MEMOIRS 
OF THE 
PEABODY ACADEMY OP SCIENCE. 
VOLUME I. NUMBER 111. 
PUBLISHED BY THE ACADEMY, 
APRIL, 1872, 
PRINTED AT THE SALEM PRESS. 
SALEM, MASS: PEABODY ACADEMY OF SCIENCE. 
-THIRD MEMOIR. - 
EMBRYOLOGICAL STUDIES 
ON 
HEXAPODOUS IPPSECTS 
BY 
ALPIIEUS S. PACKARD, Jr 
SALEM, MASS. 
PUBLISHED BY THE PEABODY ACADEMY OF SCIENCE. 
1872. ACCEPTED FOR PUBLICATION 
JANUARY, 1872. EMBRYOLOGICAL STUDIES 
ON 
HEIAPODOUS INSECTS. 
I. DEVELOPMENT OF NEMATUS VENTRICOSUS. (PLATE 1.) 
The only observations on the embryology of the Tenthredinidse, with 
which lam acquainted, are those by Prof. A. Winchell,* made in 1863, on 
Selcmdvia cerasi Harris. The eggs of Nematus ventricosus are laid along 
the principal ribs on the under side of the leaf, being simply glued to the 
surface. They are rather large, .03 long, a little over .01 inch in diameter, 
regularly cylindrical, the ends being well rounded and white, the chorion 
being white and transparent. When magnified the egg is seen to be dotted 
over with pores, appearing like the mouths of little canaliculi (PI. 1, tig. 
8). The eggs of the first brood are laid during the last week in May. My 
observations were made later in the season, on eggs laid for a second (or 
possibly a third) brood. The females continued to oviposit from the middle 
of July, until early in August. The yolk granules vary as usual in size. 
In the earliest stage observed (July 24) the blastodermic cells had just 
formed, surrounding the yolk, and crowded three deep at one end of the 
egg (tig. 1), and at the opposite end forming but a single row (fig. 1 a). 
They were of the usual appearance, with a small dark nucleus. The cells, 
after the amnion has separated from the yolk and forms a true membrane, 
arc much more crowded than usual at the anterior end of the egg, where 
they actually touch each other; otherwise the amnion does not differ from 
that of other insects, so far as observed, and remains as an independent 
membrane, not taking any part in building up the body of the embryo. 
The mode of formation of the visceral membrane (faltenblatt) was not 
observed, but just before the embryo leaves the egg, the body and limbs 
are seen to be surrounded by this membrane, as we shall sec farther on. 
The next stage observed (fig. 2) may be compared with Biitschli’s figure 
16 of the embryo of the honey bce,f and between the embryos of the two 
insects at this stage there is no apparent difference. 
* Proceedings Boston Soc. Nat. Hist., ix, p. 321,18(55. 
|O. Biitschli. Zur Entwicklungsgcschichte der Biene. Sicbold und Kdlliker’s Zcitschrift, Bd. 21, Heft 
4, 1870, p. 519, plates xxiv-xxvii. 4 
EMBRYOLOGICAL STUDIES 
The protozonites are now formed, and also the rudiments of the appen- 
dages of the head and thorax. The procephalic lobes (the clypeus not 
being yet differentiated) nearly touch the end of the egg, and are large and 
full, with the yolk filling the upper portion; the antennae and mouth-parts 
are tubercles equal in size, and only differ from the rudiments of the legs in 
being a little shorter, the second maxillae, however, being longer than the 
others and of the same size as the legs ; the abdominal legs are as well 
developed as the thoracic. As in the bee-embryo, the abdomen is not 
turned over the back, the end being opposite the end of the egg. 
A more advanced stage (fig. 3, 3a) may be compared with Biitschli’s fig. 
31a, 316, of the honey bee. The rudiments of the antennae are now much 
smaller than those of the appendages of the mouth, and the clypeus overlaps 
the ends of the mandibles, which are now larger than the maxillae. The first 
pair of maxillae arc trilobcd. The 2d maxillae are much smaller than the 
first pair. 
The abdomen, when the embryo is a little farther advanced (fig. 4, ob- 
served Aug. 2), begins to bend under the body, the end impinging against 
the end of the egg. The alimentary canal is now formed, the oesophagus 
can be distinguished, and still more clearly the termination of the intestine 
(fig. 4,7). The labrum (lb) is now differentiated from the clypeus (c) and 
the antennae are small tubercles, corresponding in their diminished size and 
retarded development, to the same organs in the bee-embryo, and I am not 
sure but that this diminutive size of the antennae, together with its peculiar 
cubical form, as compared with the rudiments of the month-parts, may be 
the readiest mark by which to identify a hymenopterous embryo, as other- 
wise it would be difficult to distinguish it from the early stages of a dip- 
terous embryo. 
Fig. 5 represents the embryo a day older, when the abdomen is bent 
under the body so that the tip is opposite the base. The thoracic legs arc 
now large and much longer than the abdominal ones, and are extended 
backwards. The antenme seen in profile arc squarish. 
When the embryo is one day older (fig. 6, 6«, 66, Gc), the abdomen 
reaches nearly to the base of the fore legs ; it is distinctly bilobed and 
bears the two finger-shaped tubercles on the upper lobe, which form after- 
wards two short spine-like hairs. The eyes now appear as a round, dark 
spot in the centre of an orbicular piece, whose diameter is twice the length 
of the conical antenna.', which arc now divided obscurely into three joints. 
The powerful muscle (m) attached to the mandibles is very distinct, and the 
oesophagus (o) is clearly seen, suddenly dilating into the proventriculus (p). 
The labium (fig. 66, iv) now forms a single plate, from the base of which 
arise the two palpi. The six joints of the legs (fig. 6cZ) are now indicated, 
and the yolk mass extends to the middle of the abdomen. 
