Biological series

mae eh ede PH EAATYS. i eSrSS veel renee ae’ 4 W
wee Aven eee iadagetoll bart ; _ ant Bnet reve Nein eam JH clan. fo te ete hy
Notice eve 4 nuepicatetainnate faeuden

haa
Ah

ee et

Sy } ary je
Tt) Raa
¥ ¥

p
Pa ls 1|

LAW ne, ve MARE ol Wee

UNIVERSITY OF TORONTO
STUDIES

BIOLOGICAL SERIES

No. 10: THE COLOUR CHANGES OF OCTOPUS VULGARIS
LMK., By E. V. Cowpry

THE UNIVERSITY LIBRARY: PUBLISHED BY
THE LIBRARIAN, ror1

University of Toronto Studies
COMMITTEE OF MANAGEMENT

Chatrman.: ROBERT ALEXANDER FALCONER, M.A., Litt.D., LL.D., D.D.
President of the University

PROFESSOR W. J. ALEXANDER, Pu.D.
Proressor W. H. Et is, M.A., M.B.
Proressor A. KIRSCHMANN, PH.D.
PROFESSOR J. J. MACKENZIE, B.A.
Proressor R. Ramsay Wricut, M.A., B.Sc., LL.D.
Proressor GEORGE M. Wronc, M.A.

General Editor: H. H. Lancton, M.A.
Librarian of the University

THE
GOLOUR GHANGES OF OCTOPUS VULGARIS LMK.

BY

E. V. COWDRY, B.A.

TABLE OF CONTENTS

Introductory Note

. Literature . ay es

. Material and methods

. Behaviour .

. Colour changes it ib MACHT SAG chs

. The rédle of the eyes and of the central nervous

system

. The effect of changes in the environment .

. The effect of changes in the water, and of light
. Discussion

. Summary of conclusions

Bibliography

Ver ae FY 1) ak YANy

iA? vi ees fi

Wes

Jah seg nee Ta

te

sted cea REY RLF te ate AE Ba A bi

Hk in OR AL MANN on

hs

De Oh |

THE COLOUR CHANGES OF OCTOPUS VULGARIS LMK.

INTRODUCTORY NOTE

The work embodied in this paper was done at the Bermuda
Biological Station for Research during the summer months
of 1909 and 1910. The object was to study the colour changes
of the Octopus from the standpoint of animal behaviour.
The expenses of the first summer were paid by the University
of Toronto. It is a great pleasure to thank Professor E. L.
Mark of Harvard University, the director of the Biological
Station, for very much kindness and for many extremely
valuable suggestions, and also Professor F. W. Carpenter,
who was the director in the summer of 1909 and under whom
the work was initiated. My thanks are also due to Mr.
Louis Mowbray, the director of the Aquarium, who allowed
me ready access to the tanks of the Aquarium at all times
and helped me in the collection of material.

I. LITERATURE

The literature on the changes in colour of the Cephalopods
has been recently (1906) so very carefully reviewed, in chrono-
logical order, by Van Rynberk that it will be quite unneces-
sary to go over the same ground in the same way. Mention
will simply be made of the condition of our knowledge of the
factors operating in the colour changes. The forms in which
they appear in the most marked degree and upon which the
great majority of the investigations have been made, are
classed among the Dibranchiate Cephalopods, chief among
which are the Cuttlefish and the Squids of the order Decapoda,
and the Octopi of the order Octopoda.

It has long been known that alterations in colour result
from the movements of the chromatophores. These struc-
tures are large pigment-containing cells distributed in the

6 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

superficial layers of the skin over the whole surface of the
body. By equal dilation of these cells the pigmented surface
is increased and the animal consequently becomes darker in
colour; on the other hand by unequal dilation in different
regions a great variety of patterns may be produced. Some
of these colour changes are very striking, so much so that
they have been observed and described by a series of
workers extending from Aristotle to the present day. The
iridocysts are light-reflecting cells of a yellowish colour dis-
tributed like the chromatophores in the dermis; they cause
the peculiar iridescent shimmer which is so often observed.
The chromatophores consist, first, of a spherical or oval body,
which contains coloured pigment, and, secondly, of slender
radially arranged processes, which connect the body to the
surrounding skin musculature.
The chromatophore is generally looked upon as being
developed from a single cell; but opinions still differ very
widely on this point. The body is, no doubt, multinucleated,
for Chun (1901), Steinach (1901), and all the later observers
come to this conclusion. Small nuclei are arranged at the
bases of the radial processes, about the periphery, and a large
central nucleus also has been described. The fine granules of
pigment in a single chromatophore are of the same colour,
but different chromatophores are often coloured differently.
In some species they may be of one colour only, while in
others there may be chromatophores of several varieties of
colour, such as purple, orange, brown, etc. The wall of the
pigment cell is predominantly elastic, and a band of tissue
has been observed about the equator in the plane of the radial
processes. Steinach (1901a, pp. 8-9), using Hansen’s modi-
fication of the Van Gieson method, finds that in a variety of
forms this band and the radial processes are stained yellow:
he concludes from this that they are muscular in nature. He
did not use any other differential connective tissue stains.
lofmann (1907a, p. 418) states that this band is formed by
the spreading out of the insertion of the radial muscles upon
the surface of the chromatophore.
Chun (1901, p. 172) believes that the radial processes are
formed from the developing chromatophore by the protrusion

Cowpry: COLOUR CHANGES OF OcTOPUS VULGARIS 7

and differentiation of a number of protoplasmic extensions.
Steinach found that in preparations stained with cochineal
and picrofuchsin these processes exhibited a longitudinal
striation, and he figures a nucleus in the wall of the chromato-
phore at the base of each. Hofmann (1907<, p. 390), working
on Loligo, calls attention to the physiological significance of
the abundant interlacing of the radial processes of neighbour-
ing chromatophores.

The question of the innervation of these structures has
been much debated; but it seems that the conclusions arrived
at by Hofmann are final, namely, that with his methylene
blue method nerve endings may be seen in the radial processes,
but not in the wall of the chromatophore itself. He calls
attention to the fact that a single nerve innervates a definite,
limited group of chromatophores, and that such areas of inner-
vation often overlap, so that certain regions may receive
nerve twigs from more than one source. He believes, further,
that there is in the skin a continuous peripheral nervous net-
work for the propagation of impulses. It will be readily seen,
therefore, that there is in the skin of Cephalopods a very
elaborate mechanism well calculated to produce a great variety
of colour changes, which may be sharply defined, or the
individual shades may grade insensibly the one into the other.

The chief credit of showing that the radial bands are of
muscular nature is due to Phisalix (1892) and Steinach (1901).
Phisalix (1892, p. 222), working on Sepza officinalis, destroyed
the central portion of the chromatophore with a fine needle
point and left the periphery intact. She found that, in such
cases, the movements of expansion were not interfered with;
but that after the destruction of the radial processes only, no

dilation was possible. Steinach, in addition to his experi-
ments with the Van Gieson stain, brings forward the following
arguments in favour of the muscular nature of the radial pro-
cesses. (a) There is a difference in the character of the move-
ments of the skin and of the chromatophores. (0b) Pulsating
chromatophores occur when the skin is in complete rest.
(c) The chromatophores are often quiet when the skin is in
motion. (d) The arms of Eledone moschata react to electrical
stimulation fifteen hours after amputation by skin contraction

.

8 CowpRy: COLOUR CHANGES OF OCTOPUS VULGARIS

and browning; but later by skin contraction only. It is,
therefore, generally accepted that the dilation of the chro-
matophores results from the contraction of the radial pro-
cesses, and that the contraction is due chiefly to the elasticity
of their walls.

Now that we have dealt with the character of the funda-
mental processes involved in the movements of the chromato-
phores, it becomes necessary to consider what is known of the
causation of these movements. There is no doubt that the
chromatophores can move independently of any impulses
from the nervous system, either central or peripheral.
This is shown by experiments in which the nerves to certain
areas have been cut and also in degenerations. In the first
instance it can be easily demonstrated that, while section of
the pallial nerve, which sends branches to the dorsal surface
of the mantle of one side, results in the suppression of all
reflex changes in colour of that region, still the chromato-
phores react to certain stimuli. In the second place, Phisalix
and Steinach have called attention to the fact that an arm
kept in sea water appears to have lost all reflex irritability
ten to sixteen hours after amputation; but that movements of
the chromatophores may be produced for about thirty-four
hours later. In connection with this direct stimulation the
valuable results of Hofmann and Hertel will be considered in
greater detail.

Hofmann (1907b, pp. 448-450) observed that an ampu-
tated Octopus arm bleaches if it is left in sea water; but if it
is exposed to the air and is not allowed to dry it becomes
brown. He thought that the bleaching in sea water was due
to the deficiency in oxygen and made experiments to prove it.

He took a piece of skin, spread it out on a glass plate, and
covered a portion of it with a tightly fitting cover-glass.
After from fifteen to twenty minutes the skin under the cover-
glass was bleached, but the remainder was still dark in colour.
He then covered the skin so that the oxygen could diffuse in
from the sides and it remained dark. And finally he covered
another piece, closely enclosing a bubble of air, beneath which
the skin became dark. Hofmann obtained the same results
with Sepia and Eledone, and argues that the concentration of

CowbDRY: COLOUR CHANGES OF OcTOoPUS VULGARIS 9

the carbon dioxide in the skin under the bubble and in the
surrounding portions is practically the same, and that the
bleaching of the skin is caused by the absence of the oxygen
only.

Hertel, in the same year, made some detailed observations
upon the physiological effect of many varieties of light rays
upon pigment cells. He experimented with dead Loligos
and found that there was an elective action of the blue rays
upon the yellow cells and of the yellow rays upon the violet-
red cells, and concludes that different cells are adapted to
receive light of different wave lengths. Further, in Loligo
and Octopus application of ultraviolet rays caused a deepen-
ing in colour and movements of flight; while blue and yellow
light caused only the change in colour. He argues that since
movement follows slowly when ultraviolet rays are applied,
particularily to regions devoid-of chromatophores, and since
illumination with yellow and blue light of skin possessing
chromatophores in the intact animal does not cause any move-
ment, it follows that the chromatophores do not play any part
in the initiation of the movements. Hertel believes that, in
this case, locomotion results from the stimulation of the
cutaneous nerves directly by the ultraviolet rays, and that
the reflex thus generated .passes through a special reflex arc
to the suckers. He therefore considers his most important
conclusion to be that the nervous substance can be directly
stimulated by ultraviolet rays, but by visible rays through
the help of the pigment only. The irregular movements and
the streaming of the individual granules of pigment are also
minutely described. He used atropine to paralyze the nerves
connected with the chromatophores, and concludes further
that light stimulates the movements of these cells directly
through the medium of the pigment.

Victor Bauer (1909, p. 183) sums up our knowledge of the
centres which control the movements of the chromatophores
very much as follows:—The centres for the play of the chro-
matophores seem to lie in the basal portions of the two last cen-
tral ganglia. Stimulation of these regions either by section or by
heat produces darkening (Phisalix, 1892, p. 216). The tracts
for coloration pass through the posterior commissure and the

10 CowprRy: COLOUR CHANGES OF OcTOPUS VULGARIS

subcesophageal ganglia, as has been shown by section experi-
ments (Uexkiill, pp. 603-604). Destruction of the pedal
ganglion results in the paralysis of the chromatophores of the
opposite side. The tracts of the chromatophore nerves ap-
parently cross here also (Phisalix, 1892, p. 215). Phisalix
postulates, further, special inhibition centres in the supraceso-
phageal ganglia, and contends that the bleaching reflex, which
consists of a diminution in size of the chromatophores, is sup-
pressed by destruction of this part: when only half is des-
troyed the reflex is retained on both sides (1892, pp. 98-99).

Hofmann (1907b, p. 420) has worked out very carefully
and accurately the innervation of the chromatophores and
concludes that the evidence for the existence of inhibitory
nerves is quite insufficient.

Two reflexes have been described by Steinach (1g9o01a,
pp. 27-28) in Eledone and Octopus. The first is from the
suckers along the centripetal nerves to the coloration centre
in the brain, and thence by the efferent nerves to the chromato-
phore muscles. He believes that certain colour changes
result from the strengthening of the peripheral tonus of the
chromatophores by impulses passing through this reflex arc.
The second is supposed to explain certain movements of loco-
motion which follow light stimulation. This reflex, he
believes (1901b, p. 40), passes from the chromatophores to
the suckers by purely muscular paths without the intervention
of any nervous elements. Hertel (1906) very rightly objects
to this idea.

There have been very few observations recorded on the
effect of alterations in the temperature and the purity of the
water upon the colour changes, and none on the results of
changes in pressure and salinity. Hofmann’s work, already
mentioned, dealt with the effect of oxygen and carbon dioxide
in the water and thus indirectly with the question of its
purity. Phisalix (1894, p. 93) records an experiment in which
she increased the temperature of the water, wherein she had
placed a Cuttlefish, so that, in about an hour, it rose twenty-
four degrees, and the animal became lighter and lighter in
colour until at last it died in extreme pallor. With a dead
animal she obtained quite different results, for it became

Cowpry: CoLOUR CHANGES OF OCTOPUS VULGARIS It

darker when the temperature of the water was raised. Stei-
nach (190Ia, p. 23) applied the heated end of a hook-formed
dissecting needle to the periphery of the chromatophores and
caused them to pulsate. Where he let the point of the
needle rest the skin became white; for the muscles were injured
by the strong and continuous heating. He found also that
continued milder warming paralyzed the radial bands and
led to bleaching of the skin. Hofmann (1907), p. 447) writes
that in sudden temperature changes, expansion results from
warming, retraction from cooling, and that the result of main-
taining the same temperature is bleaching in warm water
and darkening in cold.

Finally, reference must be made to the relation between
these colour changes and the habits of the living animals.
The idea of protective coloration originated with Aristotle.
Fredericq (1878, p. 573), Klemensiewicz (1878), and many
others believed that these animals changed colour to corre-
spond to their environment. Bauer (1909, p. 187) quotes
Klemensiewicz to the effect that the brightness of the back-
ground calls for a reflex coloration of the skin, and that section
of the optic tract shows that the eyes are the receptors;
because after this operation the chromatophores react in no
way to alterations in the colour or the brightness of the
bottom.

This conception of protective coloration has been accept-
ed by many travellers who have had just fleeting glimpses of
the animals in their natural environments. Charles Darwin
observed the habits of some Octopi in the Cape Verde Islands,
and writes in his “‘ Voyage of a Naturalist round the World”
as follows:

“They [the Octopi] seem to vary their tints according to
the nature of the ground over which they pass: when in deep
water their general shade was brownish purple; but when
placed on the land or in shallow water this dark tint
changed into one of a yellowish green.”

The recent work of Steinach appears to throw some light
upon these aileged changes in colour to correspond to the
environment. It has already been mentioned that he records
a reflex in Eledone from the suckers to the chromatophores

12 COWDRY: COLOUR CHANGES OF OCTOPUS VULGARIS

by which their tonus is increased and colour changes result.
He goes on to say that the function of this reflex mechanism is
to give to the animals in their natural surroundings that
coloration which renders them least noticeable. If they are
attached to smooth bodies, such as the coarse gravel of the
sea-bottom, or to the veined, weathered rock of the shore,
they assume a flecked or marbled appearance. On the other
hand if the animals are on the sand, where the suckers cannot
be firmly attached, the flecked coloration is reduced or absent,
on account of the reduced tonus: the skin becomes quite
light-coloured, speckled, and appears tolerably well adapted
in shade to the surface of the sand. These animals take on
a uniform inconspicuous grayish brown coloration when the
sucker apparatus is inactive, as In swimming and when, under
natural conditions, they are subjected to great danger. He
adds that in strong sunlight this adaptation, resulting from
the changes in the consistency of the bottom, does not obtain,
and that light is a second and more potent factor in the colour
changes.

The question as to whether there is such a thing as warn-
ing coloration in the Cephalopods is important. I have been
able to find this phenomenon recorded only once; _ for
Steinach (1901a, p. 24) writes that Sangiovanni (1823) came
to the conclusion that these animals changed their colour
to frighten away their enemies. He himself ridicules the idea.

Fredericq (1878, pp. 10-12) records the observation that
the changes in colour of Octopus vulgaris result from certain
emotions, such as anger and fear. On extreme irritation,
caused by the insertion of a stick in the mantle cavity, the
animal becomes furious and turns dark in colour. Move-
ment of the hand across the glass window of the aquarium
in the direction of the animal causes the chromatophores
about the eye to become active, especially those in the
longitudinal axis of the pupil, so that a dark streak appears
instantaneously at its two extremities and dilates. Phisalix
(1892, p. 218) writes that on irritation these animals
generally become quite black; but when they are repeatedly
annoyed or under the influence of fear, the dark colour fades
and the animal becomes quite pale. She goes on to say that

Cowbry: COLOUR CHANGES OF OcToPUS VULGARIS 13

it is only necessary to place a Sepiola in the presence of a
Cuttlefish to see it bleach immediately.

Hofmann (19074, p. 388) describes what he calls a terror
reflex in Sepia; this consists of a striking local expansion of
the chromatophores in a small round area on either side of
the median line on the dorsal surface of the mantle, the re-
mainder being white. He says that he has often observed
this pattern when he put his hand near the animal or cast a
shadow upon it. Hofmann also describes a ‘‘Zebrastreifung,”’
characterized by alternate dark and light bands upon the
dorsal surface of the animal, as occurring when two Sepias
approach each other.

And finally Annie Isgrove (1909, pp. 10-12), in a memoir
on Eledone, calls attention to certain colour changes which
were observed in the Eledone tank of the Plymouth aquarium.
She mentions that their general colour while at rest is buff
with flecks of cream, and continues as follows:

‘When the animal is excited the skin becomes of a very
dark reddish terra-cotta tinge. When Eledone is frightened in
any way the skin changes colour, and an intense pallor spreads
over it, causing it to become quite ghostly in appearance.
At this time the eyes stand out very prominently, because the
iris remains dark orange, as does the eyelid surrounding it, and
thus an orange patch marks out the eye on a whitened body.
However, under normal conditions, this patch does not stand
out in any way. At the same time that the pallor is seen the
animal tries to escape by rapidly swimming backwards, and
attempts to eject ink. Almost immediately the pallor is
replaced by an intense darkening or blush of terra-cotta colour
over the whole body. If allowed to come to rest now, the
colour gradually lightens until the normal condition is
reached.”’

From this brief résumé it will be readily seen how incom-
plete is our knowledge of the phenomena of colour changes
in the Cephalopods.

Il. MATERIAL AND METHODS

Octopus vulgaris, Lmk., was the only Cephalopod used in
this research. It occurs in Bermuda in fairly large numbers,

14 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

and its hardiness and great vitality render it a very suitable
animal for experimental work. Normal adults, larve, and
experimentally blinded animals were used, but in addition
to these, animals were studied in which the nerve supply to
certain regions had been cut off. The adults had an expanse
of from I to 1.5 metres, and the larve were about I mm. in
length. The general proportions of the former are shown in
the illustrations. In the male the third arm on the right is
hectocotylized.

The specimens have been oriented for description in the
manner adopted by Hoyle (1886, p. 53). The animal is
regarded as being stretched out with the arms pointing in
one direction and the apex of the mantle in an opposite one.
The tips of the arms are then said to be anterior and the apex
of the mantle posterior; while that surface which bears the
siphon and the mantle opening is ventral, and the opposite
one dorsal. The side of the arms which is turned toward the
mouth and is supplied with suckers is termed the inner and
the opposite one the outer. The same principle is applied
when the two surfaces of the interbrachial membrane or
umbrella are referred to.

A number of water-colour paintings have been made from
the living animals. In order to obtain the correct propor-
tions for these, the outlines of Figs. 1 to 6, inclusive, were
traced from a photograph of an anaesthetized animal. The
only alteration was the addition of the siphon which did not
appear in the photograph. In the case of the remaining
illustrations, Figs. 7 and 8 are free-hand sketches, and Fig. 9
is a representation of the colours used in the experiments on
the environment.

The observations were made, (a) in the natural environ-
ments, (6) in an Octopus car, (c) in the tanks of the aquarium,
and (d) in large glass aquaria in the laboratory. The natural
environments will be described in detail in the next section.
The dimensions of the Octopus car were, 1.8 metres long,
1.2 metres deep, and 1.5 metres in breadth. It consisted of a
strong wooden frame lined with galvanized wire of 1.2 cm.
mesh. A lid was arranged opening on hinges and in the con-
struction as little wood as possible was used. This cage was

CowpDRy: COLOUR CHANGES OF OCTOPUS VULGARIS 15

placed on a sandy bottom near the laboratory so that its
bottom was covered by water to a depth of about 75 cm. at
low tide, and its top to a depth of about 15 cm. at high tide.
Great care was taken throughout the experiments to prevent
the animal from seeing the observer, and to preclude any
movement or vibration of the car, or any other means of
irritation. The tanks used for the Octopi in the aquarium
were 1.5 metres long, 1.3 metres broad, and 1.3 metres deep.
The bottom and three walls were constructed of flat and
fairly smooth plaster; while the fourth side, through which
the observations were made, was of glass. The light came
in from above, so that the observer was in semidarkness and
practically invisible to the animals. The walls were covered
with a fairly even, but rather thin coating of dark green
algae. The colour of this growth is well represented by
sample no. 8 of the colour chart, Fig. 9. The tanks were
constantly supplied with fresh sea-water, which was forced
in at the bottom. The glass aquaria used in the laboratory
were cylindrical and had a capacity of about thirty litres.

III. BEHAVIOUR

This species of Octopus occurs in Bermuda among the
rocks of the shore, upon the reefs, in old wrecks, and in any
other sheltered places in depths varying from one to six
metres. It is sometimes found in tidal pools and its presence
is then detected by the water which it ejects from its siphon
when disturbed. Octopi are especially numerous, however,
among old pipes and tins in places where refuse is thrown into
the sea; and also on reefs which rise up from a bottom covered
with Eel grass (Zostera marina), and upon which mussels and
scallops are to be found. Each animal occupies a nest which
consists of a sheltered place of some description. It is often
a cranny in the reef, and may be from one to two metres in
extent. On the other hand, the nest may be simply the inside
of an old kerosene-tin. These nests are generally rendered
very conspicuous by piles of from fifty to a hundred mussel
and other shells, which the animals heap up outside of them.
Indeed, it may be determined by the freshness of these shells

16 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

whether the nest is inhabited or not. It is not known how
long one animal occupies a single nest; but it seems from the
size of the heap of shells that it cannot be for longer than a
few months, or at most a year. It could not be determined
whether they inhabit the nests all the year around or only
at certain seasons. There may be a period during which they
roam about and have no settled abode. During the months
of July and August two Octopi were never found associated
in the same nest.

The method of locomotion of these animals is very inter-
esting. It consists of either crawling or swimming move-
ments. The former is brought about chiefly by the con-
traction of the extended arms; but in some cases the arms
seem actually to push the animal along. When the motion
is in a sidelong or backward direction the water ejected from’
the siphon seems to be of assistance. They sometimes
scuttle along by using the distal portions of their arms in a
very peculiar, almost graceful, tip-toe sort of way, but they may
move by using only the proximal portions of their arms, the
tips being elevated. The interbrachial membrane serves as
a kind of parachute which enables the animal to glide easily
from one place to another, or to sink slowly to the bottom.
In such cases the arms are often waved in such a way as to
suggest that they aid in progression. Swimming is carried
out in the characteristic cephalopod manner. Water is
powerfully ejected from the siphon and as a result the animal
shoots along with the apex of the mantle in advance, and the
arms trailing out behind. When they swim for some distance
they do so in a jerky way, the accelerated periods corres-
ponding to the times when the water is forced out. An
Octopus breathing in water by expansion of the mantle is
represented in Figures 7 and 8. It is a peculiar fact that the
larve, for at least one week after hatching, can swim in
either an anterior or a posterior direction. They all died
about seven days after hatching so that it could not be de-
termined just when the ability to dart in a forward direction
was lost; but its loss was probably caused by the great in-
crease in the relative size of the arms and of the interbra-
chial membrane.

Cowbry: COLOUR CHANGES OF OCTOPUS VULGARIS 17

In Bermuda, copulation takes place during the months of
April and May. It was not observed in the natural environ-
ments, but I learn from Mr. Mowbray that in the tanks of
the aquarium, the animals pair off, and that the members of
the individual pairs do not molest one another. Furthermore,
I was told that during this period, the animals maintain,
more or less continuously, a uniform dark reddish brown
coloration, like that represented in Fig. 3. Eggs were
laid about the beginning of July and attached to the walls of
the tank, in a corner, in the form of thick, white, gelatinous
threads. They may also be laid in the month of October.
The female attached herself to the wall, so as to cover the
eggs, and from time to time kept the water in circulation
about them by the contractions of her mantle. The animal
very rarely moves away from the eggs until they are hatched,
not even to obtain food. In one instance an Octopus so
persistently annoyed one of these females that she deliberate-
ly left her nest and killed it. The colour of the animals at
this time is light gray and is represented in Fig. 1. The in-
ner surface of the arms is sometimes blotched with red. The
rate of respiration is about thirty-four times per minute.
When the larve begin to escape from the egg-capsules the
female becomes very excitable, swims about the tank a good
deal, and will dash at one’s hand if it is placed in the water.
She pulls down most of the eggs from their point of attach-
ment and scatters them about. When the animal rests she
always attaches herself to the place where the eggs were
originally deposited.

The Octopus preys on many animals, but it is particularily
fond of crabs and the large Bermuda crayfish. An Octopus
will swim after a crab for some distance and capture it by
alighting upon it from above with arms extended, or it may
creep along stealthily until it comes within striking distance
and then make a sudden dash at it. An Octopus was once
observed to capture a crab, to see another two or three metres
distant, pursue and capture it, still clinging to the first, and
so on until four crabs, each about eight centimetres in dia-
meter, had been taken, the crabs were then consumed one by
one. This is always done in a definite way; the crab is quickly

18 CowpRY: COLOUR CHANGES OF OCTOPUS VULGARIS

brought near the mouth, the legs are torn away and the
flesh of the body is extracted from the ventral side, so that the
carapace is left intact. This feeding reaction is seldom inter-
rupted; when it is once begun it is almost invariably carried
through to the end. The position assumed when feeding is
unmistakable. It is characterized by the humping up of
the mouth or head region, due to the presence of the victim
and also to the fact that the suckers on the proximal portions
of the arms are in action, the animal resting chiefly on the
distal portions of the arms. It is, therefore, quite easy to
tell from a considerable distance whether a certain animal is
feeding or not. Mussels, however, are the chief article of
food. The Octopus sallies out in search of them and brings
them back to its nest. It may make several journeys until
fifteen or twenty have been collected. The animal then settles
down at the entrance to its nest and begins its feast. It
could not be ascertained just how it opens the mussels which
have been collected in such numbers, but this is probably
accomplished by a continuous pull (through the medium of
the suckers) upon the two valves of the shell until the ad-
ductor muscle relaxes and finally gives way. This idea is
supported by the fact that the shells are seldom broken.
The Octopus will not take dead crabs or crayfish; but if it
has not had a meal for some time it will even kill and devour
other Octopi; neither will it attack a resting crab which it
has not, fifteen or twenty seconds previously, seen in motion.
This latter statement needs perhaps a little qualification. In
such cases the criteria for determining just when the Octopus
first noticed the crab were: a certain movement or gathering
together of the body of the Octopus; accompanied by a
change in colour, characterized by the production of a brown
band running longitudinally across one or both eyes (Figs. 2,
4 and 5) and the appearance of a greater or less amount of
brown mottling, the whole corresponding in time to a plainly
visible movement of the crab.

The very interesting question arises as to how the Octopus
finds its way back to its hole. This problem seems just as
difficult of solution as it is to explain how bees can find their
way back to their nest over great stretches of country. It is

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 19

not known how far an Octopus will venture away from its
nest, but they were occasionally seen at least fifty metres
distant. They certainly go out of sight of their hiding-
place and still find their way back again. The question of
memory in the Cephalopods is discussed by Uexkiill (1905)
and Polimanti (1910) both of whom worked on Eledone.

The adult Octopi have very few dangerous enemies, for
they can make themselves secure from attacks in their nests
and even in the open are well adapted to defend themselves.
However, certain large-mouthed fish, such as the Hamlets
(Epinephelis striatus) are said by the fishermen to attack
them in the open by making sudden dashes at them. With
the larve ,which are produced in great numbers, it is different.
They are at first free-swimming and it is likely therefore that
they would be rapidly devoured by plankton-feeding fish.

The adults sometimes eject ink when they are being cap-
tured or pursued. They generally throw out the ink when at
rest and then attempt to escape under cover of it, but it may
be squirted out in several jets when the Octopus is swimming.
If the animal is still pursued, after the ink has been once
ejected, it may throw out a second or even a third jet. The
coloration of the animal is very interesting at this time on
account of the belief that both it and the ejection of ink are
of use for protection. It is true that in the majority of cases
both the animal which throws out the ink and any others in
the immediate vicinity settle down, become quiet, and take on
a dark reddish brown coloration, which certainly renders them
almost invisible in the darkly tinted water. This coloration
is represented in Fig. 3 and often persists for five or ten
minutes. It must be stated, however, that the ejection of
ink is not invariably accompanied by such a darkening in
colour. It may be of interest to note that the larve of about
I mm. in length, and not more than a week after their
escape from the egg-capsules, often ejected ink when irritated.

Before this section is concluded brief mention will be made
of the method by which the Octopi were captured. The
white heap of shells is easily seen against the dark background.
If the water is more than a metre deep some common salt,
wrapped up loosely in a piece of paper, is placed well within

20 CowprRy: COLOUR CHANGES OF OcToPUS VULGARIS

the hole. The salt dissolving in the’ water irritates the
Octopus and it comes out and is easily captured in the hands
or in a landing-net. Sometimes the administration of the
salt has to be repeated before the animal is dislodged. When
the entrance to the nest can be reached without diving, the
procedure is more simple and rapid. A small quantity of
crude commercial formalin is squirted into the hole with a
syringe. This invariably brings out the Octopus immediately.
Most of the Octopi were taken in one of these two ways;
but some were brought in by the fishermen who find them in
their fish-pots.

IV. CoLour CHANGES

It is necessary to mention the structure and the distribu-
tion of the chromatophores, and of the cirri, and to discuss
the wave play and the pulsation of the chromatophores,
before the description of the colour changes is entered upon.

There are two kinds of chromatophores in Octopus vulgaris.
One is yellow and the other dark reddish brown in colour.
These two types are particularly distinct in the young larve,
for in them they are distributed quite differently. The
former is found on the ventral surface of the mantle, on
the dorsal and ventral surfaces of the head, and on the outer
sides of the arms. There is also a band of about eight chro-
matophores on the anterior dorsal lip of the mantle. The
latter, the dark reddish brown chromatophores, are seen only
on the dorsal surfaces of the body and head, on the ventral
surface of the head, and on the outer side of the arms. There
are no chromatophores of any kind on the dorsal surface of
the mantle.

In the adult the reddish brown chromatophores are about
0.1 mm. in diameter; and are slightly larger, more numerous,
rounded, and uniform in shape than the yellow cells. The
minute anatomy of these structures has already been dis-
cussed in Section I. In the living embryos, before they escape
from the egg-capsule, and in the larve for at least one week
afterwards, the radial processes and a round, clear, highly
refractile body in the substance of the chromatophore can
be easily distinguished. This round body is probably the

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 21

nucleus. In the adult neither the radial processes nor the
round body can be made out. Furthermore, it is interesting
to note that, although in these larve the chromatophores
are as large as they are in the adult, still there are only about
seventy-five of them inasingle animal. As the larve increase
in size new chromatophores must be formed until the adult
‘condition is reached.

The distribution and the abundance of the chromatophores
in the several body regions of the adult Octopus are of the
greatest importance. In the neighbourhood of the median
line on the dorsal surface of the mantle there are, on an aver-
age, about sixty of each variety per sq. mm., although there
are none in this region in the larve, while on the ventral aspect
the number is reduced to eight or ten per sq. mm. The skin
of the head, the iris, and the outer surface of the eye-ball are
also richly supplied. On the siphon and at its base thirty-
eight or more may be counted in a sq. mm., but there are no
chromatophores inside the mantle cavity. .As one passes
around the margin of the mantle opening it is seen that the
line where the chromatophores disappear is very definite and
sharply marked. They are more numerous on the outer
surfaces of the first two pairs of arms than they are in the
case of the third and especially of the fourth pair. The same
applies to the corresponding parts of the interbrachial mem-
brane. The inner side of the arms and of the interbrachial
membrane presents a peculiar condition, for here the chro-
matophores are distributed in patches. These patches vary
widely in shape and size and may be separated by as much
as 0.1 mm. The reddish brown chromatophores in these
areas in the case of the first pair of arms average about one
hundred per sq. mm., and the yellow ones only fifty. These
numbers decrease continuously as one passes to the second,
third, and fourth pairs of arms, and the intervening portions
of the interbrachial membrane. The two most striking
features in the distribution of the chromatophores are that
the relative number of the yellow variety is greatly reduced
on the inner surface of the arms; and that both types, in
general, are distributed more thickly on the dorsal surface
of the body and on the outer sides of the arms than elsewhere.

22 CowprRy: COLOUR CHANGES OF OcToPUS VULGARIS

When we remember that these structures are more numerous
on the dorsal than they are on the ventral arms we see that
the chromatophores are distributed more thickly on those -
parts of the body which are exposed to the light, to the attacks
of the enemy, and to the eyes of the observer.

In the living animals the skin is very loose and is thrown
into a number of folds and wrinkles, which become partially
straightened out when the different parts of the body are
extended. This unevenness of the skin is particularly notice-
able on the dorsal surface of the mantle, between the eyes,
and on the outer sides of the first two pairs of arms, together
with the intervening portion of the interbrachial membrane.
These wrinkles appear like fine dark lines and are illustrated
in Fig. 5.

The cirri are of especial interest because it has been sug-
gested that they have a protective function, inasmuch as
when elevated they make the animal indistinguishable in
irregularity of surface from its surroundings. ‘These struc-
tures have a constant and very definite distribution, which is
particularly represented in Figs. 5 and 6. The largest ones
are found on the dorsal surface of the mantle, between the
eyes, and on the outer surfaces of the first two pairs of arms.
Smaller elevations may be made out distributed thickly on
the sides of the mantle and on the outer surface of the inter-
brachial membrane. The shape of the cirri on the mantle is
peculiar, for when viewed from the side they present a fairly
narrow base which expands into a fan-shaped extremity; but
when seen from the front or back they look almost like spikes
with a broad base and a pointed apex. When fully extended
these structures may be from 1 to 1.5 cm. long. The cirri are
only occasionally elevated, and when they are in the retracted
condition, they are represented only by a slight puckering of
the skin. They are generally coloured in the same way as the
surrounding surface.

The cirri on one side of the body may be elevated indepen-
dently of those on the other side. This is probably an in-
stance of the bilateralism which is often apparent in the
Octopus. There is a group of three cirri over each eye, which
are arranged in longitudinal series, and are represented over

Cowpry: COLOUR CHANGES OF OcTroPUS VULGARIS 23

the left eye in Fig. 6. One or more of these may be very
much elevated over one or both eyes without being accom-
panied by the extension of any of the others. The causation
of the raising up of the cirri is very obscure. They were not
observed when the animals were in the act of swimming;
neither were cirri raised in regions from which the nerve supply
to the skin had been cut off; but it must be said that in this
case they were not particularly looked for.. Furthermore,
the cirri were elevated in animals from which either one or
both eyes had been removed, and it is perhaps worthy of note
that the Octopus was always, at the time, of a dark reddish
brown coloration. We see, therefore that it is very difficult
to discover a simple reflex the excitation of which would bring
about the elevation of these structures. Any reflex passing
through the eye has been exclided, and the only other reflex
path which seems at all possible is one which might pass from
the suckers to the erector muscles of the cirri. This possibi-
lity is favoured by the fact that cirri were never elevated in an
animal swimming freely, when obviously such a reflex could
not be functioning. The question now arises as to whether
there is any relation between the character, whether rough
or smooth, of the bottom and the raising of the cirri. Un-
fortunately no special experiments were made to determine
this point; but elevation was observed in the tanks of the
aquarium, where the walls and the bottom are firm and
smooth, and on the wire sides of the Octopus car, where the
surface was extremely irregular and afforded a very poor op-
portunity for attachment. The elevation seems to be con-
trolled by the central nervous system; for when there are
several Octopi in the tank, under the same conditions, the
cirri are only occasionally raised in some of them. Even if
this reflex does exist, it does not bring us very much nearer
an understanding of the rdle played by the central nervous
system or of why individual cirri over the eye, for instance,
are elevated.

Another peculiar phenomenon was observed. It consists
in an indefinite, indescribable quivering or vibration of the
colour pattern, which remains without any other perceptible
change. This agitation is not very noticeable and cannot be

24 CowprRY: COLOUR CHANGES OF OCTOPUS VULGARIS

observed from any distance. It does not seem to begin any-
where or to end anywhere; nor is it definitely localizable. It
occurs on the dorsal surface of the head and mantle, and on
the outer side of the arms and interbrachial membrane; but
it was not observed on the inner surface of the arms along
the peduncles of the suckers.

The changes which take place during such a vibration are
obscure. Observation of the skin with a hand lens failed to
show any definite changes; but the variations were probably
too delicate and too rapid to be detected in this way. When
a portion of the skin, out of water, was viewed in an oblique
direction the light which was reflected from it did not seem
to change in direction at all synchronously with the vibration.
It is obvious, therefore, that we are not dealing with a tremor
of the skin. Stroking the skin with the finger calls forth these
changes in a very marked degree. After a consideration of
the literature it seems probable that this strange quivering
in the colour pattern is identical with the ‘‘ Wave play”
which is so often mentioned.

The pulsation of the chromatophores is evidenced by
rhythmically occurring flashes of colour in fairly definitely
circumscribed regions, and is especially noticeable on the
inner surface of the arms, where the chromatophores, as
already stated, are distributed in patches. The chromato-
phores in one of these patches may be pulsating very vigor-
ously, while those in neighbouring areas may be completely
at rest. The rate is about twenty-five times a minute; but
it seldom lasts aslong as that. It seems to be more rapid out
of water. The pulsation was not very evident when the skin
on the outer surface of the arms was closely examined; neither
could it be made out satisfactorily on the head or mantle.
That it may take place on the mantle was shown, however,
in an experiment in which one of the nerves to that region was
sectioned (vide Section V).

The causation of the pulsation of the chromatophores is
likewise undetermined. It is certainly due to some peripheral
stimulus; for, as already mentioned, it occurs, in an even more
marked degree, when the nerve supply has been cut off. It
occurs also in regions distal to the point of section of the

CowbDRY: COLOUR CHANGES OF OCTOPUS VULGARIS 25

central nerve of the arm. As in the case of the wave play,
so also here stroking of the skin on the under side of the arm
with the finger causes pulsation of the chromatophores in
some only of the areas touched: here and on the surface of
the mantle from which the nerve supply had been cut off we
get pulsation resulting from mechanical stimulation. But still
we are no nearer an explanation of why, in the normal animal,
all the chromatophores in certain patches should pulsate;
while those in neighbouring areas, under apparently the same
conditions, remain at rest. It is true that the abundant
interlacing of the radial processes, as emphasized by Hofmann
(1907a, p. 390), may account for the fact that the chromato-
phores in any one of these patches always pulsate together as
a whole. On the other hand the possible existence of a peri-
pheral ganglion and of an individual nervous connection for
each area might offer an explanation of the phenomena.

We shall now pass to a consideration of the colour patterns
themselves. In all of them, as the result of the distribution
of the chromatophores, both the outer and the inner surfaces
of the dorsal arms are darker in colour than the corresponding
portions of the more ventral ones. The same gradation in
colour is exhibited in the interbrachial membrane. Further-
more, the dorsal surface of the mantle is always darker in
colour than the ventral surface. The colours range from
those at the red end of the spectrum to the orange and even
border on the green; but no purples, blues, or violets are ever
seen. In addition to these, shades of gray and brown are
very common. The colorations often grade insensibly one
into another; but for convenience of description they may be
arbitrarily divided into uniform, mottled, and striped phases.

1. Uniform phases—Under this heading four different
colorations will be described. They are represented in
Figures I, 2, 3, and 4 respectively. In general these color-
ations seem to be fundamentally independent of optic reflexes
(see Section V), although they may almost all be initiated by
them.

The first (Fig. 1) is a light shade of gray tinged with brown
or even very lightly with green. It is extremely variable in
this respect. The colour is often appreciably darker between

26 CowprRy: COLOUR CHANGES OF OcToPUS VULGARIS

the eyes and on the dorsal surface of the mantle. The siphon
occasionally appears quite brown against the light gray back-
ground of the arms, and the peduncles of the suckers are
darker than the remainder of the inner surface of the arms
and the interbrachial membrane. This coloration is often
complicated by the dilation of the chromatophores in some
of the patches on the inner surfaces of the arms, resulting in
an irregularly distributed red colour, or the location of the
cirri may be marked out by brown blotches. This appear-
ance generally obtains when the animals are at rest, either
at the entrance to their nests, in the tanks of the aquarium or
elsewhere. In the tanks of the aquarium they are often
attached to the walls by the suckers on the proximal portions
of theirarms only. Their eyes are frequently closed and they
do not seem to notice the movements of anything near them
in the water. The closure of the eyes is brought about by
the contraction of the pupil and the rotation of the eye-ball
ventrally, accompanied by a partial closure of the eye-lids.
This coloration very commonly occurs also when the vitality
is reduced by disease or after operation; but it likewise obtains
when the animal is swimming freely or crawling actively over
the surface of the bottom.

Another uniform coloration is represented in Fig. 2. It
is a shade of brownish red with a distinct tinge of green about
the edges of the outer surface of the interbrachial membrane
and between the eyes. The fine black lines marking the folds
in the skin are particularly prominent. There is also a sug-
gestion of a white blotching on the dorsal surface of the head
and mantle and on the outer surface of the arms and inter-
brachial membrane. The brown band at either end of each
eye-slit, which runs in a longitudinal direction, is very inter-
esting and will be discussed subsequently. The under side
of the arms and of the umbrella is lighter in colour and even
shows an almost orange tint. The peduncles of the suckers
are reddish brown. This phase of coloration is often very
strikingly modified by a great increase in the brightness and
the number of the white spots. These blotches are roughly
rounded and are larger and further separated on the head and
on the dorsal surface of the mantle than they are at the edges

Cl LO I

st

CowpDRY: COLOUR CHANGES OF OCTOPUS VULGARIS 27

of the interbrachial membrane, where they are densely
crowded together. The whole animal may therefore present
a peculiar dappled appearance, which is exceedingly bright
and is almost impossible to represent adequately in a painting.
This spotted coloration seldom persists for more than five
seconds. The original pattern, as represented in Fig. 2,
occurs when the Octopus is active, crawling or swimming, or
when it is touched by other Octopi, or witha stick. It is not
very common.

The next coloration, illustrated in Fig. 3, is the simplest
of all, for it is quite uniform. | It is one of the few phases of
colour mentioned in the literature; for Fredericq, Phisalix,
and Annie Isgrove all refer to it (see Section I). The reddish
brown colour is considerably lighter in the case of the ventral
arms_and of the corresponding portions of the interbrachial
membrane. It is also lighter on the ventral surface of the
mantle. The iris and the eye-ball share in the dark color-
ation. Cirri may or may not be elevated; but a few of them
are represented as raised in Fig. 3. Some animals show a
decided tendency toward the production of this coloration.
It sometimes appears when the Octopus is continually irri-
tated, either by being poked with a stick or when annoyed
by other animals. It is of common occurrence when the
Octopi are feeding. This uniform reddish colour may persist
and also appear when the animal is in the act of swimming, or
it may flash out for no known reason when the Octopus is at
rest. It has already been mentioned that it is fairly constant
and that it remains during the greater part of the period of
copulation; but the most interesting fact of all is that it often
appears immediately after ink has been ejected. The matter
is further complicated when we remember that the other
animals, in the immediate vicinity, take on the same color-
ation, although they have not themselves thrown out the ink.
The darkening in colour of the animal, which did not eject
the ink, obtains before it is enveloped in the cloud of pigment:
indeed before the darkly tinted water even touches it. This
points to the conclusion that it is the stimulus caused by the
sight of the ink in the water and not the exertion, or the
accompanying sensation of throwing it out, which causes
this coloration.

28 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

The pattern represented in Fig. 4 can hardly be termed
uniform, but still it may be most conveniently considered
under this heading. The dark lines which mark out the
peduncles of the suckers appear in striking contrast to the
brilliant white of the rest of the body. The brown lines along
the eye-slits are typical and fairly constant, and the siphon
is often of a light brownish shade. The iris is white. The
eminences formed by the eye-balls are not at all prominent.
This results from the lateral rotation of the eyes and the
increased distance between them. The animal sometimes
seems to crouch down; but at others it arches its arms over
its body and presents their inner surfaces as if to ward off an
attack. Phisalix and Annie Isgrove call attention to a
bleaching of this sort in Sepiola and Eledone respectively.
It occurs when an Octopus first notices the approach of an
enemy, such as a large fish, or it may flash out when one ap-
proaches it in the water, or touches it with the hand or any-
thing else; and it is peculiar in that it does not obtain in
totally blinded animals (Section V) whereas the darkening
(Fig. 3) does. It is very transient and only obtains for a
few seconds. When the Octopus is already dark in colour,
it often—especially when the stimulus is not sudden—passes
through the dappled phase (vide supra) before taking on this
coloration. This colour was never seen when the Octopus
was swimming, and it does not occur when the vitality of the
animal is lowered for any cause.

2. Mottled phases—The mottled colorations are the most
numerous and the most diversified of all the colour patterns
exhibited by the Octopus. Some of them are so brilliant and
so complicated that all attempts to reproduce them on paper
are in vain. They are more dependent than are the uniform
colorations upon optic reflexes. Two of the more simple are
represented in Figures 5 and 6.

In the first (Fig. 5) the ground colouring is light grayish
brown, and upon this a number of dark yellowish blotches
are superposed. These blotches are slightly irregular in
shape and may be confined to certain parts of the animal.
In the coloration figured they are restricted to the right side.
Their location seems to be constant: that is, they always

CowpryY: COLOUR CHANGES OF OcroPUS VULGARIS 29

appear in the same places. They are continuous from the
outer to the inner surfaces of the arms and they are also found
on the outer side of the interbrachial membrane. These
patches become smaller and more closely set together as one
passes from the base to the extremity of the arms. They are
often wedge-shaped and are placed at right angles to the long
axis of the arm. The peduncles of the suckers are of a reddish
colour. The brown eye-band is shown on the right side, and
the iris is white.

The occurrence of this colour pattern is an excellent in-
stance of the bilateralism which frequently obtains in these
colorations, for the mottling often appears only on the same
side as the eye with which the animal observes a movement
of some sort. The line of division between the coloured and
the non-coloured portions is sometimes very sharp. If the
exciting stimulus is repeated the coloration may spread to the
opposite side and the whole animal may darken considerably,
to a shade very much like that represented in the next figure.
As a coloration of this sort usually lasts for only a very short
time, it is exceedingly difficult to determine the exact location
of the blotches; but they seem to appear time after time in
exactly the same places.

In Fig. 6 the ground colouring is of the same shade, and
the blotches are of approximately the same shape and distri-
bution. They are more evident on the dorsal arms than they
are in the case of the more ventral ones, and they are the same
on both sides of the body. There are several irregular white
lines and patches on the dorsal surface of the head and mantle,
and on the outer side of the arms and the interbrachial mem-
brane. These patches are extremely variable in their number
and size, and by their increase in both may alter the whole
appearance of the coloration, merging into the dappled phase
already described. In these regions also a few elevated
cirri may be distinguished, which stand out as brown dashes
against the lighter colour of the background. The iris on
both sides is white and the siphon is light brown in colour.
The blotches or the dark reddish brown bands on the arms
are, as in the coloration just described, continued round to
their inner surfaces. In addition to this, the inner surfaces

30 Cowpry: COLOUR CHANGES OF OcTOPUS VULGARIS

of the arms and of the interbrachial membrane present a fine
mottling of brown against a light yellowish ground colour.
The peduncles of the suckers are of a light steel-gray shade.

The exact conditions under which these dark blotched
colorations appear are unknown. They are associated with
activity, not with rest, and they seldom persist for more than
five or at most ten minutes. They appear when the animal
is in the act of swimming, but they are even more transient
when in this condition. Moreover, these patterns are of
fairly common occurrence when the Octopi are devouring
their prey.

3. Striped phases—As the mottled patterns are the most
variable, so the striped are the most constant. They are more
transient than either of the other two types of coloration and
are characterized by the suddenness and the abruptness with
which they flash out and then disappear again. The stripes
are always very definite, occur in precisely the same regions,
and throughout show a marked tendency to be controlled by,
optic reflexes. Three main types will be described.

The first of these is the more circumscribed and perhaps
the least noticeable; but it is by no means the least compli-
cated. It is illustrated in Figures 2, 4, and 5, and consists of
a principal band about 25 mm. long, by 3 to 4 mm. broad,
which runs antero-posteriorly and is continuous, on either
side of the eye, parallel to the long axis of the pupil: this holds
only when the eye is in its usual position and is not rotated.
When this area is coloured the chromatophores on the iris,
at either extremity of the pupil, and occasionally those along
its dorsal margin, are also dilated, so that a similar brown
coloration results. It is interesting to note that Fredericq,
as far back as 1878, clearly described this stripe and made
some notes on its occurrence (see Section I). In addition to
this principal stripe, two other smaller areas of coloration
may appear, the location of which is indicated on the left
side in Fig. 4. They are situated dorsad and mediad of the
eye. The posterior one is the larger and the more promi-
nent, and is often continuous with the principal stripe
ventrally and laterally. It was once observed that the whole
of the skin over the orbit became dark in colour; while the

CowpDry: COLOUR CHANGES OF OCTOPUS VULGARIS 31

remainder of the animal retained its uniform light gray shade.

These bands flash out when the animal notices the ap-
proach of a fish or any other object in the water. They only
appear across the eye which notices the movement. The
stimulus does not pass over from one side to the other; but it
was noticed that there were varying degrees of reaction.
This is shown by the fact that if disturbance is slight or
some distance off the coloration is only light; but if it is re-
peated or comes nearer, the colour deepens to a very dark
brown. Indeed, it is only when the animal is strongly
stimulated in this way that the accessory bands appear. The
principal band may pulsate and actually vary in intensity
in different parts. It may be darker in front of the eye or
behind it, or either of these parts may not be coloured at all.
The accessory bands never appear when the principal stripe
is absent; and of the two, the posterior one is the more promi-
nent and may flash out independently of the other. When
the animal sees the object with both eyes the coloration
appears on both sides of the head. These patterns appear
when the Octopi are at rest, crawling about, or swimming
vigorously. They apparently exhibit the phenomena of
fatigue; for following repeated stimulation, after they have
appeared and disappeared several times, there is a period
when they do not occur. If the Octopus is now allowed to
rest for ten or fifteen minutes and is again stimulated, in the
same way, it reacts by coloration as it did in the beginning.
These stripes do not occur in animals which are not perfectly
healthy and vigorous. As one would expect they do not
obtain when the animal is out of water; for the optic reflexes
would necessarily be reduced to a minimum.

The next coloration, that illustrated in Fig. 7, may be
regarded as a modification of the one just described. it is
characterized by a dark brown stripe, extending from the
distal end of the second arm over the optic prominence to
the posterior end of the mantle. It may occur on one or both
sides. This stripe is not always of the same extent, but its
breadth is fairly constant.

This long band is one of the rarest of the colorations, and
it is peculiar in that it seems to be of more frequent occur-

32 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

rence when the animals are swimming, although it was once
observed to flash out in an Octopus which had just come
to rest. Some animals show a strong tendency to exhibit
this pattern. It is generally produced when an Octopus,
showing the principal stripe (vide supra) across the eye,
is pursued while swimming, and it may be regarded as
an enlargement and an elongation of this stripe. There is
some reason to think that there is a correspondence between
the side facing the pursuer and the side upon which this great
band appears; but this could not be determined with certainty.
If it is true, however, as one would expect, then both bands
would appear when the animal sees the enemy with both eyes,
and this seems to be the case. Like the smaller stripes, they
are transient and fade away in three or four seconds, after the
stimulation is withdrawn or when the Octopus comes to rest.
They were never observed with animals in the aquarium tanks.

The last coloration to be described is the one shown in
Fig. 8; it is perhaps the most remarkable and interesting of
them all. The general colour of the animal is dark reddish
brown, somewhat similar to that illustrated in Fig. 3. This
is strikingly modified by the presence of white stripes. The
largest of these extends along the dorsal margin of the first
arm on the right, takes in a portion of the interbrachial mem-
brane, and runs along between the eyes to the posterior end
of the mantle. The relative breadth is indicated, but it
varies considerably in different cases. In addition, there are
two lateral stripes on the dorsal margins of the second and
third arms, and on the adjacent parts of the interbrachial
membrane, which come to a termination before reaching the
head region. It could not be seen whether there was a stripe
on the fourth or most ventral arm; but its presence is highly
probable. The ventral surface of the mantle presents the
same shade of white. Sometimes both the median and the
lateral stripes are confined to the interbrachial membrane, so
that the arms participate but little in the banding.

The dark background is generally the fundamental colour
and the white stripes seem to appear secondarily; but ex-
ceptions to this rule were observed. _The long median white
band occasionally appears when the Octopus is at rest, and is

CowpDRY: COLOUR CHANGES OF OcTOPUS VULGARIS 33

generally the precursor of movement; but the lateral bands
do not appear unless the animal is in the act of swimming.
The same holds in the case of the white shade of the under
surface of the mantle. The lateral stripes disappear almost
immediately after the Octopus comes to rest, and the ventral
surface of the mantle becomes of the same shade as the rest
of the animal. This disappearance may be effected by either
a darkening of the areas or a bleaching of the whole Octopus.

As in the case of all the other striped colorations, so also
here, they seem to result directly from the excitation of a
reflex through the eye. This hypothesis is supported by the
observation that the lateral bands occur only on the side of
the eye by which the Octopus notices the approach of another
animal. Sometimes, presumably when the stimulus affected
both eyes, the two sides become simultaneously striped. In
some animals these changes take place so rapidly that only five
minutes’ observation is sufficient to show the occurrence of
the long median band, either alone, or accompanied by the
lateral stripes on one or both sides. The lateral bands never
appear independently of the median band. The matter is
complicated by the fact that this coloration is limited to
certain animals, which show it with unusual frequency. This
tendency of certain individuals to produce time after time a
particular colour pattern is very remarkable and seems to
indicate the operation of factors of which we know nothing.
It is especially liable to occur when the animal in question is
disturbed by other Octopi; but it may also be elicited in
various other ways. When one Octopus in the tank con-
stantly shows this coloration, there seems to be a tendency
for some of the others to exhibit it also; for they do not do so
when it is removed. The males show it as well as the females;
but it occurred in the most marked degree in the case of a
young female. It seems to be the same as that mentioned
in the literature as a ‘‘Zebrastreifung.’’ Hofmann (1907<,
p. 388) writes that in Sepia this ‘‘Zebrastreifung”’ is char-
acterized by the alternate occurrence of dark and light bands
of colour, and adds that it occurs when two Sepia approach
each other. This is the least understood of all the colorations
which the Octopus assumes.

