MARINE BIOLOGICAUABORATORY,

Z, S" (.T..O.

</'

Received

i

Accession No.
Given by &•
Place,

* *fK, book o* pamphlet Is to be removed from tbe lab-
oratory cuithout the permission o! the Trustees.

'

' " '" - /

OBSERVATIONS

ON THE

EMBRYOLOGY

OF

INSECTS AND ARACHNIDS

BY
ADAM TODI) BRUCE,

/;. I. PBJ.VI ETON <;,I.I.KI;I: i-i; ;•//. /'.. Joir\x H«;-A-;.VS rxift:i!*[TV.

A MEMORIAL VOLUME.

BALTIMOKH :

Pulilication Agency of Johns Hopkins University,
1887.

/ o 3 O

J. D. EHLERS & 00., PRINTERS AND ENGRAVERS,
HOEN BUILDING, BALTIMORE, MD.

This Volume is printed, after the Author's death, as a Memorial, by his
instructors, pupils and friends in Princeton, Baltimore and elsewhere. It con-
sists of his Thesis, reproduced from the copy which was accepted by the
Board of University Studies of the Johns Hopkins University, at his exami-
nation for the degree of Ph. D., in April, 1886.

THE

SCIENTIFIC WOR K

OF

ADAM TODD BRUCE.
A SKETCH

BV

W. K. BROOKS.

ASSOCIATE PROFESSOR OF MORPHOLOGY, JOHNS HOPKINS UNIVERSITY.

Adam Todd Bruce was born March 6th, 1860; he graduated at the University
of New Jersey in iSSi ; he obtained the degree of Ph. D., at the Johns Hopkins
University in June, 1886; he was appointed an Instructor in the Johns Hopkins
University in September, 1886, and his death took place in March, 1887.

During the three years of my acquaintance with Dr. Bruce, the rapid develop-
ment of his strong character was a constant pleasure to me, but while each day
added strength to my belief that a most useful and distinguished career lay before
him as a scientific investigator, I sorrow for the loss of an affectionate friend, rather
than the death of a bright and promising pupil and student of nature.

Our relation was not that of instructor and pupil, but an intimate personal
friendship; to me he had brought his pleasures and successes, and his troubles and
perplexities, and when I received, by the first mail which reached me at the Bahama
Islands, the sad news of his death, I keenly felt the isolation which added intensity
to the loss, and cut me off from a share in the expression of affection for his memory
by his friends in Baltimore.

As I was not able to attend the meeting of his friends in Baltimore, when the
news of his death was received, it is a great pleasure to have this opportunity to
speak of the value of his scientific researches, and of the great loss which science
has sustained by his untimely death, just as he was beginning to apply his skill
and his thorough training in technical methods to the solution of new problems.

As a student and investigator Dr. Bruce was eminently characterized by perse-
verance. Always alive to all that was passing around him, he was keenly interested
in man}7 things, but no outer attraction could draw him from the subject to which
he had set his hand. Difficulties only added zest to the work, and his indefatigable
zeal never flagged, even when there seemed to be little hope of success.

All the members of our part}', at Beaufort in 1885, will remember his efforts to
encourage out-door exercise in a climate where exercise is so repulsive. As I watched,
at evening, his boyish enthusiasm in athletic games, I could scarcely believe that he
was the same person who had struggled all day long, week after week, with the
perplexing obstacles which an unfavorable climate, the absence of all adequate facili-
ties, and the most refractory character of his material opposed to his study of the
Embryology of Lamulus.

In the laboratory all his strength of mind and body was bent upon the problem
in hand, but I always felt the fullest assurance that he was in no danger of becom-
ing a narrow specialist, for whenever he was not employed in research, or in the rec-
reations into which he threw himself with equal energy and enthusiasm, his mind
was occupied with the consideration of the bearing of his special researches, both
upon the broad problems of humanity, and upon his own personal training and
development.

As regards his scientific researches, I have to speak rather of his promise than
of results achieved.

Cut off as he was, just as he was beginning to show his power to question
nature for answers to new problems, most of his results were very incomplete, and
as he had worked, here and there, upon a very comprehensive subject, the embry-
ology of Arthropods, upon which he had purposed to spend several years, he had
done little or nothing to bring together his various lines of research.

When he presented himself for examination for the degree of Ph. D. he had
made many observations upon the embryology of Insects and Arachnids, and
although he was still carrying on this research, and adding daily to his store of
information, he, at my advice, wrote out, and submitted as his thesis, a statement of
the results of his study up to that time.

During the following year he continued the work, and if he had been able to
re-write the thesis, he would have made many important additions. As he was,
by nature, incapable of dwelling upon a subject after he had reached his results, he

hurried on to new fields, and made no drawings to illustrate his conclusions. As
embryological notes must be in pictorial form, most of his work is unrecorded
and as his thesis is his single complete and illustrated work, it has seemed proper to
his instructors, pupils and friends, to publish it. Although he himself would
have wished to include his later results, this is now impossible, and it is here printed
exactly as he wrote it in the winter of 1875-6.

In the summer of 1876 he made, conjointly with myself, a study of the embryology
of Limulus. This was very complete as regards the early stages, and we had intended
before publishing, to spend one or two years more upon it, and to make it an
exhaustive account of the entire embryology and anatomy of Limulus, but the publi-
cation of Kingsley's fully illustrated paper upon the later stages of development,
and my own employment with other subjects, and Dr. Bruce's desire to study other
forms for comparison, retarded the completion of the work.

The segmentation of the egg, the formation of the blastoderm and of the germ-
layers, and the anatomy of the young larva, were all thoroughly studied, and illus-
trated by nearly a hundred drawings, and I hope that some means of publication
will soon be found.

Upon these two papers Bruce's reputation among strangers, as a student of
nature, must rest, but all who knew him personally will feel how very inadequately
this will represent what a few years more might have produced.

As soon as the news of Bruce's death was received in Baltimore a large number
of his friends and pupils met in the biological lecture room of the University to take
action indicating their grief at his death and their sympathy with his relatives.

On motion of Dr. Howell, Professor Martin took the chair ; he requested
Mr. Washburn to act as secretary.

Professor Martin said: "My friends, the University has not existed for its brief
ten years, without the hand of death showing its power among us. Those who
have gone have been of almost every degree of academic dignity, from professor to
undergraduate. We sorrow when the old man dies, but we know that he has lived
to do his work and make his name and fame. We grieve when the bright lad dies,
but we know that he has been spared years of toil and struggle. Surely no death
is so sad as that of a young man, who has just completed seven or eight years of
hard work at college and university, and is beginning to enjoy the fruits of his
labors. Such was the death which is the occasion of this meeting.

"Adam T. Bruce graduated at Princeton in tSSi, and remained there for two
years after, as graduate student and instructor. He was certainly a member of one,
and I belief e of two, of the paleontological expeditions to the far West, through
which Princeton men have done so much for science. Bruce was elected a Fellow of
this University in 1883, and appointed a Fellow by Courtesy in 1885. He took his
degree as Doctor of Philosophy last June.

"Though devoted to morphological work he was not narrow in his sympathies
or pursuits. He had a great fondness for English literature, especially the older
literature, and had a very extensive knowledge of it. At Princeton, his studies were
largely philosophical while he was an undergraduate, and after coming here he did
a considerable amount of psychological work under the direction of Professor Stanle}^
Hall. From his first appearance among us he was an active and valuable member
of the foot-ball team. .

"His handsome vigorous frame, his bright pleasant face, his manly honest look,
made all who met inclined at once to like him : and those who knew him, esteemed
him more, the more they knew him. So that between those who loved him for him-
self, and those who esteemed him for his work, and those who were his comrades in
athletics, he had among us a very large number of friends, representing many
departments and many interests in the University. Outside the University he had
many friends in Baltimore, won by his bright intellect and gentlemanly bearing.

"During the last summer vacation, Bruce was offered the post of Instructor in
Osteology and Mammalian Anatomy. In accepting, he wrote to me expressing the
great pleasure it gave him to continue his connection with the University. Those
who were his pupils know with what energy and enthusiasm he entered on his work.
But, great as his strength was, he overtasked it. The previous spring he had been
reading hard for his degree examination ; he had been at Wood's Holl at work most
of the summer ; and the completion of his research on Limulus, together with his
lectures, was too much for him. Towards the close of November he found himself
obliged to give up all work for a time, and started on an extended tour with the aim
of resting and recuperating, intending finally to stay some time in Japan and study
some of the Pacific forms of life. While passing through Egypt he was seized with
a fever — and the end came.

"It seemed to some of us, that all who knew him here would be glad of an
opportunity to meet and testify to our affection for his memory and our sorrow for

his death. For that reason this meeting has been called."

Dr. A. L. Kimball, who had been a classmate of Dr. Bruce at Princeton, then
said a few words in regard to the esteem in which Dr. Bruce had there been held,
and proposed the adoption by the meeting of the following preamble and resolution :

Whereas, We have learned with profound sorrow of the death of Dr. Adam
T. Bruce, the friend of all, and the instructor of many of us — and

Whereas, He had especially endeared himself to us by his unfailing kindness
and courtesy — it is

Resolved, That, assembled here to-day, we love the memory of the pleasant com-
panionship which existed between him and us, in all relations, official and per-
sonal ; and express our grief that he was not spared to return among us — and

Resolved, That we tender to his immediate friends and relatives our sincere
sympathy in their bereavement — and

Resolved, That a copy of these resolutions be forwarded to the editor of the
University Circulars, to the Faculty of Princeton College, and to the Family of
Dr. Bruce.

Mr. Coleman, one of Dr. Bruce's pupils, seconded the motion of Dr. Kimball,
and the following gentlemen spoke of the high regard and affection with which they
remembered him, viz: Mr. Coleman, Mr. Reese, Mr. Riggs, Professor Hall,
Dr. Hartwell, Mr. Washburn, Mr. Gilpin, Mr. Burton, Dr. Jastrow, and President
Oilman.

The Preamble and resolutions proposed by Dr. Kimball were then unanimously
adopted, and the meeting adjourned.

H. NEWELL MARTIN, Chairman.

F. L. WASHBURN, Secretary.

i

It only remains for me to signify on my own part, and in behalf of the students
who were with me at Nassau, our regret that we were unable to take part in this
meeting, and our concurrence with its action, and to express our affection for the
memory of our lost friend Adam Todd Bruce.