When the embryo is two or three days older (fig. 7), and one day pre- 
vious to hatching, the dorsal vessel is seen distinctly pulsating, and the body 
and limbs of the embryo, arc enveloped by a delicate membrane (faltcnblatt) ON HEXAPODOUS INSECTS. 
5 
Avhich still surrounds the larva for a short period after leaving the egg. The 
clypcus is now seen to be divided into two portions, while the labruni is 
elongated. The antennae still retain their diminutive size and rounded con- 
ical shape, being situated just in front of the eyes, while the mandibles are 
chitinous, triangular in form, and end in four unequal teeth. The first 
maxillae.arc bilobed at the end (seen in profile), and divided longitudinally 
into two divisions, Avhile the second maxillae arc about half as large. The 
ehiAvs are formed and the body is covered Avith short spine-like hairs. 
The development of this, and probably all of the Tenthredinidae, accords 
in its main features Avith that of the honey bee, the differences in the later 
stages of the embiyo being those Avhich distinguish the caterpillar-like form 
of the saw-fly larvae from the footless maggots of the bees, and hence are 
of secondary importance. 
11. DEVELOPMENT OF PULEX CANIS. (PLATE 2.) 
The development of Pulex cams* has been partially investigated by 
Weismann in his celebrated Avork "Dio Entwicklung der Diptcren.” He 
complains that "though the chorion is not completely opaque, yet he could 
not observe the more delicate relations, Avhose investigation alone rewards a 
connected series of observations, and every attempt to remove it Avithout 
injuring the yolk-skin (dotterhaut), proved unsuccessful.” He describes 
but one stage in the development, that represented by our fig. 5, and re- 
marks that "at the first glance the great similarity Avith the embryo of the 
Tipulidae is apparent. In this case the primitive band must arise through an 
actual splitting of the blastoderm (Kcimhaut),” and it Avill be seen from 
our observations, that this must be the fact. He also remarks, and our 
observations confirm his statements; " Avhilc also the form and position of 
the primitive band Avholly correspond to their relations in Chironomus, so 
also the composition of each segment is Avholly analogous. * * * * 
The number of primitive segments agrees Avith that of the Tipulidae, and 
the last (twelfth) segment is clearly seen to be due to the union of two 
pieces pressed together, one dorsal and the other ventral, between Avhich a 
fine, pointed prolongation of the yolk projects, just as is the case in the same 
stage of Chironomus (Comp. Taf. iii, fig. 32). The similar formation of 
the abdomen, indicates a similar origin of the anal opening and rectum 
(hinterdarmspalte), and so avc are led to adopt the existence of a faltcn- 
blatt, Avhich in Chironomus is concerned in the formation of these parts. 
But these observations are not such as to afford proof of such a fact, and avc 
need only state that the embryonal development of the flea comes nearest to 
that of the Tipulidae, that the tAVo families belong to those insects Avhich arc 
developed out of a primitive band, Avhich oavcs its origin to a splitting of 
the blastoderm.” (p. 89, 90.) 
* Our eggs were taken from the cat. As several authors regar 1 P. canis Bouchd and P. fells BouclK; as 
the same species, I adopt the name P. canis in this paper. EMBRY OLOGI CAL STUDIED 
The eggs of the species of Pnlex we observed, lire short, cylindrical, 
white, the chorion thin, but the yolk unusually dense, owing to the small 
size of the granules. The eggs were .02 inch long, and .01J inch in 
diameter. 
Though I did not observe the female in the act of oviposit ion yet I have 
little doubt that the eggs when first observed, had been deposited.only one 
or two hours. At this time (plate 2, fig. 1) there was observed a vacant 
space between the yolk and the chorion at the posterior pole, the egg-con- 
tents completely filling out the opposite end. Also at this time the end of 
the egg distinctly bulges out, and in this shallow sinus arc four distinct polar 
cells apparently immersed in protoplasm, and a small indistinct one in addi- 
tion; they arc distinctly nucleated, just as in Chironomns. There seems to 
be a membrane (I suppose the yolk skin) retaining these polar cells in place. 
The yolk (y) is surrounded by a thin layer of protoplasm two or three 
times the diameter of the polar cells in thickness; this is the primitive 
blastodermic skin (Keimhautblaslem of Weismann) or as it might be termed, 
protoblasloderm. 
The egg at this stage may be compared with Weismann’s fig. 1, of Chi- 
ronomns. 
About two hours after the stage indicated by fig. 1, the vacant space in 
the posterior pole of the egg has disappeared, and the yolk and protoblasto- 
derm have pushed up against the yolk skin and polar cells (fig. 2). In half 
an hour’s time more the yolk mass had advanced half way to the polar cells. 
At this time there were no signs of blastodermic cells. 
A few hours after, probably not over thirty after the egg had been laid, 
the blastoderm cells had appeared, forming a layer around the yolk (fig. 3). 
This stage may be compared with "Weismann’s fig. 3. Soon after this the 
polar cells break down and disappear. 
The next stage observed (fig. 4) may bo compared with Weismann’s fig. 
18. The amnion (am) is as usual, lying next to the chorion; while the 
faltcnblatt (fb) may be traced enveloping closely the primitive band. The 
cephalic lobes and succeeding protozonites arc formed. A slight interval 
separates the end of the future abdomen from the procephalic lobes. The 
egg at this stage is about forty-eight hours old. Fig. 4a represents the eggs 
seen at the same period from above, the extremity of the abdomen resting 
on the upper surface of the posterior two-thirds of the yolk, and the proce- 
phalic lobes and two succeeding protozonites resting upon the anterior end of 
the yolk. The median furrow is indicated at the posterior end of the egg ; 
the abdominal portion of the primitive band in view is seen to be narrower 
at the end than at the base, and ending abruptly. 
A little more advanced stage observed in the same egg when it has been 
laid seventy-two hours, is represented at tigs. 4b and 4c. The outer yolk 
layer (oy) is very distinct, and seems to be composed of partially formed 
muscular tissue and broken down yolk cells, while the inner layer (?’y) or 
central portion, is as usual darker than the rest of the egg. The faltcnblatt, ON HEXAPODO.US INSECTS. 