34 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

From the preceding description of the colour changes it
will be noted that they are extremely variable and complex:
indeed it seems that nowhere in the animal kingdom are the
colour changes so rapid and so brilliant. When one sees a
dark red Octopus on a white background, or striped animals
swimming in their natural environments, one cannot help
thinking that colour seems almost to have gone to waste.
The colour patterns were observed in males and females alike,
and they all occurred in a great variety of environments. In
the following pages an attempt has been made to explain the
causation and the purpose of these colour changes.

V. THE ROLE OF THE EYES AND OF THE CENTRAL
NERVOUS SYSTEM

This study was carried on from the operative standpoint.
The operations consisted in blindfolding or extirpating one or
both eyes, and in sectioning deep and cutaneous nerves. The
results are summarized at the end of the section. In all
cases, the anaesthetic used was magnesium sulphate, which,
although not entirely satisfactory, serves the purpose fairly
well. When used it is added, in small quantities at a time,
to a dish of water in which an Octopus has been placed; this
is kept up until the response to mechanical stimulation ceases.
To obtain this result about 500 grams to 2 litres of water are
necessary; the amount varying of course with the volume of
water used. The solution may be employed repeatedly
until the mucus secreted by the animal renders it foul. The
colour of the Octopus under this anaesthetic usually becomes
light gray, and the respiration slow and shallow. The animal
may be kept under the anaesthetic for ten or fifteen minutes
if the water in the dish is carefully and continuously aerated.
Respiration occasionally ceases entirely and reviving may
then be difficult. In such cases the animal was placed in
fairly cool sea water (about 25 C.) and periodically, about
every two seconds,.some of the water was squirted into
the mantle cavity, the mantle being squeezed in the intervals,
so that complete artificial respiration was kept up. The
Octopi are very hardy and almost invariably recover from
the operations.

CowpRY: COLOUR CHANGES OF OCTOPUS VULGARIS 35

Blindfolding was attempted as a means of eliminating
optic stimuli. A small piece of opaque oil-cloth was tightly
sewn over both eyes so that it was impossible for the animal
to see anything; but it is possible that some light leaked in
from the sides. When the Octopus had recovered from the
anaesthetic it did not respond in any way to movements of
objects in its vicinity. If it was placed in the water of a
large glass aquarium in a dark room no change in colour from
its original light gray coloration occurred when a bright
acetylene light was flashed upon it. The Octopus responded
to jarring of the aquarium by the production of a brown
mottling on the arms. It was found very difficult to keep
the animals permanently blindfolded in this way, and it was
felt that the results obtained were indefinite and unsatis-
factory, and so extirpation of the eyes was resorted to.

In the single extirpations a longitudinal incision, parallel
to the long axis of the eye, and about 1.5 cm. in length, was
made in the skin on the dorsal surface of the optic prominence.
Some connective and muscular tissue was then cut through,
and the chromatophores on the dorsal surface of the eye-ball
were thus laid bare. Slight pressure was then applied to the
ventral surface of the eye-ball so that it slipped through the
incision on to the dorsal surface of the head. The optic
muscles and the optic nerve fibres were cut through, just
proximal to the retina, and the eye was removed. Finally
the edges of the wound were carefully drawn together by
means of one or two stitches. The advantage of this pro-
cedure is that neither the optic ganglion nor the optic gland
are interfered with, indeed the former is not even seen, and
there is practically no loss of blood.

Recovery from this operation is very rapid; for in the
course of half an hour or so the animals appear quite normal,
except of course for the direct results caused by the loss of
vision on one side. Neither are there any bad after-effects,
as the animals can be kept practically indefinitely in this
half-blinded condition. The operation was performed on two
Octopi, one of which was killed in a few days and the other
kept for two weeks. At the end of this time the animal was
quite active and vigorous.

36 Cowpry: CoLOUR CHANGES OF OcropuUS VULGARIS

The behaviour of these semi-blind animals does not seem
to be modified in any particular, except that they do not react
to visual stimuli on the side of the extirpation. In one of
them the right eye was removed and in the other the left.
The elevation of the cirri is not affected in any way. The
animals swim freely, crawl about, pursue and capture crabs
and crayfish, and exhibit all the reactions of normal animals.

The coloration is little modified. All the uniform colour
patterns appeared as in the normal animal; but in the case
of the mottled patterns the mottling seemed to appear more
readily on the side of the uninjured eye. This was noticed
particularly after mechanical stimulation, which, if long
continued, however, always resulted in mottling of equal
brilliancy on both sides. The colour of the animals while
at rest is either light gray or uniformly mottled with brown.
It is, however, in the case of the striped colorations that the
greatest change is observed. The small brown stripes repre-
sented in Figures 2, 4, and 5, appeared only in connection
with the uninjured eye. The large longitudinal brown
stripes were not observed at all. Neither was the white
striping, illustrated in Fig. 8, seen, although one of the animals
showed it before the operation.

In the case of the double extirpations the operation Was
performed in two stages. When one eye had been removed,
in the manner indicated above; the animal was revived and
after an hour or two the second eye extirpated. No Octopi
were ever lost as a result of the operation. It was performed
on one male and one female only. Recovery was very rapid,
as shown by the fact that the next day when a crab was placed
so that the Octopus could feel it, it gathered itself together
made a sudden dash at it and captured it. Three fair-sized
crabs were taken and devoured in this manner in succession.
Both of the Octopi took crabs in this way.

The behaviour of these blinded animals was altered com-
paratively little. They were generally found resting quietly
on the bottom or attached to the sides of the tank, but they
sometimes crawled about or swam freely. They remained
alive and active, and fed whenever the opportunity offered
for about three weeks, when they were killed, so that we

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 37

must credit them with reacting normally to all except optic
stimuli.

The examination of the colour changes in animals from
which all the optic reflexes have been clearly and definitely
eliminated affords very valuable resulis. Immediately after
the operation, when the animals were stimulated by being
gently poked with a rod, the colour darkened from the light
uniform gray to the dark greenish brown represented in
Fig. 2. This colour soon faded to the original gray. A few
hours after the operation the arms of one of the two Octopi
took on a brown mottled coloration, intermediate in intensity
between those illustrated in Figures 5 and 6, which persisted
almost continuously for two days. The body of the animal
was slight!y mottled also. After this time neither of the two
animals exhibited any mottled or striped colorations of any
description. When resting they were of the gray shade repre-
sented in Fig. 1, and when irritated they changed to the
uniform dark red of Fig. 3.

The conclusion is therefore justified that reflexes through
the eyes are very important factors in the causation of the
striped and also to a less degree of the mottled patterns; but
the uniform colorations are not entirely dependent upon them,
although, as already mentioned, these uniform colorations
may be initiated by visual stimuli. Furthermore, we learn
from the single extirpations, that there is a fairly ready diffu-
sion of optic stimuli from one side to the other, so that the
coloration is little altered by the operation. The behaviour
and the colour of both the semi-blind and the totally blind
animals have been studied in different environments and will
be discussed in the next section.

The nerve supply to the chromatophores on the arms was
studied and for this purpose two kinds of operations were
performed. In the first kind, an incision about 4 cm. in length
(i.e. about one third of the circumference of the arm at this
level) was made in the skin on the dorsal surface of the second
arm of the right side, at right angles to its long axis and oppo-
site the twelfth sucker. Everything was cut through down
to the circular layer of muscle. The edges of the wound were
then drawn together by stitches. Immediately after the

38 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

Octopus had been revived, it turned a uniform red colour,
with the exception of a band of skin about 2 mm. wide on
either side of the wound, which remained quite white. Sub-
sequent observation showed that the arm distal to this point
changed colour in precisely the same way as the other arms.

In the other kind of operation the large nerve running
down the centre of the arm inside the various layers of muscle
was sectioned. This was performed in the second arm on
the right side and on the left by thrusting in a sharp thin-
bladed knife in the median line on the inner surface opposite
the twelfth sucker. The result was that there were no colour
changes in either of the two arms distal to the point of section,
other than those caused by the pulsation of a few isolated
patches of chromatophores. Whatever the coloration of the
rest of the body these arms retained a bright white appearance.
They hung limp and inert, responding to mechanical stimu-
lation only by local pulsation of the chromatophores in the
regions stimulated. The nature of this pulsation will be
described in a subsequent case, where the nerve distributed
to one side of the mantle was sectioned. The suckers distal
to the point of section did not attach themselves to anything
with which they were placed in contact, although those
proximal to the cut did. There was both sensory and motor
paralysis. Subsequent autopsy showed that the central nerve
and artery in each arm had been completely cut across.

Therefore all the colorations, except those caused merely
by the isolated pulsation of the chromatophores, result from
impulses passing from the central nervous system to the
periphery. In the case of the arms the impulses to the chro-
matophores must pass along the central nerves and then
branch out to the skin at right angles.

The last experiment of this sort cons sted in sectioning
the left pallial nerve, which constitutes the nerve supply to
the left half of the mantle. A transverse incision of about
3 cm. in length was made in the skin on the dorsal surface of
the head, about 2 cm. behind the optic prominence. The
muscular and connective tissues were cut through and separ-
ated until a large nerve running in a postero-lateral direction
was laid bare. This nerve was sectioned and the edges of
the wound were then drawn together by stitches.

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 39

As soon as the Octopus had recovered from the effect of
the anaesthetic, it was noticed that the left side of the mantle
was quite white and did not change in colour. When the
Octopus was poked with a rod the characteristic red coloration
appeared over the whole surface, with the exception of the
left side of the mantle, which was very sharply and clearly
limited at the dorsal and ventral median lines. A few hours
later brownish spots, from 0.5 to I cm. in diameter, appeared
scattered over this area. After the lapse of ten days, examin-
ation showed that this region changed colour in a very
peculiar manner. When the animal was at rest and of a light
gray shade, the patches pulsated and varied considerably in
intensity of coloration. They pulsated in some cases at the
rate of twelve times in thirty seconds, and two such spots
very close together often flashed out at quite different times.
If the Octopus was handled it grew darker in colour, and the
spots increased in number. On the left side of the mantle
the prick of a pin called forth no motor responses whatever;
but when the skin just to the right of the median line was
stimulated in this way violent movements of the arms re-
sulted. The skin on the left side when scratched with a
needle showed a reddish brown streak, which came and went
several times in a rhythmical manner. It would be present
for about twenty seconds, and then again absent for the same
period, and so on. This coloration was limited precisely to
the regions stimulated, so that simple patterns, such as
crosses and squares, could be reproduced in colour.

This experiment confirms the previous one and also
furnishes some additional data bearing on the question of the
pulsation of the chromatophores, which has already been dis-
cussed in Section IV. It is to be noted that the abolition of
the colour changes is always accompanied by both sensory
and motor paralysis.

The most important conclusions to be drawn from these
various operations may be summed up as follows.

I. Optic reflexes are very important factors in the pro-
duction of the striped and also, to a less degree, of the mottled
patterns; but the uniform colorations are not wholly depen-
dent upon them.

40 Cowpry: COLOUR CHANGES OF OcToPUS VULGARIS

2. All the colour changes are brought about by reflexes
which pass from the central nervous system along the various
nerve trunks to the chromatophores; but simple pulsation
of the chromatophores may take place independently of the
central nervous system.

VI. THE EFFECT OF CHANGES IN THE ENVIRONMENT

These experiments were performed with the hope of deter-
mining whether the Octopus changes its colour to correspond
to that of its environment. There was little evidence of such
a phenomenon when the Octopi were observed in their natural
haunts, as they were repeatedly. True, they sometimes are
very hard to distinguish upon certain backgrounds; but it
seems that just as often they presented a coloration which
in no way resembled their surroundings.

The methods used in the attempt to solve this problem
were frequently altered and improved, and so it was not until
the very last that anything but purely negative results were
obtained. At first the animals were placed in large glass
aquariums which were surrounded with differently coloured
translucent paper. After a time they always assumed a
uniform light gray shade. It was felt, however, that they
were under very abnormal conditions and this method was
therefore abandoned.

In order to experiment under conditions as nearly natural
as possible, the Octopus car described in Section II was built.
A long series of experiments was then carried out, in which the
car with the Octopi in it was placed in different environments,
the colour changes being carefully observed. These environ-
ments were as follows: (1) a smooth white sandy bottom;
(2) a smooth sandy bottom covered with Eel-grass (Zostera
marina) so that its general shade was dark greenish gray. In
addition to these (3) a glossy black oil cloth and (4) some old
gunny sacks were placed under the car giving a black and a
dark brown background respectively. The observer was
hidden from the animals and great care was always taken to
prevent any jarring or movement of the car or any outside
disturbance. Under each of these conditions the colour

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 41

changes were noted when the animals were at rest and when
irritated, when crawling about, and when swimming, when
feeding on crabs or crayfish, and when they noticed the advent
of an enemy, such as a large Hamlet (Epinephelis striatus).
The Octopi exhibited practically all the colorations described
in Section IV; but the results were still purely negative; for
no constant relation between the changes and the colour or
the brightness of the environment could be determined.
Then I read of Steinach’s idea (190Ia, p. 28), that the
changes in colour are controlled by the evenness or the con-
sistency of the bottom through the medium of the suckers
and determined to change my method. Heretofore the
animals always rested on the extremely irregular wire bottom
of the cage; but in the series of experiments next tried an arti-
ficial bottom was arranged inside the cage. The observations
were all made on both male and female animals put into the
cage separately, both*in direct sunlight and in shadow. The
bottoms were as follows:—
I. Fine white sand, represented in No. 1 of the colour chart,
Fig. 9 (soft white bottom).
2. Large roofing slates, painted white, No. 5 (hard white
bottom).
3. Finely powdered hard coal, No. 2 (soft black bottom).
4. Slates painted black, No. 4 (hard black bottom).
5. Slates painted yellow in imitation of the Brain corals
(Meandra labyrinthiformis and Meandra cerebrum),
which are common on the reefs, No. 9.
6. Slates painted the same shade of red as a sponge which
is also of very common occurrence, No. 7.
7. The large green Alga (Ulva latissima), No. 3.
Octopi were left on some of these bottoms for as long as three
days, and were observed almost continually during the day-
light of that period. The coloration exhibited was not
modified in any noticeable way by the environments. On
each day they showed a great diversity of colour patterns.
The brown bands across the eyes appeared when the animal
noticed the movement of anything in the water near it; the
uniform dark reddish brown colour, when irritated in any way

42 Cowpry: COLOUR CHANGES OF OcroPpUS VULGARIS

or following the ejection of ink: moreover, the colour always
bleached when the Octopus was touched with the hand or
with the end of a stick.

It was then noticed that the Octopi always seemed to turn
dark in colour when they were returned to the tanks of the
aquarium. The colour of the walls of these tanks is dark
green and is represented in No. 8 of the colour chart. Another
series of experiments was initiated with the object of determin-
ing whether a change to correspond to the brightness of the
environment takes place when the animal first enters new
surroundings. For this purpose half of the bottom of the
cage was covered with white slates and the other half with
black slates. Slates were also leaned up against the walls.
A small Octopus, of a light coloration, was then induced to
move from the light environment into the dark, and it was
observed that when it came on the black slates, five or ten
seconds after, it turned a uniform ,dark reddish brown
colour like that illustrated in Fig. 3. When, on the other
hand, it moved over from the dark to the light bottom, it
became either of a light gray coloration, Fig. 1, or of a light
gray shade with a faint brown mottling. The same results were
obtained on different days, in sunlight and in shadow, and
with a large number of animals, including both males and
females. Yellow, red, and brown No. 6, and white slates
mottled with irregular brown blotches, about 4 cm. in diameter,
were used; and when there was sufficient contrast between
the two environments bleaching or darkening always resulted.

The change seemed to be dependent upon the brightness
of the environment only, and it persisted for ten or fifteen
minutes. It may take place as soon as the animal crosses the
boundary, or it may be deferred until it comes to rest. The
change often occurs when the animal is in the act of swimming,
and consequently not in contact with the bottom. It appears
simultaneously over the whole surface of the Octopus; but
the coloration which results is not always uniform; for it may
be either a dark or a light mottling. Furthermore, the change
does not always take place. Ina sickly animal or in a fatigued
healthy one, there may be no adaptation at all. If a healthy
and vigorous Octopus, which has adapted itself to any of the

CowprRy: COLOUR CHANGES OF OcToPUS VULGARIS 43

bottoms, be poked with a stick or touched with the hand it
invariably takes on the light coloration illustrated in Fig. 4;
but when repeatedly irritated in any manner the uniform red
depicted in Fig. 3 always appears.

The same experiments were performed with semi-blind and
blind animals. In the case of the former the results were
particularly interesting. The change with the brightness of
the bottom took place in exactly the same way as with normal
animals, with the exception that the manner of the bleaching
was modified. It constantly took place two or three seconds
later on the side from which the eye had been removed. I[n-
deed, at a certain time the halves of the body are very defi-
nitely demarcated from each other by their difference in shade.
This difference in the rate of bleaching is evident, no matter
whether the original colour was of a mottled or of a uniform
character. It is a peculiar fact that no difference could be
noticed in the rate of darkening on the two sides, and I am
unable to formulate any explanation of this. Sudden bleach-
ing, particularly in this case, where it is more rapid on the
side possessing the most direct and potent optic reflexes, and
also the instantaneous production of the ghostly white color-
ation of Fig. 4, seem to indicate that the contraction of the
chromatophores may be an active process and may not be
wholly dependent upon the elasticity of their walls. In the
semi-blind animals also, this white coloration and the red
phase illustrated in Fig. 3 could be brought about by alarming
the animal and irritating it, if such terms may be used, at any
time in any of the environments, so that we must conclude
that factors such as these are most powerful. It is difficult to
overestimate the importance of the influence of the physio-
* logical condition of the Octopus upon the colour changes.

In order to make these experiments complete and con-
vincing two totally blind animals were used. They seemed
quite healthy and vigorous; for they crawled about the car,
occasionally swam freely, and devoured crabs with as great
avidity as did normal animals. As one would expect, there
was no change in colour of any description when either of
the two moved from a dark environment into a light one or
vice versa.

44 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

The conclusion is therefore justified that the Octopi do,
in a general way, change their shade to correspond to the
brightness of the environment. There is not as yet sufficient
evidence in support of the belief that these animals become
mottled or striped in imitation of the bottom on which they
may be resting or over which they may be swimming. As it
was impossible, on account of the difficulties of the technique,
to experiment with environments of which the colour had
been accurately measured and determined, and as it is not
even known whether the Octopus can distinguish colours,
it could not be ascertained whether the colour of the bottom
plays any part in these changes. It is extremely unlikely,
however, for the coloration of the habitat of the Octopus is
mostly characterized by greens, bright yellows, and even blues,
and we know that these are the very colours which the animal
cannot exhibit; and also for other reasons. The experiments
show, further, that this change in colour depends upon the
excitation of a reflex arc, which passes from the retina through
the optic nerve to the brain and thence by the different nerves
to the chromatophores. In addition to this, the fact that
the change may take place when the animal is in the act of
swimming, indicates that tactile reflexes, due, for example,
to the physical condition of the bottom, can have no influence
in the darkening or the bleaching as the case may be. This
reflex has been traced through the retina to the brain, and if
proof were required that it passed by the various nerves to
the chromatophores, this is supplied by the observation that
the parts of the animal from which the nerve supply has been
cut off do not change, like the rest of the animal, in brightness
to correspond to the environment.

VII. THE EFFECT OF CHANGES IN THE WATER, AND OF LIGHT

Many experiments were performed with a view of deter-
mining the effect, if any, of changes in the temperature, the
purity, and the salinity of the water upon the colour changes.
With small alterations, the results obtained were almost en-
tirely negative, and this is what is to be expected when we
remember that in the natural environments, on the reefs and

CowpRyY: COLOUR CHANGES OF OCTOPUS VULGARIS 45

elsewhere, such changes practically do not occur. As, under
natural conditions, the colour changes are so vivid and so
sudden, the conclusion is warranted that they are in no way
governed by changes of this sort. Since the object of this
work is to study the colour changes under normal conditions
and their relation to the behaviour of the animal, these experi-
ments will not be described.

It seems that the effect of light upon the colour changes
has been exaggerated. Steinach (1901a, p. 28) finds that
when the sun shines down through the water the animals
(Eledone) become uniformly darker in colour and flee into
the shade. I found that when a dozen or more Octopus
larvae, contained in a glass bowl, were moved from a shaded
place into bright sunlight, they all became active and darker
in colour. This was repeated several times; but in the experi-
ments with adults on change in environment no constant
difference could be observed in sunlight or in shadow, although
they were particularily looked for. Furthermore, when sun-
light was condensed by means of a lens upon normal animals,
dead animals, and totally blinded animals, resting in glass
aquaria and protected by Io cm. of water, for periods as long
as a minute, no darkening could be observed.

VIII. Discussion

The position is a peculiar one. At the head of the In-
vertebrate phyla we have a class of animals, the Cephalopods,
which have existed for millions of years, almost unchanged
in their general structural characteristics, and are very
highly specialized. We find, further, that the members of
the order Octopoda of this class are remarkable in that they
exhibit more vivid, complicated, and rapid colour changes
than do any other members of the animal kingdom. It is
not to be wondered at, therefore, that these animals, above all
others, should be chosen as objects by means of which to
investigate the general and also the more particular ques-
tions involved in change of colour. The following discussion
will deal with the possibility of the active nature of the diminu-
tion in size of the chromatophores; and with the causation, the
purpose, and the acquisition of the ability to change colour.

46 Cowpry: COLOUR CHANGES OF OcToPUS VULGARIS

There is certainly no histological evidence of the existence
of nerve fibres which might conduct impulses resulting in the
contraction of the chromatophores and the bleaching of the
animal; but careful observation of the colour changes in the
gross seems to point to some such mechanism. It is hard to
believe that the sudden bleaching which occurs when the animal
is poked with a stick is purely passive, and results, simply,
from the elasticity of the walls of the chromatophores. Simi-
larly, the abrupt appearance of the long white stripes, the
rest of the animal remaining dark, as illustrated in Fig. 8, is
hard to reconcile with this view. Moreover the experiments
with half-blind animals demonstrated that both the darkening
and the bleaching result from optic stimuli—that is that they
are both active processes. Of course the possibility remains
that the bleaching may be fundamentally passive and result
solely from the inhibition of the impulses passing from the
central nervous system to the chromatophores; but if this
is the case we should have to assume that in the normal light
gray coloration of the resting animal (Fig. 1) the chromato-
phores are continually receiving impulses from the brain and
that consequently the radial muscles are always in action,
for we meet with a still lighter coloration as illustrated in
Fig. 4. The difference in the shade of the body and of the
outer side of the arms in the resting condition (Fig. 1) and in
the frightened state, if we may use that phrase (Fig. 4), is
sometimes even greater than is indicated. Furthermore, an
animal which has been dead for two or three days seems also
to be darker in colour than the condition represented in
Fig. 4). The conclusion is therefore justified that, although
the anatomical findings seem to show conclusively that the
diminution in size of the chromatophores is purely passive
and results from the elasticity of their walls, still close obser-
vation indicates that this may not be the case.

Steinach’s conception of the causation of the ‘colour
changes has already been mentioned. He states that, in
Eledone, the changes in colour are due to the excitation of a
reflex which passes from the suckers along the nerves to the
brain and thence by the various nerve fibres to the chromato-
phores. When the suckers are attached to a hard firm bottom

CowpryY: COLOUR CHANGES OF OCTOPUS VULGARIS 47

the chromatophores are supposed to be stimulated and changes
in colour to take place. When, on the other hand, the Octo-
pus is resting on sand the suckers are not in action and no
stimulation of the chromatophores results: so that, as the
hard rocky bottoms are in general dark in colour and the
sandy bottoms light, there would be a change in brightness to
correspond to the environment.

Steinach cites two experiments in support of this hypo-
thesis. He worked with Eledone and, in the first instance,
cut off all the arms and the suckers remaining on the stumps.
The result of the operation was that, in the majority of cases,
the animal was highly coloured for about two days, in either
a mottled or a uniform manner. Subsequently, the animal
rested quietly on the bottom of the aquarium and maintained
a uniform silvery white shade. In the second experiment one
arm and its suckers were left intact, and the animal exhibited
the various colour patterns and the spontaneous colour
changes without modification. On the basis of such experi-
ments he concludes, in addition, that the origin of the colour
changes is not central but peripheral.

This hypothesis seemed at first sight to explain the pheno- —
mena observed in a very satisfactory manner, especially as
it did not postulate any very high degree of intelligence on
the part of the animal, and as it depended only upon the
functioning of a simple reflex. It does not seem to hold, how-
ever, in the Octopus, although one would certainly expect it
to do so; for the Octopus and Eledone belong to the same
family and resemble each other in many ways. If the color-
ation, in the absence of direct sunlight, is dependent upon a
mechanism of this sort, then one would not look for a modifi-
cation of the colour changes when the optic reflexes are
eliminated; but in both semi-blind and totally blind animals
the patterns are found to be definitely altered. The most
important evidence which can be brought against this hypo-
thesis is derived from observation of the living animals under
natural conditions. Thus, there are often several Octopi in
the same tank under identical conditions clinging to the wall;
but they all may be coloured differently. An Octopus is
often seen attached to the walls of the tank by a few of its

48 Cowpry: COLOUR CHANGES OF OcToPpUS VULGARIS

arms only, the others hanging limp in the water. In such
cases, the animal is usually of a light gray shade, but some of
its suckers are in action and others are not, and if Steinach’s
idea is correct, some at least of the chromatophores ought to
be dilated; but they are not. In the open, the Octopi are very
often.of a dark red coloration when on a soft sandy bottom,
and of a light shade when they are on hard, dark rock. When
the animal is swimming out of contact with everything except
water, the suckers being unquestionably quiescent, it is
frequently coloured in a variety of ways, and, when in this
condition, it may even change its colour.

It has been suggested that there might be a peripheral
reflex mechanism which would be capable of governing the
colour changes. This theory is rendered more attractive
after a study of Hofmann’s work (1907a), dealing with a
continuous nervous network, formed by the branching of
ganglion cells, running in the peripheral musculature of
Cephalopods; and also when we consider that Hertel (1907)
has shown that light rays of different colours have a specific
action upon the chromatophores. Such a theory, however,
is obviously insufficient to explain the colour changes which
have been recorded above in Octopus vulgaris.

It has been shown that in animals generally optic reflexes
are exceedingly important in-the colour changes and a close
parallelism may be traced between the degree of development
of the eye and the rapidity and the brilliancy of these changes.
In the Octopi and the Squids the eyes are very highly develop-
ed and efficient. They are even capable of accommodation,
and are much superior to those of many vertebrates. The
present experiments, in which one or both eyes were removed,
show that there is indeed such an interdependence.

The effect of the physiological condition of the Octopus
upon the colour changes has already been referred to, particu-
larly in Sections V and VI. Thus if an Octopus is irritated or
excited it will turn red, or if it is pursued by another animal
or poked with a stick its colour will bleach, no matter what
bottom it is on. Similarly in the case where an Octopus is
pursuing a crab it will often remain of a mottled coloration
until it has seized its prey when, in consequence perhaps of

Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS 49

the excitement, it will darken to the uniform dark reddish
brown coloration so often mentioned, quite irrespective of the
brightness of the background. When the animals are actively
engaged in feeding they often present the same coloration. It
has already been stated that this colour is associated with
muscular activity. When the vitality of the animal is lowered
for any cause, of all the colorations this darkening and to a
less extent the bleaching persist the longest. Thus a sickly
animal will not change its colour with the brightness of the
bottom, but it will darken when it is annoyed and bleach when
it is touched. To repeat, the importance of the physiological
condition of the animal is illustrated by (1) the fact that the
colour changes resulting therefrom (darkening and bleaching)
overpower and replace those caused simply by optic reflexes;
(2) the persistence of the darkening, and also to a less degree
of the bleaching, when the colour changes resulting from
optic reflexes, etc., have been eliminated by lowered vitality;
and (3) the uniform character of these changes, the animal as
a whole reacting, not simply a portion of it.

Throughout these experiments a record of each animal was
kept, and no difference was found in the coloration of the
males and of the females, so it is unlikely that sex is at all an
important factor in the colour changes.

When we consider the purpose of these phenomena we
enter into a discussion which is almost purely speculative.
It has been hinted that they are purposeless and of no use to
the animal; but I find that such a conclusion is hard to
accept; for, although there are many instances of organs
which are apparently useless, still such structures are seldom,
if ever, of so high a degree of functional activity as are the
chromatophores. For this reason we may conclude that the
chromatophore-system and the accompanying ability to
change colour are of service to the animal, and it becomes
necessary to determine, if possible, just what benefits the
Octopi derive therefrom. The change with the environment
seems to be of direct utility to the animal, but the effect is
marred by the over-ruling of the emotions, if such a term may
be applied to so lowly an animal. The ejection of ink is
undoubtedly protective in that it aids the animal to escape
from its foes.

50 Cowpry: COLOUR CHANGES OF OcToPUS VULGARIS

It is very doubtful whether the Octopi exhibit any warning
colours; but when an Octopus notices the approach or feels
the bite of a small fish, it will change colour immediately
and the fish will dart away. The colour assumed in such
cases is variable; it may be either uniform or mottled. In
two cases the colour displayed was very strange; for it con-
sisted of an irregular network of dark brown lines and blotches
upon a creamy white background. This coloration per-
sisted for only a few seconds. The Octopus does not move
in such a way as to suggest this motion as the cause of the
fish’s alarm, in fact in some cases it does not move at all.

The acquisition of the ability to change colours, as we
know it in the Octopus, is very difficult to understand, and it
seems that, in this respect, we shall remain in the dark un-
til very much more detailed and accurate work has been done.
A study of the ability to change colours and of the appearance
of the various colour patterns in ontogeny might lead to some
very interesting results. The question also arises as to what
would be the effect of rearing Octopi in coloured and colour-
less environments of varying degrees of brightness, and in
this connection the work of Gamble (1910), chiefly concerned
with the Crustacea, should be mentioned. What is needed
above all, however, is an investigation into the psychology of
the Octopus. It is evident that the order Octopoda provides
a field for this most facinating kind of research.

IX. SUMMARY OF CONCLUSIONS

1. All the colour changes are brought about by impulses
which pass from the central nervous system along the differ-
ent nerves to the chromatophores; although simple pulsation
of the chromatophores may take place independently of the
central nervous system (see p. 40).

2. Optic reflexes are very important factors in the pro-
duction of the striped, and also, though to a less degree, of
the mottled patterns, but the uniform colorations are not
wholly dependent upon them (see p. 39).

3. Octopus vulgaris does, in a general way, change its

CowDRY: COLOUR CHANGES OF OcroPpUS VULGARIS 51

colour to correspond to the brightness of the bottom, parti-
cularly when it enters a new environment (see p. 44).

4. This change in colour depends solely upon the excitation
of a reflex arc, which passes from the retina through the optic
nerve to the brain and thence by the various nerve trunks to
the chromatophores (see p. 44).

5. The physiological condition of the Octopus is the most
important factor in the colour changes (see p. 49).

BIBLIOGRAPHY
Bauer, Victor.

1909. Einfiihrung in die Physiologie der Cephalopoden,
mit besonderer Beriicksichtigung der im Mittel-
meer haufigen Formen. Mitth. d. Zool. Stat. zu
Neapel, Bd. 19, Heft 2, pp. 148-286.

Chun, Carl.

1902. Uber die Natur und die Entwicklung der Chromato-
phoren bei den Cephalopoden. Verh. Deutsch. Zool.
Ges., pp. 162-182.

Darwin, Charles.

1845. Voyage of a Naturalist round the World. London,

Murray.
Fredericq, Léon.

1878. Recherches sur la physiologie du Poulpe commun
(Octopus vulgaris). Arch. de Zool. exp. et gén., sér.
I, tom. 7, pp. 535-583.

Gamble, F. W.

1910. The Relation between Light and Pigment-formation
in Crenilabrus and Hippolyte. Quart. Jour. Micr.
Sci., London, New series, No. 219, pp. 541-583.

Gariaeff, W.

1909. Zur Histologie des centralen Nervensystems der

Cephalopoden. I. Subdsophagealganglionmasse
von Octopus vulgaris. Zeit. f. wiss. Zool., Bd. 92,
pp. 149-187.

Girod, Paul.

1883. Recherches sur la peau des céphalopodes. Arch.
de Zool. exp. et gén., sér. 2, tom. I, pp. 225-266.

52 Cowpry: COLOUR CHANGES OF OCTOPUS VULGARIS

Hertel, E.
1907. Einiges iiber die Bedeutung des Pigmentes fiir die
physiologische Wirkung der Lichtstrahlen. Zeit. f.
Allg. Physiol., Bd. 6, pp. 43-69.
Hofmann, F. B.
1907a.Gibt es in der Musculatur der Mollusken periphere,
kontinuierlich leitende Nervennetze bei Abwesen-
heit von Ganglionzellen? 1. Untersuchung an Cep-
halopoden. Arch. f. d. ges. Physiol., Bd. 118,
PP. 375-412.
Hofmann, F. B.
1907b.Uber einen peripheren Tonus der Cephalopoden-
Chromatophoren und iiber ihre Beeinflussung durch |
Gifte. Arch. f.d. ges. Physiol., Bd. 118, pp. 413-451
Hofmann, F. B.
1907c.Histologische Untersuchungen iiber die Innervation
der glatten und ihr verwandten Musculatur der
Wirbeltiere und Mollusken. Arch. f. mikr. Anat.
Bd. 70, pp. 361-413.
Hoyle, W. E.
1886. Report on the Cephalopoda collected by H.M.S.
Challenger during the years 1873-1876. Challenger
Reports, Vol. XVI, Part XLIV, pp. 1-245.
Isgrove, Annie.
1909. Eledone. L.M.B.C. Memoirs. (London: Williams
and Norgate.)
Klemensiewicz, R.
1878. Beitrage zur Kenntniss des Farbenwechsels der
Cephalopoden. Sitzungsber. Akad. Wiss. Wien,
Bd. 78, math.-naturw. Kl. Abth. 3, pp. 7-50.
Phisalix, C.
1892a.Recherches physiologiques sur les chromatophores
des céphalopodes. Arch. de Physiol. norm. et
path., sér. 5, tom. 4, pp. 209-224.
Phisalix, C.
1892b.Structure et développement des chromatophores
chez les céphalopodes. Arch. de Physiol. norm. et
path., sér. 5, tom. 4, pp. 445-456.

Cowpbry: COLOUR CHANGES OF OCTOPUS VULGARIS 53

Phisalix, 'C.

1894. Nouvelles recherches sur les chromatophores des
céphalopodes. Centres inhibitoires du mouvement
des taches pigmentaires. Arch. de Physiol. norm.
et path., sér. 5, tom. 6, pp. 92-100.

Polimanti, M. Osv.

1910. Les céphalopodes ont-ils une mémoire? Arch. de

Psych., ‘tom. 10, 10.375 pp. 84-87.
Pouchet, G.

1876. Des changements de coloration sous l’influence des
nerfs. Jour. Anat. et Physiol. norm. et path.
tom. I2, pp. I-90.

Rynberk, G. van.

1906. Uber den durch Chromatophoren bedingten Far-
benwechsel der Tiere (sog. chromatische Haut-
funktion). Ergeb. der Physiol., Bd. 5, pp. 347-571.

Sangiovanni, M.

1829. Des divers ordres de couleurs des globules cromo-
phores chez plusieurs Mollusques céphalopodes;
Description de quelques espéces nouvelles, et par-
ticuliérement de l’Argonaute. Ann.des Sci. Nat.,
sér. I, tom. 16, pp. 315-336.

Steinach, E.

I901a.Studien tiber die Hautfarbung und iiber den Far-
benwechsel der Cephalopoden. Arch. f.d._ ges.
Physiol., Bd. 87, pp. 1-36.

Steinach, E.

1901b.Ueber die locomotorische Function des Lichts bei
Cephalopoden. Arch. f.d. ges. Physiol., Bd. 87,
pp. 38-41.

Uexkiill, Jacob von.

1905. Leitfaden in das Studium der experimentellen Bi-
ologie der Wassertiere. (Wiesbaden: J. F. Berg-
mann.)

PLATE I.

UNiv.OF TORONTO STUDIES BIOLOGICAL SERIES No. 9.

ere Ym
a >.
7 é ‘
i 2 bf +
+
9
4 \ _
\
‘ a
\ 7
} a
!
” eel —
u ;
; :
5 al .
»
& \
a \
Jt = 2 \
é | \
r, 3 3
2 r q
yr a | 3 ;
; 4
q
4

B Meisel lth. Bosto

E.V.C. del

UNIv.OF TORONTO STUDIES BIOLOGICAL SERIES No. 9. PLATE2.

E.V.C. del B Meisel lith. Bost

UNIV.OF TORONTO STUDIES BIOLOGICAL SERIES No. 9. PLATE3,

E.V.C. det. B. Meisel lith. Boston

UNIV.OF TORONTO STUDIES BIOLOGICAL SERIES No. 9. PLATE 4.

1

is: ;
“se University of Toronto Studies
aX _ COMMITTEE OF MANAGEMENT |
Chairman: Ropert ALEXANDER FALCONER, M.A., Litt.D., LL.D., D.D. a
President of the University
f PROFESSOR W. J. ALEXANDER, PH.D.— |
: Proressor W. H. Ettis, M.A., M.B.
oe PROFESSOR A. KIRSCHMANN, PH.D.
PRoFEsSOR J. J. MACKENzIE, B.A. -
ProFEssor R. Ramsay Wricut, M.A., B.Sc., LL.D.
3 PROFESSOR GEORGE M. Wrone, M.A.
’ General Editor: H. H. Lancton, M.A.
Librarian of the University
ELECTRONIC VERSION
x AVAILABLE all !
NG. ; Ti “

oe

University of Toronto Studies
Biological Series

THE NORTH AMERICAN DRAGONFLIES

OF THE

GENUS AESHNA

4445
Reh "fh

THE NORTH AMERICAN DRAGONFLIES

OF THE

GENUS AESHNA

BY
E. M. WALKER, B.A., M.B.

LECTURER IN ZOOLOGY IN THE UNIVERSITY OF TORONTO

UNIVERSITY OF TORONTO LIBRARY
MCMXII

. i |
Ae | Tate hee
eutteth

‘ | t\) Dy a ie

: un "ae
2
1 ‘
¥ si
ie ey :
‘ q .
mt
i
}
a r '
‘ ¥
fl /
4G j
ii :
f b |
t " »)
iy Vek)
7 io
4 j
veya if { $
\ -
j ; sd we
Wis) 7 ,
ie ‘ ~
i »
}
» a
. ' i neae Mi ¥
Oe be", 1 Atlas Ax > " SNe ea
. vi)
aes 5
‘
4
he
Pa by
) ie
i" Te VEO h ays.
AAA A a 1s
se rhea ae
Ue ii ee Ne A
ule

aioli

: ay ane eae aD)

PREFATORY NOTE.

It gives me great pleasure to state as preface to the present
monograph of the North American Dragonflies of the Genus
Aeshna, that the cost of the admirable plates illustrating it,
which might have proved a difficulty in the way of its being
issued by the Committee on University Studies, has been
generously met by Sir Edmund Walker, Chairman of the
Board of Governors of the University.

R. RAMSAY WRIGHT.

‘i,

BIOLOGICAL SERIES No. 11.
ERRATA.

Page 72, line 2, for ‘‘septeatrionnlis” read ‘‘septentrionalis.”

Page 99, insert the following paragraphs after the one on Measure-
ments :

Material determined—8 3 12 2. Nova Scotia: Pictou, Sept. 2, 1889
(Sheraton, Acad. N.S. Phil., 1 9). QUEBEC: Anticosti Island, 1902 (Dr.
Joseph Schmidt, Coll. Div. Ent., Ottawa, 1 9). ONTARIO: DeGrassi Point,
Lake Simcoe, Sept. 2, 1906 (Walker, 1 2); Temagami Forest Reserve, near
Lake Obabika, Sept. 11, 1908 (Walker, 16); Nipigon, Aug. 28, 30, 1907,
Aug. 6, 1910 (Walker, 4c’ 6 9). MuicHIGAN: Isle Royale, Aug. 8-16, 1905
(B. F. Savery and C. C. Adams, Coll. Univ. Mich. and Williamson, 3o7 2 9),
MANITOBA: Winnipeg, Sept. 9, 1910 (J. B. Wallis, 1 9).

Nymphs—Nipigon, Oni., Aug. 5-8, 1g10, 1 @ (St. F); 1 9 exuvia.

Page 205, omit fifth reference: 1909. The Insects of New Jersey—
Odonata.

TABLE OF CONTENTS

PAGE
NRO) DUCT LONG ic, fs oko ee ee ee CI? | OLR Bre ey oo 1
PURPOSE AND SCOPE: OF) PAPER 0 io. emer a cle Le eae ae 1
SOURCES OF MATERIAL AND ACKNOWLEDGMENTS................--: 1
RIEED OBSERVATIONS? = ok eat ee ole eae Laos Ae elt Sane en Mens 3
PRE NUA NTE CAS EE NAG re ec) a a teens) A hetomerelli ene Y RP SAWN ee et LS 8 ae mete 3
PREECE NIWS HATES FIN AC ee hte bie A ee a ca atlas | lk oT eR A 4
BIGADRO NOMEN heath a Webek Ase tes Us uate hr AORN Nes oo teat A hein i ea 4
Generic Characters olathe Adulte ec ees he Maa eye 5

Generic charaeters of the Nymph 233 0 )ic 5 «de esas rea: 6
SecondanyeSexuali Characters) \3cte ve mesiecteey st aectere tame nmctey: oY be 6
Characters of Specific Value and Terminology.................. 7

ARAN Ra etry We he A RAP, UMRIENONS 12 SB yet) ab tty a lay ai
Culouremabceriniy ss: ook Lu ls Maes aber ai pe Wires alas 8

Genitallianik) oh ee cto e tat ial Mimam Nae unlit AN Riv og idea ae ete 10

(a) Accessory Genitalia of Male...................... 10

(O)rGenttalracof Memales 8). orn eer a ele an eae 12

IVAN ere at eet aro ole yy et ie 2 Witethaehe sasteaties <a)s 13
Measurements and Abbreviations «0. sci. ota wipe tla 14

Genetic Relationships of the Genus Aeshna.................... 15

Mime Sibtaniile, eshte.) lates ek! oo Ale Ls a 15

SIE SENT MGR IMUTR ts 8. eye ais RLS woe Cal sa eats nts ah a+ 21

RRB REATION eset tsi fie ot nih am aoe ie tata ah aed i A CS oe a lead 25
ClimaticnVanlationsesser ester aren oN ie. weve wer See ae eat 25

Colour Vatiationvob Hemalese jn.) eo eae see aia. wr becio eons 29
Genperniled leAcCeSe ies ie ee hes tenn eee can S 29
(GENERA IOTREMELISTOR Mi) cb het, in br Ont ORE ete GERI SLUR oC olsre Ui alane tie 30

SE AMIAL MRCS AL MAME: tals. 50,a's 2 ter EMTs we ela see 2 30
Pengthvawlaiaountal fhe ars sale rata aac ee ele ete Oe ee sateen 31

Rca oatiek tS ee aoe, eth ce cen ene oinge MERE TREE LY 58 AG Re, 31
infuenceol Weather Conditions! se www eel: erick coe) okeveae see 33
Migrations ........ Pe oreecty svat dpe ot I Sn ol Cl at oes 35
Seasonal Variation in Number of Individuals.................. 35

TCHR ee Sale ae satel china sik) 1 ol he I MR DERE Ect 8 oe 35
DT STa oh Ste Sch Aas ae | oat Aine PARSE aD, AN aslo 10 gf SU pa ee UAE 36

IVEREITI OME A UES fence mere re cone oiies SReR IR CURT si) AE es ouchonaiega eta 38
MBcoraitek ta TE SMEAR 09.012, 5c) SR et UNA ode cap le taneae state 39

Comparison of Copulatory Position with that of Other Odonata.. 41

Vili TABLE OF CONTENTS
PAGE
OVIPOsition «... sis nd ee Ce, Pe re 42
The Egg « «os sv dew ybaisa tice coe eat et eee eee 46
Probable Number of ‘Nymphal'Stages..>.2 2-4 <4. seen 46
Changes at the ‘Various Bedyres:.”. 32.5.5 es oe ee ue ae 47
Length of Nymphal Eife: )2 40). Jaden as dee eee ee 49
Habitat.of Nymph? 255-4027 9 sb 2 Coed Aes pees 50
Habits-of (Nymiphis . eealane6aeecicine een ee coe fia nee ee 51
Symbiosis of Nymph and Green Alga......................... 53
Emergence of the ‘Iniago 2165, :).2:2).5.. wath ate cia! eee ee 54
THE NORTH AMERICAN SPECIES OF AESHNA...-..-.--::..-... 56
GENERAL ‘CHARACTERISTICS: 3 i tsd 02 oot nein ee ee 56
KEY TO THE NORTH AMERICAN SPECIES OF AESHNA—ADULTS
T= Males. ootsc ae reise ences Ae AO EE
I]-——-Females)6 6 yes oe SE ce. Sade ee oe 62
KEY TO THE NYMPHS OF NORTH AMERICAN SPECIES OF AESHNA:--.------ 66
THE CAERULEA GROUP. | 2005. BPR A eee eee 69
Tae. JUNCEA GROUP «6353.56 22:2 LE SIs Se eee 69
THE CLEPSYDRA GROUP 652) e522 ested ae ee Te ee 69
‘THE CYANEA GROUP 202125 1200.9, Sots ites au iy 70
‘THE (CALIFORNICA GROUP 42). i255 Pogo oe oes Oe vel
‘PHe MULTICOLOR: GROUP: 1-5 25455452005 Ce eee tae
DESCRIPTIONS ‘OF SPECIES 21s <%2. (5 oie eof = sues aise ences ee oe ee ientetre he ees eee 42
Aeshna caerulea septentrionalis: ).:3) 50. 40 .\0ae 32 AB see 72
Aeshna-sitchensis.: 20 0533s SRE ae eee. oe 77
Weshna’quncea: .'. 2!) dua 6 os. eecaeess aie fe es 2 oe 83
so ©-Subarctica = s.c0<c)s 1 POS oa de ee ee 93
7 Aunitermapta | io ee ie iN gI aR etree oe ee 100
ot a interrupta..\../2 TOR So She eae ee ae 103
y a NEVACENSIS . "sv x-.3 Ret see 111
Se re lineata. 3... Soeees By. Sah se ee eee 112
* He intériay os. Oho Oot ee Oe a ee 116
"Fe Pr@remnibass fos cys clnsleyeleusee toe ele ccc cae eed tee Se 119
“.\ Velepsydras 2h ccu So eaeiG ng xv ae Ei 129
‘*’ \CADAGENSIS. «555! LaePeless este Rodi ce oat Oe Sie aie a 135
$f * SVETEICALIS. . 05600 «a peters hoe clea RO ee ee 145
"7 ‘tuberculifera: . 360.0%. oc 22 Se See ee eee 152
fe: Mpallmdata pips s' ee 220.02, sacs Shigeo eee ee 157
“Wmiprosal sie is oa ee eee 165
. Mi umbrosa:2\. iss; Ba a8 Ach ae eee 165
* ns occidentalis. 2.9040. oe a ee 174
#F) CORSEFICEA. cists 0 tc Uiasde ots HOR ORIS Oe oan Sea Re nena een 176
$8) Wl CATMOrIICA sf. ccc are cere ees Se alc oe 184
OF CNT ET COON 060. 35 vy ose Se RSE ach Taare. Sea een 190
8° WSTUGAEAL es 4)s once ero ened he See ee 198
LITERATURE SCERED 6.0 os icc eese roe ene ie ere stn a bviceeie an ee 203

THE NORTH AMERICAN DRAGONFLIES OF THE
GENUS AESHNA.

INTRODUCTION.

Purpose and scope of paper—During the summer of
1906 while spending the latter part of the season at De
Grassi Point, Lake Simcoe, Ontario, I became interested in
the large dragonflies of the genus Aeshna Fab. by reason of
their great abundance that year. Rough observations in
the field seemed to point to the existence of more species
than were recognized in any of the writings on the Odonata
of eastern North America and the published descriptions of
our species of this genus impressed me in nearly every case
as being more or less vague or superficial. Correspondence
with the eminent Odonatist, Mr. E. B. Williamson, elicited
the fact that his observations on Aeshna in Indiana and
northern Ontario corresponded closely with my own, and
it seemed evident that the genus was in need of thorough re-
vision. Accordingly, on Mr. Williamson’s suggestion, I
undertook to make a careful study of this neglected group
with. the aim of placing our systematic knowledge of the
North American species upon a sound basis and obtaining
as much information as possible relating to their life histories,
ecology, seasonal and geographical distribution. Only the
species which have been found north of Mexico are considered
here, as the Mexican and Central American forms have al-
ready been ably dealt with by Calvert (Od. B.C.A., 1905).

Sources of material and acknowledgements—During the
four years in which I have been engaged in this study I have
examined most of the larger collections in North America, as
well as a considerable number of specimens received from
collectors not specially interested in the Odonata, and I
take pleasure in expressing publicly my deep gratitude to
all those who have thus assisted me in what would otherwise
have been a very incomplete piece of work. Especial thanks
are due to Mr. E. B. Williamson and Professor P. P. Calvert,

2 WALKER: NORTH AMERICAN SPECIES OF AESHNA

who have been ever ready with helpful suggestions and en-
couragement and most generous in the loan of material.
The entire collection of North American Aeshnae belonging
to the former, probably the largest in existence, has been in
my hands since the work was commenced. Other large col-
lections which were in my possession for a considerable period
of time are those of the United States National Museum, for
which I have to thank Dr. L. O. Howard and Mr. R. P. Currie,
and that of Miss Mattie Wadsworth, to whom I am also
much indebted. The total number of specimens examined
was about 1720.
The following is a list of the sources of the material
studied, in addition to that in my own collection, together
with the names of those to whose kindness I owe the privi-
lege of examining it:
The United States National Museum, Washington,
D.C... (Dr. LO. Howard:and Mr..R:/P. Comes
The Museum of Comparative Zoology, Cambridge,
Mass. (Mr. Samuel Henshaw.)

The British Museum, London. (Mr. G. M. Waldo.)

The Academy of Natural Sciences of Philadelphia.
(Prof, .P.. P;..Calwert:)

The Boston Society of Natural History. (Mr. C.
W. Johnston.)

The Public Museum of Milwaukee. (Mr. R. A.
Muttkowski.)

The University of Michigan. (Dr. C. C. Adams.)

The Provincial Museum of Ontario. (The late Dr.
Wm. Brodie.)

Also the private collections of Mr. E. B. Williamson,
Bluffton, Indiana; Prof. P. P. Calvert, Philadelphia; and Miss
Mattie Wadsworth, Hallowell, Me.; and numerous speci-
mens from Messrs. C. H. Kennedy, Sunnyside, Wash. (to
whom I am also indebted for valuable field-notes, and colour
sketches); J. B. Wallis, Winnipeg; Norman Criddle, Aweme,
Man.; T. N. Willing, Regina; A. G. Huntsman, Toronto;
R. C. Osburn, New York; Prof. J. G. Needham, Ithaca,
N. Y.; E. V. Cowdry, Waterford, Ont.; the late Dr. James Flet-
cher of Ottawa; Mr. K. J. Morton, Edinburgh, from whom I

WALKER: NORTH AMERICAN SPECIES OF AESHNA 8

received a fine series of British specimens of Aeshna juncea;
and Dr. F. Ris, Rheinau, Switzerland, who furnished me
with nymphs and exuviae of the same species.

Field observations—In addition to the knowledge gained
from the study of this material I have been able to observe
in the field most of the eastern species herein described and
in such cases descriptive notes on the colour-pattern and,
wherever possible, coloured drawings have been made from
the fresh material. Of nearly all the species that I have not
seen in life I have examined recently captured specimens
preserved in strong alcohol, in which the colours are retained
practically unaltered for a considerable length of time.

In regard to the earlier stages I have not been very suc-
cessful in my efforts to obtain and rear the mature nymphs,
and it was only by persistent effort and the help of others that
aconsiderable series of nymphs and exuviae representing eleven
species has been accumulated. Although it is impossible to
prove that those nymphs which have not been reared belong
to the species to which they have been referred, I am confident
that there is very little possibility of error in any of my deter-
minations.

Much kind assistance has also been rendered to me in
the work of collecting the imagines by Mr. Paul Hahn and
in collecting and rearing the nymphs by Mr. A. R. Cooper.

THE NAME AESHNA.

The name Aeshna, formerly coextensive with the modern
family Aeshnidae, was given to these insects by Fabricius
in 1775 (Syst. Ent., p. 424), but was afterwards changed to
Aeschna by the editors of “Illiger’s Magazin fiir Insekten-
kunde”’ (Bd.1,S. 126, 1822). In this form it was universally
quoted until Calvert, in 1905, restored the original spelling, in
which he has been generally followed by American but not
by European writers. Various attempts have been made to
interpret this word. The first suggestion is found in ‘‘Illiger’s
Magazin”’ (Joc. cit.) as follows:

““Aeschna, ae (nicht Aeshna) f. Schmaljungfer viel-
leicht Aeschyna von aicyvvn, Schamhaftigkeit, nach der

4 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Aehnlichkeit mit Jungfer? In Charleton, Exercitatt. de
differentiis et nominib. animal., Oxon. 1677, kommt acyve
als Name eines Insekts vor.”’

Williamson (Drag. Ind., p. 303, 1899) regarded the word
as probably derived from aicypos, “‘ugly,’’ but recently the
same writer has communicated to me a suggestion made
by Mr. R. J. Tillyard that the original spelling was a
printer’s error for Aechma (aixuy, ‘‘a spear,” in allusion to
the long slender form of the body). As, however, it
is‘ impossible to decide the question of the meaning of
this word with any approach to certainty it is unfortu-
nately necessary to fall back upon the original spelling,
for although ‘‘Aeshna’’ is impossible as a Greek word,
‘‘Aeschna,’’ in spite of its better appearance, is also meaning-
less and impossible to derive from any Greek word without
making allowance for errors. This is the more unfortunate
as the composites of Aeshna must all retain the emended
form in which they originally appeared (Basiaeschna, etc.).

THE GENUS AESHNA.

Taxonomy.

