circumference, the dorsal
thin strip of the stage preceding having regained a thickness
equivalent to the remainder of the blastoderm; its cells neverthe-
less are distinguishable from those of the remainder of the blasto-
derm by their irregular form. In general the blastoderm cells are
regularly prismatic, much shorter and more compact in structure
than in the preceding stage, and slightly constricted near their
bases (Fig. 12F). The latter are, as before, fused to form a
continuous layer, but this is now thin, dark and granular, ap-
parently composed of material derived from the inner cortical
layer. Each cell still possesses an elliptical vacuole lying centrad

40 THE EMBRYOLOGY OF THE HONEY BEE

of the nucleus, in those cells situated on the dorso-lateral faces
of the egg the vacuoles are larger and much more prominent.

Thirty-two to thirty-four hours. (Fig. 183A and B.) This
stage is distinguished from the preceding principally by two
points of difference: first, the blastoderm cells now form a
columnar epithelium, they are no longer joined by their bases
but limited at their central ends by a well defined basement mem-
brane; second, the mid-dorsal area has again become reduced to
a thin sheet of flattened cells. Its point of juncture at the an-
terior end with the thick blastoderm of the ventral side has now
become shifted to a point ventrad of the cephalic pole (Fig. 18A)..
The ventral blastoderm has therefore undergone a further short-
ening in the long axis of the egg. Of the inner cortical layer
there remains a vestige in the form of a very thin granular layer
lying near the inner side of the blastoderm.

Since this stage closes what may be called the blastodermal
period of the development, and brings the account down to the
formation of the germ layers—which is in fact inaugurated dur-
ing this stage—a brief discussion of some of the phenomena
characterising this period is in order.

Among other features, the behavior of the blastoderm cells
toward the inner cortical zone deserves especial mention. To
review this process briefly: the cleavage nuclei enter the (outer)
cortical layer accompanied by a portion of the cytoplasm belong-
ing to the cleavage cells, but leaving another darker portion
within to form a basal layer, the inner cortical layer, by which
the newly formed blastoderm cells are at first united to each
other. This is next cut off by the formation of a basement mem-
brane. After repeated divisions the cells elongate, the basement
membrane disappears and the inner cortical layer is taken up by
the cells (22-26 hours). A new basement membrane is then
formed (32-34 hours), when the cells have the arrangement and
appearance of ordinary columnar (prismatic) epithelial cells.
Turning again to the accounts of the development of Hydro-
philus (Heider 1889) and Musca (Noack 1901), similar phe-
nomena are found to exist in these forms. In Hydrophilus the
inner cortical layer is taken up into the blastoderm cells in pre-
cisely the same manner as in the honey bee. Moreover the blasto-
derm cells afterwards present the same elongated form and con-

d to form

The middle plate (MP) is also

A, Anterior part of
ine

The point of junction of the ventral blastoderm
B, Transverse section

, t
oy Se Lee Ay

$

~

mn

3)

<s)

nN

=

cB)

12)

Se] cB)

r) us
nn i
a Pes

a2 8

x

Ei She)
ps 3 a,
tia Fh a
ro) 2
3548.4
° “own
24,50 >
be FUE 0
8D ogn a
op AX.
oo 2S x
Ww nw

tet ie PB)
og hc es
asa 8&4
Goa ~S
3 Vo Sa
rt tere 2 Due > no = = MA
og OND
Msc.
PSE Sy
SS y Oey
ge ae
6 w 3 &
Bosses
EEsS.o

Te a | Pe ee age a ee a ae

42 THE EMBRYOLOGY OF THE HONEY BEE

stricted base. There is however, this difference: that in Hydro-
philus the blastoderm cells maintain continuously their connec-
tion with the inner cortical layer, and are never separated from
it by a basement membrane. In Musca, however, in certain parts
of the egg, the inner cortical layer is separated from the blasto-
derm cells by a layer of yolk granules. By the inward (centrad)
extension of the blastoderm cells, this layer of yolk, as well as
the inner cortical layer, is taken up into the bases of the cells.
This process seems to take place much more slowly and at a
later period than in Hydrophilus. In this respect, as well as in
the separation of the inner cortical layer from the blastoderm
cells, Musca may be said to approach the honey bee more closely
than Hydrophilus.

Another feature of interest is the long duration of the blasto-
dermal period—16 to 18 hours, approximately, or nearly one
quarter of the total length of time required for complete de-
velopment in the egg. In Hydrophilus (Heider 1889), the only
insect concerning which full and accurate data are given of the
duration of each stage, the period required for the formation
and completion of the blastoderm (from the close of cleavage to
the first appearance of the lateral plates) is about 27 hours. The
total time required for development in the egg is 11 days, or
274 hours. The blastodermal period then occupies a trifle more
than one-tenth of the entire time, as compared with one-fourth
in the honey bee. This comparison, however, is perhaps a little
unfair, since the larva of Hydrophilus at hatching is fitted for
an active existence and probably therefore more highly differ-
entiated than the larva of the bee at a corresponding stage;
nevertheless, the comparison is a striking one.

The significance of the long sojourn of the bee’s egg in the
blastodermal period can only be guessed at; its importance can
only be estimated by its duration. Compared with the changes
cccurring within the egg in any other equal space of time the
morphological changes during this period are insignificant. This
leads naturally to the surmise that the nature of the changes
undergone by the egg at this time may be principally physiologi-
cal, and therefore not made evident by the ordinary methods of
the embryologist. |

V

THe GERM LAYERS

1. Formation of the mesoderm

At the time when the germ layers begin to be formed, (32-34
hours) the blastoderm presents three more or less distinct di-
visions: (1) a median dorsal strip, (2) the blastoderm of the
ventral and lateral surfaces, (3) two bands, one on each side of
the dorsal strip (1), between the latter and (2).

The median dorsal strip (Fig. 18A, DS) extends to the caudal
pole, from a point slightly ventral of the cephalic pole. It is
widest near the cephalic pole, attaining here a width of about
one-third of the circumference of the egg in this region, caudad
of this point it narrows gradually and over the remaining distance
continues quite narrow (Fig. 18B). It is composed of a relatively
small number of extremely thin and flat cells, and also differs
from the remainder of the blastoderm in its close relation to the
protoplasm surrounding and permeating the yolk. As may be
readily determined in most preparations the yolk is enclosed in
a pellicle of protoplasm, derived from the inner cortical layer,
forming a part of the interstitial protoplasmic meshwork and
continuous with it. This pellicle is commonly separated from the
basement membrane of the blastoderm by a narrow space. This
relation, however, ceases at the lateral edges of the dorsal strip,
since here the protoplasmic pellicle unites and becomes continu-
ous with the lateral edges of the cells composing the strip. This
connection becomes especially evident during the formation of
the germ layers and the amnion. As a result of the disappear-
ance of the pellicle under the dorsal strip the latter supplies the
place of the pellicle and thus comes into close relation with the
yolk and the meshwork of protoplasm within it. The fate of the
dorsal strip will be dealt with later, but it may be said in advance
that it takes part neither in the formation of the embryo nor its
covering (amnion), and is therefore strictly non-embryonic.

43

44 THE EMBRYOLOGY OF THE HONEY BEE

The blastoderm covering the ventral and lateral faces of the
egg, as already described (p. 40), is a thick single layered epi-
thelium composed of slender prismatic cells. Its average thick-
ness is greatest in the anterior region of the egg and decreases
slightly and gradually toward the caudal pole. This division,
composing the major portion of the blastoderm, is the distinctly
embryonic portion, since it alone contributes directly to the for-
mation of the embryo.

On each side of the dorsal strip is a band comprising some
four to six longitudinal rows of cells whose central ends are very
clear and transparent (Fig. 19, Adm). Those cells next to the

MP
(Meso)
Fic. 19. Transverse section through the anterior region of an egg,

Stage IV, showing dorsal strip of blastoderm (DS), amnion forming cells
(Am), middle plate (MP) and lateral plates (LP), x 243.

dorsal strip are triangular in outline when seen in transverse sec-
tion, and in contact with the dorsal strip at one corner; all the
cells of this strip are less slender than the cells of the embryonic
blastoderm (2), many of them approaching the cubical form.
Those lying nearest the embryonic blastoderm intergrade with
the cells of the latter, so that the lateral limits of this division of
the blastoderm cannot be determined with precision until the
formation of the germ layers is well under way. These two

;

THE EMBRYOLOGY OF THE HONEY BEE 45

bands, which will later form the amnion together with the
strictly embryonic division, may be conveniently considered as

together constituting the ventral plate. Later, when the amnion

begins to be more clearly differentiated, the embryonic portion
of the ventral plate, which now constitutes the embryonic rudi-
ment, is commonly known as the germ band.

This stage is introduced by the appearance, on the ventral
surface of the blastoderm, of two narrow longitudinal ridges,
sharply defined on their inner margins, less so on the outer. As
seen in surface view, they appear as shown in figure IV, where
they have the appearance of gently curved and slightly irregular
dark lines with their convex sides facing one another, like re-
versed parentheses, )(, and separated from each other at the
point of closest approximation by a distance of about one-eighth
of the total circumference of the egg at this point: The distance
between their anterior ends and the cephalic pole of the egg, is
about equivalent to the egg’s diameter here; their length when
first visible from the exterior is, as shown in the figure, about
one-fourth of the total length of the egg. As development pro-
ceeds these ridges extend rapidly caudad, diverging slightly at
first and then pursuing a course nearly parallel to one another
as far as the caudal pole of the egg, embracing between them a
strip of blastoderm whose average width is almost one-sixth of
the circumference of the egg at any given point (Fig. V). This
strip is the middle plate, and constitutes the future mesoderm,
while the embryonic blastoderm laterad of it on each side forms
the lateral plates, whose mesial edges are the Jateral folds.

Simultaneous with the caudad extension of the lateral folds
(Stages IV-V) is a movement of their anterior portions toward
the mid-line. This movement begins first at about that point
where the folds originally approached one another most closely,
that is, about one-fourth of the length of the egg from its cep-
halic pole. The anterior ends of the folds meanwhile remain
stationary. The outline of the folds in consequence of their
movement toward one another near their anterior ends, forms a
figure more or less resembling an elongated flask (Fig. V). In
the corresponding stages of Hydrophilus (Kowalevski, 1871),
and Chalicodoma (Carriere and Birger, 1897), the resemblance
between the outline of the folds and that of a flask also occurs

46 THE EMBRYOLOGY OF THE HONEY BEE

and is much closer. The lateral folds continue to approach one
another until they meet, their first point of juncture being natur-
ally that of most rapid movement, a short distance caudad of
their anterior ends, which up to this time remain nearly station-
ary. The process of meeting and fusion now progresses both
forward and backward from the first point of juncture, the
anterior portions of the lateral plates being very quickly united,
while the posterior come together rather slowly. Figures V and
VI illustrate two stages in the completion of the process of clos-
ing; in the former figure a narrow cleft extending nearly one-
half of the length of the germ band separates the lateral plates;
in the latter figure this cleft is insignificant. With the final clos-
ure of this cleft the formation of the germ layers may be con-
sidered complete.

Examination of sections of the stages described shows that
the process just described consists essentially in the depression
of a median area of the ventral plate—the middle plate,—and its
overgrowth by the lateral portions of the ventral plate—the Jat-
eral plates;—which have broken away from its edges along the
line indicated by the lateral folds at the time of their first ap-
pearance. In respect to the manner of formation of the meso-
derm the observations recorded above are in complete agreement
with those of Kowalevski (1871) and Grassi (1884). As al-
ready mentioned (p. 40) the beginning of the process of meso-
derm formation (“gastrulation,” so-called) is indicated in figure
18B. In this figure a median section of the ventral blastoderm
is seen to be in process of separation from the lateral portions
by the mere displacement of the cells on each side of the bound-
ary line. That is, the cells forming the margin of the nascent
middle plate appear to be sliding inward over those forming the
edges of the future lateral places. This stage is the earliest ob-
served. Figure 19 represents a transverse section through the
anterior end of an egg at Stage IV. Here the middle plate has
become still more depressed, while the edge of the lateral plate
on the left side of the figure has already overlapped the cor-
rsponding edge of the lateral plate; on the right side of the
figure the relation between the middle and lateral plates is much
the same as in figure 18B. In sections like that represented by
figure 19 it is evident that the middle plate has been actually de-

SQ LS See 2 ee

om

—— Sr

Se

THE EMBRYOLOGY OF THE HONEY BEE ini

pressed, especially at its lateral margins, as the corresponding
deformation of the yolk shows, while the lateral plates are not
perceptibly lifted up from the yolk. Figures 20, A and B, il-
lustrate the stage next following. Figure 20A is a cross section
through the egg near the point when the lateral folds are closest
together; figure 20B is near their posterior ends. The middle
plate has at this stage changed somewhat in structure in its
anterior portions. The marginal cells, formerly like the remain-
der of the ventral blastoderm, slender and prismatic, with their
long axes directed radially, have turned laterad under the lateral
plates, becoming irregularly polyhedral in form. Corresponding
with this change the middle plate becomes much broader and
thinner, especially at its edges. Figure 20B illustrates the for-
mation of the middle plate in the posterior half of the egg. It
is evident at a glance that the process is identical with that in
the more anterior regions of the egg. There is, however, this
difference: that before the middle plate shows any signs of
separation from the lateral plates it is distinguishable by being
thinner than the lateral portions of the blastoderm, and its cells
are correspondingly shortened and widened. This is evident in
figure 20B. It is also evident that the initial width of the middle
plate is greater in the posterior than in the anterior regions of
the egg. This is probably correlated with the thinning of the
incipient ventral plate which brings about a corresponding ex-
tension of its surface.

The final stages in the formation of the mesoderm are shown
in figure 21, which is a transverse section through the middle of
an egg a trifle younger than Stage VI, showing the lateral folds
about to unite in the mid-line, the middle plate being now vir-
tually completely covered. Its width is—at this point—fully
equivalent to one quarter of the circumference of the egg; near
the ventral mid-line its thickness is approximately that of the
overlying lateral plates (ectoderm), it thins out rapidly to one-
half of this thickness at the lateral margins. Here it is composed
of but one cell layer; near the ventral mid-line it can be distinctly
seen to be made up of two layers. This arrangement of the -
mesoderm cells into two layers is more evident in sagittal sec-
tions passing somewhat laterad of the median plane (Fig. 23B).
It is also more evident in the posterior than in the anterior half

(Meso)

Fic. 20. Tranverse sections through egg, Stage V. A, near anterior,
B, near posterior end. The middle place (MP), lateral plates (LP),
amnion-forming cells (Am) and dorsal strip are shown, x 243.

THE EMBRYOLOGY OF THE HONEY BEE 49

of the mesoderm. In the rearrangement of the cells which are
to form the mesoderm, cell division plays an important part,
since cells in division are very abundant in the middle plate dur-
ing Stages V and VI.

In the formation of the mesoderm Apis differs somewhat from

(Meso)

Fic. 21. Transverse section through egg. Stage V-VI, near anterior
end, showing the middle plate (MP) virtually covered by the lateral
plates (LP), x 243.

most of the insects studied. Following Korschelt and Heider’s
classification (1891-3) the different modes of formation of the
mesoderm in pterygote insects may be distinguished as follows:

1. The formation of a tube by invagination of a median strip
of the ventral blastoderm. This tube, constituting the mesoderm,
may be thick-walled and round in section and consequently pos-
sessing only a narrow lumen, as in many Coleoptera, or thin-
walled and compressed in a dorso-ventral direction, as in Chali-
codoma (Carriére 1890). This type is probably the commonest
and best known of the three. It is found in representatives of
widely separated orders, for example: Hydrophilus (Kowalev-

50 THE EMBRYOLOGY OF THE HONEY BEE

ski 1871) Musca (Kowalevsky 1886), Pyrrhocoris (Graber
1888a), Chalicodoma® (Carriére 1890).

2. The middle plate remains nearly flat and is overgrown by
the edges of the lateral plates, which become free. This type is
relatively uncommon but is also found in representatives of
widely separated orders, as in Apis, Sphinx (Kowalevski 1871),
Pieris (Bobretsky 1878), Gasteroidea (Gasterophysa) (Lecail-
lon, 1897),° and, in a slightly modified form in Forficula (Hey-
mons 1895). | :

3. The mesoderm arises by proliferation and immigration from
a median area of the ventral blastoderm. A median groove is
often also present. This type approaches the conditions obtain-
ing in the Apterygota (Heymons 1897, Uzel 1898) and appears
to be especially characteristic of the Orthoptera, being found,
for example in Gryllotalpa (Korotneff 1885), Phyllodromia
(Cholodkowsky 1891b), Gryllus, Periplaneta (Heymons 1895).

It is accordingly evident that Apis, in the manner in which it
forms the mesoderm, while agreeing with representatives of
remotely allied groups, differs from its near relative Chalicodoima.
Very little systematic importance can therefore be accorded to
these differences. Carriére and Biirger, however, took the op-
posite view and on the basis of their observations in Chalicodoma,
declared their belief that the observations of Kowalevski and
Grassi in regard to the formation of the mesoderm were in-
correct on account of imperfect technique, and that a reinvesti-
gation of Apis with modern methods would show that the meso-
derm was formed in the same manner as in Chalicodoma. It is
now sufficiently evident that this assumption was unfounded, at
least so far as the eggs destined to form workers or queens are

*Carriére (1890, 1897) regards as the middle plate only the inner layer
of the two layers of cells which form the mesoderm, the outer layer being
considered as formed by the infolding of the mesial portions of the lateral
plates. Comparison with the corresponding stages of other insects in
which the mesoderm is found by a median invagination, as Hydrophilus
for example, indicates that this distinction is hardly justified, and that
in all cases the term middle plate may properly be considered as including
all of the infolded median portion of the ventral plate, and therefore all
of the material for the mesoderm.

* According to Hirschler (1909a) the mesoderm in Gasteroidea is
formed by infolding, as in Hydrophilus.

THE EMBRYOLOGY OF THE HONEY BEE 51

concerned. Petrunkewitsch (1902) has published a figure il-
lustrating a transverse section through a drone egg during the
stage of mesoderm formation in which is shown a distinct roll-
ing up of the edges of the middle plate while still attached to
the lateral plates. This condition tends to approach the type
found in Chalicodoma, and if it normally exists in the drone
could be readily interpreted as intermediate, connecting types
1 and 2, as illustrated by Chalicodoma and Apis respectively.

Fic. 22. Longitudinal (sagittal) sections through the anterior region
of the germ band, Stage V, showing segmentation. A is a section along
the mid-line, and includes only the middle plate; B is lateral of the
mid-line. The mesoderm is here now divided into two layers, x 387.

52 THE EMBRYOLOGY OF THE HONEY BEE

The writer has so far not studied the drone egg, and can there-
fore not make any statements in regard to it; a rolling up of the
edges of the middle plate was however not observed in any of the
preparations of the worker eggs.

Coincident with the process of mesoderm formation is another
phenomenon, that of segmentation or metamerism. ‘This is ob-
servable as early as Stage V, and on surface view is evident in
the form of alternating transverse light and dark bands or zones
extending across the ventral plate and appearing first in its
anterior region just behind the anterior mesenteron rudiment.

Three of these bands are to be seen in figure V. They make
their appearance in rapid succession from in front backwards
until at Stage VII they have reached the posterior region of
the egg, when at least fourteen of them can be counted. A
glance at the next stage, VIII, when the rudiments of the ap-
pendages and stigmata make their appearance, suffices to show
that the darker zones undoubtedly correspond to the definitive
segments of the embryo. Longitudinal sections of Stage VI
show that the alternating darker and lighter zones are only the
optical expression of alternating thinner and thicker zones in
both the middle and lateral plates; in other words the segmen-
tation affects both the future mesoderm and ectoderm (Figs.
22A and B, and 23A and B).

The segmentation of the middle plate is at best rather ill
defined, and most marked in the mid-line of those portions about
to be covered by the lateral plates (Fig. 22A). The latter show
a segmentation corresponding to that of the middle plate and ex-
pressed in the same manner, namely by a wavy contour of their
inner boundaries when seen in longitudinal section (Fig. 22B).
In figure V, where three dark zones are faintly visible at the
anterior end of the middle plate, the free edges of the lateral
plates at their anterior ends are lobed or scalloped in such a
way as to suggest segmentation, the lobes of the two opposite
sides corresponding to one another and also to the dark zones
of the middle part. The same is true of the lateral plates near
their posterior ends at the following Stage, VI. Segmentation
(metamerism) thus appears simultaneously in the lateral plates
before their union, in the flask stage, and is accompanied by a
corresponding segmentation of the middle plate. It begins first

- THE EMBRYOLOGY OF THE HONEY BEE 53

.MP LP
— 2 ARB eee

Fic. 23. Longitudinal (saggittal) sections, through the anterior region
of the germ band, Stage VI, showing segmentation. A is a section along
the mid-line, B is laterad of the mid-line. The mesoderm is here now
divided into two layers, x 387.
at the anterior end of the germ band and progresses backward,
following the rule for arthropods in general. Figure 22A il-
lustrates a median sagittal section of Stage V through that por-
tion of the middle plate lying just behind the anterior mesenteron
rudiment. Four segments are very plainly visible, much more
so than is usual. Figure 22B is taken from the same series and
passes laterad of the median plane through one of the lateral
plates. It shows very distinctly the segmentation of the future
ectoderm, four segments being also represented. The segments
are not however always sharply marked off from one another in
either the lateral or median plates and are frequently somewhat
irregular.

After the middle plate becomes overlaid by ectoderm its lateral

54 THE EMBRYOLOGY OF THE HONEY BEE

regions display a sort of secondary segmentation induced by
that of the ectoderm. The mesoderm is here of nearly uniform
thickness, in the later stage (V1) double-layered (Fig. 23B), and
thrown into a series of low transverse folds by the segmental
swellings of the overlying ectoderm. Near the mid-ventral line
the segmental swellings of the mesoderm are intact although in-
conspicuous, but instead of corresponding with those of the ecto-
derm they fit into the intersegmental depressions.

In general it may be'said that up to Stage VIII the segments
are not uniformly well defined, and their boundaries not sharp,
especially at the two ends of the germ band, so that certain
identification of individual segments is difficult if not impossible.
Those segments chosen for illustration were unusually well de-
fined. All that can be safely affirmed is the presence of seg-
mentation at this stage.

Biitschli (1870) observed and correctly interpreted these early
evidences of segmentation, although neither Kowalevski nor
Grassi appear to have noticed them. Bittschli’s account is of
sufficient interest to quote, and is as follows: “In addition to
these primary rudiments of the germinal ridges there are also
found the first indication of the segments. I had long overlooked
this precocious process, until I investigated the finer structure of
the germ band with a high magnification. In contrast to earlier
stages this shows no longer the regular cellular structure, but
transverse bands, in which the cells are pressed closely together,
alternating with the others which are elongated and with their
long axis placed in a transverse plane. On closer examination
one notices that this condition on the surface harmonizes with
the image seen in optical section, which shows swellings alter-
nately with thinner portions, where it is one layered, as well as
where it is many layered. Frequently in the contracted portion
between two swellings there appears a dark transverse line,
which seems to indicate a cleft through the entire thickness of
the germ band. The bands described, composed of closely packed
cells, form the median portions of the segments, the somewhat
larger, more elongated cells lie in the boundary between two
_ neighboring segments. By raising and lowering the microscope
tube, I have often persuaded myself that the first bands lie at
a higher level than the part lying between them so that ac-

THE EMBRYOLOGY OF THE HONEY BEE 55

cordingly the external surface of the germ band must also dis-
play evidences of segmentation in the form of a faintly wavy
contour” (p. 530).

It is not clear whether the difference just mentioned between
the superficial aspect of the intersegmental and intrasegmental
cells of the germ band applies to the ventral or lateral plates.
Both were examined in the most favorable preparations, but no
differences could be noted. In longitudinal sections of the ven-
tral plate, however, such as that represented by figure 22A, a
considerable difference between the form of intersegmental and
intrasegmental cells is noticeable, the latter being much nar-
rower at their outer ends than the former.

Both Kowalevski (1871) and Heider (1889) described a
precocious segmentation in the egg of Hydrophilus, the latter
investigator finding it expressed in the form of transverse folds
which appear even before the middle glate is found. Both
Kowalveski and Heider interpreted these as corresponding with
the future definitive segments. Carriére (1890) found that
the egg of the mason bee (Chalicodoma) also showed a precocious
segmentation essentially the same as that just described for the
honey bee. At the “flask” stage, or even earlier, segmentation
makes its appearance on the ventral plate, in surface views of
the egg, as dark transverse bands, alternating with lighter ones,
accompanied by a corresponding lobing of the lateral folds. The
anterior segments appear first, afterward those lying caudad, in
rapid succession. According to Birger’s statement (1897) seg-
mentation appears first in the lateral plates, and only later in the
middle plate. The individual segments become very sharply
marked out, much more so than in the honey bee. Since the
first rudiments of the embryo (antennae, mouth parts, etc.)
appear very early in Chalicodoma—as early as the “flask’”’ stage,—
and therefore long before the completion of the union of the
lateral folds, the early identification of the individual segments
is made possible. In both Chalicodoma and Apis it is important
to note that the segments appearing thus precociously are the
definitive segments of the insect (the so-called microsomites)
and that there is no previous separation of the germ band into
the larger divisions seen first by Ayers (1884) in Oecanthus,
afterwards by Graber (1888) in the representatives of several

56 THE EMBRYOLOGY OF THE HONEY BEE

orders, and termed by him “macrosomites,” on the occurrence of
which Graber has attempted to construct an elaborate theory of
segmentation applicable to the entire arthropod phylum.

2. Formation of the rudiments of the mid-intestine.’®

The mesenteron or mid-intestine in the honey bee is derived
from two rudiments, arising at the anterior and posterior ends
of the germ band respectively. The anterior mesenteron rudi-
ment, owing to its position, is much more readily observed and
studied than the posterior mesenteron rudiment, and therefore
will be described first.

It will be remembered that the lateral folds end abruptly
toward the cephalic pole, leaving vacant an area of the ventral
plate about as long as the egg’s diameter at this point, corre-
sponding to the anterior field of Chalicodoma (Carriére and
Birger 1897). At a stage a trifle older than Stage IV, on
surface view there may be observed a darker area along the
mid-line in this field. This darker area, narrow and rather
vaguely outlined at first, rapidly increases in size, density and
definiteness until at Stage VI it presents the appearance shown
in the figure (VI, AMR). Cf. also Figs. 24A and B). Its out-
line is that of a short ellipse, with its longer axis directed length-
wise of the egg; in stained preparations it is deeply stained and
very conspicuous. Its width approximates that of the anterior
end of the middle plate, with which, at this time, it has come
into contact. This is the anterior mesenteron rudiment. Fig-
ures 25A to D represent transverse sections through the middle
of this rudiment at four successive stages of its development.
The first of the series, A, is taken from Stage IV, just before
the anterior mesenteron rudiment becomes visible from the ex-
terior, and at its earliest recognizable stage. In the mid-line,
over an area whose breadth is about one-eighth of the circumfer-
ence of the egg at this point the blastoderm cells, hitherto long

” These have been very widely identified by embryologists as entoderm,
and accordingly termed such. Since however there is some doubt as to
the correctness of the homology of these rudiments with the entoderm
in other classes of animals it has seemed preferable to avoid the use of the
term “entoderm” in connection with the development of pterygote insects
and, for the present at least, to simply use the term “mesenteron rudi-
ments.” See discussion at end of this section.

THE EMBRYOLOGY OF THE HONEY BEE 57

at teats ering,
? my a0,

.

ots
Pd *

ace
-.

AMR

Fic. 24. Transverse sections through the anterior ends of two eggs of
Stage IV, intersecting the anterior mesenteron rudiment (AMR), illus-
trating the movement of the lateral blastoderm toward the mid-line during
the growth of this rudiment. The development of the amnion in this
region is also shown. In A the anterior mesenteron rudiment is barely
visible, in B, a slightly older stage, it forms an evident swelling, x 243.

58 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 25. Transverse sections through the anterior mesenteron rudiments
of four eggs, showing four stages in the development of the rudiment.
A is from an egg of Stage IV, B and C are from eggs of Stage IV-V,
C is from an egg of Stage V-VI, x 290.

prismatic in form, are now seen to be broader, more irregular in
form, and loosely arranged. Comparison of sections of a series
through the anterior end of an egg at this stage shows that the
area involved in the production of the anterior mesenteron rudi-
ment has the form of a narrow triangle lying in the mid-line;

a a ee ee
‘

Sr a Pe

eS ae.

_— eee FR i a erect i ~

ee

Retype

eS ee

THE EMBRYOLOGY OF THE HONEY BEE 59

its apex is near the anterior limits of the ventral plate, while

its base joins the anterior end of the middle plate.

In the next figure, 25B, taken from an egg intermediate be-
tween Stages IV and V the mesenteron cells are seen to have
increased greatly in number, now forming collectively a low
rounded swelling on the inner side of the blastoderm in the
mid-line. It is at this stage that the anterior mesenteron rudi-
ment first becomes visible from the exterior. The arrangement
of the component cells is still loose, numerous interstices re-
maining between them. In form the cells are, generally speak-
ing, ovoid or rounded, although somewhat irregular. On one
side of the nucleus a vacuole is frequently to be seen, as in the
adjacent blastoderm cells, which they closely resemble in all
respects except that of form. As the cells of the anterior
mesenteron rudiment continue to increase rapidly in number
(Fig. 25C), the rudiment increases correspondingly in thickness
and breadth, so that it spreads out laterally. At the same time,
its cells become compactly arranged and polyhedral in form,
due to mutual pressure doubtless caused by the necessary dis-
placement of the underlying yolk. Figure 26A represents a
longitudinal section through the anterior mesenteron rudiment at
about this period (Stage IV-V). The rudiment extends cep-
halad to within three or four cells of the anterior end of the
ventral plate, where it ends abruptly, its cephalic face perpen-
dicular to the surface of the blastoderm, leaving in front of the
rudiment a cavity bounded on its inner side by the yolk, on its
outer by the amnion (Am) and by the short stretch of unmodi-
fied blastoderm cephalad of the mesenteron rudiment. Caudad
the latter diminishes gradually in thickness to join the middle
plate (MP).

Figure 25D shows the rudiment after the lateral folds have
commenced to unite (Stage V-VI) being drawn from the same
preparation as figure 21. The anterior mesenteron rudiment
has now increased considerably in the number of its cells and in
its lateral extent, coming into close contact with the inner sur-
face of the lateral blastoderm on both sides of the mid-line. In
addition to its increase in size another change is becoming evi-
dent, which has the appearance of an encroachment of the
lateral unmodified blastoderm (ectoderm) on the external layers

60 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 26. Median sagittal sections through the anterior mesenteron rudi-
ments of two eggs, illustrating the development of the rudiment and its
relations to the adjacent ectoderm (Ect) and mesoderm (Meso). The
formation of the amnion (Am) in this region is also shown. A is from an
ege of Stage IV-V, B from an egg of Stage VI, x 290.

of the anterior mesenteron rudiment, thereby gradually restrict-
ing its superficial area, and leading finally to the complete cover-
ing of its external surface by ectoderm, with the exception of a
circular area near its posterior border (Fig. 26B). The precise
nature of this process is uncertain. Carriére (1890, 1897) has
stated that in the mason bee it consists in a centripetal differen-
tiation of the superficial cells of the rudiment into prismatic
ectoderm cells, but in the honey bee there are certain concomit-
ant phenomena not evident in the mason bee, which make it
seem possible that the covering of the anterior mesenteron rudi-
ment is brought about by a simultaneous mesiad movement of
the two halves of the ventral plate separated by the rudiment.
This covering by the ectoderm is completed first at the anterior

THE EMBRYOLOGY OF THE HONEY BEE 61

narrow end of the superficial part of the rudiment; it then
progresses rapidly both caudad and mesiad at the same time.
This process occupies the interval between Stages V and. VII.
It is important to note that the mesiad progress of the lateral
ectoderm over the anterior mesenteron rudiment and the ap-
proximation of the anterior ends of the lateral plates over the
middle plate takes place at the same time and goes on at about
the same rate.

The nuclei of the mesenteron cells, spherical in form from the
inception of the rudiment, now begins to be more clearly distin-
guishable from those of the surrounding tissues, being dis-
tinguished not only by their circular outline but also by their
somewhat greater size and paleness. These differences serve as
useful means of identification during the succeeding stages of
embryonic development.

The form and relations of the anterior mesenteron rudiment to
the adjacent parts at Stage VI when the ectoderm is completely
formed over the anterior half of the germ band is shown in
figure 26B, which represents a median longitudinal section
through the anterior end of the germ band. The anterior mes-
enteron rudiment is here seen to be thick lenticular in form, its
anterior edge projecting out from beneath the anterior margin
of the overlying ectoderm. Betweeen the latter and the ecto-
derm (Ect) which overlies the mesoderm (Meso) is a rather
wide gap representing a rounded break in the continuity of the
ectodermal covering of the germ band. This marks the place of
origin of the future mouth. Through this gap or orifice, cells of
the posterior edge of the anterior mesenteron rudiment are seen
to come to the external surface, filling the space with a plug of
cells. A similar relation between the anterior mesenteron rudi-
ment and the overlying ectoderm occurs in Chalicodoma (Car-
riere and Birger 1897, and between both mesenteron rudiments
and ectoderm in Gasteroidea (Hirschler 1909a). This plug
however is not composed exclusively of the cells of the anterior
mesenteron rudiment, since a few mesoderm cells, distinguished
by their smaller nuclei, are seen in its posterior portions. Pos-
teriorly the anterior mesenteron rudiment is very closely united
with the anterior end of the mesoderm, no sharp line of separa-

62 THE EMBRYOLOGY OF THE HONEY BEE

tion being visible, the only distinguishing character being the
differing size of the nuclei.

The source and manner of origin of the anterior mesenteron
rudiment remain to be considered. The possibility of its deriva-
tion from yolk cells is excluded by the simple fact that yolk
cells are never present in this region in sufficient numbers to
form such a structure, moreover, the close relation of the rudi-
ment with the blastoderm and the similarity of its cells to the
cells of the blastoderm make the assumption of any other source
than the blastoderm impossible.

Carriére (1890) and Carriére and Burger (1897) state that
the mesenteron rudiments in the mason bee arise from the blasto-
derm by proliferation. In the honey bee however during the
earlier stages, comprising the period of its most rapid growth,
when the greater portion of its mass is formed, cell divisions are
virtually absent in that part of the ventral plate from which
the mesenteron rudiment arises. This is particularly significant,
when it is considered that this period is very brief, extending
from Stages IV to V. Since, therefore, the anterior mesoderm
rudiment does not arise by proliferation it must be assumed that
it arises by immigration. This is strongly suggested by the ap-
pearance seen in sections like that represented in figure 25B. At
this stage the cells bordering the anterior mesenteron rudiment
laterally are plainly seen to intergrade with those of the blasto-
derm. This is also true of the section represented in figure
25C, but is perhaps less evident.

Two phenomena coincident with the formation of the anterior
mesenteron rudiment take on a special significance when con-
sidered in association with the facts just mentioned. These are:
first, that the appearance of the anterior mesenteron rudiment
and the depression of the middle plate are contemporaneous ;
second, that the two lateral halves of the ventral plate bounding
the mesenteron rudiment appear to move mesiad during the
period when the lateral plates are coming together to cover the
middle plate, or, to express it in a different way, the lateral
halves of the ventral plate bounding the anterior mesenteron
rudiment behave toward the latter much as the lateral plates
behave toward the middle plate. Figures 24A and B illustrate
this point. Figure 24A is drawn from a section through the

THE EMBRYOLOGY OF THE HONEY BEE 63

anterior end of an egg at Stage IV; figure 24B is drawn from
a section through the same region at a stage about half way be-
tween Stages IV and V. The mesiad movement of the two
lateral halves of the ventral plate in this region during the inter-
vening period becomes at once evident, since the area covered
by the dorsal strip and the amnion-forming cells has become
greatly extended transversely, accompanied by a flattening of
the amnion-forming cells. The possibility that this movement
is only an apparent one, and caused by a contraction of the
lateral blastoderm is excluded, since the blastoderm cells are
not perceptibly lengthened in the section taken at the later
stage, as would be the case if an actual contraction had taken
place.

It appears then, in view of the facts outlined above, that the
anterior mesenteron rudiment is derived from the cells of the
ventral blastoderm by immigration, and that this rudiment is
continuous with, and comparable to, the ventral plate, in so far
as both are formed contemporaneously from the. middle section
of the ventral plate, accompanied by a mesiad movement of its
lateral sections. Moreover, the anterior mesenteron rudiment
and the ventral plate are directly continuous with one another,
the chief differences between them being that the former arises
as a heap of cells not sharply marked off from the remainder
of the blastoderm, while the latter arises as a solid flat section of
the blastoderm, discontinuous with the lateral blastoderm and
separated from it by a sharp break.

Cell division, although apparently playing no part in the forma-
tion of the anterior mesenteron rudiment during its earlier
stages, steps in during the later stages, when mitotic figures are
frequently seen, being most abundant in the posterior part of the
rudiment. Two mitotic figures are evident in this region in
figure 26B.

The manner in which the ectoderm finally closes over the an-
terior mesenteron rudiment is, as before mentioned (p. 60), ob-
scure, and repeated efforts finally to decide this question were
without definite result, nevertheless, since it is certain that the
lateral blastoderm bounding the anterior mesenteron rudiment
moves mesiad during the early stages of the rudiment it seems
possible that the final covering of the external surface is brought

64 THE EMBRYOLOGY OF THE HONEY BEE

about in this manner, Sections like that represented in figure
25D lend color to this supposition, since here the mesial edges
of the lateral blastoderm are marked off from the mesenteron
cells with a considerable degree of sharpness.

The posterior mesenteron rudiment is much more difficult to
study than the anterior rudiment, for two reasons; first, being
formed near the posterior pole, in the course of its development
the lengthening of the germ band carries the rudiment over the
rounded posterior end towards the dorsal side, so that sections
normal to the surface of the germ band at this point are seldom
obtained, and second, the posterior end of the egg, being the
one by which it is attached to the floor of the cell of the comb,
is frequently damaged. For the first of these reasons, longitudi-
nal sections are more informing and more useful than transverse
sections, which are commonly cut at right angles to the long
axis of the egg, and therefore only rarely intersect the posterior
mesenteron rudiment at right angles to its own long axis. Asa
matter of fact, scarcely an satisfactory transverse sections
through this rudiment were obtained.

The development of the posterior mesenteron rudiment is
fundamentally identical with that of the anterior mesenteron rudi-
ment, but differs greatly in details. Unlike the anterior mesente-
ron rudiment, the posterior mesenteron rudiment is ordinarily not °
evident on preparations of entire eggs, except during the conclud-
ing stages of its development, when it appears as a deeply stained
discoid mass at or dorsal to the caudal pole. Prior to Stage V
the ventral plate extends over the caudal pole to the dorsal sur-
face as a single layer of low cells, rather irregular in form, and
rounded on their external surface. This layer is frequently in-
terrupted by short gaps which leave bare the yolk beneath. At
Stage V the first indication of the posterior mesenteron rudiment
becomes visible as a slightly thickened area of the blastoderm
just cephalad of the caudal pole, on the ventral surface (Fig.
27A, PMR). The cells composing this thickening are now long
prismatic in form, instead of low and rounded as before. The
posterior portion of the ventral plate has apparently meanwhile
contracted, at least in a longitudinal direction, since it now ex-
tends only a short distance dorsad to the caudal pole. At the
stage following, Stage VI (Fig. 27B), the posterior mesenteron

Fic. 27. Median sagittal sections through the posterior ends of three
eggs, to illustrate the formation of the posterior mesenteron rudiments

_ (PMR). A is from an egg of Stage V, B from an egg of Stage VI, C
_ from an egg of Stage VII, x 290.

66 THE EMBRYOLOGY OF THE HONEY BEE

rudiment is seen to have increased greatly in thickness. It now
forms a discoid mass lying on the ventral surface of the egg,
its posterior boundary lying almost precisely at the caudal pole.
Its extent is approximately the same as that of the anterior
mesenteron rudiment. As the illustration shows the posterior
mesenteron rudiment is now composed of long irregularly bent
and curved fusiform cells, some of which at least extend through-
out the entire thickness of the layer. At the next stage, Stage
VII (Fig. 27C), four changes are seen to have taken place. (1)
The rudiment, by a lengthening of the germ band, has been
shifted around the caudal pole of the egg to the dorsal surface.
(2) It is no longer composed of long fusiform cells, but of
polyhedral cells precisely like those of the anterior mesenteron
rudiment. (3) The ectoderm now covers its anterior half. (4)
It has increased greatly in thickness, which approximates one-
half its diameter. The question as to the manner in which the
posterior mesenteron rudiment becomes covered by ectoderm is
even more difficult to answer than in the case of its counterpart
at the anterior end. That this covering is brought about by ex-
tension of the ectoderm at the expense of the superficial cells
of the rudiment is strongly suggested by the section represented
in figure 27C. It is certain at least that the extension takes
place in an antero-posterior direction with regard to the embryo
itself,—and that the ectodermal covering of the rudiment is
absolutely continuous with that of the lateral plates, since in
one series (that from which Fig. 28 was drawn) the cleft still
separating the lateral plates can be followed around the caudal
end of the egg to the posterior limit of the ectoderm. Figure 28
represents a transverse section through the posterior end of an
embryo of Stage VII, and intersects the posterior mesenteron
rudiment about midway of its length. On the ventral side are
seen the lateral plates (LP) still separated by a narrow cleft, and
lying above them, the middle plate (MP), or mesoderm. The
dorsal half of the section is occupied by the massive posterior
mesenteron rudiment (PMR), connected on each side with the
lateral plates by a thin sheet of cells, the amnion. This figure
illustrates an important difference between the anterior and
posterior mesenteron rudiments. While the former is produced
only by a relatively narrow median strip of blastoderm, the lat-

THE EMBRYOLOGY OF THE HONEY BEE 67

Fic. 28. Transverse section through the posterior end of an egg of
Stage VII, showing the posterior mesenteron rudiment (PMR), and also
the relations of this rudiment and of the lateral plates (LP), middle
plate (MP), and amnion (Am), to each other, x 290.

ter demands for its production the entire posterior end of the
ventral plate, leaving at the sides no undifferentiated blastoderm
(ectoderm). In the case of the anterior mesenteron rudiment
the ectodermal covering progresses mesiad from the mesial edges
of the two halves of the ventral plate bounding the rudiment
laterally. In the case of the posterior mesenteron rudiment the
ectodermal covering progresses caudad from the posterior limits
of the lateral plates, appearing as if they were extended back-
ward to cover the posterior mesenteron rudiment. The external
surface of the latter is, however, never entirely covered, its
posterior end (relative to the embryo) apparently always re-
maining uncovered. Moreover the ectodermal covering forms
@ continuous and unbroken sheet to its caudal limits. There is
no break in its continuity corresponding to the future proctodaeal
invagination and comparable to the uncovered area of the an-
terior rudiment. This difference is of importance, since it fore-
shadows the differing relations of the fore- and the hind-intestine

68 THE EMBRYOLOGY OF THE HONEY BEE

to the mid-intestine in the larva, when communication. between
the mesenteron (mid-intestine) and hind-intestine is completely
cut off. .

The source of the posterior mesenteron rudiment is obviously
the posterior end of the ventral plate; its mode of origin is
much more clear than in the case of its counterpart at the an-
terior end. In the earlier stages of the formation of the pos-
terior mesenteron rudiment, as in the anterior rudiment, mitotic
figures are rare, so that cell division cannot be regarded as an
important factor in the earlier stages of its formation. The
posterior end of the ventral plate shares in the general de-
crease in breadth associated with the formation of the germ
layers, moreover, it also contracts in a longitudinal direction.
Decrease in the length and breadth of the ventral plate at its
posterior end together with a corresponding elongation of its
component cells may then be safely set down as the factors first
concerned in the formation of the posterior mesenteron rudi-
ment. During the later stages (VI-VII) cell division becomes an
increasingly important factor in its growth, as in the anterior
rudiment. At Stage VII mitotic figures are exceedingly abund-
ant, as figures 27C and 28 show.

In conclusion then it may be said that both the anterior and
the posterior mesenteron rudiments are formed from the blasto-
derm of the ventral plate by a movement inward of its cells. In
the case of the anterior rudiment the migration takes place by
the detachment of cells from a limited area in the mid-line. The
rudiment is afterward covered over either by a movement of
the lateral halves of the remaining blastoderm toward the mid-
line, or by additions to their mesial borders from the superficial
cells of the rudiment itself. In the case of the posterior rudi-
ment the entire posterior end of the ventral plate is involved,
being moulded by changes, first in the form and next in the
arrangement of its cells, to form the posterior mesenteron rudi-
ment. The ectoderm, which later covers it, is continuous with
the posterior ends of the lateral plates, and appears as a caudad
extension of them, although it cannot be stated with certainty
whether they are formed by an actual caudad extension or by
the addition of new material from the external surface of the
rudiment.

THE EMBRYOLOGY OF THE HONEY BEE 69

It cannot fail to be noticed that there is a fundamental simi-
larity between the formation of the mesenteron rudiments and
of the mesoderm, since all three arise by an inward movement of
cells of the ventral plate. Moreover, the latter is continuous
cephalad and caudad with the mesenteron rudiments during the
period of their formation.

The origin of the mesenteron in the honey bee was first dis-
cussed by Biitschli (1870). Biitschli believed it to be produced
like the blastoderm by the so-called “free cell formation.”* Re-
ferring to a stage nearly corresponding to that numbered VIII
in this paper he says (pp. 540-541) “the formation of the mid-
intestine has already commenced, yet, as I often observed in
much more advanced embryos, only on the dorsal side of the
yolk ; seen in optical section, it is a single layer of close pressed,
yellowish cells. It appears to me that this cell layer takes its
origin near the ends of the germ band, since its thickness is
here greater than in the middle region. I have not succeeded in
observing its beginning, but in regard to it must conclude with
Zaddach and Weismann that the mid-intestinal wall develops
by free cell formation and not by delamination of an inner cell
layer.” It is interesting however to find that Biitschli both noted
and figured the anterior mesenteron rudiment, although without
recognizing it as such. He says concerning it (p. 529): “In
connection with the regression of the germ band at the anterior
pole is a thickening found there, concerning the true significance
of which I am not sure; seen en face it appears rounded, in
profile, it projects inward, yet I believe that the hemispherical
swelling represented in figure 10B is not in reality such, but that
the lateral extension of the thickening has produced an optical
illusion.” The figure referred to corresponds to a stage about
midway between Stages V and VI.

Kowalevski (1871) believed that the mesenteron in the bee
owed its origin to the inner or splanchnic layer of the mesoderm.
He was evidently led to this belief by the fact that in sections the

“ This consists in the spontaneous formation of nuclei within a proto-
plasmic matrix. It is hardly necessary to say that such a process does not
exist, every cell arising by division from a preexisting cell. The belief
was due to the imperfect technique employed by the investigators.

70 THE EMBRYOLOGY OF THE HONEY BEE

mesenteron at later stages is seen lying close beneath the lateral
portions of the mesoderm.

Grassi (1884) gave an essentially correct account of the origin
of the mesenteron in the honey bee, and one which has been
generally accepted as such. His observations of the development
of the anterior mesenteron rudiment were much more satisfac-
tory and complete than those of the posterior rudiment. The
substance of his statement in regard to it is as follows: After
the formation of the mesoderm (middle-plate) has begun, the
median part of that area of the blastoderm anterior to the fur-
rows (lateral folds) becomes many layered, with the exception
of its anterior margin. Later, beginning at the lateral margin
ef the many-layered portion, and perhaps also at its anterior
margin, the superficial layer is separated from the deeper layers.
This superficial layer is continuous with the one-layered blasto-
derm at the anterior pole, and is also continuous caudad with .
the ectoderm and is itself ectoderm. The deeper layers are con-
tinuous caudad with the mesoderm and are themselves meso-
derm. The formation of the posterior mesenteron rudiment was
supposed to be similar to that of the anterior rudiment. One
entire plate (Pl. VI) containing thirty-six figures of sections is
devoted to the development of the mesenteron rudiments. It
is to be noted, however, that Grassi makes no statement as to
the manner in which these rudiments are produced, and also
failed to note the break in the ectoderm covering the anterior
mesenteron rudiment, where the mouth is formed later. More-
over, his belief that these rudiments are to be interpreted as
mesoderm is, in view of subsequent investigation, scarcely justi-
fied; nevertheless his account of the origin of the mesenteron
rudiment is probably the most important part of Grassi’s paper,
and also the one which attracted most attention.

Carriere and Buirger’s (1897) account of the formation of the
mesenteron rudiments of the mason bee is substantially the same
as that of Grassi, differing from it principally in interpretation.
According to these investigators the anterior and posterior me-
senteron rudiments arise from proliferating areas of the un-
differentiated blastoderm at the two ends of the ventral plate,
respectively cephalad and caudad to the middle plate and in-
dependent of it. These rudiments constitute two large hemi-

~

THE EMBRYOLOGY OF THE HONEY BEE 71

spherical cell masses of which the posterior is the larger, their
convex surfaces directed inward. Later the superficial layer
of these rudiments is modified to form ectoderm, with the ex-
ception of a small area situated over the center of the anterior
rudiment, which remains unmodified and actively proliferating,
and which is to form the floor of the stomodaeal invagination.
This area corresponds to a similar area over the anterior
mesenteron rudiment of the honey bee, described above. Car-
riére and Burger, in opposition to Grassi, insist strongly on the
independence of the mesenteron rudiments from the mesoderm,
and consider those parts of the ventral plate from which the
mesenteron rudiments take their rise as purely blastodermal.

Only one other paper in the social Hymenoptera remains to be
mentioned in this connection, that of Dickel (1904) on the honey
bee. This is unique in that it seeks the origin of the “entoderm”
in a peculiar discoid cell mass appearing at the anterior end of
the egg during the earlier stages in the formation of the germ
layers, and derived from yolk cells, and therefore turned “‘yolk
plug” or “yolk syncitium.” A corresponding “yolk plug’ was
assumed by Dickel to exist at the posterior end of the egg.
These cell masses are supposed to be later carried inward by
invagination and to constitute the anterior and posterior mesen-
teron rudiments. In a succeeding section the origin and fate of
the “yolk plug” of Dickel will be discussed in detail; it is suffi-
cient to state here that it has no connection with the mesenteron
rudiments. The invagination figured by Dickel at the anterior
end of the egg can readily be construed as an artifact, since such
infoldings are very common in eggs of the bee which have not
been properly handled and are frequently produced by the os-
motic pressure of a clearing agent, such as cedar oil, when
incautiously used.

The origin of the mesenteron of insects has for the past forty
years been the subject of numerous investigations and also a
prolific source of discussion, from which the partisan spirit has
not been altogether absent. In the history of embryological re-
search there is perhaps no problem about which there has been a
greater diversity of opinion, and it is a regrettable fact that even
at the present day investigators of this subject are still arrayed
against one another in opposing camps. This is still more regret-

72 THE EMBRYOLOGY OF THE HONEY’ BEE

table since it is a recognized fact that much of this want of agree-
ment rests on differences of interpretation. These various obser-
vations and interpretations have been frequently discussed, often
at great length, in the various textbooks and the papers which
have dealt with this phase of insect embryology, so that a pro-
longed review of the individual papers appears superfluous, but
in order to gain an insight into this perplexing subject another
means has suggested itself, which is embodied in the following
classified list or table of the different investigators who have dealt
with the origin of the mesenteron. It shows (1) the particular
view adopted, (2) its adherents in order of the dates of publica-
tion of their papers, (3) The genus or genera of insects on
which the observations were made. It is realized that such a
table fails in many instances to represent the differences as re-
gards details, but on the other hand these could not be brought
out except in a prolonged discussion, and in any event should
be sought in the original paper. It is also realized that this
table may not be altogether wanting in inaccuracies and oniis-
sions, but a conscientious effort has been made to reduce these |
to a minimum by consulting the original papers wherever pos-
sible. If this table does nothing more it will serve at least to
throw into relief the opposing views and interpretations, and
also to illustrate their diversity.

(I) Mesenteron derived from yolk cells.
Dohrn (1866, 1876).

Butschli (1870) Apis.

Mayer (1876).

Bobretzsky (1878) Pontia (Pieris).
Graber (1878).

Balfour (1880).

Hertwig (1881).

Weismann (1882) Rhodites.
Tichomiroff (1882) Bombyx.
Ayers (1884) Oecanthus, Telias.
Patten (1884) Neophalax.
Korotneff (1885) Gryllotalpa.”

” Korotnefft believed that only the embryonic mesenteron was formed
by yolk cells, the functional or larval mesenteron owing its origin to
blood cells.

THE EMBRYOLOGY OF THE HONEY BEE sie,

Will (1888, 1888a) Aphis.

Tichomirowa (1890) Chrysopa..

Tichomiroff (1890) Calandra.

(1892) Bombyx and Calandra.

Tichomirowa (1892) Pulex.

Heymons (1897) Lepisma and Campodea.

Claypole (1898) Anurida.

Tschuproff (1903) Epitheca and Calopteryx (Median section
of mesenteron only.)

Dickel (1904) Apis.

(II) Mesenteron derived from the lower layer (mesoderm, en-
tomesoderm, primary entoderm).

1. Derived solely from anterior and posterior sections.

Grassi (1884) Apis.

Kowalevsky (1886) Calliphora (Musca).

Graber (1889) Lucilia, Calliphora (anterior rudiment only)
Lina.

Wheeler (1889) Blatta, Leptinotarsa (Doryphora).

Ritter (1890) Chironomus.

Heider (1888) Hydrophilus.

Karawaiew (1893) Pyrrhocoris.

Kulagin (1897) Platygaster.

Escherisch (1900) Calliphora.

Petrunkewitsch (1902) Apis.

Schwangart (1904) Endromis, Zygaena.

Nusbaum and Fulinski (1909) Gryllotalpa.

2. Derived from anterior and posterior sections and also from
a median section.

Nusbaum (1886) Periplaneta.

Nusbaum and Fulinski (1906) Phyllodromia.

Hirschler (1905) Catocala.

(1909) Donacia.
(1909a) Gasteroidea (Gasterophysa).

3. Derived from anterior and posterior sections of mesoderm
and also from splanchnic (inner) layer.

Heider (1885) Hydrophilus.

Cholodkowsky (1891c) Phyllodromia (Blatta).

4. Derived from splanchnic layer of mesoderm by delamina-
tion of two lateral bands.

74. THE EMBRYOLOGY OF THE HONEY BEE

Kowalevski (1871) Apis, Hydrophilus.

Tichomiroff (1879) Bombyx.

Cholodkowsky (1888) Phyllodromia (Blatta).

Graber (1888a) Stenobothrus, Lina.

5. Derived from splanchnic layer of mesoderm by delamina-
tion and also from a median section.

Nusbaum 1888) Meloé.

6. Derived from median section of mesoderm only.

Hammerschmidt (1910) Dixippus.