7 
or visceral membrane, may now be distinctly seen passing from the pro- 
cephalic lobes to the last abdominal protozonite, showing that it is as 
complete an envelope to the embryo as the amnion. The antennas (i), 
mandibles, and two pairs of maxillae are beginning to bud out, and the 
dypeus (c) is differentiated from the procephalic lobes, forming a simple 
fold of the membrane. 
Shortly after (I did not observe the number of hours, but probably in 
less than twenty-four) the embryo appears as in fig. 5, which may be com- 
pared to Weismann’s fig. 30. The inner yolk mass is now much smaller 
than before, being confined to an oval mass. The rudiments of the appen- 
dages of the head have increased in length; and there are plainly indicated 
ten abdominal segments, but the tenth appears to be afterwards folded in to 
form the end of the body, as in Chironomus, according to Weismann’s 
observations. Fig. 5a represents a little later stage, the same as figured by 
Weismann (Tab. vii, fig, G2). 
At a still later stage (fig. G, 6«) the head is walled in; the second pair of 
maxilhe have approximated; the mandibles touch each other on their future 
cutting edges, over which the labrum laps, while the first maxilhe arc still 
remote from each other. The dorsal part of the embryo has now closed in, 
and the next stage is the folding of the end of the abdomen under the body. 
This occurs when the egg has been laid four days. The process is just as 
has been described and figured by Weismann in Chironomus. Fig. 7 gives 
an idea of the appearance of the embryo when the act is half accomplished, 
the ninth and terminal segment of the body pressing against the end of the 
egg. The next day the body has extended twice its former length, and 
the embryo is doubled upon itself so that the end of the body rests under 
the head. At this time the faltenblatt loosely envelops the embryo. The 
supposed egg-opener is distinct, and shortly before hatching the embryo 
appears just like the freshly hatched larva. It leaves the egg in about six 
days after the egg is laid. 
We have seen that in its embryology Pulex does not differ from that of 
Chironomus, with whose development that of Simulium essentially agrees; 
and though we know nothing of the embryology of the Mycetophilidre, yet 
when we come to compare the larvre of Pulex with those of the Mycetophi- 
lidre, there is a general resemblance, which does not militate against the 
view of Haliday, who first suggested that the Pulicidse arc related to this 
dipterous family. 
111. DEVELOPMENT OF ATTELABUS EHOIS. (PLATE 3, FIGS. 1-5.) 
The eggs of this species of Attelabus are readily obtained by opening the 
singular thimble-like rolls, made by the female late in June and during the 
month of July, on the alder. When about to oviposit, the female picks up a 
leaf with her mandibles, and begins to eat a slit near the base of the leaf, on 
each side of the midrib and at right angles to it, so that the leaf may be folded 8 
EMBRY©LOGICAL STUDIES 
together. Before beginning to roll up the leaf she gnaws the stem nearly 
off, so that after the roll is made, and has dried for perhaps a day, it is 
easily detached by the wind and falls to the ground. Then folding the leaf, 
she tightly rolls it up, neatly tucking in the ends, until a compact, cylindri- 
cal, solid mass is formed. Before the leaf is entirely rolled, she deposits a 
single egg, rarely two, in the middle, next to the midrib, in a little cavity, 
where it lies loose. While she is thus engaged the male may often be seen 
standing at the other end of the leaf, or on an adjoining one. I once saw a 
female pick up a leaf with her jaws, then stop, and run back to receive the 
embraces of the male, and then resume her work. In another case I saw 
the sexes unite after a roll had been half made, so that it is probable that 
union occurs several times at short intervals, as I have observed to be the 
case with Gastrojjhysa polygoni. These rolls remain on the bushes some- 
times for several days, but probably drop by the time the larva escapes from 
the egg, and it seems probable that the larva at once eats the inside of the 
roll, which thus answers to the leaf bud or seed in which other larvae of this 
family live. 
The only coleopterous insect whose development in the egg is known,* is 
Donacia, whose embryology was first described by Kollikerf and more fully 
by MelnikowJ in 1869. It will be seen that the embryology of Attclabus 
essentially agrees with that of Donacia, a member of a remotely allied 
family. The eggs of Attelahus rhois are nearly spherical, though a little 
longer than thick, .04 inch long, and .03 in diameter. The chorion is thin, 
smooth, very transparent; the yolk is made up of pale green, fine granules. 
When the egg has been laid about (probably less than) twenty-four hours, 
I could detect the blastodermic cells in course of formation (plate 3, fig. 1), 
though the earliest appearance of the nucleus at the periphery of the yolk 
(see Melnikow’s fig. 1) was not seen. But the stage at which I observed 
them may be compared with his fig. 2. The outer row of blastodermic 
cells had been formed, but in the inner row the molecules of the yolk mass 
were arranging themselves around the nucleus, and the cell walls were par- 
tially formed, being evidently as first shown by Wcismann§ filled with yolk 
molecules. 
When the egg is fully twenty-four hours old, the blastoderm, entirely 
surrounding the yolk, is fully formed, as easily seen on crushing the egg, 
and the primitive band is confined to one side of the egg. The cells of the 
blastoderm (fig. 2) are pressed side by side, being, a little elongated, and 
the dark nucleus is much more distinct than before, while the yolk is dis- 
tinctly separated from the yolk. 
Fig. 3 represents the egg after it has been laid about thirty-six hours, and 
*Joly (Ann. des Sc. Nat., 3d Ser., Tom. 2,1814, p. 5) has given a brief account of the embryology of Colaspis 
atra, but only figures the later stages just before the embryo leaves the egg. His observations are not suffi- 
cient to enable us to decide whether Colaspis is developed like other Chrysomolids or not. 
fObservationes de prima Insectorum Genesi, etc. Turici, 1842. 
t Beitriige zur Embryonalentwickelung der Insekten, Archiv fiir Naturgeschichte, vol. 35, p. 13C. 18G9. 