The genus Aeshna is here considered in its nar-
rowest sense and the species separated from it by
Williamson (’03) under the name Coryphaeschna have not
been included, although this group of forms has not been
generally recognized as having generic rank. It is true that
the characters, taken singly, which separate Coryphaeschna
from Aeshna, sensu stricto, are seemingly slight and that some
species, now referred to the latter genus, are in some respects
intermediate between these two types; nevertheless the points
of difference between such forms as Coryphaeschna ingens and
Aeshna juncea are numerous, and, taken together, fully de-
serving of generic rank, according to the usual conception
of a genus as held by modern entomologists. A number
of apparently slight differences, difficult of definition, may be
of greater phylogenetic significance than a few easily definable
ones. Thus any typical species of Aeshna may be readily

WALKER: NorTH AMERICAN SPECIES OF AESHNA 5

separated from Oplonaeschna by the unforked radial sector,
while the distinctions between the former and Coryphaeschna
are much less easily defined (vide Williamson, loc. cit.). Yet
apart from this one character, which is really a very slight one,
Oplonaeschna is a typical Aeshna, while Coryphaeschna
differs from the latter in many points, small individually
but together giving the group a distinct habitus of its own.

It must be admitted, however, that the separation of
Coryphaeschna as a genus may necessitate further subdivision
of the genus Aeshna asdefined at present, but it is not improb-
able that this would be an advantage, as the numerous
species which the latter now embraces are not all obviously
nearly related forms and possibly represent several distinct
lines of descent.

The genus Aeshna as understood here may be character-
ized as follows :

Generic characters of the adult—Eyes contiguous for a
distance not greater than the dorsal length of the frons
and frontal vesicle combined; frons not unusually produced
in advance of the eyes, the face more or less convex in profile
view; wings moderately broad, the base of the hind wing and
the anal loop well developed; males with an anal triangle and
the postero-anal margin of the hind wing angulate; membran-
ule of hind wing extending along the anal margin for a
distance equal to or greater than the distance along which
it extends on the postcosta; upper piece of arculus equal to
or longer than the lower piece; median space free; subcosta
not prolonged beyond the nodus, pterostigma distinctly
braced ;! triangles typically with two basal cells, inner side
of triangle of hind wing at least half as long as the outer
side; radial sector forking more or less unsymmetrically a
short distance before the pterostigma or at the level of its
proximal end, the posterior branch generally continuing the
direction of the main stem more nearly than the anterior
branch, two to four rows of cells (generally three in North
American species) between the forks at the level of the distal
end of the pterostigma, increasing at the margin; radial sup-

1 Those of one pair of wings are sometimes without the brace in Ae. caerulea
and sitchensis.

6 WALKER: NORTH AMERICAN SPECIES OF AESHNA

plement more or less sinuate, diverging from the radial sector
and converging again distally, the intervening space with at
least three rows of cells in its widest part ; Mia arising from
just before to just behind the stigma; M4 unbroken though
more or less sinuate distally; abdominal segments without
accessory lateral carinae;! segment 2 in the male with a
pair of auricles, segment 3 longer than segment 4, more or
less constricted before the middle, especially in the males;
ventral surface of segment Io in the female spinulose or den-
ticulate towards the posterior margin.

Generic characters of the nymph—Eyes prominent, the
antero-posterior diameter not longer, generally distinctly
shorter,than the posterior margin; hind angles of head rounded
or rarely bluntly angulate; posterior margin of head straight
or slightly excavate. Mentum of labium reaching back
to the bases of the middle legs, the lateral lobes, generally,
but not always, squarely truncate, their hind margins minute-
ly and feebly crenulate, the terminal hook minute or occasion-
ally absent. Wing-cases in full-grown nymphs usually
reaching back to the base or apex of segment 4, sometimes
over a part of segment 5, showing the curve of Rs toward the
anterior margin and its unsymmetrical apical fork, and the
radial and median supplements curving toward the posterior
margin; abdomen broadest at segment 6 or 7, without dorsal
hooks, lateral spines typically present on segments 6 to 9,
but sometimes those on segment 6 are almost obsolete, while
in others they are distinctly developed on segments § to 9.
Superior appendage a little shorter than the inferior appen-
dages, deeply excavated at apex.

Secondary sexual characters—Beside the differences in
the genitalia, abdominal appendages, and angle of the hind
wing and length of the pterostigma, the sexes in Aeshna, as
in most Odonata, differ considerably in the general form of the
body and in the relative proportions of its parts, and to a
greater or less extent incolour-pattern. The thorax is slightly
smaller in the female owing probably to the lesser devel-
opment of the muscles of flight than in the male, which is the

‘Rudimentary accessory lateral carinae are sometimes present, e.g., in
Ae. californica.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 7

more active sex. The abdomen is a little shorter, being gen-
erally more nearly equal in length to the hind wings, but
is considerably deeper, though nearly the same in breadth.
The greater depth is most marked at segment 3, which is
considerably shorter and less constricted than inthe male, and
at segments 8 and 9. The ground colour is generally paler in
the female, the abdomen being brown, darkened at the
sutures and elevated parts and around the pale areas. Of
these the dorsal thoracic bands and posterior abdominal
spots (PD) are smaller, the lateral abdominal spots larger,
than in the male. The lateral thoracic bands tend to be
slightly broader in the female. The colour of the pale areas
may be the same (homceochromatic females) or more or
less widely different (heterochromatic females) from that of
the male (vide p. 29).

Characters of specific value and _ terminology—The
abdominal appendages are generally regarded as offering
the most important specific characters in the adults, but the
accessory genitalia are at least as valuable and perhaps more
so as group characters. Other important specific characters
are found in the female genitalia, the dark markings of the
face, the thoracic colour-pattern and to a less extent that of
the abdomen. In particular cases other characters are
useful, such as the relative lengths of the occiput and line of
contact of the eyes, the colour of the membranula and certain
venational features, especially the form and position of the
fork of the radial sector and the origin of Mita.

In the nymphs the best systematic characters are found
in the form of the labium, especially of its lateral lobes, the
supracoxal processes of the prothorax, the length of the
lateral spines of the abdomen, the size of the female geni-
talia, the relative lengths of the abdominal appendages,
and the colour-pattern.

In the descriptions of both imagoes and nymphs [| have
found it convenient to introduce a few new terms.

A. Imago—The preocular band is the black ocular
margin of the frons and nasus, which is broadest above
where it crosses the head just in front of the frontal vesicle,
narrowing as it passes ventrad along the sides of the frons

8 WALKER: NORTH AMERICAN SPECIES OF AESHNA

and usually the nasus. It is confluent above with the ‘“‘T-
spot’’ in all North American species. The lateral lobes of
the nasus are the two rounded or subangulate ventro-lateral
processes of that structure. The remaining new terms all
refer to the colour-pattern of the abdomen and the genitalia.

1. Colour-pattern—The colour-pattern in the genus
Aeshna is on the whole very uniform, remarkably so in the
North American species, and in the ordinary cabinet speci-
mens the appearance of uniformity is exaggerated by the
disappearance of the natural colours and the frequent ob-
scurity of the markings. Hence Odonatists generally have
relied almost exclusively upon structural features in character-
izing the species of this genus, particularly upon the abdomin-
al appendages of the male, with the result that the species
of Aeshna have been ‘‘lumped”’ to an unusual degree and the
belief has become prevalent that the coloration, though re-
latively uniform for the genus, is very variable within the
limits of a given species.

The study of these insects in large series demonstrates
conclusively that this is not the case and that each species
has its characteristic colour-pattern, which usually varies but
very slightly, if exception be made of the wide variation in
colour alone of the females of most species, in which a ten-
dency to dimorphism may be observed (vide p. 29). I have
therefore described at considerable length the colour-patterns
of all the species and subspecies, and, as the same type of
pattern is present in all the forms, it has been found conve-
nient to give special names to the more characteristic spots
and bands of the thorax and abdominal segments.

(a). Thorax—The dorsal thoracic bands are the pale
longitudinal stripes on the mesepisternum, frequently de-
scribed as ‘‘antehumeral stripes.’’ The term /ateral thoracic
band is used in describing two oblique lateral bands or pale
areas on each side of the thorax, even when these areas cannot
be well described as ‘‘bands.’’ When other pale areas occur
on the side of the thorax they are not included in the term.

(6). Abdomen—Except the first abdominal segment, on
which there is nearly always a dorsal and generally a lateral
spot, the serial homologies of the pale areas of the various

‘WALKER: NORTH AMERICAN SPECIES OF AESHNA 9

segments in any species of Aeshna are quite obvious; and as
in all the species the pale areas of corresponding segments
are homologous the same terms are applicable to these areas
in every case.

The most typical arrangement of these spots is generally
seen on segments 4 to 7 and may be understood by reference
to text figure 1, which gives dorsal and lateral views of
segments 1 to 4 in the male of Aeshna juncea. It will be
observed that the spots can be divided into a dorsal and a
lateral series, and each of the series into anterior, middle
‘and posterior spots. These will be referred to hereafter by
the abbreviations used in text figure 1 (AD, MD, PD, AL,

Fic. 1—Aeshna juncea L, abdominal segments 1-4, showing the characteristic spots ; 1 lateral view;
2 dorsal view. D dorsal spot (of segment 1); L lateral spot (of segment 1); AL antero-
~ lateral spot ; ML medio-lateral spot; PL postero-lateral spot; AD antero-dorsal spot ; MD
medio-dorsal spot ; PD postero-dorsal spot. AML (on segment 2) is a combination of AL
an ‘

ML, PL). AD is always small and often absent in North
American species, and except for the sake of uniformity in
the terminology need not have received a special term. MD

10 WALKER: NortH AMERICAN SPECIES OF AESHNA

and PD are paired, the former generally small and yellowish
or greenish and represented on segments 2 to 7 or 8, the latter
large and typically blue, and nearly always present on all
the segments. PD or PD+PL is the ‘‘apical spot” of
Hagen. Of the lateral spots, all of which are typically blue,
AL and ML are partly or wholly separated by the trans-
verse carina and together form the ‘‘lateral divided basal
spot’’ of Hagen. ‘They are generally represented on seg-
ments 2 to 8, AL gradually decreasing in size caudad to a mere
dot on 8, while ML increases to 6 or 7, becoming smaller
again on 8. On segment 2, AL and ML are almost always
united, forming a single spot AML. PL, which is generally
small and frequently a mere ventral offshoot of PD, is the
most variable in development of all the spots. It always de-
creases in size caudad and is often represented on only a
few of the anterior segments. In females, PL and ML are
usually confluent on segment 2, and occasionally, in both
sexes, also on segment 3.

2. The genitalia. (a). The accessory genitalia of the
male—These are lodged in a deep median groove or pocket,
the genital fossa, on the ventral side of segment 2, bounded
laterally by the tergal margins, which are more or less ap-
proximated behind but divergent cephalad in the region
occupied by the anterior hamuli and anterior lamina. The
vesicle of the penis is a sac with chitinous walls lying at the
posterior end of the genital fossa and attached by its broad
posterior end to» the front of the sternum of segment 3.
Arising from the dorsal surface of the vesicle, at its distal end,
is the penis, a three-jointed chitinous tube whose lumen is
continuous with that of the vesicle. The first joint is the
longest and is bent upon itself, the proximal limb being di-
rected dorsad, the distal limb ventrad. The second joint
is directed ventrad and caudad and has a longitudinal slit-
like aperture on its convex surface, communicating with the
lumen of the penis. The third joint consists of a small
strongly chitinized piece, and two large fleshy distal lateral
lobes, which are complexly folded and lie in a marked de-
pression in the wall of the vesicle at its anterior end. When
not in use the bent penis is protected by the sheath of the pens.

WALKER: NorRTH AMERICAN SPECIES OF AESHNA II

a concave chitinous process arising from the floor of the geni-
tal fossa; and on each side of the sheath, generally concealed
by the overlapping tergal margins, is one of the posterior
hamuli, two simple chitinous processes’ of the walls of the
genital fossa bearing upon their apices a number of long hairs.

In front of the posterior hamuli, and in some species
(none North American) partly concealed by the hairs of
the latter, are the much larger and more complicated anterior
hamuli. Each anterior hamulus consists of a broad, some-
what triangular thickening of the wall of the genital fossa
and a folded, strongly chitinized mesially concave process,
arising from its sloping mesial surface. The two concave
processes of opposite hamuli form a pair of claspers for the
ovipositor during copulation and their efficiency as such is
increased by the presence, in each concavity, of a small ele-
vation thickly covered with spinules, (fig. 2, ¢). For
descriptive purposes each clasper may be divided into two
parts, the hamular process and the hamular fold. The ham-
ular process (fig. 2, hp) is a freely projecting structure, the
most anteriorly and ventrally situated part of the clasper.
The two processes together form the floor of the passage for
the ovipositor. The hamular folds (fig. 2, hf), so called
from their folded form, are the more posterior and dorsal
part of the clasper and close the passage above and behind.
Generally the hamular process and fold are not sharply
marked off from one another, the free margin of the one pass-
ing insensibly into that of the other; but in some cases the

Fic. 2—Anterior lamina and anterior hamuli of (A) Aeshna interrupta Walk., (B) Ae juncea L: al
anterior lamina ; ¢m, tergal margin of segment 2; sf spine of anterior lamina ; 4A hamula
process ; ¢ spinulose tubercle ; 4 basal part of hamulus ; 4 f hamular fold.

12 WALKER: NORTH AMERICAN SPECIES OF AESHNA

hamular processes assume highly specialized- forms and be-
come partly or entirely separate from the folds, which are
in such cases more or less reduced (fig. 2, B). Both of these
structures differ greatly in form in different species and are
of great systematic importance.

Immediately in front of the anterior hamuli is the an-
terior lamina (al), a large plate, deeply but broadly cleft in
the middle line, the cleft forming the anterior termination
of the genital fossa. The anterior lamina bears on each
side at or near the margin a more or less acute process, the
spine of the anterior lamina (sp), which is directed more or
less caudad and either ventrad or dorsad according to the
species. In form and size it exhibits great variation within
the limits of the genus, but offers excellent group and some-
times specific characters.

(b). The genitalia of the female—These are situated on
the ventral side of segment 9. Under this term are included
the ovipositor, genital valves, styli and lateral genital plates

Pic. 3—Aeshna clepsydra Say, femrle genitalia and appendages, ventral view : 6p basal plate ;
/p lateral plate ; ov ovipositor ; gv genital valve ; sf stylus.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 13

(fig. 3). The ovipositor (ov) is attached to the apex of segment
8 and consists of a pair of somewhat falciform processes,
concave internally and fitting together to form a laterally
compressed tube. Each process consists of an upper and a
lower piece fitting closely together along their adjacent
margins and sharply pointed at their distal extremities. The
outer surface of the upper piece, at the apex, is marked by a
series of fine transverse ridges like a file, while the lower
piece is quite smooth. At the base of the lower piece is a
short broad transverse plate,ithe basal plate of the ovipositor(bp).
When not in use the ovipositor is largely concealed by the
genital valves (gv), apair of broad thin chitinous plates, attached
to. the whole length of segment 9 under cover of the tergal
margins. At the base, the genital valves are widely separ-
ated, leaving part of the ovipositor exposed, but beyond
a point somewhat before the middle they are attingent and
form a complete protective sheath for the apical part of the
ovipositor. Each valve generally exhibits distinct lateral
and ventral surfaces separated by a lateral carina which
terminates at the apex of the valve. Dorsal to the apices
the valves are produced caudad and dorsad into a beak-like
process, which fits into a depression on the sternum of seg-
ment Io and serves as a sheath for the apex of the ovipositor.
Articulating with the lateral surface of each genital valve, a
little dorsad of the apices, are the stylz (st), or valvular processes,
a pair of short one-jointed cylindrical processes, each bearing
an apical pencilofhairs. The lateral geniial plates (lp) are a pair
of small folds not present in all species, occupying the proxi-
mal part of the space between the tergal margins and the
genital valves. They perhaps give additional support to
the latter.

B: ‘Nymph—The only terms that require special
mention are those used in describing certain scar-like im-
pressions and spots which are present in all Aeshna nymphs.
The lateral scars are a series of pairs of smooth areas on each
side of the abdomen, parallel to the lateral margin. The
ventral scars are two series of similar areas on the ventral
surface. The dorsal punciae consist of a number of groups of
impressed dots and transverse streaks arranged in a median

14 WALKER: NORTH AMERICAN SPECIES OF AESHNA

dorsal series, there being two pairs on nearly every segment.
The /ateral punctae are a series of dark spots or dots, one on
each side of every segment near the base and just mesad of
the lateral scars. The dorso-lateral punctae are smaller dots
mesad and distad of the lateral punctae. They are absent
from segment 9.

Measurements and abbreviations—All measurements are
given in millimetres. The terms length and breadth, when
their meaning is not otherwise obvious, refer to the antero-
posterior and transverse dimensions, respectively, of the
measured part.

The relative lengths of the occiput and line of contact
of the eyes are not accurately definable, but they are occa-
sionally of some taxonomic value. The somewhat arbitrary
point selected as the apex of the occipital triangle is the meet-
ing-point of two tangents drawn to the curves of the ocular
margins a little in front of the middle of the sides of the tri-
angle (fig. 4).

bic, 4—Method of measuring the relative lengths of the occiput (oc) and line of contact of the eyes

The latter is measured from the apex of the occipital triangle (A) to the hind margin of the

front vesicle (FV).

The measurements of the T-spot when not otherwise
indicated are those of the transverse dimension of the cross-
bar.

The lengths of the thorax, abdomen and its segments are
measured on the side along the dotted line in fig. 5.

The width of the hind wing is measured at the nodus.

The pterostigma is measured along the posterior margin.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 15

Length of Seg 3

Length of Theres. Length of Abdomen.

Fic. s—Method of measuring the lengths of the thorax, abdomen and abdominal segments.

The abbreviations used in describing the abdominal
colour-pattern have already been given (p. 9). The only
others besides those used in reference to wing-venation (vide
fig. 6) are the following: apps. =appendages; h.f.=hind femur;
h.w.=hind wing; pter.=pterostigma ; gen. v.=genital valve;
gen. 9 = genitalia of female. '

Fic. 6—Wings of Aeshna eremita Scudd. C costa; Sc subcosta; R radius; Rs radial sector ;
Rspl radial supplement; M media ; Mi, Mra, Mz, M3, Mg branches of media ; Mspl median
supplement ; Cu cubitus, Cur, Cuz branches of cubitus ; A anal vein ; at anal triangle ; al
anal loop; m median space ; cu cubital space ; CuCr cubito-anal cross-veins ; t triangle ;
spt supratriangle ; Ar arculus ; N nodus.

Genetic relationships of the genus Aeshna. (a). The sub-
family Aeshninae—The natural classification of the Aeshninae,
like that of most groups of organisms, is complicated by
the occurrence of many cases of parallel evolution which
seem to have arisen through certain orthogenetic ten-
dencies, more or less characteristic of the entire group.
As Needham (’03) has pointed out, there are five
dominant tendencies in the development of the
venation in the Aeshnidae: ‘‘(1) The similar elongation of

16 WALKER: NORTH AMERICAN SPECIES OF AESHNA

the triangle in both wings; (2) the development of strong
supplements ;|(3) the hypertrophy of two antenodal cross-veins;
(4) the development of a brace to the stigma; (5) the angula-
tion of the hind angle of hind wing in the males.’’ These char-
acters are generally least developed in the most generalized
members of each subfamily and best developed in the most
highly specialized forms, so that we find resemblances be-
tween spectalized members of different subfamilies or smaller
subdivisions which are not due to genetic relationships but
to parallel development. The brace of the pterostigma in
the Gomphinae and Aeshninae or the elongation of the tri-
angles and the forking of the radial sector in such distantly
related Aeshnine genera as Nasiaeschna and Staurophlebia i are
examples of such parallelisms.

It is therefore almost an impossible task to represent
in detail with any approach to certainty the genetic relation-
ships of the genera of such a large group as the Aeshninae.
Some approximation to the truth is all that we can expect
to arrive at.

The earliest classification of the Aeshninae that can be
regarded as an approach towards a representation of the
natural affinities of the genera in this group is the system
proposed by Karsch (’91). On venational grounds alone he
divided the subfamily into five groups; Anax, Aeschna,
Hoplonaeschna (Oplon-), Allopetalia and Brachytron, the two
dominant groups being those of Aeshna and Brachytron.
Needham (Joc. cit.) has also shown that the two main lines
of venational specialization in this subfamily are indicated
by the members of these two groups respectively, though
he excludes from the Aeshna group the genera Staurophlebia
and Neuraeschna. Karsch’s groups of Oplonaeschna and Allo-
petalia are the most generalized forms according to Needham,
while the Anax group and the genera Staurophlebia and
Neuraeschna represent independent side lines of descent.

My attempt to group these genera in a natural way is
based upon the two admirable works cited above, and I have
been able, to considerable extent, to substantiate by the
study of other characters the phylogenetic interpretations
which Professor Needham has placed upon the venational

WaLckKER: NorRTH AMERICAN SPECIES OF AESHNA 17

characters alone. In the placing of those genera which |
have not seen I have relied chiefly upon the numerous il-
lustrations in Martin’s recent work on the Aeshninae of
the Selys collection (Martin, 08).

The only important points in which my erent differ
from those of Needham are in the positions of Anax and
Hemianax and of Neuraeschna and Staurophlebia. Although
I should place the first two genera in a different group from
Aeshna, | regard them, as well as the other two genera, not
as belonging to independent side-lines of descent from primi-
tive ancestors, but as specialized offshoots from the Aeshna
line. My reasons for this belief will be given later.

The genera which represent the Aeshna line of descent,
including the four just mentioned, have the following char-
acteristics in common: (a) The radial sector is curved for-
ward under the stigma and is forked, the fork being more or
less unsymmetrical as a result of the curve. (6) The radial
supplement is bent toward the posterior margin of the wing
so that the space between it and the radial sector is widened,
being occupied at its widest part by three or more rows of
cells. (c) The median supplement is similarly curved and the
space between it and M4 similarly widened. (d) M4 is
bent away from M3, at a point about opposite the proximal
end of the radial supplement and is apparently more or less
distinctly forked at the point of deflection. In most of the
genera the trigonal supplement joins the median supplement.
To this group belong the genera Aeshna, Coryphaeschna,
Anaciaeschna, Amphiaeschna, Heliaeschna, Gynacantha, Tri-
acanthagyna, Tetracanthagyna, Platacantha, Cornacantha,
Subaeschna, Neuraeschna, Staurophlebia, Hemianax and
Anax.

To the Brachytron group belong a number of genera
which agree with those of the Aeshna group in the forked
radial sector, but differ in the absence of the other characters
given. In these the radial sector is not curved forward but
is straight and the fork is symmetrical, or nearly so, and is
generally deeper than in the Aeshna group. The radial and
median supplements are also straight, and there are only one
or two rows of cells between these supplements and the longi-

18 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

tudinal veins in front of them (Rs or M4, as the case may be).
M3 and M4 tend to be parallel throughout their course and
the trigonal supplement does not join the . median but is
separate from it by a row of cells. To this group belong
the genera Brachytron, Austroaeschna, Nasiaeschna, Epiaesch-
na, Caliaeschna, Aeschnophlebia, Telephlebia and Periaesch-
na.

Of the remaining genera which agree in the unforked
radial sector, Basiaeschna and Oplonaeschna approach the
Aeshna group, in that the radial and median supplements are
similarly curved toward the posterior margin of the wing.
Rs is slightly bent forward and Mg is more or less deflected
from M3. These features are present in but a slight degree
in Bastaeschna, in which, also, the base of the hind wing is
not much enlarged, but in Oplonaeschna the only character
which separates it from Aeshna, is the simple radial sector,
the anterior branch of which is represented, as in Basiaeschna,
by a line of cross-veins which is tending to straighten out.
In some species of Aeshna, such as Ae. caerulea, this anterior
branch is but little better developed than in Basiaeschna and
Oplonaeschna, and in many individuals of this species the
radial sector cannot be described as forked at all. The two
genera Basiaeschna and Oplonaeschna also resemble Aeshna
in all other characters and may therefore be regarded as very
nearly related to the immediate ancestors of the Aeshna
group.

The corresponding prototype of the Brachytron group
is Boyeria, in which the venation is very like that of Basiaesch-
na, but differs mainly in that Rs is still straighter, the sup-
plements straight, and M3 and Mg nearly parallel. As in
Basiaeschna the position of the trigonal supplement in relation
to the median supplement is undecided and the anterior
branch of Rs is represented by a line of cross-veins tending
to form a longitudinal vein. The median space is reticu-
lated.

The evidence of the relationships supplied by these
venational characters is supported by the characters offered
by the auricles and accessory genitalia of the males. These
structures, as found in Basiaeschna (or Oplonaeschna) and

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 19

Boyeria, are typical of the Aeshna and the Brachytron
groups respectively (pl. 12, figs. 2, 3). Inthe great majority
of the former group the auricles bear but few well separated
teeth, the hamular processes are short, not lying close to
the floor of the genital fossa, and are continuous with the
well-developed hamular folds, with which they form a pair
of concave chitinous plates. The spines of the anterior
lamina are typically well developed. In Boyeria and the
Brachytron group the auricles bear a large number of
small teeth, more or less closely crowded together,
and the anterior hamuli are strikingly different from the
Aeshna type. The bases tend to increase in size and ex-
tend caudad, the hamular folds are greatly reduced or scarce-
ly distinguishable at all (pl. 12, fig. 3) and are sharply marked
off from the hamular processes, which are large and peculiarly
shaped, being broad and elevated behind, mesially attingent
and prolonged cephalad with an acute apex which lies close
to the floor of the genital fossa. The spine of the anterior
lamina is usually less developed than in the Aeshna group.

This type of auricle, hamuli and anterior lamina, modi-
fied in detail in various ways, is present in Boyeria, Caliaesch-
na, Brachytron, Austroaeschna, Nasiaeschna and Epiaesch-
na, and presumably in other genera of the Brachytron
group. Hamuli simulating this type more or less closely are
present in Aeshna caerulea and Ae. sitchensis and in Gyna-
cantha and Staurophlebia and probably also in other genera
related to the latter two. The resemblance in these two
genera is very remarkable, and in Gynacantha is sometimes
shared to a certain extent by the auricles, which have a larger
number of teeth than are usually present in the members of
the Aeshna group. Hamuli of an intermediate character
are present in Amphiaeschna ampla, and as there is no simi-
larity in venation between these forms and the Brachytron
groups the resemblance in the hamuli is probably to be in-
terpreted as a case of parallelism not of relationship.

Of the primitive genera other than the three mentioned
Allopetalia resembles Boyeria in the straightness of the radial
and median supplements, the nearly parallel course of M3
and M4 and many venational details, but differs in the short-

20 WALKER: NortH AMERICAN SPECIES OF AESHNA

ness of the triangles and of the stigma, in the free median
space and in the much less curved course of Cu and Cu2.
In these latter characters it resembles Oplonaeschna closely,
and seems in venation to be nearly intermediate between
these two genera. As I have seen no males of this genus I
have no knowledge of the accessory genitalia: The resem-
blance between Martin’s figures of the male appendages of
A. reticulosa and Boyeria irene is perhaps worthy of note.

The three remaining genera of Aeshninae with a
simple radial sector, excluding: the Petalia group, which
is probably worthy of subfamily rank (vide William-
son,,.Proc., .\U.S.N.M,, .:. 335 Pa/27150» LOOZ),.) ake
na, Jagoria and Gomphaeschna. These are-: evi-
dently nearly related genera and differ but little from one
another in venation, except that, in the order given, the tri-
angles are successively shorter and the venation, in general,
less complex. JLinaeschna most nearly resembles Boyeria and
Basiaeschna in the comparatively complex venation and the
elongated triangles, but possesses only two cubito-anal cross-
veins and has the base of the hind wings less dilated, with a
smaller anal loop and no supplementary anal loop, which is
present in the other two genera, though imperfectly developed.
In Jagoria, and still more so in Gomphaeschna, the
number of cells in the wings is reduced, the triangles are less
elongated and the supplements less strongly developed than
in the other genera noticed. The bifid form of the in-
ferior appendage of the males present in all three of these
genera and very marked in Gomphaeschna is probably
another primitive feature. It is also present in a slight de-
gree in Allopetalia. Gomphaeschna is the only genus of this
group in which I have examined the auricles and accessory
genitalia of the male. These are quite typical of the Brachy-
tron group and much resemble those of Boyeria. It. should
not be concluded, however, that the hamuli of this group
are of a more primitive type than those of the Acshua group,
because of their presence in the primitive genus Gomph-
aeschna. On the contrary it seems more probable that the
Aeshna type, as found for instance in Basiaeschna, is the
more primitive. It is a simple folded chitinous plate borne

WALKER: NortTH AMERICAN SPECIES OF AESHNA 21

upon a Butt hase: and, apart from its Seon een hears) a
considerable resemblance to the anterior hamuli of Cordule-
gaster (pl. 12, fig.1). It seems highly probable, too, that in
Gynacantha and allied genera the hamuli, which closely
resemble those of the Brachytron group, have been derived
from those of the Aeshna type, so that here at least they are
secondary to this type. A knowledge of these structures in
‘the archaic Petalia group, the nearest living relatives of
the Aeshninae, is very desirable.
(b). The Aeshna group—Of the genera belonging to this
group Aeshna appears to be the most generalized, and is
probably the parental form of many of the other genera. ' It
is the only genus in the group in which the forking of Rs
is sometimes imperfect and we find a series of stages in the
development of this character within the genus. In Ae.
caerulea Rs is but little curved forward under the stigma and
the anterior branch is slenderer than the posterior and its
connection with the latter is not always distinct. There is
a long line of cross-veins between Rs and M2 proximal to
the fork and continued also a short distance beyond the
latter in front of its anterior branch. In Ae. sttchensis this
line of cross-veins is also present, though somewhat shorter,
and .the fork is generally distinct, but the anterior branch
comes off somewhat abruptly, the posterior branch continu-
ing the course of the main vein. In Ae. juncea the fork is
similar, but the line of cross-veins is reduced and often absent,
especially in the male. With the further development. of
the fork and the strengthening of the anterior branch the
line of cross-veins is no longer needed and is lost; Rs as a
whole tends to bend farther forward and the anterior branch
to come off at a smaller angle so that the fork becomes more
nearly symmetrical. In Ae. californica, for instance, the
fork is but slightly unsymmetrical at base and the posterior
branch no longer continues the course of the main stem.
Further in Coryphaeschna, Heliaeschna, Anax, etc., the for-
ward curvature of Rs is so marked that the anterior branch
continues the direction of the main stem, while the posterior
‘arises at a more or less distinct angle. This is most marked
in Anax and Hemianax, in which the posterior branch becomes

22 WALKER: NORTH AMERICAN SPECIES OF AESHNA

only the most distal of a series of subparallel branches arising
from the posterior side of Rs. With these progressive changes
in the radial sector are associated corresponding alterations
in other longitudinal veins. M2 becomes more strongly
curved forward, the radial supplement tends to diverge more
and more from Rs and then become convergent again; the
median supplement bears a similar relation to M4, which
tends to become more deflected from M3.

Considering now the relationships of the various genera
we must again return to Aeshna. In this large cosmopoli-
tan genus great variety exists in the characters of venation,
appendages and accessory genitalia. The latter differ so
much in different subdivisions of the genus as to suggest a
polyphyletic origin from several parent genera resembling
Bastaeschna and Oplonaeschna which, themselves, differ
considerably from one another in these structures.

Some species of Aeshna, such as californica and multt-
color, and to a less extent those of the grandis and clepsydra
groups, closely resemble Basiaeschna in the structure of the
anterior hamuli, particularly in the form and large size
of the hamular folds, while in others such as Ae. affinis,
tsoceles, cyanea, constricta, etc., there is a closer resemblance
to Oplonaeschna, in which the folds are but little developed,
than to Basiaeschna. In other species again they are of
other types and their affinities difficult to determine.

Of the other genera Coryphaeschna is a lateral offshoot
from Aeshna, in which the specialization in wing characters
has been carried a little farther than in the latter genus.
Amphiaeschna is also very near Aeshna but the median
space is reticulated. The hamuli of A. ampla are very like
those of certain species of Aeshna but approach those of
Gynacantha and allied genera in the somewhat reduced
hamular folds and the knob-like elevation of the postero-
ventral part of the processes.

Of the remaining genera (Heliaeschna, Gynacantha, Neur-
aeschna, Staurophlebia, Tetracanthagyna, Triacanthagyna,Plata-
cantha, Cornacantha and Subaeschna) I have seenexamplesonly
of the first four, but with the possibleexception of Subaeschna,
only the male of which is known, they appear to form a

WALKER: NORTH AMERICAN SPECIES OF AESHNA 23,

natural group, having much in common in their venational
plan in spite of certain apparently important differences, and
in which the membranule is much reduced and the tenth ab-
dominal segment in the female is not spinulose as in most
genera, but is armed with a few long spines or with a long
bifid process.

Needham’s objections to associating the genera Neur-
aeschna and Staurophlebia with the Aeshna group are that the
tip of the subcosta in these genera extends beyond the nodus,
the stigma lacks a brace, and in Neuraeschna (and also
Heliaeschna and Amphiaeschna) the median space is reti-
culated. These appear to be primitive characters but in
view of the otherwise close resemblance in venation and
other characters between these genera and others of the
Aeshna group, especially Gynacantha, I feel satisfied that they
belong to this group and have acquired these characters
secondarily. The presence of cross-veins in the median space
in such genera as Neuraeschna may well be due to the same
influences which have brought about the general multipli-
cation of cross-veins in this genus. It is also very doubtful
whether the apparent continuation of the subcostal vein
beyond the nodus really belongs to that vein, or is a short
series of cross-veins in line with it, as its irregular course
suggests. The point can only be settled by a study of the
early stages. Furthermore I see no reason why the brace
of the stigma should not be lost by reversion to an earlier
condition in which it was not needed. In some species of
Aeshna, as Ae. stichensis and Ae. caerulea septentrionalis,
the stigma of one pair of wings may be sometimes well
braced, that of the other pair not at all (cf. fig. of Heliaesch-
na uninervulata in Coll. Zool. Selys, XX, p. 163, fig. 164).

For similar reasons it seems to me more natural to re-
gard Anax and Hemianax as representing a highly specialized
offshoot from the Aeshna group than as constituting an in-
dependent line of descent. The position of the media at
the arculus and the rounded anal margin of the hind wing of
the male are considered by Needham to be primitive char-
acters, but in none of the otherwise primitive genera of Aesh-
nidae are these characters found, and in all other respects

24 WALKER: NorTH AMERICAN SPECIES OF AESHNA

Anax and Hemianax are very highly specialized. More-
over in the genus Anaciaeschna, which is unmistakably very
close to certain species of Aeshna (cf. figures of the wings of
Aeshna martint and Anaciaeschna triangulifera, Coll. Zool.
Selys, XVIII, p. 30, fig. 24 and p.72, fig. 70), some of the peculi-
arities of Anax are present in a rudimentary condition; the
upper part of the arculus is shorter than the lower, vein M2
is strongly curved forward behind the stigma in a manner
resembling that of Anax, the auricles are very small (absent
in Anax) and there are vestiges of the supplementary lateral
carinae, which are well developed in that genus. The ham-
uli and anterior lamina of Anaciaeschna jaspidea and of
Aeshna affinis and Ae. tsoceles, which belong to that section
of the genus which is most nearly allied to Anaciaeschna, are
very much alike and bear a decided resemblance to these
structures in Anax.

_ My views on the relationships of the generaof Aeshninae
are expressed in the accompanying phylogenetic tree. (Fig. 7.)

Oplonaeschna | Coryphaeschna Hehaeschna
s VAwaiaenchina Z-— Neuraeschna

> a
} a WT meas 0 Wile Gyna canthas (24rephlebia

’ ee
B® <= Anaciaeschna SS Plectane Het aak
Subaeschna \Cornacantha
5 Tetr thao
Gomphaeschna ae a a

Teléphlebia
Periaeschna
Caiiaeschna

Hemianax --
Anax .

Austroaeschna

Epiaeschna
Nasiaeschna P ch

Aeschnophlebia

Fic. 7—Diagram showing the relationships of the genera ot Aeshninae. ©

It is impossible to arrange these genera satisfactorily in
a linear series. The arrangement that appears to me most
consistent with the phylogenetic tree is that which is given
below.’ -It has the necessary defect that the most primitive

11 have omitted FGrster’s genera Limmnetron, Protoaeschna and Rhionaeschna,
as I have seen neither specimens nor figures of any of them. Limnetron evident-
ly belongs to the Brachytron series while the other two are apparently members
of ‘the “Aeshna series.

WALKER: NORTH AMERICAN SPECIES OF AESHNA = 25

genera of the two great groups of Aeshna and Brachytron are
widely separated from one another, e. g., Basiaeschna and
Oplonaeschna from Boyeria, Allopetalia, etc., but similar
difficulties are encountered in any other attempt to arrange
the genera naturally. The only alternative would be to
divide the genera into three main groups, the first compris-
ing the primitive genera (those in which Rs is not forked,
etc.), the other two containing the specialized members of
the Brachytron and Aeshna groups respectively. This method
is perhaps the simpler of the two, but is less exact than that
which I have adopted.

BRACHYTRON GROUP. AESHNA GROUP.

Allopetalia Bastaeschna
Gomphaeschna Oplonaeschna
Boyeria 4; Jagoria Aeshna Aeshna
Series Linaeschna Series Coryphaeschna
Boyeria Anaciaeschna
| Amphiaeschna
Telephiebia Gynacantha
Caliaeschna Triacanthagyna
Brachy- | Periaeschna Platacantha
tron } Austroaeschna Cornacantha
Series Nasiaeschna Gynacan-} Tetracanthagyna
Epiaeschna tha Series | 7eliaeschna
Brachytron | Staurophlebia
Aeschnophlebia Neuraeschna
Subaeschna
Anax Hemianax
Series { Anax
Variations.

(1). Climatic variations—On a whole the variations
among the individuals of a given species of Aeshna in a given
locality are not great. The only striking ones are found in
the colour of the females, in which there is a tendency to-
wards dimorphism. This subject will be considered later.
The abdominal appendages of the female, which are appar-

3

26 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

ently functionless structures, and the third segment of the
abdomen, are also subject to considerable variation, the form-
er in length and the latter in depth. The range of variation
in these parts in widely distributed species is found, however,
to be much greater over the whole area of distribution than
in any one locality. A correlation is also noticeable in the
variations of these structures. Thus in the females shorten-
ing of the abdominal segments, especially segment 3, is ac-
companied by deepening of the same segments and a shorten-
ing of the appendages. In the males the latter variation
does not appear, as the appendages are functional structures
and must be of a more or less definite length.

It may also be shown that these variations are depend-
ent in a great measure upon locality. Thus if we compare
specimens of Ae. eremita from the following three localities,
Bay of Islands (Newfoundland), Heyden (northern Ontario),
and Toronto (southern Ontario), we find that the abdomen
of the Newfoundland females is distinctly shorter and deeper
than in the Toronto specimens, and the appendages are
much shorter, while the Heyden females are about inter-
mediate between those from the two other localities.

Similar variations occur in the case of Ae. interrupta,
canadensis, juncea, palmata and umbrosa, and it may be fur-
ther shown that the various species from a given locality
resemble one another more or less closely as regards these
variable features. Thus in Newfoundland and the Magdalen
Islands, Aeshna juncea, canadensis, eremtta, and interrupta
all have relatively short and deep bodies and the females
have short appendages as compared with the same species in
Ontario; while the same differences in less degree are observ-
able in the specimens of these species from northern On-
tario or northern Michigan and southern Ontario respective-
ly (except in Ae. juncea which does not occur in southern
Ontario). In the case of the widespread and common Ae.
umbrosa, specimens from Anticosti, Vancouver Island and
New Bridge, Oregon, are very similar in the comparatively
short abdomen, which is relatively stout at seg. 3, while in
those from Toronto, Bluffton (Ind.), and various localities

WALKER: NORTH AMERICAN SPECIES OF AESHNA a7

in New York, Pennsylvania and Ohio the abdomen is long
and slender, particularly at seg. 3. Specimens from northern
Ontario and Manitoba come nearer the latter than the former,
though varying somewhat towards those of the first category,
especially in the females... Again in the circumpolar Ae.
juncea, specimens from the Magdalen Islands (Que:), Alaska
and Great Britain are nearly alike in form and are stouter
than specimens from Nipigon (Lake Superior), the Bighorn
Mountains (Wyoming) and other parts of the interior of
North America.

The obvious correlation of these variations with locality
suggests climatic influence as a factor in their causation, and
as the form of the body does not change during adult life
we must look for the factor in the environment of the nymph.

In an aquatic life like that of the nymph the most im-
portant climatic influence is probably that of temperature,
although light may also be of some importance. During the
winter the nymphs are obviously not influenced much by
temperature as they cannot withstand freezing and there-
fore must live in water at 0° C. wherever the ponds or streams
in which they breed are frozen over throughout the winter;
and this is the case throughout nearly all of the territory
under consideration. Moreover there is probably no growth
of the nymph during the winter as they apparently take no
food at this time, at least in the later stages. That the winter
temperature has nothing to do with the phenomena in question
is also obvious when we compare the mean temperature
of winter or of January, the coldest month, in the various
localities where the individuals of a given species of Aeshna
are similar in form. Thus of Ae. umbrosa specimens from
Treesbank, Man., and Toronto, Ont., are nearly identical
but differ considerably from Anticosti examples, although
the mean January temperature of Anticosti is nearly midway
between those of the other two localities: thus, Anticosti
(eastern point),11.9° F.; Toronto, 22.9° F; Treesbank (Winni-
peg), 2.6° F.?

*These and the following temperatures were kindly furnished by Mr. R. F.
Stupart, Director of the Meteorological Observatory, Toronto.

28 WALKER: NortTH AMERICAN SPECIES OF AESHNA

But if we compare the mean July temperatures, or better,
those of the season of growth from May to September in-
clusive, we find a decided correlation between the variations
in temperature and those exhibited by the insects, This
may be illustrated by the following tables' (1. = length, d. =
depth). (See also plate 1.)

(1). Aeshna eremita.
Mean | No. of | se | =

Locality. pee oe (mean) tae ie pede: =e
(romp. siege SS) nes a | ie | 7.3 | sha ee an
ae ee ee ' 564 5 3 | 8.31 | sai | 3.19 | 6.51
Toronto, Ont. Pe ed Ceska We 8 |} 87 | 24 | 7.75
| e

(2). Aeshna Luadence
~~ | Mean Ne, of |, .. > | segs|d> secu meen
Locality. | Temp. | Speci- 1 abd, 3 3 cee
_|May- -Sep.| mens. (mean) |(mean) a)

pied Enteps Magdalen Is.

(Temp. Grindstone, Magd. Is.) ao 2 45.5 | 6.92 4.5 4.2
pe ree -|——— | Bs
Nipigon, Ont. se > + ae is
(Temp. Port Arthur, Ont.) a ; 47.0 7.15 | 4.0 | 4.7
—}—— | _]. ae
2.75 | 5.6
|

Ottawa, Ont. | 62.4 | 2 ye 8.00

(3). Aeshna umbrosa.
Mean, } Navotl i. fo

Locality. _ Temp. | Speci- | lL abd. | I. seg. 3 be seg. 3
|May-Sep.| mens.

+o 4 ear AM i aR 48.0 1.33). |. Aa Oe
Ellis Bay, Anticosti Id., 49.2 3 8. 5 Av 8.33 Av. 3 Av.
(Temp. East Point, Migteosth ) rr 348 0 . 25) 797 29 1.97
Ai. HITLRD, Ee Lt f 50. G: *s — {A Ale
Nipigon, Ont. ns f 20 on \ Av. |g nay Av. ay,

(Temp. Port Arthur, Ont.) | ang : 51.0 90-75). 9/8 8.96 ied 1.9
Gee - }— = IS bE ose
Winnipeg Beach, Man. tee > |ol.y Av. |9.0 ) Av. {1.7) Av.
(Temp, Witnipes, Man.) | 27 2 150,550.75 |9.25/9.12|1.9F 1.8
RA AY A 152.0 ; 9.7 ae y ‘
Toronto, Ont. 63.3 3 1.0} AY. 9. a3} gins L, go | Av.

50.07 °° 1.75)

\The tables are merely illustrative, my conclusions having been based upon
the examination of a much larger series of specimens.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 29

Thus with an increase of the mean summer temperature
there is an increase in the length and a decrease in the depth
of segment 3, and an increase in the length of the female ap-
pendages. Similar but slighter variationsin some of the other
abdominal segments posterior to segment 3 are demonstrable
insome cases, but segments I and 2 seem to take little or no
part in these modifications.

1 am as yet unable to offer any suggestions as to the
operation of this factor of temperature in the production of
these variations.

(2). Colour variation of females—The females of all or
nearly all the North American species of Aeshna are very
variable in colour but the variations are all in one direction.
A certain number of females of each species are coloured like
the males in which, with very few exceptions, the abdomen
is nearly black, spotted with blue. In a number of other
females all the pale markings are yellow or yellowish green
and the wings often flavescent. These two types may be
known respectively as the homceochromatic and the hetero-
chromatic types—terms employed by Calvert (’05) to desig-
nate similar colour phases in the females of Ischnura and
other Coenagrionine genera. These two colour phases are,
however, not sharply marked off from one another, but are
connected by intermediate forms. Homceochromatic females
are generally less common than the heterochromatic and in-
termediate forms, but the proportion varies in different species.
In some species, e.g., Ae. canadensis, heterochromatic females
are never pure yellow but always greenish; while in others,
such as Ae. constricta and californica, the extreme hetero-
chromatic examples are yellow-spotted without a trace of
green. This tendency to dimorphism is doubtless of the
' same kind as that which occurs in a more perfect form in
Ischnura, Anomalagrion, etc. The so-called third form of
female, met with in some species of these genera, is probably
merely an intermediate between the other two forms.

(3). Geographical races—While some species of Aeshna
of wide distribution are very uniform over their entire range

3

30 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

of territory, others are more or less distinctly divisible into
geographical races or subspecies. The characters which
separate these races are altogether different from those
whose variability depends on climate and hence also on
geographical distribution, but are similar in kind to the
specific characters of the genus. The species in which such
races are best defined is Ae. interrupta, which ranges from
Newfoundland and the northern New England States to
Great Slave Lake, and the northwestern coast of British
Columbia, and thence southwards to Nevada and New
Mexico. It includes three dominant races, interrupta, which
inhabit the wooded regions east of the Great Plains; lineata,
the characteristic Aeshna of the Canadian prairies, found
from the Dakotas to Great Slave Lake; and interna, a moun-
tain form, ranging from the southern boundary of Alberta
and British Columbia into New Mexico.

Lineata and interna intergrade in the southeastern
Canadian Rockies; a form appears in northwestern British
Columbia extremely like the eastern interrupta but approach-
ing also the other two races; in Nevada another occurs
(nevadensis), also approaching all three of the other forms,
while in the Magdalen Islands, in the extreme east, the
species takes on certain features of colour-pattern resemb-
ling those ot the western races interna and nevadensts.

Ae. umbrosa is likewise divisible into an eastern race
umbrosa and a western one, occidentalis, the latter approach-
ing Ae. palmata,the characteristic western species of the same
group, more closely than does the race umbrosa, while the
latter in the Canadian Zone (in northern Ontario) approaches
the race occidentalis. Ae. palmata itself varies considerably
throughout its range and it would thus appear that the varied —
topography of the western half of the continent has reacted
on the species of Aeshna as on many others groups of organ-
isms with the production of many local races.

General Life History.

Season of imaginal life—In Canada and the northern
United States, the metropolis of the North American species

of Aeshna, the imagoes of most species are abroad during
a

WALKER: NORTH AMERICAN SPECIES OF AESHNA 31

the latter part of summer and in early autumn. July,
August and September are the months when they are most
abundant. Most of the species first appear in July, but Ae.
canadensis is generally on the wing before the end of June
in the Transition Zone in Ontario, while Ae. californica,
multicolor and mutata reach maturity still earlier, the two
first named species appearing early in April. Most of the
species have nearly disappeared by the beginning of October
but Ae. umbrosa, and probably others such as Ae. constricta
and Ae. verticalis, may linger on until the middle of the month,
disappearing only after severe frost.

Length of tmaginal life—As the majority of individuals
of a given species in a given locality emerge within a period
of about two weeks, the length of the imaginal life would
appear to be at least a month, probably considerably longer
in most cases. Reared specimens kept in captivity and un-
fed live only a few days and do not acquire their mature
coloration.

Habitat—The imagoes are found, as a rule, most abun-
dantly about their breeding grounds, but may often be seen
foraging in large numbers more than half a_ mile
away from any possible breeding-place. The majority of
species breed in still shallow waters, thickly grown up with
the smaller species of reeds, sedges, Sparganium, Acorus
calamus L., Equisetum fluviatile L., and other plants of simi-
lar habit. Open marshes bordering rivers or at the mouths
of sluggish streams, shallow reedy lakes, ponds or bays, are
favourable localities for most of the species, and during the
season of flight the imagoes may be seen foraging over these
marshes, generally at a height of two to five feet above the
vegetation and following as a rule, no regular course; though
they may also be frequently seen flying low and following the
water’s edge, darting in and out of the washouts or little
recesses between the clumps of reeds and rushes, on the
lookout for their prey. Ae. umbrosa is an exception to all
the other eastern species that I have observed in the field, in
that it frequents small woodland streams and ditches, or
small pools on the edges of woods, never being found asso-
ciated with the other species in open marshes. The imagoes

32 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

fly up and down such streams or pools keeping, as a rule,
near the water, and when the streams are large, close to the
margins.

Field notes made at Temagami, Ontario, Sept. 14, 1908, on the species of
Aeshna observed about a small lake between Lakes Wakimaka and Obabika,
Temagami Forest Reserve :

This lake is an expansion of the Obabika Creek which connects the two
lakes named above, and is broadly margined by an open marsh, behind which is
a dense spruce forest. Just below the lake the creek is broad and passes fora
short distance through the same kind of open marsh, thence entering the deep
woods. On the west side is an area of several acres of marsh in the midst of
which is a pond of some 300 square feet in area. It was about this pond and
between the pond and the creek that the Aeshne were most abundant. The soil
is a soft dark mud into which one would frequently sink to the knees, but, as a
rule, owing to the abundant cover of horse-tails and sedge, one could obtain a
fairly good footing. Mvoose-tracks were numerous about the margins of the
creek and along the edge of the woods. The vegetation of the marsh consisted
largely of Zguzsetum fluviatile L. with a few sedges and coarse grasses, and in
a few places where the ground was a little higher, Myrica gale L. grew in
abundance with tufts of Spiraea salicifolia L., Triadenum virginicum L. and
a few grasses.

Four species of Aeshna, Ae. eremita, interrupta, canadensis and subarctica,
were taken here,and the following additional species of Odonata: Lestes congener,
L. disjunctus, Enallagma hageni, Somatochlora williamsoni, Sympetrum scoticum,
costiferum, semicinctum and obtrusum. The Aeshne were very abundant but ex-
tremely difficult to capture. Ae. interrupta was the most abundant, de. cana-
densis coming next. Of Ae. eremita quite a number were seen but only two
captured, while of Ae. subarctica only one was taken. They were flying back and
forth following no definite course, though many would skirt the margin of the
creek for some distance and then fly inwards over the marsh or across to the
other side. They seemed to be similarin habits and character of flight. The
males, which appeared to be in the great majority, were foraging and seemed to
prey entirely upon a species of caddis-fly (Limnephilus indivisa Walk.)' which
was very abundant among the horse-tails and sedge of the marsh. The male
Aeshne were flying rather low, 3-5 feet from the ground, and were constantly
descending within a few inches of the water, making their way for some distance
among the horse-tails on the lookout for caddis-flies. Their rustling could be
heard at a distance of 6 or 7 yards. When two males came near one another
they would dash off together up into the air but generally soon separated and
continued their foraging.

Shortly after reaching maturity the imagoes may often
be found in large numbers at some distance from their
breeding-grounds, flying about the borders of woods or in
openings in them, along wood-roads, etc. Some species,
particularly Ae. canadensis, are especially attracted by open
coniferous woods and are fond of settling on the trunks in
the sunshine or hanging from twigs. Ae. umbrosa shows a
marked preference for more or less shady haunts, while its
near ally, Ae. constricta, is most often seen ranging over open
fields or bushy pastures. Late in the season they return to

their breeding-grounds.
1Determined by Mr. Nathan Banks.

WALKER: NORTH AMERICAN SPECIES OF AESHNA = 33

Influence of weather conditions — Although generally
most active in bright sunny weather the species of Aeshna
are influenced to a less degree than most Odonata by con-
ditions of light. Most if not all of our species fly readily
in dull warm weather and sometimes even during
very light rains, while Aeshna umbrosa habitually flies till
well after dusk. This habit, which is common among the
Aeshninae, has also been observed by Mr. Williamson in the
case of Ae. interrupia, eremita and canadensts.

According to Mr. C. H. Kennedy, the activity of certain
species of Aeshna is decreased in intensely hot weather. In
a letter to the writer he says of Ae. californica and multi-
color, at Sunnyside, Washington: ‘Both californica and
multicolor hang in the shade from the underside of leaves of
trees on very hot days (when the thermometer rises to 100° or
105° F.). I have not noticed them resting on days when the
temperature was less than 100.° These rests last for a few
minutes only but occur at short intervals. It is only at
such times that they are easily taken away from the water.
One very hot afternoon for a few minutes I caught them,
while hanging up, as fast as I could empty my net, and two
only fifteen feet from the back door.”

In Ontario, where the weather is very rarely as hot as
described above, this habit must be attributed to some other
cause. Indeed, it is on the warmest days that the Aeshnae
are least likely to be seen resting, while in cool bright weather
although they fly readily enough, they take very frequent
and often prolonged rests, sometimes in the shade, sometimes
in sheltered sunny places, the difference here being also
probably a question of temperature. On the other hand on
hot still days, when the thermometer is at 80° or go°F. they
are often very restless and almost ceaselessly on the wing.
On July 22, 1910, between 4 and 4.30 p.m. the writer observed
Aeshnae flying in considerable numbers at De Grassi Point,
Lake Simcoe, among the young bushy trees near the edge of
a dense low wood bordering a large area of pasture land.
The weather was very hot and still and the sun partly ob-
scured by the smoke from forest fires. It was observed that
Aeshnae kept in the small open places, a few square yards in

34 WALKER: NortTH AMERICAN SPECIES OF AESHNA

area, which were well shaded by the surrounding trees.
Generally speaking each of these small spaces was occupied
by not more than a single Aeshna which flew around the space
in a more or less regular course at a few feet from the ground,
and was not readily frightened by my attempts to capture it,
being driven away only when actually touched by the net.
A number of specimens were captured, all except a single
female Ae. constricta proving to be Ae. canadensis.