OBSERVATIONS ON THE EMBRYOLOQY

OF

INSECTS AND ARACHNIDS.

The important points to be determined in Insect embrj'ology are, the segement-
ation of the egg and formation of the blastoderm — the origin of the embryo and
embryonic membranes — the formation of the germinal layers — metameric segement-
ation and all connected with it — including number of appendages, nerve ganglia
and etc.

The embryology of Arachnids, or at least of spiders, in some important par-
ticulars resembles the embryology of Insects. These points of likeness will be
brought out in the detailed description of the embryology of spiders which will be
given later.

The Insects studied included representatives from the Lepidoptera, Coleoptera
and Orthoptera ; while a few incomplete observations were made on the embryology
of the Neuroptera and on the maturation of the ovum in Musca. The ova were
studied by means of sections. With opaque eggs the sectional method of study
certainly throws more light on important embryological points than mere super-
ficial observation.

The eggs were hardened in corrosive sublimate and in Perenyi's fluid. The
chorion was punctured to admit hardening and staining fluids. After staining
they were passed through the various grades of alcohol and were then transferred
to chloroform — they were then placed in a mixture of chloroform and parraffine, in
which they remained a shorter time. They were then embedded in paraffine and
were readily cut to any degree of thinness required

2 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

A detailed description of each species studied will be given, commencing, not
with the lower insect groups and spiders, but with the highest group studied, viz:
the Lepidoptera.

General conclusions and comparisons will be reserved for the closing paragraphs

of this paper.

LEPIDOPTERA.

The Lepidoptera were represented in this study by Thyridoptoyx ephemerai-
fonin's or the common bag worm.

The observations made on the embryology of this insect owing chiefly to
abundance of material were more complete than those made on any other insect
studied by the writer. Consequently the description of its development is given
first that it may serve to fill out as far as possible the incomplete accounts given of
the development of other orders of insects. The development of Thyridopteryx
from the earlier stage when all the important organs of the embryo had been formed
was carefully followed. The study of the embryology of this insect, if it has brought
to light nothing new, has, in the opinion of the writer at least, settled some important
points connected with the embryology of this group of insects.

The manner in which Thryridopteryx, commonly known as the bag worm, lays
its eggs, has been recently described by Professor Riley. A female kept in confine-
ment left the cocoon after laying her eggs in it, crawled a short distance and died.
Professor Riley states that in a state of nature the female leaves the cocoon, falls
to the ground and dies. The unsegmented ovarian egg of Thridopteryx (Fig. I & I')
consists of an outer protoplasmic stratum shading off into the central yolk spherules
which are probably invested with protoplasmic ramifications extending from the
outer stratum.

Referring to figure I, which represents a highly magnified portion of the section
represented by figure I, it will be seen that the egg is invested by a thick chorion
(CH in Fig). The chorion is perforated at the pole of the egg by a conspicuous
micropyle which is not represented in the figure.

Outside the chorion the ovarian epithelium (OE in Fig.) is seen. It con-
sists of flattened cells with large oval granular nuclei. The granular protoplasmic
stratum of the egg (P, in Figs.) lies beneath the chorion and extends between the

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

yolk spherules. The spherules (YS in Fig.) are frequently vaccuolated. No trace
of a germinal spot or vesicle was found in the mature egg. The former might
readily have been present and have escaped observation.

The early stages of segmentation ha\;e already been studied in the Lepidoptera
by Bobretzky.'1' In the earliest stages he found four amoebiform cells in the yolk
situated in pairs at opposite ends of the egg. Later the blastoderm is formed by a
multiplication of these cells : according to Bobretzky it is not formed simultaneously
over the entire surface of the egg, but is laid down first at one or more points on
the surface. This type of segmentation cannot strictly be called entolecithal, in as
much as the cells are not, in the earliest stages of segmentation, at the surface en-
closing the yolk. All the primitive uudifferentiated cells do not, according to
Bobretzky, reach the surface to form the blastoderm, but some remain centrally lo-
cated as yolk cells after the formation of the blastoderm. The earliest stages of
segmentation observed in Thyridopteryx showed several amoebiform cells in the
yolk in each cross section.

The egg at this stage must consequently contain a number of cells most of
which have not reached the surface. The nuclei of these centrally located cells
generally have indistinct boundaries which become more distinct and their contents
clearer on reaching the surface (Fig. III'). Figure IIP represents a cell at the sur-
face undergoing division. It has two nuclei with clear contents and well defined
boundaries.

The other cell represented in the figure has just reached the surface of the egg.
Its nucleus is indistinct and granular. Much reliance cannot be placed on histologi-
cal details like these, however, except in case of remarkably well preserved specimens.

In Thyridopteryx, as in the Lepidopterous insects studied by Bobretzky, the
blastoderm does not appear simultaneously over the entire surface of the egg but
appears first at one point on the surface. Figure III is a transverse section of the
egg at this stage. The yolk spherules are not represented in the figure. It will
be observed that the blastoderm is only formed over a portion of the surface.

Closely related insects are said to differ in their blastoderm formation, however,
and the simultaneous or unsimultaneous appearance of the blastoderm over

U) Bobretzky. Uber die Bildun;; iU-s Bl.i^lodrrms und dc-ii Kdmbl.-ittur bei drn Insecten.

4 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

the entire surface is consequently a fact of little embryological importance.

The first trace of the embryo is a thickening of the blastoderm at one point
where it becomes more than one cell thick (Figs. IV & IV). Whether this thick-
ening results from the division in a single layered blastoderm at that point, or from
a special aggregate of embryonic cells occurring there, could not be satisfactorily
determined. Figure IV represents a transverse section of an egg showing the early
embryo or ventral plate, E in fig. in cross section. Figure IV1 is a drawing of the
embryo at this stage more highly magnified. On either side of the embryo, as seen
in cross section, are seen folds of blastoderm (Am F. Fig. IV) which ultimately
grow together and unite over the median line of the embryo. On the union of these
folds, the embryo with the inner limb of the fold or true amnion becomes separated
from the outer fold or serosa, and comes to lie in the yolk (Fig. V), while the serosa
remains continuous with the blastoderm. At the stage represented by Figure IV
it should be noted that some cells have not reached the surface, but remain in the
yolk enclosed by the blastoderm (YC Fig. IV')- These are the so called yolk cells.
Concerning the origin and fate of these cells in different insect groups there is con-
siderable difference of opinion among investigators. Some investigators, notably
Balfour, in his work on Embryology, from evidence afforded by Bobretzky's work on
the Lepidoptera just referred to, and from other observations, claims that these
central yolk cells are in all cases undifferentiated cells which have never reached the
surface, andhomologizes them with the endoderm of Arachnids. On the other hand,
Patten (1) and probably Graber(-} believe that a stage occurs in the development of
the insect egg in which there is a blastoderm surrounding a central yolk-mass con-
taining no cells.

The yolk cells which are present in the yolk at subsequent stages must migrate
from the blastoderm if this view be correct. Probably both views are correct. In
the Lepidoptera from Bobret/.ky's observations, and from the stud}' of Thyridopteryx,
there seems to be little doubt that the first view is true. But from a study of some
other groups of insects it seems probable that the second view is also true of those
groups.

To return to the embryo, which it will be remembered had been separated from

U) Patten. Development of Phrynganids. Quarterly Journal of Microscopical Science, 1885.
(2) Graber. Archiv. f. Mikr. Anat. Vol. XV.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 5

the surface by the union of the amnion folds, and had come to lie in the yolk : Figure
V represents a transverse section of the egg at this stage with the embryo (E in
Fig. ) lying in the centre of the yolk. The embryo at this stage, as shown by study-
ing a series of sections, is watch-glass shaped, with its convex ventral side covered
by the amnion, which it will be remembered is the inner limb of the folds described,
and is constricted off from the surface of the egg with the embryo. In Figure V
the amnion covering the ventral surface of the embryo is not represented, but on re-
ferring to Figure VII, which is a cross section of an older embryo, the structure
and relation of the amnion (AM in Fig. VII) covering the ventral surface of the
embryo will be understood. The early watch-glass shaped embryo (Fig. V) consists
of a single layer of elongated epithelial cells with nuclei in some cases in process of
division on their ventral extremities, i.e. on the extremities at the convex or ventral
side of the embryo. A few nuclei with aggregated protoplasm about them lie on the
concave side of the embryo (YC Fig. V). These nuclei with the protoplasm about
them are probably cells about to migrate from the embryo into the yolk.

In the earliest stages it will be remembered that the embryo was more than one
cell thick in some places (Fig. IV). Probably some of the cells of this early embryo
remaining on its dorsal or concave surface are the cells just described as about to
migrate from it into the yolk. Before the embryo reaches the stage represented in
Figure V the yolk undergoes segmentation. The segmentation of the yolk is
probably due to the multiplication of yolk cells. Radial processes from the pro-
toplasm of these cells adhere to a number of yolk spherules. Consequently there
is formed a structure known as the yolk ball or yolk segment, consisting of a central
cell, radial protoplasmic processes of which hold yolk spherules in position. YB
Figure XI represent a cross section of a yolk ball. Similar cross sections will be
observed in other figures. Adjacent yolk balls appear in some cases to be united;
some are connected with the blastoderm and others with the amnion or embryo,
consequently the entire egg is probably a protoplasmic continuum with all its parts
in connection. The structure of the blastoderm cells is shown in Figure VII. They
are flat hexagonal cells with conspicuous nuclei. On the union of the amniotic
folds, the embryo with the true amnion is constricted off from the surface of the egg
as described. With the sinking of the embryo into the centre of the egg the yolk

6 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

comes to lie between the amnion and blastoderm. In the space between the amnion
and ventral surface of the embryo, however, there is no yolk. The amnion consists
of flat hexagonal cells resembling the blastoderm cells in appearance. Later the
embryo lying in the centre of the yolk becomes pear-shaped, with its broad anterior
end deeply cleft. The shape of the embryo at this stage will be understood from
an examination of Figs. VIII, VIII' & IX. Figure VIII, shows the outline of the
embryo at this stage as indicated by a horizontal section. Figure VIII' represents
diagrammatically a longitudinal section of an embryo at about the same stage of
development. The lateral portions of the broad anterior end of the embryo, sepa-
rated by the deep cleft are the so called pro-cephalic lobes (P C Fig. VIII). Figure
IX is a less diagrammatic reproduction of a horizontal section of an embryo at the
same stage of development as figure VIII. B L, in Figure IX stands for the
blastoderm. The space between the blastoderm and embryo in the section, was of
course filled with yolk balls, only one of which (Y B in Fig. IX) is drawn.