§ Die Entwicklung der Dipteren. See Tab. iv, fig. 52, 53, 53A, 538, E. ON HEXAPODOUS INSECTS. 
9 
may at this time be compared with the stage in Douacia, represented by 
Melnikow’s fig. 5. The primitive band rests entirely on the outside of the 
yolk, the thickened portion, or so-called primitive band, covering two-thirds 
of its surface. At the anterior pole of the egg it is a little depressed. The 
stages indicated by Melnikow’s figs. 6-11, were not observed, but the next 
step in the development of Attclabus, is the extension of the primitive 
band, the ends meeting so as entirely to surround the yolk. At this time 
the median furrow in the band is distinct, with the appendages just begin- 
ning to bud out. Though the falteublatt could be seen passing around the 
cephalic appendages, I could not make out its relations so clearly as shown 
in Melnikow’s tig. 12, with which this stage may be compared. 
In the succeeding stage observed (fig. 4) the procephalic lobes, clypeus, 
and appendages of the head and thorax, are well developed. The rudiments 
of the mandibles are rather larger than the other mouth-parts. The four 
protozonites of the head may bo distinctly seen, the procephalic lobes form- 
ing the first. The germ still floats upon the yolk, which fills the inner half 
of the primitive band. The yolk granules are contained in polygonal 
vesicles just as in Douacia. The end of the abdomen, consisting of eleven 
protozonites, has retreated from the procephalic lobes, allowing a portion 
of the yolk to impinge against the chorion. 
Fig. 5 represents the embryo a day older. The mouth-parts are a little 
longer, and the articulations in the legs (flg. sa) arc distinct. The cells ot 
the amnion arc very distinct, but scattered and irregular, and the falteublatt 
is seen enveloping the body and limbs, being composed of scattered cells 
like those of the amnion. This stage may be compared with Melnikow’s 
fig. 13. 
In its essential features the embryology of Attclabus is similar to that of 
Douacia, the primitive band resting on the outside of the yolk, being an 
"ectoblast,” to use Dr. A. Dohrn’s expression. It is after the stages above 
described that the embryo assumes the form peculiar to its genus and family. 
The larvce began to hatch out about the middle of July. I did not ascertain 
the exact number of days from the time of ovipositiou to the exclusion of 
the embryo, but judge that it required about a week. 
IV. DEVELOPMENT OF TELEPHORUS FUAXINT. (PLATE 3, FIGS. 6—8.) 
The development of this beetle in its earliest stages, is of remarkable 
interest, since it differs from the other Coleoptera whose development is 
known, in the primitive band floating in the centre ot the yolk, instead of 
surrounding it. 
On the 26th of June, a female which I had put in confinement hoping to 
obtain some eggs from it, laid a few loose in the bottom of the vial. They 
were broad, ovate, pale yellowish in color, with a thin, smooth, transpar- 
ent chorion. The yolk cells varied in size, from that of the molecules which 
had collected around the periphery, to a few large granules situated in the 10 
EMBRYOLOGICAL STUDIES 
centre of the yolk and about one-sixth the diameter of the egg. On the 
30th of June several of the eggs were observed with the primitive band 
(plate 3, fig. 6) forming a broad, sigmoid, yellowish streak. ‘The anterior 
end was the broader, and dilated slightly at the end, forming the procephalic 
lobes, and consisting of fine yolk granules, becoming finer posteriorly, until 
its structure could not be determined, as it had sunk in towards the centre 
of the yolk. The protozonites had not yet been indicated. At this stage the 
appearance presented by the germ was almost identical with that of the eggs 
of Lecanium, observed by myself, several years previously, and of Aspidi- 
otus nerii, which has been figured by Metznikow* (pi. xxxii, figs. 7-10, and 
also his figure of the embryo of Aphis roste). It is also similar to the 
eggs at this stage of Caloptcryx and Agrion, as figured by Brandt, f and of 
the Pediculina described and figured by Mclnikow. 
In the next stage observed, the embryo rested on the outside of the yolk 
and in no wise differed from the same stage in Attelabus (see fig. 4), the 
rudiments of the head and the protozonites, and the disposition of the 
body .on the yolk being the same. The two membranes, the amnion and 
faltenblatt, are also the same. 
When the embryo is about eight days old (fig. 7, ventral view), and 
nearly ready to hatch, the clypeus is fully differentiated, and the mouth-parts 
are as in Attelabus, the yolk still occupying a large portion of the egg. 
Fig. 8 represents the embryo seen ventrally, and ready to escape from the 
egg; it presents very much the same appearance as the Donacia embryo at 
a little later stage, figured by Kollikerf (Tab. iii, fig. vi), the differences 
being only such as we would expect to find in embryos belonging to two 
widely different families. The abdomen is bent under the body so that the 
tip reaches nearly to the end of flip labium. The yolk mass is restricted to 
a small portion of the body. The antennae arc still embryonic, being simple 
tubercles, but the mandibles meet beneath the labium.. The second maxillae 
have united to form the labium, and the head is distinct from the thorax. 
By the 6th of July all the embryos had hatched, the period required for 
the growth and hatching of the embryo being about ten days. 
Y. DEVELOPMENT OF CIIRYSOMELA (GASTROPIIYSA) POLYGONI (PLATE 3, 
FIG. 9) AND MYSIA 13-PUNCTATA. 
Our observations are very fragmentary, and they are given only to show 
that the development of these two Chrysomelidans agrees substantially with 
that of Donacia. 