(III) Mesenteron derived from proliferations of the blind
inner ends of the stomodaeal and proctodaeal invaginations.

Ganin (1874).

Witlaczil (1884) Drepanosiphum (Aphis).

Voeltzkow (1888, 1889, 1889a) Calliphora (Musca), Melo-
lontha.

Graber (1889) Lucilia, Calliphora (posterior rudiment only).

Graber (1891) Stenobothrus.'*

Graber (1891b) Gryllotalpa, Meloé.

Korotneff (1894) Gryllotalpa.

Heymons (1894, 1895) Forficula, Gryllus, Gryllotalpa, Peri-
planeta, Phyllodromia (Blatta), Ectobia.

Heymons (1897a) Bacillus.

Lecaillon (1898) Clytra, Gasterophysa, Chrysomela, Lina,
Agelastica.

Rabito (1898) Mantis.

Schwartze (1899) Lasiocampa.

Toyama (1902) Bombyx.

Pratt (1900) Melophagus.

Deegener (1900) Hydrophilus.

Tschuproff (1903) Epitheca and Calopteryx (anterior and
posterior sections of mesenteron only).

Czerski (1904) Meloe.

Hirschler (1905) Catocala.**

Saling (1907) Tenebrio.

* Graber believed that cells were also added to the mesenteron from the
splanchnic layer of the mesodern.

“* Hirschler believes that in Catocala a median section of the lower layer
also contributes cells to the mesenteron.

THE EMBRYOLOGY OF THE HONEY BEE 75

Friederichs (1906) Donacia.

(IV) Mesenteron derived, independent of the mesoderm,
from two proliferating areas of the blastoderm, one at each end
of the germ band, corresponding to the future location of the
stomodaeum and proctodaeum, respectively.

Carriere and Burger (1897) Chalicodoma, Tenebrio.

Noack (1901) Calliphora. |

(V) Mesenteron derived from cells migrating inward from
thickenings or islands of the blastoderm.

Uzel (1897, 1898) Lepisma, Campodea.

The importance accorded by all investigators of insect embry-
ology to the question of the origin of the mesenteron is due of
course to its relation to the germ layer theory. According to
this theory, which is based on a large number of observations
on the development of various animals, the material from
which the mesenteron is formed should correspond to entoderm
and the efforts of practically all investigators of this prob-
lem have been bent toward establishing this homology, and
to derive the conditions found in the insect egg from the typical
gastrula. This has proved to be an exceedingly difficult task.
Thirty years ago Weismann (1882) wrote (p. 81): “It be-
comes more and more evident, that nowhere in the entire animal
kingdom is the ontogeny so distorted and coenogenetically de-
generate, as in the insects, so that scarcely anywhere are the
germ layers so difficult to recognize as here.” Time has proved
the truth of these statements.

Prior to 1884 nearly all of the investigators of insect embry-
ology were divided into two camps; either they followed Dohrn
(1866) in deriving the mesenteron from the cells remaining in
the yolk, or followed Kowalevski in deriving the mesenteron
from the inner wall (splanchnic layer) of the mesodermic so-
mites. Only one exception is to be noted: Ganin (1874) de-
scribed the mesenteron as derived from the inner ends of the
ectodermal proctodaeal and stomodaeal invaginations. In these
three views are contained the germs of all the later theories of
the origin of the mesenteron.

In 1884 Grassi’s paper on the development of the honey bee
appeared, in which the bipolar origin of the mesenteron from
the lower layer (mesoderm) was first demonstrated. Kowal-

76 THE EMBRYOLOGY OF THE HONEY BEE

evsky in 1886 published a brief paper on the origin of the mesen-
teron rudiments in Musca. In this insect the mesenteron
rudiments were formed, much as Grassi had described them in
Apis, from the anterior and posterior ends of the middle plate
(mesoderm). Kowalevsky constructed an ingenious theory to
account for the origin of the germ layers in insects, comparing
the conditions which he found in the insect embryo with those
occurring in a marine invertebrate, Sagitta. He regarded the
insect egg, at the time of the formation of the germ layers, as
comparable to a gastrula so much stretched out or elongated
that the entoderm (mesenteron rudiments) was pulled into two
halves. The views of Grassi (see above p. 70) and Kowalevsky
were later accepted by Wheeler (1889), Cholodkowsky (1891),
and by Heider in Korschelt and Heider’s textbook, and re-
cently in a modified form, by Nusbaum and his pupils (II. 2).

These results did not however find universal acceptance, since
Witlaczil, Voeltzkow and Graber, like Ganin, contended that the
mesenteron arose from the blind inner ends of the proctodaeum
and stomodaeum. In 1895 Heymons, in his handsome mono-
graph on the development of the Orthoptera and Dermaptera,
devoted especial attention to the development of the mesenteron,
and also found that it was derived from the blind inner ends of
the stomodaeal and proctodaeal invaginations. This work has
had a wide influence and Heymons’ results have been confirmed
by a number of investigators in the Coleoptera, Lepidoptera and
Orthoptera.

Heymons recognized more fully than his predecessors the
theoretical difficulty involved in deriving the mesenteron from
ectoderm, since identification of the mesenteron of insects with
entoderm is thereby precluded. Heymons boldly met the diffi-
culty by supposing that the functional mesenteron of pterygote
insects is of comparatively recent origin, and that the original
entoderm is now represented by the yolk cells, which therefore
may be considered as constituting a vestigial or degenerate
mesenteron. This view received support by Heymons’ discovery
(1897) that in Lepisma, a primitive apterygote insect, the func-
tional mesenteron is actually formed from yolk cells and by
Madame Tschuproff-Heymons’ discovery (1903) that in the
Odonata the mesenteron is formed in part by yolk cells and in

THE EMBRYOLOGY OF THE HONEY BEE 77

part by ectoderm derived from the stomodaeal and proctodaeal
invaginations, a condition which could readily be interpreted as
constituting a transitional stage between Lepisma and the higher
pterygote insects. Recently Nusbaum and Fulinski (1906, 1909)
have reinvestigated the origin of the mesenteron in two of the
forms studied by Heymons, Phyllodromia (Blatta) and Gryllo-
talpa, and have obtained a different result, namely that the me-
senteron is formed from the two ends and also from the median
portion of the lower layer. Similar results were obtained by
Hirschler (1906, 1909) in the Lepidoptera and Coleoptera. Hey-
mons’ conclusions and interpretations have also been contested.
by Escherisch (1900) in his paper on Calliphora (Musca). Car-
riére and Biirger’s (1897) observations on the development of
the mesenteron rudiments in the mason bee have already been
mentioned (p. 71) ; Noack (1901) arrived at similar conclusions
in the case of Calliphora (Musca). In contrast to the results of
Heymons and all the other investigators of this subject Ham-
merschmidt (1910) finds that in the orthopteron Dixippus the
mesenteron is formed exclusively from the median section of
the lower layer, which in most insects produces the blood cells.

The conditions existing in the honey bee, as they have been
described in this paper, obviously lend little support to the views
of those who regard the mesenteron of insects as arising from
the ectoderm of the stomodaeum and proctodaeum (III), since
its rudiments are already formed long before the stomodaeal and
proctodaeal invaginations appear; much less do they harmonize
with the theory of the origin of the mesenteron from yolk cells
(1). The relation of the mesenteron rudiments in the honey
bee may be interpreted in either of two ways, and the one chosen
will probably depend largely on the theoretical bias of the in-
terpreter. First, the mesenteron rudiments may be referred to
the mesoderm (Il). Several facts can be cited in support of
this view, for example: the continuity of the middle plate and
the mesenteron rudiments during the earlier stages in their for-
mation, and the fundamental similarity in the manner of for-
mation of both the mesenteron rudiments and the mesoderm, all
being formed by a contemporaneous inward migration of ele-
ments of the blastoderm. Second, the mesenteron rudiments
may be considered, with Carriére (IV) as purely blastodermal

78 THE EMBRYOLOGY OF THE HONEY BEE

in origin, since their manner of formation, although essentially —
‘similar to that of the mesoderm, differs much from it in detail,
moreover the cells composing the mesenteron rudiments are from
the first distinguishable from those of the mesoderm. A final
decision between these two interpretations seems premature. The
honey bee is a highly specialized member of a specialized order,
and therefore an unsuitable form on which to base generaliza-
tions, since its development certainly presents many modifica-
tions of the type, moreover generalizations based on the study
of one form are always unsafe.

It is sufficiently evident that in spite of the numerous papers
dealing with the origin of the mesenteron in insect embryos,
there is much need of further investigation, particularly of
the more generalized types. Superficial study, however, would
be worse than useless; the type of investigation demanded is the
highest, requiring the delicate and precise methods of the cytol-
ogist, the best fixation and staining possible, a complete series
of stages, a study of the origin of the rudiments cell by cell, and
finally an eye single to the facts and regardless of preconceived
theoretical considerations.

Before leaving the subject of the germ layers it will be neces-
sary, in order to discuss the stages following, to describe briefly
the structure of an embryo at the final stage of this period, Stage
VII. This stage is illustrated by a series of transverse sections,
represented by figures 29, 30, 31 and 32.

Figure 29 shows a section passing through the extreme an-
terior end of the embryo. The ectoderm here extends over about
two-thirds of the circumference of the yolk and is much thick-
ened in its lateral portions. Within the ectoderm is a crescentic
mass of cells, the anterior mesenteron rudiment, which has
grown both cephalad and laterad to form a cap-like mass cover-
ing the ventral side of the cephalic end of the yolk. In this and
also in the next section the amnion (Am) is seen covering the
exterior of the embryo as a thin sheet of flattened cells. This
membrane will be discussed at length later. |

Figure 30 also shows a section passing through the cephalic
end of the embryo, intersecting it just caudad of the posterior
limits of the anterior mesenteron rudiment. The opening in the
ectoderm through which the anterior mesenteron rudiment reaches

THE EMBRYOLOGY OF THE HONEY BEE 79

Fic. 29. Transverse section through the anterior end of an egg, Stage
VII, passing in front of the rudiment of the stomodaeum, showing the

procephalic lobes (ProL), anterior mesenteron rudiment (AMR), and
amnion (Am), x 243.

Fic. 30. Transverse section through the anterior end of an egg, Stage
VII, passing just caudad of the rudiment of the stomodaeum, showing
the procephalic lobes (ProL), anterior mesenteron rudiment (AMR),

mesoderm (Meso) and amnion (Am), also the remains of the cephalo-
dorsal body (CB), x 243.

80 THE EMBRYOLOGY OF THE HONEY BEE

the external surface is seen, but the cells filling it belong to the
mesoderm (compare Fig. 26B), as the small size of their nuclei
demonstrates. The mesoderm (Meso) extends also a short dis-
tance laterad on both sides of the opening. The anterior mesen-
teron rudiment (AMRF) is represented in this section by a layer
of cells two or three deep lying on each side against the inner
surface of the lateral ectoderm, above the mesoderm. The
ectoderm has the same thickness and relative extent as in the
preceding section.

Figure 31 represents a third section passing through the ceph-

Fic, 31. Transverse section through the anterior end of an egg. Stage
VII, intersecting the posterior margin of the procephalic lobes (ProL)
The middle cord (VC) is shown, on each side of which is the neurogenous
ectoderm from which the ventral cord is to be formed. The mesoderm
(Meso) the amnion (Am) and two or three cells of the anterior mesen-
teron rudiment ( AMR) are also shown, x 243.

alic end of the embryo, a few sections caudad of the last. The
ectoderm here is seen to be diminishing slightly in both extent
and thickness. In the ventral mid-line is a strip of cells (MC)
whose transverse section has somewhat the outline of an hour
glass, and represents the point of juncture of the lateral folds.
The significance and.fate of this strip will be described in the
section devoted to the nervous system. Lining the ectoderm is
a single layer of mesoderm, broken at two points. High up,

THE EMBRYOLOGY OF THE HONEY BEE 81

on the right hand side of the section, in place of mesoderm, two
cells belonging to the anterior mesenteron rudiment are seen
(AMR).

Figure 32 shows a section taken through the mid-region of the

<<

Meso

Sees,

sie Rte Css Seri as

Fic. 32. Transverse section through the future thoracic region of an
embryo, Stage VII-VIII, showing the form and relations of the ectoderm
(Ect) and mesoderm (Meso), x 243.

trunk, and is representative of the condition existing through-
out its extent, including the future thorax and abdomen. The
ectoderm covers somewhat less than the ventral half of the egg,
and is composed of a single layer of closely packed columnar
cells. Within the ectoderm is the mesoderm (Meso), whose
lateral extent is somewhat less than that of the ectoderm. The
mesoderm, near the mid-line, is somewhat thinner than the ec-
toderm, but towards its lateral margins, which are rounded, its
thickness approaches that of the ectoderm. It is clearly com-
posed of two layers, continuous with one another at the lateral
margins. Near the mid-line in its thinner portions the cells
composing the two layers are flat and the line of separation be-
tween the layers is somewhat indistinct. Toward the lateral
margins of the mesoderm, however, the cells composing the two
layers become columnar, and the line of separation between the
layers is sharp. This line of separation represents the cavity of
the mesodermal somites.

A section through the caudal end of an embryo of Stage VII
has already been represented in figure 28, and described in the
part of this section relating to the posterior mesenteron rudiment.

VI

THE AMNION AND THE CEPHALO-DORSAL Bopy

1. The Amnion

The cells destined to form the amnion, as described on p. 44,
constitute the marginal portions of the ventral plate, and cover
the dorso-lateral regions of the egg, bounding the median dorsal
strip laterally (Figs. 19 and 24A, Am). At first the amnion-form-
ing cells are not sharply demarcated from the remainder of the
ventral plate, and seem to intergrade with the cells of the latter,
but as the differentiation of the germ layers progresses the amnion
cells also become differentiated, appearing more or less shortened
or flattened as contrasted with the long prismatic form of the
cells of the embryonic portion of the ventral plate. As soon as the
amnion cells become distinguishable as such, it becomes evident
that the amnion cells at the anterior end of the egg differ some-
what from those of the other regions. Over the entire cephalic
end of the egg, the amnion cells are rounded in form, closely ar-
ranged and relatively numerous, in sections resembling a string
of beads (Fig. 24A and 26A, Am). The characteristic transpar-
ency of the inner ends of the amnion cells, noted by Petrunke-
witsch (1903) is especially evident.. Caudad of the cephalic
region the amnion cells become relatively fewer and their form
also becomes more flattened (Figs. 20A and B and 21, Am).
These differences are, however, merely temporary and disappear
as the amnion increases in extent. Prior to Stage V the amnion
therefore consists of two longitudinal bands of epithelial cells,
separated by the median dorsal strip and joining the ventral plate
laterad. These amniotic bands are widest at their anterior ends,
which cover the cephalic end of the egg, and their component
cells are here more numerous and less flattened than elsewhere,
as just mentioned. At Stage V the two bands begin gradually to
widen, the cells of their inner margins creeping up over the dorsal

82

THE EMBRYOLOGY OF THE HONEY BEE 83

strip, which becomes submerged in the yolk (Figs. 20A and B, 21,
24B, 34B and C, Am). Subsequently then is a fusion of the
amniotic bands along the dorsal mid-line of the egg, beginning
first at the cephalic end, and occurring somewhat later over the
remaining extent of the egg, being completed at Stage VI.
While the amnion is thus covering the dorsal side of the yolk
it commences also to cover the ventral side. This process can
best be observed in fresh material, in which the outlines of the
amnion are beautifully clear. In fixed material, whether examined
in alcohol or stained and cleared, the outlines of the amnion are
frequently invisible, except in actual or optical sections. At Stage
V or a little earlier, on the ventral side of the egg, the amnion
together with the anterior end of the germ band separate from the
yolk (Fig. 26A). This is not primarily due to the depression of
the yolk caused by the development of the anterior mesenteron
rudiment as the figure might suggest, since the separation is al-
ready evident when this rudiment is in its earliest stages; it is
probably first brought about by an increase in the superficial extent
of the amnion in this region, as indicated by its convex or arched
form. Soon after, the amnion separates from the yolk over the
entire cephalic pole of the egg, rising up in the shape of a hemi-
spherical cap (Fig. V). Next—at Stage VI—it severs its con-
nection with the germ band around the anterior end of the latter
and slides over it in the form of a hood or cowl, thus forming the
cephalic fold (Fig. 32, tam, and 26B Am). This separation of
the amnion from the germ band progresses caudad along its
lateral margins, accompanied by the caudad extension of the ceph-
alic fold over the ventral face of the germ band in such a way
that the free edge of the cephalic fold forms a semicircular curve
with its concave side directed caudad (Fig. 32, A-C, ram). When
the cephalic fold has covered about one-half of the ventral face
of the embryo, a second amniotic fold, the caudal fold (Fig. 32
B, 2am), appears at the extreme caudal end of the germ band.
This fold is formed like the head fold, by separation of the am-
nion from the caudal end of the germ band, and first appears as a
crescentic membrane. Since the caudal end of the germ band is
now curved completely around the caudal pole of the egg, so that
the former lies on the dorsal side of the egg, the caudal fold also

84 THE EMBRYOLOGY OF THE HONEY BEE

lies on the dorsal side of the egg with its concave edge directed
toward the caudal pole. The caudal fold increases in extent in the
same manner as the cephalic fold, progressing slowly toward the
caudal pole of the egg, the lines of rupture of the caudal fold
and the germ band extending to meet those of the cephalic fold
(Fig. 32C). The two folds thus approach one another, the ceph-
alic fold moving at a much more rapid rate than its counter.
part; soon the two meet and fuse near the caudal pole of the egg
(Fig. 32C and D). This occurs slightly prior to Stage VIII. While
the amnion is thus covering the ventral side of the egg, and
consequent to its severance from the edges of the germ band,
it also separates from the yolk on the dorsal side of the egg.
This separation is directly connected with the separation of the
amnion from the edges of the germ band, since it must at the
same time also separate from the yolk at this point. The separa-
tion thus initiated is continued dorsad, the amnion being, so to
speak, peeled off from the yolk, which is thus left bare except
for the thin protoplasmic pellicle surrounding it (cf. Figs. 29 and
30). When the formation of the amnion is completed, or shortly
afterwards, the amnion therefore forms a complete envelope sur-
rounding the embryo, and free from the embryo at the two ends
of the egg, but closely applied to the embryo elsewhere. Its outline
is similar to that of the chorion, except that it is shorter, leaving
a considerable space vacant between amnion and chorion at the
ends of the egg. Since the space surrounding the egg, between —
the latter and the chorion, is filled with a watery fluid, it follows
that the space between the embryo and amnion is also filled by
this same fluid.

The cephalic fold of the amnion, as described above, is at first
composed of cells which are rounded in form (Figs. 24B, 26A).
As this fold progresses over the surface of the germ band its
cells become gradually thinner and flatter (Figs. 26B, 29, 30 and
31, Am). At its completion, at Stage VIII, its average thickness
is scarcely greater than that of the chorion. At the ends of the.
ege the amnion is somewhat thicker than elsewhere, and in these
regions the nuclei are oval in outline, forming lenticular swellings
(Fig. 29). Elsewhere, over the body of the embryo, the amnion
is scarcely thicker than the chorion, its nuclei being flattened to

:
}
¢
|

8 LS ee eS ee
4

~~“ an ee
a

THE EMBRYOLOGY OF THE HONEY BEE 85

such an extent that in sections they frequently appear as short
dark lines (Fig. 30). Comparing the extreme tenuity of the am-
nion, when completed, and the superficial extent of each of its-
component cells as compared with the thickness and slight super-
ficial extent of the amnion cells, at earlier stages, particularly
those of the head fold, it is apparent that the development of the
amnion is due principally, if not exclusively, to a mere extension
or spreading out of the original cells present at least as early as
Stage IV. This was essentially the view taken by both Biitschli
(1870) and Grassi (1884). Neither of these investigators saw
any division of the amnion cells, nor has the writer observed
them.

Biitschli (1870), Kowalevski (1871) and Grassi (1884) have
described and figured the formation of the amnion in the honey
bee. Biitschli’s account, based exclusively on observations of
fresh eggs, is full and substantially correct, recording, among
other details, the covering of the dorsal surface of the egg by
the amnion cells. Kowalevski’s account is less extended than
that of Biitschli, and while correct as regards the topographical
relations of amnion and germ band, erroneously describes the
amnion as originally composed of two layers, which subsequently
fuse to form one. This error was possibly due to Kowalevski’s
contemporaneous studies on Hydrophilus, in which, as is well
known, there are two embryonic membranes. Grassi’s account,
although based in part on actual sections, adds but little to that
of Biitschli. Grassi incorrectly describes the meeting of the ceph-
alic and caudal folds as taking place on the ventral surface of
the egg midway of its length, whereas it normally takes place at
or near the caudal pole of the egg.

Before entering upon a comparison of the embryonic envelope
(amnion) of the bee with the embryonic envelopes of the other
pterygote insects, it will not be out of place to recall the manner
in which these are formed. Briefly stated, it consists essentially
in the elevation, around the embryonic rudiment, of a fold of the
extra-embryonic blastoderm, which then extends over the embryo
from all sides, its edges finally meeting and fusing (see Kors-
chelt and Heider, Fig. 133). Contemporaneous with this fusion
is the separation of the two layers composing the fold, so that

86 THE EMBRYOLOGY OF THE HONEY BEE

a:
tee -
° _ =

arto:

. .
ee 4 cel ter
ae tdi ecee?

Fic. 33. Side views of four embryos, drawn in diagrammatic form, show-
ing four stages in the development of the amnion. Outlines drawn with a
camera lucida from living eggs. The amnion (14m, 2Am) is stippled,
with dotted outline, except when it adheres to the embryo or yolk. In B,
C and D the amnion is represented for the sake of clearness, as standing
off from the embryo further than is actually the case. A corresponds
to Stage VI, B to Stage VII, C to Stage VII-VIII, D to Stage VIII-IX,
x AI.

THE EMBRYOLOGY OF THE HONEY BEE 87

the embryo thus becomes covered by two separate layers, one
enclosing the entire egg, the serosa, and one covering only the
embryo, the amnion. The amniotic fold, the common rudiment
of the amnion and serosa, does not usually develop simultaneously
around the entire margin of the embryo, but appears commonly
first at its anterior and posterior ends, thus forming a cephalic
and a caudal fold. These conditions may be considered typical
of the majority of pterygote insects. There are, however, some
exceptions of which the honey bee is an example. Here, as is al-
ready apparent, but one envelope is present, which, since it covers
the entire egg inclusive of the germ band, is therefore comparable
to the serosa rather than to the amnion. This is the opinion held
by Heider (1891). The term amnion nevertheless has been ap-
plied by both Biitschli and Grassi to the single embryonic envelope
of the honey bee, and it therefore seemed inadvisable to change it.

In addition to the honey bee, embryonic envelopes have been
observed in the following representatives of the non-parasitic
Hymenoptera: Formica, Myrmica (Ganin 1869), Polistes, Form-
ica (Graber 1888), Hylotoma (Graber 1890), Chalicodoma.
Polistes (Carriére 1890, and Carriére and Burger 1897), and
Camponotus (Tanquary 1913). Graber states that two embryon-
ic membrances were found in all the three forms studied by him.
On the contrary Carriére and Biirger (1897) explicitly state that
in one of the forms studied by Graber, Polistes, only one embryon-
ic membrane is present. This is also the case in Chalicodoma.
In the ants but one membrane is found, according to both Ganin
(1869) and Tanquary (1913).

In some of the parasitic Hymenoptera, for example Biorhiza,
Rhodites (Weismann 1882) Platygaster (Kiulagin 1897), an
embryonic membrane is formed which is also single, but the forma-
tion of this—at least in the case of Platygaster,—is so peculiar
and so different from that found in other insects that its homology
with the embryonic membrane of non-parasitic insects is perhaps
open to question.

Concerning the fate of the amnion in the honey bee the only
data given by previous observers are those of Buitschli (1870)
who says (pp. 533-534): “This envelope persists during the
entire development of the embryo, and like the egg envelopes

88 THE EMBRYOLOGY OF THE HONEY BEE

is finally torn by the active movement of the young larva.’’ The
writer has not observed this process in the living larva, but a
study of sections confirms Bitschli’s statements. Up to Stage
XIV the amnion is intact, but at Stage XV, when the young
larva has become flexed ventrad, and has ruptured the chor-
ion, the amniotic membrane is evident only in fragments,
usually clinging to the chorion. These fragments or shreds are
very much thicker than was the amnion before its rupture, and
always contain a number of ovoid nuclei, closely grouped to-
gether, indicating that the fragments had contracted and sug-
gesting that the amnion had been under tension previous to its
rupture. In Chalicodoma, according to Carriére (1890, 1897)
some time before the hatching of the larva the embryonic envelope
becomes torn into fragments, which become thickened to form
narrow twisted bands of polygonal cells. These disappear before
hatching, being apparently absorbed. In the pterygote insects
of other orders,—Orthoptera, Coleoptera—the later history of
the embryonic envelopes is very varied, and usually associated
with a shifting in the position of the embryo relative to the yolk
(blastokinesis, Wheeler 1893). The fate of these envelopes,
briefly stated, is as follows: The serosa is either cast off or ab-
sorbed by the egg; the amnion is usually also absorbed, but in some
cases has been described as contributing to the formation of the
dorsal body wall, as in Meloé (Nusbaum 1890).

2. The Cephalo-dorsal Body

Associated with the formation of the amnion and contempor-
aneous with that of the germ layers, is a peculiar structure to
which the writer (1912) gave the name “cephalo-dorsal disk.”
This structure (Fig. 34 A-C), which attains its maximum size
about Stage V, is extremely variable in location, form and size.
On account of its variability in ‘size it would be better to term this
structure the “cephalo-dorsal body.’ It is situated on the dorsal
side of the egg usually near the cephalic pole and opposite to the
anterior mesenteron rudiment (Fig. 34A), but in one egg, which |
appeared to be normal in every other respect, the cephalo-dorsal
body was found situated almost precisely at the cephalic pole it-
self. The form of this body, while very variable, usually ap-

THE EMBRYOLOGY OF THE HONEY BEE 89

Fic. 34. Parts of sections showing the cephalo-dorsal body. ‘A, cepha-
lic part of a median sagittal section of an egg, Stage V, showing the
cephalo-dorsal body at the time when it is best developed and most con-
spicuous. B, dorsal part of a transverse section of an egg, Stage V,
showing the amnion (Am) advancing over the cephalo-dorsal body toward
the dorsal mid-line. C, the cephalo-dorsal body covered by the amnion,

x 387.

go THE EMBRYOLOGY OF THE HONEY BEE

proaches that of an ellipsoid, flattened dorso-ventrally, with its
long axis lying in the sagittal plane. In many cases in its form
and longitudinal extent it approximates the one illustrated in
figure 34A, in some it is smaller, while in two cases the cephalo-
dorsal body was tongue or club-shaped, its smaller end extending
back in the dorsal mid-line for approximately one-third of the
entire length of the egg. Its external surface, in the earlier stages
forms a part of the external surface of the egg, in later stages
it is partly or entirely covered by amnion cells. Its inner surface;
usually strongly convex, is indented by the alveolar spaces repre-
senting the yolk spheres, while extending out between the inden-
tations are delicate protoplasmic processes continuous with the
interstitial protoplasmic meshwork (Fig. 34A and C). The ceph-
alo-dorsal body is therefore, like the dorsal strip, closely asso-
ciated with the yolk. In structure the cephalo-dorsal body appears
to be a syncitium, composed of rather clear and somewhat vacuo-
lated cytoplasm, within the inner half of which numerous nuclet
are embedded. At its posterior edge it becomes continuous with
the dorsal strip, as shown in the figure (Fig. 34A).

In origin the cephalo-dorsal body appears to be little more
than a localised swelling of the dorsal strip. At Stage IV and
the stages intervening between Stages IV and V the cephalo-
dorsal body, as compared with its condition at Stage V, is located
somewhat closer to the cephalic pole, is flatter and contains fewer
nuclei, being plainly nothing more than the slightly thickened
anterior end of the dorsal strip. At about Stage V it increases
rapidly in size and in the number of nuclei contained within it,
until it attains its maximum size, as shown in the figure, but its
connection with the dorsal strip is never lost and is always readily
apparent. Moreover, in many cases smaller but similar swellings
of the dorsal strip exist at various points on the dorsal mid-line
in the anterior half of the egg. One of these is illustrated in
figure 20A.

The rapid increase of the cephalo-dorsal body in size and more
particularly in the number of nuclei contained within it, is difficult
to explain. While it is not impossible that it owes its origin in
part to cells derived from the yolk, there is no satisfactory evi-
dence that this is the case. The nuclei seen in the yolk during

a se

THE EMBRYOLOGY OF THE HONEY BEE QI

the formation of this body are usually larger and rounder than
those contained within the latter, moreover they are not present
in sufficient numbers in the immediate neighborhood of the ceph-
alo-dorsal body to make it seem probable that they contribute
largely to its formation or growth. It is also possible that the
number of nuclei may be increased by either direct division or
fragmentation of the nuclei originally present, but this could not
be demonstrated to be the case, although in many instances this
was suggested by the appearance of the nuclei. The latter are
so small and so frequently crowded closely together that it was
found impossible satisfactorily to decide this question. It is,
however, certain that the nuclei differ among themselves in size,
and that they are also very frequently lobed or constricted.

Soon after reaching its maximum size the cephalo-dorsal body
during Stage VI breaks up into amoeboid cells or small syncitia
containing one to several nuclei. Some of these wander ventrad
towards the center of the egg (Fig. 30), but many remain near
the dorsal surface of the yolk. At Stage VII the remnants of the:
cephalo-dorsal body are still recognizable at the cephalic end of
the egg as irregular branching islands of rather pale granular
cytoplasm enclosing one or more small nuclei. The outlines of
the nuclei are always irregular and often faint, and when grouped
closely together as is frequently the case, many of the nuclei ap-
year to be in process of fusing with one another (Fig. 34C). At
Stage VIII the remains of the cephalo-dorsal body are evident,
lying close beneath the anterior mesenteron rudiment, near its
posterior end (Fig. 87A,CB). From this point they travel. to-
ward the caudal pole of the egg as if drawn by the rudiment until
they reach the point of junction of the anterior and posterior
mesenteron rudiments. During this period the vestiges of the
cephalo-dorsal body usually present the appearance of a thin layer
of slightly yellowish protoplasm, containing a few nuclei which
are situated near its inner surface, as shown in figure 87A, CB.
In some preparations, however, the cephalo-dorsal body seems to
have remained undiminished in size up to Stage VIII or IX. It
is then quite conspicuous and plainly evident even at a low mag-
nification. In these cases the appearance is almost precisely like
that figured by Petrunkewitsch (1902, Fig. 8) for the drone egg,

92 THE EMBRYOLOGY OF THE HONEY BEE

the cephalo-dorsal body being here represented by the “Rz” cells.™
At Stage X the vestiges of the cephalo-dorsal body may still be
seen lying beneath the dorsal wall of the mesenteron, but they
soon become indistinguishable from the elements of the yolk, and
unquestionably like them suffer ultimate absorption. The cephalo-
dorsal body was first seen by Grassi (1884), who has represented
it in figure 1 of Plate X, but did not mention it in the text. Dickel
(1903) was the first to describe this body, which he named the
“yolk plug,’ since he regarded it as derived from yolk cells which
migrated to its point of origin, this point being designated as the
blastopore. The “yolk plug” was regarded by Dickel as the an-
terior mesenteron rudiment. This assumption was founded on the
circumstance that the ‘‘yolk plug’ had disappeared at the time
when the anterior mesenteron rudiment came into prominence.
It will hardly be necessary to say that this assumption is wholly
wrong, and must have been due to defective observation, since
any series of sections showing the “yolk plug” should also show
the true anterior mesenteron rudiment, since in fact the “yolk
plug’ becomes evident only after the anterior mesenteron rudi-
ment has begun to be formed. Moreover, the remnants of the
cephalo-dorsal body (“yolk plug’) are plainly evident long after
the anterior mesenteron rudiment is fully formed.
Petrunkewitsch (1902), in a paper on the fate of the polar
bodies in the drone egg, describes and figures a body formed in
the cephalo-dorsal region of the egg by cells—termed for brevity
“Re-cells”—which owe their origin to a single cell formed by the
union of the second polar body with the inner half of the first.
This structure formed by the “Re-cells” apparently agrees closely
in form, structure and position with the cephalo-dorsal body of the
worker egg. Petrunkewitsch, however, states that he did not
find this body in worker eggs, and gives a figure to illustrate this
statement, but this figure (4) is obviously of an earlier stage than
those in which is found either the cephalo-dorsal body or that
formed by the “Rez-cells.” The fate of the latter according to
Petrunkewitsch’s account, is as follows: It first sinks below the

* Petrunkewitsch (loc. cit.) states that mitotic figures were found in
this cell mass. The writer has never seen them in the cephalo-dorsal body
at any stage.

ee eT
a i =

.. <z, OR Ee rn

THE EMBRYOLOGY OF THE HONEY BEE Me

surface into the yolk and becomes covered by blastoderm (am-
nion). So far its history corresponds to that of the cephalo-
dorsal body. Next the “Ftz-cells” glide caudad along the dorsal
side of the yolk until they reach the middle of the abdomen, in-
creasing in number meanwhile by mitotic division. Here they
are overtaken by the advancing ectoderm and the epithelium of
the mesenteron and come to lie between these two layers. Finally
the “Jts-cells” divide into two groups, one on each side of the
embryo. These then migrate ventrad, enter the coelomic cavities
on each side of the third, fourth and fifth abdominal segments
and constitute the male sex cells.

Nachtsheim has already (1913) explicitly affirmed the identity
of the “Rez-cells” of Petrunkewitsch and the “yolk plug,” of
Dickel solely on the evidence afforded by the two papers.

The writer has not so far studied the “Rez” cells in the drone
egg, so that a final conclusion regarding the identity of the
cephalo-dorsal body and the “Rez” cells may seem premature.
Nevertheless, a comparison of Petrunkewitsch’s figure with the
corresponding stages in the worker egg, as well as with the figures
given by Dickel (1904) leads one almost inevitably to the con-
clusion that the “Rz” cells, the “yolk plug,’ and the “cephalo-
dorsal” body are identical.

The significance and possible homologies of the cephalo-dorsal
body are extremely uncertain. Its rapid disappearance and great
variability suggest that it is a vestigial structure. Hirschler
(1909) has described, in the chrysomelid beetle Donacia, a struc-
ture termed by him “the primary dorsal organ,” which in several
respects resembles the cephalo-dorsal body of the honey bee. This
“primary dorsal organ” consists of an oval area of the blastoderm
lying in the dorsal mid-line, which sinks down into the yolk and
becomes covered over by blastoderm, its component cells mean-
while sending out processes into the yolk and also showing evi-
dences of degeneration. Finally “the primary dorsal organ” de-
generates completely and is absorbed by the yolk. In position,
overgrowth by blastoderm and final absorption by the yolk this
structure corresponds quite closely to the cephalo-dorsal body of
the honey bee, but in one respect it differs: in the bee-the cephalo-
dorsal body is formed during the formation of the germ layers,

94 THE EMBRYOLOGY OF THE HONEY BEE

while in Donacia the “primary dorsal organ” is formed prior to
the formation of the germ layers. In respect to other insects
little can be said that is not pure conjecture, since the data afforded
for comparison are too scant. To those familiar with insect em-
bryology comparisons with “the micropylar organ” (Wheeler
1893) or “procephalic organ” (Claypole 1898) of Anurida and the
“indusium” of Xiphidium (Wheeler 1893) will suggest them-
selves, but only further investigations on the embryology of
insects with special regard to the question as to whether a vestigial
“dorsal organ” is of frequent occurrence in insect embryos can be
of real value in determining the homology of these various anoma-
lous structures.

a ee

—

ee ee

C—O ae PE te

a)

ee PLP

se

a=

eee

Vil

GENERAL ACCOUNT OF THE DEVELOPMENT OF THE EMBRYO WITH
ESPECIAL REFERENCE TO THE EXTERNAL ForRM

1. Changes in the Form of the Egg up to Stage VIII

The developmental changes undergone by the egg, from the
beginning of the cleavage to the close of the formation of the
germ layers, have already been described in detail, nevertheless
there is one phase of the development which so far has been left
out of account, that is, the changes in the external form of the
egg. As far as Stage VIII the original outlines of the egg are in
a large measure preserved; these have been broken only by the
amniotic folds, and in a slighter degree, by the lateral folds. On
the other hand the proportions of the egg have been somewhat
altered; during the stages described, it has appreciably shortened,
as may be seen by comparing figures I and VII. This shortening
is very evident in the actual specimens, in which the chorion is
intact, since at Stages VII-VIII there is a space of considerable
size left vacant between the embryo and the chorion at the two
ends of the egg (Fig. 35D). Examination of early stages (I)
shows that at the beginning of cleavage the egg contracts slightly
in the long axis, as shown by the appearance of a narrow space
between the egg and the chorion at the anterior end and fre-
quently at both ends. The amount of contraction does not usually
exceed about 5 per cent of the original length of the egg, as esti-
mated by the length of the chorion (Fig. 35A). This shortening
of the egg in its long axis was seen in the living egg by both
Biitschli (1870) and Kowalevski (1871) and mentioned by them
as being one of the earliest phenomena of development. No fur-
ther shortening of the egg takes place until the time of the forma-
tion of the germ layers, but during this latter period (Stages IV-
VII) the egg again shortens rapidly until at Stage VIII the egg
(embryo) is now from 20 to 25 per cent shorter than the chorion,
relatively large spaces, filled with fluid, being left vacant at both
ends of the egg, between the embryo and the chorion (Fig. 35B

95

96 THE EMBRYOLOGY OF THE HONEY BEE

A B

Fic. 35. Profile drawings of eggs showing four stages in development,
to illustrate the relative dimensions of the egg and chorion. Outlines
drawn with camera lucida from fixed material. Amnion represented by
dotted line. A, cleavage and blastoderm stages; B, Stage IV-V; C, Stage
VI; D, Stage VIII, x qr.
and C). These spaces are invaded by the amnion. A comparison
of figures 35B and D shows that this second shortening of the
egg is not only contemporaneous with the formation of the germ
layers, but that it is also accompanied by the curving of the germ
band around the ends of the egg. It is therefore evident that the
shortening which takes place has principally affected the yolk,
and furthermore that the germ band itself has actually lengthened
but little, if at all, between Stages [TV and VIII. The embryo
does not sttbsequently increase materially in length until immedi-
ately before hatching, since eggs 70-73 hours old still show spaces
of considerable size at their ends, between the embryo and the
chorion. At the time of hatching, however, the young larva, after
rupture of the. chorion, increases greatly in length, as shown in
figure X'V.

These changes in the length of the egg, with the exception of
the initial shortening, are not explicitly mentioned by any of the
previous investigators; nevertheless, the figures of both Butschli

a

iJ
=

Y
Ue
s
cy
be
ik
} <
;
al
A

THE EMBRYOLOGY OF THE HONEY BEE 97

(1870) and Kowalevski (1871), which represent the embryos of
different stages surrounded by the chorion, show that they ob-

_ served these changes precisely as they are described above.

2. The Development of the Embryo

In the study of any developmental processes a knowledge of the
end stage is necessary to an intelligent comprehension of the
earlier stages, since these are only isolated examples chosen from
what is in reality a continuous process leading to the end stage.
This is, in a sense, their goal, but in a larger sense is itself only
a more or less arbitrarily chosen stage in the continuous process
of the life cycle. The end stage of the development in the egg is
of course attained when the embryo breaks the egg envelope or
chorion and becomes a larva. This stage, Stage XV, reached in
from 74 to 76 hours after the egg is laid, is illustrated by figure
XV, representing a recently hatched larva. This larva is drawn
not from a living specimen, but (like the others of this series)
from one which has been stained, cleared and mounted. The
body wall is represented as much more transparent than it is in
reality, in order to show more clearly the organs contained within,
while the sharpness of outline of these is exaggerated.

The larva as a whole is long cylindrical in form, slightly bent
toward the ventral side, like a bow, and tapers gently toward
the caudal extremity, which is obliquely truncate. The larva is
divided by constrictions into a short and rounded head and thir-
teen segments, the latter being approximately equal in length and
collectively constituting the trunk. On the dorsal side the con-
strictions are narrower and deeper than on the ventral side. The
head is of the same diameter as the anterior part of the trunk,
and the constriction separating the two is only a trifle deeper
and broader than those separating the trunk segments. On the
ventral side of the head is the mouth (Mth). In front of the
latter and directed downwards is a blunt rounded lobe, the labrum
(Lm), while bounding the mouth laterally are two pairs of rather
large papilliform processes, the mandibles (Md) and the first max-
illae (rMx). At the rear the mouth is bounded by a single flat
lobe, the labium (Lb).

The alimentary canal is simple, consisting of a tubular oeso-
phagus (Oe) which extends scarcely beyond the neck and opens

98: THE EMBRYOLOGY OF THE HONEY: BEE

into the mid-intestine (MInt). This is a capacious cylindrical
sack which occupies the greater part of the space contained within
the trunk, extending caudad to the twelfth segment, where it ter- ©
minates blindly in a rounded end. It is succeeded by the short
hind-intestine (HInt) whose inner (cephalic) end is also blind, but
whose caudal end opens to the exterior by an opening on the
obliquely truncated caudal face of the terminal or thirteenth seg-
ment. Opening into the cephalic end of the hind-intestine are two
pairs of slender tubes, the Malpighian tubules (Mal), which lie
against the lateral faces of the mid-intestine and pursue a sinuous
course cephalad to the fifth trunk segment, where they end blindly.
The function of the Malpighian tubules is supposed to be
excretory.

Traversing almost the entire length of the trunk and lying near
the ventral body wall are a pair of straight tubes, less slender
and with somewhat thicker walls than the Malpighian tubules, the
silk glands (SIkGl). Their caudal ends are blind, at their ceph-
alic ends they contract somewhat, their walls at the same time
becoming thinner, they then unite in the neck region to form a
single median thick walled duct which opens at the end of the
labium.

The nervous system comprises a brain (Br) and a ventral nerve
cord (VNC). The brain is a large mass situated in the dorsal
half of the head, set astride, as it were, of the oesophagus. It
consists of two lateral lobes united together in the mid-line by a
slender commissure. Each lobe contracts ventrad to a slender
stalk which joins with the anterior end of the ventral cord be-.
low the oesophagus. The ventral cord consists of 12 successive
swellings or ganglia. These, when looked at from above or below
are seen to be double, and in fact each represents a pair of gang-
lia fused in the mid-line (Figs. XIII and XIV). Each of the
pair is connected with the ganglia of the same side in the immedi-
ately adjacent segments by a connective, thus the entire cord
appears as though pierced by a series of intersegmental round
openings, and therefore somewhat resembles a ladder in form
(Figs. XIII and XIV). The first ganglion is much thicker and
longer than any of the others, this is the suboesophageal ganglion
(SoeGng) and is a compound ganglion composed of the three pairs
of ganglia belonging to the three segments represented by the

en a ae a a D

5

a

THE EMBRYOLOGY OF THE HONEY BEE 99

mandibles, and the first and second maxillae, respectively. The
last or twelfth ganglion is also compound, probably also represent-
ing the fused ganglia of three segments.

The respiratory system consists essentially of ten pairs of spira-
cles (Sp) and two longitudinal tracheal trunks (LTraT), one on
each side. The spiracles are arranged in two longitudinal rows,
one on each side of the larva, each trunk segment, from the second
to the eleventh inclusive, bearing a single pair. The spiracles con-
sist of plate-like thickenings of chitin. In the center of each of
these is an opening leading into a narrow and short tubular branch
which in turn connects with the longitudinal tracheal trunk of
each side. These traverse the entire length of the body, lying
between the mid-intestine and the body wall. At their anterior
ends each tracheal trunk bends around the anterior end of the
mid-intestine above the oesophagus to join its counterpart of the
opposite side, thus forming a closed loop. At the posterior end
of the mid-intestine the posterior ends of the longitudinal trunks
form a similar loop, which is situated below the hind-intestine.
The longitudinal tracheal trunks give off both dorsal and ventral
branches which arise near those leading to the spiracles. The
larger ventral branches join with the corresponding branches of
the opposite side to form a series of tracheal loops or commissures,
one of which is found in each trunk segment, from the first to the
eleventh inclusive. The dorsal branches split up into finer branch-
es which are distributed to the dorsal half of each segment.

The heart (Ht) is a thin walled tube, slightly constricted at
the boundaries between the segments, and extending the length
of the trunk in the dorsal mid-line, above the mid-intestine.

The rudiments of the ovaries (Ov) are seen as somewhat flat-
tened elongated bodies lying above the mid-intestine in the dorso-
lateral region of the seventh and eighth trunk segments and
extending half way into the ninth segment.

After having briefly outlined the structure of the newly hatched
larva, the account of the development will be resumed with Stage
VIII, which follows immediately after the completion of the
germ layers (Stage VII).

Stage VIII, 44-46 hours (Figs. Villa and b). In comparison
With the previous stage the germ band has increased slightly in
breadth and also in length. At Stage VII its anterior end barely

100 THE EMBRYOLOGY OF THE HONEY BEE

reached the cephalic pole, and was expanded laterally to form
a pair of flat lobes, the procephalic lobes, rounded on their anterior
and lateral margins. At Stage VIII the germ band has lengthened
and its anterior end has now moved around the anterior pole of
the egg so that the procephalic lobes come to lie on its dorsal side
(Fig. VIIIb). They are now no longer simple flat expansions of
the germ band, but are subdivided into two successive pairs of
lobes, an anterior and a posterior pair. The anterior pair, the
protocerebral lobes (1Br), are the larger, and in form and posi-
tion may be compared to a saddle, placed on the dorsal side of the
anterior end of the egg. Their lateral ends are rounded and
extend ventrad half way down the lateral surfaces of the egg.
The second pair, the deutocerebral lobes (2Br), which with refer-
ence to the embryo, lie caudad of the first, are situated just a
trifle dorsad of the cephalic pole of the egg, and are also saddle
shaped but are somewhat smaller, and their lateral margins,
instead of being broadly rounded converge on each side
toward small rounded elevations; these elevations are the rudi-
ments of the antennae (Ant). In the mid-line of the embryo,
about half way between the protocerebral and deutocerebral lobes,
just dorsal to the cephalic pole of the egg, is a low elevation,
broad at its anterior (dorsal) end, narrow and cristate at its
posterior (ventral) end. This is the rudiment of the labrum
(Lm). Just below (caudad of) the labrum, at the cephalic pole
of the egg, is a shallow circular depression, the stomodaeum, the
common rudiment of the mouth and oesophagus. Immediately
behind the mouth and near the mid-line are a pair of low rounded
swellings, the tritocerebral lobes (3Br). Following these, but
nearer to the lateral margins of the germ band and in line with
the antennal rudiments are three successive pairs of low rounded
swellings. The first and largest of these are the rudiments of the
mandibles (Md), the next those of the first maxillae (1M), and
the last those of the second mavillae (2Mx). These appendages
with the segments to which they belong, complete the head of the
future larva and imago; the remainder of the germ band corres-
ponds to the trunk of the larva, and the thorax and abdomen of
the imago. On this portion of the germ band, destined to form
the trunk, eleven segments are plainly indicated in side view (Fig.
VIIIb) by narrow intersegmental light zones alternating with

THE EMBRYOLOGY OF THE HONEY BEE io1

darker ones ; the last two segments, the twelfth and thirteenth, are
however not yet clearly separated. On the ventral side the inter-
segmental light zones broaden out and unite along the mid-line,
so that here the segments are represented by dark processes
stretching out from the sides toward the mid-line. These darker
areas represent the mesodermal somites.

The germ band is of almost uniform width from the mandibu-
lar segment to the twelfth trunk segment, widening only a trifle at
the fourth and fifth trunk segments. On the three segments next
following the head, on the ventral side are seen the faint outlines
of three pairs of larger appendages, which are the rudiments of
the thoracic legs (1L, 2L, 3L).

Contemporaneous with the appearance of the rudiments of the
mouth parts and legs is the appearance of the invaginations which
form the silk glands and tracheae. The rudiments of the silk
glands (S/kG/) are plainly seen as a pair of small pale apertures
transversely elongated, situated just caudad of the bases of the
rudiments of the second maxillae. The common rudiments of
the spiracles and tracheal system appear as a linear series of ir-
regular apertures on each side of the germ band, a pair being evi-
dent on the second, third, fourth, fifth and sixth trunk segments,
the first pair being the largest and most distinct, and the last the
smallest and faintest. Since the definitive number, ten pairs, is
found in the stage next following, it is apparent that the tracheal
rudiments at Stage VIII are just making their appearance and
also that they develop successively caudad.

Stage IX, 52-54 hours (Fig. 1X). At this stage the germ band
has not altered in its general appearance or proportions but all of
the rudiments present in: the preceding stage are more sharply
defined. Those of the mandibles and maxillae are now well de-
fined ovoid swellings, distinctly marked off from one another
and plainly raised above the general level of the germ band.
The premandibular appendages or second antennae are at this
stage best developed, being well defined rounded eminences situat-
ed just in front of the mandibles and a trifle nearer the mid-line.
A line connecting them would pass just below the stomodaeal in-
vagination. The rudiments of the first pair of thoracic legs are
more clearly marked off than before, but those of the second and
third pairs are ill-defined, their position being still marked only

102 THE EMBRYOLOGY OF THE HONEY BEE |

by pale crescentic areas. Extending the entire length of the germ
band, from the stomodaeal invagination to the caudal end, is a
wide shallow groove, the neural groove. This is bounded laterally
by a pair of low ridges, the neural ridges. From the ectoderm
included in the neural groove and the neural ridges the ventral
nerve cord is formed.

The invaginations forming the silk glands have now lengthened ~

out to tubular sacs extending caudad and slightly laterad, but as
yet not reaching the posterior border of the first trunk segment.

The definitive number of tracheal invaginations on the trunk,
ten pairs in all, are now present, and appear as narrow slit-like
openings.

The stomodaeal depression has deepened, becoming funnel-
form.

The invagination forming the anus, the proctodaeum, and those
forming the Malpighian tubules, also appear at this stage, but
are not apparent from the ventral side. Their formation will be
more fully described in the sections devoted to the alimentary
canal.

Stage X. Estimated at 54-56 hours (Figs. X and Xa). At
this stage several well marked changes in the general appearance
of the embryo are noticeable. The cerebral lobes do not now
plainly show their subdivision into protocerebral and deutocere-
bral lobes, they form instead conspicuous hemispherical swellings,
one on each side of the head, which on the dorsal side are marked
off from the trunk by a deep constriction. The labrum (Lm) has
become a conspicuous conical process projecting out from be-
tween the cerebral lobes parallel to the long axis of the embryo
and is somewhat flattened dorsoventrally; its tip is seen to be
plainly bilobed.

The stomodaeal invagination (Sto) has lengthened to a short
tube, and its connection with the mid-intestine, whose outlines
are now plainly visible, is much more apparent. The tritocerebral
lobes are scarcely visible and are in fact in process of flattening
out and disappearing. The antennal rudiments and mouth parts
have changed but little, except that the latter have lengthened a
little in a dorsal-ventral direction and are beginning to take on
their characteristic forms, the mandibles being long papillate,
while the two pairs of maxillae are shorter and more blunt, some-

THE EMBRYOLOGY OF THE HONEY BEE 103

what enlarged and flattened at the tip, like a pestle. The rudi-
ments of the legs are now apparent as low rounded elevations in
line with the mouth parts. The first pair are fairly well defined,
the second and third pairs are faint, the last especially so. The
germ band as a whole has increased considerably in width, now
covering a trifle over a half of the entire surface of the yolk,
and is divided by slight constrictions into thirteen segments, the
definitive number. |

The silk glands have lengthened to slender tubes extending cau-
dad to the sixth trunk segment.

The openings of the tracheal invaginations, the future spiracles,
are now small round aperatures and by the widening of the germ
band have been carried so far laterad as to be virtually invisible
from the ventral surface. The invaginations themselves have
increased in size and extent and form flat sacs lying close against
the inner surface of the ectoderm.

The proctodaeal invagination (Pro) has made its appearance
as a funnel-shaped depression just dorsad of the caudal pole of
the egg while the Malpighian tubules (Mal) are seen as two pairs
of short blind tubes arising from its base.

Stage XI. Estimated at 56-58 hours (Fig. X1). The general
appearance of this embryo is very similar to that of the stage
preceding. The swellings representing the tritocerebrum have
_ totally disappeared. The germ band covers a little over one-half

of the circumference of the yolk. The thirteen segments of the
trunk are more sharply separated from one another, and a lateral
constriction slightly deeper than the rest marks the boundary be-

f . tween head and thorax. Both the silk glands and the Malpighian
_ tubules have increased in length. The tracheal invaginations have

each sent off tubular branches caudad, cephalad and ventrad.
Those directed caudad and cephalad have united with the cor-
responding branches in the preceding and succeeding segments to
form the longitudinal trunks, while the ventral branches are about
_ to unite with those of the opposite side to form the ventral loops
- or commissures (TraCom). At this stage the latter are seen
_ on the ventral surface as delicate sharply outlined tubes which as
yet have not quite reached the mid-line.

Stage XII. Estimated at 60 to 62 hours (Fig. XII). The
embryo now begins to resemble the larva decidedly in general

104 THE EMBRYOLOGY OF THE HONEY BEE

appearance. The head and trunk are sharply separated from
one another by a deep constriction. The head has become broader
and its sides rounder; the trunk segments are marked off from
one another by sharply defined constrictions. The germ band or
rather that part of it constituting the body wall of the future
larva has increased in width until it now virtually covers the entire
yolk, its lateral margins uniting along the dorsal mid-line. At the
same time the head has turned ventrad through an angle of about
forty-five degrees, so that the labrum, which formerly projected
nearly straight out in front, is now directed obliquely ventrad, and
the mouth, whose position up to this time corresponded to the
cephalic pole of the egg, is now brought to the ventral surface. A
similar change occurs at the posterior end of the embryo. The
proctodaeal invagination makes its first appearance on the dorsal
side of.the posterior end of the egg (see Fig. Xa, Pro), the poster-
ior end of the germ band since Stage VII having been curved
around the caudal pole of the egg, and the embryonic hind-intestine
up to this stage has accordingly had an oblique course, being di-
rected obliquely caudad and dorsad. Now its direction corresponds
with the long axis of the embryo, and the anus having come to lie
precisely at the posterior end both of the embryo and of the egg.
Since at Stage VIII both ends of the germ band were seen to
have been bent around the two poles of the egg, these changes
therefore represent a partial straightening of the embryo, and
since the whole embryo does not appear to have materially in-
creased in length, judging by its relation to the chorion, it must
be assumed that the ventral half of the embryo has contracted, at
least to a certain extent.