Similar flights of Aeshnae have on several other occa-
sions come to my notice, as shown by the following extract
from my note-book: ‘‘On August 28, 1909 (at De Grassi
Point, Lake Simcoe), Aeshnae were observed in consider-
able numbers flying over the tennis lawn and garden, a few
rods from the lake shore. The weather was cloudy but very
hot and still and the Aeshnae were for the most part flying
within a few feet from the ground, each confining its move-
ments to a more or less definitely limited area or beat. There
were apparently about fifteen or sixteen individuals flying
about the tennis lawn alone and about half that number
flying over a small vegetable-garden about 20 square yards
in area. The land surrounding the lawn and garden is cover-
ed with long grass, tall herbs, numerous bushes and young
trees and the Aeshnae were flying here too, but it was notice-
able that they preferred the more open places. They were
hawking among the multitudes of Chironomidae that were
abroad and were often observed to capture them. At 6.30
p.m. they were still numerous, but at 7.00 p.m. only one was
seen, flying rather high. The following were captured:
Ae. constricia, 32 ¢:; Ae.’ canadensis, 26, 3 9°
umbrosa, 1°.”

A similar but smaller flight was observed about ten days
before this one, about the same time of day, the weather
being likewise still and rather warm, although the sky was
clearer. As I was without a net only one was captured, a
female constricta, but others were recognized as belonging to
this species. The first flight of this kind that I remember
was on a warm still afternoon in August 1906. The insects
were observed at the same locality, about 5 p.m. or a little
later and it was estimated that about twenty-five Aeshnae

WALKER: NORTH AMERICAN SPECIES OF AESHNA = 35

were flying over the tennis-lawn at one time. A number
were captured and all proved to be Ae. constricta.

Migrations—Brown (’91) records a migratory swarm of
dragonflies in Wisconsin, one of which was captured and
determined as Aeshna eremita Scudd. This appears to be
the only case on record of a North American species of this
genus exhibiting the migratory trait, though Campbell (’85)
described a similar swarm of the European Ae. mixta,which he
observed flying along the banks of the Gironde in France.
They were first noticed about five o’clock in the afternoon,
and the swarm lasted from one and a half to one and three-
quarters hours. ‘‘ The weather was fine and warm but the
sky was clouded and rain had fallen during the day. There
was little or no wind.”

Seasonal variation in number of individuals—T he number
of individuals of some species of Aeshna varies greatly from
year to year. The season of 1906 in the vicinity of Lake
Simcoe was remarkable for the extraordinary abundance of
Ae. constricta and Ae. canadensis, while in 1907 both of these
species were comparatively scarce. Since then they have
both been common enough, but not remarkably so, until the
‘season of 1911, when very few individuals of either species
were observed. Ae. verticalis is also inconstant in numbers,
occasionally occurring inabundance. Ae. umbrosa, on the other
hand, seems to be more or less common every year, but appar-
ently never occurs in excessive numbers as in the case of the
other species mentioned.

The relative scarcity of individuals during some seasons
is not dependent upon the drying up of the waters in which
they breed. Ae. umbrosa is the only species that would be
at all likely to suffer from this cause, since its nymph often
breeds in small pools and ditches, but it is the most constant
in point of numbers of all the species whose habits are fami-
liar to me. Possibly the irregularity of appearance is de-
pendent upon parasites.

Food—The food of Aeshna consists of flying insects,
generally of small size, especially Chironomidae and other
Diptera, caddis-flies, small moths, etc. On Aug. 4, 1910, I
observed a male Ae. eremita at Nipigon, Ontario, flying about
over a small clearing on the river shore with a grasshopper

36 WALKER: NORTH AMERICAN SPECIES OF AESHNA

in its jaws. It was watched for three or four minutes,
while it flew about the edge of the woods. The grasshopper
was of about the size of Camnula pellucida and was probably
this species, which positively swarmed on the clearing,
to the exclusion of almost all other Orthoptera. I
have also observed the nearly allied Basiaeschna
janata feeding upon a somewhat teneral specimen of Gomphus
spicatus, and in the U.S. National Museum collection
there is a female specimen ofNasiaeschna pentacantha which
was taken with a worn specimen of Papilio asterias in its
clutches. The European Aeshnine Brachytron hafniense has
been observed to feed upon Libelluline dragonflies. -

A microscopic examination of the stomach contents of
a female Ae. canadensis taken while foraging revealed only
minute chitinous fragments of small Diptera and possibly
of other insects.

Enemies—In the adult state Aeshna seems very well
able to take care of itself and has but few enemies. The
most critical periods of its life are doubtless the time of emer-
gence from the nymph and while still teneral; and in the
case of females, while they are engaged in oviposition.

Tenerals are doubtless often caught in spiders’ webs,
but no instance of the kind has come under my notice. On
one occasion, however, I found a fully mature female of
Aeshna tuberculifera in the web of a spider (Argiope trifasciata
Forsk.).1. This was on the banks of the Etobicoke Creek,
near Toronto, on Sept. 30, 1908. The web was in the grass
on the steep slope of the bank about two feet from the water's
edge, and the dragonfly whose thorax was partly eaten had
evidently been recently killed as the colours were still fresh.

Doubtless tenerals also frequently fall victims to some
of the larger insectivorous birds. Mr. R. P. Currie sent me
a specimen of Coryphaeschna ingens and the male abdominal
appendages of Epiaeschna heros, both from stomachs of the
chuck-will’s-widow. The latter species has also been
recorded as having been captured by the king-bird (7 yran-
nus tyrannus) (Moore, ’0o). *
Possibly the most serious enemy of the adult Aeshna

‘Determined by Mr. Nathan Banks.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 37

is the frog, which no doubt destroys many ovipositing fe-
males. Mr. C. H. Kennedy suggests that the scarcity of
females of some species of Aeshna, notably Ae. umbrosa and
Ae. palmata, is due to this cause. In support of this belief he
has sent me the following interesting observations on these
species, made in the Blue Mountains, Oregon:

“Of Ae. umbrosa and Ae. palmata the females were a
rarity, males were much more abundant. This was inex-
plicable to me until one day I saw a female umbrosa while
Ovipositing knocked into the water by a frog. Thestreams
and ponds, particularly of the Blue Mountains, are almost
swarming with a medium-sized frog. I examined the stom-
achs of several but found only grasshoppers and water-bugs.
However, I felt certain that the abundance of frogs explained
the paucity of female Aeshnas, because in both umbrosa
and palmata, while the males are high wide fliers, the females
nearly always fly less than a foot above the water and even
lower when ovipositing.

“The above observations were made in Eagle Valley,
- all of which lies in the upper Sonoran and Transition Zones.
Later when | collected in Pine Valley, which is a higher val-
ley, | had my case against the frogs strengthened.

‘The lower end of Pine Valley lies in the Transition Zone
but the upper end is in the Canadian Zone. These frogs are
not found in the Canadian Zone. They are peculiar to the
sage-brush regions, ceasing abruptly at the lower edge of the
timber. Only males of umbrosa were seen in the lower end
of Pine Valley, where the Meee were abundant, while females
were as abundant as the males up at Carson in the Gy
or timber zone, where there were no frogs.”’

[ may add that in Ontario, also, the females of umbrosa
are apparently much rarer than the males, though my ex-
perience in rearing the nymphs and collecting the exuviae
shows that the two sexes enter the adult stage in about equal
numbers.

In this connection Mr. Kennedy also observes that the
males and females of Ae. californica and multicolor in the
Yakima Valley, Wash., occur in about equal numbers, and
he attributes this fact to the entire absence of frogs in the

38 WALKER: NorTH AMERICAN SPECIES OF AESHNA

ponds of this valley: ‘‘There are here a small ground-in-
habiting tree frog and a small toad, neither of which resort
to the water except for a short time in the spring to breed.”’

Mating habits—Pairing may take place at any time
during the season of maturity. Mr. Kennedy has observed
Ae. californica in cottu as early as May 8, and I have seen
Ae. clepsydra paired at Go Home Bay, Georgian Bay, On-
tario, before the end of July, and took a female Ae. canadensis
ovipositing at the same locality on July 17, 1907. August
and early September are, however, the months in which pair-
ing may be most frequently observed.

The pairing habits differ somewhat in the different
species of Aeshna. Of the common species that I have ob-
served, Ae. constricta isconspicuous for its habit of pairing in
open bushy places often some distance from water, for its
wild nuptial flight and the tenacious grasp in which the male
holds the female. When abundant it is very often seen in
copula, while I do not recall a single occasion on which I have
identified with certainty a pair of the still more abundant Ae.
canadensis. Ae. verticalis | have also on several occasions
taken paired a long way from its breeding haunts, but its
habits appear to be more retiring than those of constricta
and unlike the latter species the pairing individuals separate
immediately when taken into the hand.

Mr. Kennedy writes: ‘‘While in copulation pairs of
californica cling to bushes, not indulging as much in wild
nuptial flights as multicolor, which species in copulation is not
often seen hanging to bushes.’’ It is probable that the
highly differentiated male abdominal appendages of con-
stricta and multicolor, and the unusually large genitalia of
both sexes in constricta, are an adaptation to their active copu-
lating habits, in which a firmer union between the two sexes
is necessary than in such retiring species as Ae. verticalts
and californica, the males of which have comparatively simple
abdominal appendages.

In most species the males seize the females while ranging
over the reeds and rushes which grow in their breeding
places. Very frequently the females are picked up while
ovipositing. If copulation ensues, the pair usually fly off

WALKER: NorRTH AMERICAN SPECIES OF AESHNA 39

to the nearest trees, often after circling about in the air a
few times.

As might be expected from the close resemblance of
the various species to one another, the uniformity of colour-
pattern and unspotted wings, together with the marked dif-
ferences in the external genitalia and male abdominal ap-
pendages, there is no contest between the males for the fe-
males, nor are they apparently able to distinguish the fe-
males of their own species from those of other species of the
genus. Pairing between different species is probably pre-
vented by the mutual inadaptibility of the genitalia of the two
SeXes.

In the Temagami Forest Reserve, Ontario, the following notes were made
on the pairing habits of the species of Aeshna observed there (the locality and
date are the same as those quoted on p. 32):

‘‘Now and then a buzzing of wings was heard in the reeds anda pair would
emerge, flying off swiftly, sometimes making a bee-line for the neighbouring
woods, sometimes circling about at first but always flying eventually to some
sheltering bush or tree.

“Several times pairs were observed which were not i” coztu, the male simply
grasping the head of the female by the abdominal appendages. Some of these
Pairs were flying with the bodies of both sexes extended after the manner of the
damsel-flies (Coenagrionidae), no attempt at copulation being made. In others
the abdomen of the female was seen to curve upward frequently toward the male
accessory genitalia but without establishing a connection.

‘*‘While some or all of these partial unions may have taken place between
different species, two positive cases of this kind were observed. A male of Ae.
subarctica was taken with a female of Ae. canadensis and a male of Ae. interrupta
also with a female of canadensis. In the former case no union had been effected
when the pair was captured though takenin the net at the same time. It was
believed, however,that the male had taken hold of the female by the appendages.
In the latter case a connection was apparently established for a few seconds be-
tween the genitalia of the female and the accessory genitalia of the male but the
pair dropped to the ground and a scuffle ensued, during which they were
captured.

“‘The only pair zm coztu that was observed at close enough range to permit
of recognition of the species was one of Ae. interrupta, which was followed some
distance to the bushes at the edge of the wood and then approached very closely,
though not captured.”

In the first week of August, 1910, the writer observed a number of copulat-
ing pairs of Aeshna on the Nipigon River, where six species of the genus are
more or less common. Several pairs of eremita and tnterrupta were recognized
but nothing of special interest observed except a series of three individuals con-
nected in line and consisting apparently of two males and a female, the first male
holding the head of the second, which in turn was grasping the head of the
female. I also saw two males making frantic efforts to secure a single female,
but apparently not attempting to drive each other away. They were flying at a
censiderable height and the species was not recognized.

Copulatory position—In the paper by Williamson and
Calvert, on ‘‘Copulation of Odonata’‘ (Ent. News, XIII,
1906, pp. 143-150) the former writer states that in copulating

40 WALKER: NORTH AMERICAN SPECIES OF AESHNA

the male Aeshna grasps the female by the head, the inferior
appendage resting on the top of the head, the superior appen-
dages on the rear of the head. The plate accompanying
this paper (pl. VII) is from a photograph of a pair of Ae.
umbrosa (referred to constricta) in coitu, and shows the general
position of the two sexes in copulation, but the appendages of
the male are no longer in contact with the head of the female,
nor is the connection between the genitalia of the female
and the accessory genitalia of the male exactly as it is in life.
I have been fortunate in obtaining two pairs of Ae. constricta
and one of Ae. clepsydra in all of which the natural position
of the male appendages has been retained, and in one pair of
constricta the connection between the genitalia of the two sexes
has also been preserved. Both pairs of constricta which were
taken at De Grassi Point, Lake Simcoe, Aug. 20 and Sept. 2,
1909, were carried in the hand over a quarter of a mile before
the benzine necessary for killing them suddenly could be
obtained. The pair photographed (pl. 2, fig. 1) in which the
natural position has been preserved in every detail was dried
in an envelope over a coal oil lamp after being killed with
benzine. The pair of Ae. clepbsydra was taken at Go Home
Bay, Georgian Bay, by Mr. W. J. Fraser, who immediately
severed the abdomen of the male at the sixth segment,
thus leaving the distal portion and the appendages adhering
to the female.

An examination of the two pairs of constricta shows that
Williamson’s statement is correct as far as it goes, but that
in addition to the head of the female being grasped by the
superior and inferior appendages of the male, the prothorax
is held between the two superior appendages, the supero-
internal surfaces of the latter being applied to the lateral
surfaces of the pronotum (pl. 2, figs. 3 and 4). The preapical
spine of the superior appendages fits into the groove or space
between the rear of the head and the cardo or basal joint of
the maxilla, and the preapical tubercle, on the inner surface
rests in a depression just under the lateral margin of the an-
terior lobe of the prothorax.

In Ae. clepsydra, (figs. 5 and 6) the position issimilar but the
prothorax of the female is apparently less firmly grasped by

WALKER: NortTH AMERICAN SPECIES OF AESHNA 4I

the superior appendages of the male, which are much narrower
and simpler in structure than in constricta. The pointed
apices occupy the same position as the preapical spine of
constricta, with which it is homologous. Probably the slightly
elevated and denticulate apical portion of the superior car-
ina corresponds in position to the tubercle in constricta, but
this cannot be determined from the specimen.

The relative positions occupied by the accessory geni-
talia of the male and the genitalia of the female in Ae. con-
stricta are shown in plate 2, fig. 2. The ovipositor (Ov) is
unsheathed from the genital valves (GV) and lies for the
most part in the genital fossa of the male (ventral side of seg.
2). In its basal half it is grasped by the posterior hamuli
(PH) and from its position and relation to the anterior
hamuli (AH) it must pass through the channel formed by the
concavities of the hamular processes and folds, the two an-
terior hamuli thus forming a pair of claspers. The two
parts of the anterior lamina are closely applied to the ventral
surfaces of the genital valves and are turned somewhat in-
wards, the tips of the two spines (Sp) lying close to the con-
cave margin of the ovipositor and possibly braced against
the sides in life. The appendages of the female and the
styli apparently take no part in copulation.

Comparison of copulatory position with that of other
Odonata—Aeshna is the only genus of Anisoptera, so far as
known, the male of which grasps the prothorax of the female
during copulation, although many, if not all, of the other
Aeshnine genera doubtless resemble Aeshna in this respect.
In many of the Zygoptera, as Williamson has shown, the
anterior surface of the hind lobe of the prothorax is grasped
by the inferior appendages, but there is nothing comparable
in the mechanism to that which obtains in Aeshna. Of
the other subfamilies of Anisoptera, the head alone of the
female is grasped by the male appendages in the Gomphinae
and Libellulinae.

I have a pair of Gomphus spicatus taken at Go Home
Bay, Georgian Bay, in July, 1907, which well illustrates the
method in the former group. This pair which was resting
on the ground was captured by carefully placing the mouth

4

42 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

of a large net over it. The position was observed and the
pair was partially anaesthetized with chloroform and then
placed in a cyanide bottle. Unfortunately in carrying them
they came apart but the appendages of the male were easily
replaced, for the large ventral tooth on each superior appen-
dage had made a deep depression on the rear of the head of
the female. It will be seen from the figure of this specimen
(pl.2, fig. 7) thatthe superior appendages have no connection
with the pronotum. The occiput of the female is held be-
tween the two pairs of appendages at their bases, while the
curved apices on the forked inferior appendage are applied to
the head between the frons and the frontal vesicle.

It will be noticed that in the Gomphinae and Libel-
lulinae the upper surface of the superior appendages is smooth
and rounded while the under surface is commonly provided
with toothlike projections or denticles, while in the Aeshninae
it is the upper (supero-internal) surface which shows the great-
er amount of structural differentiation. In the Cordule-
gasterinae, the general form of the male appendages resembles
that of the Gomphinae and the superior pair probably also
rests upon the rear of the head of the female during copulation.

The nearest approach to the Aeshnine method of clasp-
ing thus far known is that of Petalura gigantea which has been
described by Tillyard (’09). In this species the supero-
internal surface of the superior appendages of the male were
applied to the shoulders of the female, while the inferior ap-
pendage is pressed down upon the occiput.

Oviposition—Females begin to oviposit soon after be-
coming fully mature and apparently continue to do so from
time to time throughout the rest of their life. I have found
Ae. canadensis, a species which does not often appear before
the last week of June, ovipositing at Go Home Bay, Georgian
Bay on July 17, 1907, and I have found the same species ovi-
positing at the end of August. Ovipositing females are
most frequentiy seen in August and early September.

I have watched the process of oviposition at close range
in the case of two species of Aeshna, Ae. eremita and Ae.
constricta. The former was observed in a small reedy bay
on the Nipigon River on Aug. 5, 1910. The insect when first

WALKER: NorRTH AMERICAN SPECIES OF AESHNA 43

seen was clinging to the blade of a bur-reed (Sparganium)
close to the surface of the water with about half the abdomen
immersed (pl. 3, fig.1). She was watched for about five minutes
at the end of which she suddenly flew away. During this
time the end of the abdomen was thrust against the stem every
two or three seconds and was gradually lowered until wholly
under water. Soon after this depth was reached she flew
away.

I then examined the reed and found the punctures made
by the ovipositor arranged as shown on pl. 3, fig. 2. They were
confined to the two narrow surfaces of the three-cornered
reed and the great majority were on one surface. The upper-
most were found about 1.5 cm. above the surface of the water,
the lowermost about 4 cm. beneath the surface. It will
be seen that they are not arranged in a double row as de-
scribed by Needham (’o1) for Basiaeschna janata, but tend
to be grouped in oblique rows, although the arrangement is
by no means regular. A few eggs lying near the edge of the
reed were exposed by carefully picking away the tissues on
one side of them. They were found to lie a little beneath
the surface and very obliquely placed, occupying more nearly
a vertical than a horizontal position (fig.3). The pointed an-
terior ends were in all cases outermost.

On the following day another ovipositing female of
the same species was observed at still closer range, in fact
I was almost directly above the insect while watching her.
She was supported on a collection of dead floating reeds
among a thick growth of living ones and was thrusting the
ovipositor, seemingly at random, into any piece of reed within
reach. She did not remain more than a minute or so, but
the lever-like thrusts of the ovipositor could be seen dis-
tinctly when the abdomen was turned sideways. In making
these thrusts the terminal abdominal segment served as the
fulcrum (fig. 1).

Many other individuals were seen ovipositing in the
same way in both living and dead floating reeds and it ap-
peared that in the latter case the thrusts of the insect were
always irregular, as though the insect were not satisfied with
the site for her operations.

44 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

In the case of Aeshna constricta,which I observed ovi-
positing only once, though at close quarters, the female
was seen to alight upon a sweet-flag (Acorus calamus L.)
growing at the edge of a broad sluggish creek near De Grassi
Point, Lake Simcoe, Although about two and one-half feet
above the water she at once began to oviposit, curving her
abdomen and using her large ovipositor apparently in the
same way as Ae. eremita but spending a longer time, four or
five seconds, over the deposition of each egg. Instead of
lowering the abdomen during the operation, she gradually
climbed upwards and seemed little disturbed by the very
strong breeze that was blowing. When she had climbed
about six inches and was not far from the end of the flag she
flew away.

On first examining the seat of her operations I could
find nothing to show that she had been there, but on a closer
inspection longitudinal slits in the epidermis were found
along the whole path of the operations. These were each a
few millimetres long, the length varying considerably, and
arranged somewhat irregularly in two rows (pl. 3, fig.5). An
egg was placed longitudinally under each incision, the an-
terior end upward. They lay rather loosely in the slit in
the parenchyma. The flag was taken home and the lower
end placed in a jar of water but it gradually withered and
the eggs did not hatch that season and were finally destroyed
by an accident.

Although the slits did not open with the drying of the
flag it is probable that they do so under natural conditions
and thus allow the eggs to drop into the water before hatching.
The unusually large size of the ovipositor in Ae. constricta
seems to be related to the practice of making elongate in-
cisions instead of mere punctures in the tissues of the plants.
This method of ovipositing high above the surface of the
water is unique as far as I know among the Anisoptera but
has been described by Needham (’00) in the case of the Zy-
gopterous species, Lestes unguiculatus, and L. uncatus.

A number of other species of Aeshna were also seen
ovipositing, viz., Ae. juncea, subarctica, interrupta, canadensis
and wmbrosa, and the process was in general the same as that

WALKER: NorTH AMERICAN SPECIES OF AESHNA 45

of Ae. ervemita, as nearly as could be judged, but although
reeds or reed-like plants are usually chosen by the female for
this purpose it is no invariable rule. Thus Ae. canadensis
usually oviposits in the ordinary way in aquatic plants but I
have seen it at Grenadier Pond, Toronto, performing this
function on the side of an alga-covered log, just below the
water-line. The eggs were apparently distributed quite ir-
regularly. I have also observed the same species apparently
utilizing for this purpose the fine wet sand on the shore of
Obabika Lake, near the mouth of Obabika Creek, Temagami
District, Ontario. There is a stretch of open marsh here
covered largely with Equisetum and coarse sedges, which,
earlier in the season had evidently been under water for
some distance back from the shore and the sand, especially
within a few yards of the water’s edge, was beaten hard by the
waves. It was this fine wet sand some yards back from the
water-line into which the dragonfly was thrusting her ab-
domen, as though in the act of ovipositing.

A similar act was also observed in the case of Ae. um-
brosa, a female of which was observed ovipositing in the mud
of a partly dried up bed of a short creek connecting Lakes
Temagami and Obabika. Although pools of water were
near at hand the insect chose the wet mud of the stream-
bed, clinging to the base of a sedge-stalk and thrusting the
end of the abdomen into the mud in the direction of the stem
and moving it irregularly to right and left but apparently
not touching the stem with the ovipositor. While the fe-
male was thus engaged, a male swooped down upon her and
the pair were thus captured.

Miss Wadsworth sent me a female of this species which
she observed ovipositing on the nearly dead trunk of an alder
shrub in the middle of a stream. She observed the insect
light on the trunk and ‘‘press the eggs into the wet bark of
the alder. Some were placed about half an inch from the
water line and from there down to the water line, and a very
few just below. A male soon came to her and as they flew
I got the net over both but the male escaped.”

Sometimes a female Aeshna may be seen tapping the
apices of the genital valves against a rock or other hard

46 WALKER: NorTH AMERICAN SPECIES OF AESHNA

substance, apparently adjusting their position in relation to
the ovipositor. I have observed this performance in the
case of two species, Ae. umbrosa and Ae. eremita.

While ovipositing the female Aeshna is never accom-
panied by the male as sometimes occurs in the case of Anax
junius.

The egg—The eggs of Aeshna (pl.3, figs. 4 and 6) are
elongate, smooth, cylindrical, rounded at the posterior and
pointed at the anterior end. They are pale yellow when first
deposited but may darken considerably before the nymph is
hatched (Ae. eremita). The size and exact form of the ma-
ture egg varies according to the species, e.g., in Ae. eremtta
they are 1.7-1.8 mm. long and about five times as long as
broad; in Ae. constricta they are about the same size and
general form but aresomewhat less acute at the anterior end;
in Ae. canadensis they are about 1.6-1.66 mm. long, and
slightly more acute than in eremita; while in Ae. sitchensis
the length is only about 1.1 mm., the breadth proportionately
somewhat greater than in the other species, and the anterior
ends but little pointed.

The ovaries of a female Ae. umbrosa, taken at Lake
Simcoe on Aug. 29, 1909, were found to contain 839 eggs,
nearly all of which were full-grown.

Probable number of nymphal stages—With a sufficiency
of material it is not a difficult matter to determine the number
of stages in the nymphal life after the wing-buds have begun
to appear, as with each ecdysis there is a very definite in-
crease in the relative size of these organs. Of the stages
preceding the first appearance of the wing-buds I have no
knowledge.

The material I have obtained by collecting nymphs of
Aeshna canadensis and Ae. umbrosa is sufficient for the de-
termination of the number of ecdyses that the nymph goes
through after attaining a length of about 1 cm. The results
thus obtained were partially confirmed by rearing nymphs of
various sizes through several ecdyses and comparing the
exuviae with the collected material. Nymphs of various
sizes of Ae. eremita, interrupta, juncea, californica and mul-

WALKER: NORTH AMERICAN SPECIES OF AESHNA 47

ticolor, as well as the two species previously mentioned can
in all cases be referred to a particular stage.

The earliest stage represented in the material at hand
is a nymph of Ae. canadensis, measuring 10.5 mm. in length.
The wing-cases are barely indicated by a pair of minute buds
and the antennz are five-jointed. Beginning with this stage
the nymph apparently moults eight times before emerging as
the adult insect, there being eight instars including the full-
grown nymph. Judging by the size of the egg and the rel-
ative sizes of the known instars it appears probable that
there are three or four ecdyses in addition to those observed,
making a probable total of twelve or thirteen stages. This
would make the number about the same as in the Agrionidae,
according to Balfour-Browne (’09), who found 10-14 stages in
various British species, the number varying considerably
among individuals of the same species. It is possible that
the number varies also in Aeshna but I am inclined to be-
lieve that the number of stages in a given species is constant.

Changes at the various ecdyses—As | have not been able
to determine the number of stages with certainty I have desig-
nated the various known ones as A, B, C,..H, stage H being
the mature nymph. Owing to lack of material in the earlier
stages I shall not describe in detail the changes which appear
in the nymph at each ecdysis. I shall do little more than
indicate some of the external characters by which each in-
star may be recognized. The younger instars among those
represented in the material at hand differ from the mature
nymph chiefly in the somewhat greater relative size of the
eyes, which are less prolonged mesad, the smoother surface
of the interocular region, the relative lengths of the antennal
segments, the shorter and somewhat narrower thorax, and,
of course, the smaller genitalia and shorter wing-cases, when
these are present. The colour changes also with the growth,
the younger instars being darker than the older ones. In-
stars A and B of Ae. canadensis and umbrosa, C, D and some-
times E of Ae. constricta are nearly black, except for a short
time after each month, but in the succeeding stages the
colours of the full-grown nymph are rapidly assumed.

Stage A. Antenne 5-jointed, terminal joint very long,

48 WALKER: NORTH AMERICAN SPECIES OF AESHNA

wing-cases very minute buds, probably appearing for the
first time at this stage.

Stage B. Antenne 6-jointed, terminal joint long,
wing-cases still small and tubercle-like (pl. 4, fig. 1).

Stage C. Antenne 7-jointed, the terminal joint of pre-
ceding stage having divided into two. Wing-cases in the
form of small triangular flaps, the posterior pair reaching
back about as far as the posterior margin of the metathorax
(fig. 2).

Stage D. Antenne 7-jointed, wing-cases still widely
separated mesially, the hind pair extending to about the
middle of seg. 1 of the abdomen, their front margins meeting
the hind margins of the front pair but not overlapping the
latter (fig. 3).

Stage E. Front wing-cases though still separated mesi-
ally are much more closely approximated and their hind
margins are overlapped by the hind pair, which reaches about
as far back as the apical margin of seg. I (fig. 4).

Stage Ff. Front wing-cases attingent and about half
covered by the hind pair, which reaches nearly to the apical
margin of seg. 2, or to the middle of seg. 3, and a little beyond
the middle of the hind femora (fig. 5).

Stage G. Hind wing-cases reaching the base or apical
margin of seg. 3 and nearly to the ends of the hind femora
(fig. 6).

Stage H (full- grown). Hind wing-cases generally
reach as far back as the middle or apical margin of seg. 4
and extend a little beyond the ends of the hind femora (pls.
5 and 6).

The above characters are based chiefly on the nymphs of
Ae. umbrosa and Ae. canadensis but they apply equally well
to those of all the other species seen except Ae. constricta.
Two nymphs of this species apparently belonging to stage D,
but with slightly longer wing-cases than is characteristic
of that stage in other species were found in a creek near De
Grassi Point, Lake Simcoe, in early July, 1910. After the
next moult which occurred about a week later the nymphs
were intermediate in wing-length between stages 9 and 10,
and after another moult, in August, they had all the appear-

WALKER: NoRTH AMERICAN SPECIES OF AESHNA 49

ance of full-grown nymphs, except that they were a little
below normal size for full-grown nymphs of this species.
One of them died just after the moult, but the other lived
until Dec. 24, when it died apparently from a too rapid
change of temperature in the aquarium. The genitalia in this
specimen, which is a female, are fully developed, extending
beyond the apical margin of segment 10, and the wing-cases
reach a considerable distance beyond the apices of the hind
femora. I am unable to say at present whether these pecu-
liarities are characteristic of this species or were the result
of abnormal conditions under which the nymphs were reared.

Length of nymphal life—Our knowledge of the nymphal
life of Aeshna is not sufficient to warrant a positive state-
ment as to its length, which indeed is not necessarily the
same in all species, nor in every locality. It is, however,
exceedingly probable that in southern Canada and the north-
ern United States three years is the normal length of life of
Ae. canadensis and Ae. umbrosa, if not of all the species found
there.

It seems to be a general rule that stage F is entered
upon about a year before the imago emerges. Of thirty-
seven nymphs of Ae. multicolor collected by Mr. C. H. Ken-
nedy at Sunnyside, Wash., in May, about the time when the
imagoes were emerging, eleven were mature, twenty-four
belonged to stage F and two to stage E. As the imagoes all
emerge within a period of a few weeks the nymphs belonging
to the latter two stages certainly remain over until the follow-
ing year. Stage F of Ae. canadensis is the oldest stage
that I have found in late summer after the time of emer-
gence of this species and I have never found it in the early
half of the season. I succeeded in rearing one individual
taken during this stage. It moulted in the fall and became
mature in the spring (the dates were lost), emerging on June
25. Six nymphs of this species taken at Go Home Bay,
Ontario, on Aug. 14, 1908, all belong to this stage, while of
six others taken on July 29, at nearly the same locality, four
belonged to stage D and two to stage E. Six nymphs of
Ae. canadensis taken at De Grassi Point on Sept. 6, 1910, be-
long to stage C. Two of these soon moulted but all entered

50 WALKER: NORTH AMERICAN SPECIES OF AESHNA

the winter either as stage C or D. Finally I have taken
stages A and B on July 27, 1910, and one example of stage
C on July 31.

These facts seem to show that stages E and F and often
D are passed through during the season before that in which
the imago emerges, and that A, B and C belong to a still
earlier season. The remaining early stages doubtless be-
long to the early part of the same season, as I have never been
able to find any trace of them in the summer and autumn, in
which practically all my nymph-collecting has been done.
Thus it seems very probable that in Ae. canadensis the imago
appears in the third season after the egg is deposited. This
conclusion is, of course, open to doubt, and is offered merely
as a basis for future work. And even if true in the case of
canadensis it is not necessarily so for other species, as the
following observations seem to indicate. On Aug. 8, I9gI0,
[ took ten nymphs of Ae. eremita from a marsh on the Nipigon
River, Ontario. At this time the imagoes were flying in
numbers, the great majority fully mature and some worn,
though a few tenerals were still to be seen. Of these nymphs
two were mature, three belonged to stage F, two to stage E
and three to stage B. The mature nymphs were taken home
alive and I expected them to emerge within a fortnight at
the latest, but they did not do so that season and are still alive
at the time of writing, Jan. 25, 1911. Thus it is possible
that an additional year is required by this species to complete
its life-history, at least in the northern locality where these
nymphs were found.

Habitat of nymph—Although fresh water is probably
the normal habitat of the nymphs of all the North American
species of Aeshna, Ae. californica has been found in brackish
water at Victoria, B.C. (Osburn ’06), and Mr. C. H. Kennedy
has taken the adults of Ae. interrupta interna and Ae. palmata
at an alkaline slough near Baker City, Oregon, where it is
possible they were bred.

Nymphs of Aeshna are unable to withstand freezing
even for a short time and are killed by rapid changes in the
temperature of the water. On July 29, 1908, I found a number
of young nymphs of Ae. canadensis (stages D and E) ina

WALKER: NORTH AMERICAN SPECIES OF AESHNA 51

very shallow pond with a sandy bottom. The water over
many square yards was not more than about three inches deep
and had been heated by the sun to such a degree that it was
almost uncomfortable to step into it with bare feet. The
temperature must have been at least 38° C. Nymphs of
Ischnura verticalis Say and Nehallennia irene Hagen were also
common here and all were quite as active as usual. I brought
a number of the Aeshna nymphs to the laboratory and left
them overnight in a jar of water, but all were dead the next
morning.

The nymphs of Aeshna and the Aeshninae in general
are climbers, living among reeds and rushes in still waters
usually a few inches to one-and-one-half feet deep. Still
weedy waters with a soft muddy bottom grown up with
reeds, bur-reeds (Sparganium), sweet-flag (Acorus), horse-
tails (Equisetum), coarses edges, etc., withan abundance of true
aquatic plants, such as water-milfoil and pondweed, and of
aquatic life in general, offer the best conditions for their ex-
istence. Reed beds growing in water that is subject to any
considerable wave-action may be searched in vain for any
species of this genus, nor have any North American species
been found in rapid streams where the allied genus Boyerza
sometimes occurs. Only one species of Aeshna, Ae. umbrosa,
has been found in shady woodland streams free from all reeds
or reed-like plants, although this species seems to prefer a
certain quantity of such growth. It never occurs in the
open reed-like beds frequented by the majority of the species
of this genus.

Habits of nymph — The nymphs cling to the reeds,
lying in wait for their prey. In captivity they are usually
almost motionless, except when attracted by some moving
object. Some nymphs of Aeshna which | kept in glass jars
placed upon a window sill overlooking a lawn were often at-
tracted by white clothes hanging from a clothes-line, when
moved by the wind. They usually rest with the head
downwards and in insufficiently aerated aquaria the tip of
the abdomen is kept at the surface with the valves open.
The opposite position is, however, often assumed, particu-
larly when the time for emergence is near at hand. Nymphs

52 WALKER: NORTH AMERICAN SPECIES OF AESHNA

of Ae. cyanea, which I found in abundance in small pools at
Tiibingen, Wiirttemburg,and Wiesen, Bohemia, were more
frequently seen resting upon the bottom with the abdomen
tilted slightly upwards than clinging to the reeds, though this
is also a common habit. They were quite often observed to
swim short distances with their characteristic straight jerky
movements. While swimming all three pairs of legs are
directed backwards, close to the body.

The food of the nymph of Aeshna is similar to that of
other large Odonate nymphs. Tadpoles, nymphs of smaller
dragonflies, such as Leucorrhinia, Sympetrum, Lestes, or
the smaller nymphs of their own species, nymphs of mayflies,
Notonecta, larvae of aquatic beetles, crustaceans such as Gam-
marus and Asellus, leeches, etc. According to Needham
(03a) the nymphs of Ae. constricta (=Ae. umbrosa) will eat
young brook-trout as long as themselves. This was demon-
strated by confining them together in a breeding-cage. The
trout disappeared one by one until all had been eaten.

In aquaria I have found small earthworms convenient for
feeding the older nymphs. The larger earthworms are avoided
by Aeshna. Leeches are still better, where they are easily
obtained, as they remain alive a long time in the aquarium.
A leech, however, will often succeed in slipping away from
the nymph after the latter has apparently succeeded in obtain-
ing a firm hold of his prey. I have also fed Aeshna nymphs
on flies, water-snails removed from their shells and suspended
upon a thread in front of the insects, and on the larvae of
the larch saw-fly (Nematus erichsonit), which has at times
proved very convenient on account of its abundance. |
tried also the larvae of Leconte’s saw-fly (Lophyrus lecontet)
which occurred in large numbers on young white pines at
De Grassi Point, but although seized by the nymph they were
at once rejected. Probably the strong odour of turpentine
which they give forth when crushed was disagreeable.

In the words of Professor Needham, ‘‘the Aeshna nymph
approaches its prey with the slowness and poise and stealth
of a cat till within striking distance.’’ Those who have
watched these insects in captivity will at once recognize the
aptness of this comparison.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 53

Considerable difference may be observed between dif-
ferent species of Aeshna in the boldness and voracity of their
nymphs. Those of the cyanea group are distinctly less shy
than those of the clepsydra group that I have observed.
None of our species as far as I know is as voracious as Ae.
cyanea which I reared in Germany, though Ae. umbrosa
is very like it both in appearance and habits. Ae. cyanea
would seize almost any object thrust into the water in the
breeding-jar, such as a twig or even one’s finger and would
attack earthworms of a size that would certainly be shunned
by Ae. umbrosa. According to Stefanelli (82) this European
species sometimes comes out of the water at night and de-
vours the newly emerged imagoes of its own species. Wil-
liamson has observed a nymph of Ae. constricta (umbrosa?)
kill and partly consume within an hour seven toad tadpoles
measuring about 13 mm. in length. I have seen one feeding
upon an earthworm for over half an hour. The younger
nymphs are eaten readily by the older nymphs of the same
or other species of Aeshna or by other large Aeshnine nymphs,
such as Anax junius. According to Needham (loc. cit.) nymphs
of Ae. umbrosa are eaten by brook-trout.

Symbiosis of nymph and green alga—Kammerer (’07)
has described an interesting case of symbiosis between the
nymph of Aeshna cyanea and the green alga Oedogonium un-
dulatum (Alex. Braun). The nymphs were all found in a
small pool near St. Margaretenbad, Bohemia, which had been
used for a number of years by the washerwomen of the
neighbourhood. All the nymphs of Ae. cyanea in the pool
supported a growth of the alga, while none of those found in
neighbouring pools not disturbed by the washerwomen were
thus affected. A series of experiments showed conclusively
that the association was a true case of symbiosis and that the
alga-covered nymphs unlike those not associated with the
plant were able to live in water strongly contaminated with
various impurities, such as soap, their own excreta, etc.,
by reason of the more abundant supply of oxygen provided
by the algae. They were also more resistant towards certain
ectoparasites, such as Saprolegnia, which cannot live in an

54 WALKER: NORTH AMERICAN SPECIES OF AESHNA

atmosphere rich in oxygen. Thirdly, the growth of algae
serves to hide the nymph from its enemies.

On the other hand the alga has the benefit of free
transport from place to place, which favours the processes of
assimilation, and is able to obtain a richer food-supply from
the fecal matter of the nymph and the mud, in which it fre-
quently burrows. The author also suggests that the spines ©
and sharp corners of the chitinous sclerites of the nymph offer
favourable points of attachment for the alga and that the
latter is protected from many animals which ordinarily feed
upon its filaments.

It was also shown by experiments that the association
between these two species could be brought about artificially,
using individuals originally free-living and independent of
one another, but that, with the exception of a temporary
attachment to the cyanea nymph of the alga Oedogonium cap-
illare no similar connections could be established between
other aquatic larvae and algae. Anax and Libellula nymphs,
for example, were refused as hosts by Oedogonium undulatum.

No similar associations have been recorded from among
North American Odonata and it is somewhat improbable that
they will be found under such circumstances as those attend-
ing the case described by Kammerer.

Emergence of the imago—For a week or more prior to the
time of emergence the nymph is sluggish and takes no food,
and a few days before this event takes place certain changes in
its appearance are noticeable. The eyes of the developing
imago can be distinctly seen through the integument of the
head, nearly meeting at the middle line and extending farther
back than those of the nymph. As the time for transfor-
mation approaches, the eyes become mesially attingent and
the hind margins move backwards until they nearly reach
the hindmarginofthe head. While these changes are taking
place the labium is acquiring the form which it assumes in
the imago and becomes entirely withdrawn from the larval
cuticle, which appears quite empty shortly before the imago
begins to emerge. During the last few days before emer-
gence the nymph rests at the surface of the water with the
head partly exposed.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 55

Judging from the very few occasions on which I have
observed the emergence of Aeshna or the freshly transformed
imagoes, it would seem that this process usually takes place
late in the afternoon or in the evening. On August
15, 1908, I witnessed the entire process in the case of
Ae. canadensis, a nymph of which I had been keeping in a
breeding-jar in my room at De Grassi Point. At 10 p.m. I
found that the nymph had crawled up a reed which had been
placed for the purpose in the jar. It was clinging to the reed
with the body in a vertical position (pl.5,fig.1). It remained
almost motionless for nearly ten minutes and then began to
execute slight jerky movements of the abdomen and lifting
movements of the wing-pads. These were accompanied by
a bulging of the thorax and head, the integument of which al-
most immediately split in the usual way, i.e., in an arc across
the eyes, along the mid-dorsal line of the back part of the
head and thorax to the bases of the wing-cases, and along the
upper edges of the thoracic pleura. In five minutes from the
time when the abdominal movements commenced the head
was free, and immediately afterwards the first pair of legs,
followed rapidly by the second pair, these movements being
accompanied by a backward curving of the body. The ex-
traction of the wings and then the third pair of legs followed in
less than a minute and at 10.16 p.m. the imago was hanging
back in the position shown in fig. 5, the accessory genitalia
plainly exposed and projecting, and only the last four ab-
dominal segments not yet free. In this position the insect
rested almost motionless for eleven minutes, the only notice-
able change during this time being the very slight elongation
of the wings. Then the legs began to move irregularly and
at intervals but the position remained the same for seven
minutes longer. At 10.34 p.m. there was a sudden convul-
sive movement of the whole abdomen and slight movements
of the legs, and ten minutes later the insect suddenly bent
forward, grasped the anterior part of the exuvia with the legs
and quickly withdrew the rest of the abdomen (figs. 6 and 7).
The abdomen now measured 32 mm. and the wings Io mm.
The colour was pale greenish grey, the lateral thoracic bands
faintly distinguishable. In the next twelve minutes the

56 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

wings became fully expanded but the abdomen remained very
little longer than before (fig. 8). Bobbing movements of the
abdomen ensued, followed by a gradual clearing of the wings.
Rhythmical up-and-down movements of the whole body and
irregular telescopic movements of the abdomen were next
observed. At 11.30 the abdomen had not quite reached its
full length and was still stout. The colour was practically
unchanged.

Exuviae of Aeshna nymphs are, as a rule, difficult to
find. As most of our species transform on reeds or other
plants growing closely together in shallow waters they are
not at all conspicuous and are very easily overlooked. They
should be sought during the period of emergence, or soon
afterwards, for they are soon blown into the water by the
wind. They are generally found clinging to reeds a few inches
above the water. Exuviae of Ae. umbrosa are sometimes
found on stumps or logs at the water’s edge.

THE NortH AMERICAN SPECIES OF AESHNA.
General Characteristics.

Although the number of described North American
species of Aeshna is somewhat greater than that of the
Palearctic species, the latter are more diversified in size,
structure and colour-pattern and represent a larger number
of groups. All of the North American groups except those
of californica and multicolor occur also in the Palearctic
region, where in addition to these the groups of grandis,!
mixta,? tsoceles* and melanicterat are represented. The
North American species are remarkably uniform in size,
form and coloration, the males of nearly all the species having
the abdomen brownish black, spotted with blue, no such
variety of colour-pattern occurring as in the Palearctic
species grandis, viridis, 1soceles, melanictera and cyanea.

In both regions the greater number of species inhabit
the cooler parts of the temperate zone and among these are

1 Including de. grandis L. and Ae. viridis Eversm.

* Including Ae. mixta Latr. and Ae. affinis van der Linden.

* Including Ae. tsoceles Miiller and Ae. martini Selys, and possibly others.
* Including this species alone (?)

WALKER: NortH AMERICAN SPECIES OF AESHNA 57

some of the most characteristic dragonflies of the subarctic
and arctic regions. A number of the North American species,
including two groups, are entirely restricted to the boreal
region, where they are a dominant group, both in species and
individuals.

Key to the North American Species of Aeshna.
I. MALEs.

A. Anal triangle 2-celled; spines of anterior lamina
directed ventrad.

On

B. Line of contact of the eyes little or no longer than
the occiput; lateral thoracic bands less than 1 mm. broad,
the first band sigmoid or bent twice at alternate
angles; hamular processes large, broad and elevated
behind, acute in front, their inner margins straight and
attingent or approximated ; hamular folds small, more
or less concealed by the hamular processes with which
they are connected by a low ridge... .. (caerulea group).
C. Preocular band not produced forward on each
side behind the T-spot; distance from hind margin
of occiput to frontal vesicle less than 2 mm.; spines
of anterior lamina not longer than the hamular pro-
cess, stout, straight, bluntly pointed; MD large,
isteoular .Or suqUadrates fue he Masala. leo ticle valierh «
ERG AE RST 9 FRR a eee caerulea septentrionalts Burm.
CC. Preocular band produced forward on each
side of the stem of the T-spot; distance from hind
margin of occiput to frontal vesicle about 2.5 mm.;
spines of anterior lamina longer than the hamular
process, curved ventrad and tapering to a slender
point; MD of ordinary size, triangular.. . sttchensts Hag.
BB. Line of contact of eyes distinctly longer than the
occiput; lateral thoracic band generally more than I mm.
broad, when narrower never sigmoid; hamular processes
never with straight closely approximated inner margins;
hamular folds (except in Ae. subarctica) not at all con-
cealed by the hamular processes.
D. Hamular processes long and slender, separated
from the hamular folds, which are more or less reduced;
spines of anterior lamina very long, tapering to a fine

58 WALKER: NORTH AMERICAN SPECIES OF AESHNA

point; superior appendages more or less acute at apices,
the superior carina not denticulated; a black band on
the fronto-nasal suture.............. (juncea group).
E. Lateral thoracic bands broad, the margins
straight; spines of anterior lamina curved ventrad;
apices of hamular processes acute and somewhat

EE. Lateral thoracic bands broadly excavated in
front, the first one narrowed and slightly bent a
little above the middle; spines of anterior lamina
straight, very slender; hamular processes rounded
at apices, almost concealing the small hamular folds
in a perpendicular view......... subarctica Walk.
DD. Hamular processes short, continuous with the
hamular folds, which are well developed; spines of an-
terior lamina short and straight... (clepsydra group).
F. Superior appendages without a prominent basal
inferior tubercle; seg. 10 with a pair of pale dorsal
spots (PD) which are sometimes confluent.
G. Dorsal thoracic bands reduced to a pair of small
isolated spots (often absent in dried specimens) ;
lateral thoracic bands reduced, either narrow
and nearly straight or divided into an upper and
lower spot. Mtaarising normally beyond middle of
pterostigma; spines of anterior lamina very short
and blunt; superior carina of superior appendages
generally with a few small denticles; a black line
on the fronto-nasal suture....... interrupta Walk.
GG. Dorsal thoracic bands complete, expanded at
their upper ends ; lateral thoracic bands broad,
more or less excavated or sinuate in front; Mita
arising normally before the middle of the pteros-
tigma.
H. Superior appendages with a low sub-basal
inferior prominence, expanding almost sym-
metrically from the base, the inner margins
in dorsal view not sinuate; apices rounded,
normally without a terminal spine and not at

WALKER: NorTH AMERICAN SPECIES OF AESHNA 59

all decurved; superior carina rather strongly
elevated apically, where it bears 6 or 8 well-
marked denticles; a black line on the fronto-
nasal suture (sometimes absent in arctic
specimens) ; first lateral thoracic band strongly
constricted about the middle by the deep ex-
cavation of) the front margin.) 0. MWe
evil ara SAR EMRE NY UL CURB RH eremita Scudd.
HH. Superior appendages expanding unsym-
metrically from the base, their inner margins
sinuate in dorsal view; apices acute witha
terminal spine; superior carina moderately or
but little elevated apically.
I. A black line on the fronto-nasal suture;
a large pale triangular spot immediately
in front of the humeral suture below ;
lateral thoracic bands very broad, the up-
per end of the first band narrowed and bent
cephalad ; pale spots between the bands
very large, occupying the greater part of
the space between them; outer margin of
the superior appendages in lateral view
very slightly elevated apically, with 3-5
small denticles..:/07 2 o52.Nc4 clepsydra Say.
II. No black line on the fronto-nasal sut-
ure; no triangular spot in front of humeral
suture below ; first lateral thoracic band
bent caudad at the upper end ; spots be-
tween the lateral bands, when present,
small; outer margin of superior append-
ages in lateral view more or less upcurved,
apices decurved; superior carina moderate-
ly elevated apically.
J. Lateral thoracic bands blue or green,
the first generally green below, blue
above, its anterior margin almost rec-
tangularly sinuate ; superior carina of
superior appendages with a few denticles,
apices rather abruptly decurved; ham-

60

WALKER: NortTH AMERICAN SPECIES OF AESHNA

ular processes rather long, directed ceph-
alad, subparallel, the apices converg-
ent; PL typically represented on abd.
SES) GL A canadensis Walk.
JJ. Lateral thoracic bands yellowish
green, the anterior margin of the first
band obtusangularly sinuate; superior
carina of the superior appendages not
denticulated; apices usually gently de-
curved; hamular processes directed mesad
and ventrad, each consisting of a stout
proximal and a slender distal part; PL
typically represented on abd. segs. 2-4.
AVES CHM RL REDS an AIRY QOL verticalis Hag.
FF. Superior appendages with a prominent inferior
basal tubercle; superior carina moderately elevated
apically, not denticulated ; apices rounded with a
small projecting spine, not decurved ; lateral thoracic
bands broad, not excavated in front; abd. seg. 10
wholly black h) un ONi eae nn apt tuberculifera Walk.

AA. Anal triangle 3-celled; spines of anterior lamina curved

dorsad.

K. Abd. seg. I without a distinct ventral tubercle;
dorsum of seg. 10 smooth; superior appendages without
a well-developed superior carina but with an anteapical
internal hairy tubercle and an anteapical inferior spine;
apices ‘proadly ‘rounded iMG ih Zid ows (cyanea group).

L. A black line (rarely a fine brown line) on the fronto-
nasal suture; dorsal thoracic bands I mm. or less.
broad, tapering somewhat towards each end or some-
times suddenly expanded at the posterior end; lateral
thoracic bands nearly straight and equal; a distinct
greenish yellow lateral spot on seg. 1 ; lateral carinae
of seg. 7 in ventral view, slightly or not at all sinuate;
venter of abdomen and rear of head wholly black;
generally one cell between A2 and A2 at their origin.
OPO is) Sy RAM tet Asi Ses bt SEES SER palmata Hag.
LL. Without a black line on the fronto-nasal suture;

WALKER: NORTH AMERICAN SPECIES OF AESHNA 61

dorsal thoracic bands rapidly widened caudad so as
to be triangular in form with the base just in front of
the antealar sinus; generally two cells between Az
and A3 at their origin.
M. Rear of head in part fuscous; lateral thoracic
bands rather narrow (1 mm.); straight, not widening
above the middle, surrounded by a dark margin; a
pair of large pale basal spots on the ventral surfaces
of segs. 4, 5 and 6; lateral carina of seg. 7, viewed
ventrad, strongly sinuate in its anterior two-fifths.
NAVARA A ar Ait ng UOT AeA SA A RO LT Al umbrosa Walk.
MM. Rear of head wholly black; lateral thoracic
bands broader, not surrounded by a dark margin,
widening above the middle, margins of first band,
especially the anterior, sinuate; venter of abdomen
without pale spots; lateral carinae of seg. 7, viewed
ventrad, at most feebly sinuate.....constricta Say.
KK. Abd. seg. I with a distinct spinulose ventral tuber-
cle; dorsum of seg. 10 with median and submedian
tooth-like elevations; superior appendages with a well-
developed superior carina.
N. A black line on the fronto-nasal suture; superior
appendages without an anteapical inferior spine, the
apices obtusangulate or rounded, superior carina not
angulate or denticulate; inferior appendage somewhat
less than half as long as the superior pair.............
DES Ay INO RS rad UC MOS eR Sen CML ret californica Calv.
NN. No black line on the fronto-nasal suture; superior
appendages with an anteapical inferior spine, the
apices acute and decurved, superior carina angulate;
inferior appendages slightly more than half as long as
fhe SHMEeKION Aaumuys . 8 Ie eal elas (multicolor group).
O. Abdomen (excl. appendages) less than four
times as long as the thorax; a prominent ventral
tubercle on abd. seg. 1; inferior sub-basal tubercle
of the superior appendages at one-fourth to one-
fifth the length of the appendages; height of supe-
rior carina above outer margin, in profile, not less
than the depth of the appendage directly below it;

62 “\WALKER:;NORTH AMERICAN SPECIES OF AESHNA
distance from summit of superior carina to apex
of appendage greater than one-third of the length of
the appendage; outer side of anal loop in hind
wing longer than inner side of the triangle.........
Peay ales Ad Sk iPD teh Wnred LUD SA) Nec Ale Mapas dhe fd multicolor Hag.
OO. Abdomen (excl. appendages) not less, gener-
ally a little more, than four times as long as the
thorax; ventral tubercle on abd. seg. 1 but little
elevated; inferior sub-basal tubercle of the superior
appendages at one-sixth to one-seventh the length
of the appendage; height of superior carina above
outer margin, in profile, generally much less than
the depth of the appendage directly below it; dis-
tance from summit of superior carina to apex of
appendage about one-third of the length of the
appendage; outer side of anal loop in hind wing
about as long as the inner side of the triangle.
a aN tata eal en a wih aU Cl ae ta i8 a re A mutata Hag.

I]. FEMALES.
A. Abd. seg. 1 without a distinct ventral tubercle; fork
of Rs decidedly asymmetrical at base.