The eggs of Gastrojphym polygoni arc long, cylindrical, yellow, and laid 
in patches on the under side of the leaves of different plants. The chorion 
* Siebold und Kolliker’s Zeitschrift, 1886, p. 380. 
fßeitriige zur Entwicklungsgeschiclite der Libelluliden und Ilemiptcren. Mem. Acad. Imp. des Sciences, 
St. Petersbourg, 1869. 
t Observations de prima lusectorum Genesi. Ttirici, 1812. OX HEXAPODOUS INSECTS. 
11 
is smooth and transparent. Two days after the eggs were laid the embryo 
had arrived at the stage indicated by pi. 3, fig. 9, which may lie compared 
with Melnikow’s tig. 14 of the embryo of Donacia. It also bears a striking 
resemblance to the embryo (in the same stage of growth) of Chironomus, 
as figured by Weismann (Tab. ii, fig. 21, 22), and I think it would be 
impossible for an experienced embryologist to tell whether the germ at this 
stage was destined to become a dipterous or coleopterous insect. On 
account of the large clypeus filled with yolk grannies, as in Chironomus, 
and the relation of the rudiments of the cephalic appendages and also of the 
faltenblatt (which compare with Weismann’s fig. 20), one would think it 
resembled a dipter the more. I did not observe the earlier stages so as to 
ascertain whether the primitive band is at first endo or ectoblastic. 
The eggs of Mysia 13-punctata, one of the Cocciriellidte, are oval cylindri- 
cal, not so long as those of Chrysomela, and are yellow, with a thin, smooth 
chorion. They were laid in confinement on the 16th of July, and two days 
after, a slight examination sufficed to show that the primitive baud in this 
genus rests on the outside of the yolk, the eggjit this stage appearing much 
as that of Attelabus (fig. 3), though not quite so far advanced. 
These fragmentary observations on species of the Coleopterous families, 
Lampyridte (Telephorus), Curculionidae (Attelabus), Chrysomelidoe (Dona- 
cia, by Kolliker and Melnikow, and Gastrophysa) and Cocci nellidee (Mysia), 
may bo taken as tending to show that (with the exception of Telephorus in 
its earliest condition, when the primitive band floats in the centre of the 
yolk, as in the Hemiptera and Libell.ulidae) the mode of development in the 
Coleoptera is nearly identical with that of the Diptera and also the 
Hymenoptera; the formation of the embryonal membranes seems also to be 
nearly identical. 
Newport’s observations on the embryology of Stylops * are not detailed 
enough for us to judge how far the development of this singular Coleopter 
agrees or disagrees with that of the group generally, except that he states 
that the primitive band rests on the outside of the yolk. 
VI. DEVELOPMENT OF CHRYSOPA OCULATA. (PLATE 3, FIGS. 10-13.) 
The first observations on the embryology of the family of Hemcrobidie, 
of which Chrysopa is a member, were made by Dr. H..Hagen,f on Osmylus 
maculatus Fabr. He describes the three membranes surrounding the em- 
bryo, i. e. an outer covering forming a network of polygonal pits around the 
chorion ("Umhullungshaut ”), the chorion and the " dotterhaut,” or amnion. 
"The development of the primitive band begins on the fourth-day, the ventral 
side presenting outwards, first at the head and a little later in the posterior 
abdominal segments.” From this I suppose that the primitive band does 
t Die Eutwicklung und der innere Ban von Osmylus. (Published in the “ Linnsea Entomologica,” Stettin, 
1852.) 
♦The Natural History, Anatomy, and Development of Meloe. Linn a'an Trans, xx, Tab. 11, 1817. 12 
EMBRYOLOGICAL STUDIES 
not float at first in the centre of the egg, but from the first rests on the out- 
side of the yolk. On the seventh day it presents the appearance indicated 
by my fig. 11, of Chrysopa. On the tenth day the germ is developed much 
as in my fig. 12 ; and the embryo hatches on the twelfth. 
F. Braner* has studied the later developmental history of Mantispa 
pagana Fabr. He states that the "turning of the embryo begins on the 
eighth day, and on the ninth is completed.” After this the position of the 
body and its appendages is essentially like that of Chrysopa, the .differences 
being due to the greater length of the body of Mantispa. The eggs are 
stalked as in Chrysopa, and the large micropyle is similar to that of the 
eggs of Chrysopa. He notices the chorion and the amnion ("dotterhaut”). 
At the expiration of twenty-one days after oviposition the embryo escapes 
from the egg. 
The female of Chrysopa oculata Say, the species upon which the follow- 
ing observations were made, laid between forty and fifty eggs (one in con- 
finement laid forty-six eggs ; another eighteen eggs, or egg stalks, but her 
reproductive powers were evidently impaired, or else she had laid some eggs 
before capture, as there were only nine well developed eggs in the batch, 
and nine stalks, some only half the usual height, others with the knob of 
cement at the end to which the egg is ordinarily fastened). The eggs (.04 
inch long, ,02 in diameter) are at first pale green, turning grayish just 
before the .embryo is hatched. The eggs arc evidently stuck on to the end of 
the pedial after the latter has been formed, as in one case an egg was glued 
to the stalk very much out of centre, the insect’s ovipositor not having aimed 
straight, so to speak, at the mass of cement. The chorion is not very trans- 
parent, though smooth. The micropyle, like that of Corixa, forms a con- 
spicuous button-like knob, resembling the finely milled head of a screw. 
The yolk granules are remarkably small, so that the primitive band is in 
strong contrast to the yolk in color and density. When the egg is crushed, 
the yolk does not flow out as a liquid, bearing globules along the stream, but 
in a pasty mass, somewhat, but not so marked as in the eggs of Limulus 
when similarly treated, and I would question whether the fineness of the 
yolk granules, and the denseness of the yolk be not connected in the way 
of cause and effect, with the exposed situation of the eggs of Chrysopa. 