The cleft separating the pro-cephalic lobes is a continuation of a median groove
extending along the median line of the ventral surface of the embryo at this stage.
This groove or depression on the ventral surface of the embryo causes an elevation
on its other surface, that is, on its surface towards the future dorsal wall which has
not yet been formed. Figure VII represents a highly magnified transverse section
of an embryo at a stage when the median groove deepens, beginning to pushdorsally
the median portion of the embryo (D in Fig. VII). In subsequent stages the groove
deepens, and the pushed in portion of the embryo becomes folded off, and forms the
inner layer (i in Figs). The inner layer is not strictly invaginated, for it is cut off
from the rest of the embryo before the opposite sides of the median groove have
met. After its separation the inner layer splits into two bands. — The line of sepa-
ration of these two bands being immediately beneath the blastopore. Figure X is
a transverse section of an embryo at this stage. The inner layer (I in Fig. X), it
will be observed, is divided into bands, the line of separation being a continuation
of the median groove (BH in Fig). Figure X' represents the same section as X
drawn with a lower power, showing the relation of the embryo to the yolk balls and
blastoderm (BL). On its separation into two bands or shortly after the separation
has been effected, each band of the inner layer segments or becomes divided into a

THE EMHRYOLOGY OF INSECTS AND ARACHNIDS. 7

series of somites. The somites thus formed are solid. Figure VIII' is a diagram-
matic drawing of a longitudinal section of an embryo at this stage of development.
The somites of the inner layer are seen in longitudinal section (DS, in Fig. VIII).
It will also be seen that the body somites are marked by incisions of the outer layer
or ectoderm extending between the somites of the inner layer. The inner layer ex-
tends the entire length of the embryo.

The dotted line joining the ends of Fig. VHP represents the amnion in longi-
tudinal section. The extreme ends of the embryo, in referring to figure VIII', are
seen to be curved dorsally, that is, away from the amniotic side of the embryo. The
two terminal somites thus have their body wall formed dorsally before the dorsal
surface of the intermediate portion of the embryo is closed.

At a later stage of development the ectoderm becomes thickened on each side
of the blastopore ; consequently there are formed two thickened strings of ectoderm
which are the first indications of the nervous system. Figure XI is a drawing of a
transverse section of an embryo at this stage. NS in Figure XI represents the
thickened ectoderm lying on each side of the blastopore (BH in Fig.).

Subsequently the thickened ectoderm strings which are to form the nerve cords
become separated from the superficial ectoderm, and also from the ectoderm covering
the sides and bottom of the blastopore. The ectoderm cells at the bottom of the
grove or blastopore (BH, Fig. XII) become greatly elongated. These elongated
cells lie close to the nerve cords, but do not actually join them. Figtire XII, repre-
sents a transverse section of an embryo at this stage of development. The nerve
cords (NS in Fig.) will be seen to be separated from the surface ectoderm lining the
groove. The elongated ectoderm or what might more properly be called indifferent
cells at the bottom of the groove will also be observed in the figure. When the
nerve cords have been separated from the surface ectoderm they segment; — the num-
ber of somites corresponding to the number of body segments.

Hatscheck, (1> from his studies on the Lepidoptera, thought the elongated ecto-
derm or indifferent cells lying at the bottom of the blastopore, ultimately fused with
the nerve ganglia and formed the long commissure of the nervous system.

In this, as subsequent stages in the development of Thyridopteryx show, he

111 Beitrage. Zur Entwicklung d. Lepidoptera. Jenaische Zeitschrift Bd XI.

8 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS-

was probably mistaken: for the elongated cells in question divide in subsequent
stages, and thus give rise to cells corresponding in all particulars to migratory
mesoderm cells (I' Figs. XII & XV). All the stages in the division of these cells
and their conversion into migratory mesoderm were not traced, but after a study of
the sections drawn, and from many others not drawn, there can be little doubt that
the cell mass at the bottom of the blastopore takes no part in the formation of
nervous tissue. The non-participation of these cells in the formation of nervous
tissue is also shown from a study of the development of the nerve commissures.

In Figure XV, which represents a transverse section of a well advanced embryo,
it will be seen that a granular substance, probably formed by the breaking down of
cells, has appeared on the dorsal surface of the ganglia (NS in Fig. XV). The
cells at the bottom of the blastopore take 110 part in the formation of this granular
substance. Still later in embryonic development the commissures, both longitudinal
and transverse, can be recognized as extensions of the granular material formed near
the dorsal surface of the nerve ganglia. In these later stages there is no trace of
elongated cells at the bottom of the grove, but between the nerve ganglia and form-
ing the peritoneal coat or neurilemma of the nervous system are migratory mesoderm
cells.

Figure XXIII represents two thoracic and the first two abdominal ganglia in
median longitudinal section. Between the ganglia will be seen the migratory
mesoderm cells (I') which probably arise in the manner just described, forming the
neurilemma or peritoneal coat of the nervous system, but evidently taking no part
in the formation of commissures. Only a portion of the migratory mesoderm cells
arise from cells lying between the nerve cords : the greater portion of it is probably
derived from cells of the inner layer. These migratory mesoderm cells have a round
cell body which does not stain, containing a nucleus which stains deeply. They
are similar in appearance, though apparently not in origin, to migratory mesoderm
cells described by Reichenbach for the Crustacea.

The supra-oesophageal ganglion arises differently from the other nerve ganglia.
It appears first as a thickening of the lateral portions of the procephalic lobes.
Figure XIII is a drawing of a section through the procephalic lobes ; the lateral
portions of which (I in Fig.) are thickened considerably. On the median portion

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 9

of the procephalic lobe (LB in Fig.) is the paired outgrowth which forms the upper
lip. The dorsal surface has been partly covered by the amnion (AM' in Fig.) Yolk
spherules are seen filling the body cavity.

Figure XIII' is a drawing of a section through the procephalic lobes of a more
advanced embryo. Here it will be observed that the inner cells of the lateral
portions of the procephalic lobes (I in Fig.) have become specialized as nerve cells.
Later these specialized cells are separated from the outer unspecialized ectoderm,
and form on each side the halves of the supra-cesophageal ganglion. Figure XXII
represents a good longitudinal section of the supra and subcesophageal ganglia.

The Supra-oesophageal ganglion consists of two portions ; an anterior portion
(I in Fig.) innervating the antennae, which are not represented in the figure, and a
posterior portion (No. I' in Fig.) which sends a nerve (LBN in Fig.) to the labruni
or paired lip.

Posteriorly the second division (No. I' in Fig.) of the supra-ossophageal ganglion
forms part of the circum-cesophageal commissure, which is completed by a portion
of the mandibular division of the sub-oesophageal ganglion, (No. II in Fig).

The supra-oesophageal ganglion has its opposite halves united by two com-
missures; an anterior commissure (C Fig. XXIX Plate 3) extending beneath the
oesophagus, and a posterior commissure (C" Fig. XXXI) extending above the
oesophagus.

The sub-oesophageal ganglion consists of three pairs of closely connected
ganglia innervating the mandible, first, and second maxillae, respectively (Nos. I,
II, & III Fig. XXII). The three pairs of thoracic ganglia are larger than the fol-
lowing ten pairs of abdominal ganglia. This difference in size is apparent on con-
sulting figures XVII, XVIII, XIX, which represent good longitudinal sections of
an advanced embryo. There are ten pairs of abdominal ganglia, the last or tenth
pair being smaller than the preceeding (No. 17, Fig. XVII). It is questionable
whether the terminal portion of the abdomen which forms the so called eleventh
abdominal somite is to be regarded as a true somite or not. It has no ganglion
corresponding to it, and is formed, as stated, by the dorsal flexure of the posterior
end of the embryo, and consequently has its body walls formed on all sides at an
early period in the manner described.

10 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

The nervous substance consists principally of round nuclei. The commissures
consist of the granular material before described. The nerve fibres consist of
similar granular material (Figs. XXII Si XXIII). After the separation of the
nerve ganglia from the ectoderm, the blastopore closes entirely (Fig. XV). A closer
examination of figures XVII, XVIII, XIX, may make clearer the number and
relations of the nerve ganglia.

Figure XVII is a nearly true median longitudinal section The cesophageal
invagination (OE in Fig ) will be seen directly posterior to the upper lip (LB in Fig).
Following the three large thoracic ganglia are the ten smaller abdominal ganglia,
(Nos. 8-12, Fig). The last abdominal ganglion will be seen to be smaller than the
preceeding ganglia.

The nature and origin of the last or eleventh abdominal somite has already
been described. Its dorsal wall extends forward beyond its lateral walls. This dor-
sal extension of the last somite will be seen, on referring to figures XVII, XVIII,
& XIX, to be deeply grooved. This groove is a continuation of the median groove
or blastopore over the dorsal surface. It becomes deeper, and is finally invaginated.
On being separated from the dorsal surface of the body, the invaginated portion
forms the anal invagination. The cesophageal invagination also occurs in the
middle line of the body ; unlike the anal invagination, however, it is formed not by
the infolding of a portion of the median groove, but by a simple, vertical ingrowth
(OE Fig. XVII). The cephagus is formed before the anal invagination is completed.

Back of the oesophagus there occurs an aggregate of migratory mesoderm
cells, wrongly regarded by Hatscheck as endoderm. (1> The cesophageal invagination
occuring in the median line of the body pushes in the cells of the inner layer at
that point, and extends dorsally and posteriorly, invested by these cells, (Figs.
XVIII, XXII). On the dorsal surface of the cesophageal invagination there is
formed a thick string of mesoderm cells. The outer cells of this string adhere to-
gether, while the inner cells separate from them ; thus is formed a tube containing
cells. This tube is the heart (H in Figs. XXVIII Plate 2, & XXIX, XXX, XXXI
Plate 3).