B. Line of contact of eyes little or no longer than occi-
put; lateral thoracic bands less than I mm. broad, the
first band sigmoid or bent twice at alternate angles.
Genital valves in ventral view tapering evenly to the

narrow attingent apices, lateral carinae prominent.
C. Distance from hind margin of occiput to frontal
vesicle less than 2 mm.; preocular band not produced
forward on each side above; appendages about as
long as segs.9+10, more slender proximally than dis-
tally, the apicesrounded or broadly and obscurely
perianal t)3) en caerulea septenirionalis Burm.
CC. Distance from hind margin of occiput to frontal
vesicle about 2mm.; preocular band produced forward
on each side of the stem of the T-spot; appendages
slightly longer than the dorsa of seg.9+10, tapering
equally at base and apex, the latter distinctly though
blimey pointed “si a ee ei sitchensis Hag.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 63

BB. Line of contact of eyes distinctly longer than oc-
ciput; lateral thoracic bands generally more than 1
mm. broad, when narrower never sigmoid.
D. Basal plate of ovipositor distinctly bilobate,
lateral genital plates absent; PL large but separate
from PD (sometimes narrowly connected on 2).
E. Lateral thoracic bands broad, not excavated in
front, first band gradually tapering to the upper
end, without a distinct posterior offshoot; genital
valves with a distinct ventral surface; appendages
shorter than abd. segs. 8+9 (5 mm. or less). ....
Ba ete setnaur sib) eM ogi seek. 1:2 SO 73 2-1 a juncea L.
EE. Lateral thoracic bands narrower, broadly ex-
cavated infront, first band gently widening above
the middle to the upperend from which a narrow
posterior offshoot is given off; genital valves with-
out a distinct ventral surface, appendages about
as long as abd. segs. 8+9(6-6.8 mm.)..........
ee ha at MN ie TEM ARES! 95.5 oH IDES subarctica Walk.
DD. Basal plate of ovipositor not bilobate, its hind
margin straight or slightly arcuate; lateral genital
plates present ; PL generally confluent with PD
(often separate in Ae. umbrosa).
F. Styli much shorter than the dorsum of abd. seg.
10; appendages less than 1.5 mm. broad, decidedly
slenderer in the basal than in the apical third;
broadest beyond the middle ; apices generally
rounded.
G. A black line on the fronto-nasal suture.
H. Genital valves 2-2.5 mm. long; apices not
elevated, bearing a very minute pencil of hairs.
I. Dorsal thoracic bands absent or repre-
sented each by a minute and usually ill-
defined spot; lateral thoracic bands reduced,
I mm.or lessin width, each band often di-
vided into an upper and lower spot ; Mia
normally arising beyond the middle of the
pterostigma. 22.6229... - interrupta Walk.
II. Dorsal thoracic bands distinct in well-
preserved specimens, complete, or the upper

64 WALKER: NORTH AMERICAN SPECIES OF AESHNA

end separated as a distinct spot; lateral

thoracic bands more than I mm. broad,

never divided but the first band deeply ex-

cavated in front; Mianormally arising in
front of the middle of the pterostigma.

J. Without a pale triangular spot in front

of the humeral suture; first lateral thoracic

band not bent forward at the upper end;

spots between lateral bands not unusually

large: ic. eat ee es eremita Scudd.

JJ. A large pale triangular spot in front

of the humeral suture below;; first lateral

thoracic band bent forward at upper

end; spots between lateral bands very

large, often confluent with the bands.

we io Pha d bled. ee ae ie clepsydra Say.

HH. Genital valves 3-3.5 mm. long, apices ele-

vated, without a terminal pencil of hairs; lateral

‘ thoracic bands about I mm. broad, straight, not

excavated in-dfuont. e002" 4eeie lick: palmata Hag.
GG. Without a black line on the fronto-nasal
suture.

K. Lateral thoracic bands green or yellowish
green (rarely blue) not margined with black, the
anterior margin of the first band distinctly sin-
uate, the second band elongate-triangular; geni-
tal valves 2-2.6 mm. long, apices not elevated,
bearing a minute pencil of hairs.
L. First lateral thoracic band with the anterior
margin almost rectangularly sinuate, much
narrowed about the middle, its upper end
giving off caudad a very narrow offshoot;
posterior (postero-inferior) margin of second
band generally curued ventrad at the upper
end; sulcation of ventral surfaces of genital
valves not distinctly delimited posteriorly,
appendages usually 4-6 mm. long.............
BET) sc a eet. canadensis Walk.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 65

LL. First lateral thoracic band with the anterior
margin obtusangularly sinuate,not much narrow-
ed about the middle, its upper end giving off
caudad a broad offshoot; posterior margin of
second lateral band straight; sulcation of ven-
tral surfaces of genital valves terminating
more or less abruptly some distance before the
apices; appendages 5.75-7 mm. long. .........
ba Wii Dat 3 5 Oa na RMR aes ale verticalis Hag.
KK. Lateral thoracic bands yellow, margined
with black or dark brown, about 1 mm. broad,
straight, the first band expanded a little below,
but not at all sinuate, the second band with the
margins subparallel; genital valves 2.7-3 mm. long,
apices slightly elevated, without a terminal pencil
ATES EO Vey Oh AA a aa Ben umbrosa Walk.
FF. Styli as long as the dorsum of abd. seg. 10
(1.5-2 mm.); appendages 2 mm. broad or nearly so;
basal third fully as broad as the apical, expanding
rapidly with convex margins so that the greatest
breadth is attained before the middle; apex gener-
aliy acute, never broadly rounded; no black line
on the fronto-nasal suture.
M. First lateral band with the anterior margin
straight or nearly so, gradually tapering dorsad and
not giving off a distinct posterior offshoot; second
lateral band not widened at the upper end; genital
valves 3.1-3.5 mm. long, apices broad in profile, each
bearing a minute pencil of hairs;styli 1.5 mm., spin-
ules on ventral surface of abd. seg. Io comparatively
femanihl eOaree yy 5 le eas tuberculifera Walk.
MM. First lateral band with the anterior margin
distinctly sinuate, the upper end giving off a
small posterior offshoot; second lateral band
rapidly widened at the upper end; genital valves
4-4.5 mm. long, apices slender in profile, without
a pencil of hairs; styli nearly I mm. long ; spinules
on ventral surface of abd. seg. 10 numerous, minute
SINE bes eee Es 2), hi.4. 0.) 0) mre Meets boy. constricta Say.

66 WALKER: NORTH AMERICAN SPECIES OF AESHNA

AA. Abd. seg. I with a distinct ventral tubercle; fork of
Rs nearly symmetrical at base; genital valves without dis-
tinct lateral carinae.
N. A black line on the fronto-nasal suture; supratri-
angle of fore wings clear or with a single cross-vein.
S12 Yves RAN Si eta aL Ae ey ie californica Calv.
NN. Without a black line on the fronto-nasal suture;
supratriangle of fore wings with 2-4 cross-veins.
©. Abdomen scarcely more than four times as long as
the thorax; a prominent ventral tubercle on abd. seg.
1; appendages somewhat shorter than segs. 8+9
(5-6 mm.), both margins arcuate, the curve of the inner
margin not much stronger than that of the outer;
outer side of anal loop of hind wings longer than inner ©
sidévof triangle) AAA On aa Ae multicolor Hag.
OO. Abdomen four and one half times as long as
thorax; ventral tubercle on abd. seg. 1 but little
elevated; appendages a little longer than segs. 8+9
(6.5-7.5 mm.); outer margin straight, inner margin
arcuate; outer side of anal loop of hind wings about
as long as inner side of triangle......... mutata Hag.

Key to the known Nymphs of North American Species of Aeshna.

A. Mentum of labium distinctly more than half as broad at
base as at apex; genitalia of female usually not reaching the
posterior margin of 9 and never extending beyond it.
B. Lateral spines present on abd. segs. 5-9; lateral lobes
of labium squarely truncate.
C. Postero-lateral angles of head bluntly angulate or
but, Hitde*reunded ) 00: 4en a eee eee eremita.
CC. Postero-lateral angles of head broadly rounded.
DE CO Reeee gee |. eR ee Cerone interrupta lineata?
BB. Lateral spines present on abd. segs. 6-9; postero-
lateral corners of head well rounded.
D. Femora dark with three pale annuli; abdomen
marked with irregular pale blotches on a darker
ground colour; lateral lobes of labium squarely trun-
cate, the outer apical angle scarcely rounded. .
Ng TE ee tated otc “o> lave, SRS a ele ee interrupta interrupta.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 67

DD. Femora uniform; abdomen without irregular
blotches, more or less distinctly longitudinally striped;
lateral lobes of labium, when truncate, with the outer
apical angle distinctly rounded off.
E. Lateral lobes of labium truncate; abdomen not
conspicuously striped; lateral spines on 6 rudi-
mentary, distant from the posterior margin of the
segment by at least four times their own length.
F. Punctae conspicuously marked with dark
brown, those of the lateral series distinctly larger
than the dorso-lateral; lateral spines on g not
quite reaching as far caudad as the middle of the
dorsum of 10; lateral appendages of female half
as long as the inferior pair.............. juncea.
FF. Punctae very inconspicuously marked, the
dorsal series being scarcely darker than the
ground colour, those of the lateral and dorso-
lateral series about equal in size; lateral spines on
g reaching as far caudad as the middle of the dor-
sum of 10; lateral appendages of female three-
fifths as long as the inferior pair....... subarctica.
EE. Lateral lobes of labium not truncate, but ex-
teriorly curving to a prominent internal terminal
hook; abdomen with conspicuous well-defined dor-
sal longitudinal bands; lateral spines on 6 well-
developed, distant from the posterior margin by
about their own length.
G. Lateral margins of mentum of labium in the
basal half not in the least degree arcuate, lateral
lobes terminating in a broadly curved hook; dark
median band on abdomen not deepened about the
MOrSsal DUNEEAG:. . Axscyinr amass tea ce ta clepsydra.
GG. Lateral margins of mentum of labium in the
basal half very feebly arcuate, lateral lobes
terminating in a very abruptly curved and al-
most truncate hook; dark median band on ab-
domen deepened about the dorsal punctae.
SPAR Te ae Kea hace aren aire aN te Ue. sda st OIOCH SES:

68 WALKER: NORTH AMERICAN SPECIES OF AESHNA

AA. Mentum of labium not, or scarcely more than half, as
broad at base as at apex.
H. Lateral lobes of labium tapering to a slender slightly
hooked point; legs uniformly coloured; genitalia of female
reaching caudad nearly or quite to the apical margin of
TQ. (550 9 BRE, LUO lath by he aie ce ae iy Ace constricta.
HH. Lateral lobes of labium squarely truncate; femora,
and generally the tibiae also,with alternate pale and dark
annuli; genitalia of female never reaching caudad to the
middle of ro.
I. Eyes very prominent, postocular part of head com-
paratively long with straight lateral margins,and pro-
minent, though rounded, postero-lateral angles; men-
tum of labium strongly narrowed in the proximal four-
sevenths, distal part with the lateral margins strong-
ly arcuate; genitalia of female reaching to or beyond
the posterior margin of 9.
J. Apical breadth of mentum of labium equal to
about four-fifths of its length, lateral lobes without
an internal apical tooth; supracoxal processes equal;
genitalia of female extending over the basal third of
L(! Pe PR PRT ELEC Ges ar ei oeeR ea Se palmata.
JJ. Apical breadth of labium equal to about three-
fourths of its length, lateral lobes with a small in-
ternal apical tooth; supracoxal processes unequal,
the posterior somewhat the larger; genitalia of
female extending barely to the posterior margin of
Gs RE Grey ce eae ug Dl oh vie Ren oct Re umbrosa.
II. Eyes but little prominent, postocular part of head
short with scarcely any development of the lateral
margin as distinct from the curve of the postero-
lateral angles; mentum of labium expanding almost
evenly, distal part with the lateral margins moderately
arcuate; genitalia of female not reaching the posterior
margin of 9.
K. Posterior supracoxal process distinctly longer
than the anterior; lateral spines on abd. seg. 6 dis-
tant from the posterior margin by I-2 times their
own length ; anterior wing-cases with II-I5 ante-

WALKER: NORTH AMERICAN SPECIES OF AESHNA 69

cubital veins; basal part of superior appendages
of male about as long as the basal breadth; tibiae
without, or with indistinct, annuli...... californica.
KK. Supracoxal processes subequal; lateral spines
on 6 rudimentary, distant from the posterior mar-
gin by at least four times their own length; anterior
wing-cases with 16-19 (rarely 15) antecubital veins;
basal part of superior appendages of male distinctly
longer than the basal breadth; tibiae with distinct
“LAU 01 6 Anema et PM ER ARAN RAORMAE HL as tr multicolor.

The caerulea group—A well circumscribed group of
boreal distribution, consisting of the circumpolar Ae. cae-
rulea with its two races, caerulea and septentrionalis, and the
North American Ae. sttchensis. Their origin and affinities
are obscure, their nearest allies being apparently the juncea
group, with which they share the long downwardly directed
spines of the anterior laminae and the reduced hamular folds.
In the feeble and undecided development of the anterior
branch of the fork of the radial sector, especially in caerulea,
and the long line of cross-veins between Rs and M2 the group
appears to be a primitive one, approaching the genera Oplon-
aeschna and Basiaeschna.

The juncea group—This is also a sharply defined group
of two boreal species, of which juncea is circumpolar, sub-
arctica North American. Their nearest affinities are with the
clepsydra group, to certain members of which they bear a
remarkably close resemblance, both in the adult and nymphal
state. The wide differences in the genitalia, however, show
that there is no very close relationship between the two groups.

The clepsydra group—Another circumboreal group of
eight species,! six of which occur in North America, two in
Asia and one in Europe. Their nearest allies are the Pale-
arctic species grandis and viridis, which differ chiefly in the
simpler and more generalized form of the male abdominal
appendages and the more specialized type of coloration.
The genitalia of both sexes are of precisely the same type and

‘Ae. nigrojflava Martin, from Japan, is probably also a member of this group
but as I have not seen it I prefer to leave it out.

70 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

the venational characters are also very similar in the two
groups.

The North American species of the clepsydra group are
for the most part remarkably alike and it seems probable that
they have diverged from a common ancestor at a compara-
tively recent period. Species are apparently still in the pro-
cess of evolution, as exemplified in the case of the geographical
races of Ae. interrupta. The most generalized member of the
group, however, is the Siberian Ae. crenata, in which the lateral
thoracic bands are broad and straight as in juncea and
grandis. This species closely resembles interrupta and ere-
mita in the accessory genitalia of the male, and the appendages
of this sex combine the characters of several of the American
forms such as eremita, interrupta and clepsydra. It would
thus seem very probable that the clepsydra group had a
common origin with the grandis group in the Palearctic re-
gion.

The cyanea group—This is a very well-marked group of
four species, remarkable for the form of the male appendages.
In the Palearctic Ae. cyanea thespecialization in these struc-
tures has not been carried quite so far as in the other forms
and the downwardly directed apical spine is clearly homolo-
gous with the pointed apices in the ordinary type of appen-
dage, while the hairy anteapical tubercle is less developed than
in the other species and appears to represent part of the su-
perior carina. Of the other three species palmata is in most
respects intermediate between umbrosa and constricta and
may be regarded as nearly representing the parental form
from which they have descended. This being so, it is an
interesting fact that palmata occurs in northeastern Asia as
well as western North America.

Thus it is probable that the cyanea group is also of
Palearctic origin, Ae. cyanea being the most primitive species
in the form of the male appendages, though most specialized
in coloration, while Ae. palmata is the most generalized of the
other three. In this connection it may be noted that Ae.
umbrosa has diverged from palmata more widely in the east
than in the west, the abdominal coloration being more similar
to that of palmata in British Columbia and the Western

WALKER: NORTH AMERICAN SPECIES OF AESHNA 7I

States, where both species occur together, than in eastern
North America where only umbrosa is found.

The californica groubp—The nearest allies of Ae. californica
are found in a group of species from South and Central
America and Mexico. This group as pointed out by Cal-
vert (’05b) consists of such species as Ae. diffinis, marchali,
cornigera, galapagoensts, bonariensis and confusa. Ae. haarupi
probably also belongs here. Not having studied many of
these forms closely, I am uncertain as to the limits of the
group, or whether it is sharply marked off from the following
group, to which it is certainly very closely allied. The nymphs
of californica and multicolor are remarkably similar.

The multicolor group—The peculiar form of the abdom-
inal appendages of the male seems to be the only character by
which this group differs from the preceding, of which it is
probably an offshoot. It is best developed in southern
Mexico, where Ae. dugesit, multicolor and jalapensis occur.
The only other species, Ae. mutata, is a northeastern offshoot,
occurring in the eastern half of the United States.

Besides the species of Aeshna described here, one other,
the Palearctic Ae. grandis L, has been once recorded from
North America by Hagen (’61), who observed a single in-
dividual at Bergen Hill, N.J. This specimen was believed to
have been introduced by a vessel. There is also a specimen
of grandis in the collection of the Ontario Agricultural Col-
lege, Guelph, labelled ‘‘London, Ont., J. M. Denton.”
Nothing, however is known of the history. of this specimen
and in the absence of such information I prefer to exclude
this species from the present account.

caerulea group
<< iuneea group
7 Nl Clepsydra group
ae “> grandis group
an g
* cyanea group
_ mixta group
iSoceles group

californica group
multicolor group

Fic. 8—Interrelationships of the various groups of deshna represented in North America and Europe.

72. WALKER: NorTH AMERICAN SPECIES OF AESHNA

DESCRIPTIONS OF SPECIES.
Aeshna caerulea septeatrionnlis Burmeister.

(Pl. 12, fig. 4; pl. 15, figs. 1, 1a; pl. 18, figs. 1, 1a; pl. 22,
figs. 1, 2) 3)

Aeschna septentrionalis, Burmeister, Handb. der Ent., II, 839 (1839) ; Hagen,
Syn. Neur. N.A., p. 120 (1861); Pr. Bost. Soc. Nat. Hist., XVIII, p. 33 (1875);
Kirby, Syn. Cat. Neur. Od., p. 87 (1890); Hagen, Psyche, V, p. 354-355 (1890);
Banks, Trans. Am. Ent. Soc., XIX, p. 353 (1892); Calvert, Trans,
Am. Ent. Soc. XXV, p. 54 (1898); Fyles, Ann. Rep. Ent. Soc. Ont.,
XXXI, p. 54 (1901); Williamson, Ent. News, XVII, p. 135 (1906) ; Martin,
Coll. Zool. Selys., XVIII, pp. 37-41, 79, figs. 35, 36 (1908); Gen. Ins., CXN, para
pl. 3, figs. 1, Ia. 1b (1911). }

Aeshna septentrionalis, Calvert, Occas. Pap. Bost. Soc. Nat. Hist., VII, 6, p.
23 (1905) ; Williamson, Ohio Nat., VII, p. 150 (1907); Walker, Can. Ent., XL.
pp. 386, 390, 451 (1908) ; Muttkowski, Cat. Od. N.A., p. 113 (1910).

A rather small, moderately stout species.

Male—Occiput pale yellowish, broader than long, as
long as, or somewhat longer than, the line of contact of the
eyes, which measures less than 1 mm. in length ; frontal
vesicle very large, about twice as broad as long, the front
margin straight or slightly emarginate, black with a median
yellow spot not reaching the lateral ocelli; eyes much narrowed
above and smaller than in other North American species;
T-spot with cross-bar very heavy, 2-2.5 mm., the stem very
short, about .5 mm. broad; preocular band not curved for-
ward on each side of the T-spot, but continued ventrad as a
marginal line to the bases of the lateral lobes of the nasus, ex-
panding considerably to the fronto-nasal suture, but narrow
on the nasus; a heavy black line on the fronto-nasal suture,
widened only at the outer ends; face dull yellowish or brown-
ish olivaceous with a pale submarginal band, short, the frons
being about four times, and the nasus three times, as broad
as long (vertical dimension) ; two transversely impressed areas
near the lower margin of the nasus not marked with black,
lateral lobes rotundo-rectangulate, feebly flaring; rhinarium
reddish brown;labrum greenish yellow with a black upper,
and generally a brownish lower, margin; labium dull yellow
stained with reddish brown peripherally; rear of head black.

Thorax somewhat more robust than in Ae. sitchensts,
dull olivaceous brown, darkened along the sutures and about
the lateral bands. Dorsal bands reduced to a pair of minute,

WALKER: NORTH AMERICAN SPECIES OF AESHNA 73

though distinct, elongate pale spots 1-1.75 mm. long. A pale
streak usually present along the lower half or more of the
humeral suture in front. Lateral thoracic bands narrow and
irregular, pale yellow (bluish above?). First band very
strongly sinuate, being twice bent, at alternate angles,
broadest in the middle and lower parts, but nowhere more
than I mm. and sometimes (in Labrador specimen) very
narrow and broken; a narrow posterior offshoot is given off
at the upper end. Second band nearly straight, narrowed
and sometimes divided at the middle, upper ends expanding
along the posterior margin of the thorax above and below.
Two small spots generally present between the lateral bands,
one at the metastigma, the other above it.

Legs rather pale reddish brown, darkened on the tarsi,
distal ends of the tibiae and under surface of the femora.

Abdomen about four times as long as thorax and barely
longer than the hind wing, considerably stouter than in
sttchensis, strongly constricted at 3, then rapidly expanding
and attaining nearly its full width at the apex of that segment;
base of 10 considerably narrower than the apical margin of 9.
Seg. 1 without a ventral tubercle; auricles with 4-6 teeth;
tergal margins of 2 approximated but not angular.

Spines of the anterior lamina rather stout, shorter than
in Ae. sitchensis, reaching caudad nearly to the posterior
margin of the hamular processes, and projecting a little below
the tergal margins, nearly straight, rather bluntly pointed,
directed ventrad, caudad and slightly mesad. Hamular pro-
cesses very large, their posterior margins broad, rounded and
strongly elevated, the apices bluntly acute, the inner margins
rather narrowly separated at base but approximated apically.
Outer surfaces strongly concave and almost vertical apically;
hamular folds very small and almost concealed in a direct
ventral view by the hamular processes with which they are
connected by a low ridge; closely approximated mesially.

Lateral carinae on 7 and 8 nearly straight; on 9 present,
as a rule, only on the apical half. Dorsum of 10 with a
prominent median basal tooth and a smaller one on each side
of it.

Superior appendages a little shorter than 9+ 10, moderate-

6

74 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

ly slender and slightly divergent in their basal fifth, where they
are separated by aspace 2.5 times their width, thence expand-
ing on the inner margins to the middle, where the breadth
is 1-5 to 1-4 times the length, tapering slightly in the distal
half to the well-rounded apices, which are without a terminal
tooth; outer margins moderately convex, inner margins sinuate
in the basal half. Superior carina obsolete or feeble in the
basal third, considerably elevated in the apical fourth and
strongly arcuate at apex, nearly smooth or with a few minute
elevated points. In profile the outer margin appears gently
upcurved or nearly straight but the appendage appears con-
siderably upcurved apically on account of the elevated
superior carina and arcuate inferior carina, which is formed by
the deflected inner margins. Basal part thickened, with a low
sub-basal inferior prominence. Inferior appendage triang-
ular, slightly less than one-half as long as the superiors, basal
breadth three-fifths to two-thirds of the length, sides barely
convex, apex very bluntly rounded, surmounted by a pair of
small recurved teeth. Upper margin in profile view very
gently curved, lower margin strongly curved, the sides very
broad near the base, tapering rapidly.

Abdomen brownish black, all the pale markings prob-
ably blue, varying considerably in size but always large.

Seg. 1. Dorsal spot apparently large, lateral spot vari-
able, sometimes I.5 mm. in the transverse dimension, some-
times obsolete (?Labrador specimen).

Seg. 2. AML very large, covering most or all of the lateral
surface back to the transverse carina, connected with or
narrowly separated from MD. PL and PD confluent, covering
about two-thirds of the posterior half of the segment, deeply
notched in front, broadly confluent above with its fellow and
with a median dorsal band, more than I mm. broad, which
runs cephalad nearly to the anterior margin.

Segs. 3-10. AL on 3-8, upper margins straight, no basal
offshoots ; on 3 covering almost the whole lateral surface in
front of the transverse carina. ML large, rounded poster-
iorly. MD very large, irregularly semi-elliptical or sub-
quadrate, and generally broadly confluent with ML except
on 8, where it is reduced to a mere dot; generally confluent

WALKER: NORTH AMERICAN SPECIES OF AESHNA 75

also with PD on some or all of the segments, the variation in
this regard being considerable. In the Labrador specimen
these spots are more widely separated than in the others.
PD on 3-10, varying in length from 1.5 mm. on 3 to 2mm. on
8, always well separated from its fellow, except sometimes on
10 where there may be a narrow connection next the posterior
margin; inner margins nearly straight on last 2 or 3 seg-
ments, elsewhere convex or somewhat indented and irregular.
PL usually on 3-8, on 3 and 4 only in the Labrador specimen,
connected with PD, straight and elongate where well-devel-
oped and separated from PD throughout, or narrowly con-
nected on 3 and sometimes 4; often also with ML on 3 and 4.
Wings of average size and usual form, the hind wing about
as long as the abdomen, hyaline; costal margin moderately
dark brown; pterostigma long, brown; membranule uniform
grey, extending caudad to cross-vein of the anal triangle,
which is 2-celled; one cell between A2 and A3 at
their origin. Upper branch of Rs following a somewhat
irregular course, not always distinct at its origin and, when
distinct here, coming off at a rather marked angle, at a point
opposite the first to the third postnodal cell before the pter-
ostigma ; 3 (rarely 2) rows of cells between the forks at the
level of the distal end of the stigma, and 3-4 rows of cells be-
tween the forks at the margin. Rspl nearly straight, sepa-
rated from Rs at the point of greatest divergence by 4 (rarely
5) rowsof cells. Tworowsof cells between Rs and M2 for a con-
siderable distance before and beyond the fork of the Rs, some-
times indications of a third row. Mya arising under the
stigma at a point opposite or somewhat before the middle.
Antenodals bad postnodals CuCr a Spt 3.

i-2

<

Female—Abdomen about as long as hind wing, deeper
than in the male, especially at seg. 3, the dorsal and ventral
surfaces of which appear in profile nearly straight. Seg. 3 al-
most as long as 1+2, barely longer than 4.

Genital valves a little longer than the dorsum of 9, ven-
tral margin in profile slightly convex, especially at apex;
lateral carinae percurrent, sharp and very prominent; space
between them in ventral view, suddenly widened a little be-

76 WALKER: NORTH AMERICAN SPECIES OF AESHNA

yond the base, thence tapering evenly to the narrow pointed
apices, the greatest width being about two-fifths of the
length. Ventral surface of valves moderately declivent. Styli
somewhat less than one-half as long asthe dorsum of Io. Basal
plate of ovipositor of moderate size, the posterior margin
nearly straight. Ventral surface of 10 deeply cleft by the
posterior margin. Appendages slightly shorter than 9+10,
somewhat incurved at base, lanceolate, greatest width be-
yond the middle, somewhat less than one-fourth of the length;
curve of inner margin slightly stronger than that of the outer;
apices rotundo-angulate without a projecting tooth.

The colour-pattern differs but little from that of the male.
Dorsal abdominal spots somewhat smaller, lateral spots
slightly larger than in the male. MD sometimes compara-
tively small, triangular and isolated, sometimes large, sub-
quadrate and confluent with ML and PD on some of the seg-
ments. PL on 2-7 separate from PD, but confluent with ML
on some or all of the segments.

Measurements—Thor. & 9-10, 9 9-9.5; abd. co 37.5-42,
2 35-40; seg. 3 co’ 5.8-6.7, 9 5-5.8; depth seg. 2 o" 4-4.5,
9 4-4.8 ; depth seg. 3 o& 2-2.6, 9 3.3-3.6 ; apps. o 4-4.1,
9 3.5; gen. v. 2.4-2.6 ; h.w. o 36-40, 9 34-38.5 ; width h.w.
oO I1-12.5, Q I1-12.8; pter. & 3.5-3.8, 9 4.

Type— 9, Museum Comp. Zoology—Labrador.

Nymph—Unknown.

Material determined—5 30 5 9. LABRADOR (M.C.Z. 10° 1 9, Brit. Mus. 1%).
NEWFOUNDLAND: Grand Lake, July 21, 1910 (D.A. Atkinson, coll.Williamson,
1 9). New HAmpsHIRE: White Mountains (Scudder, M.C.Z., 1 o). NoRTHWEST
TERRITORY : Fort Resolution, Great Slave Lake, 1861 (Kennicott, M.C.Z.,2 0
39).

Identity—I can find no good characters for the separa-
tion of this form from the Palearctic caerulea Strém (borealis
Zett.). The only distinctive structural character is the shorter
and relatively broader form of the appendages of the female in
septentrionalis, but as | have shown elsewhere these struc-
tures vary greatly in length in most species of Aeshna and, as
a rule, long appendages are relatively narrower than short
ones. It must also be borne in mind that very few female
specimens of septentrionalis exist in collections, and still
fewer with unbroken appendages. In fact I have seen but

two such specimens.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 77

The upper branch of Rs is generally somewhat less defi-
nite in typical caerulea than in septentrionalis,the venation in
general slightly more complex and the size slightly larger,
but none of these characters are constant enough to be of any
value in separating the two forms.

The colour-pattern is also almost identical in the two
forms but here there is one point of difference in the form of
the first lateral thoracic bands. These are somewhat wider
opposite the lowest bend in caerulea, the hind margin not being
angularly bent as in septentrionalis but only slightly curved.
This feature, however, seems too trivial to be made the basis
of a specific distinction, and may prove to be inconstant when
the species is better known.

The distinctive characters for these forms given by
Martin (’08) are incorrect. The line of contact of the eyes is
quite as short in this form as in typical caerulea. Martin’s
figure of the wings of septentrionalis belongs to some other
species, one of the clepsydra group and probably interrupta.
It may be noted that the maximum number of antenodal
veins in the fore wings as given by Martin is 18, and the
number in his figure is 19. The figures of the appendages are
also apparently from one of the races of interrupta, though
the profile view is as much like septentrionalis as any other
species.

Distribution—This is a circumboreal species, the race
septentrionalis inhabiting the Hudsonian and northern part of
the Canadian Zone from the Atlantic at least as far west as
Great Slave Lake.

Aeshna sitchensis Hagen.

Celoae. ne. 5 Ole ES. es. 2: Zao. 22, fie. 4.3
pl. 23, figs. 1;2.)

Aeschna sitchensis, Hagen, Syn. Neur. N.A., p. 119 (1861); Pr. Bost. Soc.
Nat. Hist., XVIII, p. 33 (1875) ; Kirby, Syn. Cat. Neur. Od., p. 87 (1890); Hagen,
Psyche, V, p. 353-355 (1890); Banks, Trans. Am. Ent. Soc., XIX, p. 353 (1892);
Currie, Pr. Wash. Acad. Sci., III, pp. 217, 222 (1901); Skinner, Ent. News, XV, p.
288 (1904); Osburn, Ent. News, XVI, p. 186 (1905); Williamson, Ent. News,
XVII, p. 135 (1906); Fletcher and Gibson, 38th Ann. Rep. Ent. Soc. Ont., p.
132 (1908); Martin, Cat Coll. Zool. Selys, X VIII, pp. 41, 79, fig. 37 (1908); Gen.
Ins., CXV, p. 11 (1911).

Aeshna sttchensis, Williamson, Ohio Nat., VII, p. 150 (1907); Walker, Can.
Ent., XL, pp. 386, 390, 451 (1908); Fletcher and Gibson, 39th Ann. Rep. Ent. Soc .

78 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Ont., p. 114 (1909); Muttkowski, Cat. Od. N.A., p. 114 (1910); Hebard, Ent.
News, XXI, p. 134 (1910). Gibson, 4oth Ann. Rep. Ent. Soc. Ont., p. 126 (1910).

A rather small, moderately slender species.

Male—Occiput, pale greenish yellow, broader than long,
two-thirds to fully as long as the line of contact of the eyes.
Frontal vesicle large, usually somewhat more than twice as
broad as long, pale yellow with a black border of variable
extent, usually confined to the ocular margin but sometimes
surrounding the vesicle and always including the lateral
ocelli. Eyes shorter than in the other North American
Aeshnae except Ae. caerulea septentrionalis, their colour in
the living insect unknown, in dried specimens dark olivaceous
brown. T-spot heavy, the cross-bar thick, 2-2.5 mm., stem
very short, .5--75. mm. broad; preocular band extending
laterally a little beyond the limits of the T-spot, curving a
little cephalad and terminating abruptly a little beyond the
base of the antenna, its breadth near the end being about
.5 mm. From this band a narrow black line passes ventrad
along the ocular margin of the frons and nasus to'the tips
of the lateral lobes of the latter. A black line on the
fronto-nasal suture, not very heavy in the middle but ex-
panding to form a broad black triangular area where it joins
the marginal line. Face pale, dull greenish or olivaceous,
sometimes dark brownish in dried specimens, paler next
the ocular margin. Nasus short, the vertical dimension
being about equal to one-third of the transverse, the two
depressed areas near the lower margin not marked by black
dots; lateral lobes not flaring, bluntly angulate, the outer
margin straight. Rhinarium dark reddish brown. Labrum
pale yellowish green, margined narrowly above, broadly
below, with black. Labium pale yellow, the lateral lobes
and the apical part of the median lobe reddish brown.
Rear of head black.

Thorax moderately dark brown with a bronzy lustre,
rather thinly covered with greyish hairs, venter more or
less covered with a greyish bloom. Dorsal thoracic bands
reduced to a pair of small narrow pale spots about I mm.
long, often ill-defined and sometimes indistinguishable in
dried specimens. Antealar sinus without pale markings.

WALKER: NORTH AMERICAN SPECIES OF AESHNA #79

Ventro-lateral angles of meso-episternum with a pale spot.
Lateral bands narrow and irregular, pale yellow or whitish,
generally shading above, especially the second band, into
pale bluish or blue. First band very strongly sinuate,
being twice bent, at alternate angles, broadest at about the
lower bend (.5-.75 mm.), narrowest above the second bend
(.2-.5 mm.), more or less expanded toward the upper end.
The lowest part of the band is oblique, bounding the humeral
suture, the middle part nearly horizontal, the upper part
sub-vertical. Second band nearly straight behind, con-
stricted and sometimes nearly divided by the deep excava-
tion of the front margin, the breadth below the constriction
.4-.75 mm.; expanding at the upper end to about 2 or 2.5 mm.
Behind the upper bend in the first band is a narrow sinuate
pale streak. Spots at bases of wings above pale yellowish.

Legs dark reddish brown, under side of femora nearly
black, no pale streak on fore femora.

Abdomen 4.5 times as long as thorax and slightly longer
than hind wing, rather slender, considerably constricted
before the middle of 3, thence expanding on this segment
to nearly its full width, which is attained at the apex of 4;
sides of 9 not at all divergent caudad, continuous with those
of 10.

Seg. 1 without a ventral tubercle, auricles on 2 with 3
teeth (rarely 2), tergal margins of 2 subangulate, widely
separate at base, but closely approximated a little beyond
the middle. Spines of the anterior lamina long, reaching
caudad a little beyond the hamular processes, stout at base,
and proximally directed caudad, thence tapering and curving
ventrad to the finely pointed apices, which project distinctly
below the level of the tergal margins. Hamular processes
very large, triangular; bases rounded, apices acute, directed
cephalad; inner margins straight and attingent along their
entire length, considerably elevated. Hamular folds small,
close together, nearly or quite covered in a direct ventral
view by the hamular processes, with which they are connected
by arather low ridge. Lateral carinae on 7 and 8 straight.
Dorsum of 10 with a fairly prominent median sub-basal
tooth and one or two minute ones on each side of it.

80 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Superior appendages more or less granulate, about as
long as the last two segments, their bases slender and about
one-third as broad as the space between them, expanding
gradually on the inner margin to the middle or a little be-
yond, when the breadth is about one-fifth of the length,
thence narrowing slightly to the rotundo-angulate apices,
which terminate in a small abruptly decurved spine; outer
margin moderately convex, inner margin (vertical view)
gently concave at the proximal half, straight, or nearly so,
in the distal half. Superior carina arising in a basal dorsal
tooth, partly concealed by the tergum of 10, percurrent,
though often indistinct in the basal third, gradually and
moderately elevated beyond, bearing 8 or 10 minute denticles
on the apical third. In profile the outer margin appears
nearly or quite straight, the part of the appendage internal
to the superior carina forms a fairly prominent subangulate
ventral carina, the deepest part of which is opposite the
apical fifth or sixth of the appendage. Inferior appendage
triangular, barely more than one-half to three-fifths as long
as the superior pair, breadth at base equal to three-fifths of
the length, sides straight, apex rounded; in lateral view very
gently curved, the curve of the lower margins considerably
stronger than that of the upper, tapering considerably in the
distal two-thirds.

Seg. 1 brown, dorsal spot blue, lateral spot pale greenish,
2.5-3 mm. in transverse dimension, acute above, widening
below to I mm. or a little more.

Seg. 2 brown. AL and ML pale greenish or bluish,
separated except for a very short distance above the auricle
by a black streak. AL oblong, about 1.5 mm. broad (trans-
verse), covering more or less of the upper surface of the au-
ricle, but not extending below the latter in front. ML a
narrow streak generally extending somewhat farther dorsad
than AL and nearly meeting MD. PL blue, about as long
as broad, front margin roundish and almost touching the
transverse carina. PD blue, nearly as long as PL, with which
it is narrowly connected, connate with its fellow of the oppo-
site side, concavely emarginate on each side of the median
line, along which it is continued forward as a narrow streak

WALKER: NORTH AMERICAN SPECIES OF AESHNA 81

joining AD (a small yellowish triangle) at the front
margin.

-Segs. 3-10 brownish black, the pale areas except MD
blue and somewhat larger than usual. AL on 3-8, large,
quadrate, giving off narrow basal transverse offshoots which
do not form complete rings; upper margins straight, inclined
more or less dorso-caudad (except on 3). ML present on
3-8, large, subrhomboidal, somewhat rounded behind, not
indented in front. MD on 3-8, greenish or yellowish, quad-
rate and broadly confluent with ML on 3, triangular on
4-7 and connected with ML on 3 and sometimes narrowly
on 4, small and transversely elongate on 8. PD increasing
in length from I-1.25 mm. on 3 to I.75-2 mm. on 9 and Io.
It covers about half the dorsum of 8, four-fifths or more of
g and 10 and may be more or less connate with its fellow of
the opposite side on all of these segments, though generally
separate throughout. The front margins are rounded on
the more anterior segments, somewhat irregular and indented
on the last four or five, the inner margins posteriorly sub-
parallel. PL on 3-7, rather large, connected behind on
all the segments with PD, of which it forms a broad, some-
what curved offshoot, pointed at the anterior ends.

Wings of average size and usual form, the hind wing
very slightly shorter than abdomen, hyaline, costal margin
moderately dark brown; pterostigma long, dark brown,
yellowish beneath; membranule uniform smoky brown,
extending caudad fully as far as the cross-vein of the anal
triangle, which is 2-celled.1. One or 2 cells between A2
and A3 at their origin. Rs forking under the proxi-
mal end of the stigma or the first postnodal cell before it,
in the hind wing occasionally opposite the distal half of the
second cell before it; the fork distinct but the upper branch
is slenderer and comes off at a much greater angle than the
lower branch; 2-3 rows of cells between the forks at the level
of the distal end of the stigma and 2-4 cells between them
at the margin, towards which the forks are often slightly
convergent. Two rows of cells between Rs and M2 fora

1] have a specimen from the Isle d’Orleans, Quebec, in which a minutethird
cell is present on each side, including the outer basal angle.

82 WALKER: NORTH AMERICAN SPECIES OF AESHNA

considerable distance before and after the point of forking.
Rspl gently sinuate, 4-5 rows of cells between it and Rs
where most widely separated. Mia arising under the stigma
at a point about opposite the middle.

Antenodals Satis postnodals shia CuCr oe Sp gee
10-11 II-15 5-6 1-2
Female—Abdomen about as long as the hind wing,
deeper than in the male, especially at 3, the dorsal and ventral
surfaces of which appear straight in profile. Seg. 3 about
as long as I+2 or 4.

Genital valves barely longer than the dorsum of 9,
ventral margin in profile nearly straight, apices scarcely
elevated. Lateral carinae percurrent, very prominent and
of nearly equal height throughout; the space enclosed by
them, in ventral view, widening a little from the base to
about three-tenths of the length, where the breadth is equal
to about two-fifths of the length, thence tapering evenly to
the narrow pointed apices. Ventral surface of valves strongly
declivent, distinctly grooved inside the lateral carinae; styli
about half as long as dorsum of 10 (.6 mm.); basal plate of
ovipositor of moderate size, the posterior margin straight or
indistinctly bilobed.

Appendages about as long as dorsaofg+1o, lanceolate,
broadest about the middle, where the breadth is about one-
fourth of the length. Apices tapering but little more abruptly
than the bases, rotundo-angulate with a small terminal
tooth; both margins evenly convex, the curve of the inner
somewhat stronger than that of the outer.

The colour-pattern differs from that of the male in the
usual way. Dorsal thoracic bands represented by a pair of
minute ill-defined spots, or absent; lateral bands in the New-
foundland specimen broader than in any of the males seen.
Abdominal spots PD considerably smaller than in the male
and more widely separated, especially on 8 and 9g, the inter-
vening space being somewhat more than I mm. broad on 8;
lateral spots considerably larger than in the male. PL large,
quadrate, connected narrowly with PD, and confluent on 2
with ML.

WALKER: NORTH AMERICAN SPECIES OF AESHNA = 83

The colour of the face and thoracic bands seems to have
been about the same in the Newfoundland female as in the
male. The lateral bands in the Isle Royale specimen are
wholly greenish yellow. Of the abdominal spots, PD appears
to have been a yellowish green, MD yellow and the lateral
spots green on seg. I, bluish green on the others. The wings
are hyaline in the Newfoundland and Burrough’s Bay speci-
mens, somewhat smoky in the one from Isle Royale. No
doubt considerable variation occurs in the females of this
as in other species of Aeshna.

Measurements—Thor. & 9-10, 99; abd. o& 40-44.5,
9 38.5-40; seg. 3. o’ 6.75-8, 2 5.7-6; depth seg. 2 & 4-4.4,
24.5-4.75 ; depth ‘seg: 3.\0" 1-5-2, 2 3-25 ; apps: |\o" 4-4-7;
9 3.5-3-753; gen. v. 2.2-2.25 ; h.w. co 38-40.5, 2 36.5-40;
width h.w. co II-12.3, 9 12-12.5; pter. o 3.75-4, 2 4.5-5.

Type— ov, Mus. Comp. Zoology—Sitka, Alaska.

Material determined — 15 3 62. NEWFOUNDLAND: (R. Thaxter, M.C.Z.,
2c 1 ¢); Bayof Islands, July 7, 1901 (D.A. Atkinson, coll. Williamson, 17 1
9). QuEeBEc: Ellis Bay, Anticosti Is., Sept. 12, 1910 (E. V. Cowdry, coll. Walker,
1 o'); Levis (T. W. Fyles, coll. Walker, 167) ; Isle d’Orleans, Aug. 26, 1904
(Walker, 1 o’). ONTARIO: De Grassi Pt., Lake Simcoe, July 1, 1905 (A. L. Walker,
1 oc’); Sault Ste. Marie, Sept. 14, 1907 (Donaldson, coll. Williamson, 1). MIcH-
IGAN: Isle Royale, Aug. 21, 1905 (C. C. ApAms, Univ. Mich.,1 07). MINNESOTA:
Duluth, Sept. 2 (O. Sacken, M.C.Z., 1c). ManitosBa: Winnipeg Beach, L. Win-
nipeg, Sept. 6, 1909 (J. B. Wallis, 1c") ; Westbourne, Aug. 19, 1908 (J. B. Wallis,
coll. Walker, 1 co’). SASKATCHEWAN (Scudder, Kennicott, M.C.Z., 2 & 19).
NORTHWEST TERRITORY: Fort Resolution,Great Slave Lake, July 1862 (Kennicott,
M.C.Z., 1@). “British AMERICA” (Scudder, M.C.Z., 1 9). ALASKA: Sitka,
aie Io’, type); Burrough’s Bay, Aug. 5, 1894 (J.A. Cadenhead, coll. Walker,
rie):

Distribution—Canadian and Hudsonian Zones.

Most of the records of this insect are from the southern
part of its area of distribution, where it is rare and local. The
single specimen from Lake Simcoe must be regarded as a
straggler from the north, as it is certainly not native there.
This specimen is the largest that I have seen.

Aeshna juncea Linné.
(Pita2 ae, 6. pl. 05, hes. 3, 345 pl 18, figs. 3, gars

pl. 23, figs. 3-4.)
Libellula juncea, Linné, Syst. Nat. 1, p. 544 (1758.)
Aeshna juncea, Stephens, Ill. Brit. Ent.,6, p. 84 (1835); Evans, Brit. Lib., p.
21, pl. 11, fig. 2 (1845); Calvert, Occ. Pap. Bost. Soc. N.H., VII, 6, p. 23 (1905);
Williamson, Ohio Nat.,VII, p. 150 (1907); Walker, Can. Ent., XL, pp. 385, 390,
451, pl. 10, figs. 1, 2, 5 (1908); Muttkowski, Cat. Od. N.A., p. 111 (1910).

84 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Aeschna juncea, Selys, Mon. Lib. Eur., p. 106 (1840); Hagen, Stett. Ent. Zeit.,
XVII, p. 369 (1856); Hagen, Syn. Neur. N.A., p. 120 (1861); Pr. Bost. Soc. Nat.
Hist., XVIII, p. 35 (1875); Ris, Faun. Helv., p. 25 (1885); Kirby, Syn. Cat. Neur.
Od., p. 87 (1890); Banks, Trans. Am. Ent. Soc., XIX, p. 353 (1892) ; Calvert,
Trans. Am. Ent. Soc.,XX, p. 199, 207 (1893); Wallengren, Ent. Tidsk., XV, p.
256 (1894); Calvert, Trans. Am. Ent. Soc., X XV, p. 54 (1898); Lucas, Brit. Drag.,
p. 190, pl. 16, o& @ (1900) ; Currie, Pr. Wash. Acad. Sci., III, p. 219 (1901); Wil-
liamson, Ent. News, XIII, p. 146 (1902) ; Williamson, Ent. News, XIV, p. 7,
pl. 2, fig. 1 (1903); Osburn, Ent. News, XVI, p.190 (1905); Williamson, Ent. News,
XVII, p. 134, 135 (1906); Martin, Cat. Coll. Zool. Selys., XVIII, pp. 34, 82,
fig. 29 (1908); Cockerell, Ent. News, XIX, pp. 455. 457-459 (1908); Martin, Gen.
Ins., CXV, p. 11, pl. 1, figs. 1-8b (1911).

Libellula ocellata, Mueller, Nova Acta Leop. Carol. Acad., III, p. 125 (1867).

Aeschna ocellata, Hagen, Syn. Lib. Eur., p. 54 (1840).

Aeschna rustica, Zetterstedt, Ins. Lapp., p. 1040 (1840).

Aeschna picta, Charpentier, Lib. Eur., p. 112, pl. 20, 7 9 (1840).

Aeschna picta var. caucasica, Kolenati, Mel. Ent., V., p. 114 (1846).

Aeschna caucasica, Selys, Rev. Odon., p. 300 (1850).

ae propingua, Scudder, Pr. Bost. Soc. Nat. Hist., X, p. 215 (1866) [in
part].

A rather stoutly-built species, of medium or somewhat
less than medium size, though attaining a large size in Europe.

Male—Occiput light yellow, sometimes darkened at apex,
distinctly broader than long, often twice as broad, and from
one-fifth to two-thirds as long as the line of contact of the
eyes. Frontal vesicle more than twice as broad as long,
generally wholly yellow above, except a small area around the
lateral ocelli and a narrow hind margin, but sometimes the
yellow area is confined to the anterior half or two-fifths.
Eyes emerald green above, fading below to pale yellowish
green and pale blue behind, the hind margin above greenish
yellow. Preocular band about 2.5-3 times as broad, at base of
antennz, as one of the lateral ocelli, narrowed to a thread
line on the sides of the frons and nasus. T-spot heavy, but
with a rather slender stem, the sides of which are slightly
divergent or subparallel, cross-bar 2.5-3 mm., stem .5-.6 mm.
broad in-front, .7--8 mm. broad behind. Frons and nasus
yellow or greenish yellow, more or less infuscated just below
the T-spot; a heavy dark brown line on the fronto-nasal suture,
narrowing laterally, and two impressed dark brown dots near
the lower margin of the nasus. Lateral lobes of the nasus
rotundo-angulate, somewhat flaring. Rhinarium dark red-
dish brown, with a yellowish upper margin. Labrum yellow
or greenish yellow, with upper and lower black marginal
lines. (In one male from the Bighorn Mts., Wyo., it is com-

WALKER: NORTH AMERICAN SPECIES OF AESHNA 85

pletely surrounded by a heavy black margin.) Labium light
yellow, more or less infuscated towards the margins. Rear
of head black.

Thorax rather heavily built, dark reddish brown. Dor-
sal bands pea-green, 4 mm. long .4-.75 mm.broad at the middle,
straight or slightly curved, sides nearly parallel, lower ends
pointed, separated by 4-4.5 mm., upper ends generally some-
what expanded, .75-1.3 mm. broad, separated by 1.3-2.3 mm.
A green spot in the antealar sinus and sometimes an ill-de-
fined humeral streak. Lateral bands broad, straight, green-
ish yellow, pale blue at the upper ends, surrounded by an ill-
d fined dark margin; first band rounded at lower end, near
which it is 1.25-1.5 mm. broad, gradually tapering to about
half that breadth at the upper end, from which no offshoot is
given off; margins nearly straight. Second band rounded at
the lower end, which is generally slightly narrower than the
upper end, sides straight. Between the two bands are two
elongated yellowish spots, sometimes united into an undulate
streak. Spots at bases of wings above light yellow, those on
the tergum between the wings, pale greenish blue.

Legs black, the posterior surfaces of the fore coxae and
bases of fore femora pale yellowish.

Abdomen nearly or quite 4.5 times as long as the thorax,
strongly constricted before the middle of 3, widening to about
the apex of 4 and narrowing again slightly on the posterior
4 or 5 segments, though generally expanding slightly on 7.
Seg. 1 without a ventral tubercle; spinulose area on ventral
surface of 2 generally restricted to the apical fourth; auricles
with 3-5 teeth; tergal margins angularly approximated op-
posite the sheath of the penis. Spines of the anterior lamina
long, projecting below the tergal margins, curved ventrad and
inclined caudad and slightly laterad, stout at the base but
tapering to a fine point. Hamular processes long, slender,
spine-like, slightly hooked at the tips, inclined ventrad and
mesad, the tips often overlapping a little. Hamular folds
separate from the hamular processes, and somewhat widely
separated from one another, rather small and contracted, the
outer and posterior margins meeting at a rotundo-acute angle.
Seg. 3 one-fifth to one-third longer than 1+2 or 4, the

86 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

relation depending upon climate (vide postea). Lateral
carinae on 8 not angulate, obsolete on the basal half of 9,
generally distinct on the apical half. Dorsum of Io with a
prominent median tooth between two pairs of much smaller
teeth.

Superior appendages as long as, or slightly shorter than,
g+10, their basal breadth about one-third of the space
which separates them, gradually expanding to the middle or
a little beyond (dorsal view), where the breadth is about one-
fifth of the length; thence narrowing again somewhat to the
rounded or somewhat angulate apices which bear, near the
outer margin, a terminal decurved spine of variable size,
generally small but sometimes well developed. Outer mar-
gin in dorsal view moderately curved, inner margin feebly or
scarcely at all sinuate. In profile view the outer margin is
moderately to strongly curved upwards in the distal half, the
lower margin minutely denticulate, slightly convex in the
basal fifth, somewhat concave before the middle, produced in
the distal three-fifths or two-thirds into a rather prominent
angulate or arcuate inferior carina. Superior carina dis-
tinct except at base, somewhat elevated and arcuate in the
apical fourth, smooth. Inferior appendage triangular, about
three-fifths as long as the superior pair, the breadth at base
equal to three-fifths or two-thirds of the length, sides straight
or barely convex, apex rather blunt and rounded, lateral mar-
gins terminating above in a pair of minute recurved teeth; in
profile gently curved, tapering evenly to the apices.

Seg. 1 brown, dorsal spot blue, lateral spot greenish
yellow, generally about 3mm. long and .75-1 mm. broad but
much narrower in the Alaskan specimen and generally in
European specimens.

Seg. 2 brown, AML greenish yellow in front, blue be-
hind, subrhomboidal,1.75-2 mm. broad at the middle, cover-
ing sides of auricle, partly divided by the transverse carina,
behind which it is produced dorsad as a narrow offshoot not
quite reaching MD. PL yellow in front, shading into blue,
about four-fifths as long as the posterior division of the seg-
ment, fairly broadly connected with PD, which is confluent
with its fellow of the opposite side, forming a broad blue band,

WALKER: NoRTH AMERICAN SPECIES OF AESHNA 87

whose front margin is convexly curved on each side of the
middorsal line, and from which a pale greenish median streak
extends forward to the front margin, where it is somewhat ex-
panded (AD).

Segs. 3-10 brownish black, AD and MD pale green, PD
blue, lateral spots bluish green, on 3 often blue. AL on 3-8, of
moderate to small size; on 3, subtriangular, nearly as broad
at base as long, upper margin straight, very oblique; on 4-5
or 6 giving off a small basal transverse offshoot which does
not form a complete ring. ML on 3-8, of moderate size, en-
larging to 6, angular with the upper margin very oblique on
the more anterior segments, more rounded on the posterior
segments, front margin not indented. AD represented on
3-8 by a minute spot; MD on 3-8, not connected with ML, a
transverse streak on 8, transversely elongate-triangular on the
other segments, the postero-lateral margins more or less con-
cave. PD on 3-10, I-1.25 mm. long on 3, increasing to 1.5 mm.
on 8, rounded in front; the space separating them gradually
increasing from about .4 mm. on 3 to I or 1.25 mm. on 8, then
decreasing again slightly; inner margins straight on 8 and 9,
nearly parallel on 8, slightly divergent on 9, rounded on the
other segments. PL on 3-5,-6 or -7, separate throughout from
PD or narrowly connected on 3 only, triangular on the an-
terior segments, becoming an elongate streak farther caudad,
and finally a mere dot. Ventral surface of abdomen without
pale markings.

Wings hyaline to rather strongly flavescent, costal veins
brown, ochre-yellow along the free margins, pterostigma long,
reddish brown; membranule of hind wing smoky brown, with
the basal fourth to more than one-half whitish, not reaching
thecross-vein of the anal triangle, which is 2-celled. One or 2
cells between A2and A3 at their origin. Rs forking at the level
of 1-2 (forewing) or 1-3 (hind wing) postnodal cells before the
stigma,! the upper branch strongly curved proximally; 3 rows
of cells between the forks at the level of the distal end of the
stigma, and 3-6cells between them at the margin. Three or 4
rows of cells between Rs and Rspl at the level of widest separ-

1In European specimens the fork is often before the proximal end of the stigma.

88 WALKER: NORTH AMERICAN SPECIES OF AESHNA

near the point of forking. Mua arising beyond and some-
times opposite the middle of the stigma.

Antenodale 9°T3° CuCr 20) Spt aS as *

II II-15 4-5 2(3)

Female—Abdomen slightly shorter and considerably
deeper than in the male, slightly longer than the hind wing.
Dorsal and ventral surfaces of 3 nearly straight in profile.