The labrum has meanwhile changed little in form. The trito-
cerebral swellings have completely vanished, the area formerly
occupied by them being now smoothly rounded. The antennae
are large and somewhat conical in form, and directed cephalad.
The mandibles have not perceptibly changed. The first maxillae,
however, have lengthened slightly and their peripheral portions are
plainly seen to be turned inward. The second maxillae have be-
come bluntly conical in form, and slightly flattened dorsoventrally,
lying close to the ventral surface of the head, and directed ceph-
alad. At the same time their bases have approached the ventral
mid-line, carrying with them the openings of the silk glands.

THE EMBRYOLOGY OF THE HONEY BEE tos

4 The three pairs of leg rudiments are plainly outlined and consti-
tute low rounded protuberances. The ventral nerve cord is now
_ distinctly visible, on ventral view appearing as a dark band, seg-

mentally enlarged, and pierced intersegmentally by oval openings.

The tracheal system is well on its way to completion, as indi-
cated by the completion of the ventral tracheal loops (TraCom)
which are plainly seen on ventral view as delicate sharply outlined
tubes. .

The silk glands and the Malpighian tubules, the latter no. longer
visible from the ventral side, have continued to increase in length.

Stage XIII. Estimated at 62-64 hours (Fig. XIII). The
changes of most importance at this stage concern the antennae
and mouth parts. The former are less prominent than at Stage
XII, and are in fact in process of disappearing. Both the man-
dibles and first maxillae are bluntly conical in form and directed
inwards and forwards. The second maxillae have drawn still
closer together, and are becoming united to one another by their
mesial borders. At the same time the two external openings of
the silk glands are also brought close together. The second maxil-
lae unite by their edges only, overarching, as it were, the inter-
vening space on which the silk glands open, and which, after the
completion of the fusion of the second maxillae, becomes the
common unpaired duct of the silk glands which ultimately opens
at the tip of the labium. ,

Stage XIV. 66-68 hours (Fig. XIV). At this stage the de-
velopment within the egg is virtually completed. The principal
changes to be noted are: the almost total disappearance of the
antennal rudiments, of which only the faintest vestiges remain
visible on the exterior, and the completion of the union of the
second maxillae to form the labium. At 76-78 hours, eight hours
later—the embryo will rupture the chorion and elongate, being at
the same time flexed in a direction opposite to that of the embryo,

and become a young larva (Fig. XV) whose structure has already

been described.

3. Segmentation
It is generally believed by students of invertebrate morphology
that insects as well as other arthopods are made up of a linear
series of serially homologous segments, which in the primitive

106 ©6°THE EMBRYOLOGY OF THE HONEY BEE

ancestors of the highly specialized modern representatives of the
arthropods were probably essentially similar to one another, much
as they are in the modern chilopods or centipedes. Each of these
component segments consists of an annular section of the body
wall, provided with a pair of appendages, and containing a pair
of ganglia and a pair of coelomic (mesodermal) sacs. In the
insects, however, as well as in other modern representatives of
the arthropods, a differentiation has taken place among these
originally similar segments, accompanied also by a division of
labor. In the insects, for example, a certain number of the anter-
ior segments have become united to form the head, the appendages
of some of these head segments having meanwhile become altered
to form biting or chewing organs; other segments, such as those
of the abdomen, although otherwise but slightly modified, have
totally lost their appendages. For the most part the segments
composing the insect body are easily recognizable, except in the
cephalic region. Here the modification of the primative segments
has been especially great, and the identification of these has for
the past century been the subject of research, the problem hay-
ing been attacked first from the anatomical side, later from the
embryological. A complete account of the work done in this field
cannot be attempted here, but may be found in such special papers
as those of Heymons (1895a), Folsom (1900), Comstock and
Kochi (1902), and Riley (1904).

In insect embryos the segments originally present are much
more readily recognized than in the imago, the primative condi-
tions being more or less imperfectly reproduced or recapitulated,
and our present knowledge of the segmentation of the insect is
principally based on embryological rearches. A hypothetical gen-
eralized insect embryo, combining the results of investigation on
this problem is represented in figure 36.*°

‘A primary head and a primary trunk division are recognized

Tn this figure, borrowed from Snodgrass (1910), one segment of the
abdomen is omitted, eleven only being represented. Heymons (1895) has
shown that in the Dermaptera and Orthoptera twelve abdominal seg-
ments are present. The location of the mouth in the first segment may
also be criticized, since it is the opinion of Comstock and Kochi (1902)
that the mouth opening is located on the ventral surface of the third
(tritocerebral) segment.

THE EMBRYOLOGY OF THE HONEY BEE to7

Fic. 36. Diagram of a generalized insect embryo, showing the segmenta-
tion of the head, thorasic, and abdominal regions, and the segmental
appendages. From Snodgrass (1910).
in the insect embryo, as distinguished from the secondary or
definitive head and trunk divisions of the imago. The primary
head division comprises those segments whose ganglia unite to
form the supraoesophageal ganglion or brain. The number of
segments entering into this division has been variously estimated ;
Janet (1899), basing his conclusions on a study of the structure
of the head of an adult wasp, Vespa, enumerates seven, while
Carriére finds four in Chalicodoma, but it is probably safe to say
that the majority of investigators follow Viallanes (1887), Patten
(1889), and Wheeler (1893) in recognizing but three segments.
These are: (1) the protocerebral, (2) the deutocerebral, and
(3) the tritocerebral. The first of these is without true append-
ages ; the second bears the antennae, the third bears in some of the
more primitive insects vestigal appendages probably correspond-
ing to the second antennae of the Crustacea, as the researches of

108 THE EMBRYOLOGY OF THE HONEY BEE

Viallanes and others have shown. Following the three segments
of the primary head division come three, or possibly four seg-
ments which enter into the composition of the head of the adult
insect and which bear those appendages commonly known as the
mouth parts (mandibles and maxillae). The next three segments,
bearing legs, form the thorax. The remainder, twelve in number
in primitive insects (Heymons, 1895a), of which only eleven are
apparent in the majority of the embryos of pterygote insects, as in
the honey bee, form the abdomen.

The early appearance of the definitive segments of the honey
bee has already been described, but the precise identification of
these is uncertain, even at Stage VII, since the segments of the
primary head division are not clearly demarcated, and distinguish-
ing characters, in the form of embryonic rudiments, are yet entire-
ly lacking. At the Stage following (VIII), all of the segments,
excepting the terminal segments of the abdomen, are readily dis-
tinguishable and may now be identified, since the rudiments of the
appendages, and of the silk glands and spiracles have put in an
appearance. The protocerebral and deutocerebral segments (7Br,
2Br) are plainly distinguishable, in fact more so than at any other
time in the development. The protocerebral segment consists, as
already described, of two large lobes extending half way down
the sides of the egg, and rounded at their external margins. This
segment possesses no recognizable appendages, although certain
investigators have compared the labrum (Lm) which in many
insects is more or less bilobed, with a pair of fused appendages.
Heymons (1895), however, has pointed out that true appendages
arise laterad of the nervous system, while the labrum in all cases
arises in the mid-line between the two lateral halves of the ner-
vous system; its homology with a pair of fused appendages there-
fore seems very improbable. .The protocerebral segment, in
addition to the labrum, bears the compound eyes and the ocelli,
and contains a certain amount of mesoderm, but no coelomic sacs.

The deutocerebral segment (2Br) consists also of two large
lobes. These are somewhat triangular in outline, and their ex-
ternal apices, which are, with respect to the germ band, directed
laterad and somewhat caudad, bear the papillate antennal rudi-
ments. In certain insects the antennae are post-oral in position
at the time of first appearance, and in fact, in the honey bee a

THE EMBRYOLOGY OF THE HONEY BEE _ tog

straight line drawn between them would intersect the germ band
behind the mouth, yet their obvious connection with the deuto-
cerebral lobes precludes their being seriously considered as any-
thing but pre-oral in their relations in this insect. To this segment
belong also well developed coelomic sacs, as will appear later.
The tritocerebral (second antennal, intercalary or premandibu-
lar) segment (3Br) is, in the honey bee, of more than ordinary
interest, but since Butschli (1870) in the embryo of this insect
first called attention to the presence of appendage-like swellings
in front of the mandibles, between these and the antennae, thus
occupying the region afterwards identified as that of the trito-

cerebral segment. “Close in front of the anterior angles of the

cephalic plates, between the mandibles and those angles, there
arises a conical projection, springing from each of the germinal
ridges, which attains a considerable degree of development and
appears almost like a pair of inner antennae” (p. 538). Grassi
(1884) also observed these appendages, considering them as
the first pair of mouth parts. In another place (p. 201) he
states “The first pair of mouth parts, which has an ephemeral
existence, may possibly be compared to one pair of the
antennae of the arthropods.” Tichomiroff (1882) also noticed
a pair of transitory swellings in front of the rudiments of the
mandibles of the embryo of the silk worm. Wheeler (1893) de-
scribed unmistakable vestigial appendages on the tritocerebral seg-
ment of the collembolan Anurida. Heymons (1895), although
expressing his disbelief in the presence of appendages on this
segment in pterygote insects, recognized its presence in the Derm-
aptera and Orthoptera and found that it was provided with rudi-
mentary coelomic sacs in these orders. Later researches (Uzel
1897, Claypole 1898, Folsom 1900), have shown that true append-
ages appear on the tritocerebral segment in several apterygote
insects, and Uzel has found that in certain forms these appendages
may even persist in the imago. Moreover Riley (1904) has de-
scribed traces of appendages on the tritocerebral segment in the
cockroach. In the honey bee the swellings on the tritocerebral
segment first appear at Stage VIII, and attain their greatest
development at the stage next following, Stage IX. During these
stages they form conspicuous low conical swellings situated a
trifle nearer the mid-line than the rudiments of the mouth parts,

110 THE EMBRYOLOGY OF THE HONEY BEE

while a straight line connecting them would pass close to the
posterior edge of the stomodaeal invagination. Examination of
cross sections through the region of this segment at a stage a
trifle older than Stage VIII demonstrates that the bases of these
swellings are nearly in line with the bases of the rudiments of
the mandibles and maxillae, but that these swellings do not resem-
ble the rudiments of the mouth parts in their structure. If the
appendage-like swellings on the tritocerebral segment at Stages
VIII or IX are compared with the rudiments of the antennae or
the mouth parts a difference in their histological character be-
comes at once apparent. The rudiments of the true appendages
are ectodermal evaginations composed of numerous small and
slender cells with a core of mesodermal tissue; those of the trito-
cerebral swellings on. the other hand are composed of an outer
layer of cells destined to form hypodermis, and an inner layer
of neurogenic cells, continuous cephalad with the rudiments of the
protocerebrum and deutocerebrum, caudad with those of the ven-
tral cord. In brief, the tritocerebral swellings are composed princi-
pally of neurogenic tissue, and at a later stage the tritocerebruim is
actually split off from the ectoderm of this region. With these facts
in view the probability that these swellings represent appendages
diminishes and it therefore seems more appropriate to consider
them as exaggerated ganglionic swellings. At Stage X the trito-
cerebral swellings have become almost invisible, this region of the
head having flattened out, and before Stage XI is reached the
swellings have totally disappeared. Butttschli (1870) states that
these rudiments fuse to form a sort of lower lip for the larva.
An inspection of Stages VIII to X shows that this is not the case,
but that there is on each side a fold which appears to be a con-
tinuation of the cephalic edge of the mandibular rudiments and
which extends cephalad and mesiad and terminates on the ex-
ternal side of one of a pair of rounded eminences. An inspection
of the transverse sections shows that these eminences are only
the anterior ends of the neural ridges, as shown in the figures of
these stages, and it seems also probable, judging from the rela-
tion of the parts, that these swellings belong rather to the mandi-
bular than. to the tritocerebral segment. Recently (1909)
Hirschler has described in Donacia a pair of similar swellings,
situated on the tritocerebral segment near the ventral mid-line,

THE EMBRYOLOGY OF THE HONEY BEE 111

which he terms “hypopharyngeal papillae,” and which he believes

1 - contribute to the formation of the definitive hypopharynx. These,

however, are clearly not to be considered as comparable to seg-
mental appendages, for obvious reasons.

The three following segments, the first three of the primary
trunk division, the mandibular, and the first and second maxillary
segments, are in most respects similar to those of the embryos
of other pterygote insects. The appendages (Md, 1Mx, 2Mx)
up to the time of hatching are in the honey bee simple in
form, and do not exhibit any differentiation into lobes. Folsom
(1900) has discovered in Anurida, a primitive apterygote insect,
plain evidences of an additional segment intercalated between
the mandibular and first maxillary segment and bearing on its ven-
tral surface a pair of papillae, representing a pair of rudimentary
appendages (Fig. 36, Slin). These fuse with a median tongue-
like evagination of the first maxillary segment (Lin) to form the
hypopharynx or lingua of the mature insect. This segment, with
its corresponding structures, is wanting in the embryo.of the honey
bee as is apparently also the case in other pterygote insects.

The three segments next in order bear the legs (1L, 2L, 3L)
the rudiments of which appear on the surface, at Stage X, as low
protuberances, but are withdrawn from view shortly before
hatching.

The remaining segments, twelve in number, constitute the defin-
itive abdomen of the adult in insects other than those of the
Hymenoptera. In this order, as is well known, the first abdominal
segment is usually united to the third thoracic segment (metatho-
rax) and may be said to constitute a part of the definitive thorax.
In the embryo of the honey bee, as in those of most of the higher
pterygote insects, the abdomen consists of eleven segments plus a
terminal or anal segment. Only ten abdominal segments anterior
to the terminal segment are usually visible from the exterior
(Figs. Xa and XI). Sections cut in the sagittal plane through
embryos of Stage XI-XII (Fig. 48A) show that just anterior to
the short terminal segment there is a short and much reduced
eleventh segment, from which a rudimentary pair of ganglia are
produced. The number of segments present in the abdomen of
the honey bee is therefore the same as that found by Heymons
(1895, 1895a) in the Dermaptera and Orthoptera.

112 THE EMBRYOLOGY OF THE HONEY BEE

The embryos of nearly all insects, as is well known, display at
some time during development more or less evident rudiments
of appendages on the ventral surface of the abdomen, as shown
in figure 36. Those on the first segment of the abdomen are fre-
quently very conspicuous and may, before the close of develop-
ment, attain quite a considerable size, as in Melolontha (Graber
1890). Butschli (1870) has described rudiments of appendages
on all of the abdominal segments of the embryo of the honey
bee, the last two being especially conspicuous. ‘The three thor-
acic segments are provided with short appendages projecting
laterally and posteriorly, and by careful observation of the follow-
ing trunk segments a similar but very ill-defined process may be
observed on all of them; in the twelfth and thirteenth segments
these appendages are so much developed that they can be designa-
ted as the anal appendages” (p. 537). Biitschli’s statements
however, are not well supported by his figures, since these do
not show the abdominal appendages with sufficient definiteness to
be convincing, with the exception of the last two, belonging to the
ninth and tenth abdominal segments. These are quite definitely
shown. Grassi (1884) apparently did not see the less well defined
abdominal appendages described by his predecessor, but states that
in several instances he saw two pairs of processes near the hinder
end of the embryo, at a stage when the rudiments of the mouth
parts had scarcely made their appearance. In Grassi’s drawings
(Tab. IV Fig. 10) these processes have much the appearance of
artefacts, and his statement that these appendages frequently
could not be seen, lends probability to this assumption. The
writer was never successful in finding anything which could be
safely construed as abdominal appendages. They certainly occur,
nevertheless, in certain Hymenoptera. Carriére (1890) observed
unmistakable rudimentary appendages on the first and second seg-
ments of the abdomen, and indications of them on the third and
fourth segments. All of these soon vanish. The embryo of Hylo-
toma (Graber, 1890) possesses twelve pairs of well defined abdom-
inal appendages, but this is to be expected, since Hylotoma is a
much more generalized form than Apis or Chalicodoma. Both
Wheeler (1910) and Tanquary (1913) report finding rudimentary
appendages on the abdominal segments of ant embryos.

|
1 :
:
;
E

ee a ge ge

_— A a

—-

VIII

Nervous SYSTEM

1. Newly hatched larva

The nervous system of an insect consists of a brain, ventral
nerve cord, and a complex of ganglia associated with the brain,
the stomatogastric system. The ventral nerve cord and the brain
of the young larva—Stage XV,— have already been briefly de-
scribed but before entering upon an account of their development
it will be desirable to examine their structure more closely. The
ventral cord, as already said, consists of seventeen ganglia, one
for each of the primary trunk segments. These ganglia are act-
ually double, each representing a pair of simple ganglia united in
the mid-line. The first three of these double ganglia, belonging to
the mandibular and the two maxillary segments, are closely united
to form a large elongated compound ganglion, the suboesophageal
ganglion (Figs. XV, 38, 39, 42, 44 and 45, SoeGng), situated in
the lower half of the head, below the oesophagus, and supplying
nerves to the mouth parts. Its compound nature is clearly indi-
cated by the presence of two transverse constrictions dividing the
ganglion into three subequal parts (Fig. 39). Springing from the
anterior end of the ganglion is a pair of stout connectives leading
to the brain and embracing the oesophagus, the circumoesophageal
commissures, while from its posterior end two similar connectives
Or commissures unite it to the ganglion next following, the first
thotacic. The eleven ganglia following, including the first thor-
acic, are essentially alike except as to their size, the first thoracic
being the largest, the second and third thoracic ganglia a trifle
smaller, while the abdominal ganglia diminish slightly in size to-
ward the posterior end. The last three ganglia are united to
form a compound ganglion, but the last of the series, the eleventh
abdominal, is so slight that it may be considered as vestigial. Each
of the ganglia (Fig. 37, A, C and D) is flattened in form, about

113

114 THE EMBRYOLOGY OF THE HONEY BEE

Mcl Nv
Z <P me es Se : an Sa

ord Se . " eh (r 3) ——

oe Ce SO ee ges OS @is Nntentthe Jae
Qe: () €:) CCS - > “a oes torn ; eS EE : ese = Gass
Saas SS Ol Ge NESS

ul

2

Hyp

|

:

]

)- §F-TraCom

LS

Fic. 37. Sections illustrating the structure of the ventral cord of a ~ .
newly hatched larva (Stage XV), x 900. A, transverse section through the
second thoracic ganglion, showing the lateral nerves (Nw), neurilemma 4
(Nim), commissure (Com), and neuroblasts (Nbl). B, transverse section }
through the connectives between the second and third thoracic ganglia,
showing nerve fibres (NvF) and neurilemma (Nim). C, sagittal section ;
of the third thoracic ganglion. D, coronal (horizontal) section through :

the third thoracic ganglion.

a a ee |

THE EMBRYOLOGY OF THE HONEY BEE 115

twice as broad as long, with a slight median depression on its ven-
tral surface. From its anterior and posterior ends arise thick
connectives joining it to the ganglia of the adjacent segments.
Each ganglion also sends out laterally a large nerve trunk
(Fig. 37A, Nv) which soon breaks up into smaller branches and
these are distributed to various parts of the internal organs and to
the body wall.

Histologically each ganglion is composed of a mass of cells,
the ganglion cells, each of which sends off a slender nerve fibre.
These fibres unite in the upper half of each ganglion with the
nerve fibres from other ganglia to form bundles or strands (Fig.
37A). A large longitudinal strand traverses the lateral halves of
the ganglia and passes through the center of the connectives ;
these two strands thus serve to link together all of the ganglia of
the nervous system, including those of the brain (Fig. 37B and
D). Two other smaller strands, the transverse commissures
(Fig. 37A and D, Com) pass close together side by side from one
side of each ganglion to the other thus joining its lateral halves.
The nerve fibres in a ganglion consequently form a figure like the
letter H, the cross bar of the H, however, being double (Fig.
37C). Each ganglion is thus incompletely divided into five
regions ; two situated laterad of the longitudinal fibres, and three
in the mid-line. These Wheeler (1893) following Graber, has
named the Jateral, the anterior, the central and the posterior gang-
liomeres. Other nerve fibres also pass out from each ganglion
laterally, forming the core of the lateral nerves (Fig. 37A).

An exceedingly thin and delicate cellular layer, the neurilemma,
(Fig. 37, Nim) covers the exterior of the ventral cord and also
of the brain.

The brain of an insect (Fig. 38) is to be regarded as composed
essentially of three double (paired) ganglia: the protocerebrum
(rBr), the deutocerebrum (2Br), and the tritocerebrum (3Br),
corresponding respectively to the protocerebral, deutocerebral, and
tritocerebral segments. These brain divisions consist of pairs of
more or less evident swellings, each of which is made up of a
mass of ganglion cells traversed by a thick central core of nerve
fibres continuous with those of the longitudinal connectives of the
ventral nerve cord. The protocerebrum of the young bee larva

116 THE EMBRYOLOGY OF THE HONEY BEE

“ss om! £- ide
See Se A D
oes Ne > ke

ae mM ‘
StgNv- %, '2Br(AntL)
4 ‘3B
SoeCom* / ac
“*SoeGng

ye Ps

Fic. 38. Diagram of the nervous system of an embryo of Mantis,
showing the head and first three trunk segments. From Viallanes.

includes the greater portion of the brain and is made up of a pair
of large lobes (Figs. 39-44, 7Br) situated in the upper part of
the head capsule. These lobes are somewhat ovoid in form, their
larger ends directed caudad and slightly laterad; their smaller
anterior ends lie close to the anterior wall of the head. Each
lobe is somewhat compressed at right angles to the external sur-
face. At their anterior ends the lobes are closely apposed to
each other and are here united by a thin bridge of nerve fibres, the
supraoesophageal commissure (Figs. 43 and 44, SupCom). Prob-
ably some fibres from the second brain division, the deutocere-
brum, also enter into this commissure.

Each half of the protocerebrum of insects is typically divided
into three lobes, which in many cases, as in the Dermaptera and
Orthoptera (Fig. 38) are marked off from one another by thick-
enings of the ectoderm forming the wall of the head capsule.

OO

ee ee

i ale ta Ro

—_——

eS ee. ee

te ee

THE EMBRYOLOGY OF THE HONEY BEE 117

Moreover, as well be seen by consulting the diagram represented
in figure 38 the halves of the protocerebrum are typically directed
laterad as well as cephalad. In the larva of the bee the halves
of the protocerebrum are, as mentioned above, directed caudad
and laterad (Fig. 55). This difference is due to the bending of
the anterior end of the germ band and of the resulting embryo
around the cephalic pole of the egg and must be constantly kept
in mind in comparing the development of the brain of the bee
with that of other insects. Those forms in which the development
of the brain has been most thoroughly worked out belong princi-
pally to the order Orthoptera, and in comparing the embryonic
development of the brain of the bee with that of those insects,
another important point of difference must be also considered.
In the Orthoptera, for example, the development is without the
intervention of a larval stage, that is, direct from the embryo to
a form completely equipped for active life, and provided, like the
mature insect, with functional eyes, antennae and appendages.
In the bee, on the other hand, the embryonic development termin-
ates in a larva, a form not adapted to an independent existence,
and very different structurally from the imago. Many parts of
the larva are correspondingly but slightly developed, and may
be said to linger in a more or less latent or embryonic condition
until near the time of pupation. This is especially true of the
brain. At the time of hatching, the brain of the bee is, generally
speaking, only comparable to the brain of the embryo of those
insects with a direct development.

The three lobes of the protocerebrum of the insect brain have .
the following fate, as determined by Viallanes (1891), Wheeler
(1893) and Heymons (1895) for the Orthoptera and Dermap-
tera. The first (outermost or anterior) forms the optic lobe,
the second the optic tract, while the third lobe forms those parts
of the imaginal brain included in the protocerebral lobes. On
account of its relatively undeveloped condition and the absence of
ectodermal thickenings these three divisions of the protocerebrum
of the bee are not marked off with the same clearness as in the
Orthoptera and Dermaptera, nevertheless three lobes are distin-
guishable, both on surface view (Fig. 39) and in longitudinal
section (Figs. 43 and 44). The first lobe (rBr,), morphologically

118 THE EMBRYOLOGY OF THE HONEY BEE

Br, (OpL) i Br,

7 IBr3
x
Ne _2Br(AntL)
PhyGng i
----FtGng

tity

Fic. 39. Cephalic portion of the nervous system of a newly hatched larva
(Stage XV), lateral aspect, drawn with the camera lucida from an entire
preparation, and corrected with the aid of sections, x 243.

the anteriormost (Fig 38), but in the bee the posteriormost, is the
broadest of the three, and subspherical in form. Its posterior
half, separated from the rest of the brain by a groove, is com-

Fic, 40. Transverse section through the head of a newly hatched embryo
(Stage XV), passing just in front of the mouth, showing part of the
protocerebrum (zBr) and the frontal ganglion (FtGng), x 337.

ee

THE EMBRYOLOGY OF THE HONEY BEE = 119

posed of a thick plate of cells whose cytoplasm in many prepara-
tions stains more deeply than the remainder of the brain, and
whose nuclei are unusually large, clear, and sharply stained (Figs.
39 and 44 4).

The second or middle lobe of the protocerebrum (zBr,) is
very short, its length being about one-half its width.

The third lobe (zBr,) is about one-third the length of the cor-
responding lateral half of the protocerebrum and at its anterior
end meets its mate of the opposite side and is here united to it
by the supraoesophagael commissure (Fig. 43, SupCom).

The deutocerebrum, or antennal lobes, the second division of
the brain (Figs. 39, 40 and 41, 2Br [AntL] manifests itself ex-
ternally as a pair of conspicuous rounded swellings of the antero-
lateral face of the brain just above the oesophagus and below
the protocerebral lobes. From the fibrous core of each of the

StgNv

BOE TR

1 Ten---- 2. ;

SlkDO

Fic. 41. Transverse section through the head of a newly hatched larva

(Stage XV) intersecting the circumoesophageal commissure and showing

part of the protocerebrum (Br), the deutocerebrum (2Br), the tritocere-

brum (3Br), and the anterior end of the suboesophageal ganglion
(SoeGng), x 337.

120 THE EMBRYOLOGY OF THE HONEY BEE

deutocerebral lobes is given off a thin strand of nerve fibres, the
rudiment of the antennal nerve (Fig. 41, AntNv) which passes
laterad to the antennal rudiment (Antik).

The tritocerebrum is in the larva, as in the adult, a rather ill-
defined region of the brain; its two halves are virtually continuous
ventrad with the halves of the circumoesophageal commissures,
dorsad they are fused with the antennal lobes or deutocerebrum.
Each is united to its mate of the opposite side by a long and thin
band of nerve fibres, the suboesophageal commissure (Figs. 41

coe

SIkDO SoeGng

Fic. 42. Transverse section through the posterior region of the head of
a newly hatched embryo (Stage XV), intersecting the optic lobes (1Br,
OpL), the suwboesophageal ganglion (SoeGng), and the corpora allata
(CorAll), x 337.

and 45, SoeCom), which passes beneath the oesophagus and in-—

tersects the median plane between the oesophagus and the anterior
end of the suboesophageal ganglion. The two halves of the trito-
cerebrum are in addition connected with one another and with

THE EMBRYOLOGY OF THE HONEY BEE J tai

the frontal ganglion (Figs. 39 and 40, FtGng) lying between
them, by short thick cords of nerve fibres covered on the exterior
by a layer of ganglion cells, the frontal nerves (Fig. 40, FtNv).
The two nerves composing this commissure arise on each side
from the anterior face of the tritocerebrum and at their point
of origin from the latter is the root of a nerve which passes
cephalad to the labrum, the Jabral nerve (Figs. 39 and 40 LmNv).

The histological structure of the brain is simple and essentially
similar to that of the ganglia. Each lateral half consists of a ~
mass of ganglion cells covered by the neurolemma and surrounding
a central mass of nerve fibres continuous with that of the cir-
cumoesophageal commissures. ;

The stomatogastric system of the newly hatched larva consists
of the following parts: the frontal ganglion, the stomatogastric

Fic. 43. ‘Coronal (horizontal) section through the right half of the
protocerebrum of a newly hatched larva (Stage XV), x 337.

122 THE EMBRYOLOGY OF THE HONEY BEE

1Bro 'Bry (OpL)

Fic. 44. Sagittal section of the head of a newly hatched embryo
(Stage XV), passing laterad of the median plane and intersecting the
right half of the brain (Br), the circumoesophageal commissure (OeCom)
and the suboesophageal ganglion (SoeGng), x 337.

nerve or nervus recurrens, and a pair of ganglion lying against

the sides of the posterior region of the oesophagus, the pharyngeal —

ganglia. To these may be added, for the sake of convenience, the
so-called ganglia or corpora allata of Heymons, which in fact are
probably not ganglia at all, according to Heymon’s own state-
ment (1897), (Figs. 39 and 42, CorAll).

The frontal ganglion (Figs. 39, 40 and 45, FtGng) is a large and
conspicuous pyriform ganglion with its blunt end directed ceph-
alad, lying in the median plane at the base of the labrum and just
above the oesophagus. It consists of a compact mass of ganglion
cells surrounding a core of nerve fibres. From its anterior end

nerve fibres, accompanied by a few scattered ganglion cells, run

to the tip of the labrum. At its posterior end the frontal ganglion
becomes continuous with the stomatogastric or recurrent nerve
(Figs. 41, 42 and 45, StgNv) which runs backward in the median

KS

THE EMBRYOLOGY OF THE HONEY BEE = 123

Oc ( Flint)

_StgNv
4

_-ssupCom

Fig. 45. Median sagittal section through the head of a newly hatched
larva (Stage XV), intersecting the supraoesophageal commissure
(SupCom), the frontal ganglion (FiGng), the stomatogastric nerve
(StgNv), the oesophagus (Oes), and the suboesophageal ganglion
(SoeGng), xX 243.

Fic. 46. Transverse section of the oesophagus, showing the pharyngeal
ganglia (PhyGng), x 534.

plane between the two halves of the brain just above the oesopha-
gus. This nerve consists, near its point of origin, of a thick
bundle of nerve fibres surrounded by a sheath of ganglion cells,

124 THE EMBRYOLOGY OF THE HONEY BEE

but diminishes rapidly in calibre as it passes caudad. At a point
slightly cephalad of the junction of the oesophagus and the mid-
intestine the recurrent nerve divides to form two small and rela-
tively insignificant ganglia, the pharyngeal ganglia (Figs. 39 and
46, PhyGng). These are flat and lie closely applied to the lateral
faces of the oesophagus.

The corpora allata (Figs. 39, and 42, CorAll) are two conspicu-
ous spherical masses of cells which most resemble ganglion cells,
but lack nerve fibres. These bodies lie some distance apart in the
posterior part of the cavity of the head, just below the level of
the lower edge of the oesophagus. These bodies are attached
dorsally to the thin walled sacs which constitute the coelomic
sacs of the antennal segment.

2. Development

Before entering upon a description of the development of the
nervous system of the honey bee it will be profitable and in fact
almost necessary first briefly to describe the course of develop-
ment of the nervous system in the Orthoptera, a relatively primi-
tive group, in which the intimate details of the process reveal
themselves sharply and distinctly. The broader features of the
development of the nervous system in the honey bee are simple
and readily understood; the minuter details of the origin of the
development of the nervous system, in other words its histogenesis,
is on the other hand much more difficult to follow and relatively
obscure, in fact so much so that it would be decidedly difficult of
interpretation without the assistance afforded by previous work
on this subject.

The nervous system of the insect arises, as was first shown by
Hatschek (1877), from two longitudinal thickenings or ridges
extending along the ventral side of the germ band, one on each
side of the mid-line. These ridges, the neural ridges (Fig. 47,
NIR) enclose betwen them a narrow median groove, the neural
groove (NIG), the internal infolded portion of which forms the
median cord. From the intrasegmental portion of the median
cord and the neural ridges the ganglia are formed by a process
of splitting off or delamination, the outer portion of the ectoderm
remaining as the ventral hypodermis; in the interganglionic re-

eT Pe eee

Se a ee eS

i i A i

a

THE EMBRYOLOGY OF THE HONEY BEE 125

Fic. 47. Transverse sections through the ventral ectoderm of embryos
of the grasshopper, Xiphidium. A, early stage, showing segregation of the
neuroblasts (NbI); B, late stage, showing the neuroblasts and their pro-
ducts, the ganglion cells (GngC). From Wheeler (1893).
gions the neural groove remains intact as hypodermis, the neural
ridges here forming the connectives.

The histogenesis of the neurogenic tissue constituting the neural
ridges and neural groove has been studied by several investigators,
but Wheeler (1891, 1893) was the first to give a complete and
consistent account of this process, as observed in the grasshopper
Xiphidium. The essential features of Wheeler’s results have
since been confirmed by a number of investigators of the Orthop-
tera and also of other orders, among these are Heymons (1895)
for the Dermaptera and Orthoptera, Carriére and Biirger (1897)
for the mason bee, and Lecaillon (1898) for the beetle Clytra.
Wheeler’s account is briefly as follows:

At a stage soon after the completion of the formation of the
germ layers the ventral ectoderm of the embryo is seen to consist
of two different kinds of cells; a small number of large pale
cells with vesicular nuclei, and a large number of smaller cells

1246 THE EMBRYOLOGY OF THE HONEY BEE

with small dark nuclei. The former will give rise to the nervous
system and are therefore termed neuroblasts (Fig. 47A, Nodl),
the latter form only the hypodermis and the structures associated
with it; these cells as therefore termed dermatoblasts (Dbl). A
little later, when the neural ridges and neural groove make their
appearance, the neuroblasts become arranged in four or five rows
on either side of the mid-line, lying next to the inner surface
of the ectoderm. The neuroblasts next begin to bud off by mito-
tic division smaller and darker cells from their inner surface.
Fig. 47B). These constitute the ganglion cells (GngC). Each
neuroblast thus gives off by division a number of ganglion cells
which together form a colum placed at right angles to the outer
surface of the embryo, each column having at its base the parent
neuroblast. Besides these neuroblasts lying on either side of the
mid-line there is another set, the median cord neuroblasts, which
are situated in the mid-line, at the bottom of the neural groove
(Fig. 47B.) These neuroblasts are intersegmental in their arrange-
ment, one neuroblast being placed between every two segments.
Each median cord neuroblast gives rise to a heap of ganglion
cells lying on its inner surface. These latter become displaced
cephalad and form the posterior median portion of each ganglion.
The remainder of the ganglion is formed from the invaginated
ectoderm of the neural groove. From this portion, Wheeler be-
lieves, arise the cells which form the inner and outer neurolemma.
The ganglion cells are the functional nerve cells of the ganglia
and send off delicate processes which branch and thus form the
nerve fibres or fibrillar substance which constitutes the connectives
and the commissures.

The two divisions of the nervous system, the brain and the
ventral nerve cord form a continuous whole from their inception,
but for convenience they will be treated separately, following the
usual custom. The development of the brain is slightly in ad-
vance of that of the ventral cord and, in the latter, following
the general rule among arthropods, development progresses from
the cephalic to the caudal end. In the bee, however, the de-
velopment of the ventral cord is almost simultaneous in all
of the segments, the anterior being but slightly in advance of the
posterior.

eee

ee ee ee
: — ‘ = 5

THE EMBRYOLOGY OF THE HONEY BEE 127

A. The Ventral Cord
At Stage VII, when the formation of the germ layers is vir-

-tually completed the ectoderm of the germ band caudad of the

procephalic lobes forms a uniform and rather thick layer com-
posed of prismatic cells and covers the ventral half of the egg
(Figs. 32, 48A). The cells of the ectoderm are not, however,
all alike, those comprising its middle third having nuclei consid-
erably larger and somewhat clearer than those of its lateral por-
tions ; this median area may be termed the neurogenic area, since
it includes the cells from which the ventrad cord will arise. In
the mid-ventral line there is a narrow strip, three to five cells
wide, which is quite well defined in the anterior region of the
germ band, less so in the posterior regions, the median cord (Fig.
48A, MC). With this exception the cells comprised within the
neurogenic area are at this stage precisely similar as regards form,
size, or staining reaction.

At the next Stage, VIII (Figs. 46B and C) several notable
changes are evident. The ectoderm is greatly reduced in thick-
ness laterad of the neurogenic area and also along the ventral
mid-line. The neurogenic area therefore comprises two longi-
tudinal thickenings, the primitive swellings (Fig. 48B, PriSw)
which are separated by a median furrow, the meural groove
(NIG.) The primitive swellings do not owe their formation to
cell proliferation, but to changes in the form of the cells of differ-
ent areas, the primitive swellings alone preserving the thickness
of the ectoderm and the primatic form of its cells existing at
Stage VII.

The primitive swellings, also shown in figures VIII, IX and X,
are highest in the gnathal region, and diminish rapidly in height
caudad, becoming scarcely distinguishable in the abdominal re-
gion. Moreover they are broader and slightly higher in the intra-
segmental than in the intersegmental regions, as a comparison of
figures 48B and C, and 48D and 49A will show, being thus divided
into a series of segmental ganglionic swellings corresponding to
the future ganglia.

The median cord can now be readily identified; its cells form
the floor of the neural groove. At this stage the median cord is
much broader and thinner than either at the stage preceding, or

NbI

Fic. 48. Transverse sections through the ventral ectoderm of embryos
of the honey bee, showing segregation of the neuroblasts (NVI) and the
dermatoblasts (Dbl), and the formation of the ganglion cells (GngC)
by the granddaughter cells of the neuroblasts. A, thoracic region, Stage —
VII; B, middle of second thoracic segment, Stage VIII; \C, anterior half

of second thoracic segment, Stage VIII; D, second thoracic segment, Stage
IX.

THE EMBRYOLOGY OF THE HONEY BEE - tag

at subsequent stages, and has the appearance of being stretched
laterally. This is especially marked in the preparation from
which figures 48B and C were drawn.

The cells of the primitive swellings only in part retain the
prismatic form seen at Stage VII; many of them are now seen to
be either in the act of division or to have recently divided.
These cells, fall into two classes, according to their position or to
their mode of division. First there are those in which the division
is equal, the mitotic spindle lying near and parallel to the external
surface and the resulting daughter cells occupying a superficial
position. These cells (Dbl) correspond to the “dermatoblasts’ of
Wheeler and are destined to form hypodermis only. Second,
there are cells in which the division is unequal, the spindle lying
near the internal surface and usually directed at right angles to
it, the products of division consisting of a smaller central and a
larger peripheral cell. The larger cells in these pairs corres-
pond to the “neuroblasts” (Fig. 48B and C, Nb/) the smaller to
the “ganglion cells” of Wheeler. A difference in the arrangement
of the products of the dermatoblasts and neuroblasts becomes at
once evident; those of the dermablasts form ordinary epithelial
cells lying side by side; on the other hand each neuroblast to-

‘gether with its daughter cells form a compact and more or less

ovoid nest or mass of cells (Fig. 48B and C).

At Stage IX, (Fig. 48D and 49A) the neural groove is both
narrower and shallower than at the preceding stage, while at the
same time the median cord has become correspondingly narrower
and deeper, its cells resuming their characteristic long prismatic
form. In cross section the median cord now presents somewhat
the form of the letter V, its outer end forming the point, although
in fact it is more or less truncate or flattened and constitutes the
floor of the neural groove. The cells of the median cord are
therefore long tapering in form, their inner ends always being
wider than their outer.

At this stage the neuroblasts and dermatoblasts have become
completely segregated from one another, so that each primitive
swelling is divided into an inner neurogenic layer composed of cell
nests, the lateral cord, made up of the neuroblasts and the gang-
lion cells, and an outer dermatogenic layer made up of the pro-

130 THE EMBRYOLOGY OF THE HONEY BEE

ducts of the dermatoblasts. The latter constitutes a single tier
of cells which are growing progressively smaller by division, as
evidenced by the diminishing size of their nuclei. The form of
these cells ranges from cubical, near the mid-line, to long prismatic
at the lateral edges of the primitive swellings.

The neuroblasts, which are frequently to be observed in division
(Fig. 49A) are true teloblasts and suffer no diminution in size
and are becoming increasingly conspicuous owing to the reduction
in size of the dermatogenic cells. The outermost row of neuro-
blasts of each side is now seen to differ from the remainder in
giving off its daughter cells mesiad instead of centrad. In the
honey bee the arrangement of the neuroblasts is not quite so
regular as in many other insects, the Orthoptera for example,
but as a rule there are from three to five rows of neuroblasts in
each lateral cord. The number of rows is greatest in the middle
of each segment, as may be seen by comparing figures 48D and
49A. This difference exists in Forficula (Heymons 1895) and
Donacia (Hirschler 1909), and is quite probably of frequent
occurrence.

The number of ganglion cells has now increased, but not uni-
formly, since some cell nests possess as many as four, others
only one.

At Stage X (Figs. 49B, C, and D) several changes are notice;
able. One of the most evident is the flattening of the ventral
surface of the embryo, resulting in the almost complete disap-
pearance of the external evidences of both the primitive swellings
and neural groove except in the gnathal region (Fig. X). Asa
result of the flattening of the external surface of the ventral
ectoderm the lateral cords are, so to speak, thrust inward and
project into the body cavity, especially in the intrasegmental
regions. This change in the contour of the ventral ectoderm is
probably brought about by the absorption of yolk and the con-
sequent diminution of its mass, causing its withdrawal from the
ventral ectoderm and a lowering of pressure upon the latter from
within. The mesial edges of the two halves of the mesoderm
now extend up to the lateral boundaries of the lateral cords and
about as closely against them (Fig. 49C) that it is frequently not
easy to distinguish the neurogenic from the mesodermai tissue,

THE EMBRYOLOGY OF THE HONEY BEE 131

i Renet FY NbI

R

f Fic. 49. Transverse sections through developing ventral cord. A,
| region between the first and second thoracic segments, Stage IX; B, first
7 mandibular segment, Stage X; C, first thoracic segment, Stage X; D,
posterior region of second thoracic segment, Stage X, x 600.

especially since the ganglion cells are very similar to those of
the mesoderm.

132 THE EMBRYOLOGY OF THE HONEY BEE

The separation between the lateral cords and the hypodermal
(dermatogenic) layer is more evident than before, being marked
in many sections by irregular clefts (Fig. 49C). The cells of the
hypodermal layer are still diminishing in size and now begin to
form an evident and well defined epithelium. The cell nests of the
lateral cords on the other hand, are becoming compacted together
to form the definitive ganglia. This is particularly noticeable in
the gnathal region (Fig. 50A) and at the anterior and posterior
ends of the thoracic segments (Fig. 49D).

In addition to the mitotic figures of the neuroblasts it is not
unusual at this stage to find other mitotic figures within the cell
nests of the lateral cords. Two of these mitotic figures are to
be seen in the section illustrated by figure 49D. Both the size of
these mitotic spindles and their relation to the adjacent cells make
it clear that they belong to the daughter cells of the neuroblasts.
It is not clear whether all of the daughter cells thus undergo
mitotic division but it is at least certain that many of them do.
Wheeler, in his earlier paper on the histogenesis of the nervous
system of the grasshopper (1891), believed that the daughter cells
of the neuroblasts did undergo division. This opinion was re-
linquished in his later paper (1893) on the same subject, as
regards the grasshopper, but not as regards Doryphora, in which
form, Wheeler expressly states, the daughter cells of the neuro-
blasts divide. On the other hand Heymons (1895) for Forficula
and Orthoptera, Lecaillon (1898) and Hirschler (1909), for the
chrysomelid beetles, and Escherisch (1902) for Musca, deny the
division of the neuroblast daughter cells.

At the stage next illustrated, Stage XI-XII, the entire ventral
cord—with the exception of its extreme posterior end—becomes
completely severed from the ectoderm and the ganglia begin to
assume their final shape. The separation of the lateral cords
from the overlying hypodermis was already evident at Stage X,
but at that stage the median cord was still embedded in the

hypodermis, the distal ends of the component cells of the cord -

forming the bottom of the neural groove. These cells have mean-
while been increasing slowly by mitotic division. These divisions
are equal, no evidences of median cord neuroblasts having been
observed. Meanwhile the median cord has also separated from

ae. ox Ke)

DESERT DE Om

Fic. 50. Sections through the embryonic ventral cord, x 600. A, gang-
lion and lateral nerves of the mandibular segment, Stage XI; B, ganglion
of second thoracic segment, Stage XI; C, connectives between the second
and third thoracic ganglia, Stage XI-XII; D, ganglion and lateral nerves
of second thoracic segment, Stage XIV; E, median sagittal section through
the ganglia of the first and second thoracic segments.

134 THE EMBRYOLOGY OF THE HONEY BEE

the hypodermis in the intrasegmental regions and now constitutes
the median portion of each ganglion. This separation of the
median cord from the hypodermis takes place by a progressive
attenuation of the already narrow outer ends of the cells of the
median cord until the external portion of the latter becomes re-
duced to thin strands, the hypodermis meanwhile closing in from
both sides until its lateral halves meet and unite in the mid-line.
The hypodermis along the ventral mid-line therefore owes its
origin—in the intrasegmental regions, at least,—to hypodermal
cells lying originally laterad of this region, as first noted by
Grassi (1884). Similar conditions also exist in many other in-
sect embryos, Melolontha (Graber 1890) for example. The
median cord, after its separation from the hypodermis moves
inward, frequently leaving behind it a temporary median notch
or groove on the inner surface of the hypodermis, as shown in
Fig. 50A. In this figure also are to be seen the delicate strands
of protoplasm connecting the median cord and the hypodermis.
This separation does not occur simultaneously throughout the
length of each segment, but appears to take place first in its
posterior half (Fig. 50E). In the intersegmental regions the
median cord remains an integral part of the hypodermis (Fig.
50C, (MC) ; the lateral cords here separate from both the median
cord and the hypodermis proper and constitute the connectives
Con).

In the lateral cords the number of ganglion cells has greatly
increased. In every segment a conical group of these lies along
each side of the lateral cord, laterad of the neuroblasts, while a
well defined single layer covers them on the ventral side. The
lateral cell groups apparently furnish the material for the forma-
tion of the lateral nerves, while the ventral layer provides the
material of the neurilemma. In the dorsal region of the lateral
cords a cleft is apparent, separating the uppermost tier of cells
from the remainder. These clefts are, so to speak, the forerun-
ners of the nerve fibres, which during the succeeding stages
traverse these spaces. Laterad of each ganglionic rudiment the
mesoderm (Fig. 50B, Meso) presses close against it in the middle
region of each segment, frequently, as at Stage X, making it
difficult to accurately determine the lateral limits of the ganglion
cells.

——————

7

ee ee a er

;

oF
a :
ty,
i
i
‘

4
i"

THE EMBRYOLOGY OF THE HONEY BEE 135

The final changes leading to the functional larval ganglion are
illustrated by figure 50D, Stage XIV. The most important
changes are: the development of nerve fibres, the development
of the lateral nerves, and the final incorporation into the ganglia
of the intraganglionic portions of the median cord.

The development of the nerve fibres corresponds essentially to
the accounts of Wheeler (1893) and Heymons (1895). The
ganglion cells of both the median and the lateral cords at about
Stage XII become pyriform; the smaller pointed ends of the
ganglion cells of the lateral cords being directed toward the clefts
already described, while the smaller ends of the cells of the median
cord are directed more or less dorsad. The smaller ends of the
ganglion cells now become more and more attenuated and finally
elongate into delicate protoplasmic fibres, the nerve fibres. Dur-
ing this process, in those regions where the commissures are
formed, a vacant space also appears in the dorsal half of the
median cord, in line with the spaces in the lateral cords, with
which it unites. These spaces are traversed from side to side
by the cell processes which are to constitute the nerve fibres of
the transverse commissures (Fig. 50D, NvF). These commis-
sural nerve fibres appear to arise from both the cells of the
median and lateral cords, but mainly from pyramidal groups of
cells situated immediately laterad of the median cord (see also
Fig. 37A).

The lateral nerves are apparently formed from the aggregations
of ganglion cells which at Stage XI-XII (Fig. 50B) constitute
the extreme lateral portions of the ganglionic rudiments. These
lateral cell groups are presumably derived by migration of gan-
glion cells which formerly lay within the semicircle formed on
each side by the neuroblasts, since there is no evidence that cells

are ever budded off from the neuroblasts in a laterad direction.

Moreover both the arrangement of these lateral cells, and the
fact that the number of ganglion cells lying immediately dorsad
to the neuroblasts is perceptibly diminished when the lateral
cell groups appear, tend to confirm this conclusion. The fate
of the lateral cell masses is clearest in the gnathal segments, since
in these the lateral nerves are short and straight, so that the entire
nerve together with the ganglion from which it arises may be

136 THE EMBRYOLOGY OF THE HONEY BEE

observed in a single cross section. Such a section, through the
mandibular segment at Stage XI-XII, is represented in figure
50A. Here the laterad extensions of the ganglia—composed of
ganglion cells—are seen in process of being directly transformed
into the lateral nerves innervating the mandibles. It will also
be noted that the clefts in the lateral cords, later filled by nerve
fibres, are in this section continuous with hollow spaces traversing
the centre of the lateral nerves, which spaces are also, at later
stages, filled by nerve fibres. Caudad of the gnathal region the
lateral nerves are much longer and relatively more attenuated,
and their development consequently more difficult to follow.

The median cord, up to Stage XI-XII (Figs. 49A and B), has
altered little histologically from its condition at Stage VII, being
for the most part still composed of rather long prismatic cells
with large and clear nuclei. Its cells nevertheless are slowly in-
creasing in number by mitotic division and decreasing in size.
These divisions are equal, and appear rather inconstant in direc-
tion. In figure 50B a mitotic spindle is seen, directed obliquely.
The size of the nuclei of the intraganglionic portions of the me-
dian cord, after their severance from the hypodermis, continues
to decrease, while at the same time they lose their characteristic
elongated form, until at Stage XIV their appearance is the same
as that of the ganglion cells of the lateral cords, having undergone
a similar differentiation into functional ganglion cells.

The intersegmental sections of the median cord, as already
stated, remain united to the hypodermis, except in the gnathal
segments and in the fifteenth, sixteenth, and seventeenth segments
(Fig. 51B). In these the median cord is taken up entire, the inter-
segmental portion being also severed from the hypodermis and
contributing to the formation of the ganglia of these segments,
which are fused to form compound ganglia prior to hatching. In
the gnathal region, however, a slight thickening of the ventral
hypodermis marks the boundary between the rudiments of the
mandibular and first maxillary ganglia. In the remainder of the
trunk region the intersegmental sections of the median cord separ-_
ate from the rest and form processes of the hypodermis, whose
form, prior to hatching, is illustrated by figures 50C and E. At the
time of hatching (Stage XV) the larva elongates, involving a

THE EMBRYOLOGY OF THE HONEY BEE — 137

Fic. 51. Median sagittal sections through the last three ganglia of
the ventral cord, corresponding to the 9th, roth, and 11th (true) abdom-
inal segment. A, Stage XI-XII, showing the ganglia separating from
the hypodermis of the oth, roth, and 11th abdominal segments. B, Stage
XV, showing two evident ganglia and the vestiges of a third, fused into a
compound ganlion, x 600.

stretching in the longitudinal axis of all parts of the trunk. The
effect of this process upon the intersegmental sections of the
median cord is a decrease in their height. At Stage XV they
have consequently the appearance of being mere transverse folds
of the ventral hypodermis (Figs. 37B and D). The number of
ganglia (or pairs of ganglia) in the ventral chain of the honey
bee is seventeen. Of these the first three are fused to constitute
the suboesophageal ganglion, and the last three similarly united to
form a compound ganglion, which in the young larva (Fig. XV)
is an oblong mass lying beneath the hind-intestine. This mass,
in sagittal section (Fig. 51B) is seen to be subdivided by con-
strictions into three successive swellings, the anterior two of
which are of nearly equal size while the third is much smaller.

138 THE EMBRYOLOGY OF THE HONEY BEE

The first two of the three swellings each possess two evident
‘commissures, clearly demonstrating that these two swellings
represent distinct ganglia, the third swelling on the other
hand, shows no evidence of commissures. In favorable trans-
verse sections, however, evidences of a few commissural fibres
may be seen in this swelling also. The best evidence for its
title to be considered as representing a pair of ganglia is obtained
in sagittal sections of Stage XI-XII (Fig. 51A). Here the ter-
minal swelling, like the two preceding it, is plainly seen to be
derived from a distinct trunk segment, the seventeenth (the
eleventh of the trunk), which is marked off by a well defined
constriction. There are therefore three pairs of ganglia repre-
sented in the terminal swelling of the ventral cord, the last pair
being rudimentary.

In the embryos of insects generally either sixteen or seventeen
pairs of ganglia are recognizable; in the latter case the last pair
is more or less rudimentary. Among the Hymenoptera Grassi
(1884) in the honey bee found sixteen pairs; Graber (1890) in
Hylotoma found sixteen pairs and the evident rudiments of a
seventeenth pair; Carriére and Burger (1897) in the embryo of
Chalicodoma found the rudiments of seventeen pairs of ganglia.
The investigators last named state that the rudiment of the last
(seventeenth) pair of ganglia is somewhat shorter than the others.
In figure XLV of their paper is a representation of the posterior
end of an extremely young larva, showing clearly the last three
pairs of ganglia of the ventral cord. These are here seen to be
subequal in size, the seventeenth pair appearing to be nearly if
not quite as large as the other two pairs.

The origin of the lateral cords in insects is fairly well estab-
lished. In every insect embryo thus far studied they arise, as
Hatschek (1877) discovered, from the inner layers of the primi-
tive swellings. Wheeler (1891, 1893) called attention to the
role played by the neuroblasts in the development of the lateral
cords. Subsequent investigators of the development of the
nervous system of insects have uniformly observed similar telo-
blastic cells whose behavior corresponds, with the exception of
minor differences, to the account given by Wheeler.

At the end of the embryonic period the neuroblasts undergo

THE EMBRYOLOGY OF THE HONEY BEE 139

degeneration in the Orthoptera (Viallanes 1891, Wheeler 1893,
Heymons 1895) and also in the Dermaptera (Heymons 1895).
In Donacia, according to Lecaillon (1898), they simply disappear.
Birger (Carriére and Biirger 1897), says (p. 370) in regard
to Chalicodoma: “I do not think that the cells I have designated
neuroblasts degenerate. Later they can not be distinguished
from the cells which they have produced.” In the honey bee
the neuroblasts do not degenerate before hatching, since at Stage
XV they are conspicuously visible at the periphery of the ganglia
(and brain as well) and are frequently dividing mitotically
(Figs. 37A, 37C, Ndi, see also Fig. 41).

The development of the median cord, unlike that of the lateral
cords, appears to differ considerably in different insects. That
it is derived from a median strip of the ventral ectoderm, forming
the floor of the neural groove, seems to be at least certain. The
ultimate fate of this strip is less uniform. All investigators of
the subject—with the exception of Wheeler—agree with Hatschek
(1877) that the intraganglionic sections of the median cord con-
tribute at least a large part—if not all—of the median portions
of the ganglia, including the transverse commissures. To this
opinion, however, Wheeler (1893) takes exception. This in-
‘vestigator, while admitting that in Xiphidium the intersegmental
regions of the median cord—the progeny of the median neuro-
blasts—are taken up into the central portions of the ganglia to
form functional ganglion cells, does not believe that the median
cord cells in the intrasegmental regions became ganglion cells,
but believes that they are used up in the formation of neurilemma.
In the same group to which Xiphidium belongs, the Orthoptera,
Heymons (1895) later found that the anterior and central median |
gangliomeres were actually formed by the median cord, much as
in other insects. With this exception, the differences in regard
to the development of the median cord center principally about
the fate of the intersegmental (interganglionic) sections. Hat-
schek (1877) stated that these remained in connection with the
ectoderm and contributed nothing to the ganglia.