The eggs were laid in confinement on the 3d of July, and I did not suc- 
ceed in watching the formation of the amnion, or seeing the earlier steps in 
the formation of the primitive band, owing to the opacity of the chorion 
and the density of the yolk. PI. 3, fig. 10, 10«, 10&, indicates the embryo 
when three days old, the eggs being observed July 6th. At this time the 
amnion was a closed sac entirely surrounding the yolk; the cells were 
unusually round and distinct, with a well defined nucleus, and were less 
crowded than usual, and they may be viewed when looking at the egg as an 
opaque object, appearing as greenish opaque dots, probably reflecting the 
*Bcitragc zuv Kenntniss des inneren Banes mid dor Verwaudlnng der Ncuroptcrcn. Vcrhaudhnigen des 
Zbol.-Bot. Vercins in Wien. Bd. v, p. 701.1853. OX lIEXAPODOUS INSECTS. 
13 
green tint of the yolk. The cells are crowded thickly at each end of the 
egg, being seen more readily when the yolk thins out. 
The primitive band is fully formed, the protozonites being distinctly 
marked, the transverse impressed lines indicating the primitive segments 
being distinct, and the median furrow easily discerned. The procephalic 
lobes arc well formed, being regularly convex in front, broader than the rest 
of the band. The abdominal portion of the band is a little more than half 
as wide as the anterior portion. The end of the band seen on the dorsal 
side of the egg forms a distinct region of the band which may be called the 
primitive postabdomen ; it forms a rounded portion separated by a suture 
from the rest of the abdomen, beyond which suture the median furrow does 
not pass. 
There arc no indications of limbs. The germ lies on the outside of the 
yolk, the ventral side facing outwards, the end of the abdomen curving 
around the yolk and resting on the dorsal surface of the egg, as seen in tig. 
9b. In profile (fig. 10) the under or dorsal side of the germ sinks into the 
yolk as in all insects, but the line between the band and the yolk mass is 
remarkably distinct; the under side of each protozonite is convex, especi- 
ally, in those segments nearest the procephalic lobes, and these latter are 
broad, orbicular, seen in profile. The primitive band is pale yellow, grow- 
ing paler towards the edges; and in this respect, though perhaps unimport- 
ant, the germ differs remarkably from that of any other insects with which I 
am acquainted, and also the eggs of any Crustacean or Arachnidan referred 
to by authors or noticed by myself. Hagen describes colored spots in the 
primitive band of Osmylus, like those noticed farther on in the present 
insect, so that this high color of the primitive band may be a characteristic 
of the Hemerobidse (in the broad sense, including the Ascalaphi and Man- 
tispac). This stage may be compared also with that of Chironomus, repre- 
sented by Wcismaim’s pi. ii, fig. 18, the differences apparently being due to 
the greater density in the yolk and primitive band of Chrysopa, rather than 
to any morphological differences. It also more closely agrees with the stage 
of Corixa indicated by Melnikow’s tig. 15. 
Figure 11 represents an egg older than the other, and somewhat flattened 
by pressure, the primitive band being distorted; the sutures are, however, 
clearly indicated. At this time the procephalic lobes are pale yellow, while 
the rest of the primitive band is greenish yellow, and the postabdomen is 
brownish, and there arc brownish spots at most of the sutures. 
In another egg which had been laid between three and four days (laid 
July 4-sth, and studied July 7th), the appendages had appeared (fig. 12). 
While the amnion is distinct, the faltenblatt is also distinctly seen closely 
surrounding the head and the end of that part of the abdomen in sight, the 
end being still folded on the back of the yolk, so that there is no doubt but 
that the whole body is enveloped by it, as in the Colcoptera. Owing to the 
opacity of the yolk, I could not trace it along the side of the cephalic and 
thoracic portion of the band. How far its relations to the yolk mass, at the 14 
EBBRYOLOGICAL STUDIES 
period of turning of the embryo, is like that described by Brandt in his me- 
moir on the development of Caloptcryx and Agrion, I could not determine. 
The rudiments of the clypeus now appear as a tubercle arising from 
between the procephalio lobes, which hang over the area destined to form 
the mouth. The antennae (i) arc plainly seen to depend from near the edge 
of the lobes. The three succeeding pairs of appendages (ii-iy, the mandi- 
bles, and Ist and 2d maxillae) arc alike in size and appearance, while the 
three pairs of legs are directed obliquely backwards and meet on the median 
line of the band. The germ is now more highly colored than before, being 
quite bright yellow, while at the base of the abdominal portion of the band 
are two large, rounded, brown spots, which meet on the median furrow, and 
this band is spotted with brown posteriorly, as seen in fig. 10. 
At this time the germ is very similar to the germs of Diplax and Calop- 
teryx, and there is apparently no essential difference, except that the germ 
of Chrysopa docs not sink into the yolk as do those of the Libelluliche. 
I am inclined not to regard this as of much importance, as even in the 
Hcmiptera as shown by Metznikow and Brandt, the primitive band varies in 
degree of immersion in the yolk, being most marked in Aspidiotus and least 
so in Corixa, the primitive band in the latter being surrounded by the yolk 
only at the earliest moments of its formation (see Metznikow’s fig. 7, in 
which the band is only half as long as the egg), and it is not improbable that 
the primitive band in Chrysopa is at first surrounded by the yolk, while in 
Chironomus (fig. 18 of Weismann), the germ sinks partially down into the 
yolk. (See Metznikow’s fig. 15 of Corixa, with which the present stage 
maybe compared.) We are also strikingly reminded of the primitive band 
of the Araneina, as it is seen floating on the surface of the yolk, with the 
rudiments of the mouth-parts and limbs projecting outwards as tubercles 
(compare Claparede’s fig. 33*). 