Apices of genital valves not at all elevated, compressed
and rather broadly rounded as seen in profile, inner surface
obliquely concave. Lateral carinae obscure and rounded at
base, prominent in distal half, in profile sinuate, being more or
less concave before the middle, convex beyond; in ventral
view, sinuate, usually strongly so, being convex in the proxi-
mal three-fifths or two-thirds, slightly concave or straight
distally. Space enclosed by lateral carinae broadest just
before the middle,where the breadth is about three-sevenths of
the length, narrowing to about half this breadth before the
apices. Ventral surfaces of valves feebly sulcate. Basal
plate of ovipositor large, consisting of two rounded lobes, one
on each side; lateral plates absent. Styli scarcely more than
half as long as the dorsum of Io (.6 mm.). Appendages about
as long as 9+10, expanding from the moderately slender
bases by the curve of the inner margin to about three-fifths
of the length, where the breadth is about one-fifth of the length,
tapering rather abruptly to the rounded or rotundo-angulate
apices, which usually terminate in a minute, barely projecting
tooth. This tooth is occasionally well developed or may
sometimes be absent. Outer margin nearlystraight, inner mar-
gin more strongly curved.

The dorsal thoracic bands are usually narrower than in
the male and, as a rule, divided into an anterior longitudinal
streak, and a posterior shorter transverse spot at
the antealar sinus. The lateral thoracic bands and the lateral
abdominal spots tend to be a little broader than in the male,
particularly PL,which is represented on 2-7, and tends to be
quadrate in form, and is separated from PD except sometimes
on 2, where they may be narrowly confluent. On the same
segment PL is broadly connected with AML and is sometimes

, postnodals

WaLkKER: NorTH AMERICAN SPECIES OF AESHNA 89

narrowly joined with ML on 2. AL on 2, quadrate, connected
below with ML. PD smaller than in the male, generally con-
nate on 2, but separate in specimens from Alaska and in one
from the Magdalen Islands.

The ground colour is like that of the male except that
the abdomen is reddish brown, darkened about the margins,
sutures and pale areas. Considerable variation exists in the
colour of the latter. I have taken both blue and yellow
females at Nipigon, Ont., the colour of the former differing
but little from that of the male except in the paler shade of
blue. In the heterochromatic type all the pale markings are
green or yellow. Both types are also represented in a series
of British specimens received from Mr. K. J. Morton of Edin-
burgh. In this series PD varies from blue to greenish yellow
and the lateral spots from pea-green to pale yellowish green.

The wings are sometimes hyaline, sometimes more or less
flavescent and may be quite deeply suffused with brown. The
pterostigma is considerably longer than in the male.

The following colour notes were made on the day of
capture from a female taken at Nipigon, Ont.:

Occiput and frons greenish yellow, shading into light
yellow slightly tinged with green on the lower part of the nasus
and labium; eyes grey-green with a greenish yellow hind
margin. Thorax olive-brown, paler beneath; lateral thoracic
bands yellowish green, the lower part of the first band yellow.
Abdomen brown, lateral spots on 1 and 2 yellowish green, on
the remaining segments pea-green. PD yellowish green, be-
coming pale green on g and Io.

Measurements of North American specimens—Thor. o&
10.5-II, 9 10-10.2; abd. co 45.5-49.7, 2 42.5-48 ; seg. 3. oS
7.4-8.6, 2 6-7.1; depth seg. 2 @ 5-5.75, 2 5-6; depth seg. 3
2 3.75-4.33; apps. do’ 4.8-5.5, 9 4-4.75; gen. v. 2-2.33; h.w.
o' 42-46, 2 39.5-46.5; width hw. @ 12-13, 9% 12.5-13.3;
pter. o 3.6-4.25, 2 3.7-4.9.

Measurements of British spectomens—Thor. o 10.5-11.5,
9 10.5-12; abd. & 47-54.5, 2 46-53-53 seg. 3 o 7-5-9, @
6.5-8.33; depth seg. 2 o’ 5-6.5, 9 6-6.5; depth seg. 3 @ 4.25-
4.8; apps. co 4.6-5.5, 2 4.33-5.5; gen. V. 2.3-2.5; hw. o 43.5-

7

g0 WALKER: NORTH AMERICAN SPECIES OF AESHNA

47, 9 43-47; width h.w. o& 13-14, 9 13.5-15 ; pter.
oS 4-4.75, 2 4-33-5-4-

Nymph (pl.6,fig,1 ; pl.8,fig.1; pl.10,fig 1) —Eyes moderately
prominent, the posterior margin about equal in length to the
antero-posterior diameter; lateral margins of head somewhat
more oblique than in clepsydra, canadensis, etc., the straight
part but little longer than the width of the interocular space;
postero-lateral corners of head broadly rounded, posterior
region barely emarginate. Mentum of labium seven-elevenths
as broad at base as at apex, the apical breadth a little
more than two-thirds of the length; proximal three-fifths
barely widening distad, the sides feebly arcuate, distal two-
fifths with the sides rather strongly arcuate; middle lobe some-
what narrow, obtusangulate; lateral lobes narrowing slightly
distad, truncate but with the outer apical angle much rounded,
the inner with a minute tooth.

Supracoxal processes well developed, equal in length,
acute, the posterior a little stouter than the anterior, the inter-
vening space rectangular or slightly narrower than aright angle.

Abdomen of average form, broadest at seg. 6; lateral
spines present on segs. 6 to 9, those on 6 rudimentary, on 7
reaching half-way to the posterior margin, on 8 two-thirds of
the way to the margin, on 9 not quite to the middle of Io.
Lateral appendages ratherslender, half as long as the inferior
pair (9), or slightly more (co); basal part of superior ap-
pendage in the male about four-fifths as long as the lateral
pair, its basal breadth a little less than its length, sides feebly
emarginate, apex bluntly pointed. Genital valves nearly one-
half as broad as long, not very deeply declivent, nearly or
quite reaching the apical margin, the ovipositor slightly
shorter.

Head and thorax almost uniform brownish, sides of
head and lateral margins above sometimes paler ;
legs concolorous. Abdomen brownish; dorsal surface with
two longitudinal pale bands, fairly well defined on the an-
terior four or five segments, but fading caudad and becoming
very faint or disappearing altogether at seg. 9.; median dark
area between the bands but little or no darker than the gen-
eral ground colour, but exhibiting a more or less distinct pale

WALKER: NorRTH AMERICAN SPECIES OF AESHNA QI

median line in the anterior half or more of the abdomen;
dorsal punctae dark but discrete; sides of abdomen brown,
slightly paler outside the lateral scars, which form a pale inter-
rupted wavy band, with a dark inner edge; lateral and dorso-
lateral punctae dark brown, well marked; no pale mottlings
between the lateral scars and the pale dorsal bands.

Measurements—Length of body 39.5-42.8; mentum of
labium 6.2-6.9; h.w. 8.4-9.8; h.f. 6.2-7.3; inf. apps. 4.2-4.8;
gen. v. 1.9-2.4; width of head 7.75-8.5, of abdomen 7-8.

Material determined —17 o& 19 9 (excl. Palearctic). QUEBEC: Grand
Entry, Magdalen Is., June 30, r901 (D. A. Atkinson, coll. Williamson, 2 ¢&
19). NEw HAmMpsHIRE: White Mts., Aug. 22, 1862 (Scudder, M.C.Z., 1 o).
OnTARIO: Nipigon, July 10, 1907, Aug. 28, 30, 1907, Aug. 4-8, 1910 (Fletcher,
Walker, 6 o& 4 9). ALBERTA: Banff, July 17, 1902 (Osburn, N.B. Sanson,
29). British CoL_umBiA: Inverness, Aug. 1890 (Brit. Mus. 1 9). NorTH-
WEST TERRITORY: Ft. Resolution, July, 1862 (Kennicott, M.C.Z.,2 @ 1 @).
ALASKA: Bethel, Kuskoquin River (Acad. N.S. Phil., 167); Kodiak, July 20, 1899
(T. Kincaid, Harriman Exped. U.S.N.M., 1 o); Unga Is., Shumagins, July 17,
1899 (W. E. Ritter, Harriman Exped. U.S.N. M., 19). Wyomina: Shell Creek,
Bighorn Mts., July 15, 1896 (Currie, U.S.N.M. 1 o). CoLorapo : “‘South-
eastern part,’’ 10,000-11,000 ft. (Thos. S. Gillin, Acad. N.S. Phil., 2 @ 2 9).
Corfi, Colorado Plains, 1872 (M.C.Z., 2 9). YELLOWSTONE : 1872 (Hayden,
M.C.Z., 3 9). Arctic AMERICA: (Brit. Mus.,1 9). AlsoIc'! 9, without
data. British specimens: 10 o 8 9. Merionethshire, England; Capel
Cureg, Wales; Glasgow, and Glen Lochay, Scotland; and Emydale, Ireland.

Nymphs—Cierfs, Switzerland, 1900 metres, July 27, 1909, 2 9, one trans-
forming, the other about to transform; exuviae, 4 o § 9, one of the latter
with teneral imago (Dr. F. Ris).

Distribution—Circumboreal; in North America apparently
most abundant in the Hudsonian Zone and not known from
the Transition Zone. In Europe it ranges into warmer
latitudes than in North America.

Geographical vartations—Specimens from Alaska and
the Magdalen Islands are somewhat smaller than those from
Nipigon, the Bighorn Mts. and southwestern Colorado.
The wings in the former specimens are slightly broader and
the abdomen perceptibly shorter, particularly seg. 3, which
is also relatively deeper. This variation in seg. 3 is most
marked in the females and is correlated with variation in
the length of the appendages, both being dependent upon
climatic conditions (vide p. 26).

The northern specimens are indistinguishable in form
from British examples, but are, on the whole, decidedly
smaller. The largest American specimens I have seen are
from Ontario, Wyoming (Bighorn Mts.) and Colorado and

92 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

are about the size of average European individuals. Speci-
mens from Kamtchatka and other parts of northeastern
Asia are also smaller than those from Europe and are quite
similar to Alaska specimens.

No constant differences can be found between European
and North American specimens of this species, but what is
typical of the former is not always typical of the latter. The
venation of the European specimens is more complex, as is
indicated by the following analysis: ;

1. Rs tends to fork a little farther distad in European
specimens, often under the proximal end of the stigma.

2. The number of rows of cells between the forks of Rs
at the level of the distal end of the pterostigma is typically
3 in American, 4 in European specimens. In a female from
Alaska, however, there are 4 regular rows in each wing
and in several British specimens there are only 3 rows.
When this is the case, however, there are usually indications
of a fourth row.

3. In American specimens, there are 3 or 4 rows of cells
between Rs and Rspl at the level of widest separation,
there being 3rows in about 65 per cent. of cases. In European
specimens there are typically 5 rows, often 6 and only occa-
sionally 4 rows.

4. In American specimens there are 3 rows of cells
between M4 and Mspl where these veins are most widely
separated, in European specimens typically 4, though some-
times 3 or 5 rows.

In colour-pattern the differences are but slight. In
the European specimens the yellow of the costal veins is
more extensive and brighter in colour, the thoracic bands are
generally somewhat narrower, the curved spot between
the bands and the lateral spot of abdominal seg. 1 smaller.
The front margin of the first lateral band is frequently
slightly excavated above the middle, a feature not seen in
American specimens.

Habits—On the Nipigon River, where this species is
fairly common, it flies over the open marshes and shallow
reed-grown waters which border the river below the rapids.
Specimens taken July 10 were teneral; on Aug. 4-8 they

WALKER: NORTH AMERICAN SPECIES OF AESHNA 93

were in good condition, while on Aug. 28-30, on which dates
three females were captured, one was still fresh the other two
somewhat the worse for wear. These were taken while
Ovipositing among the reeds. The weather was dull, and
one was taken in flight during a very light rain.

Ris (’85) has given an interesting account of the habits
of this insect in Switzerland, where they are apparently
quite similar to their habits in Canada.

Aeshna subarctica Walker.

(Pl. 12, fig. 7; pl. 15, figs. 4, 4a; pl. 18, figs. 4, 4a;
pl. 23, figs. 5, 6.)
Aeschna juncea, Fletcher, 35th. Ann. Rep. Ent. Soc. Ont., p. 104 (1906).
Aeshna subarctica, Walker, Can. Ent., XI, pp. 385, 390, 451 (1908); Mutt-
kowski, Cat. Od. N.A., p. 114 (1910).
Aeshna subarctica, Martin, Gen. Ins., CXV, p. 11 (1911).

Male—Occiput lemon-yellow, 1.5-2 times as broad as
long and 1-3.5 times aslong as the line of contact of the eyes.
Frontal vesicle more than twice as broad as long, lemon-
yellow, with a black posterior border. Eyes sea-green or
grey-green to light brownish olivaceous with bluish grey
reflections; a small dark transverse streak above, the pos-
terior border yellowish green above, whitish or very pale
green laterally. Preocular band 2.5-3 times as broad, at
the base of the antenne, as one of the lateral ocelli, narrow-
ing on the sides to little more than half its dorsal width,
except at the fronto-nasal suture, where it is generally ex-
panded. T-spot heavy, 2.8-3 mm., stem .3-.7 mm. broad
in front, 1 mm. broad behind, the sides divergent. Frons
and nasus bright yellow, often with an olivaceous tinge,
paler along the ocular margin and on upper surface of frons.
A heavy black line on the fronto-nasal suture, broadest
in the middle. Two black impressed dots above the lower
margin of the nasus. Lateral lobes of nasus somewhat
flaring, rotundo-obtusangulate. Rhinarium dull dark brown
or plumbeous. Labrum light yellow, margined narrowly
above, more broadly below, with black. Labium pale dull
yellow or greenish, the middle lobe with an ill-defined median
blackish spot next the margin. Rear of head black.

Thorax fairly robust, dark brown with an olivaceous
tinge deepening to nearly black around the lower parts of

94 WALKER: NORTH AMERICAN SPECIES OF AESHNA

the lateral bands. Dorsal bands yellowish green, complete,
somewhat curved, strongly divergent below, about 4 mm.
long and .7-.8 mm. broad at the middle, tapering towards
each end, the lowerends acute and separated by 4.5 mm., the
upper ends suddenly expanded, I-1.5 mm. broad, and separ-
ated by about 1.5 mm. An ill-defined elongate greenish
spot generally present just in front of the humeral suture.
Lateral bands of moderate width, light blue above, greenish
yellow below; first band rounded at the lower end, where the
breadth is 1.3-1.5 mm. thence narrowing to less than .5 mm.
a little above the middle, where it is bent upwards at a slight
angle, thence widening to about I mm. at the upper end
from which a narrow posterior offshoot arises. Front margin
obtusangularly excavated, hind margin nearly straight.
Second band rounded or subangulate at the lower end, a
little above which the breadth is 1.1-1.35 mm., narrowing
to the middle or upper third by a broad rounded excavation
of the front margin and widening again to the upper end to
a breadth of about 1 mm., or slightly more; hind margin
straight. Between the two lateral bands are two yellow
spots, one larger one partly enclosing the metastigma, and
a smaller one above it. Metasternum with two pale yellow-
ish antero-lateral spots. Interalar spots pale blue, spots
at bases of costal veins of the wings yellow.

Legs reddish brown, tibiae and tarsi and underside of
femora darker; a pale streak upon the outer surface of the
fore trochanters and bases of the fore tibiae.

Abdomen a little more than 4.5 times as long as thorax,
of average build, considerably inflated at base, strongly
constricted before the middle of 3, thence expanding to the
apical margin of 4; sides of 9 feebly divergent caudad. Segs.
1+2 about as long as 4; seg. 3 rather more than one-third
longer. Sternum of 1 without a tubercle. Auricles with
3-5 teeth. Tergal margins approximated on each side of the
penis. Spine of the anterior lamina long, slender and sharp-
pointed, directed ventrad and curved slightly caudad, the
apices reaching well below the level of the tergal margins.
Hamular processes having the form of elongate, flattened
and slightly curved plates, rounded at the apices, rather

WALKER: NORTH AMERICAN SPECIES OF AESHNA 95

longer than the spines and directed ventrad, mesad and slightly
caudad. Hamular folds separate from the processes, difficult
to see, except in a strong light, as they are sunk deep in the
genital fossa and are partly concealed by the hamular pro-
cesses in a direct ventral view. The anterior part of the
fold is very small, but the posterior portion is large and pro-
longed caudad, forming with its fellow of the opposite side,
with which it is closely united, a somewhat rounded promin-
ence (pl. 12, fig. 7). Lateral carinae of 8 nearly straight,
those of 9 more or less obsolete in the basal half. Dorsum
of 10 with a prominent but not large median tooth, and one
or two minute ones on each side of it.

Superior appendages slightly longer than 9+10, moder-
ately slenderat base, the breadth here being one-third to four-
ninths that of the intervening space; expanding to about
the middle or a little beyond; the slender proximal portion,
which is about one-fourth to one-third of the total length,
passing almost insensibly into the broader distal part, which
is one-sixth to one-fifth of the total length ; apices
rounded or angulate, terminating in an abruptly decurved
tooth of variable size, being sometimes minute, sometimes
fairly well developed; outer margin gently and regularly
convex, inner margin gently sinuate, being proximally con-
cave, distally convex; a prominent dorsal tooth at the ex-
treme base under cover of the tergal margin of 10; superior
carina distinct, except on the slender basal part, the apical
fourth or third somewhat elevated and gently arcuate,
without any indications of denticles. In profile view the ap-
pendage is more or less distinctly undulate in the basal half (in
one specimen from Nipigon the undulation is scarcely notice-
able), curved gently upwards distally ;internal part in thedis-
tal three-fifths to two-thirds bent downwards forming an arc-
uate inferior carina. Inferior appendage triangular, barely more
than half to three-fifths as long as the superiors, basal breadth
three-fifths of the length, sides straight, the dorso-lateral
margins terminating in a pair of minute preapical teeth,
apex rounded; in profile curved as in juncea, tapering con-
siderably from the middle to the apex.

Colour-pattern of abdomen—Seg. 1 brown, dorsal spot

96 WALKER: NORTH AMERICAN SPECIES OF AESHNA

narrow, light blue, lateral spot pale greenish yellow to
bluish green, elongate triangular, 2.5-3 mm. long, .7-.8 mm.
broad.

Seg. 2 brown, AML greenish yellow to pale blue, or
yellow in front of the transverse carina, blue behind, cover-
ing upper surface of auricles, partly divided by the trans-
verse carina, in front of which it is I-1.5 mm. broad, and
extending dorsad behind the suture as a band 2.5-3 mm.
long, which is not confluent with MD. MD of the usual
form, yellow. PL greenish yellow to pale blue, 1.3-1.5 mm.
long, rounded in front, narrowly connected with PD, or
barely separated from it. PD light blue, somewhat shorter
than PL, its front margin rather deeply concavo-emarginate
on each side of the middle, along which a narrow greenish
streak passes to the front margin.

Segs. 3-10 dark brown, the pale markings variable in
colour. AL on 3-8, greenish blue to pale blue, rather small;
on 3 triangular, about 2.5 mm. broad at base, apex reaching
the transverse carina; on the succeeding segments semi-
elliptical, not connected with a basal ring but giving off a
short basal offshoot. ML on 3-8, greenish blue to pale
blue, fairly large, increasing in size to 6 or 7, subrhombic
on the anterior segments, becoming rounded caudad, slightly
emarginate in front. AD represented by a small yellow
dot on 3-8. MD on 3-8 or -9, bright yellow, separate from
ML, somewhat L-shaped on most of the segments but re-
duced to a pair of short transverse spots on 8 and of minute
dots when present on 9. PD on 3-10, pale green to pale
blue, 1 mm. long on 3, increasing to 1.25-1.75 on 8, sub-semi-
circular on 3-7, the rounded inner margins separated by about
.6mm.; triangular orsubtriangular on 8 and separated poster-
iorly by about I mm., subovate and about the same distance
apart on 10. PL on 3-8, unconnected with PD, except some-
times narrowly on 3, straight and elongate on the more an-
terior segments, dwindling caudad to a dot.

Wings hyaline, costal veins edged with ochre-yellow;
pterostigma rather long, dark brown; membranule of hind
wing dark smoky brown, the basal fourth or less pale yellow-
ish, not quite reaching the cross-vein of the anal triangle,

WALKER: NoRTH AMERICAN SPECIES OF AESHNA 97

which is 2-celled. One cell between A2 and A3at their origin. Rs
forking at the level of 1-3 postnodal cells before the stigma,
the upper branch proximally rather strongly curved; 3 (rarely
2) rows of cells between the forks at the level of the distal
end of the stigma and 3-6 cells between them at the margin.
Three or 4 (rarely 5) rows of cells between Rs and Rspl where
most widely separated. Mra arising behind the middle or
distal half of the eae

Antenodals ae as. I postnodals slat “ai, CuCr o7 Some 34.
=E5 13-16 5-5 2-3

ese lcainiy an about 4.5 times as long as the thorax
and about as long as the hind wing. Seg. 3 rather deep,
moderately constricted, about equal in length to 1 and 2 and
about 1 mm. longer than 4.

Genital valves short, not longer than the dorsum of
g, strongly compressed on the distal half, lateral carinae
distinct only in the distal three-fifths, in profile arcuate,
especially distally, in ventral view parallel in the distal half,
divergent and indistinct proximally. The ventral surface of
the valves is not developed in the distal half and proximally
slopes: outward, passing almost insensibly into the lateral
surfaces. Apices strongly compressed and _ blade-like, in
profile broad and rounded, with a very minute tuft of hair.
Styli .5-.6 mm. long. Basal plate of ovipositor unusually
large, consisting of a pair of rounded lobes, one on
each side. Appendages about as long as 8+g9, lanceo-
late, considerably slenderer at base than at apex, greatest
width at three-fifths to two-thirds the length, equal to one-
fifth or three-tenths of the length ; curve of outer margin
very slight, that of the inner margin much stronger ; apices
rounded, with or without a small terminal tooth.

The pterostigma is somewhat longer than in the
male. Ks forks at the level of 1-2 postnodal cells before the
stigma or beneath its anterior end. Two cells are present
between A2 and A3 at their origin.

The dorsal thoracic bands are narrower than in the male
and are sometimes complete but generally divided into an
upper shorter and lower more elongate spot, or the latter only

98 WALKER: NORTH AMERICAN SPECIES OF AESHNA

may be present. Lateral bands of similar form to those of the
male but slightly broader, the intermediate spots often con-
fluent. PD somewhat smaller than in the male, all the other
spots larger, AL and AD often connected by complete basal
rings. AL on 3 trapezoidal, confluent with ML. PL large,
subquadrate, separate throughout from PD, but generally
confluent with ML on 2 and sometimes also on 3. PD ir-
regularly quadrate on 8 and 9.

Dichromatism occurs in this species and the colour vari-
ations are very similar to those of Ae. juncea. In several
specimens seen in life the coloration was almost like that of
the male, the lateral abdominal spots being pale blue or blujsh
white and PD varying from very pale blue to greenish. In
others all the pale markings are yellow or greenish yellow.
The eyes vary from grey-green to Indian yellow and the
wings from hyaline(generally in homceochromatic individuals)
to deeply flavescent(generally in heterochromatic individuals).

Measurements— f@ 10-1, @ 9.5-10.5; abd. o@ 47-51.5,
2 43.5-49; seg. 3 co’ 8.7-9.5, 2 7-7.6; width seg. 2 o 5-5.5, 2
5.25-5.7; width seg. 3 co 3-3.75; apps. o& 5.3-6, 9 5.33-6;
gen. v.1.8-2; h.w. o& 42-46, 9 39.5-44.5; width h.w.o 12-12.5,
9 12-14; pter. co’ 3.6-4.2, 9 4.2-4.7.

Types— of and 9, coll. Walker—Nipigon, Ont.

Nymph—Among the nymphs and exuviae which I col-
lected at Nipigon, Ont., in August, 1910, there are two species
of Aeshna which I had not met with before. As the nymphs
of only two of the six species of this genus that occur at Nipi-
gon are unknown, viz., Ae. subarctica and Ae. interrupta, it is
reasonable to refer the two unknown nymphs to these two
species.

The form referred to Ae. subarctica is placed under this
species for the following reasons:

1. It resembles the nymph of juncea, the only near ally
of subarctica, very closely in both structure and colour-pat-
tern, and particularly in the details of the labium. One of
the most marked differences is the distinctly longer lateral
appendages of the female. As these organs are the rudiments
of the appendages of the adult this is what we should expect

ae -

WALKER: NorTH AMERICAN SPECIES OF AESHNA 99

in view of the similar difference in the length of the appendages
of the adult females of these two species.

2. Of the two nymphs only one exuvia of the form refer-
red to subarctica was found, against fiveof the form referred
to interrupta, and of the adults of these two species interrupta
is by far the more abundant.

3. The form referred to interrupta most closely resembles
that of eremita, the nearest relative of this species.

Head similar to that of juncea, except that the postero-
lateral angles are more prominent, though well rounded off,
being nearly as in interrupta. Labium indistinguishable from
that of juncea.

Supracoxal processes well-developed, subequal, the an-
terior considerably slenderer than the posterior.

Abdomen a little slenderer than in juncea, broadest at
segs. 6 and 7. Lateral spines present on segs. 6-9, those on 6
mere rudiments, those on 7 distant from the posterior margin
by about their own length and on 8 by about one-fifth their
length, on 9 reaching to the middle of 10. Lateral appendages
three-fifths as long as the inferior pair. Genitalia of female
barely reaching the posterior margin of seg. 9, the genital
valves more steeply declivent and somewhat narrower than
in juncea.

Colour-pattern similar to that of juncea, differing only in
the more faintly marked punctae, which are but little darker
than the ground colour, and in the somewhat larger size of
the dorso-lateral punctae.

In the young nymphs there is a very dark brown band on
the sides of the head and prothorax (also distinguishable in the
exuvia on the prothorax and anterior part of the mesothorax)
and a very distinct medio-dorsal pair of narrow dark brown
longitudinal streaks. These may also be seen in the nymph
of juncea.

Measurements—Length of body 40 ; mentum of labium
6; h.w. 9; h.f. 7; inf. apps. 4.5; gen. v. 2.2; width of head 7.5;
width of abdomen 7.

Distribution, etc-— This species inhabits the Canadian
Zone, from the Atlantic coast to Manitoba. The single female
from Lake Simcoe was probably a wanderer from the north, as

100 WALKER: NORTH AMERICAN SPECIES OF AESHNA

I have collected there for many seasons and have never seen
another specimen.

Ae. subarctica is closely related to Ae. juncea, the most
important differences being found in the accessory genitalia of
the male. Apart from these, however, the two species may
be readily separated by the differences in the form of the later-
al thoracic bands, and by the length of the appendages,
particularly in the female.

These two species fly together at Nipigon, Ont., and I
was unable to distinguish them in flight, nor could I detect
any differences in habits.

Aeshna interrupta Walker.

With the typical form of this species I have united Ae.
lineata, interna and nevadensis,as the study of a large series
of specimens seems to indicate that these forms are but geo-
graphical races of one and the same species.

The chief specific characters are as follows: frontal
vesicle rather large, a black or dark brown line on the
fronto-nasal suture, dorsal thoracic bands reduced to small
isolated spots,or absent, lateral bands onemillimetre, or less;
in breadth, either straight and sometimes reduced to a mere
line or divided each into an upper and lower spot. Sternum
of abdominal segment 1 without a tubercle but with numerous
spines near the posterior margin. Abdominal spots well
developed or but little reduced, blue, except, as a rule, MD,
which is generally greenish or yellowish. Lateral spot on I
very small or absent. Venter wholly black. Rs forking at
the level of 1-2 postnodal cells before the pterostigma, Mia
arising opposite or beyond the distal end of the stigma.

Male—Anal triangle 2-celled, auricles with 3-5 teeth, ter-
gal margins bounding the genital fossa not approximated op-
posite the sheath of the penis, considerably elevated: behind.
Spines of the anterior lamina very short and blunt, directed
ventrad and caudad, their apices sunk below the level of the
tergal margins and often concealed by them. Hamular
processes directed ventrad and mesad, short and very broad,
their internal margins rounded and terminating in a minute
but well-rounded tubercle. Hamular folds continuous with

WALKER: NORTH AMERICAN SPECIES OF AESHNA IOI

the hamular processes, well developed, ear-like, closely ap-
proximated mesially, their lateral margins regularly curved and
passing insensibly into those of the hamular processes, apices
rounded.

Superior appendages expanding -unsymmetrically from
the base, the inner margins more or less sinuate, the superior
carinae gradually and moderately elevated, usually with a few
minute denticles near the apices, which are typically rounded
and not at all decurved. Terminal tooth small or absent.
Inferior appendage triangular, three-fifths to-two-thirds as
long as the superior appendages.

Female—Genital valves extending caudad about as far
as the apical dorsal margin of 9, apices not at all elevated;
lateral carinae percurrent, very prominent, in laterdl view
irregularly arcuate concealing the ventral surface except near
the base, highest a little beyond the middle, straight or slightly
depressed before the apices; in ventral view divergent in the
proximal third, thence converging to the small squarely cut
apices; breadth of enclosed space about one-third of the length.
Ventral surfaces of valves sulcate throughout, most deeply
towards the apices. Styli rather more than half as long as the
dorsum of Io (.66 mm.). Basal plate of ovipositor of moderate
size, posterior margin straight, lateral plates distinct. Ap-
pendages slightly shorter to slightly longer than 8+9, gradu-
ally expanding from the slender base to three-fifths or two-
thirds the length, where the breadth is about one-fifth of the
length, thence narrowing somewhat tothe rounded apices,
which only rarely have a minute terminal tooth. Inner mar-
gin more strongly curved than the outer margin.

This species may be separated into four geographical

races by the following table:
A. Lateral thoracic bands each divided into a superior and
inferior spot (in the female sometimes only strongly con-
stricted); superior appendages of the male without any in-
dication of a ventro-internal basal tubercle; their apices
generally without a terminal tooth; appendages of the female
typically as long as, or longer than, 8+9, their apices broadly
SENT Nha hee As Uae... do al tnterrupta.
AA. Lateral thoracic bands not divided.

102 WALKER: NORTH AMERICAN SPECIES OF AESHNA

B. Superior appendages of the male without a ventro-
internal basal tubercle.
C. Lateral thoracic bands about I mm. broad near the
lower end, somewhat narrowed at the middle; inner mar-
gin of superior appendages of the male in profile slight-
ly concave before the middle, in dorsal view strongly
sinuate, the breadth of the appendage increasing
rapidly from the basal fourth, so that at the middle it
is fully twice as broad as at base, thence scarcely
narrowing to the well-rounded apices, which bear, near
the outer margin,a prominent tooth; inferior appen-
dage slightly more than half as long as the superior
appendages. Female unknown...........mevadensis.
CC. Lateral thoracic bands linear, their breadth rarely
more than .75 mm.;inner margin of superior appen-
dages of male in dorsal view very gently sinuate, in
profile straight or slightly convex before the middle,
the inferior carina rather low, its angle rounded or
obsolete; breadth of appendage at middle scarcely
twice that at the extreme base, thence narrowing
gradually to the rounded or bluntly angulate apices,
which are normally without a terminal tooth; superior
carina but little elevated, finely denticulated or smooth,
inferior appendage three-fifths to two-thirds as long as
the superior appendages. Appendages of the female
usually somewhat shorter than 8+9, their apices
less broadly rounded than in tnterrupta....... lineata.
BB. Superior appendages of the male with a low
basal tubercle on the ventro-internal surface ; inner
margin in dorsal view gently sinuate, in profile
slightly concave beyond the middle, beyond which
it forms a prominent, more or less obtusangulate,
inferior carina; breadth at middle about twice that
of the extreme base, thence narrowing gradually to
the more or less acute apices, which terminate in a
small spine ; superior carina moderately elevated
before the apex, with a few minute denticles ; lat-
eral thoracic bands about 1 mm. broad below, nar-
rower above ; appendages of the female as in lineata.
To 5 ahr ERR «: w'-3 <a Sa ge ee a ee oe ee interna.

WALKER: WITH AMERICAN SPECIES OF AESHNA 103

Aeshna interrupta interrupta Walker.

(PL 1, figs..1, 5; pl. 13, fig. 1; pl. 15, figs. 5, ‘5a;,pl., 20,
figs...1, Ia;) pl. 22, figs, 6-8; pl. 24, figs. 325.4.)
Aeschna propinqua, Scudder, Pr. Bost. Soc. Nat. Hist., X, p. 215 (1886) [in
‘ap
Aeschna clepsydra, Calvert, Ent. News, V, p. 11, fig. 5 (1894); Walker, 36th
Ann. Rep. Ent. Soc. Ont., p. 69 (1906) [in part] ; Willisiiasn, Ent. News, XVII,
Pp. 133-135 (1906) fin part]; Walker, Can. Ent., XX XVIII, p. 50 (1906) [in part].
Aeschna ‘‘w’’, Williamson, Ohio Nat., VIL, p. 146 (1907); Walker, Ott.
Nat., XXII, p. 55 (1908).
Aeshna interrupta, Walker, Can. Ent., XL, pp. 381, 387, 450; pl. 10, fig. 3

(1908); Muttkowski, Cat. Od. N.A., p. 11 (1910).
Aeschna tnterrupta, Martin, Gen. Ins., CXV, p. 11 (1911).

Occiput pale dull yellow, somewhat broader than long,
two-thirds to one-half as long as the line of contact of the
eyes; frontal vesicle greenish yellow, broadly margined with
black, the pale area not reaching the lateral ocelli; eyes vary-
ing in colour from olivaceousgrey witha bronzy lustre to green
or bluish grey with green reflections, a dark curved trans-
verse streak above, posterior border yellow or green, preocular
band about twice as broad at base of antenne as one of the
lateral ocelli, continued ventrad on the sides to the lower
margin of the nasus where it is scarcely half as broad as upon
the frons. T-spot heavy, 2.8-3 mm., stem .5-I mm. in front,
I-1.4 mm. behind, sides straight or a little convex. Face
yellowish green to olive, pale yellow on the sides and dorsal
surface of frons. A heavy black line on the fronto-nasal
suture, narrowing laterally to a fine line. Lateral lobes of
nasus obtusangulate, more or less rounded, not flaring.
Rhinarium dull greenish yellow (reddish brown in dried
specimens). Labrum pale green, margined narrowly above,
more broadly below, with black. Labium pale yellowish,
obscured with reddish brown; lateral lobes dull bluish or
plumbeous. Rear of head black.

Thorax dark olive-brown, becoming more or less trans-
parent in dried specimens, owing to the thinness of the cuticle
and the shrinkage of the tissues away from it; the clothing of
hair moderately thick, much longer below the apex of the
median carina than above it. Dorsal bands reduced to a
pair of lanceolate green spots about 2-2.5 mm. long and some-
what less than 1 mm. broad, no spots in the antealar sinus and

104 WALKER: NORTH AMERICAN SPECIES OF. AESHNA

no humeral spots. Lateral bands each divided into a superi-
or blue and an inferior yellowish green spot, surrounded by a
diffuse dull blackish margin; the inferior spots ovate, about 1
mm., or somewhat less, broad, the anterior somewhat broader,
the posterior somewhat narrower than the corresponding
superior spot. There are also two small yellowish spots be-
tween the lateral ‘‘bands,’’ one just below the spiracle, the
other, a mere dot, farther dorsad.

Femora dark reddish brown, black beneath, the anterior
pair often with a pale streak on the basal third or half of the
outer surface, tibiae and tarsi black.

Abdomen moderately slender, being, on the whole,
slightly broader than in Ae. canadensis, clepsydra and verticals,
narrower than in Ae. constricta. Structural features, accessory
genitalia, etc., as given under “‘specific characters.’

Superior appendages generally a little shorter than 9 + 10,
sometimes slightly longer, slender in the proximal fourth, the
basal breadth of which is about one-third of the intervening
space, thence expanding mesad to about two-fifths of the
length, where the breadth is about twice that at base, and
narrowing again very slightly tothe broadly rounded apices,
which are sometimes armed with a minute projecting tooth;
outer margin gently and evenly convex, inner margin rather
strongly sinuate; superior carina in dorsal view mesad of the
middle line, obsolete in the basal fourth, gradually elevated
distally, where it bears 2-6 minute denticles. Outer margin,
in profile, curved gently upwards, inner (lower) margin
feebly concave before the middle, beyond which it forms a low,
somewhat obtusangulate, inferior carina. Inferior appen-
dage about three-fifths as long as the superior appendages and
twice as long as the basal breadth, triangular with straight
lateral margins and bluntly rounded apex, surmounted by a
pair of small dorsal tubercles; in profile moderately and evenly
curved, rather broad in the middle but tapering considerably
to the apex.

Colour-pattern of abdomen—Seg. 1 brown, dorsal spot
well developed, blue; lateral spot green, 1-2 mm. long (trans-
verse), .5 mm. or less broad.

Seg. 2, brown; AD a minute yellowish dot, AML a blue

WALKER: NoRTH AMERICAN SPECIES OF AESHNA 105

band about .7 mm. broad, continuous above with MD, pass-
ing ventrad along the transverse carina to the auricle, then
bending cephalad to the posterior margin, where its breadth
is about 1.5 mm. MD green ; PL confluent with PD, forming
a broad blue band covering two-thirds of the posterior di-
vision of the segment above the auricle, the anterior margin
deeply excavated laterally and indented on each side of the
middorsal line, along which a narrow streak runs cephalad
nearly to the front margin.

Segs. 3-10 brownish black; MD green, the other spots
azure blue (PD on 10 sometimes pale greenish). AL
on 3-8; on 3 separated above at base from its fellow by
somewhat more than I mm., the upper margin pass-
ing obliquely downwards so that the breadth at the
transverse carina, where it ends, is only about I mm.; on 8a
mere dot; on the remaining segments less than .5 mm. broad
and connected with, or narrowly separated from, a narrow
basal ring; upper margins oblique, lower margins indented.
ML on 3-8, of moderate size, irregularly quadrate, rounded
behind on 6-8, front margins angularly indented. MD dis-
tinct and comparatively large on 3-7, generally represented by
a pair of dots on 8, on 3 broadly connected with ML, on 4-7
triangular, the postero-lateral margins angularly excavated.
PD about 1 mm.long on 3,increasing to 1.75 mm. on 7 and 8,
slightly smaller on 9, inner margins on 3-5 or -6 rounded, on
the remaining segments more or less straight, becoming widely
separated (Imm.) on 8-10, and somewhat divergent cephalad
on 8 and g. PL on 3-4, -5 or -6, small, more or less indistinct
posteriorly; on 3 broadly confluent with PD, on 4 narrowly
confluent or separate,on the other segments always separate.

Wings hyaline, costal veins pale brownish yellow, pter-
ostigma dark smoky brown above, pale greenish brown be-
neath, membranule smoky brown with the basal fifth or
third white, extending to the cross-vein of the anal triangle.
One or 2 cells between A2 and A3at their origin. Rs forking
at the level of 1-2 postnodal cells before the stigma, 3 or 4 rows
of cells between the forks at the level of the distal end of the
stigma, and 3 to 5 cells between the forks at the wing-margin.
Four or 5rowsof cells between Rsand Rsplat thelevel of widest
separation.

8

106 WaLKER: NortTH AMERICAN SPECIES OF AESHNA

Antenodals a at postnodals oa CuCr 3 7 Spt pls

I-16 10-15, ar 2-3

Female—The abdomen, which is slightly shorter than in

the male, varies considerably in form, especially in the length,
depth and amount of constriction of seg. 3 (vide postea).

Appendages as long as, or slightly longer than, 8+9
(slightly shorter in specimens from Newfoundland and the
Magdalen Islands), outer marginnearly straight, inner margin
evenly curved, apices broadly rounded, without a terminal
tooth.

Dorsal thoracic bands still more reduced than in
the male, being minute indistinct spots, often wholly
invisible in dried specimens and perhaps also in living ones.
Spots representing the lateral thoracic bands somewhat larger
than in the male, those of the first band often barely connect-
ed. PD smaller than in the male, especially on the anterior
segments, not confluent above on 2, and separated on 8 and 9
by I-1.5 mm. PL on 3-6, much larger than in the male, con-
fluent throughout with PD, and on 2 and 3 sometimes also
with AL.

The range of colour variation in this species is complete,
but the pure homceochromatic female seems to be rare. Most
of the individuals I have observed in the field were interme-
diate forms, but examples of extreme heterochromatism are
not at all uncommon.

In the common intermediate form most of the light
markings are more or less greenish yellow, the lateral spots of
the abdomen often somewhat bluish, as shown in the speci-
men figured (pl. 24, fig. 4).

In aheterochromatic female from Nipigon, Ont., the
coloration noted just after death was as follows: Thorax and
abdomen dark olive-brown. Eyes deep yellowish green,
with a yellow hind border; face ochre-yellow. Spots of thor-
ax and abdomen bright cadmium-yellow, tinged with green.
ML on 5-9 and PD on 8-10 pale dull greenish.

In another specimen from the same locality the spots of
the abdomen were all yellowish olive.

WALKER: NorTH AMERICAN SPECIES OF AESHNA 107

I have not seen a female with deeply flavescent wings
such as are sometimes met with in the races /ineata and interna
but sometimes a yellowish cast is present and it is probable
that deeply flavescent wings sometimes occur.

Measurements (exclusive of specimens from Newfound-
land and the Magdalen Islands) :

Thor. o& 10-11.5, 9 10.5-11; abd. o& 45.3-51, 9 47-51.5;
seg. 3 co’ 8.1-9.5, 9 7-9; apps. o15-5.6, 9 6.5-7.5; gen. V. 2.25-2.5;
hw. o 43-47.5, 9 46-49; width hw. o 12.5-13.6, 9
13.6-14.33; pter. co 3.33-4, 9 3.75-4.5-

Measurements of specimens from Newfoundland and the
Magdalen Islands—(1¢', 49). Thor. o& 11, 9 10-10.5; abd.
o' 49, 2 45-48; seg. 3 co 8.6, 9 6.75-7.5; apps. o' 5 mm., 9
5.25-5.66; gen. v. 2.33-2.66; h.w. o&' 47, 2 44.5-46.5; width
h.w. o' 14, 9 13.5-14; pter. co 3.5, 2 4.45.

Types—o' 9 , coll. Walker—Nipigon, Ont.

Nymph (pl. 6 fig. 2; pl. 8, fig. 2; pl. 10, fig.2) As mentioned
on p. 98 two unknown Aeshna nymphswere found at Nipigon,
Ont., which were referred to Ae. subarctica and Ae. interrupta
respectively. The form described below is referred to inter-
rupta (1) by elimination of all other species occuring at Ni-
pigon; (2) because it comes next tothe nymph of eremita in
point of numbers, as is the case with the adults of these species;
(3) because of its close resemblance to the nymph of eremuta,
the nearest relative of interrupta.

Eyes shaped as in juncea but a trifle more prominent;
lateral margins of head slightly arcuate, postero-lateral
angles broadly rounded, posterior margin feebly concave.
Mentum of labium slightly narrower at base than in Ae.
juncea, the basal breadth equal to about three-fifths of the
apical, the latter equal to about five-sevenths of the length ;
proximal three-fifths slightly but distinctly widening distad,
the sides straight or very nearly so, sides of distal two-fifths
strongly arcuate; middle lobe somewhat broader but less
produced than in Ae. juncea, broadly obtusangulate; lateral
lobes scarcely narrowed distad, squarely truncate, the outer
apical angle little or not at all rounded, the inner angle with a

108 WALKER: NortTH AMERICAN SPECIES OF AESHNA

minute tooth. Supracoxal processes moderately prominent,
the anterior slightly longer and more slender than the poster-
ior, the interval a little less than a right angle. Abdomen
broadest at seg. 6, shaped as in juncea, seg. 9 nearly three
times as broad as long; lateral spines present on segs. 6-9, not
at all divergent on 8 and 9g, extending towards the posterior
margin of the corresponding segments as follows : on 6 one-
third to two-fifths of the distance, on 7 three-fourths of the
distance, or quite, to the margin, on 8 a little beyond the mar-
gin, and on 9g as far as the proximal three-fifths of 10. Lateral
appendages slender, a little more than half as long as the in-
ferior pair, basal part of superior appendage of male some-
what more elongate than in juncea, the basal breadth equal
to about three-fourths of the length, sides distinctly concave,
apex rounded. Genitalia of female not quite reaching the
apical margin of 9, genital valves about three times as long as
broad, more steeply declivent than in juncea.

Head behind the eyes with a pair of roundish scars
surrounded by a pale margin, a dark submarginal blotch
on each side and a broad pale marginal. band, extending
from the eyes to the posterior margin of the pronotum.
Dorsum with a pale sigmoid streak in front of the bases of the
hind wing-pads. Femora brownish, with three pale rings, a
basal, a median and a preapical; tibiae and tarsi almost con-
colorous. Abdomen brown, varying much in depth of shade
and in the extent of the pale mottlings, the arrangement of
which, when not greatly obscured, is as follows: dorsal sur-
face with a series of pairs of pale spots, largest on the-anterior
segments and generally running together on the first three or
four segments to form two irregular bands, but separate on
the posterior segments and confined to the basal part of each
segment; another series of subcrescentic pale blotches, just
laterad of the dorso-lateral punctae, diminishing in size
caudad and disappearing at seg. 7 or 8; a series of large sub-
marginal pale blotches enclosing the lateral scars, which ap-
pear outlined in brown and touching the lateral punctae; and
a series of pale subovate median dorsal spots, very small on
the basal segments but increasing in size to seg. 7 and absent
beyond. There is also on several of the middle segments, in

WALKER: NORTH AMERICAN SPECIES OF AESHNA 109

front of the transverse carina and the lateral scars, a pair of
whitish patches alternating with darker ones. Punctae dark
brown, the dorsal series sometimes obscured by dark brown
blotches. Spines pale, with the apices dark.

Measuremenits—Length of body 39-41.5; mentum of
labium 5.9-6.4; h.w. 9-10; h.f. 6.5-7.2; inf. apps. 4.5-4.9;
gen. v. 2; width of head 8-8.3, of abdomen 7.25-8.

Material determined — 46 oc 31 9. NEWFOUNDLAND: Grand Lake, July
pate ae (D. A. Atkinson, coll. Williamson, Io); Bay of ewer July 7, E70!
(D. A. Atkinson, coll. Williamson, 2 9). Nova Scotia: (M.C.Z., 1 9, Bri
Mus., I <). MAINE : Portland, Aug. 9, 1870 (Jones, M.C.Z., 1 9). Viewuont
Aug., 1907 (Ellis Frost, coll. Williamson, Io’). MASSACHUSETTS: (Uhler, M.C.Z.,
I o'). QUEBEC: Grand Entry, Magdalen Is., June 30, 1901 (D. A. Atkinson, coll.
Williamson, 2 2 teneral); Chicoutimi, Aug. 23, 1901 (Calvert, I o”). ONTARIO:
Ottawa, Meach Lake, July 21, 1907 (A. Gibson, 1 9); Oxtongue Lake, Muskoka,
Aug. 22, 1904 (Walker, 1 9); Algonquin Park, Aug. 14,1903 (Walker,2 @'1I ?);
Temagami Forest Reserve, Lake Obabika, Sept. 11, 12, 1908 (Walker, 8 @ 3
9); Temagami, Aug. 15, 1906 (P. Hahn, 1 o); Sudbury (J. D. Evans, 2 o);
Sault Ste. Marie, Algoma, Sept. 18, 20, 1907 (Donaldson, coll. Williamson, 26);
Heyden, July 31-Aug. 4, 1906 (Williamson IIo? 4 9) ; Searchmont, Aug. 6-9,
1906 (Williamson, I1o'29); Nipigon, Aug. 28-30, 1907, Aug. 6-8, 1910 (Walker
7072); Fort William, Aug. 2, 1910 (Walker, 16°19). NEW York: Cats-
kill Mts., Aug. 28 (Acad. N.S. Phil., 17); Lake St.Regis, Franklin Co., Sept. 5-22,
1890 (J. P. Moore, Acad. N.S. Phil., 16° 2 9). MucuiGAn : Isle Royale, July
13, 1905 (Gleason, Coll. Univ. Mich., 1 9); Minnehaha Falls, Emmet Co., Aug.
23, 1907 (Williamson, 10’); Douglas Lake, Topinabee, Aug. 20, 1910 (Miss A.
O’Brien, 1c’). British COLUMBIA: Inverness, Aug. 1890 (Brit. Mus., 2 @ 2

4 Nymphs—Nipigon, Ont., Aug. 5-8, 1910, Ic" (St. F); exuviae, 4 719.

Identity—Scudder’s description of Ae. propinqua was
evidently based partly upon specimens belonging to this
species, as the lateral thoracic bands are described as some-
times divided each into an upper and lower spot. The types
of propinqua now in existence do not, however, include any
specimens of znterrupta, but belong to juncea and verticalis. I
have therefore let the name propinqua remain in the synony-
my and have described the present species under the new name
interrupta.

Distribution—The race interrupta is a plavactewetic in-
habitant of the Canadian Zone east of the Great Plains. It
occurs only occasionally in the Transition Zone, except in its
most northern parts, and is apparently absent from the Upper
Austral Zone. It is not known north of the limits of the Can-
adian Zone. The occurrence of this eastern race in the ex-
treme northwestern part of British Columbia needs confir-
mation. The four specimens in the British Museum from

110 WALKER: NORTH AMERICAN SPECIES OF AESHNA

this locality are, however, not quite typical in the form of the
superior appendages of the male, these structures*approach-
ing those of the races lineata and interna. It is thus possible
that the distribution of the British Columbia form is not
continuous with that of typical znterrupta.

Geographical variations—Considerable variation exists
among females in the length and depth of the third abdominal
segment and in the length of the appendages, and these vari-
ations are dependent to a great extent, though not wholly,
upon locality. These structures are shortest and the third
segment deepest in specimens from Newfoundland and
the Magdalen Islands, longest in a female from Ottawa, Ont.
(pir; fig. Te):

Females from the two former localities also differ from
the typical form in the coalescence of the spots on the sides
of the thorax and in the specimens from the Magdalen Islands
the lateral bands scarcely differ from those of the western
race, interna (vide pl. 22). Of the two females from this lo-
cality, both tenerals, the appendages are present in only one
and differ from those of any other specimen of the race 7n-
terrupta that I have seen in their slenderness and the acuteness
of the apices. It may be that this form is a local race. The
appendages of both male and females from Newfoundland are
typical in form and the spots of the lateral thoracic bands are
separate in the male.

From Nova Scotia westward to Fort William, on the
north shore of Lake Superior, no geographical variations could
be detected, except those which are apparently the effect of
temperature. The lateral thoracic bands are constantly di-
vided in the male and the second band also in the female.
The British Columbian specimens, as already mentioned,
differ only very slightly in the form of the male appendages,
which approach those of interna and lineata. In one of the
males the terminal tooth of the superior appendage is dis-
tinct though very small, in the other it is barely indicated.

Habits—This species flies over shallow waters grown up
with reeds, in bushy pastures, etc. I have observed nothing
in which its habits differ from those of juncea, canadensis, etc.
Mr. Williamson, however, has taken this species flying over

WALKER: NorRTH AMERICAN SPECIES OF AESHNA III

water after sundown, a habit which I have observed only in
Ae. umbrosa.

Ae. interrupta can be distinguished in flight from the
other species with which it is associated by the dark sides of
the thorax.

Aeshna interrupta nevadensis Walker.

(Ph 15; figs..6, Ga; pl. 22,0fig.!0:)
Aeshna nevadensis, Walker, Can. Ent., XL, p. 382, 451 (1908) ; Muttkow-
ski, Cat. Od. N.A., p. 113 (1910).

Male—Similar in form to the race interrupta, differing
chiefly in the form of the lateral thoracic bands and the su-
perior abdominal appendages.

Colour and markings of the head similar to those of
interrupta but the black line on the fronto-nasal suture is
narrower than is usual in znterrupia and in one specimen is
very narrow.

Hairs covering the thorax somewhat paler than in in-
terrupta, dorsal bands pale yellow, sometimes well defined,
about 2 mm. long and .3 mm. broad. Lateral thoracic bands
not divided, nearly straight, pale yellow below, bluish above.
First band rounded at the lower end where it is about I mm.
broad, narrowing to.5 mm., or less, at the middle, and re-
maining about this width or expanding again slightly to the
upper end. Second band .66-1 mm. broad below, generally
constricted about the middle and expanding again to about
I mm. at the upper end.

Colour-pattern of abdomen nearly identical with that of
interrupta. MD on segs. 3-5 tends to be larger and to lose
the triangular form and on 3, and sometimes also on 4, may be
broadly confluent with ML. AL and ML are also sometimes
confluent on 3. AL is narrowed as in znterrupta but is not, or
less distinctly, indented below. PL present on 3-7, connected
throughout with PD, of which it forms a falciform ventral
offshoot.

Superior appendages somewhat shorter than 9+10, dif-
fering from those of interrupta in the somewhat greater
length of the slender basal portion, which is nearly or quite
one-third of the length of the appendage, in the more strongly
and angularly sinuate inner margin and consequently more

112. WALKER: NoRTH AMERICAN SPECIES OF AESHNA

rapid expansion of the broad distal part of the appendage,
which is slightly broader and less tapering distally than in
interrupta, the apices being more broadly rounded and term-
inating in all the specimens examined in a minute tooth.
Superior carina with 3-5 minute teeth. In profile view the
superior appendages are practically identical with those of
interrupta. Inferior appendage about three-fifths as long as
the superior appendages, similar to that of interrupta.

Wings similar tothoseofinterrupta except that the mem-
branule is uniform greyish brown. Rs forking at the level
of 1-2 postnodal cells before the stigma; 3 rows of cells be-
tween the forks at the distal end of the stigma, and
3-4 cells between them at the margin. Three or 4 rows of
cells between Rs and Rspl where most widely separated. Mta
arising beyond the middle of the stigma, generally under the
distal end. Two cellsbetween A2 and A3 at their origin.

Antenodals 29’, postnodals 2) Cur 38. Spt. Ei
12-13 II-13 4-5 I-2

Female unknown.

Type— oo, Mus. Comp. Zoology—Reno, Nevada.

Measurements (male)—Thor. 10-10.5; abd. 46-48; seg. 3
8-9; apps. 4.8-5; h.w. 45-47; width h.w. 13.6-14; pter.
3-3-5-

Nymph—Unknown.

Material determined—NEVADA: Reno, 1878 (H. K. Morrison, M.C.Z., 8c).
Only two of these, kindly lent me by Mr. Samuel Henshaw, were studied in de-

tail, and the venational characters given above are based on these two speci-
mens alone.

Aeshna interrupta lineata Walker.
(PL46, fies. 1,.1a; pls 24, dies.5) 60

Aeshna lineata, Walker, Can. Ent., XL, pp. 382, 388, 450 (1908); Muttkowski,
Cat. Od. N.A., p. 112 (1910).

Male—Head as in race interrupta, the frontal vesicle a
little shorter, the naso-frontal line and the black lateral
margins of the race generally a little narrower.