Probably the inner cells of the tube form the blood corpuscles and plasma.

Beyond the cesophageal invagination the heart does not appear to be formed
from a solid string of cells, but is apparently formed from separate mesoderm cells

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. II

which come together on the dorsal surface, over the alimentary tract, flatten out, and
adhere at their edges (Fig. XXXI). A more detailed description of figures XXVII,
XXXIII will make preceeding references clearer.

Figures XXVII & XXVIII represent sections through the head. The head
ganglia are not observed in these figures, because the sections were made through
the ventral surface of the head outside the nervous substance.

The oesophagus (OF in Figs.) is represented in cross section, as well as the
mandibles (MDj and labrum (LB).

The cesophageal epithelium is represented as surrounded by mesoderni carried
in with it as it grew inwards from the median line. The mesoderm surrounding
the cesophageal ingrowth, on its dorsal surface, will be seen to be hollow, thus form-
ing what appears to be the first trace of the heart in the head region (H in Figs).

A line joining numbers I and II in figure XVII, will represent approximately
the plane of the sections represented by figures XXVII and XXVIII. Succeeding
sections of the same series represented by figures XXIX, XXX and XXXI, were
cut in planes parallel and internal to the imaginary line joining numbers I and II
in figure XVII. In figure XXIX portions of the supra, and sub-oesophageal ganglia
are represented (Nos. I and II in Fig.) The sub-oesophageal commissure of the
supra-cesophageal ganglion is also represented (C in Fig. XXIX.)

The structure marked GL in the figures is a prolongation of the salivary gland
which originates at the base of the mandibles (GL, Fig. XXVIII). Figure XXX
represents a section through a plane internal to that represented by the preceeding
figure. The section figured has been carried through the base of the brain, cutting
the circurn-cesophageal commissure (C' in Fig.)

Figure XXXI represents a section internal or dorsal to the preceeding, showing
the supra-cesophageal commissure of the supra-cesophageal ganglion (C" in Fig.)

Figure XXXII is a drawing of a cross section of the embryo back of the
cesophageal ingrowth. The epithelium of the midgut (IE in Fig.) is represented
enclosing yolk cells and yolk spherules (YS, YC in Fig.)

It may be here in place to describe the sense organs and appendages of the
head. The head which includes the portion of body containing the supra and sub-
cesophageal ganglia is distinctly separated from the thorax.

12 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

Figures XXIX' and XXX' Plate II are diagrammatic representations of the
head viewed from its under surface and laterally. The upper lip is seen to be dis-
tinctly bilobed. The first maxillse are longer than the other cephalic appendages.

The sense organs are what may be termed compound simple eyes. They are
represented in figures XXVII and XXVIII. Each ocellus, (OC in figures,) appai'-
ently arises as an involution of cells from the inner surface of the ectoderm. These
involutions form sacks apparently containing no cells ; but the lumen or cavity of
the sack is filled with granular material. Migratory mesoderm cells probably in-
vest these ocelli.

. It does not appear that any part of the ocelli is formed from the nervous
system. The observations which were made were not conclusive on this point
however.

An invagination (GL, Figs. XXVII & XXVIII) occurs at the base of the
mandible which apparently forms the salivary gland.

It remains now to describe the origin of the endoderm and the closure of the
dorsal surface of the body. The appendages and tracheal invaginations arise in the
customary manner.

A portion of the inner layer on each side of the embryo becomes separated
from other parts of the inner layer (IE Figs. XV, XXIV, XXV.)

These portions of the inner layer which may be called endoderm grow together
and imite first on what is the ventral surface of the alimentary tract (Fig. XXV.)
They then extend dorsally and enclose the yolk and the yolk cells which lie in the
body cavity.

Figure XXXII is a cross section, already explained, of an advanced embryo in
which the epithelium of the midgut is fully formed. The yolk cells lie, with the
yolk, in the digestive tract and certainly do not form any considerable portion, if
any, of the endoderm. A similar formation of endoderm has been described by
Tichomiroff ' for other Lepidoptera. His conclusions have been disputed, notably
by Balfour, who claims that the yolk cells are the true endoderm. Before its
closure by the eudoderm, the yolk is enclosed by migratory mesoderm cells (I'
Fig. XXIV.) Inside these cells the endoderm grows round and encloses the yolk.

(i) Zool. Anzeiger. II Jahr. no. 20.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 13

The migratory mesoderm cells surrounding the digestive tract may have been
wrongly regarded by Hatscheck as endoderm since, as already stated, he seems to
have mistaken other migratory cells for endoderm.

The closure of the dorsal surface of the body is interesting. The amniotic
folds (AM' Fig. XXV) grow dorsally more rapidly than the ectoderm of the body
walls. These opposite amniotic folds finally unite and the inner limb of each fold
forms a portion of the dorsal surface of the body. The outer limbs of the amniotic
folds unite to form a sack in which the entire embryo lies. This sack contains no
yolk, but is apparently filled with fluid. Just before the union of the opposite amni-
otic folds, there is formed what may be termed a dorsal organ (Fig. XXVI,) though
it does not correspond to the dorsal organ described for some other insects.1

A study of figures XV, XXIV, XXV, XXVI, will make clearer what has been
said concerning the origin of the endoderm and the union of the amniotic folds on
the dorsal surface of the body.

Figure XV has already been described. It represents a transverse section of a
well advanced embryo. The tracheal invagiuations are shown in the figure (TR) to be
iuvaginations of ectoderm. The appendages are outgrowths of the body cavity
occurring between the tracheal invaginatious and the nerve ganglia.

On comparing figure XV with figure XI or with preceediug figures it will be
seen that in the former, which represents a transverse section of a more advanced
embryo than the latter, the amniotic folds have extended farther towards the dorsal
surface of the figure. Figures XXIV, XXV, XXVI, are drawings of an older
embryo than that represented by figure XV. Here it will be seen that the inner
portions of the amniotic folds (AM' in Figs.) are growing together and forming the
dorsal wall of the body as described.

Figure XXVI represents a section of an advanced embryo just before the union
of corresponding portions of the amniotic folds of opposite sides, when the outer
and inner portions of each fold are united by intermediate amiiiou which forms a
sort or dorsal organ.

Inasmuch as no yolk lies between the embryo and the amnion it will readily be

01 Brandt. Beitrage zur Entvvicklungsgeschichte d. Libellula.

14 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

seen on referring to the figures that on the union of the outer portions of the oppo-
site amniotic folds the sack enclosing the embryo will contain no yolk.

How it happens that folds are formed by the dorsal growth of the amnion will
be understood by referring to figure XV. In describing the origin of the endoderm
reference was made to figures XV, XXIV, XXV, as illustrating stages in the forma-
tion of the rnidgut.

In figure XV the most dorsal portion of the inner layer (IE in Fig.) on oppo-
site sides begins to be constricted off from the rest of the inner layer.

The process of constriction has been completed in figure XXIV. The portions
of the inner layer thus constricted off (IE Figs. XXV, XXVI), on each side grow
together forming first the ventral surface of the digestive tract, thence they ex-
tend dorsally, and shut in the digestive tract on all sides.

To sum up the embryology of Thyridopteryx : It was found that on the form-
ation of the blastoderm some of the cells, probably, do not reach the surface, but re-
main in the yolk as yolk cells, which take little, if any, part in the subsequent form-
ation of endoderm.

The nervous system arises in the customary manner. The median ingrowth
between the nerve ganglia takes no part, as Hatscheck thought, in the formation of
the commissures. The supra-cesophageal ganglion is double. It has a double
commissure uniting its opposite halves.

The ocelli probably arise from ectoderm independently of the nervous system.
The true amnion on the union of its folds forms a portion of the dorsal surface of
the body. Before the union of its opposite folds what might be described as a dorsal
organ is formed.

NEUROPTERA.

Chrysopa was the representative of this group which I studied. The obser-
vations made on the embryology of this insect, owing to lack of material, were very
incomplete.

Figure XXXIII represents a transverse section of the head showing the position of
the upper lip and the antennae. The upper lip is distinctly bilobed. Its two lobes
are not, however, well shown in the figure.

The abdomen of the neuroptera, according to Packard, consists of eleven somites

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 15

including the last somite or the so called post abdomen. The number of somites is
not well shown in figure XXXIV, which is not strictly median. The last or eleventh
abdominal somite is covered dorsally and corresponds to the last somite of Thyridop-
teryx. According to Packard, the last somite, or post abdomen, consists in Diplax, n>
of several somites.

This segmentation of the post abdomen was not observed in Ch^sopa through
the observations made on this insect were confessedly very incomplete.

COLEOPTERA.

The embryology of Meloe, the parasitic beetle, was studied as representing the
Coleoptera. The earliest stage obtained showed a surface blastoderm enclosing
central cells (Fig. XXXV). Later the blastoderm cells become more columnar,
(Fig. XXXV) and there are apparently no cells in the yolk; though unless a com-
plete series of sections of a very well preserved egg be obtained, it is impossible to
be positive on the latter part. Apparently, however, a stage occurs in which there
are no cells in the yolk, consequently the yolk cells, numerous in latter stages,
probably arise from the blastoderm or from the embryo. The earliest embryo (Fig.
XXXVI) showed, in cross sections, incipient amniotic folds on each side with a
median blastopore.

It becomes separated from the surface on the union of the amniotic folds, but
does not sink into the centre of the egg as does the embryo of Thyridopteryx, (Figs.
XXXVII, XXXVIII).

Amcebiform cells (YC' in Figs.) can be seen on the dorsal or yolk side of the
embryo at this stage. These are probably yolk cells which arise by division from
the embryo and subsequently migrate into the yolk. The median invagination be-
comes constricted of and forms the inner layer, (Fig. XXXIX).

No observations were made on the origin of the nervous system or on the origin
of the endoderm. The nerve ganglia become separated from the surface ectoderm,
and in advanced embryos all trace of the blastopore has disappeared as in Thyri-
dopteryx (Fig. XL).

(!) Packard. Guide to the study of Insects pp. 56-57.

l6 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

The upper lip is double and the antennae have the same position as in Thyri-
dopteryx and Chrysopa.

ORTHOPTERA.

Mantis and the grasshopper were the orthopterous insects studied.