Of course the limbs did not appear at once in this stage, as seen in this 
egg; their rudiments must have appeared synchronously with the segments 
seen in the preceding stage, though they could not be detected, owing to the 
difficulty of examining the eggs from the opacity of the chorion. The next 
steps in the growth of the embryo were not observed, including the bending 
of the abdomen beneath the body. The embryo, however, does not turn 
around in its shell, as will be seen by looking at fig. 13 (which rudely 
and very imperfectly represents the embryo when between five and six days 
old), where the head remains next to the micropylc, as in the earliest period 
of its development. The process of turning is not, then, essential in insects, 
and for this reason I am inclined to think that as in the Coleoptera, the 
faltenblatt is not drawn in to form a yolk bag as described by Brandt in the 
Libellnlidee. 
The embryo is hatched in six or seven days from the time of oviposition, 
just about half the time occupied in the development of Osmylus, as stated 
by Dr. Hagen, and one-third the time occupied by Mantissa jpagana. 
* Rechcrches sur I’Evolution lies Araignees. Utrecht, 18C2. ON HEXArODOUS INSECTS. 
15 
While, ns wo have seen, Chrysopa in its earliest stages, i. e. in the mode 
of formation of the primitive band, probably agrees closely with Corixa, it 
afterwards, at the time when the limbs bnd out, agrees more closely with 
the Libellulidie, the difference being that the germ of the Libellulidae floats 
within the yolk, while that of Chrysopa (like that of Corixa) rests upon the 
outside of the yolk. This difference of position of the germ seems to be of 
but slight importance, especially when we bear in mind the great diversity 
in the position of the yolk in the different genera of the Hemiptera, there 
being every grade from Corixa, in which the germ is immersed only at the 
first beginning of the primitive band, up to Aspidiotus, where it forms a 
large distinct S-shaped band. We have also seen that in the Colcoptera the 
germ of Attelabns is, so far as we were able to find, never immersed in 
the yolk, while in Telephone it has the S-shapc, characteristic of Aspidi- 
otus.* 
The embryology of Chrysopa docs not essentially differ, then, from that 
of the Libellulidie. We now know that the development of Osmylus, and 
probably Mantispa is similar to that of Chrysopa, and hence we may sup- 
pose that the other members of the family Hemcrobidie are developed in the 
same manner. The Libellulidie arc considered a group of the Pseudoneu- 
roptera, while the Hemerobidce arc types of the ' true” Neuroptcra. But 
having shown that the mode of development of these two families is as 
similar as in two families of the Hemiptera, we would ask what becomes 
of the distinction between the "Pseudo” Ncuroptera and "true” Neuroptcra? 
Never believing that the differences were quite enough to separate the 
Linnsean Ncuroptera into two independent orders or suborders (whichever 
we may choose to call them), we would now ask if embryology does not 
give independent testimony as to the close alliance at least of the Libcllu- 
lidie and Hemcrobidie, even if we go no farther? 
We have shown that the development of Nematus agrees with that of 
Apis, as studied by Biitschli, while that of the Formicidie is much the same, 
as shown by Ganin. The development of the egg-parasites (Proctotrypidie) 
is in many respects very extraordinary, and seems obviously connected 
with the exceptional mode of life of the larva. Of the development of the 
Lepidoptera we know little beyond Herold’s researches. Ganin,f however, 
states that the development of the Hymenoptera and Lepidoptera, as regards 
the mode of formation of the amnion, its relation to the embryo, as also the 
physiological role it plays during embryonic development, is entirely similar 
in the two groups. He farther says that the amnion is developed in Apis 
(according to WeismaniTs investigations) exactly as in the Formicidie and 
Lepidoptera, and adds that such is the case in many Ichncnmonidie and 
* For the reason that the primitive band varies so much in its position in relation to the yolk, closely allied 
insects varying greatly in this respect, while in some genera of parasitic Ichneumon Hies (Teleas, Polynema, 
Platygaster, the primitive band is not formed until the second larval stage is reached, according to Ganin), 
I have been led to think that Dr. Dohrn’s classification of insects into entoblasts and endoblasts (see Ent. 
Zeitung, Stettin., 1870, p. 244) is quite artificial. 
f Über die Erabryonalhalle der Hymenoptercn uud Lepidopteren Bmbryonen, 18GP. 16 
EMBRYOLOGICAL STUDIES 
Cynipidse. He adds farther that the mode of development of the embry- 
onal membranes is the same in all insects with a complete metamorphosis. 
In all these insects the amnion takes no part in building up the body of the 
embryo, while in the ametabolons lice they do, and thus, he says, in earlier 
times this may have been the case with all ametabolons insects. 
We have also seen that the Diptera and Coleoptcra are apparently devel- 
oped in much the same manner, and that they agree substantially with the 
Hymeuoptera. The greatest divergence is presented in the Hemiptera and 
Neuroptera (the Phryganeidae and Pod nr idm are similar in their embryology 
and resemble that of Attelabus in perhaps the most important features, 
though with some peculiarities of their own). These differences are not 
improbably owing (at least this explanation may be better than none at all) 
to the fact that the larvae arc born with a peculiar form, their metamorphosis 
being incomplete. Indeed in several of the instances in which the embry- 
ology of insects has been studied, we find that the divergences from the 
normal mode of development are correlated with important biological differ- 
ences of the larval or adult insect, and I think it will be one of the points to 
be established in the future, i. e., how far the differences in the embryology of 
animals arc due to differences in the habits and instincts of the larvae and 
adults. 
Thus embryology will not prove an infallible guide in the classification of 
animals, but is to be used with the same caution as anatomical and biologi- 
cal characters. So also with the postembryonic changes in insects; the 
diversity of the mode of metamorphosis in the families of the Neuroptera 
and Orthoptera for example, is perhaps of less importance in taxonomy 
than many will allow. 