Thorax slightly heavier than in interrupta, colour similar
but appearing paler on account of the longer and paler
growth of hair, which gives it a greyish brown appearance.
Dorsal bands generally even more reduced or entirely absent.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 113

Lateral bands linear, nearly straight, greenish yellow below,
more or less bluish or blue above, the first band .5 mm., or
less, broad at the lower end, tapering dorsad to a fine streak;
second band still narrower below, somewhat more oblique and
gently curved backwards, sometimes interrupted below the
middle.

Abdomen slightly shorter and stouter than in interrupta ;
auricles with 3 teeth (occasionally 4); accessory genitalia as
described under “specific characters.’’

Superior appendages barely shorter than the dorsa of
9+10, showing no very constant differences from those of 1-
terrupta, but typically darker in colour, expanding more
gradually from the base to three-fifths of the length, where the
breadth is generally somewhat less than in interrupta; apices
rounded, usually smaller than in znterrupta, sometimes with a
very minute terminal tooth; curve of outer margin stronger,
inner margin less sinuate, denticles of superior carina usually
very minute and occasionally absent altogether ; appendage
in profile more strongly curved upwards. Inferior ap-
pendage three-fifths to two-thirds as long as the superior
appendages, the apex smaller than in interrupia, otherwise
similar in form.

Colour-pattern of abdomen similar tor that of interrupta
but the blue spots are somewhat larger.

Seg. I as in interrupta.

Seg. 2. AML broader than in znterrupta and less distinctly
divisible into vertical and horizontal parts, produced ventrad
along the front margin to the ventral surface; PL and PD
shaped as in interrupta but slightly broader.

Segs. 3-10. AL generally larger throughout than in in-
terrupta, broadly connected on 3, and sometimes narrowly
on 4. PD as in interrupta, or slightly larger on the anterior
segments. PL on 3-5, -6 or -7, falciform, connected behind on
each segment with PD.

Hind wings about aslong as the abdomen, the breadth
averaging somewhat greater than in interrupta; membranule
brownish grey, generally paler than in interrupta, basal fifth or
third whitish. Oneor2 cells between A2 and A3 at their origin.
Rs forking at the level of 1 postnodal cell (in hind wing sometimes

114. WALKER: NorRTH AMERICAN SPECIES OF. AESHNA

2) before the stigmaor behindits proximalhalf. Three or 4rows
of cells between the forks of Rs at the level of the distal end
of the stigma and 2-4 cells (rarely 5) between them at the
margin; Mta arising ator beyond the level of the distal end
of the stigma pace in 8o per cent. of cases).

Antenodals 142° -———, postnodals at CuCr 27 Spt =
9-13 9-12’ 4-6 2-4

Female—The differences in form exhibited by the females
of lineata and interrupta correspond with those of the males.
Appendages usually slightly shorter than 8+ 9 (a little longer
in I 9 from Regina, Sask.) |Apices somewhat less broadly
rounded than in interrupta, often with a feebly indicated
apical tooth, otherwise as in the latter race.

Dorsal thoracic bands generally absent, sometimes in-
dicated by a minute and usually indistinct spot. Lateral
bands as in the male or a little broader. PD some-
what smaller than in male, not confluent on 2. PL much
larger than in male, generally confluent with AML on 2 and
frequently with AL on 3.

Ground colour of abdomen light reddish brown, darker
about the margins and pale areas, the latter varying greatly
in colour, sometimes closely resembling those of the male,
sometimes strongly heterochromatic. In a Winnipeg ex-
ample of the latter type, the face is rather dark olivaceous,
lemon-yellow laterally and on the upper surface of the frons;
the dorsal thoracic spots are distinct though minute and are
chrome-yellow, as are also the lateral bands and the spots be-
tween the wings above. The abdominal spots are somewhat
discoloured except those on 1 and 2, which are yellow. The
other abdominal spots were apparently of the same colour. I
have another similarly coloured female from the Swan River,
Man., and both of these individuals are remarkable for their
flavescent wings, a character which I have seen in no other
specimens of this race.

Measurements —Thor. o& 10.25-11, 9 10-11; abd. o& 45-
48, 9 44-46 ; seg. 3 o’ 8.5-9.3, 9 7-7.5 ; width of seg. 2 of
5, 9 5-5.8; apps. co’ 4.5-5, 2 4.5-6.5 ; gen. v. 2.3-2.6; h.w.
S 44-46, 9 44.5-46; pter. co 3.5-4.25, 9 3.75-4.5; width of
h.w.o’ 13.5-15, 9 14.2-15.

WALKER: NorTH AMERICAN SPECIES OF AESHNA II5

Types—o 2, coll. Walker—Regina, Saskachewan.

Nymph (?)—In the Cabot Collection in the Museum of
Comparative Zoology, Cambridge, are two Aeshna nymphs
that differ somewhat from any other nymphs that I have seen.
They are labelled ‘‘Mouth of the Red River of the North,
Scudder, 1860,” and were originally preserved in alcohol, but
are now dry and somewhat shrunken and the colour-pattern
has practically disappeared. They seen to be nearest in-
terrupta interrupta, though I was unable to compare them
directly with specimens of that form. They differ principally
in the presence of small lateral spines on segment 5 and in the
somewhat larger labium.

As these nymphs come from the territory inhabited by
Ae. interrupta lineata, it is probable that they belong to this
race. They are in any case a member of the clepsydra group;
and the other members of this group, except Ae. eremita and
canadensis, whose nymphs are known, are not known from
the Red River country and probably do not occur so far north.

Eyes rather prominent, antero-posterior diameter slightly
shorter than the posterior margin; lateral margins of head
nearly straight, two-fifths as long as the posterior margin,
which is considerably emarginate; postero-lateral angles well
rounded. Labium large, apparently extending somewhat be-
hind the bases of the middle legs; proximal margin of mentum
a little more than half as broad as the distal margin, which is
equal to about five-sevenths of the length; sides subparallel in
the basal two-thirds, then diverging, the distal third strongly
arcuate; medium lobe short, feebly arcuate; lateral lobes
broad, squarely truncate, internal apical angle with a some-
what prominent tooth. Abdomen broadest at seg. 6. Later-
al spines present on 5-9; on 5 reaching half way to the margin,
on 6 two-thirds of the way, on 7 about to the margin, on 8
beyond the margin by two-thirds of their length, on 9 reaching
just beyond the posterior margin of 10 (probably a little be-
yond the middle in exuvia). All the spines are curved slight-
ly inwards, the curve continuous with that of the margin.
Lateral appendages ( 2) barely more than half as long as the
inferior pair. Genitalia not quite reaching the apical mar-

gin of 9g.

116 WALKER: NorTH AMERICAN SPECIES OF AESHNA

Colour-pattern indistinct, probably like that of interrupta
and eremita. Femora with broader sub-basal and narrower
preapical dark dorsal (not seen distinctly in @).

Measurements (?)—Length of body 36 (shrunken) ;
mentum of labium 7; h.w. 10; h.f. 8; width of head 9; of ab-

domen 8. .

Material determined—30 ¢& 30 9. NortH Dakota: Aug. 13 (Coues,
M.C.Z., 1 &); La Roche Percée, Aug. 16, 1873 (Dr. Mack, M.C.Z., 1 o); Fargo,
June, 13, 1902 (N.G. Orchard, coll. Williamson, 1 9). MAnitoBa: Winnipeg,
July 6-24, Sept. 7, 1908 (J. B. Wallis, 5 @ 5 9); Winnipeg Beach, Lake
Winnipeg, Aug.-Sept. 6, 1909 (J. B. Wallis, 2 # 3 9); Westbourne, July
28-Aug. 19, 1908 (J. B. Wallis, 5 @ 6 9) ; Aweme, July 20, 1906, Aug. 10,
1907, Aug. 16, 1908 (N. Criddle, 2 co 2 9); Swan River, Sept. 8, 1906 (W.
J. Alexander, coll. Walker, 19). SAsKATCHEWAN: (Scudder, M.C.Z., 10°19);
Regina, July 18, 1905 (T. N. Willing, 4 @ 2 2); Carlton, July 28, 1900,
July 22, 1907 (J. Fletcher, T. N. Willing, 3 @, teneral) ; Duck Lake, July 22,
1907 (J. Fletcher, T. N. Willing, 2 <7, teneral); Goose Lake, July 21, 1907 (T.
N. Willing, 107); Parkside, July 24, 1907 (T. N. Willing, 1 o, teneral); Meota,
July 8, 1907 (T. N. Willing, 10° 19); Moose Jaw, Aug. 24, 1903 (A. N. Cau-
dell, U.S.N.M.,19). ALBERTA: Banff, Aug. 4, 1906, Aug. 16, 1908 (R. P. Currie,
U.S.N.M., 1 9, N. B. Sanson, 1 o) ; Waterton Lake, Aug. 7-10, 1908 (E. V.
Cowdry, coll. Walker, 2 @ 3 92). Nortuwest Territory: Great Slave
Lake, July 1862 (Kennicott, M.C.Z., 19). Also ‘‘Upper Missouri” (Hagen,
M.C.Z., 1 9); and ‘‘ British AMERICA” (Scudder, M.C.Z. 1, 7 19).

é Nonbhs (?)—‘‘ Red River of the North, Scudder, 1860” (M.C.Z.) 1 2910
(St. F).

Distribution—This race of Ae. interrupta is probably the
most characteristic dragonfly of the Canadian Prairies,
ranging over the whole of this region from Manitoba to the
foothills of the Rocky Mountains and northward into the
wooded country as far as Great Slave Lake. South of Canada
its range extends into North Dakota and to the Upper Miss-
ouri. It is thus characteristic of the Canadian Zone, in the
middle part of the continent, reaching the northernmost
limits of that zone; while southward it ranges well into the
Transition Zone.

Geographical variations—No variations of geographical
significance have been detected in this race, except where it
intergrades with the race interna (vide postea).

Aeshna interrupta interna Walker.

(PLMGG.«fies. 2, 2a; pl. 22. fesota, 51).
Aeschna interna, Hagen, Pr. Bost. Soc. Nat. Hist., p. 35 (1875) (no descrip-
tion); Martin, Gen. Ins., CXV, p. 11 (1911).
Aeschna clepsydra, Calvert, Trans. Am. Ent. Soc., X XIX, p. 43 (1903).
Aeschna juncea, Currie, Pr. Ent. Soc. Wash., VII, p. 16 (1905) ; Osburn, Ent.
News, XVI, p. 190 (1905).

~ WALKER: NoRTH AMERICAN SPECIES OF AESHNA_ 117

Aeshna interna,Walker, Can. Ent., XL, pp. 381, 388, 450 (1908); Muttkowski,
Cat. Od. N.A., p. 112 (1910).

Closely related to Ae. interrupia lineata with which it
intergrades where the territories inhabited by the two races
meet. It differs from lineata in the form of the superior ap-
pendages of the male and in the greater development of the
pale areas of the thorax and abdomen.

Head indistinguishable from that of lineata except that
the paler area of the frontal vesicle is generally larger, oc-
cupying nearly the whole of its upper surface, though not
reaching to the lateral ocelli.

Dorsal thoracic bands in the male 2-2.5 mm. long and .5-.6
broad; in the female 1.25 mm. long or less, very narrow and
sometimes indistinct or even absent; lateral bands blue above,
pale green or yellow below, somewhat variable in width, in
typical specimens distinctly broader than in lineata; first band
.75-I1 mm. broad below, narrowing to about half that width a
little above the middle, then remaining equal or expanding
slightly to the upper end. Second band of about the same
width below, curved ‘gently caudad, equal, or slightly con-
stricted at the middle and expanding slightly on the upper
half.

Abdomen similar in form to that of lineata. Superior ap-
pendages of male somewhat contracted at the extreme base,
the proximal three-tenths rather slender, equal; thence ex-
panding gently on the inner margin to about three-fifths the
length, where the breadth is somewhat less than twice that of
the slender basal part. Distal two-fifths narrowing somewhat
to the rounded or angulated apices which bear, externally, a
distinct spine. Outer margin curved as in lineata, more
strongly than in interrupta, inner margin more distinctly
sinuate than in the former. In profile the superior carina ap-
pears more elevated apically than is usually the case in either
of these races, the denticles very minute or absent, outer
margin less curved upwards, lower (inner) margin sinuate,
being proximally slightly convex, concave before the middle,
thence produced ventrad to form a somewhat prominent ob-
tusangulate inferior carina, which is usually deeper than in
interrupta and lineata. A rather low sub-basal ventro-in-

118 WALKER: NORTH AMERICAN SPECIES OF AESHNA |

ternal tubercle appears in an oblique view from above or be-
low.

Appendages of the female as long as the dorsa of 8+9,
or slightly shorter, shaped as in lineata, apices with a small
terminal tooth.

Colour-pattern of abdomen similar to that of lineata but
all the spots average somewhat larger.

Seg. 1. Lateral spot 2.5-3.5 mm. long, .5 mm. or less in
breadth. Seg. 2 asin lineata ; PL sometimes confluent with
AML in the female, sometimes separate.

Segs. 3-10. Spots all light blue, almost identical in form
with those of lineata; MD considerably larger and the others
averaging somewhat larger, though no good differential
character can be based upon them. AL and ML as in
lineata; MD larger, more or less quadrate in the male, at least
on 3-5, triangular in the female, generally broadly confluent
with ML on 3 and 4 in the male, separate throughout in the
female. PL larger than in lineata, represented on 3-8, and con-
fluent throughout with PD, becoming recognizable on 8 only
as the lower part of the latter. PD as in /ineata or barely
larger, about 1.25 mm. long on 3, increasing to nearly or quite
2 mm. on 7 and 8.

As in races lineata and interrupta, the female exhibits both
types of coloration. One female from Utah and one from Ore-
gon are apparently homceochromatic, or at least partially so,
with hyaline wings; while onefrom Baker City, Ore., is strongly
heterochromatic, all the pale markings being yellow, and the
wings, except beyond the pterostigma, strongly flavescent.

Measurements—Thor. o& 10.2-11, 9 10-10.5 ; abd. o& 46-
48, 9 42.5-46; seg. 3 o& 8-9, 2 6.8-7; apps. o& 4.8-5.25, 9
5-5-33; h.w. & 43-45, 2 42.5-44; pter. o& 3-3.8, 9 3-5-4;
width of h.w. o& 13-13.6, 9 13.4-14.

Types—o 9, Acad. Nat. Sc. Philadelphia—Lamb’s
Canon, Salt Lake Co., Utah.

Nymph—Unknown.

Material determined—32 3 22 9. British CoLtumBiA: (Crotch, M.C.Z.,
1 9); Kaslo, Aug. 7, 1903 (R. P. Currie, U.S.N.M., 1 o*). OREGON: Baker City
and vicinity, July 27, 30, 1908, Aug. 7, 9, 1909 (C. H. Kennedy, 5 o& 2 9 im
cop., July 30). WASHINGTON: Spokane, July 22, 1882 (S. Henshaw, M.C.Z., 1
9); Colville, July 23-25, 1882 (S. Henshaw, M.C. Z., 2 &@ 4 9). UTAB:
Ogden (Cyrus Thomas, M.C.Z., 1 &2 9); Summit Cavern, Aug, 8. 1875

WALKER: NortTH AMERICAN SPECIES OF AESHNA_ II9Q

(M.C.Z., 2 c& 8 9); Lamb’s Cafion, Wasatch Mts., Salt Lake Co., July 23,
1899 (Browning, coll. Calvert, 1 &@ 1 Q); City Creek Cajion, July 5, 1899
(coll. Calvert, 1 9). New Mexico: Beulah, Aug, 15. 1901 (coll. Calvert, I o).
Also forms intermediate between races interna and lineata as follows: AL-
BERTA: Banff, Sept. 6, 1906, Sept. 1, 10, 1908 (N. B. Sanson, 1 @ 2 9).
British CoLumBiA: Peachland, Aug. 4-21, 1909 (J. B. Wallis, 18 @).

Distribution—British Columbia to New Mexico, a
mountain form.

Geographical variations—All the intermediates between
this form and lineata, that I have seen, with the probable
exception of a single damaged male from Fort Collins, Col.,
come from British Columbia. No other geographical vari-
ations were noted, specimens from Oregon and Washington
being quite like those from Utah and New Mexico.

Aeshna eremita Scudder.
Ot, figs. 2. 6. plo. figs. 1-4;.pl13, fig. 2; pl. 46,
figs. 3, 3a; pl. 19, figs. 1, 1a; pl. 24, figs. 1, 2.)

Aeschna eremita, Scudder, Pr. Bost. Soc. N.H., X, p. 213 (1866); Hagen,
Pr. Bost. Soc. N.H., XV, p. 376 (1873); Harvey, Ent. News, II, pp. 73, 75 (1891);
Brown, Ins. Life, III (1891); Hagen, Rep. U.S. Geol. Surv. Terr., VI, p. 727
Pay Calvert, Trans. Am. Ent. Soc., XX, p.206 (1893); Ent. News, V, p. 9
(1894).

Aeshna eremita, Walker, Can. Ent. XL, pp. 383, 388, 451 (1908); Muttkowski,
Bull. Wis. N.H. Soc., VIII, p. 56 (1910); Cat. Od. N.A., p. 110 (1910).

Aeschna eremitica, Hagen, Pr. Bost. Soc. N.H., XVIII, p. 34 (1875); Cabot,
Mem. M. C. Z., VIII, p. 23, pl. 2, fig. 2 (aymph) (1881).

Aeschna hudsonica, Selys, Ent. M. Mag., p. 242 (1875); Martin, Cat. Coll.
Selys, XVIII, p. 35, fig. 30 (1908); Gen. Ins., CXV, p. 11 (1911).

Aeschna crenata, Bergroth, Ent. Nachr., VII, p. 86 (1881) [in part]; Calvert,
Ent. News, V, p. 9 (1894).

Aeschna clepsydra, Calvert, Trans. Am. Ent. Soc., XX, p. 248 (1893); Ent.
News, V., pp. 9-13, figs. 6and 7 (1894); l.c., p. 243 (1894); Currie, Pr. Wash. Acad.
Sc., III, pp. 217, 223, (1901); Ruthven, Ec. Surv. N. Mich., p. 101 (1906); Walker,
36th Ann. Rep. Ent. Soc. Ont., p. 69 (1906) [in part]; Walker, Can. Ent.,
XXXVIII, p. 50 (1906) [in part]; Williamson, Ent. News, XVII, pp 133-135
(1906) [in part].

Aeshna ‘‘x,’’ Williamson, Ohio Nat. VII, pp. 145, 146 (1907).

A species of large size and average or rather stout build.

Male—Occiput light yellow, black in front, equilateral
or somewhat broader than long, two- to three-fifths as long as
the line of contact of the eyes; frontal vesicle rather narrow,
emarginate in front, the anterior one- to two-thirds light
yellow, this pale area sometimes divided mesially into two
spots and not reaching the lateral ocelli. Eyes varying in
colour from pale greyish green with bluish reflexions to bright

grass-green, hind margin pea-green to yellowish green. Pre-

120 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

ocular band more than twice as wide, at base of antenne, as
one of the lateral ocelli; reduced to a narrow marginal line on
the sides of the frons and nasus, reaching the lower border of
the latter. T-spot 3-3.3 mm., regularly but not very strongly
arcuate in front, stem rather short, .4-.8 mm. in front, .66-
1.66 mm. behind, the sides straight and parallel or somewhat
divergent caudad. Frons and nasus light green or yellowish
green, pale yellow laterally and on upper surface of frons, naso-
frontal line black, very heavy in front but narrowing to a fine
line before the lateral margin. Lateral lobes of nasus slightly
flaring, rectangular, the apices rounded. Labrum greenish
yellow or light green, margined above and usually also below
with black. These marginal lines are generally narrow and
sometimes the lower one is reduced to a fine brown line, but
occasionally the lower one is the broader and in an example
from Temagami, Ont., it is very broad and confluent mesially
with the upper margin, thus dividing the pale area into two
parts. Labium lemon- to dull clay-yellow, lateral lobes blue-
green to “robin’s egg’’ blue. Rear of head black. .
Thorax rather robust, dull greyish brown with an oliva-
ceous tinge, greyish below; darker along thesutures and around
the lateral bands, dorsal carina and edges of antealar sinus
black. Dorsal bands bluish green, 4-5 mm. long, .75 mm.
broad at the middle, lower ends separated by about 4 mm.,
taper-pointed; upper ends separated by about 2 mm., some-
what expanded and truncated, about 1 mm. broad, the ex-
panded portion occasionally separated from the rest of the
band as a distinct spot; outer margin straight, inner margin
convexly curved. Lateral bands rather broad but irregular
in form, first band blue, often pea-green in the lower half or
margined below with green, rounded at the lower end, 1.75-2
mm. broad in the lower half, greatly constricted just above the
middle by the deep rounded excavation of the anterior margin,
the breadth here being generally less than 1 mm., thence ex-
panding again to I or I.5 mm. at the upper end, from which a
short posterior offshoot is given off, or may be separated as a
distinct spot; posterior margin gently sinuate. Second band
blue, hind margin sometimes green below, expanding from the
acutangulate lower end to the upper end, which is 2-3 mm.

WALKER: NortH AMERICAN SPECIES OF AESHNA I2I

broad; hind margin straight, front margin excavated in the
upper two-thirds, but less deeply than in the first band. A
small green or blue elongate spot just behind the constricted
part of the first band and one or two spots about the metastig-
ma, no pale humeral spot. Spots between the wings azure
blue.

Legs very dark reddish brown, the femora paler above;
anterior pair generally with a pale greenish streak along the
outer surface, extending from the base sometimes to the apex.

Abdomen 4.5 to nearly 5 times as long as the thorax, the
length and stoutness, especially of seg. 3, apparently
depending to some extent upon locality (vide postea). The
constriction at this segment is as usual in this group, the ab-
domen expanding behind it to the apex of 4, thence remaining
equal as far as 6, 7 or 8, narrowing slightly on the remaining
segments.

Ventral surface of 1 without a distinct tubercle, but
bearing numerous minute spines on the posterior half. Auri-
cles with 4 teeth. Tergal margins behind the hamuli some-
what elevated, straight or somewhat divergent. Spines of
anterior lamina well developed for the clepsydra group,
reaching as far back as the bases of the hamular processes,
straight, sharp-pointed, directed ventrad, caudad and laterad.
Hamular processes directed ventrad, mesad and cephalad,
rather short and thick, ventro-lateral surface somewhat con-
cave, internal margin thick, strongly convex, apices produced
into a distinct papilliform process smaller than that of ver-
ticahs, but larger than the apical tubercle of interrupta.
Hamular folds well developed, closely approximated mesially,
their lateral margins continuous with those of the hamular
processes, moderately thick, somewhat divergent caudad,
forming an acute angle with the well-rounded apices.

Seg. 3 about 2-7 times longer than 1+2o0r4. Lateral cari-
nae of 8 rather strongly angulate about the middle. Dorsum of
10 with a prominent median and two smaller lateral teeth.

Superior appendages about one-sixth shorter than 9+10,
basal breadth about one-third or two-sevenths of the inter-
vening space, expanding gradually and almost symmetrically
to about the middle,where the breadth is more than twice that

9

122 WALKER: NORTH AMERICAN SPECIES OF AESHNA

at base and about one-fifth of the length, and narrowing again
very slightly in the distal half to the broadly rounded apices
which only very rarely bear a minute terminal tooth; outer
and inner margins in dorsal view very gently and almost
equally curved, the latter not being at all sinuate. Superior
carina arising in a prominent basal tubercle, percurrent though
feeble in the proximal fourth, apical three-fifths ratherstrongly
elevated, arcuate, bearing 6 to 8 well-marked denticles. In
profile view the outer margin appears moderately curved,the
basal fourth of the appendage thickened by the presence of a
rather large but low and rounded sub-basal ventral tubercle,
whose surface bears a number of minute denticles. Inner
margin produced downward beyond the middle into a fairly
prominent rotundo-obtusangulate inferior carina. This car-
ina together with theelevated superior carina just above it gives
the appendage the appearance of being rather strongly bent
upwards in its apical third. Inferior appendage six-elevenths
to three-fifths as long as thesuperior pair and nearly three-fifths
as broad at base as long, triangular, apex blunt and rounded;in
profile moderately curved, tapering considerably in the distal
two-thirds.

Colour-pattern of the abdomen—Seg. 1 brown, dorsal spot
azure blue, lateral spot greenish blue or green, 2.5-3.5 mm. long
(transverse), .75-I mm. broad.

Seg. 2 brown ; AML blue, sometimes greenish yellow or
green in front, rhomboid, 2.5-3 mm. broad at front margin,
which is not produced below the level of the auricles, con-
fluent above with MD; PL and PD united, forming a blue
band, which is two-thirds (above), four-fifths (below), as
broad as the posterior division of the segment, the anterior
margin deeply excavated laterally and confluent above with
a fairly large median dorsal blue spot.

Segs. 3-10 brownish black, the spots rather large, azure
blue, except MD, which is dull yellow. AL on 3-8, average
size; on 3 separated by a dark brown dorsal band which is
about I mm. broad in front, 2 mm. broad behind; on 8 minute,
on the remaining segments rounded and not indented below,
straight above and connected with a basal ring, which be-
comes very narrow and sometimes obsolete at the middorsal

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 123

line. ML fairly large, being about twice as high on 6 and
7 as AL, quadrate, somew hat rounded behind on the posterior
segments, front margin on most of the segments with a small
oblique acute indentation. MD on 3-8, fairly large, con-
fluent with ML on 3, transversely elongate-triangular,
posterior margins concave. PD about 1.25 mm. long on
3, increasing to 2 mm. on 8, front and inner margins more
or less rounded on 3 and 4, nearly straight on the remaining
segments, the inner margins narrowly separate except on 8
and 9, parallel except ong, where they are divergent cephalad.
On I0 they are connate in about 65 per cent. of individuals, but
may be separated by I mm. PL on 3-9, straight, narrowly
connected behind with PD on all the segments.

Wings distinctly shorter than abdomen, hyaline, costal
veins brownish ochraceous, pterostigma smoky brown above,
ochre-yellow below, membranule of hind wing dark smoky
grey, not reaching the cross-vein of the anal triangle, which
is 2-celled. One or 2 cells (2 in 75 per cent.) between A2 and A3
at their origin. Rs in front wing forking at the level of 1-3
postnodal cells (2in65 per cent.) before thestigma, in hind wing
at the level of 1-4 cells (3in50 per cent.) with 3 (rarely 4) rows
of cells between the forks at the level of the distal end of the
stigma and 2-6 (usually 4 or 5) cells between the forks at
the margin; 3-6 (4 or 5 ing2 per cent.) rows of cells between Rs
and Rspl where most widely separated; Mra arising before
the middle of the stigma, sometimes before the proximal
end.

Antenodals 15°22, postnodals 22°, Cucr £2, spe +4.
II-15 Tosa 4-6 1-3
Female—Genital valves about as long as 9, lateral
carinae broadly rounded and swollen at base but elsewhere
sharp and fairly prominent, the apical half almost straight
in profile view, in ventral view subparallel or slightly diverg-
ing to about two-fifths of the length, then converging and
soon becoming subparallel, the greatest width of the enclosed
space being equal to about two-fifths of the length; ventral sur-
face of valves feebly sulcate, strongly declivent in the basal half,
squarely cut; in profile nearly rectangulate, not elevated,

124 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

bearing a minute pencil of pale hairs. Styli about .75 mm.
long; basal plate of ovipositor of usual size, posterior margin
straight; spinulose area on ventral surface of 10 rather large,
with numerous rather coarse spines. Appendages varying
much in length according to locality, sometimes slightly
shorter, sometimes a little longer, than 8+9 (vide postea),
much slenderer at base than apex, greatest breadth a little
beyond the middle, equal to one-fifth to two-sevenths of the
length; curve of inner margin stronger than that of the
outer margin; apices broadly rounded, sometimes subangulate
but without a projecting tooth.

The dorsal thoracic bands are reduced and divided
into an anterior elongate spot, 2-2.5 mm. long and .5
mm., or less, broad, and a posterior, small, more or
less transverse spot, corresponding to the expanded upper
end in the male. Often the posterior spot is indistinct or
absent altogether. Lateral abdominal spots somewhat larger
than in the male, PD slightly smaller. The size of the
lateral spots depends to a considerable extent on the depth
of the abdomen, which is deeper in northern than in southern
specimens (vide postea). PL is always connected with PD
but is nearly always separate throughout from ML, though
rarely narrowly confluent on 2.

In colour a complete range of variation occurs, from
the pure homceochromatic type to a marked degree of hetero-
chromatism. The majority of females that I have seen
were of the latter type, the pale markings being usually
yellowish green throughout, but in one example from Go
Home Bay, Georgian Bay, Ont., the lateral thoracic bands
as well as the abdominal spots (except MD) are blue, as
in the male. Heterochromatic females often have strongly
flavescent wings.

Measurements—Thor. o& 11.5-12.5, @ 10.5-12; abd.
oO 51.5-58.5, 9 48.3-52 ; seg. 3 co’ 9-11, 297-9 ; depth seg. 3
Q 2.4-4.9; apps. co’ 5.5-6.3, 95-7.75; gen. v. 2.2-2.5; h.w.
3 45-52, 2 46-49.5; width hw. o@ 13-14.5, 2 13.6-15;
pter. o& 3.5-4, 9 4-4.6.

Type— oo’, Mus. Bost. Soc. Nat. Hist.—Hermit Lake, N.H.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 125

Nymph (pl. 6, fig. 3; pl. 8, fig.3; pl. 10, fig. 3) Eyes as in
canadensis and juncea; lateral margins of head moderately
oblique, straight or nearly so, postero-lateral angles obtus-
angulate, sometimes rounded off a little; posterior margin
nearly straight. Mentum of labium about three-fifths as
broad at base as at apex, the apical breadth about seven-
ninths of the length, proximal three-fifths slightly expanding,
the sides very feebly arcuate, distal two-fifths much expanded
the sides strongly arcuate; middle lobe very little produced,
very broadly subarcuate; lateral lobes broad, equal, squarely
truncate, outer angle scarcely rounded, inner angle with a
minute tooth. Supracoxal processes somewhat longer than in
mlerrupta and juncea, about equal in length and similar in
form, rather slender, acute and somewhat curved away
from one another, the interval nearly rectangular. Ab-
domen broadest at seg. 6 or 7, slightly stouter than in juncea;
lateral spines present on segs. 5-9, usually more spreading
than in other related species but somewhat variable in this
respect, reaching back towards the posterior margin of the
corresponding segments as follows: on 5 one-fifth or one-
fourth of the distance, on 6 half way, on 7 three-fourths or
quite to the margin, on 8 to the margin or a little beyond,
on g as far as the proximal three-fifths of 10. Lateral ap-
pendages one-half (9), or slightly more than one-half (<7),
as long as the inferior appendages; basal part of superior
appendage of male about as broad at the base as long, about
four-fifths as long as the lateral appendages, lateral margins
feebly concave, apex bluntly pointed. Genitalia of female
not quite reaching the posterior margin of 9; genital valves
about two-fifths as broad as long, somewhat less steeply
declivent than in interrupta and canadensis.

In life the nymph is usually of a very dark greenish
colour with paler mottlings, the pattern of which is al-
most identical with that of <interrupta. As seen in the
exuviae the pattern is as follows :

Head behind the eyes with an oblique subtriangular
brownish patch on each side and a broad pale marginal band
extending from the eyes to the posterior edge of the pronotum
and sometimes continuing in a sigmoid course to the bases of

126 WALKER: NORTH AMERICAN SPECIES OF AESHNA

the wing-pads. Femora brownish with three pale rings, a
basal, a median and apreapical. Tibiae and tarsi almost con-
colorous. Abdomen varying much in depth of shade and
extent of the pale mottlings, which are almost identical with
those of interrupta; dorsal surface with a series of pairs of pale
blotches, which run together on the proximal five or six seg-
ments, forming two irregular zigzag longitudinal bands,
but separate on the posterior segments, where they are con-
fined to the basal part of each segment; another series of pale
blotches just laterad of the dorso-lateral punctae, often en-
closing them and confluent with the longitudinal pale bands,
diminishing caudad and disappearing at 8; a series of large
irregular submarginal pale blotches enclosing the lateral
scars which appear as dark brown rings; and a transverse
series of alternately pale and dark basal spots in front of the
lateral scars on several of the middle segments; lateral and
dorso-lateral punctae comparatively large, dorsal punctae
dark brown, discrete; lateral spines whitish, black-tipped.

Measurements—Length of body 41-47.7; mentum of
labium 6.3-8 ; h.w. 8.4-11; h.f. 7.51-8.75; inf. apps. 5-5.9;
gen. 9 2.2-2.4; width of head 8-9.5; width of abdomen
8.1-9.5.

This nymph is the same as the one attributed to this
species by Cabot. Although I have not yet succeeded in
rearing it, there is no possibility of doubt as to its identity.
In one specimen which died during emergence the charac-
teristic form of the lateral thoracic bands of eremita (or can-
adensis) can be made out. Moreover, its size is too large for
any other species and it has always been found in localities
frequented by adults of eremita. It was by far the common-
est nymph at Nipigon in 1910, where the imago was also much
more abundant than any other species of Aeshna.

Material determined — 118 & 37 9. LABRADOR: Hopedale (S. Weiss,
M.C.Z., 1 o' 5 9). NEWFOUNDLAND: (Brit. Mus., 19, Thaxter, M.C.Z., 1 9);
St. John’s (Peary Exped., Ac. N.S. Phil., 1 o); Bay of Islands, July 7, 1901
(D. A. Atkinson, coll. Williamson, 2 o 1 9); Grand Lake, July 21, 1901 (D.A.
Atkinson, coll. Williamson, 1 9). NortHwest TERRITORY : Atik River, 45
miles from mouth (Hudson Bay Slope), Aug. 31, 1907 (W. J. Wilson, coll. Walker,
1 o'); Fort Resolution, Great Slave Lake, July 1862 (Kennicott, M.C.Z., 13 ¢@
4 9). ‘“ Arctic America”’ (Brit. Mus., 3 co’). ALASKA: (U.S. N. M.,2 o) ;
Bethel, Kuskoquin River (Acad. N.S. Phil., 1 co). MAtne: Six Ponds, Piscataquis
Co., Sept. 14, 1898 (F. L. Harvey, U.'S.N.M., 1 @ 1 9). NEW HAMPSHIRE:

WALKER: NORTH AMERICAN SPECIES OF AESHNA 127

Franconia (Mrs. A. T. Slosson, U.S.N.M., 2 o") ; Hermit Lake, Mt. Washing-
ton, Aug. II, 25, 1862, Aug. 2, 1890 (Scudder, M.C.Z., and Calvert, Acad. N.S.
Phil., Bost. Soc. N.H. and coll. Williamson, 17 co 1 9). New York: Loch
Bonnie, near Lake Placid, Adirondack Mts., Sept. 10, 1905 (Calvert, 4 <*);
Bone Pond, Saranac Inn, Adirondacks, July 26, 1900 (Needham 2 o’). QUEBEC,
Chicoutimi, Aug. 23, 1901 (Calvert, 1 o). ONTARIO: Toronto, Sept. 25, 1891,
Sept. 15, 1906 (Brodie, Walker, 2 & 1 9); Go Home Bay, Georgian Bay,
Aug. 17, 1907 (Huntsman, I o& 1 9); Dwight, northern Muskoka, Aug.
23, 1903 (Walker, 2 o 2 9); Temagami Forest Reserve, Kokomo Lake,
Aug. 19, 1907 (P. Hahn, 1 <); id., near Lake Obabika and Cross Lake, Sept. 3,
11, 1908 (Walker, 2 o); Heyden, Algoma, July 30, Aug. 4, 1906 (Williamson,
20 o& 4 9); Searchmont, Algoma, Aug. 7, 1906 (Williamson 1 @ 1 Q);
Nipigon, Aug. 28, 1907 (Walker, 1 @ 1 9); id., Aug. 4-8, 1910 (Walker, 10 o
4 9). Micuican; Marquette (U.S.N.M., 1 9); Oden, July 25, 1906 (J. H.
Williamson, 1 co); Minnehaha Falls, Emmet Co., Aug. 23, 1907 (Williamson,
I o&); Isle Royale, July 26-28, 1905 (Adams, Gleason, Wood, coll. Univ. Mich.,
9 o'). _Manitosa: Husavick, July 8, 1910 (J. B. Wallis, 1 @). SASKATCHEWAN:
1860 (Kennicott, M.C.Z., 6 @ 3 9). ALBERTA: Banff (Sanson, 1 Go).
BritTisH Cotumsia: Peachland, Aug. 3,1909 (Wallis, 1 9, reared); id., 2500 ft.,
Aug. 7, 1909 (Wallis, 1 9); Vancouver (Brit. Mus.,1 9). ‘‘ British AMERICA”
(Scudder, 4 oo’). Wyominc: Shell Creek, Bighorn Mts., July 15, 1896 (R. P.
Currie, U.S.N.M., 1 o). Also3 @ 1 2 without data.

Nymphs—Nipigon, Ont., Aug. 6-8, 1910, 1 oi 1 @ (St. H), 3 #2 Q
(St. F), 1 (St. E), 3 (St. B); exuviae 6 8 9. Temagami, Ont., 1 o exuvia
(P. Hahn). Go Home Bay, Ont., Aug. 1907, 1 9 (emerging); Aug. 1, 1908, 1 9
(St. H). Midland, 1 o& exuvia.

Identity—This species was thought by Hagen (’75) to
be the same as the Siberian Ae. crenata Hag., and Calvert (’94),
accepting this view, attempted to show the identity of Ae.
eremita Scudd. with Ae. clepsydra Say. This study was
based upon forty male specimens, including at least four species,
but, although a careful piece of work, it takes nocognizance of
the important characters found in the accessory genitalia of
the males and the thoracic colour-pattern. A detailed study
of all these forms in much larger series leaves no room for
doubt as to the specific distinctness of eremiia and the com-
parison of this form with the types of crenata likewise shows
that these two forms are also quite distinct.

Ae crenata differs from Ae. eremita chiefly in the follow-
ing characters:

Abdomen of male slenderer, sides of 9 slightly divergent
caudad (subparallel in eremita); posterior margin of 10 more
strongly rounded, superior appendages somewhat narrower,
the outer margin convexly curved at base, thence straight to
the apices, inner margin strongly sinuate, apices slenderer than
in eremtta, with a small projecting tooth; superior carina per-
current and more gradually elevated apically, with larger

128 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

but fewer denticles (3-5). Profile view of superior appendage
straight, scarcely atall elevated apically, the inferior carina less
prominent, its deepest part close to the apex. Spines of the
anterior lamina somewhat shorter and blunter. Appendages
of female much narrower, with a more prominent mid-carina,
apices very acute. Lateral thoracic bands broad and straight,
shaped like those of juncea, with a distinct dark margin and
not at all excavated in front. Blue spots on segment 10 in
the male smaller and more widely separated than in eremita.
Rspl less strongly curved and less widely separated from Rs.

Martin’s (’08) figure of the male appendages of Ae.
hudsonica Selys shows unmistakably that hudsonica and ere-
mita are synonymous, the latter name having priority.

Distribution—Ae. eremita has an extensive distribution
throughout the Hudsonian and Canadian Zones, from At-
lantic to Pacific, and occurs more locally and in smaller
numbers in the Transition Zone. It appears to be the com-
monest and most generally distributed species of Aeshna in
the far north. I found it more abundant than any other
species at Nipigon, north shore of Lake Superior, in the first
week of August, I910, but at the end of the same month in
1908, it was outnumbered by Ae. interrupta interrupta and
by Ae. canadensis.

Geographical variations—The relative length and depth
of the abdominal segments and of the female abdominal ap-
pendages is subject to considerable variation. In specimens
from Newfoundland, Mount Washington and the far
north, seg. 3 is distinctly shorter and deeper than in those
from middle and southern Ontario, while in those from Isle
Royale, Mich., and northern Ontario, it is about intermediate
(vide p. 26).

Habits—I have observed nothing distinctive in the
habits of this species. It flies about shallow, reed-grown
margins of lakes and bays in company with other species of
Aeshna, and may also be met with in bushy pastures and open
woods in the neighbourhood of such bodies of water.

Williamson (’07a) has taken this species together with
Ae. interrupta interrupta and canadensis flying about pools in a
lumber-yard after sunset.

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 129

Brown (’91) records a migratory swarm of Ae. eremita
in Wisconsin.

Most of the nymphs I found in the Nipigon River were
taken in. rather deep water (15-18 inches). The two mature
nymphs which were taken on Aug. 7th, 1910 did not trans-
form that season, but lived until the following January.
They fed readily until about November but took nothing dur-
ing the winter.

Aeshna clepsydra Say.

OAL 2, hes. 5, 63. pl, 14, fies 23 ol 1G)" figs. 4) caa's) pl:
IO; figs. 2, 2a, pl 25) ngs. 5, ;2-)

N.B.—Undoubted references to the species described here as clepsydra are
marked with an asterisk.

Aeshna clepsydra, Say, Journ. Acad. Phil., VIII, p. 12 (1839) ; Calvert, Occ.
Pap. Bost. Soc. N.H., VII, 6, p. 23 (1905); Od. B.C.A., p. 183 (1905). Walker,
Can. Ent., XL, pp. 383, 388, 451 (1908).* Wilson, Pr. U.S.N.M., XXXVI
p. ety ; Smith, Ins. N. J., p. 78 (1909) ; Muttkowski, Cat. Od. N.A., p.
109 (1910).

Aeschna clepsydra, Hagen, Syn. Neur. N.A., p. 122, (1861); Pr. Bost. Soc.
N.H., XV, p. 271 (1873); l.c., p. 35 (1875); Kirby, Syn. Cat. Od., p. 89 (1890).
Beutenmuller, Prel. Cat. Od. N.Y., p. 163 (1890); Harvey, Ent. News, II, p. 73,
75 (1891) ; Calvert, Ent. News, V, p. 243 (1894); Banks, Trans. Am. Ent. Soc,
XIX, p. 353 (1892), Calvert, Trans. Am. Ent. Soc., XX, p. 248 (1893) ; Banks,
Can. Ent., X XVI, p. 77 (1894) ; Calvert, Journ. N.Y. Ent. Soc., III, p. 46 (1895);
Kellicott, Journ. Cinc. Soc. Nat. Hist., XVIII, p. 212 (1895); l.c., XVIII, p. 114
(1896); Calvert, Journ. N.Y. Ent. Soc., V, p. 93 (1897) ; Kellicott, Od. Ohio, p. 84
(1899); Williamson, Drag. Ind., p. 304, pls. 7, figs. 12, 13 (1900); Calvert, Ins. N.J.
Od., p. 71 (1900); Burnham, Pr. Manch. Inst. Arts Sc., I, p. 32 (1900); Needham,
Bull. 47, N.Y. State Mus., pp. 469, 470 (1901); Williamson, Pr. Ind. Ac. Sc., VIII,
p. 124 (1901)*; Needham and Hart, Bull. Ill. Lab. N.H., VI, pp. 40, 42 (1901);
Williamson, Ent. News, XIV, p. 7 (1903) ; Osburn, Ent. News, XVI, p. 186,
(1905); Martin, Cat. Coll. Zool. Selys, XVIII, p. 36 (1908); Muttkowski, Bull.
Wis. N.H. Soc., VI, p. 91 (1908) ; Needham, Rep. Geol. Surv. Mich., App. III,
p. 252 (1908); Martin, Gen. Ins., CXV, p. 11 (1911).

A medium-sized species of slender form, remarkable for
the variegated pattern of the sides of the thorax.

Male—Occiput lemon-yellow, equilateral, two to three-
fifths as long as the line of contact of the eyes; frontal vesicle
more than twice as broad as long, the yellow area reaching
nearly to the lateral ocelli. Eyes bluish grey, pale yellow be-
hind, with a broad dark brownish transverse stripe above;
preocular band somewhat wider at the base of the antenneze
than one of the ocelli, narrowed to a fine line on the sides where
it extends to the lower margin of the nasus. T-spot very
heavy, the cross-bar rather strongly convex in front, 2.5-3
mm., stem .75-1.3 mm. broad, sides straight and parallel or
nearly so. Face pale bluish or brownish green; a heavy deep

130 WALKER: NORTH AMERICAN SPECIES OF AESHNA

brown band on the fronto-nasal suture, which fails to reach
the lateral margins by nearly 1 mm. Rhinarium dark reddish
brown, margined above with pale green; labrum pale green,
very narrowly bordered above and below with dark brown.
Labium pale yellowish or greenish, the middle lobe stained
apically with brown, lateral lobes blue or green. Rear of head
black.

Thorax of average size, ferruginous, very dark in the
humeral region and upper part of lateral surface, distinctly
paler on the dorsum between the dorsal bands, covered ven-
trally and on the lower part of the lateral surface with a
greyish bloom, the ventral surface being quite pruinose.
Dorsal bands light green, 3.5 mm. long, lower ends pointed
and turned outwards and separated by about 4.5 mm., middle
breadth about 1 mm., inner margins strongly convex below,
curving mesad above, outer margins concave, upper ends at
the antealar sinus about 3 mm. broad, separated at the middle
line only by the dorsal carina. Lateral bands broad and ir-
regular in form, pale green below, bluish above; first band
rounded at its lower end, about 1.5 mm. broad in its lower
half, where itis separated by the humeral suture from a large
triangular antehumeral spot. About the middle it is bent
abruptly backwards and then immediately dorsad, the breadth
somewhat reduced, terminating in a slender spur which is bent
sharply cephalad. Just behind this band and frequently
confluent withit are two spots of the same colour, a large sub-
triangular or rhomboid patch enclosing the metastigma and a
smaller one above it. Second lateral band large, triangular,
sometimes reaching the suture in front, lower end acute, upper
end 2.5-3 mm. broad, front margin often a little excavated
near the middle, hind margin straight except where it curves
downward at the upper end. Antealar sinus green.

Legs reddish brown; femora somewhat darker; no pale
streak on outer surface of fore femora.

Abdomen a little more than four and one-half times as
long as the thorax, very slender. Seg. 3 considerably con-
stricted. Seg. 1 without a ventral tubercle; auricles with 3
(rarely 4) teeth. Spines of anterior lamina short and blunt,
directed ventrad and caudad, not projecting below the tergal

WALKER: NORTH AMERICAN SPECIES OF AESHNA 131

margins. Hamular processes directed mesad and cephalad,
short, with bluntly pointed apices and without a distinct apical
tubercle, elevated considerably above the adjacent margins
of the hamular folds, surface divided by a ridge into anterior
and postero-ventral areas, the latter rhomboid and somewhat
sulcate. Hamular folds continuous with the processes but
distinctly marked off from them, ear-shaped, the outer
margins straight, parallel or slightly divergent, inner margins
closely approximate. Tergal margins on each side of penis
straight. Seg. 3 one-half, seg. 4 very slightly, longer than
segs. I+2. Lateral carinae of 8 sinuate at base in ventral
view, those of 9 obsolete in the anterior half. Dorsum of 10
with an abruptly elevated median tooth and two minute
lateral teeth on each side of it.

Superior appendages barely longer than 9+10, slender in
the proximal fifth or sixth, the basal breadth of which is about
one-third of the intervening space, thence expanding mesad to
about one-third or two-fifths of the length, where the breadth
in a vertical dorsal view is one-seventh to one-sixth of the
length; sides beyond parallel, converging again in the distal
fourth to the rotundo-angulate apices, which are produced
into a rather prominent sharp-pointed backwardly directed
spine. In dorsal view the outer margins are nearly straight,
the inner margins sinuate in the proximal half, nearly straight
beyond. In lateral view the outer margin is also nearly
straight, feebly curved downwards at apices, inner (lower)
edge beyond the basal fifth forming a well-marked arcuate
inferior carina. Superior carina obsolete in the slender basal
part, elsewhere moderately sharp, straight, scarcely elevated
apically, with 3-5 minute denticles on the apical third. In
lateral view it barely rises above the lateral margin before the
apex. Inferior appendage elongate-triangular, three-fifths as
long as the superior pair, the breadth at base about one-half
the length, sides straight, apex acute; in profile view moder-
ately curved, rather narrow, evenly tapering to the apex
which is surmounted by a pair of minute spines.

Colour-pattern of abdomen—Ground colour dark brown,
distinctly paler in front of the transverse carinae and darken-

132 WALKER: NORTH AMERICAN SPECIES OF AESHNA

ing posteriorly on each segment; blue spots paler than in
most of our species of the genus.

Seg. 1 brown; dorsal spot blue, lateral spot 3-3.5 mm.
long, 1.5 mm. broad, giving off a yellowish offshoot to the
front margin of the segment.

Seg. 2 brown; AML pale blue, somewhat V-shaped, the
anterior limb at the front margin 2.5-3 mm. long, posterior
limb somewhat narrower and not confluent with MD. PL
blue, almost or quite crossing the posterior division of the seg-
ment, rather narrowly confluent with PD, the latter about
two-thirds as wide as the posterior part of the segment,
moderately broadly connected with its fellow of the opposite
side and giving off cephalad to the front margin a diffuse
median band.

Segs. 3-10. Pale spots, except MD, blue; AL on 3-8, well
developed on the anterior segments, connected by a narrow
basal ring, except on 7 and 8, and curving mesad along the
transverse carinae. Upper margin concave, lower margin
notindented. ML on 3-8, generally slightly indented in front,
rounded behind, confluent with MD on 3-6 or -7 and often with
PL on 3. MD on 3-7, yellowish, much elongated trans-
versely, narrowed laterad except on 3 and 4. PD about 1
mm. long on 3 increasing to 2 mm. on 8, connate behind on 10,
with a V-shaped anterior emargination, triangular on 8 and 9,
the inner margins separated by a V-shaped space; elsewhere
rounded in front, the inner margins straight and separated
only by the dorsal carina. PL on 3-6 or -7, an anteriorly di-
rected offshoot from PD, broader and nearly straight ante-
tiorly, curved and slender posteriorly. There is also a pair
of pale basal ventral spots on segs. 4-7.

Wings hyaline, costal nerves brownish ochraceous, pter-
ostigma dark smoky brown above, paler yellowish brown
beneath; membranule of hind wings wholly dark smoky
brown not reaching cross-vein of anal triangle, which is 2-
celled. Twocells between A3 and A2attheirorigin. Rs fork-
ing at the level of 2-3 (fore wing) 3-4 (hind wing) postnodal
cells before the stigma; 3 (rarely 4) rows of cells between the
forks at the level of the distal end of the stigma and 3-6 cells
between the forks at the margin; 4 (rarely 3 or 5) rows of cells

WALKER: NORTH AMERICAN SPECIES OF AESHNA 133

between Rs and Rspl when most widely separated; Mia
arising before the middle of the stigma.

bates postnodals 29, CuCr 2%, Spt 24.
II-14 10-16 4-6 2-3

Female—Dorsum of 9 about five-sevenths of the length of
8. Genital valves in profile slightly arcuate, the apices not
elevated; lateral carinae prominent, percurrent, in profile some-
what irregularly arcuate, in ventral view somewhat sinuate,
enclosing a space, which is broadest at about two-fifths the
length, the breadth here being about one-third of the length;
thence narrowing to the somewhat widely separated apices,
each of which bears a minute terminal pencil of hairs. Ven-
tral surface of valves feebly sulcate, inner margins not at all
elevated, meeting when the ovipositor is withdrawn, only a
very short distance about the middle. Styli scarcely more
than .5 mm. long. Appendages longer than 9+ 10, but shorter
than 8+9, more slender in the basal than in the apical half,
greatest width at or alittle beyond the middle, equal to nearly
one-fifth of the length; inner margin more strongly curved
than the outer, which is nearly straight; apices rounded,
sometimes with a very minute terminal spine.

Colour-pattern—The contrast in the depth of the ground
colour of the abdomen before and behind the transverse
carinae is somewhat more marked than in the male. The
pale markings are very similar, the differences being of the
usual kind. PD is shorter, especially on the anterior segments,
and is less distinctly or not at all triangular in form on 8 and
g. Lateral spots all somewhat larger, PL confluent with ML
on 2 and generally also 3. Legs somewhat paler reddish
brown.

In the two females seen in life one (the specimen figured)
resembled the male except that the pale markings were duller
and PD green; in the other the colours were not quite mature.
It is impossible to determine the range of colour variation
from the dried material examined.

Measurements—Thor. @ 10-10.5, 9 9.5-10; abd. o 45.5-
49, 9 46-48; seg 3 of 8-9, 2 7-8.5; apps. co 5.2-5.4, 2 5.33-
5.6; gen. v. 2-2.33; h.w. o 40-47, 9 43-44; width hw. o&
12-13.33, 9 13-13.5; width seg. 2 o& 5-5.7, 9 5.5-5.7-

Antenodals

134 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Neotype— Harris Collection (Harris Cat. No. 45), Bost.
Soc. Nat. Hist.—Massachusetts.

Nymph (pl. 8, fig. 4)—-Very similar to that of Ae. can-
adensis from which it differs as follows: Eyes and postero-
lateral angles of head very slightly more prominent, the
latter similarly or somewhat less broadly rounded. Mentum
of labium about three-fifths as broad at base as apex, the
apical breadth five-sevenths of the length, somewhat narrower
in the proximal three-fifths than in canadensis, the sides not at
all arcuate except in the distal two-fifths; middle lobe moder-
ately produced, subarcuate; lateral lobes rather slender
terminating in a somewhat strongly curved hook. Supracoxal
processes somewhat shorter than in canadensis, bluntly coni-
cal, about equal in length, the posterior one a little the stouter,
interval less than aright angle. Lateral spines of abdomen
somewhat longer thanin canadensis, those on seg. 6 extending
half way tothe posterior margin, on 7 to the margin, on 8 one-
fourth to one-third beyond the margin, on 9g as far as the
proximal two-thirds of 10. Appendages similar to those of
canadensis, the superior pair a trifle shorter, being scarcely
more than one-half as long as the inferior pair in the male, and
about half as long in the female.

The colour-pattern is very much like that of canadensis,
but the pale lateral bands in the head and thorax are broader
throughout their length; the longitudinal bands of the abdo-
men are continued to the apical margin of ro (at least in the
exuvia) and their margins are almost straight throughout
their length. The dark median band is generally paler but
more even in depth of shade, not being deepened about the
dorsal punctae, which are not marked with darker brown.
The dorso-lateral and lateral punctae are also inconspicuously
marked.

Measurements—Length of body 36.5-40; mentum of
labium 6-6.33; h.w. 8.2-9; h.f. 6-6.2 ; inf. apps. 4.1-4.2; width
of head 7.5-8 ; width of abdomen 7-7.5.