The early stages in the development of Mantis were obtained and the latter
stages in the development of the grasshopper. The developmental history of the
group derived from the study of both insects is not complete, though some obser-
vations of importance were made.

At the earliest stage obtained, the egg of the grasshopper consisted of large
angular yolk masses like those of the spider (Fig. XLV). The yolk is enclosed by
a very thick membrane with concave hexagonal depressions on its outer surface
(CH, Fig. XLV, XLVI). Within this outer chorion there is another thinner mem-,
brane.

In what is apparently the earliest stage, a portion of the yolk is much vesicu-
lated (E' Fig. XLV). Near or in this portion of the yolk a few nuclei occur, (YC,
Fig. XLV). At what appears to be a later stage, the yolk consists of pyramids with
their apices at the centre of the egg. On the bases of the pyramids at the surface
of the egg are nuclei or nucleated cells. Within the yolk at this stage, no nuclei
were found, though they might have been present and have escaped observation, as
the sections were in many cases considerably broken.

Later in development the yolk pyramids break up. This breakage is probably
effected by the vesicles in the pyramids uniting and consequently causing a sepa-
ration of the yolk substance along the line of their union.

These pyramids recall the yolk pryamids of Astacus, described by Reichenbach.(1)
They might also be compared to the yolk columns described by Ludwig ('~' for the
spider's egg. Other investigators, however, have not confirmed Ludwig's obser-
vations, but find that all the cells are not at the surface, as he claimed, but that some
remain centrally located in the yolk. The yolk of the Mantis egg is very like the
yolk of the grasshopper egg. The yolk of the higher orders of insects studied, con-
sisted of rounded spherules which were often much vesiculated.

(i) Die Embryoanlage u. erste Entwicklung d. Flusskrebses. Zeit f. wiss. Zool. Vol. XXIX.
(-) Ueber die Bildung des Blastodermes bei den Spinnen.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 17

The earliest trace of the embryo was obtained from sections of the eggs of the
Mantis. At this early stage the embryo is a mass of undifferentiated cells lying on
the surface of the egg. (Fig. XLJ, E.) The blastoderm is not formed at this stage.
Some cells remain in the yolk; whether they are all at the surface at some subse-
quent stage could not be determined.

Figures XLJI & XLJII, show two stages in the formation of the amniotic
folds of Mantis. It will be seen that the amnion, as seen in cross sections,
arises on each side of the embryo as folds of blastoderm, which meet and unite
over its middle line.

When the union of the folds is effected, and the embryo is separated from the
surface and covered veiitrally by the amnion, the inner layer is formed, as in Meloe
and Thyridopteryx, as an ingrowth from the median line of the embryo. Whether
the inner layer gives rise to both mesoderm and endoderm could not be determined
in the case of Mantis as no advanced embryos of this insect were studied. Figure
XLIV is a drawing of a transverse section of an early Mantis embryo, showing the
origin of the inner layer ' ' from the median groove. From a study of more
advanced grasshopper embryos, however, it seems probable that the yolk cells do not
take part in the formation of the endoderm ; consequently that layer is formed in
the grasshopper as in Thyridopteryx from the inner germ layer. Ayers u/) in his
memories on CEcanthus, an Orthopterous insect, described undifferentiated cells
which never reach the surface to form blastoderm but remain in the centre of the
egg as yolk cells. Unlike the corresponding cells of Thyridopteryx, these cells,
according to Ayers, take part in the formation of the endoderm. The serosa or
outer embryonic membrane is taken into the body cavity of the embryo through its
enclosed dorsal surface, and also forms parts of the endoderm. Nothing at all cor-
responding to such an absorption of embryonic membranes was observed in the
advanced grasshopper embryo. This embryo is perhaps best understood by following
the series of cross sections represented by figures XLVII-LXV. Most of the sec-
tions figured were selected from a single series from an advanced embryo. They
were drawn in order, from before backwards. Figure XLVII, represents a cross
section of the head of an advanced grasshopper embryo. Reference to the meaning

(*> Ayers. .Memoirs of the Boston Society of Natural History, Vol. Ill no. VIII.

l8 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

of the letters will explain the details of the figures. C" in the figure, is the cross
commissure of the supra-cesophageal ganglion, crossing above the oesophagus. At
a point on the dorsal surface of this region of the embryo the yolk spherules lose
their definite outlines, and appear to run into a homogeneous mass, (YS in Fig.), in
which are cells, which are, apparently, migratory mesoderm cells.

Figure XLVIII represents a section of the supra-cesophageal ganglion back of
the section just described. It shows the position of the antennae and labrum.
These correspond in every way to the antennae and labrum of other insects which
have been described.

The amniotic folds of opposite sides have not met in this region. It will be
observed that the amnion at this stage does not cover the ventral surface of the
embryo, but has been apparently absorbed at all points except the dorsal extremity
of the body wall, where it persists as shown in the figures.

Figure XLJX represents a section, posterior to those previously described,
through the anterior part of the oesophagus. I in the figure represents the inner
layer given off from the median line at this point. It has been carried inwards by
the oesophageal imagination.

Figure L is a drawing of a succeeding section in which the circum-cesophageal
commissure of the supra-cesophageal ganglion is represented as surrounding the
oesophagus. In succeeding sections this commissure disappears.

In figure LI it will be observed that the mesodermic ingrowth pushed in with
the inward growth of the oesophagus forms a hollow thickening on its dorsal surface.
This mesodermic thickening may be the rudiment of the heart in the head region,
as it apparently corresponds to a similar thickening which forms the head portion
of the heart in Thyridopteryx.

Figures LII-LVII represent sections through the mandibular region. In figures
LV, LVI, LVII, it will be noticed that the amniotic folds have met and united on
the dorsal surface of the body. In figure LV the thickening of mesoderm on the
dorsal surface of the oesophagus previously referred to as the possible mass from
which in the head region the heart originates, is again shown.

In figures LV, LVI, LVII & LVIII, the ganglion, marked No. II, is the man-
dibular portion of the sub-oesophageal ganglion. Figures LIX, LX, LXI, represent

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 19

sections through the first pair of maxillae. In this region the oesophagus terminates.
The amniotic folds and ectoderm have not covered the dorsal surface of the
body, which is, however, closed by mesoderm cells.

The supra-oesophageal ganglion extends through the first maxillary region of
the head. It will be noticed that the ganglion is separated from the superficial
ectoderm which is particularly thick on its dorsal surface. Possibly the compound
eyes are formed from this thickened portion of ectoderm.

The maxillae have each two lobes outside of and at the base of the main axis
of the appendage. (XL- in Fig.)

These lobes recall, though they are probably not homologous with, the exopo-
dite and epipodite of the Crustacean appendage. Similar lobes have been described
by Patten for the maxillary appendages of Blatta.

Figure LXII is through the region of the second maxilla. The body wall is
here closed dorsally by mesoderm and the amniotic folds extend farther towards the
ventral surface than in preceding sections.

Figure LXIII represents a section through the maxillary region of what was
probably a somewhat older embryo.

Figure LXIII is a portion of the same section more highly magnified. No. Ill
in the figure represents the ganglion of the second maxilla. I represents the meso-
derm which here as elsewhere covers the dorsal surface before the latter is closed by
the arnuion and ectoderm.

Beneath the dorsal mesoderm lie large cells with granular protoplasm and large
deeply granular nuclei. These may be described as blood cells (BC. in Fig.) In
the spider somewhat similar cells form the plasma and corpuscles of the blood.

Figures LXIV and LXV represent sections through the thoracic region. These
figures hardly require explanation. It will be observed that the amnion, though
more extended than in preceeding figures of this series, does not cover the ventral
surface of the embryo.

To sum up briefly the observations made on the embryology of the Orthoptera, —

At one stage all the cells are probably on the surface at the bases of the yolk

pyramids. Yolk cells must then arise by migration from the surface cells. The

yolk cells probably take no part in the formation of the endoderm ; for when the

2O THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

body wall was closed dorsally by mesoderm no yolk cells were observed in the body
cavity, though they were present in the yolk. The early embryo of Mantis is a
mass of uiidifferentiated cells lying on the surface of the yolk like the primi-
tive cumulus of spiders.

The amniotic folds and inner layer arise as in other insects described. At
later stages the amnion is incomplete, not covering the ventral surface of the embryo.

The maxillae have additional lobes at the base of the main axis of the appendage.

Korotneff (1) has made a stiidy of the embryology of Gryllotalpa by means of
sections. An abstract of his work is given here. On comparing it with the results
arrived at from the study of the Orthopterous insects, mantis and grasshopper,
there appear to be some important points of difference which can hardly be due
entirely to the more complete developmental history obtained by Korotneff.

In the earliest stage observed by Korotneff there were four amoebiform cells in
the yolk. On the multiplication of these cells and their migration to the surface, the
blastoderm is formed. A stage occurs in the formation of the blastoderm during
which there are no cells in the yolk. Yolk cells subsequently migrate from the
surface; Korotneff states that before the primitive cells, are converted into blastoderm
cells, they have for a time no nuclei at all. No such disappearance of nuclei was
observed by me in the Orthoptera studied, or in the blastoderm formation of other
insects. It might, however, have been overlooked. The granular ill-defined nuclei of
the undifferentiated cells appear, from observations made in Thyridopteryx, to
become less granular and more vesicular when these cells reach the surface.

The mesoderm of Gryllotalpa, according to Korotueff, arises, not by the separa-
tion of a median ingrowth from the outer layer, but by delamination on each side of
the median groove. This is certainly not the mode of origin of this layer in mantis.
Its origin in different Orthopterous insects ma}- differ, however, for Korotueff's
observations seem to have been carefully made. The embryonic membranes of
Gryllotalpa arise, as in mantis and other insects described, as folds of blastoderm on
each side of the embryo, which meet and unite over its middle line. The united
inner limbs of these folds or the true amnion folds rupture in late embryonic life and
are absorbed.

(') Korotneff. Die Embryologie der Gryllotalpa. Zeit. f. wiss. Zool. 1884.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 2 I

It will be remembered that the amnion, in the late stages of development of the
grasshopper, is absorbed, with the exception of a portion attached to the dorsal
extremities of the body walls of the embryo.