So far as our present investigations have gone, we may conclude that 
there is a remarkable uniformity in the mode of development of the Hexa- 
poda, as much so perhaps as in the Decapodous Crustacea, and it is difficult 
to determine what embryological characters may bo set down as distinguish- 
ing even the different suborders. These characters, whatever they may be, 
do not probably reside in the embryonal membranes, or in the relation of 
the primitive band to the yolk. It is more probable that they will be found 
in the form of the advanced embryos. For example, we now know that the 
embryos of the Isopod Crustacea only differ from those of the Amphipods 
while in the egg, by having the end of the abdomen bent over the back, 
while in the latter (Amphipods) it is curved beneath the body, as pointed 
out by Fritz Muller. We may find that the cylindrical, more or less cater- 
pillar or maggot-like form of the Hymenopterous, Lepidoptcrous, and Dipter- 
ous lame, which I have called* cruciform, have a common mode of develop- 
ment, in distinction from the larvae of the Neuroptera and Orthoptera, which 
do not have a worm-like body, the head being large, the legs usually like 
those of the adult, and the body elongated and flattened, or oval, more or 
less like a mite. This general type I have called Jejiliform, and believe that 
♦American Naturalist vol. iv, p. 75G, y. p. 51, 5G7,1871. OX HEXAPODOUS INSECTS. 
17 
it is the primitive form of the paleozoic larvae, in accordance with Fritz 
Muller’s opinion "that the 'incomplete metamorphosis’ of the Orthoptera is 
the primitive one, inherited from the original parents of all insects, and the 
'complete metamorphosis’of the Coleoptera, Diptera, etc., is a subsequently 
acquired one.”* It should be noticed in this connection that the maggot- 
like larvae of the weevils (Attelabus), which are rather cruciform than lepti- 
form, are developed much like the Diptera, while the somewhat cruciform 
larvae of the Phryganeidae also resemble the Diptera in form. (The Podu- 
ridae form an apparent exception, as they develop like the Phryganeidae, 
though decidedly leptiform when hatched.) But such speculations may 
prove to be premature, our present knowledge of the embryology of the 
Hexapods being far less perfect than that of the embryonic stages of the 
Crustacea. 
*Fiir Darwin, Eng. Transl., p. 121. EXPLANATION OE PLATES. 
Figure 1. Anterior and postex-ior end of egg, showing formation of blastoderm. 
Plate 1. Nematus ventricosus. 
Figux-e 2. Primitive baud, with rudiments of appendages of head and thorax. I, antenna ; 
11, mandible; 111, maxilla; IV, 2d maxilla. (The lettering is the same in all the plates.) 
Figux-e 3. Ventral view of the embryo a little fai'ther advanced. 3a, the same enlarged, 
. showing the clypeus (c) overhanging the labrum {lb); the maxillae are tx-ilobed. 
Figure 4. The body has increased in length, the abdomen beginning to bend down. 
Figure 5. The head separate from the thoi’ax, and the abdomen bent beneath the body. 
Figure G. The embryo still farther advanced, the abdominal legs being distinct. 6a, 66, 
views of the head; o, oesophagus; p, proventriculus; in, muscles attached to the mandi- 
bles; 6c, tubex-cles on the end of the abdomen, destined to form hairs; 6d, a leg. 
Figure 7. Embryo one day previous to hatching, the end of the abdomen reaching to the 
anterior feet, dl, faltenblatt. 
Figure I. End of egg after having been laid one or two hours. 
Figure 2. The same, two houi’s after the stage represented by fig. 1. 
Figure 3. Egg showing layer of blastodermic cells. 
Plate 2. Pulex canis. 
Figure 4. Blastoderm, with the procephalic lobes and protozonites indicated, also the fal- 
tenblatt (/6) and amnion lying next to the chorion. 4a, dorsal view of the same; 46, 
side view of end of the egg, when a little older, showing the head and body enveloped 
by the faltenblatt (/6) and the inner and outer yolk layers; 4c, the same enlarged; 
am, amnion. 
Figure 6. Embx-yo about three days old; sa, a little later stage. 
Figure 6. Head completely walled in. Embx-yo four days old; 6a, head enlax-ged. 
Figure 7. Embryo about a day before hatching. 
Plate 3. Figs. 1 to 5, Attelaxxus rhois; Figs. 6 to 8, Telfpiiorus fraxini; Fig. 9, 
Chrysomela polygoni; Figs. 10 to 13, Chrysopa oculata. 
Figure 1. Portion of egg, showing layer of blastoderm cells, {am) destined to form the am- 
nion ; y, half formed primitive baud. 
Figure 2. The same when the egg has been laid about 24 hours. 
Figux-e 3. Egg, 36 hours after being laid. 
Figure 4. Protozonites, and rudimentary appendages of head and thorax. 
Figure 5. The embryo a day older; sa, legs showing articulations. 
Figure 6. Eggs with the S-shaped primitive band floating in the yolk. 
Figui’e 7. The embx-yo about eight days old. 
Figure 8. The embx-yo just ready to hatch. 
Figure 9. Embryo two days old. 
Figure 10. Eggs three days after being laid; 10, side view of primitive baud; 10a, ventral 
view of the band; 106, dorsal view of the egg, showing abdominal end of the germ. 
Figure 11. Ventral view of germ (the yolk cells are represented three times too large, as 
the largest oxxly measure 1-4000 of an ixxch in diameter). 
Figure 12. Embryo three or four days old. 
Figure 13. Embx-yo ready to hatch. 
[The observations were made with Zentmeyer’s 1 1-2 and 4-10 inch objectives.] MEMOIRS PEABODY ACAD SCIENCE. 
3 v MEMOIR PLA7T; I 
r'. A.WALK.E.R, 
PACKARD'S MEMOIR GIST THF. DEVELQPMENT OP INSECTS MEMOIRS PEABODY ACAD. SCIENCE 
■3? MEMOIR PLATE 2 
CA.WALKEA 
PACKARD’S MEMOIR ON THE DEVELOPMENT OP INSECTS. MEMOIRSFEABODT ACAI). ■ 'CKNCI 
■‘MfruV.-fi j’UVF.S 
C.A.WALKER so 
PACKARD'S MEMOIR ON THE DEVELOPMENT OP INSECTS.