Graber (1890) found that in Melolontha the median cord was
separated from the hypodermis throughout its entire length, but
that the intersegmental portions later divide transversely, each

1430 THE EMBRYOLOGY OF THE HONEY BEE

half being then drawn cephalad and caudad respectively into the
ganglia adjoining. In Hydrophilus, Lina and Stenobothrus on
the other hand the median cord was not observed to separate inter-
segmentally from the hypodermis.

Wheeler’s account of the development of the nerve cord of
the grasshopper Xiphidium has already been outlined at the be-
ginning of this section. The interganglionic sections of the median
cord were each represented by a deep invagination, forming a
part of the floor of the neural groove, and also by a median
neuroblast, the median cord neuroblast. The former, the ecto-
dermal invagination, produced no nerve tissue, but remained in
connection with the hypodermis and later formed, in the thoracic
region, the furcae, apodemes for the attachment of muscles; in
the abdomen these invaginations also occurred but persisted only
for a short time and then disappeared. The products of the
median cord neuroblasts on the other hand become displaced
cephalad and contributed to the formation of the posterior me-
dian neuromere.

Heymons (1895) in his researches on the Dermaptera and
Orthoptera has virtually confirmed Wheeler’s results, as concerns
the interganglionic portions of the median cord. Heymons
however found instead of one interganglionic neuroblast, several
of these cells. In Lepisma Heymons (1897) found that a con-
tinuous median cord was set free from the hypodermis and
present in the newly hatched nymph, extending the entire length
of the ventral cord.

Carriére and Biirger’s (1897) statements concerning the fate
of the median cord in Chalicodoma are contained in the following
paragraphs (pp. 371-372): “My investigations essentially con-
firm the account given by Heymons (1895). Nevertheless
I have not been able to determine, that the floor of the neural
groove becomes split up into a dermatogenic and neurogenic layer.
According to my observations all of its cell material goes to form
the median cord, while its covering is produced by the union of
the hypodermis formed in the region of the primitive swellings. —

“The complete sundering of the median cord from the hypo-
dermis takes place about the end of development. Its intragan-
glionic portions separate from the hypodermis at about the same

THE EMBRYOLOGY OF THE HONEY BEE = 1q1

time as the delamination of the lateral cords, its interganglionic
portions remain in connection with it, after the nerve fibres have
become evident in the lateral cords.”

Escherisch (1902, 1902a) has given a very complete and in-
teresting account of the development of the median cord in
Jfusca, in which the conditions recall those found in Lepisma. In
Musca a continuous median cord is separated from the ectoderm.
Within the limits of the ganglia the median cord contributes their
median portions, as in other insects; in the interganglionic re-
gions it presents swellings of considerable size, one being situated
directly caudad of each ganglion. From each of these swellings,
near its posterior end, a pair of lateral processes are given off
which extend to the neighborhood of the stigmata. These pro-
cesses are regarded as nerves. Escherisch points out the close
resemblance which the median cord bears to the unpaired median
nerve described by several investigators of the anatomy of insects.
This discovery, as well as that of Heymons (1897), shows that in
some insects the median cord may form a more or less continuous
median nerve. The presence of such a nerve in Lepisma suggests
that its occurrence represents a primitive condition.

Hirschler (1909) in the chrysomelid beetle Donacia, has in part
confirmed Graber’s statements regarding Melolontha, since he
finds that the median cord is at first completely severed, through-
out its extent, from the hypodermis, but that afterward the inter-
segmental sections are added to the ganglia, forming in each the
posterior median gangliomere. In their fate these interganglionic
sections correspond to the median cord neuroblasts and their
products in the Dermaptera and Orthoptera.

In the development of the lateral cords it is evident that the
honey bee conforms to the general rule and that such differences
as it presents are of relatively minor importance. In the origin
and fate of the median cord it conforms to Grassi’s account and
is also similar to many of the Coleoptera, as, for example, Hydro-
philus (Graber 1890), in so far as the median cord is not de-
veloped in the interganglionic regions, but remains here united
with the hypodermis. These interganglionic spaces nevertheless
are, up to the time of hatching, of very slight extent in an antero-
posterior direction, as figure 50E shows; moreover the anterior

142 THE EMBRYOLOGY OF THE HONEY BEE

and posterior commissures are close together, so that it is obvious
that the anterior and posterior median gangliomeres, as well as
the central gangliomere, are formed from the median cord.

B. The Brain

The rudiments of the brain, the procephalic lobes (see p. 100),
appear at Stages VI-VII. These together form a heart-shaped
expansion of the anterior end of the germ band which at its
widest part embraces a trifle over two-thirds of the diameter of
the egg. The antero-lateral margin of each lobe is rounded, and
the two lobes are separated from one another at the anterior end
of the germ band, by a median notch or indentation. Caudad the
lobes narrow rather gradually to join the remainder of the germ
band. The anterior limit of the procephalic lobes is at this stage
slightly ventrad of the cephalic pole of the egg.

The structure of the procephalic lobes at Stage VII is shown
in figures 29, 30, and 31. As these show, the lobes (ProL) are
composed of a single thick layer of long prismatic cells which
rises slightly above the level of the surrounding blastoderm,
(Fig. VII). On comparing the ectoderm of the procephalic
lobes with that of the neurogenic area of the trunk it,is at once
evident that the ectoderm of the lobes is of much greater thick-
ness, particularly in their anterior half (Figs. 30, 31). The
nuclei of the cells composing the procephalic lobes are similar
to those of the neurogenic area of the trunk ectoderm both in size
and clear appearance.

In cross sections through the posterior region of the proto-
cerebral lobes, caudad of the point where the anterior mesenteron
rudiment comes to the external surface, there is seen on each
side a group of cells (Fig. 31, Hyp) whose nuclei have the small
size and dense appearance characteristic of the dermatogenic ecto-
derm bordering the neurogenic area of the trunk. By following
this group caudad it is found to be actually continuous with this
portion of the ectoderm, which may therefore be conceived as
sending a tongue-like prolongation forward on each side into the
neurogenic portion of the protocerebral lobes. The fate of this
group of cells is uncertain, but its position suggests that it may
represent the antennal rudiment, which at the following stage

THE EMBRYOLOGY OF THE HONEY BEE = 143

(VIII) is actually composed of cells of just this character (Fig.
52, Ant). If these groups of small cells represent the antennal
rudiments, then the antennae originate caudad of the future
mouth, since its location on the germ band is marked by the point

Fic. 52. Transverse section through the head of an embryo, Stage IX,
showing the three principal divisions of the brain (1Br, 2Br, 3Br). and the
antennal rudiment (Ant), x 387.

where the cells of the anterior mesenteron rudiment come to the
surface. |

‘At Stage VIII the procephalic lobes have undergone considera-
ble change both in position and form. The entire germ band has
increased in length, thereby bringing its anterior end around the
cephalic pole of the egg, so that the procephalic lobes are now
curved in a semicircle, their anterior ends lying on the dorsal
side of the egg and directed toward the caudal pole. The
procephalic lobes have meanwhile become subdivided into
three lobes, the three pairs constituting respectively the proto-
cerebrum, deutocerebrum and tritocerebrum, each of which rises

144 THE EMBRYOLOGY OF THE HONEY BEE

above the external surface of the egg as a more or less evident
swelling. Of these the protocerebral lobes (Figs. VIII, VIIa,
rBr) are the largest. They are relatively flat with a rounded
external margin and have previously been described as resembling
a saddle in shape. The deutocerebral lobes (2Br) are more con-
vex, do not extend so far laterad, and are each tipped by a small
papilliform projection, the antennal rudiment (Ant). The trito-
cerebral lobes (3Br) are small hemispherical elevations situated
on a line with the posterior border of the stomodael depression
and somewhat farther apart than are the neural ridges to which
they are joined. They are in fact nothing more or less than the
much discussed “second antennae” of Biitschli and the “transitory
anterior appendage” of Grassi (see p. 109). A section transverse
to the long axis of the egg and passing through these three pairs
of lobes are shown in figure 52.

On each side are seen the papilliform antennae (Ant), imme-
diately dorsad of which are the deutocerebral lobes (2Br). Above
these and divided from them by a slight depression are the anter-
ior ends of the protocerebral lobes (zBr) which extend almost
to the mid-dorsal line, being here joined together by a bridge of
small dermatoblastic cells (Hyp). Ventrad of the antennae are
the tritocerebral lobes (3?Br), which are clearly seen to be essen-
tially similar to the remainder of the ganglionic swellings of the
neural ridges. Between these lobes, on the ventral side, is a pair
of high rounded elevations separated by a rather shallow median
depression. These elevations are merely the extreme anterior
ends of the neural ridges, which, at this and the stage next suc-
ceeding terminate anteriorly in a pair of rounded swellings situa-
ted just behind the mouth, and which are very evident in figures
IX and X.

During Stage X the external elevations marking the two anter-
ior divisions of the brain become less evident, the outer surface
of the head becoming relatively smooth in contour. The swellings
forming the tritocerebrum tend to lose their prominence and at
Stage X (Fig. X) have almost disappeared. This flattening out
of the external evidences of the rudiments of the neuromeres of
the brain takes place at the same stage and corresponds with the
flattening out of the neural ridges in the trunk, the rudiments of

THE EMBRYOLOGY OF THE HONEY BEE = 145

all of the neuromeres now rising above the internal instead of the
external surface of the ectoderm.

The lateral aspect of the head at Stage X shows little evidence
of the two anterior brain neuromeres. Its surface appears hemi-
spherical, smooth and unbroken except for the button-like anten-
nae. The dorsal aspect of the head however enables one to gain
a conception of the general form of the protocerebral lobes at this
stage. This is illustrated by Fig. 53, taken from a camera draw-

Lm

Fic. 53. Dorsal view of embryo, Stage XII, showing outlines of proto-
‘cerebrum (r1Br), x 112.
ing. The protocerebral lobes are now seen to be oval in outline,
their smaller ends directed backward and outward, their long axis
diverging at any angle of about fifty degrees,

The prismatic cells composing the ectodermal thickening which
constitutes the future brain, at Stage VIII are precisely similar
to those of the corresponding neurogenic area of the trunk. Dur-
ing Stage IX and X the brain rudiment undergoes a differentiation
corresponding with that of the ventral cord. Near the inner sur-
face of the ectoderm of all three segments of the brain appear the
clear and rounded nuclei of neuroblasts (Fig. 54, A and B, NO/).
The future history of these cells appears to be essentially similar
to those of the ventral cord, although their study in the brain is
much more difficult than in the ventral cord because of the way in
which the brain is bent about the anterior end of the egg. Never-
theless, in favorable sections the same phenomena may be observed.
Figure 54A shows the unequal teloblastic division of a neuroblast
in the rudiment of the deutocerebral segment, and also two ad-

146 THE EMBRYOLOGY OF THE HONEY BEE

jacent neuroblasts, one of which is seen accompanied by two of
its progeny. Figure 54B taken from the tritocerebrum shows three
parallel rows formed by the neuroblasts and their descendants.

Fic. 54. Parts of sections through the brain of embryos, to show the
neuroblasts (Nbl) and their products. A is from a transverse section of
Stage IX, passing through the antennal rudiment (Ant), and shows
the rudiment of the deutocerebrum (2Br). B is from a sagittal section
of Stage IX, passing through the tritocerebrum. ‘C. includes a portion
of the dorsolateral surface of the head of an embryo of Stage X-XI, and
shows small spindles of two different sizes, x 600.

THE EMBRYOLOGY OF THE HONEY BEE = 147

Figure 54C taken from the protocerebrum shows three spindles,
the two larger being plainly those of neuroblasts, while the smaller
is that of a daughter cell of the first generation.

A minor point in which the histogenesis of the brain differs
from that of the ventral cord is the greater irregularity of the
groups formed by the neuroblasts and their progeny. In the
ventral cord these form more or less regular rows, lying for the
most part in the transverse plane of the embryo, while in the
brain the different neuroblasts and their progeny often appear
to form a confused mass. This is especially true of the earlier
stages, and is at least in part to be ascribed to the varying planes,
with regard to the morphological long axis of the embryo, in which
the brain is intersected. In the later stages more or less regular
pyramidal groups with a neuroblast at their outer and larger ends
are frequently seen. Some of these are shown in figure 56.

Like the neuroblasts of the ventral cord, those of the brain
persist until after the hatching of the young larva. Probably they
remain active much longer than this. Sections through the brain
of a larva about two days old show cells with large clear nuclei
situated about the periphery of the brain and having all the ap-
pearance of the embryonic neuroblasts, except that they are larger.
Two of these larval cells were observed undergoing an unequal
and tangential division precisely like that of the embryonic
neuroblasts.

At Stage X, when the brain has nearly reached its ‘ultimate
embryonic dimensions, it begins to separate from the hypodermis.
This takes place in the same way as in the ventral cord, and the
hypodermis has for the most part the same character, being thin
and made up of flat cells. The deutocerebrum and tritocerebrum
separate from the hypodermis first. The protocerebrum separates
from the hypodermis more slowly, the separation beginning at
the mesial border of these lobes at Stage XI, and slowly progress-
ing laterad, being completed as Stage XII (Figs. 52 and 55).

The nerve fibres make their appearance at Stage XI (Fig. 55),
appearing simultaneously in the three divisions of the brain. Their
mode of development is precisely the same as in the ganglia of the
ventral chain, the formation of the nerve fibres being preceded
by the appearance of clefts beneath the innermost row of cells of

148 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 55. Transverse section through head of embryo, Stage XI-XII,
showing the formation of the suboesophageal commissure (SoeCom) and
the stomatogastric ganglion (StgGng), x 387.
the brain rudiment (Fig. 55). These clefts then widen out and.
become occupied by the fibres (Fig. 56).

While the greater portion of the brain is formed by neuroblasts,
parts of both the deuto- and protocerebrum are not formed in this
way. ‘These exceptions include the optic lobes, whose develop-
ment will be described later, and a pair of small groups of pris-
matic cells, situated, when first seen, one on each side, just above
the base of the antennal rudiment. These groups of cells becomé
plainly noticeable at Stage XI as a subspherical cluster of large
prismatic cells (Figs. 55, and 57, y). They apparently do not
divide, at least up to the time of the hatching of the larva, but
are covered over by the hypodermis derived from the base of the
antennal rudiment as shown in figure 57. At Stage XV they
are readily recognizable as a spherical group of relatively large
cells lying embedded in the deutocerebrum close to its outer sur-
face (Fig. 41, y).

The optic lobes of the insect brain constitute the first, and with

THE EMBRYOLOGY OF THE HONEY BEE 149

Fic. 56. Transverse section through head of an embryo, Stage XIII-
XIV. This section passes just caudad of the antennal rudiments, x 387.

reference to the rest of the brain, the outermost of the three
divisions of the protocerebral lobes (see pp. 116-117). Since in the
honey bee these lobes are, with reference to the poles of the egg,
directed caudad and laterad, the optic lobes may be similarly
described as derived from the lower and posterior border of the
protocerebral lobes, although this is actually their upper (dorsal)
and anterior border. As in other insects then development differs
from that of the remainder of the brain.

At Stages VIII and IX the ectoderm destined to form the optic
lobes is not distinguishable from the other neurogenic ectoderm
of the brain rudiment. During the succeeding stages the greater
part of the latter becomes transformed into neuroblasts and gang-
lion cells and the regular palisade-like arrangement of its cells
altogether disappears. This does not occur in the region of the

150 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 57. Part of a transverse section of Stage XI-XII, passing through
the antennal rudiment (Ant), showing one of the pair of spherical cell

groups of unknown significance (Y), which enter the deutocerebrum
(2Br), x 600.

future optic lobes. Here the regular palisade-like arrangement
characteristic of the neurogenic epithelium at earlier stages is
preserved, the rudiment of the optic lobes appearing as a thick
epithelial plate composed of slender prismatic cells with large
nuclei (Fig. 58A, OpL). Only a limited number of neuroblasts
are produced from this area. These, together with the progeny
of adjacent neuroblasts form a mass of cells underlying all but
the extreme lateral borders of the epithelial plate. On its ex-
ternal rounded margin the epithelial plate is sharply demarcated
from the dematogenic ectoderm immediately adjacent, which
equals the epithelial plate in thickness and is composed of narrow
prismatic cells of the smaller size characteristic of the dermato-
genic ectoderm (Fig. 58A, OpP1l). This dermatogenic layer be-
comes thin at its edges where it meets the optic lobes, the thin
edge overlapping and covering their margins.

The epithelial plate destined to form the optic lobes is at Stages

a,
io Piece
D

+

i ~ Ree eae
ee ee ee
et ee SS

THE EMBRYOLOGY OF THE HONEY BEE J itsr

XI and XII plainly seen to be marked off into two nearly equal
parts by a narrow furrow which runs parallel to the outer margin
of the protocerebral lobes. A similar furrow or depression on the
opposite side of the epithelial plate combines with the. first to
form a constriction which separates the plate into two nearly equal
parts, one of which is central, lying next to the second division
of the protocerebrum, while the other is marginal or peripheral
(Fig. 58A). This half projects laterad nearly free from the
mass of cells constituting the adjacent parts of the protocerebrum.

During Stages XII-XIII the superficial furrow deepens into a
cleft, the two halves of the epithelial plate simultaneously bending
inward in such a way as to form an invagination of which the
cleft is the lumen (Fig. 58B). The cells at the bottom of this
invagination are considerably deformed, the lateral two-thirds of
the epithelial plate being, so to speak, doubled up into a form
resembling that of the letter U (Fig. 58C). The peripheral half
of the epithelial plate together with the adjacent dematogenic ecto-
derm, forms a double fold constituting the lateral boundary of
the invagination. During Stage XIII, the two layers of this
double fold separate from one another, the outer dermatogenic
layer gliding over the optic lobe to join the thin hypodermis
formed from the adjacent parts of the brain by delamination
(Fig. 58C). This thicker hypodermis overlying the optic lobes
now constitutes the optic plate, from which the receptive portion
of the compound eye is derived (Fig. 58A, B, and C, OpP/). The
optic lobe of each side now appears in either transverse or coronal
sections as an oval mass composed of a double layer of large
columnar cells, and retains this appearance at least well into the
larval period.

A conception of the superficial extent and direction of the in-
vagination concerned in the formation of the optic lobes may be
obtained from certain favorable tangential sections of the head.
Such a section is represented in figure 58D. The double fold
of neurogenic ectoderm forming the optic lobe (OpL) is at Stage
XV seem to extend across the head at approximately right angles
to the long axis of the protocerebral lobes. These two folds are
continuous with one another at both ends of the slit-like lumen
of the invagination, so that each optic lobe may be compared to

Fic. 58. Sections through the otic lobe of embryos. A and B, trans+
verse sections intersecting the left optic lobe of two embryos of Stage
XII; C, of Stage XIII-XIV. D, tangential section through one of the
optic lobes of a newly hatched embryo (Stage XV), x 290.

THE EMBRYOLOGY OF THE HONEY BEE 153

an ordinary cup-like invagination greatly drawn out and flattened.
At stages prior to hatching the long axis of the invagination is
oblique to the long axis of the embryo, so that both coronal and
transverse (Fig. 58A-C) sections intersect the long axis of the
invagination. At hatching the dorsal flexure of the embryo is
exchanged for a ventral flexure, and the head becomes turned
ventrad, so that the long axis of the invagination is brought to a
position approximately at right angles to the long axis of the
young larva (Fig. 39). |

The accounts of the formation of the optic ganglion and optic
plates of insects differ to a considerable extent. In Mantis (Vial-
lanes 1891) the optic ganglion is produced by “Cellules ganglio-
genes” or neuroblasts like the remainder of the brain, the over-

lying hypodermis splitting off from the optic ganglion to

constitute the optic plate. In Xiphidiwm (Wheeler 1893) the
optic ganglion and optic plate are formed by a simple separation
of two cell layers, as is Mantis, but the inner layer is single and
contains no neuroblasts. In Forficula (Heymons 1895) the pro-
cess is much the same, but Heymons reports having observed a
few neuroblasts in the layer forming the optic ganglion. In none
of these forms is there any invagination of the ectoderm forming
the optic plate. There is present however, an ectodermal invagin-
ation or proliferation just posterior to the embryonic optic lobe
(Fig. 38, Jgl), but this is outside of the limits of the latter and
situated between it and the second protocerebral lobe, and is
moreover purely hypodermal in character, so that—as will appear
later—there seems to be no good ground for comparing this in-
vagination, the intergangionic thickening of Wheeler (1893), with
the invaginations of the optic lobe described by Patten (1889),
as Viallanes (1891) and Wheeler (1893) have done. In the re-
maining accounts, which relate exclusively to the orders Coleop-
tera and Hymenoptera a conspicuous ectodermal invagination is:
concerned in the development of the optic lobe. Patten (1887)
was the first to observe this in the case of the wasp Vespa. Here
the rudiments of the eye are formed somewhat as in the honey
bee, the optic plate lying at first laterad of the rudiment of the
optic plate, and later extending mesiad over it. The separation
of the optic plate from the optic lobe however takes place early,

154 THE EMBRYOLOGY OF THE HONEY BEE

and the ectoderm forming the optic lobe is not folded to any
extent, being merely bent inward to form a cup-shaped cavity.
At the beginning of the process, the optic plate is connected with
the optic lobe by a strand of cells which is said later to constitute
the optic nerve. Nothing of this kind was seen in the honey bee.

In Patten’s next paper (1889), on the eyes of the beetle Acilius
a long curved slit-like invagination, or rather three such invagina-
tions, almost continuous, were described as concerned in the
formation of the optic lobe. One of these invaginations was
considered as rudimentary. The other two seem together to
correspond very closely to the invaginations forming the optic
lobes of the bee in their form, situation and relation to the sur-
rounding parts. As in the honey bee, the ectoderm destined to
form the optic plate lies at first external to the future optic lobe.

Heider (1889) described a similar condition in Hydrophilus.
Heider’s text figure 5 representing a transverse section shows
an infolding strikingly similar to 58B. All the relations are
apparently the same as in the honey bee. Moreover the ectoderm
from which the optic lobe is formed appears to also consist of a
single layer of columnar cells.

All the cases thus far studied fall broadly into two classes:
those in which the optic lobe is formed by delamination, and those
in which it is formed by infolding. To the first class belong
representatives of the order Orthoptera and the nearly related
order Dermaptera. To the second class belong representatives of
the Coleoptera and Hymenoptera. In this class the optic plate is
formed from ectoderm lying outside of and immediately sur-
rounding that destined to form the optic lobes, in other words
the optic lobe and optic plate are formed from separate areas
of the ectoderm, while in the first class they are formed from the
same area. In the first class (Orthoptera, Dermaptera) the optic
lobes undergo rather a complex series of changes in order to
arrive at their ultimate or imaginal form. The postembryonic
changes undergone by the optic lobes in insects belonging to the
second class are unknown so that a basis for a definite comparison
is lacking, and only a guess is possible as to the meaning of the
different parts. A comparison of the optic lobes of the bee at
Stage XV with those of Mantis or Xiphidium at a stage shortly

THE EMBRYOLOGY OF THE HONEY BEE 155

prior to hatching suggests that the folding of the optic lobe in the
bee—and in the Coleoptera—may correspond to the subdivision
of the optic lobe in the forms first mentioned into the ganglionic
layer and the external medullary mass. This is however only
a surmise.

Both the supra- and suboesophageal commissures have been
compared by writers on the insect brain to the commissures of
the ventral cord, and are supposed to be formed in the same way,
that is, by a median ingrowth of the ectoderm comparable to the
median cord. Heymons has observed this in both commissures
in Forficula. In the honey bee the cellular portion of the supra-
oesophageal commissure is formed from a median thickening of
the ectoderm of the head. Figure 59B, taken from a transverse

SoeCom

Fic. 59. A,median sagittal section through the mouth of an embryo,
Stage X, showing the suboesophageal commissure (SoeCom). B, part of a
transverse section of an embryo, Stage XIII-XIV, showing the formation
of the supraoesophageal commissure, x 600.

section of Stage XIII-XIV sufficiently illustrates this point. As
in the commissures of the ventral cord it is the cellular portion
principally which is furnished by the median ingrowth, the bridge
of nerve fibres arising principally from the ganglion cells of the
paired ganglia lying on each side of the mid-line.

The origin of the suboesophageal commissure, or even a part
of it, from the ectoderm of the mid-line, is much less clear. As
figure 41, Stage XV, shows, the commissure (SoeCom) consists
principally of an extremely thin band of nerve fibres connecting

156 THE EMBRYOLOGY OF THE HONEY BEE

together the two halves of the tritocerebrum, and is accompanied
by a few small cells, apparently ganglion cells. This band of
fibres is formed at Stage X-XI by processes of the ganglion cells
of the tritocerebral lobes (Fig. 55, SoeCom). These fibres lie in
a minute channel at the bottom of the reentrant angle formed by
the junction of the hinder wall of the oesophagus and the ventral
ectoderm and lie immediately above the extreme anterior end of
the median cord (Fig. 59A, SoeCom). During Stage XI after
the suboesophageal ganglion separates from the ventral hypoder-
mis the suboesophageal commissure also leaves the ventral body
wall, although remaining in contact with that of the oesophagus.
During the stages succeeding, the hypodermis directly caudad of
the mouth grows out to form the lower lip or hypopharynx, in-
creasing the space below the oesophagus and consequently the
distance of the circumoesophageal commissure from the ventral
surface. Up to Stage X the nerve fibres of the suboesophageal
commissure and the anterior end of the median cord, now an
integral part of the suboesophageal ganglion, remain in close ap-
position (Fig. 59A). At Stage XITI-XIV (Fig 45) these become
separated by a slight cleft traversed by a few muscle fibres (dila-
tors of the pharynx) which pass from the median part of the
tentorium to the posterior or lower wall of the pharynx.

In view of the close apposition of the fibres of the circumoeso-
phageal commissure and the anterior end of the median cord, and
of the serial homology of the commissures of the ventral cord
with those of the brain, it is not impossible that the cells accom-
panying the fibres of the suboesophageal commissure are derived
from the ventral cord. This point could not however be definitely
determined, in spite of repeated efforts. While it may be true
that these cells are actually derived from the median cord, they
have rather the appearance of having migrated mesiad from the
two lobes of the tritocerebrum.

Summary

In summing up the foregoing observations on the development
of the brain of the honey bee, with reference to their bearing on
the development of the brain of other insects, it may be said,
first, that in a broad sense they confirm the results of Viallanes,

THE EMBRYOLOGY OF THE HONEY BEE 157

Wheeler, Heymons and others, especially with regard to the
division of the insect brain into three segments or neuromeres.
Second, with regard to details, the brain of the bee shows several
points of difference when compared with the few accounts avail-
able. These accounts relate principally to the development of the
brain in Orthoptera, and the nearly allied Dermaptera, as exem-
plified by Forficula. In fact since the publication, in 1895, of
Heymons’ monograph, there has appeared no detailed account of
the embryonic development of the insect brain. Even Carriére
and Burger (1897), in their otherwise quite complete account of
the development of the mason bee, give a disappointingly brief
account of the brain, concluding with the statement that “An
extended account of the development of the brain may be omitted,
since it would only serve to confirm the more recent investigations
on this subject.”

The points in regard to which the development of the brain of
the bee differs from the Orthoptera and Dermaptera may be
summed up as follows: 3

(1) The brain is flexed around the cephalic pole of the egg in
such a way that the morphological anterior ends of the proto-
cerebral lobes are directed toward the caudad pole of the egg.

(2) The cells of the second generation from the neuroblasts,
instead of the first, form the definitive nerve cells. This is the
same as in the ventral cord.

(3) The three subdivisions of the protocerebrum are not at
first plainly marked off from one another and are never separated
by hypodermal ingrowths.

(4) The optic lobes are formed, apparently independent of the
agency of neuroblasts, by a deep invagination of the neurogenic
ectoderm, which has no counterpart in the Dermaptera or Orthop-
tera, but seems to correspond more or less closely with an invagin-.
ation of the optic lobe described by Patten (1889) in Acilius and
Heider (1889) in Hydrophilus. Patten (1887) has also described
an invagination concerned in the formation of the optic lobe in
Vespa which seems to be similar to that in the honey bee.

C. The Stomatogastric System

The stomatogastric system, as in other insects is formed from
the dorsal wall of the stomadaeal invagination. The rudiments

58 THE EMBRYOLOGY OF THE HONEY BEE

of this system first make their appearance at Stage IX, but are
more sharply marked at Stage X. At this stage three evaginations
of the dorsal wall of the stomodeaum wall are visible (Fig. 60).

Meso f

= ‘
i eee
——

Ze os :
EGS :
aw: a

Fic. 60. Median sagittal section through the stomodaeum of an embryo,
Stage X, showing the three evaginations of the dorsal wall of the stomo-
daeum which form the stomatogastric system, x 387.

On the lower surface of this wall these appear as sharply pointed
clefts, on the upper surface as rounded swellings (Fig. 55, FtGng).
The first (anteriormost) of these is relatively insignificant, while
the other two are very noticeable. The first is situated not far
from the end of the labrum, the second close behind the first,
while the third is halfway between the second and the point where
the oesophagus joins the mid-intestine. The first evagination lib-
erates a few scattered ganglion cells which later are found ac-
companying the labral nerve, and then quickly disappears. The
second forms a large ovoid mass of nerve cells lying just above
the oesophagus and between the two halves of the brain, and is
easily recognized as the frontal ganglion (Fig. 55, FtGng). This
mass is at first hollow, with the cavity of the evagination extend-
ing up through it, but soon becomes solid. The third mass of

oo Se ee

THE EMBRYOLOGY OF THE HONEY BEE 159

nerve tissue thus formed is lower and longer in an antero-poster-
ior direction than the rudiment of the frontal ganglion, and is in
contact with the latter. This third mass forms the pharyngeal
ganglia of the stomatogastric system and possibly a part of the
frontal nerve connecting these ganglia with the frontal ganglion.

The frontal ganglion is known to arise from a median evagina-
tion of the dorsal wall of the stomodaeum in Orthoptera (Vial-
lanes 1891, Wheeler 1893, Heymons), Forficula (Heymons 1895),
Coleoptera (Heider 1899), Hymenoptera (Carriére 1890) so that
it is safe to assume that this mode of origin is typical for the
insects. The accounts of Heymons and of Carriere and Burger
of the origin of the frontal ganglia and its associated structures
are the most complete and circumstantial. In Forficula and sev-
eral of the Orthoptera the stomatogastric system consists of the
frontal ganglion, close behind which is the elongate occipital
ganglion connected with the frontal ganglion by the short nervus
recurrens (stomatogastric nerve). From the occipital ganglia two
nerves run dorsad and caudad for a short distance to the paired
ganglia pharangea (pharyngeal ganglia), and these in turn send
out nerves which run to the posterior termination of the
oesophagus.

The stomatogastric system in Chalicodoma (Carriére and Bur-
ger (1897) is essentially similar to that of the honey bee, consist-
ing simply of a frontal ganglion, a recurrent (stomatogastric
nerve) and a pair of pharyngeal ganglia.

On comparing the stomatogastric system of these two repre-
sentatives of the Hymenoptera with that of Forficula and Gryllus
it becomes evident that the principal difference between them
lies in the apparent absence of the occipital ganglion in the repre-
sentatives of the Hymenoptera. In both the mason bee and the
honey bee the stomatogastric nerve is surrounded by ganglion
cells for some distance caudad of the frontal ganglia, and this
part of the nerve may therefore readily be homologized with the
occipital ganglion, as Carriére and Biirger have already stated.

The mode of origin of the stomatogastric system in Forficula
and Gryllus differs slightly from that in the honey bee. Three
evaginations are found on the dorsal wall of the stomodaeum.
From the first of these is formed the frontal and occipital ganglia,

160 THE EMBRYOLOGY OF THE HONEY BEE

from the second the pharyngeal ganglion, and from the third the
nerves passing caudad from the pharyngeal ganglia. The first
evagination in Forficula therefore corresponds apparently to the
second in the honey bee, the second to the third, while the last
is wanting.

D. Neurilemma

The neurilemma, which at Stage XV constitutes a thin cellular
membrane covering the external surface of the nervous system
(Fig. 37A, B and C, Nim), first becomes evident in the ventral
cord at Stage XI-XII. At this stage, it will be remembered that
a considerable shifting and rearrangement of the cells of the
lateral cord takes place as shown in figure 50A and B. A large
number of the daughter cells of the neuroblasts have, in the intra-
ganglionic regions, shifted laterad, while others appear on the
ventral surface of the lateral cords where they form a single
layer. A similar layer is already present at the dorsal surface.
From these superficial cells the neurilemma is evidently derived.
In figures 50A and B, on the ventral side of the lateral cords,
the neurilemma may be seen to be in process of formation, certain
of the superficial cells having already assumed a flattened form
(Nim). The development of the neurilemma on the dorsal side
of the ventral cord is less readily observed, owing to the number
of small cells crowded together in this region, but it seems reason-
able to assume that its origin is the same here as on the ventral
surface. The cells of the neurilemma at later stages (XI-XV)
vary much in size, indicating division subsequent to their assump-
tion of a superficial position.

In regard to the origin of the neurilemma in insects there has
been a considerable difference in opinion. Nusbaum (1883,
1886) and Korotneff (1885) traced its origin to wandering blood
cells. Wheeler (1893) believed that it arose from the intragan-
glionic sections of the median cord. Heymons (1895, p. 45) states
that in Forficula “The outer neurilemma apparently arises from
cells which during the segregation of the neuroblasts from the
dermatogenous layer were separated off from the latter.”

In the honey bee it seems highly improbable that the neuri-
lemma arises from blood cells, since the cells which are destined to
form the neurilemma are when first evident—at Stages XI-XIIJ—

TEE Sy a ee

ab

a er ge eS

ee ee

a a eo

ee
ae

THE EMBRYOLOGY OF THE HONEY BEE 161

so closely associated with the ganglion cells of the ventral cord
and so like them in appearance. It is still less probable that they
owe their origin to the median cord since at the time they are
first evident the median cord is just separating from the hypo-
dermis, and moreover its cells are still relatively large and pre-
serve their original elongate form. ‘There accordingly remain
two possibilities; either the neurilemma cells are, as Heymons
suggests, split off from the dermatogenous layer; or else they are
merely transformed ganglion cells. The former view has much
to commend it, from a theoretical standpoint and, as far as the

‘writer’s observations go, can not be totally excluded. The evi-

dence at hand however seems to favor the latter view. Prior to
Stages XI-XII the line of separation between the lateral cords
and the future hypodermis is coincident with the outer boundary
of the neuroblasts, in other words there is no direct evidence that
any considerable number of dermatogenous cells are split off
from the peripheral layer; on the other hand at the time the
lateral cords are definitely split off a rearrangement of the cells
takes place in the intraganglionic regions of the lateral cords, at
which time a number of small cells appear on their ventral sur-
faces. These cells closely resemble the ganglion cells. It seems
probable therefore that in the honey bee the neurilemma owes its
origin to the ganglion cells themselves.

E. Corpora Allata

The origin and development of the corpora or ganglia allata in
the honey bee corresponds very closely to the account given by
Heymons (1895) for Forficula. In both insects these bodies arise
as ectodermal ingrowths located in front of the bases of the first
maxillae. In the honey bee this is also the location of the tubular
invaginations which form the apodemes, of the adductor muscles
of the mandibles. Each of these ingrowths at Stage X has already
become a long hollow finger-like structure curving dorsad and
caudad. The mouth of this invagination is wide, invading the
base of the maxilla on its anterior and lateral sides. At Stage
X there may be found near the outer and caudal end of each of
these openings a small tubular invagination extending mesiad
into the base of the maxilla. These invaginations are the rudi-

162 THE EMBRYOLOGY OF THE HONEY BEE

4

CorAll”

C

Fic. 61. Transverse sections through the head of three embryos of dif-
ferent stages, to illustrate the development of the corpora allata, (CorAll).
A, Stage X; B, Stage XI; 'C, Stage XITI-XIV, x 387.

THE EMBRYOLOGY OF THE HONEY BEE _ 163

ments of the corpora allata (Fig. 61A, CorAll). At the next
stage (XI) the lumen of these invaginations has nearly disap-
peared and the corpora allata now appear as irregular cellular
outgrowths springing from the mesial side of the mandibular
apodemes (Fig. 61B). These cell masses are next constricted
off from the apodemes, but still remain in close contact with them.
Meanwhile the corpora allata are carried dorsad by the growth
of the mandibular apodemes and at Stage XII lie between the
latter and the inner ends of the posterior arms of the tentorium
where these join the anterior arms to form the central body. At
Stage XIII the corpora allata become attached to the ventro-
lateral angles of the antennal somites (Fig. 61C) and soon after
lose their attachment to the mandibular apodemes. This is their
position at the time of hatching and for at least a considerable
period of the life of the larva. The corpora allata acquire their
characteristic globular and compact form during the final stages
of embryonic development.

Heymons (1895) was the first to call attention to the bodies
named by him “ganglia allata,” describing their development in
some detail in Forficula, and more briefly in Gryllus. Three years
later Carriére and Birger (1897) found that in the mason bee
(Chalicodoma) the ganglia allata have the same origin as in
Forficula, but these bodies do not, however, later fuse to form a
median body as they do in Forficula but remain attached to the
ventro-lateral margins of the antennal coelomic sacs, as in the
honey bee. These investigators expressed a doubt as to the ner-
vous nature of the “ganglia allata,”’ since no nerve fibres could
be observed within them. In 1899 Heymons published a brief
paper on the structure and development of the corpora allata
of a walking stick (Bacillus). In this form these bodies lie one
behind the other, above the oesophagus and caudad of the pharyn-
geal ganglia, from which they receive nerves. In this insect the
corpora allata have the form of capsules within each of which
is a concentric layer of chitin secreted by the capsular wall. Their
development was found to be the same as in Forficula.

The function and homology of these bodies is unknown. Hey-
mons (1899) surmised that they had a static function, but exci-
sion of the corpora failed to produce any disturbance in the loco-
motion of the insect (Bacillus). Janet (1899a, 1900) discusses at

1644 THE EMBRYOLOGY OF THE HONEY BEE .

some length the probable homology of the “corpora incerta” and
arrives at the conclusion that these bodies probably represent the
tracheal rudiments of one of the cephalic segments and that they
furnish material for the tracheae of the head. No evidence to
support this view has been seen in the honey bee. On the con-
trary the corpora allata and the tracheae develop independently
of one another. In the honey bee the tracheae supplying the head
are derived from tracheal invaginations located on the second
maxillary segment (see p. 172).

A possible clue to the homology of the corpora allata appears in
an interesting observation, apparently hitherto unnoticed, by
Toyama (1900) on the embryo of the silkworm (Bombyx). In
the section devoted to the endoskeleton of the head Toyama states :
“In the mandibular segment three pairs of invaginations take
place; the most anterior (between the labrum and the mandible)
becomes the first tentorium, the second pair gives rise to the seat
of the extensor mandibulae, while the last becomes the flexor
mandibulae and salivary gland.’’* The flexor mandibulae and
salivary gland therefore arise from a common invagination, which
Toyama’s figure (woodcut Fig..1) shows to be situated at the
posterior margin of the base of the mandible, corresponding quite
closely to the point of origin of the flexor mandibulae and corpora
allata in the honey bee. This suggests that the corpora allata may
represent glands, vestigial in the honey bee and many other insects,
but functional in at least some of the Lepidoptera. Further
research in this direction is much needed.

F. Degenerating Cells

A phenomenon apparently peculiar to the bee, since it is not
mentioned by the investigators of other insects, is the frequent
occurrence of degenerating cells within the embryonic nerve tis-
sue. These cells occur isolated and in small number within the
ventral cord but in the brain they are abundant and to a certain
extent localized in definite regions. The largest, most conspicu-_
ous and most constant of these include a pair of wedge-shaped
sections of the brain, one on each side, situated near the juncture
between the proto- and deutocerebrum, and including a part of

Poe DOF.

THE EMBRYOLOGY OF THE HONEY BEE 165

“each (Fig. 55, DegCl). Degenerating cells become evident in

these regions as early as Stage IX, and are easily recognizable
at low magnifications by the presence among the brain cells of
deeply stained granules. Under a high magnification these gran-
ules prove to be the remains of the chromatic contents of the
nuclei of degenerating cells (Fig. 62). Such nuclei are spherical

Fic. 62. Part of the section represented by fig. 56, showing degenerat-
ing cells in the protocerebrum, x 600.

in form and smaller than the adjacent nuclei, the nuclear mem-
brane often faint. The chromatin generally appears to be con-

- densed into one or more relatively large granules, and large and

deeply-stained spherical nucleoli are also commonly present.
Janet (1907) has observed very similar nuclei in the degenerating
wing muscles of an ant (Lasius). The cytoplasm of the degen-
erating cells becomes broken up into a number of minute spherules,
producing an appearance suggestive of an emulsion (Fig. 62).
This process of cell degeneration continues up to the time of
hatching, and probably even later, but the number of degenerating
cells present in the brain reaches its maximum at Stages XIII
and XIV (Figs. 56 and 62). At this time not only do the regions
referred to appear to be crammed with degenerating cells, but a
large number of the superficial cells of the protocerebral lobes,
as far back as the optic lobes, are also in a state of degeneration.

;

166 THE EMBRYOLOGY OF THE HONEY BEE

A considerable number of the degenerating cells are not re-
tained within the brain tissue; as early as Stage X the debris
of degenerating cells, consisting of the shrunken nuclei and the
spherules just described, may be seen on the dorso-lateral surface
of the brain, and this debris remains here beneath the amnion
apparently unaltered up to the time of hatching, as figure 62
shows. Degenerating cells continue to be extruded from the brain
up to Stage XV and possibly even later. At Stages XIII and
XIV, when the maximum number of degenerating cells are found
in the brain, the characteristic debris of these cells is visible within
the space between the hypodermis and the brain in almost every
section, and is especially abundant over the dorso-lateral region
between the proto- and deutocerebrum, where degenerating cells
are most abundant (Fig. 62).

At Stage XV, that is, just subsequent to the hatching of the
larva, the number of degenerating cells has greatly diminished,
these having either been expelled or absorbed by the brain. A
few, however, are still visible. The debris lying on the outside
of the brain, within the head capsule, has also disappeared, having
probably been washed away by the blood current.

The significance of this extensive cell degeneration is not appar-
ent. That these cells are not artifacts is shown by the uniformity
with which they occur, being present in all series of sections
examined of embryos which were older than Stage IX. Their
presence may accordingly safely be considered as normal. Vial-
lanes (1891), Wheeler (1893) and Heymons (1895) state that
the neuroblasts, or at least the major part of them, degenerate at
the close of embryonic development—a kind of senile degeneration
of the cell. This, as already stated, is not the case in the honey
bee. Moreover the degenerating cells begin to appear at a period
when the neuroblasts are just beginning to be differentiated. The
possibility that some of the neuroblasts degenerate is however by
no means excluded. At Stages XIII and XIV, many of the
superficial cells on the dorsal side of the protocerebrum are in a
state of degeneration, but these are in most cases at least, peri-
pheread of the neuroblasts. This seems to indicate that these
degenerating cells are derivatives of the neuroblasts, which have
wandered peripherad of the latter, as do the neurilemma cells
(p. 160), but their precise origin is unknown.

— e

ee ae i
a Ss os =

_

IX

TRACHEAL SYSTEM, ENDOSKELETON AND HypopERMIS
1. Tracheal System

The essential features of both the structure and development
of the tracheal system have been correctly described by both
Butschli (1870) and Grassi (1884). The tracheal system of the
larva consists of a pair of longitudinal trunks (Figs. 63, 64 and
75, TraTr, Fig. XV, LTraT), each of which traverses virtually
the entire length of the trunk close beneath the lateral hypoder-
mis. These trunks are connected to the ten spiracles (Sp) of
each side by slender short branches. At their anterior and pos-
terior ends, respectively, the two lateral trunks are united by
semicircular loops or commissures, one of which is situated ceph-
alad of the mid-intestine, above the oesophagus (Figs. 63 and
64, ATraL, Fig. XV), the other caudad of the mid-intestine,

just ventrad of its juncture with the hind-intestine (Fig. 63

PTraL, Fig. XV). The tracheal trunks are also united between
these points by eleven other commissures, segmentally arranged.
These commissures arise from the ventral side of the lateral
trunks, nearly opposite the point of origin of the branches to
the stigmata, thence they pass downward, close to the hypoder-
mis and beneath the ventral cord (Figs. 63 and 64, TraCom,
Fig. XV). The first and second of these loops join the main
trunk close to the anterior boundary of the second and third
trunk segments, and intersect the ventral mid-line between the
first and second trunk segments. The third and fourth loops
are united at their bases, joining the main trunk in the anterior
half of the fourth trunk segment. These loops intersect the
ventral mid-line near the posterior margin of the third and
fourth trunk segments. The remaining seven loops are con-
tained entirely within the limits of the corresponding segments.
Just in front of each of the spiracles the tracheal trunk sends
off one or more branches which pass dorsad, breaking up into
finer branches supplying the dorsal region of the trunk. Small

167

168 THE EMBRYOLOGY OF THE HONEY BEE

aN S SH \ iS T om £
BBS Foil Sn

Fenton

i

TI

/N\\\
au

mts
43

s
2
>

SR

Fic. 63. Side view of larva about two and one-half days old, treated
with caustic potash, showing tracheal system.

branches are also sent off by the ventral commissures. At the
anterior end of each of the lateral trunks a large branch arises
which passes cephalad and breaks up into two smaller branches,
both of which run cephalad, one in contact with the inner sur-
face of the optic lobe, the other parallel to the mandibular apo-
deme and close to it. Mesiad of this large branch is a smaller
branch which is sent off by the anterior tracheal commissure.
This also extends cephalad supplying the dorsal part of the head.
A third branch arises from the ventral side of each of the lateral
trunks just cephalad of the first pair of stigmata and dividing,
supplies both the base of the labrum and the ventral portion of
the first trunk (thoracic) segment (Fig. 64).

The tracheae are simple in structure, being merely tubes com-
posed of a single layer of epithelial cells continuous with those
of the hypodermis at the spiracles and lined by chitin thickened
in the form of fine transverse threads, more or less spirally ar-
ranged, the taenidia (Fig. 65, Tae).

THE EMBRYOLOGY OF THE HONEY BEE _ 169

TraTr.

Fic. 64. Side view of anterior end of larva about two and one-half days
old, treated with caustic potash, showing the tracheae and the endoskelleton
of the head.

Fic. 65. Section through a spiracle (Sp), spiracular branch (SpBr)
and tracheal trunk (TraTr), of a newly hatched larva (Stage XV),
x 600

The spiracles are twenty in number, ten on each side, ranged
wy in a row along the lateral surface of the trunk, and correspond to
a the second to the eleventh trunk segments, inclusive. All of the
spiracles are situated in the anterior half of the segment to
which they belong, the first three pairs lying close to the an-
terior margin of the corresponding segments. The form and
Structure of a spiracle, together with the branch connecting it

-

170 THE EMBRYOLOGY OF THE HONEY BEE

with the main trunk is illustrated by figure 65, taken from a
larva just hatched. The spiracle is seen to be minute, in fact
scarcely perceptible. In older larvae it is relatively much larger.
This opening leads into a spherical chamber, as shown in the
illustration, and this in turn into a contracted passageway, the
spiracular branch (SpPBr), which widens out just before reaching
the main trunk (TraTr). The taenidia (Tae) are evident in
the main trunk and at the inner end of the connecting branch,
but are replaced by a smooth chitinous cuticle for the remainder
of the course. This chitin becomes thickened as it approaches
the spherical chamber and is continuous with the cuticle of the
body at the external opening. Surrounding the cavity leading
inward from the spiracle is a single layer of long prismatic cells,
continuous with those of the adjacent hypodermis, and forming
a conical mass which may be considered as the stigma, and
which passes imperceptibly into the cellular portion of the con-
necting branch, the cells becoming progressively shorter up to
the junction with the main trunk.

The rudiments of the tracheal system make their appearance
at Stage VIII (Figs. VIII, VIIla) as two rows of pit-like in-
vaginations of the ectoderm, each row being situated about half
way between the lateral edges of the germ band and the ventral
mid-line. There are on the trunk ten pairs of these invaginations,
corresponding to the second to the eleventh trunk segments. The
two or three most anterior pairs appear first, the remainder in
rapid succession. The mouths of these invaginations are at first
irregular, the first pair being slightly larger than the others. At
Stage IX (Fig. IX) these have acquired the form of transverse
slits, the first pair however being turned obliquely to the long
axis of the embryo. Shortly after this—at Stage X (Fig. X)—
they have contracted to minute circular apertures which persist
as the spiracles. The invaginations themselves, at first shallow
depressions of the ectoderm, increase rapidly in extent and at
Stage IX assume the form of flat sacs wedged in between the
mesoderm and ectoderm (Fig. 67, Trlmv). At Stage X, each ~
of these sacs, with the exception of those of the second and
eleventh trunk segments, sends out four outgrowths or diverti-
cula, one of which is anterior, one posterior, one dorsal, and one

THE EMBRYOLOGY OF THE HONEY BEE 171

ventral. The anterior and posterior outgrowths of the sacs of
the same side meet and fuse with one another to constitute the
longitudinal trunk (Fig. 66, TraTr), while the ventral outgrowth
{TraCom) of each sac continues to extend ventrad until it meets
with the corresponding one of the opposite side. These then
fuse to form the transverse commissures (compare Figs. XI and
XII). The dorsal outgrowths continue to extend in the same
direction, at the same time breaking up into ramifying twigs, and
form the dorsal branches (Fig. 66). The anterior outgrowths

Fic. 66. Sagittal section through the rudiments of the tracheal system
of one side in trunk segments 2-4, of an embryo Stage X-XI. Developing
oenocytes (Oen) also shown, x 387.
of the first pair of tracheal sacs take a course cephalad and
dorsad over the mid-intestine and meet and fuse at the dorsal
mid-line to form the anterior tracheal loop. The posterior out-

172 THE EMBRYOLOGY OF THE HONEY BEE

growths of the last pair of tracheal sacs similarly unite and fuse
with each other beneath the hind-intestine to form the posterior
tracheal loop.

The outgrowths forming the various parts of the tracheal
system are at first stout and thick-walled, but as they lengthen
they decrease rapidly in diameter until they attain their final
form and distribution.

The chitinous lining of the tracheae is formed between Stages
XIII and XIV.

In addition to the ten pairs of tracheal invaginations found in
the trunk, a pair of tracheal invaginations occur also on the first
maxillary segment. These make their appearance at the same
time—Stage VIII—as the other tracheal invaginations, and are
situated on the lateral surface of the anterior half of the segment,
above the base of the rudiment of the second maxilla of each side,
and only a few sections caudad of the boundary between the first
and second maxillary segments. The location on the segment of
this pair of tracheal invaginations is therefore the same as that of
the remaining ten. Like the other tracheal invaginations, those
of the second maxillary segment are at first shallow and cup-
shaped depressions. They soon lose this form and, developing
with surprising rapidity, become sacs with a narrow mouth,
directed obliquely caudad. The bottom of the sac quickly spreads
between the ectoderm and yolk (Fig. 67, TrInv) and at the same
time sends off four branches or diverticula. One of these is
directed caudad, one dorsad and the other two cephalad. Of
these last two branches one passes above the mouth of the invagi-
nation and one below it. All of these changes take place between
Stages VIII and IX. Even before the close of Stage IX the
posterior branch of each of these tracheal sacs may be traced
back to a juncture with the anterior prolongation of the tracheal
sac of the corresponding side of the second trunk (thoracic) seg-
ment. The anterior ends of the tracheal trunks therefore owe
their origin to the tracheal invaginations of the second maxillary
segment. The mouths of the invaginations have by this time con-
tracted to narrow ducts while over them (Fig. 67) on each side is
seen a well marked fold of ectoderm obviously caused by the
formation of the tracheal sac, which raises the overlying ectoderm,
on account of the resistance offered by the yolk. This fold is

THE EMBRYOLOGY OF THE HONEY BEE = 173

; ie ‘s cooteee-eMint

ge,
-
-
-

Fic. 67. Transverse section through the anterior half of the second
maxillary segment of an embryo, Stage X, showing one of the pair of
tracheal invaginations (Tr/nv) and the mesoderm (Meso) belonging to
this segment, x 387.

clearly shown in figure Xa, but not labeled. At the beginning of
Stage X the opening of the tracheal sac to the exterior has already
become contracted, and by the end of this stage or the beginning
of the next (XI) is completely closed, leaving behind no trace of
its existence. At the same time the ectodermal folds also disap-
pear, their disappearance being in part at least due to the shrink-
age and withdrawal of the yolk from this region.

The further history of the three remaining diverticula of each
of the tracheal sacs remains to be considered. All of these elon-
gate rapidly, assuming a tubular form. The dorsal diverticulum,
which is the largest of the three (Fig. 68, ATraL), continues to
extend dorsad, taking a slightly cephalad course, skirting the
posterior margin of the cerebral lobes, along the line of junction
of the head and trunk. At Stage XIII the distal ends of these
branches reach the dorsal mid-line, when they unite to form the
anterior tracheal loop or commissure (Figs. 63 and 64, ATraL).
Of the two remaining diverticula the ventralmost forms the larger

174 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 68. Anterior end of a sagittal section laterad of the median plane,
from an embryo of Stage XI-XII, showing the development of the deriva-
tives of the tracheal invaginations of the second maxillary segment.
Hypodermis and brain represented in outline only. Drawn from two
sections, X 243.
of the two branches previously described (p. 168), supplying the
brain and first maxillae, the other forms the smaller branch
supplying the dorsal part of the head (Fig. 64).

Hatschek (1877) described what he thought were tracheal
invaginations on the gnathal segments of embryos of Bombyx, but
examinations of Hatschek’s plates show that these invaginations
are merely those of the tentorium and mandibular apodemes, and
this is the interpretation of subsequent embryologists. With this
exception there is no record in the literature of the insect embry-
ology of tracheal rudiments existing in the gnathal region. It
may therefore be inferred either that this phenomenon is peculiar
to the honey bee, or else that it is of more or less general occur-
rence, but up to the present time has been overlooked. Of these
two inferences the latter seems the more probable. If it were not
so, it would be difficult to understand why the honey bee, a
specialized member of a group generally regarded as highly

‘

ee

Sa

Se et! Nee I ge

}
‘vil
A a
Ms
hi

a

<a S

THE EMBRYOLOGY OF THE HONEY BEE 17s

specialized, alone displays a developmental feature highly gen-
eralized in its nature, since it can hardly be regarded as a new
formation. Moreover the rapidity of the development of these
tracheal sacs and the brief duration of their external openings
make it still more probable that they have escaped the notice of
the investigators of the embryology of other insects. Finally, it
is a trait of human nature to see only what is looked for, in proof
of which it may be said that the discovery of the tracheal invagi-
nations of the second maxillary segment in the honey bee was
scarcely more than a happy accident.