The outer membrane or serosa, which is equivalent to the original blastoderm
minus that portion which forms the amnion, is also absorbed. But before this
absorption occurs a layer of yolk cells is formed beneath it. This layer becomes
specially thickened on the dorsal surface of the yolk and forms a dorsal organ. A
similar dorsal organ has been described for other insects. It is, however, totally
different from the dorsal organ described by Brant for Libellula. It will be remem-
bered that no corresponding structure occurs in the development of Thyridopteryx.
The dorsal organ becomes concentrated at one point on the dorsal surface. Its cells
multiply and migrate into the yolk which they digest or prepare for digestion. The
body walls are then closed dorsally; the dorsal organ for a time remains as a tube
on the surface of the body. A similar behavior of the dorsal organ has been noticed
in other insects. In the latest stages which I studied, it will be remembered that
in the grasshopper the serosa or blastoderm persisted on the surface of the yolk ;
and there appeared to be nothing present or in process of formation which corres-
ponded to a dorsal organ.

According to Korotneff, the mesoderm is distinctly divided into splanchnic and
somatic portions The former enclosing in the body cavity the yolk which is
arranged in pyramids with yolk cells at their bases. No marked separation of
mesodermic layers was observed in the grasshopper.

Migratory mesoderm or mesenchyme cells arise from the mesoderm. These
resemble similar cells described in Thyridopteryx. Blood cells arise from mesoderm
on the median ventral surface. Their histological structure is not described by
Korotneff. They may, however, be like the cells described as blood cejls in the
advanced grasshopper embryo.

The heart is formed from folds of the somatic layer of mesoderm growing
together over the dorsal surface somewhat in the manner described by Dohrn. "' The
nervous system arises according to Korotneff as a single string which becomes sep-

di Notzien zur Kenntniss d. Insectenentwicklung.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

arated by a median depression, which is formed where the first groove or primitive
furrow had previously occurred.

The cells forming the floor of the depression separating the nerve strings take
part in the formation of the nervous substance. This does not agree with my obser-
vations on the origin of the nervous system in Thyridopteryx, where the median
groove did not disappear before the formation of the nervous system. The latter
did not appear first as a single string. The cells lining the median groove separa-
ting the nerve strings did not apparently form any of the nervous substance.

DIPTERA.

Before going to the embryology of spiders, it may be well to insert here some
observations made on the maturation of the egg in Musca (domestica?).

Each ovarian tube of the fly's ovum is divided, at certain stages at least, into
five divisions, viz : the end chamber (EC Fig. L,XVI,) and four following chambers,
the largest (CB, IV, Fig. LXVI.) at the end of the tube.

The ovarian tube consists externally of a membraneous peritoneal sheath in
which nuclei are embedded (FE in Fig. LXVI.)

The end chamber contains nuclei, the histological structure of which could not
be determined.

Korschelt, from observations on Musca vomitaria, finds that the egg cells and
nutritive cells which form the contents of succeeding chambers arise from the larger
nuclei contained in the end chamber, while the smaller nuclei of the end chamber
assume a superficial position in succeeding chambers, enclose the germinal vesicle
and nutritive cells and become the nuclei of the ovarian epithelium.

Will, (1 in an elaborate article on the origin of the yolk in insects, claims
that the nuclei of the end chamber (the so-called ooblasts) give off portions of their
substance, which form in some cases the nuclei of the ovarian epithelium and in
others the nuclei of the nutritive cells.

The remains of the nucleus of the ooblast then form the germinal vesicle and spot.

From the incomplete observations made on the embryology of Musca, it seems
that Korschelt's view is the correct one. There does not appear to be any budding

") Will. Zeit f. wiss Zool. 1885.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 23

from the nucleus of the ooblast to form ovarian epithelium and nutritive cells ; on
the contrary, the nuclei after leaving the end chamber are arranged in circular
masses, the outer nuclei of which are smaller than the enclosed nuclei. The latter,
in more advanced egg chambers, can be recognized as the nutritive cells, while one
differing from the nutritive cells in appearance becomes the germinal vesicle.

The germinal vesicle was not observed in the younger egg chambers, that is,
those nearest the end chamber of the tube. Each chamber of the tube with the ex-
ception of the end chamber forms a single mature egg.

Figure LXVII represents the last chamber of the egg tube in a more advanced
stage than that represented by CBIV Fig. LXVI.

The ovarian epithelium (OVE in Fig ) has become quite columnar at the end
of the egg chamber towards the outer end of the ovarian tube. The nutritive cells
at this end of the chamber have broken down, forming yolk, (Y in Fig. LXVII).
The yolk consists of small spherules, in every respect similar to spherules found
in the nuclei of the nutritive cells (N, C in Fig.) The yolk of the mature fly's egg
is of a similar character. In all probability then the yolk consists of the broken
down nuclei of the nutritive cells. In the mature ovum it is surrounded by a proto-
plasmic sheath which is probably derived from the protoplasm separating the nuclei
of the nvitritive cells.

Figure LXVIII represents a longitudinal section of a more advanced stage in
which the germinal vesicle is seen lying in the yolk. It consists of finely granular
protoplasm which, unlike the nuclei of the nutritive cells, does not stain. At this
stage, the vesicle has a definite boundary. It is eccentric in position, lying near the
ovarian epithelium, which at this stage, has begun to excrete the chorion (CH in
Fig.) on its inner surface. The ovarian epithelium apparently takes no part in the
formation of the yolk. After the formation of the chorion it adheres as flattened
hexagonal cells to the outer surface of the latter even when the egg is laid.

Figure LXIX is a drawing of a cross section through an egg chamber at the
same stage of development as that represented in figure LXVIII. In this figure the
deeply stained germinal spot is represented in the centre of the germinal vesicle
(GS in Fig).

The germinal spot is not homogeneous throughout, but has in its centre what

24 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

appears to be a swelling with a depression on its apex. However that may be, the
central portion of the spot has a different refractive index from the peripheral portions.

Figures LXIX and LXIX' represent a later stage in the maturation of the egg
drawn with a low and a high power.

The germinal vesicle at this stage has lost its definite boundary and shades off
into the yolk spherules.

The ovarian epithelium has grown round and nearly closed the outer pole of
the egg.

Through the opening still left by the ovarian epithelium the remaining un-
absorved nutritive cells are probably taken in. The germinal spot, at this stage,
differs from the same structure in earlier stages. This, however, may be due to the
different action of hardening fluids or to differences in preservation, etc.

Sections of the mature egg showed no traces of the germinal vesicle or spot.
The latter, as we may believe from Hertwig's (1) observations on the maturation of
the ovum, might readily persist and yet be overlooked, owing to its small size and
to the confusion resulting from the number of yolk spherules. Will (~} in the article
referred to describes the migration of the vesicle of this egg to the periphery where
it loses it boundary and runs into the yolk. He thinks the germinal spot also

disappears.

ARACHNIDA.

The observations made on the embryology of spiders were not complete, but
they have brought to light some interesting points not heretofore noticed or suffici-
ently emphasized. No observations were made on the early stages of segmentation.
The observations of Ludwig on the early stages of development do not, as before
stated, fully accord with those of other observers. Schimkewitsch agrees with
Balfour in thinking that on the formation of the blastoderm some undifferentiated
cells remain in the yolk cells. The observations recorded here do not settle this
point however.

Before the formation of the blastoderm, the so-called primitive cumulus is
formed on the surface of the egg.

At this stage there are cells undergoing division in the yolk (YC Fig. LXXV).

O Hertwig. Morphologishes Jahrbuch Vol. III.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 25

Balfour says the cumulus appears after the blastoderm has been fully formed. This
is not true of the species of spider studied by me. The primitive cumulus con-
sists of a mass of undifferentiated cells, extending well into the centre of the egg.
(Fig. LXXY. ) Their histological structure is well preserved in the section which
figure LXXV represents. The cells have large granular nuclei. The protoplasm
is marked by granular radii extending from the nucleus to the periphery of the cell.
These cells are primitive and unspecialized. They resemble the yolk cells which
are certainly like the cells occurring in the early stages of segmentation before
specialization of tissue has taken place. It is not important to enquire whether the
cumulus results from the division of surface cells or is formed by the accession of
cells from the yolk. It is very probably formed both by the division of cells which
have reached the surface, and by the addition to these of yolk cells. It is important
to note, however, that it consists of undifferentiated cells, and is formed, in the
species of spider which I studied, before the blastoderm is completed. In Limulus
the first trace of the embryo is also a mass of undifferentiated cells lying at the sur-
face of the egg."1

Figure LXXVI represents the next stage obtained after the cumulus stage. It
is a longitudinal section showing nine or ten rnesoblastic somites which are not
hollow at this stage.

It seems that the cumulus must take part in the formation of some if not all of
these somites.

In Limulus the mesoderm arises in part at least from the inner cells of the
cumulus.

Balfour states that the somatic mesoderm arises from the ectoderm and the
splanchnic mesoderm from the yolk cells. In the early stages, however, the meso-
derm is not divided into two layers. Figure LXXVI I represents yolk cells near the
surface undergoing what appears to be endogenous division. This corresponds to
an endogenous division observed by Reichenbach as taking place in the yolk cells of
Astacus.

Figure LXXVIII represents a longitudinal section of an embryo more advanced
than that represented by figure LXXVII. Here the six thoracic and four provisional
abdominal appendages have appeared.

U) University Circulars, 1885.

26 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

The mesodermic somites are hollow, the somatic portion lining the cavities of
the appendages, the splanchnic portion not entering the appendages. At this stage
there is posterior to the last thoracic appendage a swelling marked OP in figure
LXXV1II. Although this was not observed in all cases, it is apparently a normal
structure, for it is present in later stages. It corresponds in position to the
operculum of Limulus. Figure LXXIX is a drawing of a longitudinal section of
a still more advanced embryo. The first two abdominal appendages ( A & B in Fig.)
are not well marked but the mesoblastic somites corresponding to them are well de-
fined (A' & B' in Fig.) Figure LXXIX represents a more highly magnified portion
of the same section.

The two abdominal appendages, A & B, are better defined in reality than in the
figure.

It will be seen that the ectoderm covering the two appendages, A & B, is
columnar, each cell having a well defined chitinous boundary and a square nucleus.

These cells correspond, in histological structure, to those forming the laminae of
the lung book of the adult spider as described by McLeod.'1' The histological
structure of the cells of the gill lamellae of Limulus is quite similar.

On the anterior surface of the appendage, A, will be seen a fold (I in figure.)
If we imagine the appendage to be pushed further in and come to lie entirely in the
lung cavity L, B, the fold on the anterior face of the appendage and others which
may arise there will then lie on the anterior wall of the lung cavity, but these folds
will then be directed backwards and not forwards as before the involution of the
appendage and will consequently correspond in every way to the laminse of the lung
book. All the stages in the involution of the appendage were not traced, but there
can be little doubt that the lung book of the spider results from the involution of
embryonic abdominal appendages. Such an involution of appendages to form the
lung book has been suggested by Lankester ('2) on theoretical grounds, but not from
observation. The involuted appendages are covered by the structure described as
the operculum. (OP, Fig. LXXIX'.)

Some remarks on the supra-oesophageal ganglion and structures connected with
it will conclude the observations made on the embryology of spiders.

(1) Archiv. de Biologic, 1884.

(2) Quarterly Journal 1885.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 27

The supra-cesophageal ganglion was studied by transverse and longitudinal
sections. Figures LXXI, LXXII, LXXIII, LXIV, represent transverse sections of
the brain of an advanced embryo cut backwards from the anterior part of the
cesophageal imagination. On each side of the brain will be seen folds (AM in
figure) which correspond to the amniotic folds of insects. In the last transverse
section (Fig. LXXIV,) the outer limbs of the folds have separated from the inner
limbs and have united. The thicker inner limbs of these folds have not united
but in other respects they correspond completely to the true amnion of the insect
embryo.

It will be seen from a study of longitudinal sections (Figures LXXX, LXXX'
LXXXI, LXXXIP, LXXII) that these folds occur in the head region only, and cover
but a part of the supra-cesophageal ganglion.

Figure LXXX, is a drawing of a longitudinal section of an advanced embryo
laterad of the median line. Figure LXXX' represents the cephalic portion of the
same section drawn with a higher power. AM in the figure represents the so-called
amniotic folds in cross section. No. I is the anterior portion of the supra-oesopha-
geal ganglion. No. I' is the posterior division of the same ganglion. MD repre-
sents the mandible. The portion of the brain above it represents the mandibular
division of the supra-cesophageal ganglion.

The large cells (B, C, in the figure) with granular protoplasm and well defined
nuclei may be termed blood cells. On the fusion of these cells their protoplasm
appears to form blood plasma; and their nuclei, with an investment of protoplasm
in some cases, the blood corpuscles.

Figure LXXXI represents a section of the same series farther towards the
median line of the embryo.

LXXXI' is a drawing of the same section more highly magnified.

Here it will be seen that the inner amniotic fold has separated from the outer
and adheres closely to the anterior portion of the supra-cesophageal ganglion. The
outer limb of the amnion fold extends over the supra-cesophageal ganglion to its
mandibular division.

The blood cells are kept from passing out between the amnion folds by meso-

28 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

dermic strings extending from the niesoderm which invests the supra-oesophageal
ganglion to the outer amnion fold.

Figure LXXXII represents a section of the same series close to the median line.
Here it will be observed that the inner amniotic fold has separated from the outer
and lies close to the supra-oesophageal ganglion, to which it is apparently attached.
Blood cells have broken down between the two folds and have thus formed blood
plasma and blood corpuscles. Blood cells are numerous in the dorsal surface of the
embryo, where they generally lie between the two layers of mesoderm, indicating
perhaps that the blood once circulated through the body cavity. Balfour describes
the groove regarded here as amnion, as a depression on the anterior and lower sur-
face of the procephalic lobes. Metschinkoff,0) in his article on the embryology of
scorpions, has described a fold which closes in the ventral surface of the brain of the
scorpion, and is probably similar to the amnion of spiders. If this fold were
extended the whole length of the embryo it would correspond in every way to the
amnion of insects.

The observations presented here on the embryology of spiders can be very
briefly summarized.

The primitive cumulus, consisting of undifferentiated cells, appears before the
blastoderm is fully formed.

It probabl}' forms a considerable part of the mesoderm. According to Balfour
it occupies a position where the caudal lobe of the embryo subsequently appears.

The mesoblast, or part of it at least, must then grow forward from this posterior
part of the embryo.

Amniotic folds appear in the head region of the embryo.

In the species of spider studied, probably two abdominal appendages are invagi-
nated to form each lung book.

The supra-oesophageal ganglion is indistinctly divided into two portions.

CONCLUSION.

It may be well to close this paper with some remarks on the relations of
tracheates suggested in part by the observations given.

0) Embryologie des Scorpions. Zeit. f. wiss. Zool. Bd. XXI.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 29

Peripatus and Myriapods, from the absence of wings and other primitive char-
acters, may fairly be considered the most primitive tracheates.

The position of Peripatus is uncertain ; but some Myriapods show indications
of a hexapod stage in their development. They may therefore be related to the wing-
less Hexapods. The large number of body segments and appendages in Myriapods,
and perhaps in Peripatus as well, is probably only the vegetative reproduction of
homologous parts.

In Myriapods and in Peripatus as shown by the studies of Kennel" 'and
Sedgwick ('2> on the latter, and by observations of other observers on the former, the
segmentation is total. From the accounts given by Sedgwick and Kennel it appears
that the gastrulatiou differs in different species of Peripatus.

The mode of origin of the endoderm is not, however, very important for classi-
ficatory purposes, inasmuch as it is very likely to be modified by the presence or
absence of food yolk.

The mesoderm, in the development of Peripatus, grows forwards from an
undifferentiated cell mass at the posterior end of the embryo. The mesoderm
arising from the primitive cumulus of spiders also grows forwards from an undiffer-
entiated cell mass at the posterior end of the embryo.

The endoderm of Peripatus, Myriapods and Spiders is derived from the inner
layer of the gastrula. How the inner layer of the gastrula arises is unimportant.
Consequently Peripatus and spiders are quite alike in the formation of germinal
layers.

This alone, however, does not indicate any close relationship, for in Crustacea
the endoderm arises from the inner layer of the gastrula while the mesoderm grows
forward from the posterior end of the embryo as in Peripatus and spiders.

In the higher insects the yolk cells, from their mode of origin, probably repre-
sent the inner layer of the gastrula and are consequently equivalent to the endoderm
of lower forms. The true endoderm is functional only during embryonic life in
absorbing the yolk. It takes little or no part in the formation of the digestive
tract. In these higher insects, as already shown, the inner layer, which from its

(1) Kennel. Entwicklungsgeschichte der Peripatus Edwardsii. Sempers' Arbeiten 1884.

(2) Sedgwick, Development of Peripatus Capensis. Quarterly Journal, 1885.

30 THE EMBRYOLOGY OF INSECTS AND ARACHNIDS.

position and segmentation corresponds to the mesoblast of Arachnids and Peripatus,
has usurped the functions of the true endoderm.

In Aphides from the studies of Witlaczil it appears that the intestine is formed
exclusively from the proctodaeal and stomadaeal invaginations.

In order to separate the different divisions of the arthropod phylum anatomical
characters as well as embryological phases must be considered.

The posession of a single well developed pair of antennae, of tracheal invagina-
tions and of embryonic membranes, and the existence of a hexapod stage in their
development afford sufficient ground for regarding myriapods as lowly organized or
degenerate insects. Peripatus would perhaps come under the same category through
the embryonic membranes of Peripatus do not appear to correspond fully to those of
insects.

Arachnids, from the absence of antennae and the histological structure of the
abdominal appendages, and from other characters, anatomical and histological, must
certainly be included, with Limulus, in a distinct group of arthropods. The small

*

seventh pair of thoracic appendages of Limulus is perhaps an interpolated append-
age, or perhaps a corresponding appendage may be discovered in the development
of spiders and scorpions.

Arachnids, probably, never possessed antennae, since all their appendages, like
those of Limulus, are at one period post oral, and are not innervated by the supra-
oesophageal ganglion. Trilobites, possibly the ancestrial form of Limulus, from
evidence afforded by the Cincinnati specimen, probably possessed no antennae.
If antennae were ever present in the group we would expect to find them in these
old forms.

The antennae of insects from their innervation correspond to the first pair of
crustacean antennae. The bilobed upper lip of insects is innervated from the second
division of the supra-cesophageal ganglion which forms part of the circum-cesopha-
geal commissure. In the nauplius stage the second pair of crustacean antennse is
innervated from the circum-cesophageal commissure. From their similar innerva-
tion a comparison may then be fairly drawn between the paired upper lip of insects
and the second pair of crustacean antennae.

THE EMBRYOLOGY OF INSECTS AND ARACHNIDS. 31

The antennae of insects and Crustacea are probably homologous structures and
ally the two groups.

The amnion of insects and arachnids is probably homologous, and allies these two
groups; consequently insects, having characters common <o both arachnids and
Crustacea, may be placed between arachnids on the one side and Crustacea on
the other. It does not follow from this that insects, from the possession of common
characters, are the most primitive form, but rather that they are the most hetero-
geneous and highly evolved of the arthropod groups.

The three arthropod groups may not have arisen one from the other ; probably
each group arose independently from a common source.

The tracheae of insects and spiders are probably analogous, not homologous
structures.

This may fairly be concluded from the fact that the tracheae of spiders can be
derived from the lung books which it will be remembered are involuted appendages.
Consequently we could not expect to find in insects tracheal invaginations occurring
in appendage-bearing segments of the body if the tracheae of the two groups were
homologous structures. It is needless to say that we do find tracheal invaginations
in insects occurring in appendage-bearing segments. The tracheae of insects and
spiders are therefore not homologous.

EXPLANATION OF PLATES.

REFERENCE LETTERS.

Am. — Amnion.

Am. F. — Amniotic folds.

Am. — Portion of amnion folds which

forms dorsal surface of body.
App. — Thoracic appendages.
BL. — Blastoderm.
BH.— Blastopore.
C". — Sub-oesophageal commissure of su-

pra-oesophageal ganglion.
C'. — Supra-cesophageal commissure of

supra-oesophageal ganglion.
C. — Circum-oesophageal commissure.
CB. — Chambers of ovarian tube of Fly.
CH.— Chorion.
E. — Embryo.