The presence of a pair of tracheal invaginations in the second
maxillary segment assists materially to attain a correct concep-
tion of the significance of the other invaginations of the cephalic
ectoderm. Palmen (1877), Hatschek (1877), Wheeler (1889)
and particularly Carriére and Biirger (1897) contended for the
homology of the silk glands, the tentorial invaginations and the
mandibular apodemes with tracheae. According to Carriére and
Birger the first pair of tentorial invaginations were to be con-
sidered as the first pair of stigmata, the mandibular apodemes the
second, the second pair of tentorial invaginations the third, etc.
Now that a pair of tracheal sacs is known to exist in the second
maxillary segment, the homology of the second pair of tentorial
invaginations with the stigmata of the second maxillary segment
is completely excluded, and the homology of similar invagina-
tions with those of the tracheae is made decidely problematical,
especially when one considers that the tracheal invaginations of
the second maxillary segment are situated some distance dorsad
to the bases of the appendages of that segment. The basis for
the conjecture of Janet (1899a, 1900) that the corpora allata
furnish material for the tracheae of the head, is likewise removed.

Aside from the occurrence of a tracheal invagination in the
gnathal region, the development of the * hea! sv:zem of the
honey bee possesses no points of specisl “1.teresi, ..ut conforms
to the type found in the majority of insects, and tl: efore requires
no further comment.

2. Tentorium and Mandibular Apodemes.

The larval tentorium (Fig. 69B) consists of a narrow chitinous
bar or central body, compressed in a dorso-ventral direction,

176 THE EMBRYOLOGY OF THE HONEY BEE

placed transverse to the long axis of the body and situated
between the oesophagus and the suboesophageal ganglion (Fig.

Fic. 69. A, side view of the head of an embryo, Stage XI, drawn in
outline, showing two of the four invaginations which form the tentorium
(1Ten, 2Ten), and also the one of two invaginations which form the
mandibular apodemes (RAp), x 243. B, ventral view of the head of a
larva about three days old, treated with caustic potash, showing the form
and relations of the tentorium (1Ten, 2Ten), x 200.

THE EMBRYOLOGY OF THE HONEY BEE 177

45 Ten). From the ends of this bar spring two pairs of arms,
one pair anterior (17en) and the other posterior (2Ten).
These arms join the chitinous cuticle of the head. The tentorium
thus serves as a set of braces to support and strengthen the
cranial capsule as well as to serve for the attachment of muscles.
The posterior arms (27en) together with the central body
(Ten) form a flat arch, its convex side directed cephalad. The
outer ends of these arms are attached to the cranial capsule low
down on its lateral aspect, near its junction with the trunk. The
anterior arms, which are much more slender than the posterior,
arise also from the ends of the central body and are here about
equidistant both from one another and from the lateral wall of
the head. Each takes a course dorsad, and only slightly laterad,
to a point on the anterior cranial wall just above the base of each
mandible. About midway between the central body and the anter-
ior cranial wall the anterior arms give off a short pointed spur,
directed laterad and cephalad (Fig. 69B). In all of the more
advanced larvae studied the anterior arms are much longer than
the posterior, but in recently hatched larvae (Fig. XV) the arms
are nearly equal in length. As is evident in figure XV the anter-
ior arms together with the central body form a U-shaped struc-
ture which embraces the oesophagus and the circumoesophageal
commissure.

The tentorium serves for the attachment of two pairs of large
muscles, besides a group of small muscles near the mid-line.
Both pairs of large muscles are inserted on the wall of the head.
The first of these is attached to the tentorium at the ends of the
central body, and from here each member of the pair runs ven-
trad, ectad and caudad, to an attachment of the hypodermis at
the base of the first maxilla. The muscles of the second pair are
attached to the anterior arms of the tentorium, just cephalad of
their junction with the central body, and run obliquely dorsad,
laterad and cephalad to a broad area of attachment on the cranial
wall, laterad of the cephalic lobes. It is presumably the tendon
of this muscle which forms the spur on the anterior arms (Fig.
69B). Muscle fibres also pass from the anterior margin of the
central body to the posterior (ventral) wall of the pharynx and
doubtless serve as dilators of the latter.

178 THE EMBRYOLOGY OF THE HONEY BEE

In insect anatomy the tentorium, as well as other portions of
the endoskeleton, is commonly regarded as a structure which is
wholly chitinous. In sections through the head of a larval bee
however the tentorium appears as a system of tubes, composed of
a single layer of epithelial cells, which are continuous with those
of the hypodermis at the four points where the tentorium joins
the head capsule (Figs. 41 and 42,Ten, Fig. 62, 2Ten). The
chitinous lining of these tubes constitutes the tentorium proper.
This lining is a mere pellicle in recently hatched larvae (Stage
XV), but somewhat thicker in older larvae, but never much
thicker than the larval cuticle with which it is continuous, and
consequently it never attains any considerable degree of rigidity.

The mandibular apodeme, the second member of the endo-
skeleton of the head, is a slender spine, arising near the base of
each mandible at its mesial margin and directed cephalad and
dorsad (Figs. 63 and 64, RAp). In recently hatched larvae (Fig.
XV) the mandibular apodeme arises somewhat cauded of the
base of the mandible. Like the tentorium, the mandibular apo-
deme is hollow and consists of an epithelial tube lined with
chitin (Fig. 41, RAp). The flexor muscle of the mandible, of
which the mandibular apodeme is only the tendon, is a large fan-
shaped muscle having a broad origin on the dorso-lateral wall
of the cranial capsule, near its junction with the neck and just
behind the optic lobes (Fig. 42, RMcl). This muscle is more
evident in recently hatched than in older larvae.

The origin of the tentorium and the mandibular apodeme is
simple. At Stage XI, or a little earlier, three pairs of rounded
or oval ectodermal invaginations may be observed at the bases
of the mouth parts (Fig. 69A). The first pair (rTen) is sit-
uated just in front of the bases of the mandibles close to the
antennae, the second (Fig. 61A and B, Fig. 69, RAp) between
the bases of the mandibles and first maxillae. The third (Fig.
69, 2Ten) between the bases of the first and second mazxillae.
These invaginations deepen and grow inward as slender tubes,
blind at their inner ends (Fig. 61C, RAp). The first pair be-
comes directed caudad, the second caudad and dorsad, the third
cephalad and mesiad. At Stage XII the first and second. pair
meet and coalesce to form the tranverse bar or central body

— a ee

eee

ae eS

EEE

a ee ee

|
;
)
3
i

THE EMBRYOLOGY OF THE HONEY BEE = 179

of the tentorium. The second pair constitute the mandibular
apodeme.

Heider (1889), Heymons (1895), Carriere and Biirger (1897),
Toyama (1900) and Riley (1904) have given complete accounts
of the development of the endoskeleton of the head, and with
these the account just given is in substantial agreement, and may ©
possibly be considered therefore as representative of the typical
mode of development of this structure.

The tentorium of the honey bee is unmentioned in all previous

" accounts except that of Grassi (1884). In this are described and

figured the two pairs of invaginations which ultimately unite to
form the tentorium, but in later stages Grassi lost sight of the
anterior pair and consequently failed to grasp their significance,
supposing that only the posterior pair were concerned in the
formation of the tentorium.

3. Hypodermis

The hypodermis is formed of the ectoderm which remains after
the subtraction of the various tissues and organs of ectodermal
origin. At Stages VII and VIII, the dorsal half of the blastoderm
is stripped from the yolk to form the amnion. This portion of
the yolk is therefore left entirely bare. A little later—at Stage
IX—a delicate cellular membrane is found covering this previ-
ously nude area, except in the cephalic region, which now is
covered by the cephalic end of the germ band. This membrane
is as thin and delicate as the amnion itself, and is continuous later-
ally with the ectoderm, which therefore is presumably responsible
for its production. This thin dorsal sheet, for the sake of conven-
ience in description, may be termed the extra-embryonic ectoderm,
in contradistinction to the embryonic ectoderm, which overlies the
mesoderm, and from which are derived the oenocytes, the tracheal
system and the ventral nerve cord. The subtraction of this
material from the embryonic ectoderm, which takes place during |
Stages IX-XI Figs. 78 and 79), materially reduces the thickness
of the embryonic ectoderm, particularly in its median part, from
which the ventral cord is derived. At Stage XI (Fig. 80) the
entire embryo is seen to have begun to increase in breadth, this

180. THE EMBRYOLOGY OF THE HONEY BEE

increase naturally affecting the embryonic ectoderm, which always
remains slightly wider than the underlying mesodermal tissues.
During the two succeeding stages, XII and XIII (Figs. 81 and
82), the embryo continues to increase in breadth, its lateral limits
being marked by the lateral—now dorsal—margins of the em-
bryonic ectoderm which is progressing upward over the sides
of the egg toward the dorsal mid-line. This growth of the em-
bryonic ectoderm, as the figures indicate, takes place in part at
the expense of the extra-embryonic ectoderm, which, with the
increasing breadth of the embryonic ectoderm, undergoes a cor-
responding decrease. Meanwhile the embryonic ectoderm is
steadily diminishing in thickness, and at its margin it passes
rather gradually into the thinner extra-embryonic part. Between
Stages XIII and XIV, the lateral margins of the embryo meet
and unite along the dorsal mid-line, the extra-embryonic ectoderm
being now entirely absorbed by the embryonic ectoderm, now the
definitive hypodermis (Fig. 82).

The head capsule is formed as follows: During the interval
between Stages VII and VIII, when the cephalic end of the germ
band, consisting principally of the protocerebral lobes, travels
around the cephalic end of the egg, the interval between the
lateral margins of the procephalic lobes and of the adjoining
margins of the gnathal segments remains continuously filled by
ectoderm. The interval between the procephalic lobes is, at
Stage VII, filled by a thick layer of ectoderm (Fig. 29, Ect), and
after the procephalic lobes reach the dorsal side of the egg,
this part of the ectoderm constitutes the dorsal part of the head
capsule, comprising the triangular area between the cerebral
lobes. The remainder of the head capsule is formed from the
superficial layer of that part of the cephalic ectoderm concerned
in the formation of the brain.

ee es
“

+ oe

~ oS

*“< Fr

Xx

THE OENOCYTES

The name oenocytes was given by Wielowiejsky (1886) to
certain large wine-colored cells observed embedded in the fat body
of the larvae of Corethra, and which he also found in other
insects belonging to several different orders.

This investigator was, however, not the first to note these cells,
for as Graber (1891) has stated, similar cells had been previously
observed and described by other investigators, as for example,
by Tichomiroff (1882), in the embryos of the silk worm. After
Wielowiejsky called attention to the oenocytes they were observed
in the larvae and imagoes of insects belonging to so many different
orders that they are now generally assumed to be common to all
pterygote insects. They differ much in different orders; the prin-
cipal features which they have in common appear to be their
relatively large size and their ectodermal origin. To this it may
be added that after leaving the ectoderm they have never been
seen to divide. They are always found more or less closely
associated with the cells of the fat body, and often attached to
tracheae. They may be scattered or arranged in clusters, in which
case a pair of clusters is usually situated in each of the first eight
abdominal segments, as in Corethra. Little is known of the func-
tions of the oenocytes, although several attempts have been made
to solve this problem, the latest being that of Glaser (1912), who
has studied the oenocytes of the larvae of the leopard moth and
finds evidence that they contain oxydising enzymes.

The development of the oenocytes has been observed by several
investigators. Graber (1891) calls attention to the fact, that prior
to the appearance of Wielowiejsky’s paper, Tichomiroff (1882)
and Korotneff (1885) had described the development of cells to
be regarded as identical with oenocytes, and that later Heider
(1889) had also observed an ectodermal proliferation which was
probably concerned in the development of oenocytes, but without
recognizing its significance. Graber also described the develop-

181

oS ARs 8 EMBRYOLOGY OF THE HONEY BEE

ment of the oenocytes of the grasshopper Stenobothrus. Here
the oenocytes are derived by delamination from the inner surface
of the lateral ectoderm. In the beetles Hydrophilus, Melolontha
and Lina Graber finds presumptive evidence that the oenocytes
are derived from the ectoderm near the eight abdominal stigmata,
including special “metastigmatic” invaginations, one of which is
situated directly caudad to each of the eight abdominal stigmata.*®
In Hydrophilus and Melolontha these invaginations are situated
near the posterior border of the segments some distance behind
the spiracles, in Lina they are close behind them.

Wheeler (1892) has reviewed the work of previous investiga-
tors and described briefly his observations on the development
of the oenocytes in the representatives of the Orthoptera, Ephe-
meridea, Hemiptera, Neuroptera and Lepidoptera. In all cases
they were found to arise by either delamination or migration from
the lateral ectoderm, and the metastigmatic invaginations of
Graber were found not only in representaives of the Coleoptera
but also in representatives of the Orthoptera (Blatta, Xiphidium),
but Wheeler believes that they play only a minor part in the pro-
duction of the oenocytes.

Heymons (1895) studied the development of oenocytes in
Forficula and Gryllus. He states “that by means of simple
migration or through the intervention of Graber’s metastigmatic
invaginations they find their way into the interior of the.
body and thence into the already formed fat body. Here for a
long time they form sharply marked, metamerically arranged
groups of cells. Later they become more irregularly distributed
throughout the abdomen, and also find their way into the thorax,
but are nevertheless to be here easily distinguished from the true
fat cells principally by their size.’ So far this corresponds
closely with Wheeler’s account (1892) of Blatta and Xiphidium.
Heymons was, however, able to add the important fact that
oenocytes are developed from all of the eleven abdominal seg-
ments, and not merely from those bearing spiracles, as had been
previously assumed to be the case. Moreover these “postspiracu-
lar’ oenocytes differ from the remainder by their larger size.

Wheeler (1892) states that Patten (1889) had previously observed
these in Acilius.

THE EMBRYOLOGY OF THE HONEY BEE 183

Hirschler (1909) finds that in Donacia the oenocytes arise in
two ways. A part of them arise by migration from restricted
areas of eight of the abdominal segments, immediately caudad
of the stigmata, thus forming metamerically arranged groups.
Another part is formed by a sort of diffuse or irregular immigra-
tion from various parts of the ectoderm, but particularly from
the vicinity of the pericardial septum and the proctodaeum.

The oenocytes of the honey bee were first described by Wie-
lowiejsky (1886) and have been studied in detail by Koschev-
nikov’® (1900, 1905), and—in connection with the metamorphosis
—by Anglas (1900). These investigators agree that in the larval
honey bee the oenocytes are represented by “certain cells which
are of great size, not vacuolated, their protoplasm staining strongly
with borax carmine, and having large regular oval nuclei... .
They are to be found in the depths of the body surrounded on all
sides by the fat body and sometimes packed close against the
walls of the Malpighian tubules. I never observed two nuclei in
any of these cells. During histolysis, they float free in the body
cavity. The largest of them are truly gigantic, measuring 176
micra with a nucleus 56 micra in diameter.’*° One of these cells,
together with the surrounding fat cells, in a larva four days old,
is illustrated in figure 70A. In younger and smaller larvae the
oenocytes are smaller, but their size, as compared with that of
the fat cells, remains about the same. An oenocyte surrounded
by three fat cells, from a recently hatched larva, is represented in
figure 70B. It will be noted that in this case the magnification
is 1107 diameters, while in that of the older larva (Fig. 70A) the
magnification is only 580 diameters. A fair conception of the

* Glaser (1912) has expressed doubt in regard to identity of the cells
described by Koschevnikov with the oenocytes of other insects on the
ground of Koschevnikov’s statement that he has observed the oenocytes
in the act of engulfing fat cells. Without attempting to weigh the merits
of this objection, it may be remarked that the oenocytes of Koschevnikov,
in the larval stages at least, correspond morphologically with the oenocytes
of other insects more closely than any other cells within the body cavity,
and, if these are not truly oenocytes, it would be necessary to assume that
such cells are either non-existant in the bee, or else that they are morpho-
logically different from those of other insects. It is of interest to note in
this connection that Anglas (1900) states that he has never seen them act
as phagocytes.

® Koschevinkov (1909).

184 THE EMBRYOLOGY OF THE HONEY BEE

great increase in size of the oenocytes and fat cells between the
newly hatched larva and the larva four days old may be gained
by comparing figures 70A and 75, Oen, these two figures being
drawn at the same magnification.

As in older larvae the oenocytes at Stage XV are principally
distinguished by their size which, although somewhat variable, is
only equalled by the neuroblasts and some of the blood cells.
Their form is largely determined by their location. When free
they are more or less spherical as in figure 70B; when in close
contact with other cells, polyhedral, as in figure 70A. The nucleus
is usually more or less spherical and its chromatin content is in
young larvae not especially different from that of the cells of
other tissues. The cytoplasm presents a rather more homogen-
eous appearance than that of the fat cells. The oenocytes are
rather irregularly scattered throughout the body cavity and are
frequently enmeshed in the fat cells, as in figure 70B. At Stages
XIV and XV oenocytes are rare in the first three trunk seg-
ments, in other words the oenocytes during these stages are found
principally in the last twelve trunk segments, corresponding to
the imaginal abdominal segments of insects of other orders than
the Hymenoptera.

In surface views of many embryos of Stage X small clear spots,
like those of the rudiments of the spiracles, may be seen to alter-
nate with the last eight spiracles of each side, so that the latter
have the appearance of being double. Transverse sections show
that there are actually at this stage pit-like invaginations of the
ectoderm precisely in line with the spiracles and that each of
these invaginations is situated half way between the adjacent
spiracles.

Each pair of spiracles is, however, situated, not in the middle,
but in the anterior half of the corresponding segment, so that
each pair of the invaginations just described belongs to the pos-
terior half of the same segment as the pair of spiracles next in
front (cephalad of it). Accordingly these invaginations are
located on the third to the eleventh trunk segments, or, regarding
only the first three trunk segments as belonging to the thorax,
the first eight abdominal segments. These invaginations corre-
spond in form, size and location with the “metastigmatic” invagi-

THE EMBRYOLOGY OF THE HONEY BEE 185

Fic. 70. Oenocytes (Oen), surrounded by fat cells. A, from a larva
about four days old, x 387; B, from a newly hatched larva, x 1107.

nations of Melolontha as described by Graber (1891). Figure
71 shows half of a tranverse section through the region between
the fourth and fifth trunk segments and intersecting the first
metastigmatic invagination (Oen), showing its position and
relation to the surrounding tissues. In figure 72A is shown one of
these invaginations more highly magnified. There are in all eight
pairs of these, as in those insect embryos examined by Graber
(1891) and Wheeler (1892). The first pair is the largest, those
taudad successively smaller, the last being faint and not easy to
discern. Each invagination is funnel-shaped externally. Inter-
nally it is formed by a cluster or rosette of long pyriform ectoderm
cells as shown in the figure, the entire structure strongly suggest-

18 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 71. Right half of a transverse section through the posterior por-
tion of the fourth trunk segment of an embryo. Stage X, showing the
location and relative size of the cell-clusters which produce the oenocytes,
x 367.

ing a sharp contraction of the ectoderm at this point. Even at
this stage some of the cells of the clusters are seen to be detach-
ing themselves and migrating inward. These cells are in all
probability the embryonic oenocytes (Oen). At Stage XI the
external evidences of an invagination have almost disappeared,
only a minute depression marking the point where it had existed.
The oenocytes are now seen to be slightly larger than the cells
of the ectoderm and are moving inward into the body cavity. This
is fairly well shown by figure 72B. Certain other sections illus-
trate this particular point somewhat more clearly, but the section
also shows another important but also unfortunate circumstance,
that is, the similarity of the oenocytes at this stage to the adjacent
mesoderm cells, particularly those of the fat body (zF, 2F). This
similarity is so close that the continuity of the cells seen at Stage
X with the oenocytes as they appear in the late embryonic stages
(XIII, XIV), as well as of the larval stages could not be demon-
strated, in. spite of repeated attempts. Graber (1891) encoun-
tered the same difficulty in Hydrophilus. Nevertheless, in view
of the developmental history of the oenocytes in those forms in

Fic. 72. A, transverse section through the lateral ectoderm of the
fourth trunk segment of an embryo, Stage X, showing the formation of
oenocytes (Oen). B, a similar section from the third trunk segment of
r an embryo, Stage XI-XII. C, section through the lateral ectoderm of the
a fourth trunk segment of an embryo, Stage XIII, showing the point where
?, the oenocytes arose. D, the same, from the sixth trunk segment of an
‘a embryo, Stage XIII-XIV, x 600.

18 THE EMBRYOLOGY OF THE HONEY BEE

which it may be continuously followed, as in Stenobothrus or
Xiphidium, one can scarcely doubt that the metameric cell clus-
ters seen in the bee embryo also give rise to the oenocytes of
the larva.

At Stages XII-XIV the point of origin of oenocytes is marked
by a minute pad or tuft of cells which stain somewhat less deeply
than the other hypodermal cells (Figs. 72C and D). In some sec-
tions the cells of this tuft are so long as to suggest the possibility
of the separation of some of these from the hypodermis (Fig.
72C), but no direct evidence of this was obtained.

The definitive oenocytes of the larva become evident at Stage
XITI-XIV, owing to their large size as compared with the other
cells in the body cavity. From this time until hatching their size
remains practically constant and is the same as that of the cells of
the segmentally arranged clusters of Stages X and XI, which are
assumed to be the embryonic oenocytes, so that the identification
of the oenocytes at the later stages is accordingly made possible by
the differentiation of the other cells in the body cavity and the
accompanying reduction in their size, rather than by any morpho-
logical change in the oenocytes themselves.

XI

Muscies, Fat Bopy AND CIRCULATORY SYSTEM

In the young larva each muscle is composed of a varying num-
ber of parallel fibres, which constitute the contractile portion of
the muscle. These fibres range from round to flat in tranverse
section, stain deeply and are themselves composed of fine fibrillae
(Figs. 73A and B). Each fibre is continuous from its origin to
its insertion and remains of nearly the same diameter throughout
its length, tapering only slightly toward the two ends. In the
longitudinal trunk muscles, however, which are attached to the
intersegmental hypodermis, instead of terminating at these points,
the fibres appear to be continuous through at least two segments,
and probably more (Fig. 73A). A layer of undifferentiated
protoplasm, the sarcoplasm, enwraps all of the fibres which con-
stitute a single muscle. At the middle of a muscle the sarcoplasm
is common to all of the fibres of the muscle, binding them together,
but towards the two ends of the muscle the sarcoplasm divides, a
small portion, constantly diminishing, following, each fibre to its
attachment (Fig. 73A). Embedded in the sarcoplasm are the
muscle nuclei, arranged in rows along the fibres. Usually they
lie all on one side of the fibre—in the case of the longitudinal
trunk muscles the inner side—but this rule is subject to many
exceptions. In young larvae none of the muscle fibres are trans-
versely striated. In the imago, on the other hand, the greater part
of the muscles are striated, while the relation of the sarcoplasm
and the muscle fibres is the reverse of that existing in the larva,
the sarcoplasm and its nuclei in the imago constituting, so to
speak, the core or central portion of the muscle, about which are
arranged the muscle fibres. It is of interest to note that the con-
dition existing in the muscles of the bee larva is quite similar to
that found in the muscles of the vertebrates.

The muscles may for convenience be divided into three groups:
the muscles of the head, the muscles of the trunk, and the muscles
of the alimentary canal.

189

1909 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 73. A, muscle fibre, from a longitudinal section of the dorsal
longitudinal trunk muscles, between trunk segments 1 and 2. Hypodermis
shown in outline. B, transverse section through one of the ventral longi-
tudinal muscles, x 840. :

in)

7 =

THE EMBRYOLOGY OF THE HONEY BEE igi

The muscles included in the head comprise the gnathal muscles,
moving the mouth parts, the muscles attached to the tentorium
(already described under the section relating to that structure)
and certain muscles which although having their origin on the
endoskeleton of the head belong properly to the alimentary canal.

The muscles of the newly hatched larva are by no means so
easily identified as in older larvae, being less well differentiated.
Moreover in so small an object the study of sections is the only
method available. In older larvae both mandibles and maxillae
are provided with flexor and extensor muscles, those of the mandi-
bles being much the larger. In newly hatched larvae, however,
the only gnathal muscles clearly evident are the flexor. muscles
of the mandibles. These are relatively large muscles, each of
which is inserted on the mandibular apodeme, whence it passes
dorsad, caudad and ectad, spreading out fanwise meanwhile, to a
broad origin on the inner surface of the dorso-lateral face of the
head capsule behind the optic lobes and near the junction of the
head and trunk (Fig. 42, RMcl).

The principal muscles of the trunk comprise dorsal and ventral
longitudinal muscles and oblique muscles. The dorsal longitu-
dinal muscles (Fig. 75, DLMcl) are broad flat sheets compris-

ing one layer of fibres, and lie on either side of the heart,

beneath the dorso-lateral hypodermis extending far down on the
side. A pair of these are situated in every segment, their ends
being attached to the intersegmental hypodermis, as shown in
figure 73A.

The ventral longitudinal muscles (Fig. 75, VLMcl) correspond
in their arrangement and function to the dorsal longitudinal mus-
cles, but are only about half as wide. They extend laterally from
the ventral nerve cord about half way up to the level of the
longitudinal tracheal trunk.

The oblique muscles (Fig. 75, OMcl) comprise a pair of mus-
cles spanning the lateral wall of each segment and arranged in
such a way that if seen in side view they would together form a
figure resembling the letter X. The smaller of these lies close to
the hypodermis, and runs dorsad and cephalad, while the larger is
situated a short distance within and runs ventrad and cephalad.
In addition to the oblique muscles there are delicate muscles,

192 THE EMBRYOLOGY OF THE HONEY BEE

lying close to the inner surface of the hypodermis and consisting
of but three or four fibres, which take their origin on the hypo-
dermis a short distance below the stigmata, on the ventral side
of which they are inserted. These muscles are so insignificant
that they are easily overlooked except in favorable tangential
sections.

The muscles of the alimentary canal comprise a well defined
layer of minute circular fibres enwrapping the whole alimentary
tract, including the fore- and hind-intestines, in addition to which
is a layer of longitudinal fibres, much fewer in number and more
scattered than those of the circular fibres (Fig. 74 MclEnt).

Fic. 74. Part of a sagittal section through the dorsal wall of the mid-
intestine of a newly hatched larva (Stage XV), showing the muscular
coat (MclEnt), x 600.

These muscles, both longitudinal and circular, may conveniently
be termed the enteric muscles. They are best developed in the
fore-intestine, and here the longitudinal underlie the circular
fibres. In the case of the mid-intestine this relation, as in the
imago, is reversed, the longitudinal fibres lying outside the
circular. This is also true of the hind-intestine. The fibres mak-
ing up the enteric muscles are extremely delicate, and the muscle
nuclei minute. On the mid-intestine the circular muscle
fibres are fairly numerous and regularly arranged, the longitu-
dinal fibres, on the contrary, are scattered and of such extreme
delicacy that it is only in favorable cases that they can be seen
at all (Fig. 74). Muscle fibres also run from the anterior (dor-
sal) wall of the labrum to the anterior (dorsal) wall of the

f
}
ay
on
é
+a
iy
y
q
;
‘ti
"eg

Oe ee a a

— se

ie ie!

THE EMBRYOLOGY OF THE HONEY BEE 193

pharynx, while a few fibres run from the anterior margin of the
central body of the tentorium to the posterior wall of the pharynx;
both of these sets function as dilators of the pharynx.

The muscles associated with the circulatory system will be
described under that heading.

The fat body of the recently hatched larva has a decidedly dif-
ferent aspect from that in the old larva. In the latter the fat body
forms a compact mass filling the spaces of the body cavity and
composed of cells loaded with large fat globules, in figure 69A
represented by the empty spaces in the cells surrounding the
oenocyte. In the young larva, on the other hand, the fat body
never constitutes a compact mass, but is made up of a loose net-
work of branching cells (Fig. 69B) in only some of which a
minute fat globule can be made out. In the thorax the fat body is
somewhat more compact than in the abdomen, in which the fat
cells are much scattered, as shown in figure 75, 1F, 2F, 3F. As in
Chalicodoma (Carriére and Birger 1897) the fat body is more
or less incompletely divided into three sections: a dorsal section,
situated above the dorsal diaphragm, the pericardial fat body
(Fig. 75, 3f), and two sections situated laterad on the mid-
intestine (z7F, 2F) and divided from one another by the longi-
tudinal tracheal trunk.

The circulatory system consists of the heart, the dorsal and
ventral diaphragms, and the blood corpuscles.

The heart (Figs. XV, 75, 76, Ht) is a slender tube lying in the ~
mid-line between the mid-intestine and the dorsal hypodermis. It
extends from the twelfth trunk segment forward to the head
where it is continued as the aorta between the cerebral lobes,
lying immediately above the oesophagus. The anterior end of
the aorta opens into the body cavity. The posterior end is
closed. The lateral walls of the heart are relatively thick,
and contain the nuclei while the ventral and dorsal walls are thin,
the latter being in some sections so tenuous as to be scarcely
perceptible, the heart having the appearance of being closed dorsad
by the hypodermis (Fig. 75). Along the ventral side of the
heart is a narrow strip of cells, to which, in trunk segments seven,
eight and nine, the ovaries are attached. On its dorsal side the
heart is slightly indented by the intersegmental constrictions, while

194 THE EMBRYOLOGY OF THE HONEY BEE

DDph_s:

( ; ; : DE CEORRS Shenae
Bi ai Sree ee <a> en
i 5 {Oy

= S. a

3 s S25 D

Cy

VDph Gng — Hyp

Fic. 75. Transverse section through the 6th trunk segment of a newly
hatched larva, x 387.

in the middle of each segment its lateral walls are sharply infolded,
reducing the lumen at these points to a narrow vertical slit. These
folds evidently correspond to the ostia, or lateral valvular openings
(see Snodgrass, 1910, p. 107), although no opening in the lateral
walls was observed at Stage XV.

In tranverse sections an exceedingly delicate membrane is seen
to extend right and left from the ventral wall of the heart, the
dorsal diaphragm (Figs. 75 and 76, DDph). This ends free in the

THE EMBRYOLOGY OF THE HONEY BEE 195

body cavity, except in the intersegmental regions, where it extends
laterad to the hypodermis, to which it is attached (Fig. 76,

Fic. 76. Transverse section through the dorsal region of the trunk,
between the 5th and 6th trunk segments, showing the attachment of the
dorsal diaphragm (DDph) to the hypodermis, x 600.

DDph). This relation is clearest in the third to the ninth trunk
segments inclusive, but undoubtedly also exists in the first two
(thoracic), although masked here by the crowded condition of the
parts and the large number of pericardial fat cells. The muscle
fibres constituting the dorsal diaphragm of most insects, includ-
ing the honey bee (imago), are commonly described as radiating
fanwise from the points of origin to their attachment on the
heart (vide Snodgrass, 1910, p. 108), but in the young larvae no
muscle fibres could be seen in the dorsal diaphragm, which appears
in both longitudinal and tranverse sections as a continuous and
extremely thin membrane.

The cells constituting the dorsal diaphragm are of two kinds;
numerous flat cells with minute nuclei (Fig. 70), appearing to be
the diaphragm cells proper, and pale lenticular cells whose size
approximates that of the fat cells (Fig. 76 ParC). An examina-
tion of the text and figures of Carriére and Biirger’s (1897)
account of the dorsal diaphragm of Chalicodoma makes it evident
that the cells just described are identical with those constituting
the “paracardial cellular cord” (paracardiale Zellstrang). This
term was first used by Heymons (1895) to designate segmentally

196 THE EMBRYOLOGY OF THE HONEY BEE

arranged groups of cells lying laterad to the heart among the
fibres of the dorsal diaphragm of Forficula. Although at first
sight it may seem somewhat doubtful whether the scattered cells
of the dorsal diaphragm in the larvae of the honey bee are to be
regarded as the homologues of the segmentally arranged cells
-composing the “paracardiale Zellstrang” of Forficula and Chali-
codoma, nevertheless—as will appear later—they have precisely
the same origin in all three insects, so that their homology can
scarcely be open to question. In view of the scattered distribu-
tion of these cells in the honey bee it will be more convenient to
refer to them simply as the paracardial cells.

The aorta into which the anterior end of the heart opens, is
merely the space around the oesophagus bounded externally by
the inner walls of the coelomic sacs of the antennal segment (Fig.
42, Ao). It opens cephalad and ventrad into the cavity of the
head.

The ventral diaphragm is composed of a layer of transversely
arranged muscle fibres which combine to form a more or less con-
tinuous membrane (Fig. 75, VDph) overarching a space just
dorsad of the ventral cord, the ventral sinus. The ventral dia-
phragm extends from the first to the neighborhood of the twelfth
trunk segments, its lateral margins being attached to the mesial
margins of the ventral longitudinal muscles and also to the silk
glands. In the mid-line its dorsal surface is in contact with the
ventral wall of the mid-intestine.

The blood corpuscles are large rounded cells which at Stage
XV are for the most part confined to the lumen of the heart. In
size they approach the oenocytes, but their general appearance is
quite different. Both nucleus and cytoplasm are pale, the latter
being much vacuolated and frequently enclosing deeply stained
granules. A group of blood corpuscles are shown within the heart
in figure 75.

Since the organs and tissues which form the subject of this
section are all derived from the mesoderm, a consideration of the
development of this germ layer naturally comes next in order.
The development of the mesoderm in the head, including the
enathal segments, differs from that of the trunk, and since the
latter may be regarded as representing a less modified and there-
fore more typical condition, it will be considered first.

THE EMBRYOLOGY OF THE HONEY BEE 197

At Stage VII a tranverse section through the middle of the
trunk shows the mesoderm (Fig. 32, Meso) as a double layer of
cells lying close beneath the ectoderm, between the latter and the
yolk. This condition is typical of the mesoderm from the gnathal
region at least as far as the end of the eleventh trunk segment.
The inner of the two layers is the visceral, the outer the somatic
layer. In the neighborhood of the ventral mid-line the cells of the
mesoderm are rather irregular in form and arranged in but a
single layer. At the lateral margins of the mesoderm where the
two layers are continuous with one another, they are relatively
thick, well defined and separated from one another by a narrow
slit-like space, while the component cells of both layers are col-
umnar in form. A similar condition obtains in Chalicodoma, and
to these lateral regions, composed of columnar cells, Carriére and
Burger (1897) applied the name “mesodermal tubes” (Mesoderm-
rohre) as descriptive of their form and since these sections of the
mesoderm are marked off rather sharply from the remainder both
in their form and their behavior, this term will for convenience be
adopted in the following description. Between the mesodermal
tubes and the single-layered median strip the mesoderm cells are
flattened and somewhat irregular in form, particularly those of the
visceral layer. In Chalicodoma, Biirger (1897) found that the
mesodermal tubes were divided intersegmentally by thin parti-
tions, moreover, the mesoderm mesiad of the mesodermal tubes
is divided into pairs of flat sacs (mesodermal sacs), a pair to
each segment, and communicating only intrasegmentally with the
mesodermal tubes of the corresponding side. No well marked
evidence of segmentation of the mesoderm could be found in the
honey bee. The two layers of the mesoderm on each side of the
mid-line are virtually continuous from the second maxillary seg-
ment to the eleventh trunk segment, although a careful exami-
nation of sagittal sections of Stages VIII and IX seems to indicate
that the mesoderm mesiad of the mesodermal tubes is slightly
constricted intersegmentally. At Stage VIII-IX (Fig. 77, Meso)
while the mesoderm has changed but little in its general appear-
ance, it may be noted that the cells in the immediate vicinity of
the mid-line have increased in size, and assumed a rounded form.
These (BIC) are to form the blood corpuscles.

At Stage IX (Fig. 78) the mesoderm begins to show important

198 THE EMBRYOLOGY OF THE HONEY BEE

Mc
Fic. 77. Transverse section through the future basal region of the

abdomen of an embryo, Stage VIII-IX; showing the structure of the
mesoderm (Meso), x 387.

Fic. 78. Transverse section through the fourth trunk segment of an
embryo, Stage IX, intersecting one of the tracheal vaginations (TrInv),
showing the formation of blood cells (BIC), x 387.

THE EMBRYOLOGY OF THE HONEY BEE gg

changes in form. The cells of the mesodermal tubes have lost
their columnar shape and from the inner layer a ridge now pro-
jects abutting against the yolk, directed mesiad and ventrad, the
splanchnic layer or rudiment of the enteric muscles (MclEnt).
Near the mid-line the yolk has receded a trifle and the mesoderm
as such has disappeared, leaving a narrow space between the
ectoderm and yolk, the epineural sinus (Heymons 1895) which is
partially filled with large loose cells, the blood corpuscles (BIC).

At Stage X (Fig. 79) the rudiments of all the mesodermal

Fic. 79. Transverse section through the third trunk segment of an
embryo, Stage X, x 387.

organs and tissues may readily be identified. The splanchnopleure
or rudiment of the enteric muscles (MclEnt) has changed from a
mere ridge to a well defined layer pressed close against the yolk.
The cells of the remainder of the visceral layer have for the most
part lost thir epithelial character and show a looser arrangement,
in anticipation of their destiny, which is to form the two main
divisions of the fat body (rF, 2F). Moreover a break in the
visceral layer is seen opposite the rudiment of the tracheal trunk
(TraTr). At the mesial edge of the mesoderm, close to the
ganglion of this segment (Gug), two groups of extremely small
cells are seen; these are the rudiments of the ventral longitudinal

200 THE EMBRYOLOGY OF THE HONEY BEE

muscles (VLMcl). The somatic layer of the mesodermal tubes
now appears wedge-shaped in section, its broader end directed
mesiad and ventrad, and is moreover split longitudinally into an
inner thin single layer of cells, the rudiment of the dorsal dia-
phragm (DDph), and an outer portion composed of several
layers of cells, which are to form dorsal longitudinal muscles
(DLMcl). The yolk has continued to contract, drawing away
from the mesoderm over its entire extent so that the epineural
sinus is extended by the addition of two lateral cavities, which in
turn are in connection with the coelomic cavity by means of the
break in its visceral wall. These spaces together represent the
rudiment of the definitive body cavity.

The process of differentiation goes steadily forward during the

: “Gh SG
oe WS

Hyp
Fic. 80. Transverse section through the third trunk segment of an
embryo, Stage XI, x 58o.

THE EMBRYOLOGY OF THE HONEY BEE - 2or

- next stage (XI). The rudiment of the enteric muscles (Fig. 80.
MclEnt) increases in extent and decreases in thickness. The
rudiments of the first and second sections of the fat body become
more clearly defined, the first section almost losing its connection
with the splanchnopleure (Mc/Ent), while the second breaks up
into its constituent cells. The rudiment of the dorsal diaphragm
now finally parts company with the remainder of the somatic
layer of the mesodermal tubes, and together with the rudiment
of the enteric muscles, forms in section a figure comparable to an
inverted V. The apex of the V is formed by a group of small
cells, the cardioblasts (Cbl), whose fate is that of forming the
two halves of the heart. In the section figured the rudiments
of three sets of muscles are plainly evident by reason of their pale
appearance and small nuclei. The dorsalmost and largest of
these is that of the dorsal longitudinal muscles (DLMcl), below
this, lying between the tracheal trunk (TraTr) and the hypo-
dermis, is the rudiment of the oblique muscles (OMcl), while next
to this, ventrad of the silk gland (S/kG/) is the rudiment of the
ventral longitudinal muscles (VLMcl). Between this last and the
hypodermis, laterad of the ganglion of this segment is a small
group of pale cells, which constitute the mesodermal portion of the
leg rudiments (Meso3L). The yolk continues to contract and its
ventral half assumes a three-sided form, and the future body
cavity becomes greatly increased in extent. The rudiments of all
the mesodermal tissues and organs are now well differentiated
from one another and during this stage and the one following
(XII), they increase in size and together with the hypodermis
extend rapidly dorsad, this movement leading finally to the com-
plete enclosure of the yolk by the germ band at Stage XIV. The
most noticeable change at Stage XI is seen in the rudiment of
the enteric muscles (MclEnt). At the preceding stage this rudi-
ment covered only an insignificant part of the yolk. The rudi-
ment of the mesenteron (M/Jnt) has, since Stage IX (Fig. 78)
covered the dorsal side of the yolk, but not until Stage XI does it
come into contact with the rudiment of the enteric muscles. At
this stage however it begins to extend rapidly ventrad over the
yolk—which now is cylindrical in form—insinuating itself between
the latter and the muscle rudiment. At Stage XII (Fig. 81,

202 THE EMBRYOLOGY OF THE HONEY BEE

Meso3L

Fic. 81. Transverse section through the third trunk segment of an
embryo; Stage XII, x 387.

MInt) it covers the lateral face of the yolk and has now reached
the ventral surface. The rudiment of the enteric muscles has
followed this ventrad growth, but only imperfectly so. A short
distance back from the ventral edge of the rudiment of the mesen-
teron are four scattered muscle nuclei, while dorsad of this the
forming muscle cells form a continuous layer which is joined to
the cardioblasts. The appearance of the section suggests that the
cells of the rudiment of the enteric muscles had been forcibly
dragged apart by the rapidly growing epithelium of the mid-
intestine. }

A section through an embryo of Stage XIII-XIV (Fig. 82)
shows the process of differentiation nearly completed. The hypo-
dermis extends on each side up to the dorsal mid-line. The

THE EMBRYOLOGY OF THE HONEY BEE 203

F Cb | Mint
- 3 a \ = ie bs ae SST Pitan
DDph, A “ues is Ds : (yy =
ip CS (Ea) : TENS =e <
sy & ie ee S

>

te
AS
rete So):) Ped

OMel....-| NE

S
oe i je
Sa + a co
SEE
‘
'
Gr

Fic. 82. Transverse section through the third trunk segment of an
embryo, Stage XIII-X1V, x 387.

cardioblasts (Cb/) however are still some distance laterad of this
point, but have lost their connection with the rudiment of the
enteric muscles, which now uniformly cover the lateral surface
of the mid-intestine (M/nt). Above the dorsal diaphragm and
laterad of the cardioblasts are two small cells evidently not
muscle cells; these are fat cells belonging to the third section of
the fat body (3F) or pericardial fat cells. These are derived from
the dorsal margin of the somatic layer of the mesoderm. The
other tissues and organs derived from the mesoderm have vir-
tually attained their final form, and may be recognized without

difficulty.

204. THE EMBRYOLOGY OF THE HONEY BEE

The final stages leading to the completion of the heart are shown
in figure 83A and B, representing cross sections of the region

Sy ie. SCY iF

aN =a: if 3 Va
SRS MclEnt

ty : Mlnt

Fic. 83. Transverse sections through the dorsal region of the fourth
trunk segment of late embryos, illustrating the final stages in the develop-
ment of the heart. A, Stage XIII-XIV; B, XIV, x 600.

of the heart in the fourth trunk segment of Stages XIII-XIV
and XIV. In the first figure (83A), the cardioblasts (Cbl) form
two strips or strands of cells, each situated a short distance laterad
of the mid-line, and with their mesial edges directed slightly
dorsad. The rudiment of the dorsal diaphragm (DDph), consist-
ing of a single layer of cells, projects on each side from the
lateral border of the cardioblasts. On the right hand side of this
figure, a narrow ridge projects out from the cardioblasts, just
below the dorsal diaphragm. This is the last vestige of the
former union of the cardioblasts and the remainder of the splanch-
nopleure. In the next figure (83B) the mesial edges of the cardio-
blasts of each side have approached one another until they are
on the point of meeting. The former lateral margins of the
cardioblasts have meanwhile also moved rapidly mesiad and have
been turned about or “tucked under,” so that they also are now
directed mesiad and almost in contact with one another. In
transverse sections therefore the cardioblast strand of each side is

THE EMBRYOLOGY OF THE HONEY BEE — 205

crescentic in outline. The formation of the heart is obviously
completed by the fusion of the two cardioblast strands (see
Fig. 76).

The two rudiments of the dorsal diaphragm (DDph) are seen
to be formed from a single layer of more or less rounded cells,
originally of uniform size. These, at Stages XIII and XIV
(Figs. 83A and B, DDph) are seen to have lost this uniformity,
cells of two sizes being now present, some of them having nuclei
whose size approaches that of the fat cells, while others have
nuclei only about half as large. The larger cells moreover have
a rounded contour, the smaller on the contrary are thin and flat.
These are to be considered as constituting, in a strict sense, the
diaphragm, while the larger rounded cells are the paracardial cells.
This is in harmony with the results of both Heymons (1895) and
Carriére and Biirger (1897) and justifies the statement made on
a previous page in regard to the homology of these cells.

The ventral diaphragm is formed from muscle fibres arising
near the ventral longitudinal muscles, which extend out toward
the mid-line to join those of the opposite side.

The structure and mode of development of the mesoderm just
described obtains throughout the region of the trunk, extending
from the middle of the second maxillary segment to the thirteenth
trunk segment. As has already been said, it is less modified than
in the head or in the extreme posterior abdominal region. In the
anterior region of the head, at Stages VIII and IX, there is
found a mass of somewhat loosely arranged mesoderm cells
filling the head cavity, the more central of which are applied to
the surface of the stomodaeal invagination (see Fig. 52 Meso).
Caudad of this region the mesoderm cells divide into two lateral
groups, and these in turn give place in the antennal region to a
pair of thin-walled sacs, each composed of a single layer of flat
cells. These are the coelomic (mesodermal) sacs of the antennal
segment (Fig. 84, AntMeso). Each of these coelomic sacs sends
off laterad a solid prolongation into the cavity of the antennal
rudiment, filling the latter like a plug. The antennal coleomic sacs
diminish in size caudad and each finally terminates in a flattened
strand of mesoderm which traverses the mandibular and first
maxillary segment and in the second maxillary segment becomes

206 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 84. Transverse section through the cephalic region of an embryo,
Stage IX, intersecting the antennal rudiments (Ant) and their mesodermal
sacs (AntMesa). Ectoderm shown in outline only, x 243.

continuous with the mesodermal tubes. The cavities of all of the
rudiments of the gnathal appendages are stuffed full of meso-
dermal cells, which in the first and second maxillary segments are
joined to the lateral strand on each side by a thin bridge of cells
(Fig. 67, Meso).

The fate of the mesodermal elements of the head is briefly as
follows: The central part of the anterior mass of mesoderm
enveloping the stomadaeum becomes the layer of oesophageal
muscles. The remainder of this mass is consumed in the produc-
tion of those muscle strands which traverse the labrum and adja-
cent parts of the head capsule. The coelomic sacs of the antennal
segment, at first quite small (Fig. 84, AntMeso) expand rapidly
as the yolk is withdrawn from the head region and at Stage X
become thin walled sacs which nearly fill the posterior half of the
head capsule. Their mesial walls are however thicker than the
lateral walls. As these sacs expand they acquire, in tranverse
section, a more or less crescentic outline, their concave sides being
turned towards one another and enclosing the future oesophagus
(Fig. 58A and B, AntMeso). At Stage XIII the extreme mesial
edges of these coelomic sacs, corresponding to the dorsal points of

THE EMBRYOLOGY OF THE HONEY BEE 207

the crescents, meet and coalesce in the posterior region of the
head, forming here an annular blood lacuna around the oesopha-
gus (Fig. 42). On the dorsal side the line of coalescence extends
back to and joins the dorsal wall of the heart, so that the latter
empties into the annular lacuna, which now constitutes the aorta.
This condition obtains only in the extreme posterior region of
the head, since the ventral line-of coalescence is short, so that the
aorta soon comes to open on its ventral side into the general
cavity of the head (Fig. 58C). The history of the remainder of
the cephalic mesoderm offers little of interest. The mesoderm
cells increase by division and arrange themselves gradually into
masses corresponding in form to the various muscles of the
head, which they are destined to produce. The mesodermal mass
which constitutes the principal adductor muscle of the mandible
is seen in figure 52 Meso. With the exception therefore of the
antennal coelomic sacs, the mesoderm of the entire head, includ-
ing the gnathal region, is consumed in the production of muscu-
lar tissue.

The mesoderm of the posterior trunk segments, that is beyond
the eleventh trunk segment, has not been studied in detail, owing
to the difficulty presented by the flexure of the posterior end of
the embryo about the caudal pole of the egg. A flattened mass
of mesoderm cells was however observed lying close to the anter-
ior side of the proctodaeum (with reference to the axis of the
embryo), at the time of appearance of the latter. This mass
probably belongs to the terminal or fifteenth trunk segment. At
a slightly later stage the proctodaeum is enveloped by mesoderm
cells, evidently derived from this mass, and this envelope of
mesoderm obviously represents the rudiment of the muscular
layer of the proctodaeum. This is in agreement with the observa-
tions of Heymons (1895) on Forficula.

The successive phases of development undergone by the meso-
derm, differ to a considerable degree among the representatives
of the different orders of insects, nevertheless a fundamental simi-
larity exists, and in the development of the mesoderm of the major-
ity of insects, subsequent to its establishment asa germ layer, a
common type is discernible, more or less modified, according as
the insect—speaking generally—is specialized or primitive. In
the myriopods, which probably stand as close to the ancestors of

208 THE EMBRYOLOGY OF THE HONEY BEE

the insects as any existing group of the arthropods, the mesoderm,
shortly after its formation, becomes broken up into segments, the
mesodermal somites, and these in turn give rise to a series of
paired sacs, the coelomic sacs, a pair of these corresponding typi-
cally to each segment. Each of these sacs sends out a hollow
prolongation into the appendage of the corresponding side of the
segment. The presence of paired coelomic sacs is probably primi-
tive since it is also characteristic of the annelid worms. A pre-
cisely similar condition is found in the trunk region (thorax and
abdomen) of Lepsima (Heymons 1897) and several members
of the Orthoptera (see Heymons 1895). In all these forms the
walls of the mesodermal sacs are thin and the various derivatives
of the mesoderm are not formed im situ, as in the honey bee and
many other forms, but from infoldings of the walls of the sacs.
In Forficula (Heymons 1895) the coelomic sacs have thicker
walls than in the forms previously mentioned and their deriva-
tives are formed directly from them, without the intervention of
folds. In the Coleoptera (e.g. Hydrophilus Heider 1889) a still
more modified condition exists. Here the coelomic sacs of the
trunk are thick-walled, and their lumen is relatively small, being
restricted to the extreme lateral region of the mesoderm, a
condition recalling that found in the honey bee, except that in
the latter the cavities are continuous longitudinally, constituting
the mesodermal tubes. In the muscids (Graber 188q) the cavi-
ties of the mesodermal somites are entirely absent, the mesoderm
being solid throughout. This is without question a highly special-
ized condition.

In the majority of instances the cavities of the coelom are
formed secondarily by clefts which appear in the previously solid
mesoderm. Heider (1889) however has maintained that the
coelomic cavities are produced by a separation of the two layers
of mesoderm previously formed by the longitudinal folding of
the blastoderm (see p. 49). This view was strongly opposed by
Graber (1891) and has since not been confirmed by other investi-
gations on the development of the Coleoptera; on the other hand
Carriere (1890) has found that the coelomic cavities in Chali-
codoma are formed in precisely this manner.

In all cases there is a median layer of mesodermal cells connect-
ing the two lateral rows of mesodermal somites.

THE EMBRYOLOGY OF THE HONEY BEE — 209

In the honey bee, the mesoderm, as described by Kowalevski
(1871), Grassi (1884) and the writer, the coelomic cavity is
represented by a narrow cleft, near the lateral margins of the
mesoderm. Here the splanchnic and visceral layers are well
defined and thick, being composed of long epithelial cells, but
mesiad of this point these layers become rapidly thinner and are
poorly defined. In Chalicodoma (Carriére and Birger 1897) this
distinction is even more sharply marked (see p. 197). In the
honey bee no coelomic sacs, as such, are distinguishable, the two
mesodermal layers on each side being continuous, longitudinally,
as well as the cavity bounded by them. In Chalicodoma the thin-
walled portion of the mesoderm, lying mesiad of the mesodermal
tubes is distinctly divided into segmentally arranged sacs (‘“‘meso-
dermal sacs”), while Biirger (1897) found that the mesodermal
tubes were at a certain stage divided by faint partitions into
chambers corresponding to the “mesodermal sacs.” In the case of
this insect therefore well developed coelomic sacs are present,

‘although flattened dorso-ventrally so that the somatic and vis-

ceral layers are in contact with one another, except at their lateral
margins. Here the cavities of the coelomic sacs are best developed
but are on the other hand virtually fused in a longitudinal direc-
tion. In the honey bee this fusion has progressed to such an
extent that the adjoining walls of the coelomic sacs of each side
have been completely lost and their dorsal and ventral walls have
become continuous throughout the entire length of the trunk.

As regards the fate of the different parts of the trunk meso-
derm, modern investigators are in fairly substantial agreement.
From the outer or somatic layer of the coelomic sacs are pro-
duced the trunk muscles and the major portion of the fat body,
from the inner or visceral layer are produced the muscles of
the mid-intestine and the genital ridges. The heart is formed
from cells situated at the external margin of the coelomic sacs,
where an angle is formed by the junction of the somatic and
visceral layers. The median layer of mesoderm forms the blood
cells.?+

*Nusbaum (1886, 1888), Nusbaum and Fulinski (1906) and Hirschler
(1905, 1909, 1909a) claim that a portion of the mid-intestine is formed
from this median strip, which is therefore regarded by these investigators
as entoderm (see pp. 73).

210 THE EMBRYOLOGY OF THE HONEY BEE

The formation of the definitive or secondary body cavity is
inaugurated by the appearance of the epineural sinus, which is
produced mainly by a withdrawal of the yolk from the embryo
along the ventral mid-line. As the yolk continues to withdraw
from the embryo the epineural sinus becomes extended laterally.
Next the dorsal or visceral wall of the coelomic sacs becomes
broken through so that their cavities become continuous with the
epineural sinus, thus forming the definitive body cavity.

The development of the mesoderm in the honey bee conforms to
the above ideal scheme quite well, except in one rather important
particular: the fat body—with the exception of the pericardial fat
cells—is formed, not from the somatic wall of the coelomic sacs,
but from their visceral wall, or rather that part of it not used up in
the production of the enteric muscles. This peculiarity is also
shared by the mason bee, Chalicodoma, so that it seems probable
that it may be peculiar to the Hymenoptera in general. There
appears to be no evident reason for this divergence and it would |
be futile to speculate concerning it.