E.' — Vesiculated portion of yolk.
EC.— Ectoderm.
FE. — Folicular epithelium.
GL. — Salivary gland.
GV. — Germinal vesicle
GS. — Germinal spot.
H.— Heart.
I.— Inner layer.
I'. — Migratory cells of inner layer.

IS. — Somites of inner layer.

le. — Endodermic portion of inner layer.

LB. — Labrum.

LB. N. — Labial nerve.

M. — Muscle.

MD.— Mandible.

MX.— Maxilla.

MX'. — Second maxilla.

MXL.— Maxillary lobes.

NS. — Nervous system.

NC. — Nutritive cells.

OVE. — Ovarian epithelium.

PC. — Procephalic lobes.

S. — Cuticle.

YC.— Yolk cell.

YB.— Yolk ball.

YS. — Yolk spherule.

YS'. — Homogeneous yolk mass.

No. I. —Anterior portion of supra-oeso-
phageal ganglion.

No. I'. — Posterior portion of the same.

No. II. — Mandibular portion of sub-osso-
phageal ganglion.

PLATE I.

Fig. I. Transverse section of ovarian egg of Thyridopteryx. OE — Ovarian
epithelium. YS — Yolk spherules. P — Protoplasmic stratum surrounding yolk.
CH — Chorion.

Fig. I'. Portion of same section more highly magnified.

Fig. II. Segmenting egg of Thyridopteryx. YC — Yolk cells most of which
ultimately reach the surface.

Fig. III. Transverse section of egg of Thyridopteryx with blastoderm
partially formed about yolk. Yolk spherules are not represented.

Fig. III'. Yolk cells at this stage reaching the surface of the egg.

Fig. IV. Transverse section of egg of Thyridopteryx showing completed
blastoderm (BL); early embryo (E), and commencing amniotic folds (AMF).

Fig. IV. Highly magnified embryonic area of same section.

Fig. V. Transverse section of egg of Thyridopteryx showing embryo, (E)
in cross section situated in centre of yolk. Amnion covering its ventral surface is
not represented. The yolk is segmented.

Fig. VI. Hexagonal blastoderm cells.

Fig. VII. Transverse section of embryo of Thyridopteryx showing formation
of inner layer (I).

Fig. VIII. Horizontal section of embryo of Thyridopteryx showing proce-
phalic lobes (PC) at anterior end.

Fig. VIII'. Longitudinal section of embryo at same stage showing segmen-
tation of inner layer (IS).

Fig. IX. Horizontal section of embryo at same stage as that represented in
Fig. VIII showing its relations to blastoderm BL.

Fig. X. Transverse section of Thyridopteryx embryo showing separation of
inner layer (L) beneath the blastopore BH.

Fig. X'. Same section less highly magnified showing its relations to blasto-
derm (BL).

Fig. XL Transverse section of older embryo showing origin of nervous
system (NS).

Fig. XII. Nervous system separated from outer ectoderm. Elongated cells
at bottom of median groove.

Fig. XIII. Transverse sections through region of procephalic lobes.

Fig. XIV. Transverse section through same region of older embryo. Inner
cells of lateral portions specialized as nerve cells.

BRKCE. Insects and Ai.-n-

. ' ;

<£•%•

-

• '•

-^ *& ?

>,K%.®#%'

m

.— ,

:rm^-^^-^i\ Q •VJ
-yc- , i <p»u^.\ „..
i ) ( ; K»\\«« oc, ^ ' ' • ' ^\

^ . ' ^

AM'

^-T^TN •-» n?.xui

' y~v*3 C© , -v \CW.- ^ -"' a <"v<

-CL_ o^

AM"

*M~

A Kufl. >.

PLATE II.

Fig. XV. Transverse section of advanced Thyridopteryx embryo. Nervous
system (NS) separated from the ectoderm. Amnion (AM) growing dorsally. Por-
tion of the inner layer (IE) becoming specialized as endoderm.

Figs. XVI-XXI. Longitudinal sections of Thyridopteryx embryo.

Fig. XVII. Is nearly median. Numbers 1-17 indicate somites which have-
ganglia corresponding to them.

Fig. XXII. Cephalic portion of XVII highly magnified.

Fig. XXIII. Last two thoracic (Nos. VI, VII) and first two abdominal ganglia
(Nos. VIII, IX) of figure XVIII, highly magnified.

Fig. XXIV. Transverse section of advanced Thyridopteryx embryo showing
approximation of amnion folds (AM) on dorsal side of embryo. Endodermic por-
tion of inner layer (IE) beginning to shut yolk into body cavity. Migratory meso-
derm (I') growing round yolk before the endoderm.

Figs. XXV, XXVI. Transverse sections of same embryo, to show the approxi-
mation of amniotic folds on dorsal side of embryo.

Figs. XXVII-XXVIII. Transverse sections of the head of an advanced
Thyridopteryx embryo. The plane of the section is approximately an imaginary
line joining numbers i and 2 in Fig XVII.

Fig. XXIX. A diagramatic representation of the under surface of the head
of an advanced embryo.

Fig. XXX1. A diagramatic lateral view of the head of an embryo at a corre-
sponding stage of development.

BR1 CE. Insects .mil . \rurhmils

PLATE III.

Figs. XXIX, XXX, XXXI. Transverse sections of the head of an advanced
embryo They are from the same series as figures XXYII, XXVIII.

Fig. XXXII. A transverse section of the same series as the preceeding,
through the posterior part of sub-cesophageal ganglion (No. II).

Fig. XXXIII. Transverse section of head of Chrysopa showing relation of
last abdominal somite to preceeding abdominal somites.

Fig. XXXV. Transverse section of egg of Meloe showing formation of blasto-
derm.

Fig. XXXV. Structure of blastoderm.

Fig. XXXVI. Transverse section of the egg showing origin of inner layer
from median groove (BH), and amniotic folds (AMF).

Fig. XXXVII. Transverse section of a more advanced embryo of Meloe.
The amniotic folds have united and the embryo has consequently been separated
from the surface.

Fig. XXXVlIi. Transverse section of embryo at same stage but through
a different region.

Fig. XXXIX Transverse section of more advanced embryo of Meloe in which
inner "layer (I) has been constricted off.

Fig. XL. Transverse section of advanced embryo ot Meloe. The nervous
system (NS) has been separated from the surface.

Fig. XI, 1 Transverse section of egg of Mantis showing embryonic area E.

i s — ' .s to f "(3 x™\ CN

^(f^,^:^

< >I.V- *$<«?£• Sit noV

.

PLATE IV.

Fig. XLII. Transverse section of early Mantis embryo (E) with amniotic
folds (AM) on each side

Fig. XLIII. Transverse section of more advanced Mantis embryo almost sepa-
rated from surface by union of amuiotic folds.

XLIV. Transverse section of still older Mantis embryo showing origin of
inner layer (I).

Fig. XLY. Transverse section of early stage of segmentation of grasshopper
egg, showing a few cells (YC) in yolk.

Fig. XLVI. Transverse section of a later stage of grasshopper egg showing
yolk pyramids with cells at their bases.

Fig. XLYII. Transverse section of head of advanced grasshopper embryo
showing commissure (Cj of supra-cesophageal ganglion.

Fig. XLVIII Succeeding section of same series showing antennae (ANT) and
labrum (LB).

Fig. LXIX. Transverse section through commencing cesophageal ingrowth.

Figs. L-LXI1. Succeeding transverse sections of same embryo.

Figs. LII-LVIII. Sections through the rnandibular region.

•

&§&

m

-'oe
f-g$

ii

PLATE V.

Figs. LXI-LXII. Sections through the first and second maxilla.

Fig. LXIII. Transverse section of an older embryo through the first maxillae.
It is closed dorsally by mesoderm; beneath this lie the blood cells.

Figs. LXIY-LXX . Transverse sections of the thoracic region Body wall
completed dorsally by mesoderm ; the amnion is incomplete.

Fig. LXVI. Longitudinal section of ovarian tube of fly; CB'-CB n represent
sections of successive chambers of the tube.

Fig. LXVII. Longitudinal section of a more advanced terminal chamber (CIV).

Fig. I. XVIII. Longitudinal section of a more advanced terminal chamber (CIN )
showing elongated epithelium (OYE) excreting chorion (CH) and germinal vesicle
(GV).

Fig. LXIX. Transverse section of terminal chamber at same stage as pre-
ceeding through germinal vesicle (CV) and germinal spot (CS).

Fig. LXX. A longitudinal section of terminal chamber showing germinal
vesicle (GS) without boundary.

Fig. LXX'. Portion of same section more highly magnified.

Figs. LXXI— LXXIV. Sections through the cesophageal region of a spider em-
bryo showing amniotic folds (AM) on each side of the supra-cesophageal ganglion,
the outer limbs of which in the most posterior section (Fig. LXXIY ) have united.

liKl'I'K. hisculs ;n\i-l Ar;irhnids

PLATE VI.

Fig. LXXV. Transverse section of spider's egg through primitive cumulus (CX

Fig. LXXV. Cumulus of same section highly magnified.

Fig. LXXYI. Longitudinal section of spider embryo showing solid mesoblastic
somites.

Fig. LXXVII. Yolk cells (YC) of advanced embryo apparently undergoing
endogenous division.

Fig. LXXVIIP. Longitudinal section of advanced spider embryo with append-
ages and hollow mesoblastic somites.

Fig. LXXIX. Longitudinal section of more advanced embryo. First two
abdominal appendages (A & B) are less distinct, and are being folded into form
lung book.

Fig. LXXIX'. Abdominal portion of same figure magnified.

Fig. LXXX. Longitudinal section of advanced spider embryo showing amni-
otic folds (AM) of brain.

Fig. LXXX'. Cephalic portion of same section highly magnified.

Fig. LXXXI. Longitudinal section of same embryo nearer median line.

Fig. LXXXI'. Cephalic portion of the same highly magnified.

Fig. LXXXII. Highly magnified longitudinal section of brain near median
line,

IJNI'i'K. lns<-.:ls iinil Arachnids

SK

-^

X

1

3

-^

'* '•' '.', :

fe,

/

. ., 6

'>

-

c . ' '"«! ' <•

i »

•;

''

/ m

'

.

1

•,.•• *

^^

'®a ' __-^^
^. —

3

crn

t.-^

» Ji

)«

'

0