In that region of the insect embryo which is to constitute the
definitive head, the segments in this region are naturally much
more modified than in the trunk. In Forficula (Heymons 1895)
besides a paired mass of mesoderm in front of the stomodaeum,
there is a well developed pair of coelomic sacs in the antennal and
the three gnathal segments, and in addition a pair of unmistakable,
although reduced coelomic sacs in the premandibular segment. In
other Orthoptera (see Heymons 1895) a similar condition obtains,
except that the coelomic sacs of the premandibular segment are
reduced to mere groups of cells. In the embryos of the Coleoptera
the accounts of the mesoderm of the head are somewhat con-
flicting. Heider (1889) states that in Hydrophilus the coelomic
sacs are wanting in the cephalic region, but appear suppressed in
the mandibular segment, and their development is delayed in the
first maxillary segment. In the most recent account of the devel-
‘opment of a Coleopterous insect (Donacia), Hirschler (1909)
finds, in the cephalic region, coelomic sacs only in the intercalary
(premandibular) and second maxillary segments. In the Hymen-
optera, according to Carriére and Biirger’s (1897) description of
the head mesoderm of Chalicodoma (pp. 392-393) :

Sa ae

%e

_——_—_
4,

et salle

~ _ ~ , oe o “Oe tow a SS 6 ao a =F “ —- Ae
Ee . Se i
x s « : = 5 ae ~_————- Ss =

a Os
ee

a ee

THE EMBRYOLOGY OF THE HONEY BEE ait

“Tn an embryo, in which the mesoderm tubes are present we can
generally follow the mesoderm forward into the antennal segment.
We can determine, that it fills the antennae and all the jaw rudi-
ments, inclusive of the premandibular rudiment, forming their
cores. Its greatest development is attained by the mesoderm
in the sides of the head segments, where it is represented by sev-
eral cell layers, it is wanting in the middle or only forms, as in
the antennal and premandibular segments, a small bridge con-
sisting of a single sheet of scattered cells.

Possibly even before the appearance of the cavities which rep-
resent the beginnings of the primitive body cavities, a cleft
becomes evident in the head region of the germ band, between
ectoderm and yolk (precisely between the rudiments of the ventral
ganglionic chain and the yolk), in this cleft the mesoderm cells
are much scattered. This space later is continuous with the
epineural sinus, and the mesoderm cells scattered, that is floating
in it, are to be regarded as blood cells. It later extends greatly,
although the mesoderm contained within it can scarcely be said to
increase and in general remains limited to the appendages and
their bases. Therefore in the head section of Chalicodoma there
is no formation of coelom. The antennal segment however is an
exception.

A little later coelomic cavities appear in the antennal segment.
This condition is virtually paralleled in the honey bee.

The fate of the head mesoderm as gleaned from the various
accounts, is briefly as follows: In Forficula (Heymons 1895)
the mesodermal mass anterior to the antennal segment forms the
muscular coat of the fore-intestine. The antennal coelomic sacs
become greatly extended, particularly in a longitudinal direction,
while their mesial walls become thickened, and finally join in
such a way as to form a tube, the aorta, continuous with the heart
at its caudal end. The mesoderm in the remaining segments is
consumed in the production of muscles, except in the premandi-
bular segment. Here the mesoderm forms the “suboesophageal
body” described by Wheeler (1893). In the Coleoptera as repre-
sented by Donacia (Hirschler 1909) the muscular coat of the
fore-intestine is apparently formed by the mesoderm situated in
the anterior head region, judging by the figures. The aorta is
formed by coelomic sacs in much the same way as in Forficula,
except that Hirschler confidently asserts that these belong not to

the antennal, but to the premandibular segment. In Chalicodoma

the course of events is much like that in Forficula, except that no

212 THE EMBRYOLOGY OF THE HONEY BEE

suboesophageal body is formed in the premandibular segment.
To this the behavior of the head mesoderm of the honey bee cor-
responds closely. The only difference of any consequence is in
the formation of the aorta. In Chalicodoma, as in Forficula, late in
the development the walls of each coelomic sac become apposed,
obliterating the lumen. It is not clear however that a tubular
lumen is formed, but from the figures given it appears that this
is not the case, but that the aorta is simply a lacuna or blood
space bounded above by the walls of the antennal coelomic sacs,
but opening below into the cavity of the head.

XII

SEx ORGANS—THE OVARIES

The rudiments of the ovaries, in the newly hatched larva consist
of two elongate masses of cells, situated close to the dorsal sur-
face of the mid-intestine, near the mid-line, and extend through
the seventh to the ninth trunk segments inclusive. In their gen-
eral form the ovarian rudiments are elongate fusiform, compress-
ed in a direction at right angles to the surface of the larva (Figs.
XV, 85, Ov), and are about twice or three times as wide as thick.
The contour of the rudiments is however actually irregular, as the
figure shows. In transverse section the outline of the ovarian
rudiments is, generally speaking, that of an elongate ellipse (Fig.
85, Ov). Six to ten rows of nuclei are included in the long axis
of the ellipse, in its short axis not more than two rows, and fre-
quently only one. A somewhat irregular layer of flattened cells
surrounds the ovarian rudiment, and constitutes the rudiment of

Fic. 85. Dorsal part of a transverse section through the posterior half
of the eighth trunk segment of a newly hatched larva (Stage XV), inter-
secting the ovaries (Ov), x 567.

213

214 THE EMBRYOLOGY OF THE HONEY BEE

the ephithelial envelope of the ovary. The cells of this layer are
joined together to form an epithelium, but as yet only an imper-
fect one, since there are frequent gaps between the cells, and there
are also many cells destined to form epithelium, which have not
yet become completely differentiated. This is evident in figure 85.
At the mesial borders of the ovarian rudiments the ovarian epithe-
lium is fairly well formed, and here it is united by a slender point
of attachment to the ridge, previously described, which extends
along the ventral side of the heart, thus holding the ovaries in
place. The component cells of the ovarian rudiments present
nothing peculiar in character. Their nuclei are large as compared
with those of the adjoining tissues, approaching the size of those
of the mid-intestine. As will appear later the cells of the ovarian
rudiments are not perceptibly different from the undifferentiated
mesoderm cells of earlier stages.

The ovaries of the honey bee are derived from the genital ridge,
which is formed from the visceral wall of the mesodermal tubes
in the fifth to the tenth trunk segments inclusive. At Stage X
(Fig. 86A.) the visceral wall of the mesodermal tubes has already
become divided into two layers; an inner single layer of cells,
lying in contact with the yolk, the rudiment of the enteric muscles,
or splanchnopleure, and an outer, thicker portion, in which the
cells are somewhat irregularly but compactly arranged and which
constitutes a part of the visceral layer of the mesoderm. This is
the portion destined to form the genital ridge (Fig. 86A, Ov). In
tranverse section it approximates an ellipse in outline, and in its
narrower diameter includes from one to two layers of cells which
in appearance do not perceptibly differ from those of the adjacent
derivatives of the mesoderm. Even at this early stage the genital
ridge has become narrowed at its point of attachment to the neigh-
boring structures, the cardioblasts and the splanchnopleure.

Prior to Stage X, the cells destined to constitute the genital
ridge may be seen in active mitotic division and during this period
it increases slightly in thickness. Its nuceli seem also to increase
slightly in size. At the next stage (XI) the remainder of the vis-
ceral layer, now composed of loosely arranged cells, becomes
detached from the genital ridge. The outer or somatic wall of
the mesoderm is meanwhile becoming differentiated into the rudi-

THE EMBRYOLOGY OF THE HONEY BEE 215

(=) j----Mint

Foe

----MclEnt

Fic. 86. Transverse sections through the left rudiment of the ovary
M (Ov) of four different stages. A, Stage X-XI; B, Stage XI; C, Stage
Z XIII-XIV; D, Stage XIV, x 567.

ments of the dorsal diagraphm (DDph) and of the trunk muscles
(DLMcl) while at the juncture of the two layers the cardioblasts
(Cbl) are found.

216 THE EMBRYOLOGY OF THE HONEY BEE

The remaining cells of the visceral wall of the mesoderm mesiad
of the genital ridge constitute a loose mesenchymatous tissue the
greater part of which is transformed into fat cells. At an early
period—Stage XI (Fig. 86B)—a few of the dorsalmost cells of
this mass group themselves about the genital ridge, pushing up
on both sides of it, and become applied to its outer surface pre- .
paratory to forming the epithelial envelope of the ovary (Fig.
86C and D).

At Stage X, the genital ridge, as has been said, extends from
trunk segments five to ten inclusive. At about Stage XII it loses
its connection with the cardioblasts and the splanchnopleure and
during the succeeding stages it becomes gradually shorter and
thicker (compare Fig. 86B, C and D), until at Stage XIV it
extends only from near the anterior end of the seventh trunk
segment into the anterior end of the ninth, which is its position at
hatching (Fig. XV). Meanwhile the mesoderm cells grouped
about the ovarian rudiments have gradually assumed the flattened
form characteristic of epithelial cells and are plainly seen to be
forming an envelope for the ovaries (Fig. 86, C-D). The attach-
ment of this envelope to the heart takes place during Stage XIV.

The development of the genital organs of the honey bee was
first studied by Bitschli (1870), but owing to imperfect technique,
his observations are of little value, since he succeeded only in
seeing and identifying the ovaries in a late embryo—about Stage
XIII. The account of Grassi (1884), although brief, is more
satisfactory. This investigator describes and figures the develop-
ment of the ovaries, with the aid of sections, and finds that these
organs are derived from the mesoderm. He seems however to
have failed to observe their origin from the inner (visceral) wall
of the mesodermal tubes. Moreover his statement that the genital
organs extend almost from the fourth to the eighth abdominal
segments is an evident error since at the stage in question (X or
XI) they extend much further cephalad. Moreover, the epithelial
covering of the genital organs was missed entirely. Petrunke-
witsch has more recently (1901, 1902) devoted two papers to an
account of the development of the sex organs of the honey bee.
In these papers he describes conditions which are to say the least,
somewhat peculiar and much at variance with the accounts given

et i a i

THE EMBRYOLOGY OF THE HONEY BEE 217

by other investigators of the genital organs of insects. According
to this investigator the development of the male and female organs
of sex differ greatly. In the male the sex cells are derived origi-
nally from a cell produced by the fusion of the central product of
the division. of the first polar body with the second polar body.
The cell thus produced gives rise by division to a considerable
number of cells which migrate in two groups to the dorsal side of
the embryo, where they again unite. From here they migrate
caudad, still increasing in number, to the abdominal region, where
they penetrate into the mesodermal tubes. Having arrived at this
point they become massed together to constitute the testes. The
ovaries, on the other hand, are formed from mesodermal cells, but
not from those of the mesodermal tubes, but from the loosely ar-
ranged cells derived from the mesoderm lying mesiad of the tubes.
An examination of Petrunkewitsch’s figures (1903, Fig. 17) shows
that the mesoderm cells described as forming the ovaries are
identical with those which the present writer finds constituting the
epithelial envelope of the ovaries. The only remaining account of
the development of the genital organs of a hymenopterous insect,
aside from the more or less fragmentary accounts relating to the
parasitic forms, is in Carriere and Biirger’s description of the
development of the mason bee (1897). In this form the rudi-
ments of the genital organs first become evident as thickenings
of the inner or visceral wall of the mesodermal “sacs” of the third,
fourth and fifth abdominal segments. Since the sex rudiments
thus belong to the wall of the mesodermal “sacs,” they are
situated mesiad (or ventrad) of the mesodermal tubes. The rup-
ture of the visceral wall of the mesoderm takes place, as in the
honey bee, at the juncture of the mesodermal tubes with the re-
mainder of the inner wall, consequently the sex rudiments are
separated from the mesodermal tubes and subsequently come to
lie free within the secondary or definitive body cavity. The
epithelial envelope of the sex glands is formed by mesodermal
cells which lie laterad of the rudiments of these glands, and it
is by means of this envelope that they become attached to the
heart, as in the honey bee. Toward the close of development the
rudiments of the sex glands contract longitudinally so that at the
time of hatching they lie entirely within the limits of one segment,

218 THE EMBRYOLOGY OF THE HONEY BEE

the fifth abdominal. Two points of difference are thus to be noted
between the mason bee and the honey bee. In the former the
rudiments of the sex glands are formed from the visceral wall
of the mesoderm mesiad (ventrad) of the visceral wall of the
mesodermal tubes, instead of from the visceral wall of the tubes
themselves. Moreover the rudiment originally extends through
but three of the abdominal segments instead of five, and contracts
so as to lie in but one segment, instead of three.

Up to the year 1895 the belief prevailed among embryologists
that the genital cells in insects wereasa rule derived directly from
mesoderm cells. To this rule however several important excep-
tions had already been noted; in all these cases the genital cells
were derived not from the mesoderm, but from special cells
set aside at an early period in the development. As early as 1865
Leuckart and Metschinkoff discovered that the genital organs in
Cecidomya were formed from a group of cells (the “pole cells” of
Robin and of Weismann) derived from a single cleavage cell
originally situated at the posterior pole of the egg and readily
distinguishable from the other cells during the early stages of
cleavage. This discovery was confirmed by Balbiani (1882, 1885 )
and Ritter (1890) in Chiromomus. Metschnikoff as early as 1866
found that the cells destined to form the genital organs were dis-
tinguishable at an early period of development in Aphis, and
this observation has been confirmed by later investigators (Wit-
laczil 1884, Will 1888a), while Woodworth (1889) and Schwan-
gart (1904) obtain similar results in the case of the butterflies.
In 1895 Heymons described the sex cells of Forficula and several
members of the Orthoptera as being readily distinguishable at
about the time of the formation of the mesoderm, and located
at the posterior end of the germ band. They subsequently migrate
cephalad along the somatic layer of the mesoderm to their defini-
tive position. Soon after the publication of Heymons’ researches,
Lecaillon (1897a) announced that in certain chrysomelid beetles
the sex cells could be distinguished during the formation of the
blastoderm as a group situated at the posterior pole of the egg.
and differing from the ordinary blastoderm cells in their larger
size and darker stain. Their subsequent migration is similar to
that of Forficula and the Orthoptera. These results have been

ST ee OT as

=e
~

THE EMBRYOLOGY OF THE HONEY BEE 219

subsequently confirmed by several investigators (Friederichs
1906, Hirschler 1909, Hegner 1908).*?

In the light of all of these observations, which relate to members
of five of the orders of the insects, it seems probable that in all
insects the germ cells are segregated at an early period. This is
also in harmony with modern investigations on certain other ani-
mal forms, as for example those of Boveri (1887) on the round
worm Ascaris, and Hacker (1897) on the crustacean Cyclops.
In all of those insects, however in which an early segregation of
the germ cells has been directly observed, these cells are more or
less readily distinguishable by means of differences in size, clear-
ness, etc. This is quite evident from the figures given by Wheeler
(1893) and Heymons (1895), and is especially well shown by
an original figure by Henneguy (1904, Fig. 378). In many
insects the sex cells closely resemble the other cells of the embryo.
This was found by Heymons to be the case in Periplaneta orien-
talis and in Gryllus. In these instances however the early segre-
gation of the germ cells can be safely inferred by a comparison
with closely related forms in which this difficulty was not en-
countered. In the honey bee there is, as a rule, little difference
between the cells of different tissues and organs at the earlier
stages. As has already been stated, the cells which are to consti-
tute the ovary are at first indistinguishable from those of the meso-
derm. It does not however follow that they are of mesodermal
origin, even although they seem to constitute a portion of the
mesoderm, since it is not at all unlikely that the germ cells may
be set aside at an early period in development, and afterwards
migrate into the visceral wall of the mesodermal tubes, and that
such a migration may take place unobserved, on account of the
similarity of the sex cells and mesoderm cells. A solution of this
problem is perhaps to be found in the discovery of some constant
difference hitherto unobserved, between the sex and other cells.
An approximate solution is probably easier. Ifa closely allied form
should be found in which the germ cells can be recognized with-
out difficulty, then the behavior of the germ cells in the honey bee
could be inferred by analogy.

* A complete review of the literature on this subject may be found in a
recent paper by Hegner (1914).

XIII

ALIMENTARY CANAL

In the larva the alimentary canal comprises a short and narrow
oesphagus, a capacious mid-intestine, and a short and but slightly
curved hind-intestine. The mouth opening in the young larva is a
transverse slit located just behind the labrum in the area between
the bases of the mouth parts (see Figs. XIV and XV). At the
posterior margin of the mouth is a flattened papillate process
formed by the cephalad prolongation of the hypodermis forming
the junction between the posterior (ventral) wall of the oeso-
phagus and the ventral hypodermis. Just beyond the mouth the
oesophagus dilates slightly to form a somewhat ill-defined pharynx,
which is furnished with dilator muscles. From here the oesopha-
gus curves uniformly dorsad and caudad and joins the mid-
intestine just caudad of the point where the head and trunk
join. The mode of junction is illustrated in figure 45, which
shows that the fore-intestine (FJnt) is invaginated into the lumen
of the mid-intestine to form a structure corresponding in
form to the proventricular valve of the imaginal bee. A similar
oesophageal valve is found in many other insects. The lumen of
the oesophagus, in its anterior portion, is crescentric in section,
owing to the presence of a dorsal longitudinal fold (Fig. 41). In
its posterior half four folds are present (Fig. 42), so that here
the lumen has in section the form of a cross.

The mid-intestine is relatively large and occupies the greater
part of the body cavity, extending from the first to the eleventh
trunk segment (Fig. XV, M/nt). Its form is that of an elongate
cylinder with rounded ends. At its anterior end it communicates
with the oesophagus, as just described, but its posterior end is
completely closed. Its walls are relatively thin, but the cells
composing them are somewhat larger than the majority of the
other tissue cells. They are cubical in form, with a rounded
nucleus (Fig. 75, MJnt). The cytoplasm of these cells is granular
and dark staining, and each cell commonly contains one or more
large vacuoles.

220

THE EMBRYOLOGY OF THE HONEY BEF 221

The hind-intestine is a short tube, of about the same diameter
as the oesophagus, bent in a gentle sigmoid curve, and tra-
verses the three terminal trunk segments (Fig. XV, H/nt). Its
blind anterior end rests against the posterior end of the mid-
intestine ; its posterior end opens to the exterior at the posterior
end of the trunk. Its structure is much the same as that of the
oesophagus, except that its lumen is circular and relatively
smaller.

The Malpighian tubules, four in number, open into the anterior
blind end of the hind intestine (Fig. XV, Mal). From here they
extend cephalad in a winding course to about the sixth trunk
segment. Their diameter is slight, somewhat less than that of
the lateral tracheal trunks, and is uniform throughout their length
(Fig. 75, Mal). Their lumen is relatively small and circular in
section.

The mid-intestine or mesenteron is derived exclusively from the
anterior and posterior mesenteron rudiments. At Stage VII
(Figs. 26B and 29, AMR) the anterior mesenteron rudiment is a
thick convex disk of cells resting on the ventral side of the yolk,
at the cephalic end of the egg. This disk is covered by the ecto-
derm with the exception of a small area at its posterior margin,
where the cells of the disk extend through a gap in the ectoderm
to the external surface (Fig. 26B). During the interval between
Stages VII and VIII the germ band rapidly increases in length,
the effect of this being to cause the cephalic end of the germ band
to travel around the cephalic pole of the egg, carrying with it the
anterior mesenteron rudiment, which assumes a position on the
cephalo-dorsal face of the yolk. In this position the morpho-
logical posterior margin of the rudiment, where its cells reach the
external surface, lies approximately at the cephalic pole of the
egg. This rudiment, as a whole, has at first the form of a thick
cap, but by the time Stage VIII is reached, it has become elongated
in a longitudinal direction. Its form may then be compared with
that of a spoon or scoop (Fig. 87A). The part corresponding to
the handle is situated at the cephalic pole of the egg, where the
stomodaeum (Stom) has already put in its appearance, and here
it penetrates the ectoderm to the exterior, forming the bottom of
the stomodaeal depression, while the remainder, corresponding to

222 THE EMBRYOLOGY OF THE HONEY BEE

the bowl, extends over the dorsal surface of the yolk toward the
caudal pole of, the egg. Meanwhile the posterior mesenteron~
rudiment has also become altered in form. At Stage VII (Fig.
27C, PMR) it is disk-shaped, but between Stages VII and VIII
it rapidly elongates, sending out a thin tongue-like process over
the dorsal region of the egg toward the cephalic pole (Fig. 87B).
At the same time the ectoderm grows caudad (with respect to the
embryo) around the caudad pole, covering over that part of the
posterior mesenteron rudiment which lies closest to the caudal
pole of the egg (compare Figs. 27C and 87B, PMR). With the
development of the tongue-like extension of the posterior mesen-
teron rudiment the rudiment as a whole becomes thinner, espe-
cially in the mid-line. The relatively rapid changes in form under:
gone by both mesenteron rudiments during Stages VII and VIII,
appear to be due principally to changes in the form and arrange-
ment of their component cells, and not to an increase in their
number or volume.

In the example of Stage VIII represented by figures 87A and B,
the anterior mesenteron rudiment extends over the dorsal surface
of the yolk about one-third of the length of the egg. The thin
tongue-like process of the posterior rudiment on the other hand
extends towards the cephalic pole nearly one half of the length
of the egg. The opposite edges of the two rudiments are therefore
not far apart. During the interval between Stages VIII and IX
both mesenteron rudiments continue to spread over the dorsal
surface of the yolk until their opposite margins meet and fuse
in the dorsal mid-line. The point of meeting is usually nearer
to the cephalic than to the caudal end of the egg, on account of
the more rapid growth of the posterior rudiment along the median
dorsal surface of the yolk. By Stage IX the cells of the united
mesenteron rudiments have become distributed in a single layer
over the dorsal surface of the yolk (Fig. 78, M/nt, 88A, AMR),
The cephalic end of the yolk is also covered by the epithelium
of the anterior mesenteron rudiment, part of which forms the
floor of the stomodaeal invagination, as shown in figure 88A,
AMR. The caudal end of the yolk on the other hand is still
uncovered (Fig. 88B, PMR). The thickness of the newly formed
epithelium of the mesenteron is at first not uniform, being greatest

PMR

Fic. 87. Anterior (A) and posterior (B) ends of a median sagittal
section through an embryo, Stage VIII, showing the anterior (AMR)
and posterior (PMR) mesenteron rudiments, which are about to unite on
the dorsal surface of the egg. The stomodaeal invagination (Stom) is
just appearing. The relation of this to the anterior mesenteron rudiment

is clearly shown, x 243.

224 THE EMBRYOLOGY OF THE HONEY BEE

at about the point of junction of the two rudiments. Near the
posterior end of the yolk the epithelium is especially thin, as
shown in figure 88B, PMR. At Stage X (Fig. 79 MInt) the
posterior end of the yolk is covered by the future epithelium of
the mid-intestine which now extends ventrad on each side to meet
the splanchnic layer of the mesoderm. During the next three
stages (Figs. 80, 81 and 82) the epithelium of the mid-intes-
tine extends steadily ventrad on each side, until, at Stage XIII- |
XIV, the two edges meet and unite on the ventral mid-line (Fig.
82) thus completing the formation of mesenteron. The union ap-
pears to be virtually simultaneous throughout its extent. Mean-
while the rudiments of the fore- and hind-intestines, the stomo-
daeum and proctodaeum, are also developing. The stomodaeum
puts in its appearance at Stage VIII in the form of a cup-shaped
depression located at the cephalic pole of the egg. The floor of this
invagination is formed by cells of the anterior mesenteron rudi-
ment, as shown in figure 87A. Between Stages VIII and IX the
depression deepens rapidly, becoming funnel-shaped, while at the
same time the yolk retreats from the anterior end of the egg, its
withdrawal keeping pace with the lengthening of the stomodaeum.
The latter thus soon becomes a short tube, composed of a single
layer of prismatic ectodermal cells, closed at its inner end by the
cells of the anterior mesenteron rudiment (Figs. 52 and 60, Stom).
During the earlier phases of the development of the stomodaeum,
its cavity narrows gradually from its outer to its inner end, and
the cells of the mesenteron closing the latter form a wall almost?
as thick as that of the stomodaeum itself, containing several
nuclei (Fig. 60). Towards Stage XIV the inner end of the
stomodaeum widens out, acquiring a flaring form, and the stomo-
daeal wall becomes much thinner just anterior to its junction with
the mesenteron (89A). Careful examination of preparations of
this stage shows that this condition is associated with the forma-
tion of a double fold of the stomodaeal wall preparatory to the
invagination or intussusception of the inner end of the stomo-
daeum into the cavity of the mid-intestine. Correlated with these
changes is a thinning of the cellular diaphragm-like wall closing
the inner end of the stomodaeum, the central portion of this wall
becoming reduced to a thin membrane, in whch now no nuclei are

THE EMBRYOLOGY OF THE HONEY BEE = 225

Fic. 88. Anterior (4) and posterior (B) ends of a median sagittal
section through an embryo; Stage IX, showing the anterior (AMR)
and posterior (PMR) mesenteron rudiments, which have now united along
the dorsal surface of the egg. The stomodaeal invagination (Stom) is
well developed, while the protodaeal invagination (Proc) is indicated by
a slight depression, x 243.

present (Fig. 89A). The next and final step, which takes place
about the time of hatching, involves the actual invagination of the
inner end of the stomodaeum—now the oesophagus—into the
mid-intestine, and the rupture of the closing membrane. This
rupture seems to be caused mechanically by pressure of the inva-

226 THE EMBRYOLOGY OF THE HONEY BEE

ginating oesophageal wall; at least this is the impression given by
the sections studied, in which the peripheral part of the torn mem-
brane is seen still attached to the wall of the mid-intestine( Fig.
89B). <A similar proventricular valve appears to be formed in
much the same way in Forficula and Periplaneta (Heymons,

1895).

Fic. 89. Sagittal sections through the junction of the fore and mid-
intestines of two embryos, showing the formation of the oesophageal
valve (OeV lv). A, Stage XIV; B, Stage XV, x 567.

At the time of its first appearance the stomodaeum occupies a
position corresponding to the cephalic pole of the egg, and during
its earlier stages its lumen coincides with the long axis of the egg.
At Stage X the embryo begins to shorten, drawing the external
opening of the stomodaeum—the future mouth—to its final posi-
tion on the ventral surface of the egg, the oesophagus therefore
acquiring a curved course. At the same time the hypodermis
forming the angle bounding the future mouth on its morphologi-

THE EMBRYOLOGY OF THE HONEY BEE = 227

cal posterior side, lengthens out to constitute the papillate or
tongue-like process previously mentioned (p. 220). This is plainly
shown, although not lettered, in figure XIII, and lies behind the
labrum, bounded laterad by the mandibles and first maxillae.
The development of the different portions of the digestive
canal from their respective rudiments corresponds essentially
to that found in other insects having bipolar mesenteron rudi-
ments, except in one particular. In all the other insect embryos
thus far studied, in which bipolar mesenteron rudiments occur,
including the mason bee, each of the rudiments sends out a pair
of epithelial bands, one on each side of the ventral mid-line. These
grow toward each other, meet and fuse. The narrow space left
vacant betwen them on the ventral side is next filled by cells de-
rived from the bands. The ventral surface of the yolk is therefore
covered first, and the dorsal surface is covered subsequently by
gradual dorsad growth and extension of the bands. In the honey
bee, as Grassi (1884) discovered, it is the dorsal surface of the
yolk which is first covered by the extension of the mesenteron
rudiments ; moreover, these extensions do not have the form of a
pair of bands. although certain statements made by Grassi suggest
this, so that Carriere (1897) has interpreted Grassi as stating that
each of the two mesenteron rudiments gives rise to “‘paired lateral
arms.” Grassi’s statement with reference to the anterior “ento-
derm’” rudiment is as follows: “It is always more advanced
on the sides than in its median part, so that two or three
tranverse sections are always found in which it fails to occur
in the mid-line.” Further on in the same chapter he makes a
similar statement regarding the posterior “entoderm” rudiment.
Examination of series of tranverse sections of Stage VIII have
failed to completely confirm these statements. During this period
of rapid growth the epithelial layers formed by the two rudiments
of the mesenteron are extremely irregular and variable in thick-
ness, especially in the region of the rapidly advancing thin edges
on the dorsal side of the yolk. In two of the four preparations
of the same stage as figure 87 the statement of Grassi seemed
to hold, while in the other two the advance seemed greatest along
the mid-line. The point is however a trivial one since Grassi’s
statements cannot be fairly construed to indicate the presence of

paired bands. Evidences of these, as such, seem to have almost

228 THE EMBRYOLOGY OF THE HONEY BEE

completely disappeared. The posterior mesenteron rudiment,
gives perhaps the best indication of the former existence of paired
bands in the incomplete bilateral subdivision of its thicker basal
portion at Stage VIII .

The rudiments of the Malpighian tubules appear at Stage
VIII, preceding, as Grassi (1884) discovered, the formation of
the proctodaeum, which does not appear until Stage IX. The
rudiments of the Malpighian tubules therefore for a short time
open directly on the external surface of the embryo. A similar
relation obtains in Microgaster (Kulagin 1892), Leptinotarsa
(Doryphora) (Wheeler 1889), Chalicodoma (Carriére 1890,
Carriere and Burger 1897), and Gasteroidea (Hirschler 1909).
It is only in Apis and Chalicodoma however that the Malpighian
tubules are actually formed before the proctodaeum; in the others
they appear on the external surface of the ectoderm around the
already formed proctodaeum. In the honey bee the Malpighian
tubules appear as four equidistant pit-like depressions at the
extreme posterior end of the germ band, on the dorsal surface of
the egg. According to Grassi‘s account the depressions are at
first entirely distinct from one another, but subsequently the two
lying on the same side of the mid-line merge with one another,
a pair of longitudinal grooves being thus formed. Examination
of preparations of embryos of the proper stage (VIII) failed to
show the invaginations as distinct from one another; in all cases
in which they were sufficiently well defined to be recognizable
they had coalesced on each side of the median line to form a pair
of slightly curved grooves, whose concavities faced one another
(Fig. 90). Examination of sections shows also that the pair

Fic. 90. Dorsal surface of posterior end of egg, Stage IX, showing the
crescentic grooves from which the Malpighian tubules are formed.

THE EMBRYOLOGY OF THE HONEY BEE = 229

of invaginations lying closest to the caudal pole of the egg—the
future ventral pair of Malpighian tubules—are deeper than the
anterior pair, and this difference is evident up to a late stage.
At Stage X the ectodermal area included between the grooves
described above becomes depressed, carrying with it the rudi-
ments of the Malpighian tubules (Fig. 91, Mal). This area

Y

Fic. 91. Transverse section through the posterior end of an embryo,
Stage X, intersecting the proctodaeal invagination (Proc). The posterior
(ventral) pair of Malpighian tubules (Mal) are also shown.

therefore constitutes the floor of the proctodaeum. The open-

ing of the latter is at first on the dorsal side of the egg, on account

of the caudal flexure of the embryo, but with the gradual
straightening of the embryo the external opening of the procto-
daeum becomes terminal with respect to the egg. This takes
place during the next stage (XI). From the time of its inception
the proctodaeum elongates slowly and steadily, the yolk retreating
from the posterior end at a corresponding rate. The inner end
of the proctodaeum comes into contact with the epithelium of the
mesenteron from the first, but the relation does not become more
intimate than that of mere contact. The growth of the Mal-
pighian tubules is extremely rapid; they push forward along the
ventro-lateral wall of the mid-intestine as slender, nearly straight
tubes, until their anterior ends reach the neighborhood of the
sixth trunk segment. Here their cephalad extension seems in
some way hindered, and their further elongation causes them to
be thrown into a series of curves and loops (Fig. XV, Mal).

XIV

YOLK AND YOLK CELLS

In the chapter on “Cleavage” it is stated that not all of the
cleavage cells migrate to the periphery of the egg, a few of these
remaining behind in the yolk to form the yolk cells. In order
to distinguish them from cells entering the yolk at a later period
these yolk cells will be referred to as primary yolk cells, since
they are primary in respect to the time of their origin. At first
the primary yolk cells preserve the appearance of ordinary cleay-
age cells and divide simultaneously with their sister cells, which
are about to form the blastoderm, exhibiting mitotic figures
precisely like those of the cleavage cells (Fig. 93, A and B).
Soon after the blastoderm has become established mitotic divi-
sions of this character cease and are supplanted by mitoses of a
more or less irregular character. In these the spindles are as
a rule short and the chromosomes frequently agglomerated to
form irregular masses (Fig. 93, C-G). This period lasts until
after the nuclei forming the blastoderm have become arranged
in two layers (18-20 hrs.). The large number of cells found
in the yolk at this time is therefore to be attributed to mitotic
divisions, since the latter are found abundantly until some time
after the blastoderm has been formed. This includes both the
earlier regular type as well as the later irregular type. Many
of the mitotic figures of the later type resemble those found in
pathological tissues—such as cancer cells—and often have the
appearance of being unequal (Fig. 93D), or multipolar (Fig.
93, F and G), but the mitotic figures are so minute and the sur-
rounding cytoplasm often so dark that a critical study of them
was not attempted. The probability that both the unequal and
multipolar types exist is strengthened by the subsequent frequent
presence of nuclei of different sizes within the same mass of
crytoplasm (Fig. 92A). (These irregular figures might be con-
sidered actually pathological or abnormal, were it not for their
great abundance in several different preparations of eggs of the

230

THE EMBRYOLOGY OF THE HONEY BEE 231

nap Ra 3 ee j
a C
Fic. 92. Yolk cells from eggs eighteen to twenty hours old. A, large

multinucleate cell, showing degenerating nuclei. B, small binucleate cell;
one of the nuclei appears to be dividing by amitosis. C, small uninucleate
cell, x 840.

same age, which are in all other respects perfectly normal. More-
over Lecaillon (1897a) has reported irregular mitotic figures in
the chrysomelid beetle Clytra during cleavage. It is true that
these figures, as stated in an earlier chapter may quite possibly be
caused by too slow a fixation and are therefore to be considered
actually abnormal. Friederichs (1906) however reports that
irregular mitotic figures are of constant occurrence in the yolk
cells of another chrysomelid beetle (Donacia) and are therefore
strictly normal. This case is precisely parallel to that of the bee.
Dickel (1904) also reports the occurrence of mitotic figures in
the yolk cells of the honey bee, but neglects to say at what stage
they are found. The single small figure given by him (his Fig.
4) indicates however that this is one of the small irregular spin-
dles which succeed those of the regular type.

Examination of the text and figures of Petrunkewitsch’s two
papers (1901 and 1902) indicates that he also probably observed
the small type of mitotic figures in the yolk cells but without
recognizing their nature. The “double nuclei” in the “Rich-
tungsplasma” have much the appearance of the small mitotic
figures when seen at a relatively low magnification. On page 28
of the first paper (1901) occurs the following significant para-
graph: “I should here also mention an observation. At the

232 THE EMBRYOLOGY OF THE HONEY BEE

Fic. 93. Yolk cells, showing mitotic division. A, cell showing a regu-
lar symmetrical spindle of the ordinary (normal) type. B, telephase of the
same type. C, cell showing small but symmetrical spindle. Dy, cell showing
a small and slightly unsymmetrical spindle. E, cell showing chromatin
aggregated into dense masses; telephase. F, G, irregular and asymmetrical
spindles, x 1107.

time when the blastoderm has just been completed, the yolk
is almost free of cleavage cells. These are found isolated

ae ee ee ee ee eee ee

THE EMBRYOLOGY OF THE HONEY BEE = 233

only here and there. However when the head fold?* is formed,
a multitude of these cells appear in the yolk, which are agglomer-
ated to the number of 8, 9 or even more in a single mass; some
of these nuclei are now like the double nuclei described above,
although not so clearly divided into two halves. At this time I
could no longer discover the double nucleated cells. Could those
(just described) have arisen from these?” It seems highly
probable that the imperfectly double nuclei just described are the
small mitotic figures ; the resemblance of these which are found in
the true yolk cells to those seen in the “Richtungsplasma” is espe-
cially interesting.

During the period of irregular mitotic divisions of the yolk
cells small and densely stained bodies appearing to be composed
of nuclear material (chromatin) are found within the cytoplasm
of the yolk cells (Fig. 93G, upper left hand side). The precise
origin of these bodies is unknown, but it seems that like the
similar bodies seen later, they represent degenerating nuclei, and
are also probably in some way causally related to the irregular
mitotic divisions.

In preparations of eggs eighteen to twenty hours old, when the
nuclei of the blastoderm are arranged in a double layer over the
greater part of the blastoderm, the cellular contents of the yolk will
be found to present a richly varied appearance, and one which
differs not a little in different eggs; nevertheless the following
description will apply to the majority of examples of this stage. -
Scattered throughout the yolk are to be seen numerous irregular
islands of dark-staining cytoplasm of the most various shapes
and sizes containing one to several nuclei. These cells—or syn-
citia, since they are nearly all multinucleate—are most numerous
along the central line of the yolk, and are also rather more numer-
ous in the anterior than in the posterior half of the egg, although
syncitia of considerable size are frequently to be seen lying in
contact with the inner surface of the blastoderm. The range
in size of the syncitia may be gathered by comparing figures
92A and 92C. In the anterior half of the egg, always near the
central line of the yolk, there are usually one or two syncitia of

* Reference to the figures shows that by this is meant only the dorsal
strip of thin blastoderm.

234 THE EMBRYOLOGY OF THE HONEY BEE

relatively enormous size, like that shown in figure 92A. In a crude
and general way, both the number and size of the nuclei vary
with the size of the syncitia to which they belong. The largest
syncitia accordingly are commonly found to contain a consider-
able number. of relatively enormous nuclei; as shown in figure
9g2A. All of the yolk syncitia possess branching processes which
are continuous with the threads of the delicate protoplasmic
network remaining within the yolk subsequent to cleavage, and
are thereby placed in direct connection with one another. Ina
broad sense therefore the entire protoplasmic contents of the
yolk forms a single syncitium, directly comparable to that of the
cleavage cells.

A large proportion of the nuclei of the yolk syncitia—in the
narrow sense—show plain evidences of degeneration. A survey
of the preparations covering this period shows that the changes ©
are as follows: The nucleus decreases in size and the entire
contents of the nucleus assumes a darker shade, while the chro-
matin at the same time becomes gradually agglomerated into two
or three irregular dense masses, which finally become united into
one as the nucleus continues to contract. The last observed stage
in this process is a densely stained and relatively minute sui-
spherical mass, as shown in figure 94, A-D. A similar agglomer-
ation of the chromatin in the degenerating nuclei of the yolk
cells has been observed by Wheeler (1889) in Leptinotarsa -
(Doryphora) and by Friederichs (1906) in Domacia. Since the
nuclear membrane is intact at all stages of degeneration, it neces-
sarily follows that the karyolymph is lost by diffusion, passing
out into the surrounding cytoplasm which often takes on a lighter
hue about such nuclei. The masses of chromatin formed during
nuclear degeneration are located in various parts of the nucleus,
in some cases at the center, in some around the periphery, in
others at one side. It thus happens that in some instances a
densely stained spherule is seen lying on the internal wall of a
small vacuole (Fig. 94E), the latter representing the empty
space left within the nuclear membrane. In the majority of cases ©
the degenerating nucleus for some unknown reason lies at first
close to a normal nucleus, so that during the late stages of degen-
eration the appearance suggests that the products of degenera-

THE EMBRYOLOGY OF THE HONEY BEE 235

Fic. 94. Degenerating yolk nuclei. A, yolk cell, showing one large
vesicular nucleus, and three nuclei in each of which the chromatin has
become aggregated into a spherule. B, spherule leaving cell. C, spherule
lying just outside of cell. D, two spherules lying inside body of cell. E,
F, nuclei lying in the meshes of the protoplasmic reticulum, x 840.

tion, the spherular masses of chromatic material, may have
emanated from the normal nucleus (Fig. 94, A and B). Fried-
erichs (1906) in the course of his investigations on the embry-
ology of the chrysomelid beetles, found that during the earlier
stages of development the nuclei of the germ cells of the ectoderm
and of the mesoderm, gave rise, by a peculiar kind of direct divi-
sion to structures resembling degenerating nuclei, which he termed
“paracytoids.” These were not however thus formed in the yolk
cells, but since the nuclei of the yolk cells form precisely similar
bodies by ordinary nuclear degeneration, these too were termed
“paracytoids.” In spite of this last mentioned circumstance and
in view of the frequent apposition of the dark-stained spherules
and normal nuclei in the yolk cells of the honey bee, it seemed
advisable to determine definitely whether any of these spherules
“vere derived by the emission of chromatic material from adjoin-
ing nuclei. A careful examination under high powers of all the
available sections of the stages during the formation of the
blastoderm shows conclusively that there is no evidence of a

236 THE EMBRYOLOGY OF THE HONEY BEE

direct genetic connection between the dark-stained spherules
and the adjoining nuclei, since in every case a small but appre-
ciable space separates the two kinds of nuclei. Moreover the
nuclear membrane of the normal nucleus is always intact at
every point. It may be assumed therefore that all of the sub-
spherical deeply staining masses observed within the yolk are
formed by ordinary degeneration of yolk cells. At the stages
beyond eighteen to twenty hours up to the time of the formation of
the germ layers, densely stained chromatic spherules are commonly
found isolated within the yolk, sometimes appearing as though
caught in the meshes of the protoplasmic network (Fig. 94E),
sometimes surrounded by a fragment of cytoplasm (Fig. 94F).
That these are derived from yolk cells is evident by comparing
them with those still enclosed within the yolk cells (Fig. 94D).
In addition, it is possible to find a spherule in the act of leaving
a yolk cell (Fig. 94B), as well as other stages in the emigration
of the spherules out into the yolk (Fig. 94, A and C). The
chromatin spherules are therefore equivalent to the “entodermal
paracytoids” described by Friederichs (1906) in Donacia, since
they have precisely the same history.

Shortly after mitotic division in the yolk cells ceases, and when
the nuclei of the ventral region of the blastoderm have become
arranged in a double layer, (18-20 hours) not a few of these
nuclei may frequently be detected in the central margin of the
blastoderm, at some distance from their fellows. Others may
be seen half in the blastoderm, and half in the remains of the
peripheral protoplasmic layer lining the blastoderm while still
others are entirely within this layer (Fig. 95). Favorable prep-
arations show a considerable number of nuclei in all these stages.
The migrating nuclei apparently carry with them little or no
cytoplasm and after leaving the blastoderm become embedded in
the dark-stained vestiges of the peripheral layer of cytoplasm.
Henceforward they can no longer be distinguished with certainty
from the smaller representatives of the primary yolk cells. The
yolk cells thus produced by migration from the blastoderm may ~
be termed the secondary yolk cells** by reason of the later time of

* This term has already been applied by Cholodkowsky (1891) to certain
cells in the cockroach embryo which he believed were derived from the

THE EMBRYOLOGY OF THE HONEY BEE 237

Ag . Sale.
er Ps oupss WAAL
ats RIAR gis et Ey Sate
ets sh
: f yy Bis

Fic. 95. Section of the blastoderm of an egg eighteen to twenty hours
old, showing nuclei leaving the blastoderm and entering the yolk, x 840.

their formation. The immigration of secondary yolk cells from
the blastoderm into the yolk is most evident in preparations of
eggs eighteen to twenty hours old. It may be observed in eggs
as old as twenty-four to twenty-six hours; but in these it is
rare. In many preparations of eggs twenty-two to twenty-
four hours old there are present scattered through the yolk,
nuclei almost devoid of cytoplasm corresponding in size and
general aspect to the secondary yolk cells. (Some of these may
be seen in figure 16). While the identity of these cannot be
demonstrated it is reasonable to suppose that they are actually the
secondary yolk cells, since in their uniform size and poverty in
cytoplasm they differ from the majority of the primary yolk
cells. These cells,—presumably the secondary yolk cells,—are not
infrequently aggregated into clusters, a few of these containing
as many as ten or a dozen nuclei. A little later, in eggs twenty-
eight to thirty hours old, these almost naked nuclei have disappear-
ed. In their place are yolk cells of moderate size, resembling in
every respect the smaller representatives of the primary yolk
cells. It seems probable therefore that these are the secondary
yolk cells which have acquired additional cytoplasm, since there
is no other explanation possible of the facts observed. The means
by which this increased amount of cytoplasm is acquired is not
clear. It may have been manufactured by the metabolic activities
of the cell, building cytoplasm out of the food substances of the
yolk, or it may have been drawn directly from the protoplasmic

primary yolk cells by division, and which afterwards contributed to the
formation of the fat body and the blood. Heymons (1895) states that
these are nothing but blood cells, and are therefore not “yolk cells”
at all.

238 THE EMBRYOLOGY OF THE HONEY BEE

complex within the yolk. This complex of course includes the
yolk cells, the latter, as already stated, being linked together by fine
processes to form a complex virtually constituting a syncitium.
Moreover, although the nuclei of the yolk cells degenerate, the
cytoplasm of these cells is never observed in a state of degenera-
tion. It seems not improbable therefore that this may be trans-
ferred by streaming movements to other cells less abundantly
supplied.

During Stages IV to VII the general aspect of the yolk changes
but little. The cells of the cephalo-dorsal body enter the yolk at
the anterior end of the egg at Stage VII, as described in a pre-
vious section. In preparations of Stage IV multinuclear yolk
cells are still to be seen (Fig. 19) but they are wanting in those of
later stages. Aside from the cells of the cephalo-dorsal body,
the cells present in the yolk, up to Stage VII, are quite uniformly
distributed and of moderate size, each containing a nucelus whose
diameter seldom exceeds twice that of the nuclei of the blasto-
derm or its derivatives. The cytoplasm accompanying these
nuclei is relatively small in quantity and lacks the dark-staining
qualities and consequent opacity seen at earlier stages (see Figs.
19 and 20).

At the close of the period of the formation of the blastoderm,
when the rudiments of the mesenteron and the mesoderm are
beginning to show themselves (Stage IV), the cells of the blasto-
derm become sharply demarcated from the yolk. Up to this time
the bases of the cells of the blastoderm have been intimately con-
nected with the protoplasmic network of the yolk, and form a
zone of dark-stained and much vacuolated cytoplasm, the inner
cortical layer (see p. 31). This central portion of the blas-
toderm cells now definitely separates from the blastoderm, remain-
ing in connection with the yolk. The inner surface of the
blastoderm and the outer surface of the yolk then acquire a sharp
and definite outline. The inner cortical layer now forms a thin pro-
toplasmic pellicle around the entire yolk mass enclosing it like a sac
(Figs. 18, 19, 24, 26). The peripherally situated yolk nuclei
naturally enter into a close relation to the surface layer, and this
relation is maintained to near the close of embryonic development.
These nuclei are, however, at first few and scattered, but after the

es Spe, eae aD

pees a a
SEF

: ">

oS

THE EMBRYOLOGY OF THE HONEY BEE 239

epithelium of the mid-intestine has covered the dorsal surface of
the yolk they begin to appear in greater number, and at Stage XI
are seen applying themselves to that portion of the peripheral layer
which is in contact with the ventral margins of the epithelium of
the mid-intestine, as shown in figure 75. In this figure yolk cells
are also seen lying close to the dorsal surface of the yolk. Such
peripheral accumulations of yolk cells are noticeable only in the
anterior half of the mid-intestine. At Stage XII (Fig. 81) many
of the peripheral yolk cells appear to have traveled ventrad in
company with the advancing ventral margins of the epithelium of
the mid-intestine and are now situated close beneath the un-
covered area of the yolk, where they remain until covered over
by the walls of the mid-intestine (Fig. 82). A little later these
nuclei are seen to become shrunken and distorted and soon after
this they disappear completely leaving no trace behind, having
been digested and absorbed by the wall of the mid-intestine.

In addition to the changes described above relating principally
to the yolk cells, other changes take place in the yolk between
Stages IV and XIV. These concern the vitelline bodies and the
protoplasmic meshwork. As early as Stage IX many of the
vitelline bodies are seen to have become much enlarged. Careful
scrutiny of sections under a high power shows that the vitelline
bodies are undergoing a process of fusion, several joining together
to form a larger body and such larger bodies in turn joining
with one another to form still larger bodies. These compound
bodies when first formed are lobate, the lobes corresponding to
the vitelline bodies which have taken part in its formation. Later
the lobes disappear and the bodies become spheroidal or ovate in
form. This process of fusion goes on continuously up to Stage
XIII-XIV when the vitelline bodies are but few in number but of
large size, some of them attaining a diameter of eight to nine micra.
Soon after this they disappear, presumably being digested, and
are absent in examples of Stage XV.

At Stage IX the protoplasmic meshwork of the yolk also begins
to show changes, its meshes here and there being seen to have
increased in size, indicating coalescence of the vitelline spheres.
This coalescence continues slowly but steadily up to Stage XIV,
when the protoplasmic network of the yolk is made up of large

246 THE EMBRYOLOGY OF THE HONEY BEE

but coarse meshes. Soon it becomes ragged, irregular and finely
granular, ultimately, at Stage XV acquiring the appearance of a
pale and finely granular precipitate.

The yolk cells have been given but slight consideration by most
of the investigators of insect embryology. The reason for this
is not far to seek. Many of the earlier embryologists looked
upon the yolk cells as the material from which the mesenteron
was formed, and therefore corresponding not only theoretically
but actually to the entoderm of other animal forms. When, how-
ever, it was found that the mesenteron in the pterygote insects
was derived from other sources, interest in the cells of the yolk
waned, and consequently published observations on them are, so
to speak, sporadic and scattered.

The origin of yolk cells from cleavage cells which do not
migrate to the periphery of the yolk, but remain behind, was es-
tablished by Bobretzsky (1878) in the case of the butterfly Pieris.
This mode of origin of the primary yolk cells was confirmed by
many other investigators. Grassi’s observations on the honey
bee (1884) seem to be entitled to a place among these, although
his statements on this point are not perfectly clear. The origin
of yolk cells in this manner has subsequently proved to be the
rule among pterygote insects, although exceptions were early
noted. Thus Patten (1884) in Neophylax and Korotneff (1885)
in Gryllotalpa, found that all of the cleavage cells migrate to the
surface of the egg. Later some of these wander back into the
yolk to form the primary yolk cells. This is commonly taken
to be a primitive condition, based on the assumption that the
primary yolk cells represent the entoderm. This retrograde
migration must then be regarded as a modified process of gas-
trulation like that found in arachnids.

A second characteristic which the yolk cells of the bee possess
in common with certain other insects is the large size of these
cells. It is certain that this is true of the Dermaptera and Orthop-
tera, since Heymons (1895) calls especial attention to it and it is
also true, according to Lecaillon (1898) for the chrysomelid
beetles Chrysomela, Lina and Agelastica, but not for Clytra. It
seems doubtful how far this relation obtains among insects in
general. Definite statements on this point are wanting and an

re es

Dye ee ae ee ee

THE EMBRYOLOGY OF THE HONEY BEE = 2ar

inspection of the plates given by various investigators has proved
inconclusive, although it seems probable that in most insects the
primary yolk cells are at least of more than average size.

The formation of “nests” of nuclei occurs in many insects and
has been especially mentioned by Heider (1889) in Hydrophilus
and by Heymons (1895) in the Dermaptera and Orthoptera, more
recently by O. Dickel (1904) and Nachtsheim (1913) in the
honey bee and by Friederichs (1906) in chrysomelid beetles.
Dickel and Nachtsheim state that they found mitotic figures in
the yolk cells (see p. 230), but both Heymons and Friederichs
found multiplication of nuclei to be due principally to amitotic
(direct) division.

The degeneration of the primary yolk cells at an early stage in
the development of the embryo seems to have been noted only by
Lecaillon (1897) and Friederichs (1906), in the chrysomelid
beetles. The latter investigator has made a special study of the
cell elements of the yolk and finds that (1) the primary yolk cells
multiply at first by mitosis, later by amitosis, (2) that many of
their nuclei soon degenerate, forming chromatin spherules which
find their way out of the cells themselves into the yolk. These he
terms “entodermal paracytoids” in order to distinguish them from
similar bodies formed by cells of the tissues of the embryo. As
previously stated these conditions are closely approximated in the
honey bee.

The yolk itself, in at least three of the large orders of insects,
the Orthoptera, the Coleoptera and the Lepidoptera, becomes
split up into polyhedral blocks or segments soon after the forma-
tion of the germ band. Each of these so-called segments con-
tains one or more yolk nuclei. In the examples of the Diptera
studied this process appears to be lacking, as it is also in the
mason bee (Carriére and Biirger 1897), as well as in the honey
bee. In the latter insect at least the absence of yolk segmentation
is probably due to the fluidity of the yolk, which would make such
a breaking up a physical impossibility.

The immigration of small cells or small nuclei from the blasto-
derm or the embryonal tissues into the yolk to form secondary
yolk cells has been described by a number of investigators. Hey-
mons (1895) and Friederichs (1906) have reviewed this subject

242 THE EMBRYOLOGY OF THE HONEY BEE

quite completely, so that a prolonged discussion would be out of
place here.

A. migration of cells from the germ band into the yolk was first
claimed by Graber (1871, 1878), and later (1888a) described
more completely in the beetle Melolontha. Similar cells were also
observed by Korotneff (1885) in Gryllotalpa, and Wheeler (1889)
in Leptinotarsa (Doryphora). Voeltzkow (1889) and Noack
(1901) have found that in the muscids also cells enter the yolk
from the blastoderm. According to Noack some of these sec-
ondary yolk cells form true yolk cells as in the bee, while others
degenerate. Mecznikow (1866) and Will (1888) mention the
migration of cells into the yolk from the posterior pole of the egg
in the aphids, but the relation of these to the secondary yolk cells
of other insects is not clear. In the Lepidoptera, Schwarze
(1899) states that in Lasiocampa cells detach themselves from
the anterior end of the mesoderm and wander off into the yolk.
Later they leave the yolk and enter the body cavity becoming blood
cells. Schwangart (1904), states that the greater part of the
anterior mesenteron rudiment becomes resolved into its component
cells ; these enter the yolk and become concerned in its absorption.
Later they contribute to the formation of the mesenteron. Hey-
mons (1895) has paid special attention to the secondary yolk cells
in Forficula, Gryllotalpa and Gryllus, and which he names
“nparacytes.” These cells are first formed from the lateral por-
tions of the blastoderm, but may later be seen arising from all
parts of the germ band, especially the mesoderm. The one un-
mistakable characteristic, distinguishing these cells from normal
embryonic cells consists, according to Heymons, in “the dis-
solution of the nucleus, and particularly in an extremely charac-
teristic separation of the chromatin from the other substances
contained in the nucleus, especially the nucleolus.” In other
insects true “paracytes” have apparently thus far been observed
only by Schwarze (1899) in Lasiocampa and by Friederichs
(1906) in Meloé. The relation of the paracytes to yolk cells of
secondary origin observed in other insects is at present uncertain.

Friederichs (1906) has made a especially intensive study of the
cell elements of the yolk in certain beetles. Some of his con-
clusions regarding the primary yolk cells have already been men-

THE EMBRYOLOGY OF THE HONEY BEE 243

tioned. In the chrysomelid beetles true “paracytes” were not
found. Instead, there are found issuing from the cells of the
germ band—ectoderm, mesoderm, sex cells—small globular bodies
formed by the emission of chromatin from certain of the
nuclei of the tissues concerned. These bodies Friederichs calls
“paracytoids.” After entering the yolk they disappear. In ap-
pearance they resemble so closely the chromatin spherules formed
by the degeneration of the nuclei of the (primary) yolk cells, that
they are not distinguishable from these, and the term “paracytoid”’
is made to cover chromatin spherules from both sources. Both
are morever supposed to be identical in function, which is pre-
sumably that of assisting in the digestion and assimilation of
the yolk. In Meloé typical “paracytes” were found; these later
degenerate in the yolk and form “paracytoids.”’

The aggregation of the yolk cells beneath the wall of the mid-
intestine in embryos of the honey bee was noticed by Kowalevski
in 1871, but this condition, as well as the occurrence of nuclei in

. the peripheral layer of protoplasm surrounding the yolk at earlier

stages, is found in a far more pronounced form in the mason bee,
Chalicodoma (Carriére and Biirger, 1897). Buirger’s descrip-
tion is as follows (p. 358):

“The yolk cells also participate in the formation of the epi-
thelium of the mid-intestine. Not directly indeed, in the sense
of affording material for its construction, but in that they doubt-
less afford nourishment to the entoderm bands.

“At the time of the appearance of the entoderm bands the yolk
cells at the periphery, regularly distributed and connected to-
gether, form a complete sac, or if you will, a primary mid-
intestinal epithelium.

“Tf the series of sections, which are intended to elucidate the
development of the wall bee, ceased just after the time of the
breaking through of the mouth, one would be inclined to consider
seriously whether the yolk cells did not form the definitive epi-
thelium of the mid-intestine. Possessing other stages, it is how-
ever evident to us, that the nuclei of the yolk cells degenerate,
forming a crumbling mass, which is absorbed together with the
remaining yolk. In the oldest embryo which I have sectioned
the intestine is empty of yolk and also of every trace of yolk
cells and their nuclei.”

244 THE EMBRYOLOGY OF THE HONEY BEE

This rather remarkable condition does not appear in other
insects, and may be peculiar to the Hymenoptera. It recalls the
condition found by Madame Tschuproff-Heymons (1899) in the
Odonata and suggests that possibly the behavior of the yolk cells in
the Hymenoptera may have a phylogenetic significance, the epi-
thelium-like layer formed by the yolk cells at the periphery of
the yolk representing the vestiges of the ancient miid-intestine,
which undoubtedly was primitively formed by the yolk cells. If
this were the case however, similar conditions should be found
in the Orthoptera and other primitive orders. In Gryllus and
Gryllotalpa, according to Heymons (1895) the yolk cells form
an epithelial layer beneath the wall of the mid-intestine, but this
does not occur until shortly prior to hatching. In Dixippus, ac-
cording to Hammerschmidt (1910) the yolk cells form an epithe-
lial layer over the ventral surface of the yolk at a relatively early
stage. This layer Hammerschmidt regards as the “primary
entoderm”, a view which is in harmony with the suggestion out-
lined above.

Pe ee ee ee

ers

' Se oe

Sse

rer ee

XV

DURATION AND RATE OF DEVELOPMENT

Early in the course of the present work, efforts were made to
determine the rate of development, that is, the length of time
required by the egg to reach a particular stage. Unexpected
difficulties were encountered, and the problem was found much
less easy of successful solution than was at first anticipated. This
was doubtless due to ignorance of the proper method, and much
time was consumed in unsuccessful experimentation. In general
two methods were tried. The first is a modification of those
employed by Petrunkewitsch (1902) and Dickel (1904). It con-
sists in removing a frame from the brood nest of a vigorous
colony, and replacing it by an empty frame, upon which the queen
is placed. The hive is then closed for a period of two or three
hours. The experiment was varied by using two or more empty
frames, examining them every two hours. This method, after
many trials, was abandoned, since it was seldom successful. The
failure of this series of experiments was probably due in part, at
least, to the fact that they were undertaken too late in the season,
after the main honey flow had ceased. Continued attempts how-
ever were made to remedy this unfavorable condition by feeding
thin sugar syrup by. means of a “pepper box”’ feeder placed above
the frames. Next a single frame observation hive was used, the
queen being watched continuously in the hope that some time dur-
ing the day the queen would deposit during a two-hour period a
number of eggs sufficient for the purpose in hand. This ex-
periment was a total failure since not once during the several days
she was under observation did she deposit more than a few eggs,
and these at long and infrequent intervals, practically all of the
eggs being laid during the night. It may be added that the ob-
servation hive was kept in a warm but shaded position out of
doors, in one of the large breeding cages used by the entomolo-
gists of the Department. The third set of experiments was mod-
erately successful. The method used in these was the following:

245

246 THE EMBRYOLOGY OF THE HONEY BEE

A strong colony was shaken on eight full frames of foundation
and two empty combs, which were placed in the center of the hive.
The two combs were after two hours usually found to contain
a fair number of eggs. These combs were then marked and
placed in the second story of another strong colony, above a
queen excluder, combs of brood being taken from the first
story and placed on each side of the combs containing eggs,
in order to make sure that these would not be deserted. After
the proper interval the eggs were removed from the comb,
fixed, and afterwards stained and mounted for examination.
This method is open to the serious objection of being very
demoralizing to the colony subjected to it, and cannot safely
be tried more than once on the same colony unless a con-
siderable interval elapses between trials.

The data in regard to the rate of development, obtained by the
foregoing methods is summarized in the accompanying table.
This shows the approximate time in hours in the left hand
column, the principal development changes observed at this
age in the middle column, and the corresponding illustration
in the right hand column.

Recorded data relating to this subject are scarce. The follow-
ing may be gleaned from Dickel’s paper (1904). At twenty hours
the blastoderm is completely formed, and its nuclei arranged in a
single layer (Fig. 1). According to the writer’s data the nuclei
do not become arranged in a single layer until a few hours later.
An egg twenty-four hours old-is shown in figure 2. In this the gap
between the dorsal and ventral blastoderm, located just dorsad of
the cephalic pole, is plainly shown, as is also the case in figure 3,
which is about twenty-six hours old. In the figure last mentioned
secondary yolk cells are evident, distinguishable by their lack of
cytoplasm. These two figures. of Dickel agree fairly well with
figure 16 of the present paper, representing a sagittal section
through an egg twenty-four to twenty-six hours old. Dickel’s
figure 10 represents an egg thirty-five hours old in which “the
three germ layers are differentiated.” It apparently corresponds
to Stage V or VI, although the cephalic fold of the amnion is
not represented. This would agree with the data in the table
given below.

=

7 cS
* Ss

Sn ee

—

THE EMBRYOLOGY OF THE HONEY BEE 247
TABLE SHOWING RATE OF DEVELOPMENT OF EMBRYO
AGE IN
HOURS DEVELOPMENTAL CHANGES OBSERVED FIGURES
1-6 Cleavage proceeds rapidly from one to many| I and II

cells.

8-10 Cleavage cells reach the surface of the egg. Ill
[Estimated]

14-16 Cleavage cells cover the entire surface of the
yolk. Appearance of primary yolk cells.

18-20 Nuclei of the cells of the blastoderm become
arranged in two layers. Appearance of sec-
ondary yolk cells.

20-30 Cells of blastoderm become arranged to form
a single layer of prismatic cells. Nuclei of
primary yolk cells degenerating.

32-34 Differentiation of middle plate begins. IV

34-36 Appearance of anterior mesenteron rudiment.

Cephalic fold of amnion formed. ve

36-38 The lateral folds and the posterior mesenteron
rudiment are formed.

38-42 Approximation of lateral folds. Caudal fold VI
of amnion formed.

42-44 Closure of lateral folds completed. Amnion VII
becomes a closed sac.

44-46 Formation of the stomodaeum and _ procto- VIII
daeum.

Appearance of the rudiments of the labrum,
antennae, gnathal appendages, silk gland and
stigmata.

48-50. Formation of stomodaeum. The invaginations IX
which are to form the silk gland and stig-
mata are still wide slits.

52-54 Formation of proctodaeum and Malpighiari x
tubules. Labrum, cerebral lobes and rudi-
ments of the mesodermal somites now well
defined. Formation of tracheal trunks and
their branches. Silk glands become elon-
gated tubules.

58-60 Approximation of second maxillae to form XII
labium. [Estimated.]

66-68 Second maxillae united to form labium. Heart XIII
formed. Development nearly complete.

74-76 Embryo breaks chorion and becomes a larva. XV

248 THE EMBRYOLOGY OF THE HONEY BEE

Petrunkewitsch (1902) gives the following table for drone
(male) eggs:

Age in hours Stage of development
5 a WERE ESTES Tree UNC: first cleavage spindle.
Bae Rea ome cleavage.
fm RE movement of the cleavage nuclei towards
the periphery.
HOO aks Back eats formation of the blastoderm.
8 PE he tie blastoderm.
BRCCGs cre Tictisbnieee es gastrulation.
SES Sly BAS AME end of gastrulation; rudiments of the

mesodermal tubes formed.

According to these data it appears that the development of the
drone egg is more rapid than that of the worker (female) egg, —
since “gastrulation,” that is, the formation of the mesoderm and
the mesenteron rudiments, commences at least five hours earlier.
Unfortunately the time required for total development is not
known.

Among other insects, the only data which are sufficiently com-
plete to be of value for comparison are those of Heider (1889)
for Hydrophilus. These are extremely full and explicit, and
include all of the more important phases of development of the
insect. They are arranged in the form of a table which covers
two and a half folio pages and therefore far too elaborate to be
reproduced here. It will nevertheless be profitable to compare
Hydrophilus and the honey bee, using for this purpose certain
more or less arbitrary periods into which the development may
conveniently be divided.

Kowalevski (1871) distinguishes three periods in the develop-
ment of Hydrophilus. They are as follows: The first extends
to the complete formation of the germ band and of the
embryonic envelopes. The second includes the formation of
appendages and the origin of the glandular layer of the intestine
(“Darmdriisenblatt”), up to the formation of the dorsal plate and ~
the rupture of the embryonic envelopes. The third includes the
complete differentiation of the embryo and the dorsal wall of the
mid-intestine. Heider (1889) interprets these periods as follows:

as a ee

ae es

ey ee

ha

THE EMBRYOLOGY OF THE HONEY BEE — 249

“In the first period are included: the formation of the embryo,
of the rudiments of the cephalic lobes and of the embryonic
envelopes up to their complete overgrowth of the embryo.
At the end of this period the completely segmented germ
band has extended its entire length over the surface of the
yolk. To the second period belongs: the invaginations con-
stituting the proctodaeum and stomodaeum, the formation of the
rudiments of the appendages and of the central nervous system,
and the tracheal invaginations. Into this period falls the sep-
aration of the entoderm from the mesoderm together with the first
rudiments of the mid-intestine, the appearance of the primary seg-
ments, the development of the definitive body cavity and the
segmentation of the yolk. The rupture of the embryonic mem-
branes and their withdrawal to the dorsal side of the egg, and an
invagination, introducing the formation of the dorsal organ, form
the limits between the second and third embryonic periods. .. .
No new organs are produced, this period serves principally for
hisiclogical differentiation.”

In dividing the development of the bee into three corresponding
periods, indicated in the table given above by double lines, it is
easy to determine the limit between the first and second periods,
which falls between Stage VII, forty-two to forty-four hours, and
VIII, forty-four to forty-six hours. The limit between the sec-
ond and third periods is not so easy to find, since there is in the
bee but a single embryonic membrane, which is not ruptured until
hatching and which forms.no “dorsal organ.” Moreover in the bee
egg there is no movement of the embryo corresponding to “revolu-
tion” or “blastokinesis’” (Wheeler, 1893). By referring to Hei-
der’s table however, it is seen that at the end of the second period,
which falls between the fourth and fifth days of development, the
tracheal invaginations, formed but a few hours previously, are still
cleftlike. During the fourth day moreover the invaginations of
the proctodaeum and stomodaeum are formed. In the bee there-
fore the limit between the second and third periods has been fixed
between Stages IX and X, corresponding respectively to the ages
of forty-eight to fifty hours and fifty-two to fifty-four hours,
which is probably sufficiently close for purposes of comparison.

In Hydrophilus the time required for complete development is

250 THE EMBRYOLOGY OF THE HONEY BEE

eleven days, or 264 hours, while that of the bee is only seventy-six
hours. The only way in which the length of time consumed by each
period in the two insects can be compared is by expressing it as a
fraction of the time required for total development. This is most
conveniently expressed by decimals. The time of each period
except the first, is calculated from the close of the preceding
period. The result is shown in the following table:

Period I I] Ill
Hydrophilus 0.272 0.091 0.636
A pis 0.579 0.078 0.342

The discrepancy between these figures is striking. In Hydro-
philus slightly over one-fourth of the time required for total devel-
opment is taken up by the first period; in the bee over one-half is
consumed. Period II consumes relatively but a trifle more time
in Hydrophilus than in the bee, while period III consumes nearly
twice as much. If the time required for the formation of the
blastoderm be similarly compared, the result is expressed by the
fractions 0.166 for Hydrophilus, and 0.39 for Apis. Expressed
in general terms, it may be said that as compared with Hydro-
philus, the earlier development of the bee is greatly delayed or
conversely, that the later development is greatly accelerated. So
little is known of the vital forces underlying development that
the reason for these differences can only be surmised. It seems
not improbable however that one factor is the relatively slight
degree of differentiation of the bee larva, as compared with that
of Hydrophilus. The larva of the latter insect is fitted to lead an
active existence, finding, catching and killing its prey, and at the
same time avoiding its enemies. It is therefore well equipped
with functional organs of locomotion and perception together with
a well developed central nervous system, most of which in the bee
larva are present only as rudiments, and do not become functional
to any considerable degree until after metamorphosis.

XVI

TECHNIQUE

Except when timed stages were desired, a frame of brood comb,
containing a considerable number of eggs was selected. The eggs
were then lifted out of the cells and transferred singly to the
fixing fluid. A small fine-pointed brush, slightly moistened,
was found most convenient for this purpose. Other inves-
tigators have made use of a bent needle, but with this instrument
there is considerable danger of damaging the rather delicate egg.
A considerable difference was noted between different lots of eggs
in regard to the ease with which the eggs could be removed. In
some lots a light touch with the brush was all that was needed
to loosen the egg from its attachment to the bottom of the cell, in
others the eggs were stuck so tight as to require considerable
manipulation to bring them out of the cells without damage.
Although no records were kept it seems not unlikely that this
difference in the adhesiveness of the eggs was traceable to a differ-
ence between the mother insects, the queens, some secreting more
adhesive, or else of a better quality, than others.

Five fixing fluids were used: Bouin’s picro-formol, one-half
to eighteen hours”®; acetic-alcohol (Carnoy’s second formula),
one-half to one hour; Gilson’s mixture, twenty-four hours; Pet-
runkewitsch’s modification of Gilson’s formula, twenty-four
hours; and Kleinenberg’s picro-sulphuric mixture, three hours.
The last named was used but little. All of the others gave many
satisfactory preparations and some superlatively good ones. In
quality of the fixation, as regards the tissues of the embryo, there
is little room for choice. The effect of the different fixing fluids
on the yolk is however quite different. This has been described in
the chapter on the organization of the egg (p. 12). The effect
on the cells in the yolk is likewise different, eggs fixed with picro-
formol showing these to the best advantage. The question as to

The best material came from lots of eggs left in this for five to six

hours or longer.
251

252 THE EMBRYOLOGY OF THE HONEY BEE

which fixing fluid gives the truest pictures of these cells can how-
ever only be determined by careful comparison with living eggs.

Picro-formal proved to be the best medium for fixing eggs in-
tended to be studied entire, and since the other media showed no
decided advantage in other respects, it was most frequently em-
ployed. Its reason for superiority appears to be due to the
circumstance that it does not coagulate the yolk spheres, conse-
quently the yolk remains clear and transparent. Embryos fixed
in this fluid and mounted entire have a brilliancy and transparency
not seen in specimens fixed in other fluids.

The eggs were in all cases preserved in eighty per cent alcohol.

Of the five fixing fluids used, Gilson’s and Petrunkewitsch’s
fluids harden most, eggs fixed in these media becoming rather
brittle, while acetic alcohol hardens the least. Eggs treated with
this mixture are quite elastic and may be manipulated without
much danger of breaking, although they are rather easily distort-
ed. This is true also of eggs treated with picro-formol, although
not to the same extent.

The chorion of the egg of the bee, although excessively thin, is
nevertheless relatively impermeable, especially to aqueous fluids.
After fixing and hardening, and before staining or clearing is
attempted it is therefore necessary to puncture the chorion, which
lies close to the egg except at its two poles. The chorion may be
punctured or ruptured here, but the most satisfactory solution of
the difficulty is to puncture the egg in one or two different places
with a fine cambric needle, rubbed on an oil stone to an extremely
fine point. This is best done by watching the progress of the
sharpening under a binocular microscope. The punctures were
usually made along the dorsal mid-line of the egg since this region
is not occupied by the embryo until near the close of development.

Mayer’s carmalum was used for staining eggs intended to be
mounted entire. The eggs were stained from two to forty hours,
and thoroughly destained in acid alcohol. This method yielded
uniformly good results. Iron haematoxylin was found most sat-_
isfactory for sections, used with or without a counter stain.
Mayer’s paracarmine (alcoholic) was used for staining eggs pre-
liminary to embedding.

It is absolutely necessary that clearing be accomplished grad-

oe gh a ee ig ee en

THE EMBRYOLOGY OF THE HONEY BEE 253

ually. If this is not done, distortion will almost inevitably result.
In cases where the chorion is not opened, or opened insufficiently,
a transfer from alcohol to clearing fluid will cause the egg to be
completely flattened by osmotic pressure, owing to the slight re-
sistance of the yolk. Clearing may be quickly and easily accom-
plished by the following method of flotation, suggested by Dr.
Petrunkewitsch: A small quantity of pure cedar oil is poured
into the bottom of a narrow vial. On the top of this are poured
successive layers containing an increasing proportion of absolute
alcohol. In practice these were: pure oil, two-thirds oil, one-third
oil. On the top layer the embryos were placed surrounded by
pure absolute alcohol. They then descend gradually to the bottom
of the vial, when they may be taken out, washed in pure cedar
oil and transferred to balsam, if they are to be mounted for study
entire. If intended for sections they were placed in xylol before
transfer to paraffin. Embedding was done in watch crystals.
Paraffin of fifty-five degrees C. melting point was used, the eggs
remaining in the paraffin from one-half to an hour to three hours:
It is extremely desirable to orient the embryo accurately. This
was accomplished by placing the watch crystal containing the
melted paraffin and the embryo on a piece of clear glass, slightly
warmed, and placing the whole on the stage of a binocular micro-
scope. A layer of partially cooled paraffin, sufficient to support
the embryo in any position soon forms at the bottom of the watch
crystal, and the embryo may be placed in this in the position
desired. In practice the ventral face was placed uppermost. The
watch crystal was then floated on a vessel full of cold water, and
the paraffin subsequently hardened by complete immersion.
Sections were cut by means of a Minot microtome, five to seven
microns in thickness, according as the conditions were favorable

or otherwise.
SUMMARY

The elongate cylindriform egg is slightly curved in the long
axis and, with reference to the future embryo, presents a slightly
larger cephalic end, and a convex ventral, as opposed to a con-
cave dorsal surface. Two membranes cover the egg, an outer,
the chorion, and an inner, the vitelline membrane. The chorion is
extremely thin, transparent, and covered with minute papillae

254 THE EMBRYOLOGY OF THE HONEY BEE

arranged in the form of a fine network. A thickened area at the
cephalic end of the egg appears to represent a micropylar area.
The vitelline membrane is extremely thin and appears to be
structureless. The contents of the egg comprise a large quantity
of deutoplasmic material or yolk, and a small quantity of proto-
plasm. The yolk is principally in the form of spherical globules,
the vitelline spheres, composed of a transparent fluid of an un-
known chemical nature, not fat nor oil. Within the yolk spheres
are much smaller relatively solid bodies, the vitelline bodies. The
protoplasm surrounds the yolk spheres, filling the’ interstices be-
tween them, and also forms a thin cortical layer over the surface
of the yolk. Within the protoplasm are numerous minute bodies,
possibly identical with the Blochmann’s corpuscles of certain other
insects.

The cleavage cells at first form a rounded group near the
cephalic pole of the egg. They multiply rapidly and soon form
an elongated hollow fusiform figure, its smaller end extending
toward the caudal pole. As the cells continue to increase in
number the figure also increases in size, until the cells at its larger
end, on the ventral side near the cephalic pole, enter the cortical
layer and begin to form blastoderm. The majority of the cleavage
cells finally reach the cortical layer except a few which remain in
the yolk to form the primary yolk cells.

As the cleavage cells approach the periphery of the egg their
nuclei assume a peripheral position in the cells. The latter embed
themselves in the cortical layer, the latter, together with the
cleavage cells, constituting the blastoderm. The central ends of
the cleavage cells, however, remain at first united together below
the cortical layer thus forming the inner cortical layer. Later
this inner cortical layer is cut off from the remainder of the
cleavage cells by a structure having the appearance of a basement
membrane. Meanwhile the cells of the newly formed blastoderm
multiply by mitotic division, the division planes being at first nor-
mal to the surface of the egg, but later become oblique, the cells
thus becoming wedge-shaped or pyriform. Finally they assume a
prismatic form, the nuclei all coming to lie at the same level;
the basement membrane disappears, and the greater part of the
inner cortical layer is absorbed by the cells of the blastoderm,

THE EMBRYOLOGY OF THE HONEY BEE 255

their inner ends becoming sharply defined. The remnant of the
inner cortical layer clings to the yolk over which it now fotms a
pellicle, this being separated from the blastoderm by a narrow
space. The blastoderm is at first of nearly uniform thickness
around the circumference of the egg, but soon becomes differen-
tiated into a thicker ventral and a thinner dorsal portion. Along
the dorsal mid-line is a strip of cells differing from the remainder
in being especially thin and flat and in maintaining a close relation
with the yolk.

The mesoderm is formed from a median area of the ventral
blastoderm, the middle plate, which separates from the blasto-
derm on each side of it and which constitutes the lateral plates.
The ventral plate is then covered over by the lateral plates, which
approach one another, and finally become united along the ventral
mid-line to form the ectoderm. The rudiments of the mesenteron
are formed by the immigration of blastoderm cells, a discoid
swelling being thus produced at each of the two ends of the
middle plate, but outside of its limits. These rudiments later
become covered by ectoderm. The anterior mesenteron rudiment
does not however become completely covered, a small circular
area near its caudal margin remaining uncovered and later con-
stituting the floor of the stomodaeal invagination. During these
developmental changes the median dorsal area of the blastoderm,
composed of thin flat cells, also becomes depressed and is over-
grown by the dorsal margins of the lateral plates. Meanwhile
the cells of this dorsal strip have becme aggregated in the cephalo-
dorsal region of the egg to form a more or less discoid mass, the
cephalo-dorsal body, the so-called “yolk plug” of O. Dickel
(1904). During the formation of the so-called “germ-layers”
both the middle and lateral plates show plain evidence of segmen-
tation. The segments thus indicated appear to correspond to the
definite segments of the embryo.

The amnion is single layered and is formed from the dorsal
half of the blastoderm. This separates both from the yolk and
from the margins of the ventral or embryonic half of the blas-
toderm, the ventral plate. This separation does not take place
simultaneously, but occurs first at the cephalic end of the egg, a
cap-like fold being formed which grows rapidly caudad. A sim-

256 THE EMBRYOLOGY OF THE HONEY BEE

ilar but slighter fold is later formed at the caudal end of the egg.
The two folds meet and coalesce near the caudal pole. The am-
nion probably corresponds to the serosa of other pterygote insects.

Twenty-one segments were found in the bee embryo, including
an anal segment or telson. Appendages were observed on the
antennal segment, the three gnathal segments, and the three tho-
racic segments. No abdominal appendages were found. The
antennal rudiments and those of the three thoracic legs became
reduced to hypodermal thickenings prior to hatching. The sup-
posed appendages of the premandibular segment are only exag-
gerated ganglionic swellings. No evidence of the presence in the
bee of a superlingual segment (Folsom) was found. As is
usual in insect embryos, the second maxillae fuse to form the
labrum.

The rudiments of the nervous system take the form externally
of two longitudinal swellings of the ectoderm, the primitive swell-
ings, one on each side of and close to the ventral mid-line. At the
oral region these diverge, and in the cephalic region of the embryo
expand into the broad procephalic lobes. The primitive swellings,
including the procerebral lobes, are divided by slight interseg-
mental constrictions into twenty neuromeres corresponding to the
segments of the embryo. Two of these, formed from the
procerebral lobes, are preoral. The first, the protocerebrum,
corresponds to the primary head segment, the second, the deuto-
cerebrum, to the antennal segment. The two lobes constituting
the protocerebrum become rather obscurely subdivided into three
lobes, the most anterior of which become the optic lobes. During
the formation of the primitive swellings the germ band lengthens,
so that its cephalic end, including the procerebral lobes, becomes
curved around the cephalic pole of the egg. The protocerebral
lobes thus come to lie on the dorsal side of the egg with their
cephalic ends directed toward its caudal pole. Although the germ
band shortens somewhat at a later stage, the protocerebral lobes
always remain directed caudad, and not cephalad, as in more prim-
itive insects (e.g. Orthopetra). The third, neuromere, correspond- —
ing to the premandibular or intercalary segment, is postoral, since
the commissure connecting its two lateral halves (ganglia) passes
below the stomadaeum. These three neuromeres become united
to form the supraoesophageal ganglion or brain. The neuromeres

THE EMBRYOLOGY OF THE HONEY BEE 257

of the three gnathal segments unite to form the suboesophageal
ganglion. The remainder, fourteen in number, become the gan-
glia of the ventral nerve cord. The cells of the ectoderm con-
stituting the primitive swellings and protocerebral lobes become
segregated into two layers, an outer layer of small cells, the
dermatoblasts, destined to form hypodermis only, and an inner
layer of large cells, destined to form nerve tissue, the neuroblasts.
These divide unequally and teleblastically, giving of centrad sev-
eral cells smaller than themselves. These last mentioned cells
divide equally, the products becoming differentiated to form
ganglion cells. The latter are therefore the granddaughter cells
of the neuroblasts. The optic gangliz are not however produced
by the agency of neuroblasts, but are formed as simple infoldings
of the ectoderm. Beginning at the anterior margin of the man-
dibular segment, and extending to the last segment of the trunk
is a narrow median strip of ectoderm, the median cord. In the
intrasegmental regions this contributes the central portions of
the ganglia, in the intersegmental regions, it constitutes a series
of thickenings of the hypodermis. The supraoesophageal com-
missure is formed at least in part from the median ectoderm, but
this could not be established in the case of the suboesophageal
commissure. The stomatogastric nervous system is formed from
three median evaginations of the dorsal stomodaeal wall. The
first of these furnishes a few cells contributing to the formation
of the labral nerve, the second produces the stomatogastric gan-
glion, and the third the pharangeal ganglia. An outer neurilemma
only is present in bee embryos. This is formed from cells having
the same origin as the ganglion cells, and which migrate to the
external surface of-the brain and ventral cord. The so-called
“ganglia” or corpora allata are formed from invaginations of the
ectoderm between the bases of the rudiments of the mandibles and
first maxillae, and are closely associated with the invaginations
which produce the apodemes of the retractor muscles of the
mandibles. The invaginations producing the corpora allata soon
become solid masses of cells, lose their connection with the ex-
ternal ectoderm and migrate mesiad and dorsad, finally becoming
attached to the ventro-lateral angles of the coelomic sacs of the
antennal segment. Degenerating cells are found in the rudiments

258 THE EMBRYOLOGY OF THE HONEY BEE

of the brain of all embryos, being especially abundant in the
region between the second and third lobes of the protocerebrum.
They appear soon after the brain begins to form and are evident
up to the time of hatching of the embryo. Their significance is
not known.

The tracheal system is formed from eleven pairs of invagina-
tions of the lateral ectoderm. The first of these, situated on the
second maxillary segment, by the formation of four diverticula,
produces the anterior ends of the main tracheal trunks, including
the anterior tracheal commissure or loop, and also the tracheae
supplying the head. The ten pairs remaining are situated on the
second and third thoracic, and the first eight abdominal segments.
These also form each four diverticula. The anterior and pos-
terior diverticula become united along each side of the embryo,
thus forming the longitudinal tracheal trunks, the ventral diver-
ticula fuse with those of the opposite side of the same segments
to form the tracheal commissures, the dorsal diverticula form
branches supplying the dorsal region of the larva. The openings
of the tracheal invaginations remain as the spiracles.

The tentorium is formed from two pairs of ectodermal in-
vaginations. The first pair of these is situated in front of the
bases of the mandibles, the second behind the bases of the first
maxillae. The invaginations belonging to the first pair grow
caudad and mesiad, those belonging to the second pair cephalad
and mesiad. All four meet in the median plane to form a struc-
ture having the form of an X, extending across the head capsule
between the oesophagus and the suboesophageal ganglion. An
invagination situated immediately caudad of the base of the rudi-
ments of the mandible produces the apodeme for the adductor
muscle of the mandible.

The silk glands appear at an early period as tubular invagi-
nations of the ectoderm caudad and mesiad of the bases of the
second maxillae. These invaginations rapidly lengthen to form
long slender tubes extending the length of the trunk. As the
second maxillae approach one another, during the later stages of
development, they carry with them the openings of the silk glands,
which are thus brought to the ventral mid-line. The second
maxillae are then united in such a way that a median unpaired

THE EMBRYOLOGY OF THE HONEY BEE 259

duct is formed by them for the silk glands, opening near the tip
of the labrum.

The oenocytes are produced by immigration of cells from local-
ized areas of the lateral ectoderm. There are eight pairs of these
situated on the first eight abdominal segments, in line with the
openings of the tracheal invaginations.

The mesoderm, soon after its formation, becomes differentiated
laterally into two layers, an outer somatic and an inner splanch-
nic layer, while along the ventral mid-line it remains single lay-
ered. Separate coelomic sacs are not present, the somatic and
splanchnic layers of each side being continuous longitudinally
throughout the trunk. At the lateral margins of the mesoderm
the two layers are continuous and in this region are composed of
long columnar cells. The fate of the various parts of the
mesoderm is as follows: The median single-layered section
breaks up into the rounded blood cells. The somatic layer forms
the trunk muscles, both longitudinal and oblique, the pericardial
fat cells, and the dorsal diaphragm, including the pericardial cells.
The splanchnic layer sends off from its dorsal border a mesial
layer which forms the muscular layer of the mid-intestine. The
remainder of this layer is principally concerned in the formation
of the two main divisions of the fat body. The heart is formed
by the union, along the dorsal mid-line, of two rows of cells, the
cardioblasts, which are derived from the angle formed by the
union of the somatic and splanchnic layers at the lateral margins
of the mesoderm. A mass of mesoderm cells, forming the
anterior end of the mesoderm, and evidently belonging to the
primary head segment, closely surrounds the stomodaeum at its
appearance, and later forms the muscular layer of the fore in*
testine. A similar mass at the posterior end of the embryo forms
the muscular layer of the hind-intestine.

The ovaries are apparently derived from the cells constituting
the genital ridges; these cells are not visually distinguishable from
undifferentiated mesoderm cells. The two genital ridges are
formed from the dorsal portion of the splanchnic layer, in the
fifth to the tenth abdominal segments inclusive. This portion
becomes detached from the remainder of the splanchnic layer.
During the development of the embryo the genital ridge gradually

260 THE EMBRYOLOGY OF THE HONEY BEE

shortens, finally occupying a position in the seventh to the ninth
segments inclusive. Meanwhile it loses its attachment to the
dorsal splanchnic layer, at the same time receiving an investment
of cells from the splanchnic mesoderm lying immediately ventrad
of it. This investiture, composed of flat cells then, contracts an
adhesion with the ventral border of the heart.

The mid-intestine is formed from the anterior and posterior
mesenteron rudiments in the following manner. The discoid
anterior mesenteron rudiment becomes transferred from the ven-
tral to the dorsal side of the cephalic pole of the egg by a lengthen-
ing of the embryo. At the same time the rudiment increases in
superficial area, covering the cephalic end of the yolk like a cap.
Its caudal margin now extends rapidly caudad over the dorsal
surface of the yolk. Meanwhile the posterior mesenteron rudi-
ment has been similarly transferred to the dorsal side of the
caudal end of the egg. It now sends out a thin tongue-like pro-
cess cephalad over the dorsal surface of the yolk. The caudad
extension of the anterior mesenteron rudiment, and the cephalad
extension of the posterior mesenteron rudiment next meet on the
dorsal surface of the yolk about one-third of the length of the
egg from its cephalic pole. The epithelial strip thus formed
extends rapidly ventrad over the sides of the yolk until the latter
is completely enclosed, the two margins of the epithelium meeting
and uniting along the ventral mid-line of the yolk a short time
before hatching. Both the fore and hind intestines are formed,
as is usual, by ectodermal invaginations. The invagination for
the fore intestine, the stomadaeum, is however, not completely
ectodermal, since its floor is formed by cells belonging to the
anterior mesenteron rudiment, which is not covered by ectoderm.
The hind intestine is exclusively ectodermal. The lumen of the
stomodaeum (oesophagus) becomes connected with that of the
mid-intestine shortly before hatching. A proventricular valve
is also formed at this time by folding of the stomodaeal or oesoph-
ageal wall. The lumen of the proctodaeum (hind-intestine) is at
no time in connection with that of the mid-intestine, both the
cephalic end of the hind intestine and the caudal end of the mid-
intestine being blind. The four Malpighian tubules are formed
as ectodermal invaginations which make their appearance prior to

THE EMBRYOLOGY OF THE HONEY BEE 261

the formation of the protodaeum, grouped around the point
where the proctodaeum is to appear. Four separate invaginations
have not been observed, the pair situated on each side of the mid-
line being connected by a shallow crescentic groove.

The primary yolk cells, which are derived from cleavage cells
remaining within the yolk, multiply by mitosis, the mitotic figures
being at first similar to those of the cleavage cells. A little later
irregular mitotic figures are found. These are usually minute,
and have the appearance of being in some cases multipolar,
and in others unequal. Multinucleate cells soon become abundant,
and some of these cells are of large size. Degeneration of the
nuclei of the yolk cells soon becomes frequent, such nuclei dimin-
ishing in size and finally becoming reduced to minute deep-staining
spherules which leave the cell body and enter the yolk. Sec-
ondary yolk cells are formed by the immigration of nuclei from
the blastoderm into the yolk. These soon become indistinguish-
able from the primary yolk cells. Yolk cells are found distributed
through the yolk until shortly before hatching. They are fre-
quently seen clustered under the epithelium of the mid-intestine,
during the time when the latter is engaged in covering the lateral
faces of the yolk. Yolk cells and yolk disintegrate at the time of
hatching, being presumably digested.

The total time normally required for the development of the
egg is seventy-six hours. This is divided approximately as fol-
lows: Cleavage, fourteen to sixteen hours; formation of the
blastoderm, fourteen to sixteen hours; formation of mesoderm,
rudiments of mesenteron and embryonic envelope, twelve to four-
teen hours; remainder of development, including differentiation of
tissues and organs, thirty-two or thirty-four hours. The earlier
stages, including the formation of the so-called ‘germ layers,” the
amnion, etc., occupy considerably over one-half of the time re-
quired for total development.

LIST OF ABBREVIATIONS

In making up this list, the abbreviations used by Snodgrass
(1910) were retained to indicate the corresponding part or its
rudiment in the embryo. In every case the symbols were designed

to suggest the name of the part, in order to avoid frequent refer-
ence to the subjoined list.

Am, amnion.
Iam, cephalic fold of amnion.
2am, caudal fold of amnion.
Ant, antenna.
AMR, anterior mesenteron rudiment.
AntL, antennal lobe (of brain).
AntMeso, mesoderm of antennal somite.
AntNu, antennal nerve.
AntR, antennal rudiment.
Ao, aorta.
ATraL, anterior tracheal loop.
BI, blastoderm.
BIC, blood corpuscles.
1Br, protocerebrum.
1Br,, 1Br., tBrs, the three lobes of the protocerebrum.
2Br, deutocerebrum.
3Br, tritocerebrum.
CB, cephalo-dorsal body.
Cbl, cardioblasts.
CC, cleavage cells.
Com, commissure.
Con, connective (of ventral nerve cord).
CorAll, corpora (ganglia) allata.
CL, cortical protoplasmic layer.
Cil, cuticle.
Dbl, dermatoblasts.
DDph, dorsal diaphragm.
DegCl, degenerating cells.
DLMcl, dorsal longitudinal muscles.
DphCl, cells of dorsal diaphragm.
Ds, dorsal strip (of blastoderm).
Ect, ectoderm.
FInt, fore intestine (oesophagus).
FiCom, frontal commissure.
262

THE EMBRYOLOGY OF THE HONEY BEE

FtGng, frontal ganglion,
FtNv, frontal nerve.

Gl, gland.

GngC, ganglion cells.
Ht, heart.

HiCl, pericardial cells.
HInt, hind intestine.
Hyp, hypodermis.

ICL, inner cortical layer.
IL, 2L, 3L, legs (rudiments).
Lb, labium.

LC, lateral cords.

LF, lateral folds.

Lm, labrum.

LWNz, lateral nerve.

LP, lateral plate.

LTraT, longitundinal tracheal trunk.
M, micropylar area.

Mal, Malpighian tubules.

MC, median cord.

MclEnt, muscular layer of alimentary canal.
Md, mandible.

MdNv, mandibular nerve.

Meso, mesoderm.

Meso 1-3L, mesoderm of leg rudiments.
MInt, mid-intestine.

MP, middle plate.

Mth, mouth.

1Mx, first maxilla.

2M x, second maxilla.

Nol, neuroblasts.

NIG, neural groove,

Nim, neurolemma.

NIR, neural ridges.

Nv, nerve.

NvF, nerve fibres, “punksubstanz.”
Oe, oesophagus, fore intestine.

OeCom, circumoesophageal commmissure.
Oen, oenocytes.

OMcl, oblique trunk muscles.

263

OpL, optic lobe, equivalent to the first lobe of the protocerebrum 1Br,.

OpPI, optic plate. ;
Ost, ostium.

Ov, ovary (rudiment).

ParC, paracardial cells.

_ Phy, pharynx.

204 THE EMBRYOLOGY OF THE HONEY BEE

PhyGng, pharyngeal ganglia.

PMR, posterior rudiment of mid-intestine.
PP, polar protoplasm.

Pro, Proc, proctodaeum.

ProL, procephalic lobes.

PriSw, primitive swellings.

PTraL, posterior tracheal loop.

RAp, apodeme of flexor (adductor) muscle of mandible.
RMcl, flexor (adductor) muscle of mandible.
SIkD, common duct of silk glands.

SlkDo, opening of duct of silk glands.
SIkGI, silk gland.

SoeCom, suboesophageal commissure.
SoeGl, suboesophageal ganglion.

Sp, spiracle.

SpBr, spiracular branch.

StgNv, stomatogastric nerve.

StgR, rudiment of stomatogastric nervous system.
Sto, Stom, stomadeum.

SupCom, supraoesophageal commissure.

Tae, taenidia.

Ten, central body of tentorium.

1Ten, 2Ten, anterior and posterior arms of tentorium.
Tra, trachea.

TraBr, tracheal branch.

TraCom, tracheal commissure.

Tralnv, tracheal invagination.

TraTr, tracheal trunk.

VDph, ventral diaphragm.

VLMcl, ventral longitudinal muscles.

VNC, ventral nerve cord.

x, posterior tier of cells of optic ganglion.
Y, yolk.

YC, yolk cells.
y, subspherical group of cells in lateral region of deutocerebrum.

BIBLIOGRAPHY

Anctas, J. 1901. Observations sur les métamorphoses internes de la guepe
et de l’abeille. Bul. Sci. France et Belg. XXXIV, pp. 363-473. Planches
XIX a XXIII et 8 fig. dans le texte.

Ayers, Howarp. 1884. On the development of Oecanthus niveus and its
or Teleas. Mem. Boston Soc. Nat. Hist. III, pp. 225-281, Plates
18-25. :

BALBIANI, E. G. 1882. Sur la signification des cellules polaires des in-
sectes. C.R. Acad. Sci. [Paris] XCV, pp. 927-920.

1885. Contribution a l’étude de la formation des organes sex-
uels chez les insectes. Rec. Zool. Suisse, II. pp. 527-588, Planches XVI
bis et XVII.

Baxrour, Francis M., M.A., F.R.S. 1880-1881. A treatise on comparative
embryology, London, Vol. 1, 1880, Vol. II, 1881.

BLocCHMANN, F. 1884. Ueber die Reifung der Eier bei Ameisen und
Wespen. Festschrift zur Feier des Fiinfhundertjahrigen Bestehens
der Ruperto-Carola, dargebracht von dem Naturhistorisch A Medicin-
ischen Verein zu Heidelburg, pp. 143-172. Mit einem Doppeltafel.

1886. Ueber eine Metamorphose der Kerne in den Ovarialeirn
und tber den Beginn der Blastodermbildung bei den Ameisen. Ver-
handl. Naturhist-Med. Ver. Heidelberg. pp. 243-247.

1887. Uber das regelmassige Vorkommen von bacterienahn-
lichen Gebilden in den Geweben und den Eiern verschiedener Insecten.
Ztschr. Biol. XXIV, pp. 1-15. Mit Tafel 1.

1887a. Ueber .die Richtungskoérper bei Insekteneiern. Biol.
Centbl. VII, pp. ro8-111.

1889. Ueber die Zahl der Richtungsk6rper bei befruchteten und
unbefruchteten Bieneneiern. Morph. Jahrb. XV, 1, pp. 85-96, mit
Tafel V.

1892. Ueber das Vorkommen von bakterienahnlichen Gebilden in
den Geweben und Eiern verschiedener Insekten. Centbl. Bakt. Parasit.
XI, pp. 234-240.

Bopretzky, N. 1878. Ueber die Bildung der Blastoderms und der Keim-
blatter bei den Insecten. Ztschr. Wiss. Zool. pp. 195-215, mit Tafel XIV.

Bovert, TH. 1887. Ueber Differenzierung der Zellkerne wahrend der
Furchung des Eies von Ascaris meg. Anat. Anz. II, 22, pp. 688-693.

BraAnpt, ALEXANDER JUN. 1869. Beitrage zur Entwicklungsgeschichte der
Libelluliden und Hemipteren, mit besonderer Beriicksichtigung der
Embryonalhiille derselben. Mém. Acad. Imp. St. Petersb, VII® sér.,
XIII, 1. pp. 1-33. Mit 3 Tafeln.

Bruce, A.T. 1887. Observations on the embryology of insects and arach-
nids. Baltimore. 31 pp. 6 plates.

Bitrscuur, O. 1870. Zur Entwicklungsgeschichte der Biene. Ztschr.
Wiss. Zool. XX, 4, pp. 510-564. Mit Tafel XXIV-XXVII.

1888. Bemerkungen iiber die Entwicklungsgeschichte von Musca.
Morph. Jahrb. XIV. 1, pp. 170-174, 3 Abbildungen im Text. _

Carriere, Justus. 1800. Die Entwicklung der Mauerbiene (Chalicodoma
muraria Fabr.) im Ei. Arch. Mikros. Anat. XXXV, 4, pp. 141-165.
Tafel VIII.

und Bircer, Orto. 1897. Die Entwicklungsgeschichte der
Mauerbiene (Chalicodoma muraria Fahr.) im Ei. Abhandl. K. Leo-

365

266 THE EMBRYOLOGY OF THE HONEY BEE

Se Bali Deut. Akad. Naturf. LXIX, 2, pp. 253-420. Mit 13

afel.

CuHoLopkovsky, N. A. 1888. Ueber die Bildung des Entoderms bei
Blatta germanica. Zool. Anz. XI, 275, pp. 163-166. 2 Figg. im Text.

——— 1889. Studien zur Entwicklungsgeschichte der Insecten.
Ztschr. Wiss. Zool. XLVIII, 1, pp. 89-100, Mit Tafel VIII.

1891. Ueber die Entwicklung des centralen Nervensystems bei
Blatta germanica. Zool. Anz. XIV, pp. 115-116.

1891a. Ueber einige Formen des Blastopors bei mesoblastischen
Eiern. Zool. Anz. XIV, pp. 159-160.

1891b. Zur Embryologie der Insecten. Zool, Anz. XIV, pp. 465-
470.

1891c. Die Embryonalentwicklung von Phyllodromia (Blatta)
germanica. Mem. Acad. Imp. Sci. St. Petersb. XXXVIII, 5, pp. 1-
120. Mit 6 Tafeln.

CLAYPOLE, AGNES Mary. 1808. The embryology and odgenesis of
Anurida marituma (Guér.) Jour. Morph. XIV, 2, pp. 219-300, pls.
XX-XXV.

Comstock, JOHN Henry, and Kocui, CHujrro. 1902. The skeleton of the
head of insects. Amer. Nat. XXXVI, 421, pp. 13-45. Figs. 1-29 in the
text.

ConkKLIN, Epwin Grant. 1897. The Embryology of Crepidula. A contri-
bution to the cell lineage and early development of some marine
gasteropods. Journ. Morph. XIII, 1, pp. 1-205. With 13 text figures
and 9 plates.

Czersk1, S. 1910. Die Entwicklelung der Mitteldarmanlage bei Me/oé
violaceus Marsch. Arch. Polon. Biol. Med. II. pp. 259-284. Taf. VIII.

DEEGENER, PAuL. 1900. Entwicklung der Mundwerkzeuge und des Darm-
kanals von Hydrophilus. Ztschr. Wiss. Zool., LXVIII, 1, pp. 113-
168. Mit Tafel VIII-X.

DicKEL, Orro. 1904. Entwicklungsgeschichtliche Studien am Bienenei.
Ztschr. Wiss. Zool. LXXVII, 3, pp. 481-527. Mit Tafel XIX und XX
und 46 Figuren im Text.

DourRN, ANTON. 1866. Zur Embryologie der Arthropoden. Centbl. Med.
Wiss., LIV, pp. 849-851. Berlin.

1876. Notizen zur Kenntniss der Insectenentwicklung. Ztschr.
Wiss. Zool. XXVI, 1, pp. 112-138.

EscHERIScH, K. 1900. Ueber die Bildung der Keimblatter bei den Musci-
den. Abhandl. K. Leopold-Carolin. Deut. Akad. Naturf. LXXVII,
4, pp. 303-364. Mit 3 Doppeltafeln und 10 Figuren im Text.

— 1900a. Uber die Keimblatterbildung bei Musciden. Verhandl.
Deut. Zool. Gesell. pp. 130-134.

——— 1901. Das Insekten-Entoderm. Ein Beitrag zur Keimblatter-
lehre. Biol. Centbl. XXI. pp. 416-431.

1902. Ueber den sogen. “Mittelstrang” der Insekten. Biol.
Centbl. XXII, 6, pp. 179-181. Mit 4 Figuren in Text.

1902a. Zur Entwicklung des Nervensystem der Musciden, mit
besonderer Beriicksichtigung des sog. Mittelstanges. Ztschr. Wiss.
Zool. LXXI, 4, pp. 525-549, mit Tafel XXX.

Fotsom, J. W. 1900. The development of the mouth-parts of Anurida_
maritima Guér. Bul. Mus. Harvard Col. XXXVI, 5, pp. 87-157, Pls.
I-VIII.

Friepertcus, Karr. 1906. Untersuchungen iiber die Entstehung der
Keimblatter und Bildung des Mitteldarns bei Kafern. Nova Acta.
Abhandl. K. Leopold-Carolin. Deut. Akad. Naturf. LXXXV. Nr. 3,
pp. [265]-382. Mit 7 Tafeln Nr. XXV-XXXI und 26 Fguren im Text.

THE EMBRYOLOGY OF THE HONEY BEE 267

Ganin, M. 1869. Ueber die Embryonalhiille der Hymenopteren-und
Fee SeCemtyONEN. Mém. Acad. St. Pétersb. (7). XIV. 5, Mit
1 Tafel.

1869a. Beitrage zur Erkenntniss der Entwicklungsgeschichte bei
den Insecten. Ztschr. Wiss. Zool. XIX, 3, pp. 381-451, mit Tafel
XXX-XXXITI.

1874. Ueber das Darmdriisenblatt der Arthropoden. War-
schauer Universitatsberichte.

GLAsER, R. W. 1912. A contribution to our knowledge of the function of
the oenocytes of insects. Biol. Bull. XXIII, 4, pp. 213-224.

Graper, VEIT. 1871. Vergleichende Lebens-und Entwicklungsgeschichte
der Insecten. Naturkrafte, XXII, Miinchen bei Oldenburg.

1878. Vorlaufige Ergebnisse einer grésseren Arbeit uber ver-
gleichende Embryologie der Insekten. Arch. Mikros. Anat. XV. 4, pp.
630-640.

—. 1888. Ueber die primare Segmentirung des Keimstreifs
der Insekten. Morph. Jahrb. XIV, pp. 345-368, Mit Tafel XIV-XV
und 4 Holzschnitten.

1888a. Vergleichende Studien itber die Keimhillen und die
Ruckenbildung der Insekten. Denkschr. Math-Naturwiss. Kl. Kais.
Akad. Wiss. Wien. LV, 2. pp. 109-162. Mit 8 Tafeln and 32 Text-
figuren.

1889. Vergleichende Studien tiber die Embryologie der Insek-
ten und insbesondere der Musciden. Denkschr. Math-Naturwiss.
Kl. Kais. Akad. Wiss. Wien. LVI, pp. 257-314. Mit 1o colorirten
Tafeln und 12 Text Figuren.

1890. Vergleichende Studien am Keimstreifen der Insekten.
Denkschr, Math.-Naturwiss. Kl. Kais: Akad. Wiss. Wien. LVII,
pp. 621-734.Mit 12 colorirten Tafeln und 38 Textfiguren.

1891. Beitrage zur vergleichenden Embryologie der Insekten.
Denksch. Math.-Naturwiss. Kl. Kais. Akad. Wiss. Wien. LVIII, pp.
803-866. Mit 7 Tafeln und 12 Text figuren.

1891a. Ueber die embryonale Anlage des Blut-und ‘Fettgewebes
der Insecten. Biol. Centbl. XI, 7-8, Leipzig, pp. 212-224.

1891b. Zur Embryologie der Insecten. Zool. Anz. XIV, pp.
286-291. Summary in Jour. Roy Micr. Soc. 1891, p. 729.

Grassi, Battista, Dr. Pror. 1882-4. Studi sugli artropodi. Intorno allo
sviluppo dell Api nell’ uovo. (Letta nella seduta ordinaria del 2
Marzo 1884). Atti dell’ Academie Gioenia di scienze naturali in
Catania. Serie 3, Vol. XVIII. pp. 154222. Tav. I-X.

Hacker, VALENTINE. 1897. Die Keimbahn von Cyclops. Arch. Mikros.
Anat. XLIX, 1, pp. 35-91. Mit Tafel IV und V und 5 Figuren im

: Text.

q Hattez, Paut. 1886. Loi de l’orientation de l’embryon chez les Insectes.

r C. R. Acad. Sci. [Paris]. CIII, pp. 606-608.

HAMMERSCHMIDT, JOHANN. 1910. Beitrage zur Entwicklung der Phasma-
tiden. Ztschr. Wiss. Zool. LXLV, 2, pp. 221-242. Mit Tafel IV und V.

HatscHek, BerTHOLD. 1877. Beitrage zur Entwicklungsgeschichte der
Lepidopteren. Jen. Ztschr. Naturw. XI, (4) pp. 115-145. Tafel VII-
IX, 2 Figg. im Text. :

Hecner, R. W. 10908. Effects of removing the germ-cell determinants
from the eggs of some chrysomelid beetles. Preliminary report.
Biol. Bul. XVI, 1, pp. 19-26. é

1914. Studies on germ cells. I. The history of the germ cells
in insects with special reference to the keimbahn determinants. Ef.
The origin and significance of the keimbahn-determinants in animals.

268 |. THE EMBRYOLOGY OF THE HONEY BEE

Journ. Morph. XXV, 3, pp. 375-509. With 26 text figures and 10
single plates.

Heer, Kart, 1885. Ueber die Anlage der Keimblatter von Hydrophilus
piceus L. Abhandl. Kgl. Preuss. Acad. Wiss. Berlin. pp. 1-47, Tafel I-II.

1889. Die Entwicklung von Hydrophilus piceus L. Erster Theil.
Herausgegeben mit Unterstiitzung der Kgl. Acad. d. Wissensch. zu
Berlin. Jena. Gustav Fischer.

Henxinc, H. 1892. Untersuchungen iiber die ersten Entwicklungsvor-
gange in den Eiern der Insekten. III. Specielles und Allegemeines.
Ztschr. Wiss. Zool. LIV, pp. 1-274, mit Tafel I-XII, und 12 Holzsch-
nitten.

Hennecuy, L{ouis] Fétix. 1904. Les insectes. Morphologie—reproduc-
tion—embryogenie. Lecons recueillies par A. Lécaillon and G.
Poirault. Avec. 622 figures en noir et 4 planches en coulour hors
texte. -Paris, Masson et cie, editeurs. pp. XX-804.

Hertwic, R. 1881. Anlage der Keimblatter bei den Insecten. Jen. Ztschr.
Naturw. XIV. Supp. 1, pp. 124-128.

Herymons, RicHarp. 1894. Ueber die Bildung der Keimblatter bei den
Insekten. Sitzber. Akad. Wiss. Berlin. 1894, I. pp. 23-27.

1895. Die Embryonalentwicklung von Dermapteren und Orth-
opteren unter besonderer Berticksichtigung der Keimblatterbildung.
Monographisch bearbeitet. Mit 12 lithographischen Tateln und 33
Abbildungen im Text. Jena. Verlag von Gustav Fischer.

1895a. Die Segmentirung des Insectenkérpers. Anhang zu
Abhandl. Akad. Wiss. Berlin, pp. 1-39, mit einem Tafel. Summary by
Verhoeff, in Zool. Zentbl. II, p. 502.

1897. Entwicklungsgeschichtliche Untersuchungen an Lepisma
saccharina L. Ztschr. Wiss. Zool. LXII, 4, pp. 583-631.

1897a. Ueber die Organisation und Entwicklung von Bacillus
rossi Fabr. Sitzber. Akad. Wiss. Berlin. pp. 363-373. Mit einem Fig.
im Text.

1897b. Ueber Bildung und Bau des Darmkanals bei niedern Im-
sekten. Sitzber. Gesell. Naturf. Freunde Berlin. pp. 111-110.

1899. Ueber blaschenformige Organe bei den Gespenstheu-
schrecken. Ein Beitrag zur Kenntniss des Eingeweidenervensystems
bei den Insecten. Sitzber. K. Preuss. Akad. Wiss. (Berlin), 1899. pp.
563-575. Mit 2 Figuren im Text.

HirscHier, M. J. 1905. Badania embryologiczne nad motylem Catocala
nupta L. (Lepidoptera) [Embryological studies on Catocala nupta L.
Bull. Internat. Acad. Sci. Krakow. 10, pp. 802-810.

HirscHier, JAN. 1909. Die Embryonalentwicklung von Donacia crassipes

_L. Ztschr. Wiss. Zool. XCII, 4, pp. 627-744. Mit 15 Figuren im Text
und 5 Tafeln.

1909a. O rozwoju listkow zarodkowych i jelita u Gasteroidea
viridula Deg: (Coleoptera). [On the development of the germ layers
and the alimentary canal of Gasteroidea viridula Deg. (Coleoptera) ].
Bull. Internat. Acad. Sci. Krakow. 1900, 2, pp. 284-308. (Planche IV).

Janet. C. 1899. Essai sur la constitution morphologique de la tete de
l’Insecte. Paris. 74 pp., 7 pls. Review by Heymons, Zool. Centralbl.
VII, pp. 800-804.

18o9a. Sur les nerfs céphaliques, les corpora allata et le ten-
torium de la fourmi (Myrmica rubra L.) Etudes sur les fourmis,
les guepes et les abeilles, Note 20. Mém. Soc. Zool. de France. XII,
pp. 295-336. Planches III-VI. 3 fig. dans le text.

1900. Recherches sur l’anatomie de la fourmi et essai sur la

<a

ee A _ cat

Se

aig

-~ * ee = - = > 'd
——— = ae os ee a

THE EMBRYOLOGY OF THE HONEY BEE 269

constitution morphologique de la téte de insecte. Con planches 1
a 15, et 50 figs. dans le text. Paris, 8vo. pp. 1-205.

1907. Histolyse des muscles de mise en place des ailes, aprés le
ees nuptial, chez les fourmis. C. R. Acad. Sci. [varis]) 145, pp. 1205-
1208.

Karawalew, W. 1893. [On the embryonic development of Pyrrchocoris
apterus.| Zapiski Kiev. Cbsch. XIIi, pp. 1-34. [Russian].

Korotnerr, A. 1885. Die Embryologie der Gryllotalpa. Ztschr. Wiss.
Zool. XLI, 4, Taf. XXIX-XXXI. ,

— 1804. Zur Entwicklung des Mitteldarms bei den Arthropoden.
Biol. Centlbl. XIV, 12, pp. 433-434.

KorscHELt, E. und Heiper, K. 1891-1893. Lehrbuch der vergleichenden
Entwicklungsgeschichte der wirbellosen Thiere. Jena.

1895-1900. An English translation under the title “Text-book
of the embryology of invertebrates,” London.
- 1910, Allgemeiner Theil. Vierte Lieferung. VIII. Kapital. Keim-
blatterbildung. E. Insekten. pp. 380-416. Mit Figuren 266-287.

KoscHEvNikov, G. A. Ueber den Fettkérper und die Oenocyten der Honig-
biene (Apis mellifera L.) Zool. Anz. XXIII, 337-353. Summary by
Heymons, Zool. Centlbl. VIII, pp. 172-173.

1900. [Contributions to the natural history of the bee.] Zool.

Div. Imp. Soc. Nat. Hist. Anthrop. and Ethnog. XCIX. Contrib.

Zool. Div. XIV, pp. 1-144. With 37 text figures and 3 plates. [Russian.[

Kowatevsk1, A. 1871. Embryologische Studien an.Wiirmerm und Arth-

-‘ropoden. Mém. Acad. Impér. Sci. St. Petersbourg (7) XVI, 12. 1-70.
Mit 12 Tafeln.

Kowa tevsky, A. 1886. Zur embryonalen Entwicklung der Musciden. Biol.
Centbl. VI, pp. 49-54. [Summary in Jour. Roy. Micr. Soc. (2) VI, pp.
429-430. ] 7

Kuracin, Nicotaus. 1892. Zur Entwicklungsgeschichte der parasitischen
Hautfliigler. Zool. Anz. XV, pp. 85-87. Summary in J. R. Micr. Soc.
1892, 470.

— 1807. Beitrage zur Kenntniss der: Entwicklungsgeschichte von
Platygaster. Ztschr. Wiss. Zool. LXIII. 2, pp. 195-235. Mit Tafel X
und XI.

Lécartton, A. 1897. Note preliminaire relative aux feuillets germanitifs
des Coléoptéres. C. ‘R. Soc. Biol. [Paris]. pp. 1014-1016.

1897a. Contribution a l’études des premiers phénoménes du
développement embryonnaire chez les insectes particuliereément chez
les ‘Coléoptéres. Arch. Anat. Micros. I, pp. 205-224. Planche XI.

Lécamton, A. 1808. Recherches sur le développement embryonnaire de
quelques chrysomélides. Arch. Anat. Micros. II, pp. 118-176. Planche
VI; pp. 189-248, planche IX. Reviewed by Heymons. Zool. Zentbl.
V, 1808, pp. 813-816.

LeucKkart, R. 1865. Die ungeschlechtliche Fortpflanzung der Cecidomy-
ienlarven. Arch. Naturgesch. XXXI, 1: X, pp. 286-303, Hierzu Tafel
XII.

MarsHatt, W. S., and Dernent, PAut H. 1905. Contributions toward
the embryology of Polistes pallipes (Hymenopteron). I. The forma-
tion of the blastoderm and the first arrangement of its cells. Ztschr.
Wiss. Zool. LXXX, 1, pp. 122-154. ‘

Mayer, Paut. 1876. Ueber Ontogenie und Phylogenie der Insecten. Jen
Ztschr. Naturwiss. X. (III) pp. 125-220. Hierzu Tafel VI u. VI
a, b, c. ;

Mecznixorr, Cann. [EttAs.] 1865. Ueber die Entwicklung der Cecido-

270 THE EMBRYOLOGY OF THE HONEY BEE

myienlarve aus dem Pseudovum. Vorlaufige Mittheilung. Arch.
Naturgesch. XXXI, 1, pp. 304-310.

MeEcznikow, ExiAs. 1866. Embryologische studien an Insekten. Ztschr.
Wiss. Zool. XVI, 4, pp. 380-500. Mit Tafel XXIII-XXX.

Mercier, L. 1906. Les corps bactériodes de la blatte (Periplaneta
orientalis) : Bacillus Cuenoti (n. sp. L. Mercier). (Note preliminaire).
C. R. Soc. Biol. Paris LXI, pp. 682-684.

NACHTSHEIM, HANS. 1913. Cytologische Studien itber die Geschlechts-
bestimmung bei der Honigbiene (Apis mellifica L.) Arch. Zellforsch.
XI, 2, pp. 169-241. Mit 6 Figuren im Text, 4 Tafeln und einer
Tabella.

Netson, JAs. A. 1911. The origin of the rudiments of the mesenteron in
the honey bee. Science, N. S. XXXIII, 842, pp. 273-274.

1912. A peculiar structure in the embryo of the honey bee.
Science, N. S. XXXV, 890, p..475.

1914. A pair of tracheal invaginations in the second maxillary
segment of the embryo of the honey bee. Science, N. S. XXXIX,
1003, DP. 437.

Noack, W. 1001. Beitrage zur Entwicklungsgeschichte der Musciden.
Ztschr. Wiss. Zool. LXX, 1, pp. 1-56. Mit Tafel I-V und 10 Figuren
im Text.

NussAuM, JoseF. 1883. Vorlaiifige Mitteilung itber die Chorda der Arth-
ropoden. Zool. Anz. VI.

1886. The embryonic development of the cockroach. In Miall
and Denny, “The structure and life-history of the cockroach
(Periplaneta orientalis) ,” pp. 181-204 with figures 104-118.

1888. Die Entwicklung der Keimblatter bei Meloé proscarabaeus,
Marsham. Biol. Centbl. VIII, 15, pp. 490-452. 2 Figg. im Text Summary,
J. R. Micr. Soc. 1887, p. 942.

1890. Zur Frage der \Riickenbildung bei den Insectenembryonen.
Biol. Centbl. X. 4, pp. I10-114.

NusspAauM, JOsEF, und FuLinski1, BENEDYKT. 1906. Uber die Bildung
der Mitteldarmanlage bei Phyllodromia (Blatta) germanica L. Mit
15 Figuren. Zool. Anz. XXX, pp. 362-381.

1909. Zur Entwicklungsgeschichte des Darmdriisenblattes bei
Gryllotalpa vulgaris Latr. Ztschr. Wiss. Zool. XCIII, 2, pp. 306-348.
Mit 2 Tafeln und 5 Figuren im Text.

PatmeEn, J. A. 1877. Zur Morphologie des Tracheensystems. Leipzig.

PatteN, WILLIAM, 1884. The development of Phryganids, with a prelim-
inary note on the development of Blatta germanica.. Quart. Journ.
Micr. Sc. XXIV, pp. 549-602. Pl. XXXVIa, XXXVIb, and XXXVIc.

1887. Studies on the eyes of arthropods. Development of the
eyes of Vespa, with observations on the ocelli of some insects. Journ
Morph. I, 1, pp. 193-226. Plate I.

1889. Studies on the eyes of arthropods. II. Eyes of Acilius.
Jour. Morph. II, 1, pp. 97-190. Pl. VII-XITI, text figs. 1-5.

PETRUNKEWITSCH, ALEX. 1901. Die Richtungskérper and ihr Schicksal
im befruchteten und wunbefruchteten Bienenei. Zool. Jahrb. Abt.
Anat. u. Ontog. Thiere. XIV, 4, pp. 573-608, Tafel 1-4.

1902. Das, Schicksal der Richtungskérper im Drohnenei. Ein
Beitrag zur Kenntniss der natiirlichen Parthenogenese. Zool. Jahrb.
Abt. Anat. u. Ontog Thiere. XVII, 3, pp. 481-516 Tafel 1-3.

Pratt, H. S. 1900. The embryonic history of imaginal discs in Melo-
phagus ovinus L., together with an account of the earlier stages in the
development of the insect. Proc. Boston Soc. Nat. History. XXIX,
13, Pp. 241-272. With 7 plates.

THE EMBRYOLOGY OF THE HONEY BEE 271

Rasito, Leonarpo. 1898. Sull’ origine dell’ intestino medio nella Mantis
religiosa. Nat. Siciliano. II, 5-8, pp. 181-183.

Ritey, WititiAm A. 1904. The embryological development of the skeleton
of the head of Blatta. Thesis presented to the Faculty of Cornell
University for the degree of Doctor of Philosophy. Amer. Nat.
XXXVIII 455-456, pp. 777-810. 12 text figs. Printed as separate by
Ginn and Co. Pub. The Athenaeum Press, Boston.

Ritter, RicHarp. 1890. Die Entwicklung der Geschlechtsorgane und des
Darmes bei Chironomus. Ztschr. Wiss. Zool. L, 3, pp. 408-427. Mit
Tafel XVI.

Rizzi, Marco. 1912. Sullo sviluppo dell ’uovo di “Bombyx (Sericaria)
mori L.” nel primo mese deposizione. “Redia.” VIII, 2, pp. 323-350.
‘Con tavole I1-V.

Rtckert, J. 1895. Ueber das Selbststandigbleiben der vaterlichen und
miitterlichen Kernsuwbstanz wahrend der ersten Entwicklung des be-
fruchteten Cyclops Eies. Arch. Mikros. Ant. XLV, 3.

SALING, THEODOR. 1907. Zur Kenntnis der Entwicklung der Keimdrtsen
von Tenebrio molitor L. Mit 2 Tafeln und 14 Figuren im Text-
Ztschr. Wiss. Zool. LXXXVI, 2, pp. 238-303.

ScHWANGART, F. 1904. Studien zur Entodermfrage bei den Lepidopteren.
Ztschr. Wiss. Zool. LX XVI, 2, pp. 167-212. Mit Taf. XII und XIII
und 4 Figuren im Text.

ScHWARTzE, Dr. ErtcH. 1899. Zur Kenntnis der Darmentwicklung bei
Lepidopteren. Aus dem zoologischen Institute der Universitat Berlin.
Ztschr. Wiss. Zool. LX VI, 3, pp. 450-496. Taf. XXXI-XXXIV.

TANNREUTHER, Geo. W. 10907. History of the germ cells and early em-
bryology of certain aphids. Zool. Jahrb. Abt. Anat, XXIV, 4,
pp. 609-642. With plates 49-53.

Tanguary, Maurice Cote. 1913. Biological and embryological studies on
Formicidae. III. Studies on the embryology of Camponotus hercu-
leanus var ferrugineus Fabr. and Myrmica scabrinodis var sabuleti
Meinert. Bull. Ill. State Lab. Nat. Hist., IX, art IX, pp. 454-477. With
pl. LVII-LXIV.

TicHomirorFr, A. 1879. Ueber die Entwicklungsgeschichte des Seiden-
wurms. Zool. Anz. II, 20, pp. 64-67.

1882. [On the ontogeny of the Bomby mori within the egg.]
(Russian), Izvyest. Imp. Obshch. Lyubit. Estest. Antrop. i. Ethnog.
Moscow, XXXII, 4.

1885. Chemische Studien iiber die Entwicklung der Insecteneter.
Ztschr. Physiol. Chem. IX, 4-5, pp. 518-532.

—_—— 1800. Ueber die Entwicklung der Calandra granaria. Biol.
iCentbl. X, 13-14, p. 424.

TicHomirow, Dr. ALEXANDER, 1892. Aus der Entwicklungsgeschichte der
Insekten. Festeschrift zum 70. Geburtstage Leuckarts. Leipzig.
pp. 337-346. Mit Tafel XXXV. ;

Ticpomirowa, O. S. 18090. Zur Embryologie von Chrysopa.. Biol.
Centbl. X, 13-14, p. 423. (Der VIII Kongress russischer Naturforscher
und Aertzte in St. Petersburg. II. Sektion fiir Zoologie. 1. Sitzung
am 29. Dez. 1889. [9 Jan. 1800.]).

TrkHomirowA, Orca. 1802. Sur Vhistoire du dévelopment de Chrysapoda
perla (L’Origine du mésoderme des cellules vitellines) Cong. Internat.
Zool. 2, sess. I. pp. 112-119.

Toyama, K. 1902. Contributions to the study of silk worms. I. On the
embryology of the silk-worm. With plates VII-XI. Bul. Col. Agr.

Tokyo Imp. Univ. V, pp. 73-1 18.

272 THE EMBRYOLOGY OF THE HONEY BEE

TSCHUPROFF, HELENE VON. 1903. Ueber die Entwicklung der Keimblatter
bei den Libellen. (Aus dem zoologischen Institut der Universitat
Berlin), (Volfraufige Mitteilung.) (Mit 1 Figur.) Zool. Anz. XXVIJI,
PP. 29-34.

UzeL, Dr. HernricH. 1897. Vo6rlaufige Mittheilung itber die Entwicklung
der Thysanuren. Zool. Anz. XX, 528-529. pp. 125-132.

1898. Studien iiber die Entwicklung der apterygoten Insecten.
Mit 6 lith. Tafeln und 5 Abbildungen im Text. 58 pp.

VIALLANES, H. 1887. Etudes histologiques et organologiques sur les
centres nerveaux et les organes des sens des animaux articulés. 4te
mémoire. Ann. Sci. Nat. Zool. II, pp. 5-100.

1891. Sur quelques points de T’histoire du développement
embryonnaire de la mante religieuse (Mantis religiosa). Ann. Sci.
Nat. Zool. (7), XI, pp. 283-328. Con planches 12 et 13.

VoELTzKow, ALFRED. 1888. Vorlaufige Mittheilung tiber die Entwicklung
im Ei von Musca vomitoria. Zool. Anz. XI, 278, pp. 235-236.

1889. Entwickelung im Ei von Musca vomitoria. Arb. Inst.
Wurzb. IX, pp. 1-48. Mit Tafeln I-IV.

—— 188 9a. Melolontha vulgaris. Ein Beitrag zur Entwicklung im
Ei bei Insecten. Arb. Inst. Wurzb. IX, pp. 49-64. Mit Tafel V.

WEISMANN, Dr. Aucust. 1863. Die Entwicklung der Dipteren im Ei,
nach Beobachtungen an Chironomus spec., Musca vomitoria und Pulex
canis. Ztschr. Wiss. Zool. XIII, 1, pp. 107-158, u. 2, pp. 159-220. Mit
Tafel VIII-XIII.

1882. Beitrage zur Kenntniss der esten Entwicklungsvor-
gange im Insektenei. Beitrage zur anatomie und embryologie als
Festgabe Jacob Henle zum 4. April 1882. Bonn 1882. Hierzu Tafel
X, XI, XII.

WHEELER, WILLIAM Morton. 1889. The embryology of Blatia germanica
and Doryphora decemlineata. Jour. Morph. III, 2, pp. 291-384. Plates
XV-XX. 16 text figures.

1891. Neuroblasts in the arthropod embryo. Jour. Morph. IV, 3,
PP. 337-343.

1892. Concerning the blood tissue of the insects. Parts
I-III, 1 plate in No. 193. Psyche, VI, 190, 191, 193. Pp. 216-220,
233-236, 253-258.

1893. A contribution to insect embryology. Jour. Morph. VIII,
I, pp. 1-160, pls. I-VI.

1893a. The primitive number of Malpighian vessels in insects.
Psyche, VI, pp. 457, 485, 497, 500, 539, 545, 56I.

1910. Ants. their structure, development and behavior. Colum-
bia Univ. Press, N. Y. 1910.

WIELOWIEJSKI, Dr. HetnricH RITTER von. 1886. Ueber das Blutgewebe
der Insekten. Eine vorlaufige Mittheilung. Ztschr. Wiss. Zool.
XXXIV, 3, pp. 512-536.

Wit, Lupwic. 1888. Zur Entwicklungsgeschichte der viviparen Aphiden.
Biol. ‘Centbl. VIII, 5, pp. 145-154.

1888a. Entwicklungsgeschichte der viviparen Aphiden. Zool.
Jahrb. Abth. Anat. Ontog. Thiere, III, pp. 201-286. Hierzu Tafel
VI-IX.

Wittaczit, Dr. EMANUEL. 1884. Entwicklungsgeschichte der Aphiden.
Ztschr. Wiss. Zool. XL, 4, pp. 550-606. Mit Tafel XXVITI-XXXIV.

WoopwortH, CHARLES W. 188. Studies on the embryological develop-
ment of Euvanessa antiopa. In: Scudder, Samuel Hubbard. Butter-
flies of New England, I, pp. 95-104, pl. 63.

INDEX

Names of authors are in small capitals, generic names in italics.

Abdominal appendages, 112
Acetic alcohol, 251
Acilius, optic lobe, 154, 157
Adductor (flexor) muscle of man-
dible, 178, 191, 207, 258
Agelastica, mesenteron, 74
yolk cells, 240
Ailmentary canal, development, Ioo,
102, 103, 104, 221-229
larva, 97, 98, 220, 221
muscles, 192, 199, 201, 202, 203,
205, 206, 207, 209, 210, 259
Amnion, 43, 45, 59, 63, 66, 78, 82-88,
96, 179, 253
ANGLAS, oenocytes, 183
Antennae, 100, 102, 104, 105, 120,
142, 143, 144, 156
Antennal coelomic sacs, 205, 206
Antennal lobes, see deutocerebrum
Antennal nerve, 120
Antennal somites, 163
Anterior field, 56
Anthophora, blastoderm, 34
cleavage, 24
cortical layer, 14
Ants, abdominal appendages, 112
Anurida, mesenteron, 73
premandibular appendages, 109
procephalic organ, 04
segmentation of head, III
Anus, 104
Aorta, development, 196, 207
larva, 193
Aphids, yolk cells, 242
Aphis, genital cells, 218
mesenteron, 73, 74
Apodeme of mandible, 161, 163,
178, 179
Appendages, on abdomen, 112, 255
on antennal segment, 107, 108,
255
on tritocerebral (premandibu-
lar) segment I09-III, 255
Apterygota, mesoderm, 50
Ascaris, germ cells, 219
AYERS, mesenteron, 72
segmentation, 55

Bacillus, corpora allata, 163
mesenteron, 74

273

BALBIANI, genital cells, 218
BALFoUR, mesenteron, 72
Biorhiza, amnion, 87
cleavage, 25
Blastoderm, 44, 236, 237, 238, 242
formation of, 27-43, 254
Blastokinesis, 88
Blatta, mesenteron, 73, 74, 77
oenocytes, 182
BLOcHMANN, blastoderm, 34
corpuscles, 9, 13, 14, 15, 254
maturation of egg, 3
Blood corpuscles, development, 197,
199, 259
larva, 106
BoprETZSKY, mesenteron, 72
mesoderm, 50
yolk cells, 240
Body cavity, 210
Bombyx, corpora.allata, 164
invaginations on gnathal seg-
ments, 174
mesenteron, 73, 74
Boveri, germ cells, 219
Brain, degenerating cells, 164-166
development, 100, 102, 103, 142-
157, 255
larva, 98, 113-121
BUrcer, mesoderm, 197, 209
neuroblasts, 139
segmentation, 55
BUTSCHLI, I
abdominal appendages, 112
amnion, 85, 87, 88
cleavage, 26
contraction of egg, 95, 96
genital organs, 216
mesenteron, 69
premandibular appendages, 109,
IIO
segmentation, 54
tracheal system, 167

Calandra, mesenteron, 73
Calliphora, mesenteron, 73, 74, 75;
77
Calopteryx, mesenteron, 73, 74
Campodea, mesenteron, 73, 75
Camponotus, amnion, 87
Blochmann’s corpuscles, 14

274 INDEX

cleavage, 25
cortical layer, 15
Cardioblasts, 201, 203, 204, 205,
215, 259
Carmalum, Mayer’s, 252
CARRIERE, abdominal appendages,
112
amnion, 87, 88
anterior mesenteron rudiment,
60
frontal ganglion, 159
mesenteron, 77, 227
mesoderm, 50, 208
segmentation, 55, 107
CARRIERE AND BURGER, 3
amnion, 87
anterior mesenteron rudiment,
60, OI
blastoderm, 34
brain, 157
corpora allata, 163
dorsal diaphragm, 195, 205
fat body, 193
ganglia of ventral cord, 138
genital organs, 217
Malpighian tubules, 228
median cord, 140, I41
mesenteron, 70, 7I, 75
mesoderm, 50, 197, 200, 210, 211
polarity of egg, 4
tentorium, 179
tracheae, 175
yolk, 241
yolk cells, 243
Catocala, mesenteron, 73, 74
Cell division, blastoderm, 35
cleavage cells, 21-23
yolk cells, 230-232
Cephalo-dorsal body, 88-94, 255
Cephalo-dorsal disk, 88
Chalicodoma, abdominal append-
ages, II2
amnion, 87, 88
anterior mesenteron rudiment,
60, 61
blastoderm, 34
cleavage, 24, 25, 26
corpora allata, 163
cortical layer, 34
dorsal diaphragm, 195, 196
fat body, 193
ganglia in ventral nerve cord, 138
lateral folds, 45
Malpighian tubules, 228
median cord, 140
mesenteron, 75
mesoderm, 49, 50, 51, 107, 208,
209, 2II, 212

polarity of egg, 4
segmentation, 55, 107
stomatogastric system, 159
yolk, 241
yolk cells, 243
Chironomus, genital cells, 218
mesenteron, 73
(CHOLODKOWSKY, mesenteron, 73,
74, 76
mesoderm, 50
yolk cells, 236
Chorion, 6, 253
in technique, 252
Chorionin, 7
Chrysomela, Blochmann’s corpus~
cles, 14
chemical tests of eggs ,I5
mesenteron, 74
yolk cells, 240
Chrysopa, mesenteron, 73
Circulatory system, 193 °
‘Circumoesophageal commissures,
113, 156
CLAYPOLE, appendages on tritocere-
bral segment, 109
mesenteron, 73
procephalic organ, 94
Clearing, 252, 253
Cleavage, 16-26, 254
Cleavage cells, 13, 230, 240, 254
arrangement, 16, 254
division, 21-26
Clytra, blastoderm, 31
cleavage, 25
irregular mitoses, 231
mesenteron, 74
yolk cells, 240
Coelomic sacs, 208, 257, 259
Coleoptera, amnion, 88
frontal ganglion, 159
median cord, I41
mesenteron, 76, 77
mesoderm, 49, 208
oenocytes, I82
optic lobe, 153, 154, 155
optic plate, 154
yolk, 241
Commissure,
113, 156
suboesophageal, 120, 155, 156, 257
supreoesophageal, 116, 155, 257
Commissures, tracheal, 103, 105,
167, I71, 173, 258
transverse, I15, 135
Compound eye, I51
CONKLIN, cleavage nuclei, 22
Connectives, development, 134
larva, 98, 113, I15

circumoesophageal,

INDEX 275

Contraction of egg, 95
Corethra, oenocytes, 181
Corpora allata, development, 161-
164, 257
larva, 122, 124
Cortical layer, 10, 27, 20, 31, 34, 35,
254
inner, 31, 34, 35, 38, 40, 42, 238,
254, 255
Crepidula, cleavage nuclei, 22
CZERSKI, mesenteron, 74

DEEGENER, mesenteron, 74
Degenerating cells, 164-166, 257
Dermaptera, brain, 157
inner cortical layer, 31
optic lobes, 154, 157
protocerebrum, 116, I17
tritocerebral segment, 109
yolk cells, 240, 241
Dermatoblasts, 126, 129, 130, 257
Deutocerebral segment, 107, 108
Deutocerebrum (deutocerebral
lobes), 115
development, 143, 144, 147, 148,
164, 166, 255
larva, 100, 102, II9
Deutoplasm, 9, 15
Diaphragm, dorsal, development,
200, 201, 204, 205, 215, 259
larva, 193, 194, 195
ventral, development, 205
larva, 193, 196
DICKEL, 3
~ blastoderm, 33
cleavage, 26
irregular mistoses, 231
mesenteron, 7I, 73
method of obtaining eggs, 245
rate of development, 246
yolk cells, 241
“yolk plug,” 71, 92, 93
Dilators of pharynx, 156, 177
Diptera, yolk, 241
Dixippus, mesenteron, 74, 77
yolk cells, 244
Dourn, mesenteron, 72, 75
Donacia, appendages on tritocere-
bral segment, I10, III
irregular mitoses, 231
median cord, 141
mesenteron, 73, 75
mesoderm, 210, 2II
neuroblasts, 130, 139
oenocytes, 183
primary dorsal organ, 93, 94
yolk cells, 235, 236

Dorsal diaphragm, development,
200, 201, 204, 205, 215, 259
larva, 193, 194, 195
Dorsal longitudinial muscles, de-
velopment, 200, 201, 215
larva, IQI
Dorsal organ, 94
Dorsal strip, 38, 43, 82, 83, 90, 255
Doryphora, Malpighian tubules,
228
mesenteron, 73
neuroblasts, 132
yolk cells, 234, 242
Drepanosiphum, mesenteron, 74 °
Drone egg, blastoderm, 36
cleavage, 26
mesoderm, 51, 52
polar bodies, 92, 93
Duration and rate of develop-
ment, 245-250

Ectobia, mesenteron, 74
Ectoderm, 78, 80, 81, 124, 170, 178,
222, 255
embryonic and non-embryonic,
179, 180
Egg, changes in form, 95, 97
color, 4
contraction, 95
membranes, 7
mode of deposition, 5
polarity, 4, 253
shape, 4
size, 4
Embedding in paraffin, 253
Embryo, external changes in form,
97-105
Endoskeleton, 175-179
Endromis, mesenteron, 73
Enteric muscles, development, 199,
201, 202, 203-210
larva, 192
Entoderm, 56, 75, 76
Ephemeridea, oenocytes, 182
Epineural sinus, 199, 200, 211
Epitheca, mesenteron, 73, 74
EscCHERISCH, mesenteron, 73, 77
neuroblasts, 132
External changes in form of em-
bryo, 97-105
External medullary mass, 155
Eye, compound, I51

Fat body, evelopment, 199, 201,
203, 200, 210, 259
larva, 193
Fat cells, 193, 216
pericardial, 203, 259

276 INDEX

Fat globules, 13
Fibres, muscle, 189
Fibres, nerve, development, 134,
135, 155, 156
larva, II5
Fibrillae, muscle, 189
Flexor (adductor) muscle of man-
dible, 178, 191, 207, 258
Fotsom, appendages of head, 111
appendages on tritocerebral seg-
ment, I09
Fore-intestine, development, 244,
260
larva, 220
muscles, 192, 206
Forficula, commissures of. brain,
155
corpora allata, 161, 163
dorsal diaphragm, 1096
frontal ganglion, 159
mesenteron, 74
mesoderm, 50, 210, 211
muscles of hind-intestine, 207
neurilemma, 160
neuroblasts, 130, 132
oenocytes, 182
optic ganglion, 153
proventricular valve, 226
sex cells, 218
yolk cells, 242
Form of embryo, changes in, 95-
105
Formation of blastoderm, 27
Formation of mesoderm, 43
Formation of rudiments of mid-
intestine, 56
Formica, amnion, 87
cleavage, 25
polarity of egg, 4
FRIEDERICHS, Blochmann’s corpus-
cles, 14
chemical tests of eggs, 15
irregular mitoses in yolk cells,
231
mesenteron, 75
paracytes, 242
sex cells, 219
yolk cells, 234, 235, 236, 241, 242,
243
Frontal ganglion, development,
158, 159
larva, I2I, 122
Frontal nerve, development, 159
larva, I21
Ganglia allata, 122, 161
Ganglia of ventral nerve cord,
development, 132, 136, 137

larva, 98, 113
number, 137, 138
Ganglia, pharyngeal, development,
159
larva, 122, 124
Gangliomeres, 115, 130, I41
Ganglion, frontal, development,
158, 159
larva, I2I, 122
Ganglion, occipital, 159
Ganglion, optic, 153, 257
suboesophageal, 98, 113, 257
pharyngeal, 122, 124, 159, 160,
257
Ganglion cells, 126
development, 129, 130, 134, 135,
155, 150, 161, 257
larva, II5
Ganglionic layer of brain, 155
GANIN, amnion, 87
cleavage, 25
mesenteron, 74, 75, 76
polarity of egg, 4
Gasteroidia, Malpighian tubules,
228
mesenteron, 61, 73
mesoderm, 50
Gasterophysa, mesenteron, 73, 74
mesoderm, 50
Gastrula, 75, 76
Gastrulation, 46
Genital ridge, 214, 216, 259
Germ band, 45
Germ layers, 43-81
GILsON’s mixture, 251
GLASER, oenocytes, 181, 183
GraBER, abdominal appendages, 112
amnion, 87
median cord, 134, 139, 140
mesenteron, 72, 73, 74, 76
mesoderm, 50, 208
oenocytes, 181, 182, 185
segmentation, 55, 56
ventral nerve cord, 138
yolk cells, 242
GRASSI, 2
abdominal appendages, 112
amnion, 85
cephalo-dorsal body 92
cleavage, 26
Malpighian tubules, 228
mesenteron, 70, 72. 75, 76, 227
mesoderm, 46, 50
ovaries, 216
premandibular appendages, 109
tentorium, 179
tracheal system, 167

INDEX 277

ventral cord, 134, 138
yolk cells, 240
Gryllotalpa, mesenteron, 72, 74, 77
mesoderm, 50
yolk cells, 240, 242, 244
GRYLLUs, corpora allata, 163
mesenteron, 74
mesoderm, 50
oenocytes, 182
sex cells, 219
stomatogastric system, 159

yolk cells, 242, 244

HACKER, germ cells, 219
HALtez, law of, 4
HAMMERSCHMIDT, mesenteron, 74,

oe es
yolk cells, 244
HATSCHEK, nervous system, 124,
138
tracheae, 175
tracheal invaginations on gna-
thal segments, 174
Head, muscles of, 177, 191, 206, 207
Head capsule, 180
Heart, development, 201, 204, 205,
259
larva, 90, 193
HEGNER, sex cells, 219
Hetper, blastoderm, 34, 40, 42
frontal ganglion, 159
mesenteron, 73, 76
mesoderm, 208, 210
oenocytes, 181
optic lobe, 154, 157
rate of development, 248, 249
segmentation, 55
tentorium, 179
yolk cells, 241
Hemiptera, oenocytes, 182
HENKING, cleavage, 25
cortical layer, 14
HENNEGUvY, chorion, 7
cleavage, 25
sex cells, 219
HERTWIG, mesenteron, 72
HeEymons, brain, 157
commissures of brain, 155
corpora allata, 122, 161, 163
dorsal diaphragm, 195, 205
frontal ganglion, 159
median cord, 139, 140
mesenteron, 73, 74, 76, 77
mesoderm, 50, 208, 210, 211
muscles of hind intestine, 207
nerve fibres, 135
neuroblasts, 130, 139, 166
oenocytes, 182

optic ganglion, 153
protocerebrum, 117
proventricular valve, 226
segmentation, 108
sex cells, 218, 219
tentorium, 179
tritocerebral segment, 109
yolk cells, 240, 241, 242, 244
Hind-intestine, development, 244,
260
larva, 98, 220, 221
muscles, 192, 207
HIRSCHLER, appendages on trito-
cerebral segment, 110
ectoderm, 61
Malpighian tubules, 228
median cord, 141
mesenteron, 61, 73, 74, 77
mesoderm, 50, 209, 210, 211
neuroblasts, 130, 132
oenocytes, 183
primary dorsal organ, 93
sex cells, 219
Historical review of literature, 1-3
Hydrophilus, blastoderm, 34, 40, 42
cleavage, 25
lateral folds, 45
median cord, I41
mesenteron, 73
mesoderm, 49, 208, 210
oenocytes, 182, 186
optic lobe, 154, 157
rate of development, 248, 240,
250
segmentation, 55
yolk cells, 241
Hylotoma, abdominal appendages,
112
amnion, 87
Hymenoptera, abdominal append-
ages, II2
amnion, 87
frontal ganglion, 159
mesoderm, 210
number of ganglia in ventral
nerve cord, 138 ,
optic lobes, 153, 154
optic plate, 154
yolk cells, 244
Hypodermis, 134, 179, 180
Hypopharyngeal papillae, 111
Hypopharynx, 156
Indusium, 94

Inner cortical layer, 31, 34, 35, 38,
40, 42, 238, 254

Intercalary segment, 109

Tron haematoxylin, 252

278 INDEX

JANET, corpora allata, 163, 175
degenerating cells, 165
segmentation of head, 107
tracheae of head, 175

KARAWAIEW, mesenteron, 73
KoroTNEFF, mesenteron, 72
mesoderm, 50
neurilemma, 160
oenocytes, 181
yolk cells, 240, 242
KorscHELT and HEIDER, amnion, 85
mesoderm, 49
KoscHEVNIKOY, oenocytes, 183
KOWALEYSKI, I, 2
amnion, 85
blastoderm, 27, 36
cleavage, 26
contraction of egg, 95, 97
mesenteron, 60, 74, 75
mesoderm, 46, 50
rate of development, 248
segmentation, 55
yolk cells, 243
KowALEVSKY, mesenteron, 73, 76
mesoderm, 50
KULAGIN, amnion, 87
cleavage, 25
Malpighian tubules, 228
mesenteron, 73

Labium, 97
Labral nerve, 121, 158
Labrum, 100, 102, 104, 108
Larva, structure of, 97-99
Lasiocampa, mesenteron, 74
paracytes, 242
yolk cells, 242
Lasius, cleavage, 25
cortical layer, 14
degenerating cells, 165
Lateral cords, 129, 130, 132, 134,
135, 130, 138, 141, 160, 161
Lateral folds, 45, 46, 62
Lateral nerves, II5, 135, 136
Lateral plates, 45, 46, 47, 50, 51, 52,
53, 55, 254
Layer, cortical, 34, 35
inner cortical, 31, 34, 35, 38, 40,
42, 248, 254, 255
LECAILLON, blastoderm, 31
chorion, 7
cleavage, 25
irregular mitoses, 231
mesenteron, 74
mesoderm, 50
neuroblasts, 132, 139
sex cells, 218
yolk cells, 240, 241

Legs, rudiments, 101, 103, 105, III,
255
Lepidoptera, corpora allata, 164
mesenteron, 76, 77
oenocytes, 182
yolk, 241, 242
Lepisma, median cord, 140
mesenteron, 73, 75, 76, 77
mesoderm,
Leptinotarsa, Malpighian tubules,
228
mesenteron, 73
neuroblasts, 132
yolk cells, 234, 242
LeucKart, genital cells, 218
Lina, mesenteron, 74
oenocytes, 182
yolk cells, 240
Longitudinal tracheal trunks, de-
velopment, 171
larva, 99, 167
Lucilia, mesenteron, 73, 74

Malpighian tubules, development,
103, 105, 228, 220, 260
larva, 98, 183, 221
Mandibles, 97, 100, IOI, 104, 105
muscles, 178, I9I, 207
Mandibular apodomes, 178, 170
Mantis, mesenteron, 74
optic ganglion, 153
optic lobe, 154
MARSHALL and DERNEHL, blasto-
derm, 31, 34, 36
cleavage, 25
cortical layer, 14
location of egg in cell, 6
polarity of egg, 4
Mason bee, see Chalicodoma
Maxillae, development, 100, I01,
102, 104, 105, 255, 258
larva, 97
Mayer, mesenteron, 72
Mecznikow, yolk cells, 242
Median cord, 124, 127, 129, 132, 134,
135, 136, 137, 139, 140, I4I,
142, 156, 161, 257
Median dorsal strip, 38, 43, 82, 83,
90, 255
Meloé, amnion, 88
mesenteron, 74
paracytes, 242, 243
Melolontha, abdominal append-
ages, I12
median cord, 134, 130, 141
mesenteron, 74
oenocytes, 182, 185
yolk cells, 242

INDEX a0

Melophagus, mesenteron, 74
Membranes of egg, 6, 253
Mercier, Blochmann’s corpuscles,
14
Mesenteron, larva, 98, 220
later development, 222-228, 260
rudiments, 56-82, 221, 222, 224,
227, 255
Mesoderm, 80, 81, 130, 131, 196
formation, 43-56, 254
head, 205-207
later development, 196-219
legs, 201
Mesodermal sacs, 197, 209, 217
tubes, 197, 201, 206, 214, 216, 217,
218, 219
Metamerism, 52
METSCHNIKOFF, genital cells, 218
Microgaster, Malpighian tubules,
228
Micropylar area, 7, 8, 253
Micropyle, 7, 8
Microsomites, 55
Mid-body, 23
Middle plate, 45, 46, 47, 50, 51, 52,
53, 55, 62, 255
Mid-intestine, larva, 98, 220
later development, 222-228, 260
rudiments, 56-82, 221, 222, 224,
227, 255
Mitosis, blastoderm, 35
cleavage cells, 21-23
yolk cells, 230-232
Mouth, 97, 220, 226
Multipolar spindles, 23
Musca, blastoderm, 31, 34, 42
cleavage, 25
inner cortical layer, 31, 42
mesenteron, 73, 74, 77
mesoderm,
neuroblasts, 132
Muscids, yolk cells, 242
Muscles, development, 199, 200,
201, 202, 203, 205, 207, 209, 210,
21I, 250
larva, 177, 189-193
Myrmica, amnion, 87

NACHTSHEIM, cephalo-dorsal body,
93

mitosis in cleavage cells, 21
yolk cells, 241

Neophalax, mesenteron, 72
yolk cells, 240

Nerve, antennal, 120

Nerve cells, 157

Nerve cord, development, 105, 124-

142, 157, 160, 257
larva, 113-115
Nerve fibres, development, 134,
135, 155, 156
larva, 115
Nerve, frontal (stomatogastric),
I2I, 159
labral, 121, 158
Nerves, lateral, 115, 135, 136
Nervous system, development, 124-
106, 255
larva, 113-124
Nervus recurrens, see frontal
nerve
Neural groove, 102, 124, 125, 126,
127, 129, 130, 132
Neural ridges, 102, 110, 124, 144
Neurilemma, development, 134,
160, I61, 257
larva, 115
Neuroblasts, 126, 129, 130, 132, 135,
138, 130, 145-150, 153, 157,
160, 257
Neurogenic area, 127, 142, 145
Neuromeres, 144, 145, 255
Neuroptera, oenocytes, 182
Noack, blastoderm, 31, 34, 40
mesenteron, 75, 77
yolk cells, 242
Nuclei, blastoderm, 28, 29, 31, 33,

35
degenerating cells, 165
yolk cells, 234-236, 241
NusBAuUM, amnion, 88
mesenteron, 73, 74, 76, 200
neurilemma, 160
NusBAUM and FULINSKI, mesen-
teron, 73, 77, 200
Oblique muscles, I9I, 201
Occipital ganglion, 159
Odonata, mesenteron, 76
yolk cells, 244
Oecanthus, mesenteron, 72
segmentation, 55
Oenocytes, 181-188, 259
Oesophageal muscles, 192, 206
Oesophagus, development, 225,
226, 260
larva, 97, 220
Optic ganglion, 153
Optic lobes, 148, 149, 150, I5I, 153-
155, 157, 255
Optic plate, 151, 153, 154
Organization of egg, 4
Orthoptera, amnion, 88
brain, 116, 117, 157
frontal ganglion, 159

280 INDEX

inner cortical layer, 31
median cord, 139, 140
mesenteron, 76
mesoderm, 50, 208, 210
nervous system, 124-126, 130, 139
neuroblasts, 126, 132, 139
oenocytes, 182
optic lobe, 117, 153, 157
protodaeum, 116, 117
sex cells, 218
tritocerebral segment, I09
yolk, 241
yolk cells, 240, 241
Ostia, 194
Ovaries, development, 214-219, 259
larva, 990, 213, 214
PALMEN, tracheae, 175
Paracardial cells, development,
205
larva, 196
Paracardial cellular cord, 195, 196
Paracarmine, Meyer’s, 252
Paracytes, 242
Paracytoids, 235, 236, 241
PATTEN, mesenteron, 72
oenocytes, 182
optic lobe, 153, 154, 157
segmentation of head, 107
yolk cells, 240
Pellicle, protoplasmic of yolk, 43,
84, 254
Periplaneta, mesenteron, 73, 74
mesoderm, 50
proventricular valve, 226
sex cells, 219
PETRUNKEWITSCH, 3
amnion, 82
blastoderm, 36
cleavage, 26
fixing fluid, 252
genital organs, 216
mesenteron, 73
mesoderm, 51
method of securing eggs, 245
polar protoplasm, Io
rate of development, 248
“RZ” cells, 92, 93
yolk cells, 231
Pharyngeal ganglia, development,
159, 160, 257
larva, 122, 124
Pharynx, 220
Phyllodromia, mesenteron, 73, 74,
77
mesoderm, 50
Picro-formol, 251
Picro-sulphuric, acid, 251
Pieris, mesenteron, 72

mesoderm, 50
yolk cells, 240
Platygaster, amnion, 87
cleavage, 25
mesenteron, 73
Polar protoplasm, 10
Polarity of egg, 4
Polistes, amnion, 87
blastoderm, 34, 35
cleavage, 25, 26
cortical layer, 14, I5
polarity of egg, 4
Pontia, mesenteron, 72
PRATT, mesenteron, 74
Premandibular appendages, 101,
109, IIO
Premandibular segment, 109, I10
Primary head division, 106
Primary yolk cells, 230, 240, 254,

260
Primative swellings, 127, 129, 130,
138, 255
Procephalic lobes, 100, 142, 143,
255

Procephalic organ, 94
Proctodaeum (proctodael invag-
ination), 67, 102, 103, 104, 224,
225, 228, 229, 260
Protocerebral segment, 107, 108
Protocerebrum (protocerebral
lobes), 115, 116
development, 100, 102, 142, 143,
144, 145, 147, 148, 149-155, 157,
164, 165, 166, 180, 255
larva, 117-119
Orthoptera and Dermaptera, 116
Protoplasm, of egg, 9, 10-13, 239,
240, 254
Protoplasmic pellicle of yolk, 43,
84, 254
Proventricular valve, 220, 226, 260
Pulex, mesenteron, 73
Pyrrhocoris, mesenteron, 73
mesoderm, 50
RABITO, mesenteron, 74
Rate of development, 245-250, 261
Recurrent nerve, 122
Respiratory system, 99
Rhagonycha, Blochmann’s corpus-
cles, 14
chemical tests of eggs, 15
Rhodites, amnion, 87
cleavage, 25
mesenteron, 72
Ritey, appendages on tritocerebral
segment, 109
tentorium, 179

INDEX 281

Ritter, genital cells, 218
mesenteron, 73
RUcKERT, cleavage nuclei (of Cy-
clops), 22
SALING, mesenteron, 74
Sarcoplasm, 189
SCHWANGART, mesenteron, 73, 242
genital cells, 218
yolk cells, 242
SCHWARTZE, mesenteron, 74
yolk cells, 242
Second antennal segment, 109
Secondary head division, 107
Secondary trunk division, 107
Secondary yolk cells, 236-238, 241,
242, 261
Segmentation, 52, 53, 54, 55, 105,
112, 255, 256
Serosa, 87, 255
Sex organs, 213-219
Silk glands, development, 101, 102,
» 103, 104, I05
larva, 98
Sinus, ventral, 196
Somites, mesodermal, 101
Sphinx, mesoderm, 50
Spiracles, 90, 103, 167, 169, 170,
258
Spiracular branch, 170
Splanchnopleure, 199, 201
Stenobothrus, mesenteron, 74
oenocytes, 182, 188
Stigmata, IOI, 170
Stomatogastric ganglion, 257. See
also frontal ganglion
Stomatogastric nerve, I2I, 159
Stomatogastric nervous system,
development, 157-160, 257
larva, I2I
Stomodaeum (stomodaeal invagin-
ation), I00, 102, 157, 205, 221,
222, 224, 226, 255, 260.
Suboesophageal body, 211
Suboesophageal commissure, 120,
155, 156, 257
Suboesophageal ganglion, 98, 113,
137, 256
Supraoesophageal commissure, 116,
155, 257
Supraoesophageal ganglion, 256
Syncitia, yolk cells, 232, 234, 238
Syncitium, cleavage cells, 16

Taenidia, 168, 170
TANQUARY, abdominal appendages,
112
amnion, 87

Blochmann’s corpuscles, 14
cleavage, 25, 26
cortical layer, 15
Technique, 251
Telias, mesenteron, 72
Tenebrio, mesenteron, 74, 75
hig development, 178, 170,
25
larva, 177, 178
TicHomirorF, chorion, 7
mesenteron, 72, 73, 74
oenocytes, 181
premandibular appendages, 109
TICHOMIROWA, mesenteron, 73
ToyaAMA, corpora allata, 164
mesenteron, 74
tentorium, 179
Tracheal commissures, develop-
ment, 105, 170-175, 258
larva, 103, 167-170
Tracheal system, development, 170-
175, 258
larva, 99, 167, 170
Tritocerebral (premandibular) seg-
ment, I07-II0
Tritocerebrum (tritocerebral
lobes), 115
development, 100, 102, 103, 143,
144, 146, 147, 156
larva, 120
TSCHUPROFF, mesenteron, 73, 74,

244
yolk cells, 244

UzeEL, appendages on tritocerebral
segment, 109
mesenteron, 75
mesoderm, 50

Ventral diaphragm, development,
205
larva, 193, 196
Ventral longitudinal muscles, de-
velopment, 199, 200, 201
larva, IQI
Ventral nerve cord, degenerating
cells, 164
development, 105, 124-142, 157,
160, 257
larva, 98, I13-II5
Ventral -plate, 45, 46, 55, 60, 62,
63, 64
Ventral sinus, 196
Vespa, optic lobe, 153, 157
segmentation of head, 107
VIALLANES, brain, 156
frontal ganglion, 159

282 INDEX

neuroblasts, 166

optic ganglion, 153

protocerebrum, 117

segmentation of head, 107
Vitelline bodies, 9, 239, 254
Vitelline membrane, 8, 253, 254
Viteline spheres, 9, 12, 254

not fat or oil, 15, 254
VoELTZKOW, mesenteron, 74, 76

yolk cells, 242

WEISMANN, I
amnion, 87
cleavage, 25
cortical layer, 10
inner cortical layer, 31
mesenteron, 72

WHEELER, abdominal appendages,

112

blastokinesis, 88
Blochmann’s corpuscles, 14
brain, 157
frontal ganglion, 159
germ cells, 219
indusium, 94
Malpighian tubules, 228
median cord, 139, 140
mesenteron, 73, 76
micropylar organ, 94
nerve fibres, 135
neurilemma, 160
neuroblasts, 132, 138, 166

oenocytes, 182, 185
optic ganglion, 153
premandibular appendages, 1090
protocerebrum, 117
segmentation, 107
suboesophageal body, 211
tracheae, 175
olk cells, 234, 242
IELOWIEJSKY, Oenocytes, 181, 183
WILL, genital cells, 218
mesenteron, 73
yolk cells, 242
WITLACZIL, genital cells, 218
mesenteron, 74, 76
WoopwortTH, genital cells, 218

XIPHIDIUM, indusium, 94
median cord, 139, 140
oenocytes, 182, 188
optic ganglion, 153
optic lobe, 154

YOLK, 9, 199, 200, 201, 206, 210,
224, 227, 230-244, 254, 201
Yolk bodies, 13
Yolk cells, 19, 230-244, 254
primary, 230, 240-244, 254
secondary, 236-238, 241, 242, 261
Yolk plug, 71, 92, 93, 255

Zygaena, cleavage, 25
mesenteron, 73

Figs. I-XV. Selected stages in the development of the egg.
In the text the stage represented by each of the figures is
designated by the Roman numeral of the corresponding figure.
The drawings are made from eggs which have been fixed,
stained, and mounted in balsam, and the eggs are therefore
treated as semitransparent objects. The stages during the form-
ation of the blastoderm are omitted, since nothing of importance
is visible on the exterior of the egg, x.

Fic. I. Lateral view. The cleavage cells form a small group a}
at the anterior end of the egg. The. polar protoplasm (PP) |
is evident as a deep staining disk.

Fig. II. Lateral view. The cleavage cells have increased 3
in number and form a conical figure.

Fic. III. Lateral view. The cleavage cells. have further
increased in number, and the conical figure formed by them
has increased in size, so that in the anterior half of the egg the ~
cleavage cells lie on the surface of the egg and are here beginning
to form blastoderm.

Il re II

ee oe ae

a

EE SS eS. a! Lr OU

‘Fic. IV. Ventral view, 32-34 hours. The jewee fo
have just made their appearance.

Fic V. Ventral view. About 36 hours. The lateral
have greatly lengthened, and the lateral plates (LP)
proximating each other. The head fold of the amnion
and the anterior mesenteron ruin (AMR) are now

Cy

Fic. VI. Ventral view. asas hours. The lateral ;
(LP) have become united along the greater portion of t
length. The posterior mesenteron rudiment a is >
visible.

’

Bie. Vil, Vera ewe
(LP) are almost setae ‘ane Th ad
amnion (Am) covers about one e quarter of the ent

Fic, VIla. Lateral view of the same egg as. tie

_ Fig. VII, showing the outlines of the germ: & nd, at
Sass tn aes (ProL). ;

sirens. ‘silk ‘glands, om crachedl ae

Vill

Vila

Pe X. Ventral view, 52-54 hours. “tne tude nents
silk glands (SkIG/) are about one-half of e length
embryo. Pe pe =, oe

VIlla

‘

Fic. Xa. Lateral view of the egg represented in Fig. X. The
spiracles (Sp) have become contracted to minute ‘circular aper- |
tures; the stomodaeum (Sto), proctodaeum (Pro), a8 the
Malpighian tubules (Mal) are now evident. |

: t-
y |

a.

Fic. XI. Ventral view. The silk glands (S/kGl) now reach —
into the posterior segments of the abdomen, the rudiments of —
the tracheal commissures (TraCom) of each are about to unite :
on the ventral mid-line.

Fic. XII. Ventral view. Estimated at 58-60 hours. The —
tracheal commissures (TraCom) are completed, the second —
maxillae (2M) are beginning to approach one another, and the
ventral nerve cord (VNC) is visible from the surface. :

Bit

me ee XU Seehieal view. 66-6 I
(eMxy have begun to unite to tote» oe
a Pepetaer the previously separa: external

pha (SIkGI).- ee Oe am

] ‘ =

‘ a
Aran 6 _ +

Fie. XIV. ‘Ventral eee ) 72-76, hen - i
_ virtually completed. a0 as Oe ae Sa

Fic. XV. Lateral view of newly hatched larva.

4i4er 4424 LILA

Ey ge
Raye

Ae

oo
shee

a

aoe, . mally
Le ee e

: OT! aque 6 SS OLL

A0yu NY

Fs

Sanat bapae tata penenaon
ee ST, TT LT TTT EE TE ES Oa TE a TIT 2 ee ra —
~ = . ~ eleceetone edineerssenarntcpmeaiomierseseatboeeneas =
ine —— ————— or _— —
— ——— Seereoenccrneen oe monasccoimmeenn
ree nna Se > ee i
— = em *
—_ = 7 ioe
—— - —
a reer : ene
—_ (net ee ane eet gente
pemanamapamit — rv (ee an Sree at mene meee a
a ve es wn
oe errr aoe er = a ET Oe A ateere eaten pabeeciotataickenadsceentbamatioeen
ae — OO eC a a — ANNAN IRIN S oe
— ——— oe ee — Semon. oe ear sate wa re ea ae amen nem ~~ wAanin
ee ——— ee ee
—— Se nee nner a ee a
- —_ a —_ - ee Fee ee ee
enema anes Ft ne ee eee eee
= = SSS SSS
eT a nonapenan
a Seateneaaeremen at RN Re att ne mee ee en
mane —enn aaa Saraienanuenaeinen anapanaannanninenee angen ae 7 eR ee ee ee eee Fe ee ean
ane — a Wn ee ——— I ee ke rece ne e
—— eo PRA a Bo 5 et rrr ae
a a NN AR RA A Fa a ee ee
SS SS ESS = ao ES
om a a Rt eee eee = ee —
: en A OE, enn arrears ms — Sen ebtentestatees ems
~~ Pe = eye, ~ a a a cee nen Ong te! onan emma
-— Cet et TT en, ee an ee ntti roe — et etme ceee- ce emmpnomrs ras Seer « + +00 erebaoncaree
Re A a nn —-- SS a eae _
- ne tO a ee eee a eta See
— A AAA a A = . — meneraed - hs ane, tne A pene nent
en me al meg erento ent Be ee emt > —i aw. 2 er - . co - cae ea eee ¢.
——— ee = 2 erp ae hema emrananeery Same
Nin nr a nn ne eect en Re er oS a nen nernes : * a 28 any mara
= nee iar er rm re ome pel Lavette - sbepases-qoum oum-temaanmineh + ope noen
aaa OS CE CAAT PAA A SA PTT et —cveseenvereunoreciaoaens ee = Soeocane
ES ee net ee ~ ae bs
~ ee a ol RR ot nena Let aaron
= ae SN ne eh EE CL a at a tn =
~.
a

- aon -- = ~ eer rs rare tte uyecions r : - aopcnasinageyaannen
: eS ————————xx_E ae a : = So = : EEE ES
Ate AY ee ECT De te A A ROT a Re NS ew ~ - ~s mera oes = Pe OP ayaa “ ae
- ne i eee ert iinet iat i ice i ~ me “ . ind Poy 0 =an A 0 et AGS FID S-PU I ce Satine > >» Sek oneh. Oe = me =