Material determined—38 @ 11 9. MAINE: Manchester, Aug. 10, 1908,
Sept. 7-13, 1907, Sept. 18, 1908, Sept. 2-8, 1909, Sept. 12, I910, Oct. 1, I910
(Miss Wadsworth, Io oh I Q, incl. I o, U. S.N.M.). MASSACHUSETTS: (Bost.
Soc. N.H., Harris coll. 1 o); (Caudell, U.S.N.M., « o@); Brookline, (Shurtleff,
Bost. Soc. 'N.H. ., 1 o'); Provincetown, Aug. 6, 1899 (J. E. Benedict, jr., U.S.N.M.,
1 9); Wilbrahim, Aug. 3, 1902 (Needham, 1 ¢?); Boston. 1858 (Uhler, M.C.Z.

_ Wacker: NorTH AMERICAN SPECIES OF AESHNA 135

1 6’); Salem (Joseph True, M.C.Z., 1 co") ; Natick, 1863 (Sanborn, M.C.Z., 1 o).
Ontario: De Grassi Point, Lake Simcoe, Aug. 24, 1908, Aug. 19, 1910 (Walker,
1 &@ 1 9) ; Go Home Bay, Georgian Bay, Aug. I, 5, 1908, Aug. 18-26, 1907,
fey 19, 1910 (Huntsman, Fraser, Cooper, Walker, 14 co’ 5 9); Point Pelee,
ake Erie, Aug. 7,1901 (Walker, 1 co’). MicuiGAn: Detroit, (M.C.Z., 1 3, 19).
INDIANA : Shriner Lake, Aug. 24, 25, 1901 (Williamson, 2 o); Crooked Lake,
eeuben Co., Sept. 1, 1905 (Waugh, coll. Williamson, 1 ); also 3. & without
ata.
Nymphs—Go Home Bay, Ont., July, 1910, 2 co, about to emerge; exuviae
2 o', emerged July 25, 28, 1910 (Cooper); July 31, 1907, 1 9; Aug. 13, 1908,
I o';also2 oI Y without data. Midland,1 o.

Identity—Say’s type which was formerly in the Museum
of the Boston Society of Natural History no longer exists and
it is impossible to say with certainty which species of the
clepsydra group it belonged to. It seems clear, however,
that the species described here is the one which has the best
right to retain Say’s name, for the type locality is Massa-
chusetts, and all the specimens of the clepsydra group that I
have seen from that state,with a single exception, belong either
to this species or to Ae. verticalis. One of the former is the
single specimen labelled clepsydra in the Harris collection.
This specimen was probably determined by Say and should
be regarded as the neotype.

Distribution — Ae. clepbsydra is an eastern species in-
habiting the Upper Austral and Transition Zones from New
England to Indiana and southern Michigan. It appears to
be most common in the eastern part of its range.

Habits—At Go Home Bay this species flies over shallow
reed-grown bays and adjacent open marshes, its habitat
being apparently similar to that of Ae. canadensis. It begins
to appear, however, about a month later, the earliest date
of capture recorded being July 19, 1910. Two specimens
emerged at Go Home on July 25 and 31, 1910. The two
Lake Simcoe specimens were taken from tree trunks on the
edge of a wood, about half a mile from the-nearest possible
breeding-place.

Aeshna canadensis Walker.

(Pl ifies 3) 7 pl o:5; figs: 128\;) phvag, fig. 4; pl. 16,

Aesi5))5a;:pb. 19, figs..3, 3a; pl. 25, figs. 3,/4-
Aeschna clepsydra, Selys, Ent. M. Mag., II, p. 242 (1875); Calvert, Trans, Am.
Ent. Soc., XX, p. 248 (1893) [in part]; Ent. News,V., pp. 9-13, figs. 1-4 (1894);
Can. Ent., XXVI, p. 318 (1894) ; Harvey, Ent. News, XIII, p. 8 (1902); Ruthven,
Ec. Surv. N. Mich., p. 101 (1905) [in part]; Fletcher and Gibson, 38th Ann. Rep.

136 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Ent. Soc. Ont., p. 132 (1908); Muttkowski, Bull. Wis. Nat. Hist. Soc., (2) VI, p.
gt (1908); Martin, Cat. Coll. Zool. Selys., XVIII, p. 36, fig. 36 (1908).

Aeschna juncea, Williamson, Ent. News, XIII, p. 146 (1902) [in part].

Aeschna verticalis, Walker, 36th Ann. Rep. Ent. Soc. Ont., p. 69 (1906) ;
Can. Ent., XX XVIII, p. 50 (1906) [in part].

Aeschna ‘‘y,”’ Williamson, Ohio Nat.,VII, pp. 145, 146, (1907); Walker, Ott.
Nat., XXII, p. 54 (1908).

Aeshna canadensis, Walker, Can. Ent., XL, pp. 384, 389, 451 (1908) ; Mutt-
kowski, Bull. Wis. Nat. Hist. Soc., VIII, p. 57 (1910); Cat. Od. N.A., p. 109
(1910); Martin, Gen. Ins., CXV, p. 11 (1911).

Male—A medium-sized species of slenderform. Occiput
pale yellow, margined laterally with black, equilateral or con-
siderably broader than long, two- to three-fifths as long as the
line of contact of the eyes. Frontal vesicle more than twice as
broad as long, the yellow area not reaching the lateral ocelli.
Eyes grey-green, without blue reflexions, pale greenish or
yellowish behind with a dark brownish transverse stripe above.
Preocular band somewhat wider at the base of the antennze
than one of the lateral ocelli, becoming very narrow on the
sides of the frons but broadening again upon the nasus. T-
spot heavy, rather strongly convex in front, 2.5-3 mm., stem
-75 to 1.3 mm. broad, sides generally straight and parallel or
slightly divergent behind. Face pale bluish or sometimes
yellowish green; the frons yellowish to brownish olivaceous
just beneath the T-spot, pale yellowish or whitish next the
preocular black band; a fine brownish line on the fronto-nasal
suture; lateral lobes of nasus rotundo-angulate, scarcely
flaring. Rhinarium dark brown, pale reddish above and
usually below and along the median line. Labium dull yellow
or drab to dull bluish, stained distally with reddish brown;
lateral lobes more or less distinctly tinged with greenish blue.
Rear of head black.

Thorax of average size, chocolate-brown with a greyish
bloom beneath. Dorsal bands rather dull pea-green, some-
times bluish above, nearly straight, 3.5-3.75 mm. long, .75-1
mm. broad at the middle, lower ends pointed and separated by
about 4.5 mm.; upper ends at the antealar sinus separated by
about I mm., expanded, 1.25-1.5 mm. broad, often separated
from the rest of the band as a distinct spot. Agreenor bluish
green spot in theantealar sinus; generally an obscure streak just
in front of the humeral suture. Lateral bands broad but ir-
regular in form; the first band sometimes wholly blue, but

WALKER: NortTH AMERICAN SPECIES OF AESHNA _ 137

generally pea-green in the lower half, often wholly green,
rounded at the lower end, 1.33-1.5 mm. broad in the lower
half, greatly constricted just above the middle by the deep al-
most rectangular excavation of the anterior margin, the
breadth here being .33-.75 mm., thence expanding to about
I mm.near the upper end, from which a rather narrow offshoot
passes caudad. Anterior margin above the middle straight,
posterior margin slightly sinuate. Second band blue, often
more or less green below, occasionally wholly green, about 1
mm. broad at the middle, generally narrowed at the lower
end, expanded at the upper end, where the breadth is 1.5-2
mm.; anterior margin more or less sinuate, posterior margin
straight except at the upper end. There is also a pair of
greenish spots at the metastigma and a small yellow spot
above them. Interalar spots pale blue.

Legs dark reddish brown, the tarsi and under side of the
femora and tibiae darker. Front femora with a pale streak
on the proximal half of the posterior surface.

Abdomen 4.5 to nearly 5 times as long as the thorax,
slender, strongly constricted before the middle of seg. 3,
without a ventral tubercle. Auricles with 3 or 4 teeth.
Spines of the anterior lamina short, not reaching back to the
bases of the hamular processes, directed ventrad and caudad
and not projecting below the tergal margins when in the usual
position. Hamular processes somewhat elongate, proximally
parallel and directed cephalad, the apices approximated,
bluntly pointed. Hamular folds continuous with the pro-
cesses, the outer margins divergent, meeting the posterior
margins at a rotundo-acute angle, inner margins closely ap-
proximated. Tergal margins on each side of the penis
straight. Seg. 3 about one-third longer than I+2 or 4;
lateral carinae of 8 in ventral view irregularly bent before the
middle; feebly developed or obsolete on 9. Dorsum of 10
with an abruptly elevated sub-basal conical median tooth be-
tween two pairs of much smaller teeth.

Superior appendages about as long as 9+10 or barely
longer, their basal breadth one-third to two-fifths of that of the
intervening space, slender and slightly widening in the basal
fourth, then expanding mesad to about two-fifths of the

10

138 WALKER: NORTH AMERICAN SPECIES OF AESHNA

length where the breadth viewed dorsad is one-seventh to
one-sixth of the length; and narrowing again slightly to the
somewhat abruptly decurved apices, which terminate in a
short sharp spine; outer margin, in dorsal view, gently convex,
inner margin rather strongly sinuate, appearing more or less
concave beyond the middle. In lateral view the outer mar-
gin is curved gently upwards except towards the apex; inner
margin bent downwards beyond the basal fourth, forming an
arcuate inferior carina. Superior carina feebly marked in the
slender proximal part, somewhat elevated in the apical third,
where it bears five or six small denticles. Inferior appendage
elongate-triangular, about three-fifths as long as the superior
pair, the basal breadth half or barely more than half the
length, sides straight or feebly approximated about the middle,
a pair of minute recurved spines just before the bluntly
pointed apices; in profile view moderately curved, rather
narrow, tapering evenly to the apex.

Colour-pattern of abdomen—Spots all blue except MD,
which is dull yellowish.

Seg. 1 fuscous; dorsal spot azure blue; lateral spot a
pale blue streak along the posterior margin, 3 mm. long and
generally 3 mm. or less in breadth, though sometimes widening
below to I mm.

Seg. 2 fuscous ; AML azure blue, nearly 2 mm. broad in
the middle, produced a short distance ventrad along the
front margin and continued dorsad along the transverse carina
as a broad band which is narrowly confluent with MD. PL
and PD confluent, forming an azure blue band, which is
nearly as broad below (PL), or from one-half to two-thirds as
broad above (PD),as the posterior part of the segment; front
margin angularly excavated on each side. A blue streak of
variable width runs forward from this band along the median
line to the front margin of the segment.

Segs. 3-10 brownish black. AL on 3-8, azure blue,
giving off on most of the segments a basal transverse offshoot,
which fails to form a complete ring with its fellow of the oppo-
site side; upper margins straight. ML on 3-8, azure blue, en-
larging to 6, rhomboid on 3-4 or -5, rounded behind on the
remaining segments, anterior margin more or less angularly

WALKER: NORTH AMERICAN SPECIES OF AESHNA 139

indented. MD on 3-7, often represented also on 8 by a pair
of dots, of moderate size, on 3 more or less quadrate and con-
fluent with or narrowly separated from ML, on the remaining
segments triangular and separate from ML. PD azure blue, on
IO sometimes very pale blue or even yellowish ; about 1.3-
1.5 mm. long on 3, increasing to 2 mm. on 8; narrowly connate
behind on Io in about 22 per cent. of cases, rounded in front
and mesially and rather narrowly separated on 3-7, subtriangu-
lar on 8, and triangular on9g,theinner margins straight and more
or less divergent and more widely separated on these two seg-
ments than on the others, thoughsometimes narrowly connate
behind on 9. PL on 3-6 or -7, decreasing, separated from 5 or
6 caudad, elsewhere forming a curved offshoot from PD.
Ventral surface of abdomen with a pair of pale blue basal spots
on 4-6.

Wings hyaline, costal veins brownish ochraceous, pter-
ostigma dark smoky brown above, pale yellowish brown be-
neath; membranule of hind wings wholly dark smoky brown,
not reaching cross-vein of anal triangle, which is 2-celled.
Two cells between A2and Agat their origin. Rs forking at the
level of 1-3 (fore wing) 2-4 (hind wing) postnodal cells be-
fore the stigma; 3 or 4 rows of cells between the forks at the
level of the distal end of the stigma and 3-7 cells between the
forks at themargin. Threeor4g (rarely 5) rows between Rsand
Rspl where most widely separated. Mla arising before the
level of the middle of the stigma, usually just beyond the
proximal end.

Antenodals 122) postnodals 23, CuCr 50 Speteee.
I 4-6 2-3

Female—Abdomen very slightly shorter than the hind
wing, its depth, especially at seg. 3, varying greatly according
to locality (vide postea).

Apices of genital valves not elevated, lateral carinae per-
current, feeble in the basal third or fourth, prominent be-
yond; in profile subangulate or gently arcuate, in ventral view
somewhat sinuate, enclosing a space which is broadest in the
proximal third or two-fifths, the breadth here being about one-
fourth the length, thence narrowing very slightly to the some-

-15 1i-

140 WALKER: NORTH AMERICAN SPECIES OF AESHNA

what squarely cut, well-separated apices, which bear each a
minute pencil of whitish hairs. Ventral surface of valves
slightly sulcate, sloping laterad in the proximal three-fifths,
more or less mesad in the distal two-fifths. Inner margins of
valves when ovipositor is ensheathed attingent only for a
very short distance. Styli about .66 mm. long, or half the
length of the dorsum of 10. Basal plate of ordinary size, the
free margin straight. Appendages not longer than 8+9, gen-
eraly distinctly shorter, the length varying with locality
(vide postea), more slender in the basal than in the apical half;
greatest breadth a little beyond the middle, equal to one-
sixth or one-fifth of the length, curve of inner margin stronger
than that of outer margin, apices rounded or rotundo-angulate
without a terminal tooth.

Colour-pattern—Dorsal thoracic stripes narrower than in
the male, often obscure in colour, each divided into an anterior
more elongatespot, about .5 mm broad, anda posterior shorter
and broader spot, just in front of the antealar sinus. Lateral
thoracic bands similar to those of the male. Lateral abdom-
inal spots somewhat larger than those of the male, especially
PL, which is present on 3-7 and is only rarely separated from
PD. PD slightly smaller thanin the male, separate on 2,
quadrate on 9g, the basal rings for the most part complete,
though narrow. In colour there is great variation among the
females of this species, typical homoeochromatic and hetero-
chromatic individuals as well as intermediate forms being met
with.

The pure homceochromatic form is rare. At De Grassi
Point, Lake Simcoe, I captured a female of this form, differ-
ing in colour from the male only in the heavy brownish suf-
fusion of the wings. The dorsal thoracic bands and the lower
part of the first lateral thoracic band are green, the upper
part of the latter, the second lateral band and the abdominal
spots (except MD) blue. Another female taken at Toronto
had the lateral thoracic bands wholly blue and the wings en-
tirely clear, and one from the Porcupine Mts., Mich., is also
of this type. In the heterochromatic form the occiput is
bright yellow, the eyes yellowish green, and bright yellow
along the hind margin, face olive-green, labium dull yellowish,

WALKER: NorTH AMERICAN SPECIES OF AESHNA I4I

the lateral lobes plumbeous. Thoracic bands and the inter-
alar spots yellowish to grass-green. Abdominal spots green,
the lateral spots, particularly AL and ML, usually of a more
bluish green than PD and PL.

The legs are somewhat paler than those of the male and
the wings frequently flavescent, though often quite hyaline.
The flavescence apparently deepens with age, but it is not
entirely dependent upon it, as tenerals may have fairly deep-
ly coloured wings, while those of old battered individuals may
be quite hyaline. Flavescent wings are apparently most
common in northern individuals, in which they may be of
quite a deep shade of brown.

Measurements—Thor. of 10-11, 9 9.5-10.5; abd. o 46-
51.5, 9 45-49; seg. 3 o 8-9.4, 2 6.75-8.25 ; depth seg. 2 of
5.25-5-75, @ 5.5-6; depth seg. 3, 9 2.5-4.5; apps. co! 5.25-6,
9 4-5.6 ; hw. co 43-46.5, 9 42.5-47.33 ; width h.w. o 12.5-
14, 9 12.75-14.4; pter. oO! 3.1-3.6, 9 3.7-4.3.

Types—co and 9, U.S. National Museum—De Grassi
Point, Ont.

Nymph (pl. 6, fig. 4; pl. 8, fig. 5; pl. ro, fig. 4.)—
Eyes moderately prominent, shaped as in juncea ; lateral
margins of head moderately oblique, nearly straight, about
one-half longer than the interocular space; postero-lateral
angles of head rotundo-obtusangulate, more rounded than
in eremita but decidedly less so than in juncea and interrupta;
posterior margin straight or very nearly so; mentum of labium
about seven-elevenths as broad at base as at apex, the apical
breadth a little more than five-sevenths of the length; proxi-
mal three-fifths slightly widening distad, the sides barely arcu-
ate, distal two-fifths expanded, the sides more strongly arcuate,
though less so as a rule than in juncea; middle lobe broadly
rotundo-angulate or subarcuate, but little produced; lateral
lobes rather slender, narrowing somewhat distad, the apices
not squarely truncate, outer apical angle well-rounded, inner
angle with a distinct tooth. Supracoxal processes well
developed, about equal in length, acute, the posterior slightly
stouter, interval slightly less than a _ right angle.
Abdomen broadest at seg. 6 or 7; lateral spines present
on 6-9, not spreading, extending towards the posterior mar-

142 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

gin of the corresponding segments as follows: on 6 two-fifths
to half the distance to the margin, on 7 to the margin or nearly
so, on 8 to the margin or slightly beyond, on 9 about as far as
the middle of 10. Lateral appendages slender, about three-
fifths as long as the inferior appendages ; basal part of supe-
rior appendage of male about one-fifth shorter than the
lateral pair, somewhat less broad at base than long, sides
straight or nearly so, apex acute. Genitalia of female nearly
reaching the posterior margin, genital valves about two-fifths
as broad as long, more steeply declivent than in juncea and
eremtita.

Head behind eyes with a rather narrow pale mar-
gin, which is continued over the pronotum and meso-
thorax, becoming more diffuse on the latter; thorax otherwise
nearly uniform brownish, legs concolorous. Abdomen longi-
tudinally striped, a dark median band between two pale ones,
these becoming somewhat less distinct caudad and more
or less obsolete on 10. These bands are for the most
part somewhat conspicuous and_ well-defined, with
the margins nearly straight and subparallel. The median
band is solid and dark on the more anterior segments and
sometimes on all but the last one or two segments, but is
generally broken up posteriorly into a series of subcontinuous
dark blotches on the basal part of each segment, obscuring
the dorsal punctae. A more or less distinct pale median line

runs throughout the band. Pale longitudinal bands as

broad or slightly broader anteriorly than the median band,
narrower posteriorly, outer edges nearly straight posteriorly,
somewhat undulated anteriorly. Sides of abdomen brown,
much paler than the median band, the posterior external to
the lateral scars, paler than the rest; lateral scars forming a
pale undulating line not distinctly outlined laterally, lateral
punctae distinct, dorso-lateral punctae almost obsolete.

Measurements—Length of abdomen 35-39.5 ; mentum of
labium 5.5-6 ; h.w. 8.5-9.5; h.f. 6-6.6; inf. apps. 3.8-4.3; gen.
9 1.8; width of head 7.8-8; width of abdomen 7-7.5.

Material Determined —190 o& 67 9. Nova Scotia: Pictou, July 26,
1889 (W. Sheraton, coll. Calvert, 1 ay, New Brunswick: St. Andrews, 1910
(Huntsman, 19). MAINE: Bradley, Aug. 7, 1891 (F. L. Harvey, U.S.N.M., 3
co’); Manchester, July 28-Sept. 22, 1889, 1898, 1907-10 (Miss Wadsworth, 18

WALKER: NORTH AMERICAN SPECIES OF AESHNA 143

mGue sinc. 1c" in U:S.N.M.); ‘Norway. 4S, J.) Smith. MiG-Z.,) 2. ot):
Orono, July 28, 29, 1898 (F. L. Harvey, U.S.N.M.,8 oI 9). New Hamp-
SHIRE: White Mts. (Shurtleff, M.C.Z., 1 o&*) ; White Mountain House, Aug. 22,
1889 (Calvert, I o’); Franconia (Calvert, M. aw a Io’); Fabyans, Aug. 22, 1889
(Bost. Soc. N.H., 1 oc). VeERMonT: (Ellis Frost, coll. Williamson, 1 oc’); Newport
(Mrs. Slosson, U. S.N.M., 1 @). MASSACHUSETTS: (Needham, I 9). NEw
York: Ithaca, Aug. 28, 1889 (Banks, U.S.N.M., 1 o); Lake St. Regis, Franklin
Co., Sept. 20, 22, 1890 (J. P. Moore, Calvert, 2 o'). QUEBEC: Quebec (Prov-
ancher, M.C. Z., Io’); Grand Entry, Magdalen Islands, June 30, 1901 (D.A. At-
kinson, coll. Williamson, 2 9—I reared). ONTARIO : Ottawa, Meach Lake,
July 21, 1907 (A. Gibson, 1 2); Toronto, Sept. 12, 1907, SEBe 27,1909 (Walker,
Hahn, 7 o 2 9); Scarboro, Sept. 7, 1907 (Huntsman, 1 3&1 9); De Grassi
Point, Lake Simcoe, June 25-Sept. 6, 1903-10 (Walker, 20 e102);
Midland, Aug. 8, 1908 (Walker, 4 <); Go Home Bay and vicinity, Georgian
Bay, June 28- Sept. 8, 1907-10 (Huntsman, Cooper, Walker, 25 o'\ 6 @);

Bala, Sept. 15, 1907 (Huntsman, 1 o’); Muskoka (Hahn, 1 @ 19); Dwight,
Northern Muskoka, Aug. 23, 1903 (Walker, 11 9); Temagami Forest Reserve,
Aug. 14, 1906, Sept. 1-11, 1908 (Walker, Hahn, 9 o' 5 @); Tobermory, Bruce
Co., Aug., 1901 (Walker, Io’); Searchmont, Aug. I-9, 1906 (Williamson, 2 o);
Heyden, July 30, Aug. 3-4, 1906 (Williamson, 6 o 2 2); Nipigon, Aug. 28-30,
1907, Aug. 6-8, 1910 (Walker, 8 & 6 9); Ft. William, Aug. 2, 1910 (Walker,
I o'). MicuiGan: Detroit (Hubbard and Schwarz, M.C.Z., 1 o’); Oden, Emmet
Co., Aug. 11-24, 1906-07 (Williamson, 26 @ 3 9); Porcupine Mts., Ontona-
gon Co. (A. G. Ruthven, Univ. Mich., 3 co 4 9); Isle Royale (Gleason, Univ.
Mich., 1 9); St. Croix River (Muttkowski, 1 o); Douglas Lake, Topinabee,
Aug. 1g, 20, 1910 (Miss A. O’Brien, 2 o& 3 9). Itirnots: ‘Northern Illinois,”
1860 (Uhler, M.C.Z., 1 o); Rhode Island, 1860 (Walsh, M.C.Z., 1 o&); Lake
Forest, June, 1904 (Needham, 1 co’). INDIANA: Graveyard Lake, Aug. 27,
1911 (Ray, coll. Williamson, 1c’). Missouri, St. Louis, July 10, 1903 (C. L. Mer-
rick, coll. Muttkowski, 1 9). WuIScONsIN: (J. T. Brown, U.S.N.M., 1 o);
Nagowicka Lake, Waukesha Co., Aug. 17, 1900 (C. E. Brown, U.S.N.M., 1 9);
Fox Lake, Dodge Co., Aug. 9, 1908 (Muttkowski, 2 o”) ; Portage Co., Aug. 20,
1905 (Muttkowski, 1 ?); Oostburg, Sheboygan Co., July 8, 1908 (Muttkowski,
1 9); Milwaukee Co., June 14, 1908, July 1-Sept. 26, 1903 (Muttkowski, 203
9) ; Low Rock, Aug. 15, 1906 (J. D. Hood, 1 9). Murnnesota: Duluth, Sept.
2 and 3 (O. Sacken, M.C.Z., 2 o&\ 19). ‘‘DAKoTa” : (Dr. Mark; coll. Calvert,
I 9), MANniToBA: Westbourne, Aug. 24, 1908 (J. B. Wallis, 1 o’). BritisH
COLUMBIA: Vancouver (Brit. Mus., I co’); Victoria, July (Crotch, M.C.Z., 10°).
WASHINGTON: Loon Lake, Colville Valley, July 22, 1882 (S. Henshaw, M.C.Z.,
4 0).

Nymphs—Magdalen Islands, P.Q.,1 9 exuvia with imago, emerged June 30,
1901 (D. A. Atkinson). Go Home Bay, Ont., and vicinity, June 20, 25, 1907,2 ¢
about to emerge; Aug.9, 1908, Aug. 10, 14, 1905, 4 o' 6 @ (St. F); Aug. Io,
1905,4 o' (St. E), 16’ (St. D), 2 spec. (St. B) ; July 15, Aug. 20, 1909, exuviae
1¢'1 9 from reared imagoes. Giant’s Tomb Island, Georgian Bay, Ont. July 29,
1908, 1¢' 1 @ (St. E),3 7% 492 (St. D). Midland, Ont., Aug. 8, 1908, 1 (St.
F), 1@ (St. E). De Grassi Pt., Ont., Aug. 30, 1910 (St. C and D); Sept. 6,
1gcg (St. C); July 27, 1910 (St. B and A) ; June 22, 1909, and Aug. 5, 1910 20°
exuviae from reared imagoes; July 15, 1910,exuviae 2¢' 29. Toronto, exuviae
2d’.

Identity—This is the species to which most of the refer-
ences to Ae. clepsydraSay belong, and is the one to which Cal-
vert (’94) referred as ‘‘presumed clepsydra.’’ Its most
distinctive feature is the form of the anterior hamuli, which
are more elongate cephalad than in other species of the group.

144 WALKER: NorRTH AMERICAN SPECIES OF AESHNA

It is most likely to be confused with Ae. eremita and Ae.
verticalis, both of which it resembles in coloration. The
males may be readily separated by the form of the appendages
and hamuli-but the females are more difficult to distinguish.
Eremita may be almost invariably known by the larger size
and the black line across the face, but badly preserved fe-
males of verticalis and canadensis may give trouble to those
who try to separate them, though, in the great majority of
cases, the lateral thoracic bands can be seen well enough to
render a diagnosis possible.

Distribution—This species inhabits the Transition and
Canadian Zones from Atlantic to Pacific, but is apparently
much less abundant on the western than on the eastern half
of the continent. In the east it occurs throughout the Tran-
sition and in the lower part of the Canadian Zones. It is not
known how far north it ranges but no specimens have been
seen from beyond Lat. 50. N. Occasional examples are
taken from within the limits of the upper Austral Zone.

Next to Ae. umbrosa this is the commonest North Amer-
ican species of Aeshna met with in collections.

Habits—This species breeds among reeds and sedge in
shallow sluggish creeks, lakes and bays. With the exception
of Ae. californica and the multicolor group it is the earliest
species to appear in the adult state. Individuals may some-
times be seen as early as the middle of June but the usual
time for their appearance in the Transition Zone in Ontario is
during the last week of June and the first week of July. By
the middle of July they are numerous and may sometimes oc-
cur in very large numbers about the sunny borders of woods,
especially coniferous woods. Here they may be seen sunning
themselves on the trunks of trees or hanging from the twigs.
On hot sultry days they fly restlessly to and fro in small
openings among the trees, on the lookout for their prey. In
August they are less frequently seen in the woods. They
appear to return to their breeding-grounds, where they may
be observed in large numbers, gliding over the reeds or
skirting the water’s edge. Such individuals are all males and
may often be observed to drop down among the reeds and
then emerge with a female, im copula. Apparently copu-

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 145

lation does not take place far from the water as commonly
occurs in constricta and other species.

Ae. canadensis is still common at the beginning of Sep-.
tember and may sometimes be found until nearly the end of
the month.

Aeshna verticalis Hagen.
(Pl. 13, fig. 5; pl. 16, figs. 6, 6a; pl. 19, figs. 4, 4a; pl.
25, figs. 5, 6.)

Aeschna verticalis, Hagen, Syn. Neur. N.A., p. 122 (1861) ; Pr. Bost. Soc. N.
H., XVIII, p. 34 (1875); Provancher, Nat. Canad., IX, p. 43 (1877); Maclaughlin,
Trans. Ott. Field Nat. Cl., II, pp. 334, 335 (1887); Wadsworth, Ent. News, I,
Pp. 37 (1890); Kirby, Syn. Cat. Neur. Od., p. 89 (1890); Beutenmuller, Prel. Cat.

d. N.Y., p. 163 (1890); Harvey, Ent. News, II, p. 73 (1891); Banks, Trans.
Am. Ent. Soc., XIX, p. 353 (1892); Kellicott, Journ. Cinc. Soc. N.H., XVII, p.
212 (1895);1. c. XVIII, p. 114 (1896); Kellicott, Od. Ohio, p. 84 (1899); Williamson,
Drag. Ind., p. 304 (1899); Burnham, Manch. Inst. Arts Sc., I, p. 31 (1900); Need-
ham, Bull. 47, N.Y. State Museum, p. 469 (1901); Needham and Hart, Bull. III.
State Lab., VI, p. 46 (1901); Fyles, 31st Ann. Rep. Ent. Soc. Ont., p. 54 (1901);
Harvey, Cat. Bibl. Od. Me., p. 8 (1902) ; Williamson, Ent. News, XIV, p. 7
(1903); Needham and Anthony, Pr. N.Y. Ent. Soc., XI, p. 121 (1903) ; Walker,
Can. Ent., XX XVIII, pp. 149, 150 (1906); Muttkowski, Bull. Wis. N. H. Soc.,
VI, page 96 (1908); 1 c , p. 167 (1908).

Aeshna verticalis, Williamson, Ohio Nat., VII, p. 150 (1907); Walker, Can.
Ent., XL, pp. 385, 389, 451 (1908) ; Muttkowski, Cat. Od. N.A., p. 114 (1910);
Bull. Wis. N.H. Soc., VIII, p. 174 (1910).

Aeschna juncea verticalis, Calvert, Trans. Am. Ent. Soc., XX, p. 284 (1893);
Ent. News, V, p. 11 (1894) ; Pr.Cal. Ac. Sc., IV (2), p. 503 (1895); Jr. N.Y. Ent.
Soc., III, p. 45 (1895); p. 93 (1897); Davis, Jr. N.Y. Ent. Soc., p. 196 (1898);
Calvert, Ins. N.J., Od., p. 71 (1900) ; Weith, Ent. News, XI, p. 641 (1900);
Martin, Cat. Coll. Zool. Selys, XVIII, p. 35 (1908); Smith, Ins. N.J., p. 78 (1909);
Martin, Gen. Ins., CXV, p. 11 (1911).

Aeshna juncea verticalis, Calvert, Occ. Pap. Bost. Soc. N. H., VII (6), p. 23
(1905); Wilson, Pr. U.S.N.M., XXXVI, pp. 656, 659, 665, 667 (1909).

Aeschna propinqua, Scudder, Pr. Bost. Soc. N.H., X, pp. 214, 215 (1866) [in
part]; Hagen, Pr. Bost. Soc. N.H., XV, p. 376 (1871).

Male—A medium-sized species of slender form. Occi-
put greenish yellow, equilateral or somewhat broader than
long, two- to three-fifths as long as the line of contact of the
eyes. Frontal vesicle twice or slightly more than twice as
broad as long, the yellow area not quite reaching the lateral
ocelli. Eyes green or grey-green, varying considerably in
shade, pale yellowish green next the hind margin. Pre-
ocular band, at the base of the antenne, rarely less than twice
as broad as one of the lateral ocelli, but reduced to an ex-
tremely narrow line on the sides of the frons and nasus.
T-spot heavy, 2.5-3 mm., the stem generally short with di-
vergent sides, .6-.7 mm. broad in front, I-1.33 mm. broad be-
hind. Face yellowish green to olivaceous with a pale yel-
lowish ocular border; a fine brown line on the fronto-nasal

146 WALKER: NORTH AMERICAN SPECIES OF AESHNA

suture; lateral lobes of nasus not flaring, nearly rectangular,
with the apices more or less rounded. Rhinarium reddish
brown. Labrum pale dull green, or brownish green, often witha
plumbeous tinge, margined very narrowly above, more
broadly below with black. Labium varying in the living
insect from dull greyish obscured with reddish brown to
ivory-yellow behind, bluish white more or less obscured by
brownish in front, the lateral lobes pale blue or plumbeous,
stained apically with reddish brown, Rear of head black.
Thorax of average size, dark reddish brown, with a
slight greyish bloom beneath. Dorsal bands conspicuous,
pea-green or yellowish green, 4 mm. long, expanding dorsad,
.66-1.25 mm. broad at the middle, lower ends pointed and
separated by 4-4.5 mm., upper ends at the antealar sinus
separated by about I mm., expanded, I.5-1.75 mm. broad.
There is also a small green spot upon the antealar sinus some-
times connected with the corresponding dorsal band, as well
as a lagger green spot just in front of the humeral suture
near its middle. Lateral bands fairly broad, yellowish
green or pea-green, the second band often more or less blue
in the upper half, narrowed just above the middle to .7-.9
mm. by the obtusangulate excavation of the anterior margin,
then widening again very slightly (1 mm.) and giving off at
the upper end a posterior offshoot of nearly the same breadth.
The posterior margin of the band is more or less sinuate, but
less strongly so than the anterior margin. Second band
widening gradually from the pointed lower end to the upper
end, which is about 2 mm. broad, middle breadth I mm.;
anterior margin curving outwards at the upper end, posterior
margin straight. Interalar spots blue-green and green.
Femora dark reddish brown above, black beneath, the
posterior surface of the fore trochanters and basal fourth or
third of the fore femora pale yellowish, tibiae and tarsi black.
Abdomen nearly 4.5 times as long as the thorax, slender,
strongly constricted at 3. Seg. I without a ventral tubercle.
Spinulose area on ventral surface of 2 restricted to the apical
third; auricles with 3 (2-4) teeth. Spines of anterior lamina
rather large for the clepsydra group, straight, acute, directed
ventrad, caudad and somewhat laterad. Hamular processes

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 147

directed ventrad and mesad, consisting of a short broad
proximal and a slender papilliform distal part. Hamular
folds continuous with the hamular processes closely approxi-
mated above; outer margins divergent, meeting the hind
Margins in a rather broad curve. Seg. 3 about one-third
longer than 4, which is very slightly longer than 1-+2.
Lateral carinae on 8 more or less angulate near the middle.
Dorsum of 10 with an abruptly elevated median sub-basal
tooth, between two pairs of much smaller teeth.

Superior appendages about as long as 9+10 or slightly
longer, their basal breadth one-fourth or two-sevenths that
of the intervening space; slender but slightly widening in the
basal fifth, then expanding more rapidly mesad to about two-
fifths or one-half of the length, the breadth here, viewed dor-
sad, being one-sixth to one-fifth of the length, and narrowing
again to the more or less acute and slightly decurved apices,
which terminate in a short sharp spine. Outer margin,
viewed dorsad, gently convex; inner margin gently sinuate,
usually appearing nearly straight in the distal half before the
curve of the apex. In profile the outer margin is curved
gently upwards except apically, the inner margin bent down-
wards forming an arcuate inferior carina. Superior carina
feebly marked or obsolete on the slender basal part, distinct
beyond and somewhat elevated and gently arcuate in the
apical two-fifths ; normally without denticles.

Inferior appendage elongate-triangular, barely more than
half as long as the superior pair and about twice as long as the
basal breadth, sides straight; a pair of minute recurved spines
just before the bluntly pointed apices; in profile view moder-
ately curved, tapering evenly to the apex.

Colour-pattern of abdomen — Seg. 1 fuscous; dorsal spot
blue, 3.5 mm. broad; lateral spot pea-green, rarely bluish, 3.5-4
mm. long and about I mm. broad near the lower end, taper-
ing dorsad.

Seg. 2 fuscous; AML bluish green to blue, 1.5 to nearly
2 mm. broad at the middle, giving off a narrow prolongation
ventrad along the anterior margin and another dorsad along
the transverse carina, the latter quite widely separated from
MD. PL blue, very narrowly connected with, or barely

_

148 WALKER: NorTH AMERICAN SPECIES OF AESHNA

separated from, PD, which forms with its fellow of the op-
posite side a blue band about 1.3 mm broad, from which a
more or less ill-defined median green streak passes cephalad.

Segs. 3-10 brownish black; all the spots azure blue except
MD, which is pale green. AL on 3-8, of moderate size,
dwindling to a mere dot on 8, rounded below, giving off, on
3-6 or -7,a basal transverse offshoot, which on most of the
segments forms a complete though narrow ring with its
fellow of the opposite side; upper margins straight or slightly
concave. ML on 3-8, enlarging to 6, rhomboid on 3-4 or -5,
oblong and subangular or rounded behind on the other seg-
ments; front margin feebly or not at all indented. MD on
3-7, triangular, the postero-lateral margins concave; often
more or less quadrate on 3 and 4, and occasionally confluent
with ML on 3. PD I-1.33 mm. long on 3, increasing to 1.75
or 2 mm. on 8, rounded mesially and rather narrowly separ-
ated on 3-7, more widely separated on 8 and 9, the inner mar-
gins straight and more or less divergent cephalad, particularly
on 9; well-developed on 10, subquadrate and separated by a
variable interval, but neverconnate. PL usually represented
on 3 and 4 only, but often also on 5 and rarely on 6. On
3 it is a well-marked offshoot from PD, on 4 smaller and more
curved, generally an isolated dot when present on 6 or 7.

Wings hyaline, costal veins brownish ochraceous, pter-
ostigma dark smoky brown above, paler and more yellowish
beneath, membranule of hind wing smoky brown, uniform
or slightly paler at base, usually not reaching the cross-vein of
the anal triangle which is 2-celled. Two cells between A2and A3
at their origin. Rs forking at the level of 2-3 (fore wing)
2-4 (hind wing) postnodal cells before the stigma; 3 rows be-
tween the forks at the level of the distal end of the stigma and
3-6 cells between them at the margin. Three or 4 (rarely 5) rows
of cells between Rs and Rspl, where most widely separated.
Mia arising before the level of the middle of the stigma,
generally near the proximal end.

Antenodals thd postnodals ae. CuCr 7 Spt 35,
12-15 13-17 5-7 2-4
Female—Abdomen about 3.5 times as long as the thorax

and barely or no shorter than the hind wing.

WALKER: NorTH AMERICAN SPECIES OF AESHNA 149

Genital valves as long as the dorsum of 9, lateral carinae
very prominent, except at the base, varying considerably in
profile view, but more or less gently arcuate with a tendency
to be subangularly bent about the distal third; in ventral
view nearly straight, bounding a space which is broadest in
the proximal two-fifths, the breadth here being about one-
third of the length, thence narrowing to the moderately
broad squarely cut apices, which bear each a minute pencil of
whitish hairs. Ventral surfaces of valves with a more or less
well-defined sulcate area occupying about the middle third,
distally narrowed and sloping mesad, feebly sulcate; inner
margins touching only at a point near the middle, beyond
which they are somewhat divergent. Styli .6-.7 mm. long, or
half as long as the dorsum of 10. Basal plate of ovipositor
with the hind margin straight. Lateral plates distinct.
Appendages about as long as 8+9, slender, expanding to
about three-fifths of the length, where the breadth is one-sixth
of the length, thence tapering slightly to the rounded apices;
curve of inner margin slightly stronger than that of the outer.

The dorsal thoracic bands vary considerably in
development but are usually complete, though some-
what narrower than in the male (middle breadth .5-.8 mm.)
and more constricted below the expanded upper ends, which
are sometimes separated as distinct transverse spots. Lateral
bands averaging slightly broader than in the male. Abdom-
inal spots shewing the usual differential characters of the sex,
PD being smaller, the lateral spots larger, than in the male. —
PD is particularly reduced on 3, and on 2 is but narrowly
connected with, or separate from, its fellow of the opposite
side. PL on 2-6 or -7, relatively large, subtriangular on 3-5,
generally connected throughout with PD and broadly con-
fluent with AML on 2, and often narrowly so with ML on 3.

The extent of colour variation is as yet uncertain. The
individual figured on pl. 25, fig. 6, wasstrongly heterochromatic,
the spots being all decidedly green or greenish yellow, but in
the majority the lateral spots are more or less bluish green, the
thoracic bands and dorsal abdominal spots yellowish green.
I have never seen a purely homceochromatic female.

The wings vary from hyaline to a moderate degree of

150 WALKER: NORTH AMERICAN SPECIES OF AESHNA

flavescence. The stigma is, as usual, somewhat paler than in

the male.

The following colour notes were made from two specimens

taken at De Grassi Point, Lake Simcoe.

Specimen I.

Occiput and frontal Greenish yellow.

vesicle.
Eyes

Face

Labium.

Thorax.

Abdomen.
Seg. I.
Seg. 2.

Segs. 3-10.

Greyish brown, pale yellow
hind margin.

Olivaceous, pale yellow lat-
erally.

Ivory-yellow, below middle
lobes bluish white, lat. lobes
pale dull blue.

Dark reddish brown; bands
pea-green. Spots between
wings yellowish green.

Dark reddish brown.

D pale blue, L pale yellowish
green.

AD, MD light yellow. PD,
pale yellowish green. AML,
PL pale green.

MD light yellow. PD pale
yellowish green, becoming
more bluish caudad. Lateral
spots all very pale green in-

Specimen 2.
Lemon yellow.

Brown, light yellow hind
margin.
Pale yellowish olivaceous.

Pale drab with a pinkish
tinge, brownish about mar-
gins.
Chocolate with a grey bloom;
bands and spots between
wings bright yellowish green.
Dark reddish brown.

D and L yellowish green.

All the spots bright yellow-
ish green.

MD, PD and PL yellowish
green. AL and ML pale
green. Ventral spots pale
greenish.

clined to bluish.

Measurements—Thor. o& 10.5-11.5, 9 10-11; abd. of
46.5-51.3, 9 43.25-50; seg. 3 co’ 8.6-9.2, 9 6.5-8.1; depth
seg. 2 o’ 5-5.8, 9 5-6; depth seg. 3 9 2.9-3.1; apps.c" 5.33-
5-75, @ 5-75-73 gen. v. 2-2.6; h.w. co 45.5-49.5, 2 43.2-48;
width h.w. o -13-13.9, 9 13-5-15; pter. o& 3-3-4, 2 3-33-4.

Type—o', Mus. Comp. Zoology.

Nymph—Unknown.

Material determined—80 oc 33 9. Matne: Manchester, July 18, 1890
(1 o*), Aug. 15, 1895, Aug. 22-Sept. 8, 1908, Aug. 23-Sept. 3, 1909, Sept. 2-7, I910
(Miss Wadsworth, 11 co 8 2) ; West Beach (Hagen, M.C.Z., 1 oc). MaAss-
ACHUSETTs: (Riley, U.S.N.M., Méring, M.C.Z., 2 co); Beverley, Sept. 11,
1870 (M.C.Z., 1 co); Cambridge, (M.C.Z., 1 9); Nahant, 1874 (MGring, M.C.Z.,
Io 3 2); Salem (Joseph Lane, M.C.Z.,1 &); Saugus, (M.C.Z., 1 9); Sher-
bourne, Sept. 17, 1864 (M.C.Z., 1 co”); id. (Babcock, Ac. N.S. Phil., 1 o); Wal-
pole, Aug. 30, 1896 (F. H. Sprague, M.C.Z., 1 9); Westborn, Aug. 30, 1894
(Needham, 1 9 ); Wollaston, Sept. 20, 1896 (F. H. Sprague, M.C.Z., 1c’). NEW
York: Long Island; Sept. (O. Sacken, M.C.Z., 1 9). PENNSYLVANIA: Akron, Sept.
28, 1897 (J. S. Hine, 1 o); Folsom, Delaware Co., Oct. 4, 1890 (Calvert 1 o);
Ridley Township, Del. Co., Oct. 11, 1890 (Calvert, 1 9); Neville Isl., Sept. 27,
1908 (J. L. Graf, 4 o&); Squaw Run, Alleghany Co., Sept. 6, 1908 (J. L. Graf,
coll. Williamson, 1 o’'). District oF CoLuMBIA: Washington (O. Sacken, M. C.Z.,
1 ¢'). ONTARIO: Toronto, Sept. 12, 1907, Sept. 27, 1909 (Walker, 10 3
9); Scarboro’ Jc., Sept. 7, 1907 (Huntsman, 1 co’); De Grassi Point, Lake
Simcoe, July 17, 18, 1904 (Walker, 2 co) ; id., Aug. 21, 28, 1901, Sept. 5, 1906,

WALKER: NorTH AMERICAN SPECIES OF AESHNA I5I

Aug. 18-25, 1908, Aug. 30-Sept. 6, 1909, Aug. 27, Sept. 7, 1910 (Walker, 12
5 2); Go Home Bay, Georgian Bay, Aug. 26-30, 1907 (Huntsman,2 o I 9).
MicuiGAN : Detroit, M.C.Z., 3 9) ; Petroskey (Mrs. Godfrey Ashbaucher, coll.
Williamson, 1 o). Itiinots: Chicago, June 13, 1906 (J.D. Hood, 10%); id.
(Bolter, M.C.Z.,1 9); InpIANA: Fort Wayne, Aug. 23, Sept. 13, 1908 (William-
son, 2 6’); id., Sept. 17, 18, 1906, Sept. 15, 1907 (Williamson, 16 @' I 9);
Bluffton, Sept. 17, 1906 (L.A. Williamson, 1 o&); Crooked Lake, Steuben Co.,
Sept. 1, 1905 (Williamson, 19). Wusconsin: Milwaukee Co., June 14, 1908
(Muttkowski, I o’, teneral); id., Aug. 10, 1908 (Muttkowski, 1 9) ; Magnolia,
Sept. 9, 1880 (Hagen, M.C.Z.,1 @i 9).
Also 3 o&' 52 without data.

Identity—This species bears a superficial resemblance to
Ae. juncea \.., especially in the coloration and form of the
male appendages, and was placed by Calvert (’93) and Martin
(08) as arace of that species. It is not, however, at all close-
ly related to juncea but finds its nearest relative in Ae. can-
adensis. From this species it differs very obviously in the
form of the anterior hamuli, and may be easily distinguished
from it in the field by the much brighter thoracic bands. The
lateral bands differ but slightly in form from those of can-
adensis but this difference is very constant. The undenti-
culated superior carina of the male appendages is an almost
constant feature but occasional individuals occur in which a
few slight denticles are present. This is the case in the two
specimens from Sherbourne, Mass., which were referred by
Calvert (’94) to Ae. clepsydra.

Distribution—Alleghanian and Carolinian Zones from
the Atlantic States to Wisconsin. In addition to the lo-
calities given above it has been recently reported by Mutt-
kowski from Florida (‘10 b). Some of the older records, such
as the one from California (Calvert, ’95), need confirmation, as
more than one species have been quoted under the name
verticalis. |

Habits—All my captures of this insect were made about
the borders of woods or in small open spaces surrounded by
trees. Inits flight and its habit of sunning itself on the trunks
of trees or hanging from the branches it resembles Ae. can-
adensis, with which it is sometimes associated, though it
usually makes its first appearance at a time when Ae. can-
adensts is leaving the woods and returning to its breeding
grounds. Although two specimens, one a teneral, are re-
corded above as having been captured on June 13 and 15, the
normal time of appearance of this species is during the latter

152. WALKER: NorTH AMERICAN SPECIES OF AESHNA

half of July, while it is most numerous towards the end of
August and in early September. On Sept. 12, 1907, it appear-
ed in very large numbers about the borders of a wood near
Toronto.

Two pairs were taken here im copula and I found a third
pair in an open wood at De Grassi Pt.,Ont., on Aug. 27,
1910. They were resting on the branch of a tree and by ap-
proaching very quietly I was able to pick them up with the
fingers, but they separated immediately.

I have never discovered the breeding grounds of this
species nor do I remember ever to have seen it in the immedi-
ate neighbourhood of water... Mr. Williamson, however,
has taken it at Fort Wayne, Indiana, flying over an open
marsh, in company with Ae. constricta.

Aeshna tuberculifera Walker.

(Pi: 23, fig. .6;: pl..17, figs: f, Tas plize, figs. 2. 2a;
pl. 26, figs. 1, 2.)

, ae juncea verticalis, Calvert, Occ. Pap. Bost. Soc. N.H., VI, p. 23 (1905)
in part
Aeshna tuberculifera, Walker, Can. Ent., XL, pp. 385, 387, 451 (1908).
Aeschna tuberculifera, Muttkowski, Bull. Wis, N.H. Soc., VI, p. 167 (1908);
Martin, Gen. Ins., CXV, p. 11 (1911).

Male—A slender species of medium or rather large size.
Occiput pale bluish green, the antero-lateral margins blackish,
usually distinctly broader than long, one-third to two-fifths
as long as the line of contact of the eyes. Frontal vesicle
reniform, more than twice as broad as long, the yellow area
not reaching the lateral ocelli. Eyes in alcoholic specimens
deep yellowish olivaceous shading into bright yellow behind,
with a small transverse dark spot above. Preocular band not
more, generally less, than twice as broad as the first antennal
joint, narrowed ventrad on sides of frons, becoming obsolete
at the fronto-nasal suture, but broadening again on the nasus.
T-spot 2.33-3 mm., stem .4-.75 mm. broad in front, .8-1 mm.
behind, sides generally straight, more or less divergent.
Face (alcoholic) dull olivaceous with a slight bluish tinge, in
dried specimens yellowish olivaceous; paler above and on the

1 Since the above was written I have seen a few males hovering over small
pools on the borders of woods near Toronto. One of these pools is a favourite
breeding spot for Ae. umbrosa but asa rule it is not frequented by other species of
the genus.

WALKER:. NoRTH AMERICAN SPECIES OF AESHNA_ 153

sides next the preocular line. A fine brown line generally
present on the fronto-nasal suture. Lateral lobes of nasus
rotundo-angulate, scarcely flaring. Rhinarium reddish brown,
the upper part green. Labrum light green with very narrow
upper and lower marginal black lines. Labium pale ob-
scure yellowish or olivaceous stained with reddish brown to-
ward the fore margins. . Rear of head black.

Thorax moderately slender, dark reddish brown, deepen-
ing to black around the lateral bands, with a slight greyish
bloom beneath. Dorsal bands pea-green, nearly straight,
3.5-4. mm. long, .75-1 mm. broad at the middle, lower ends
pointed, separated by about 4 mm., upper ends reaching the
antealar sinus, expanded, about 1.5 mm. broad, separated by
I-I.5 mm. Antealar sinus pea-green. Lateral bands blue
above, fading into pale green below (pale yellowish in dried
teneral specimens), straight, the first band rounded at the
lower end, I-1.5 mm. broad in lower third, thence narrowing
dorsad to about I mm. in the upper half and widening again
slightly at the upper end, from which a very small offshoot
passes caudad; anterior margin sometimes slightly broken
above the middle, posterior margin nearly straight. Second
band tapering somewhat at the lower end, the margins
straight and subparallel; breadth at middle 1.3-1.5 mm.,
slightly broader at upper end. No spots, or at most a minute
dot, between the bands. Spots between and at base of wings
blue: : .

Legs. dark reddish brown, nearly black on the tibiae,
tarsi and underside of the femora; fore femora without a pale
streak on the posterior surface.

Abdomen slender, five times as long as the thorax,
strongly constricted at 3. Sternum of1 slightly elevated at
the hind margin, but without a distinct tubercle. Auricles
with 4 or 5 teeth. Tergal margins of 2 opposite sheath of
penis subangularly approximated. Spines of the anterior
lamina short and blunt, not reaching back to the hamular
folds. Hamular..processes short and very. broad, directed
mesad, the ventro-anterior surfaces flat, postero-internal
Margins arcuate, apices bluntly obtusangulate, with no in-

dication of a tubercle. Hamular folds continuous with ham-
II

154 WALKER: NORTH AMERICAN SPECIES OF AESHNA

ular process, not so closely approximated behind as in most
species of the clepsydra group, margins thickened, outer
margins parallel. Seg. 3 about one-third longer than 1+2
or 4. Lateral carinae on 8 more or less subangulate or ir-
regularly sinuate, on 9 distinct except for a short distance near
the middle, where they are more or less obliterated. Dorsum
of 10 with a prominent median and two pairs of small lateral
teeth.
Superior appendages usually slightly shorter than 9 + 10,
sometimes as long, contracted at the extreme base, moder-
ately slender in the proximal fourth, being about half as broad
as the intervening space, thence expanding meso-caudad to
-about two-fifths of the length, the width here being about one-
fifth of the length, narrowing very little distally, the apices
being broadly rounded with a small deflexed terminal tooth
near the outer margin. Outer margin ‘viewed dorsad,
gently convex proximally, more strongly so distally; inner
margin rather markedly sinuate, being convex just beyond
the base, just before the middle and before the apex. Super-
ior carina obsolete in the proximal third, gradually and moder-
ately elevated beyond, gently arcuate before the apex, but
with no indication of denticles. Profile view: slightly curved
upwards; a prominent basal dorsal tubercle and a well-marked
sub-basal ventral tubercle, which appears most prominent
when viewed obliquely either from above or below. Inner
margin in the distal four-fifths bent downwards, forming
a fairly prominent arcuate inferior carina, apices slightly
decurved. Inferior appendage triangular, with nearly straight
sides, one-half to three-fifths as long as the superior pair, the
basal breadth one-half to three-fifths of the length; a pair of
minute recurved dorsal spines before the bluntly pointed
apices; in profile very gently upcurved, tapering evenly to the
apex.
Colour-pattern of abdomen—Segs. 1 and 2 fuscous, the
others brownish black, pale spots all slightly reduced.
Seg. 1. Dorsal spot blue, lateral spot wanting.
Seg. 2. AML pale green to azure blue, L-shaped, the
horizontal part I-1.5 mm. broad just above the auricle and
extending downwards a short distance along the front margin

WALKER: NortH AMERICAN SPECIES OF AESHNA_ 155

to the ventral surface; vertical part somewhat narrower, ex-
tending dorsad 2-2.5 mm. above the auricle, ending abruptly
and not confluent with MD. PL pale green to azure blue,
only about 1 mm. broad, connected rather narrowly with PD,
which is about 1.3 long (antero-posterior dimension), con-
nate with its fellow of the opposite side, and in front
with a narrow pale median greenish streak.

Segs. 3-10. AL on 3-8, pale green to blue, smaller than
usual, diminishing caudad to a dot, subtriangular, with the up-
per margins oblique;on 3 about 2.5 mm. broad in front, on the
next 2 or 3 segments sometimes giving off dorsad a short nar-
row basal offshoot. ML 0n3-8, pale green to blue, increasing in
size caudad to 6; more or less rhomboid on 3, and sometimes
4, rounded on the succeeding segments, not confluent with
MD. MD on 3-7, pale green, small and subquadrate on 3
and reduced to a pair of dots on 7, elsewhere triangular, but
little elongated transversely. PD on 3-9, azure blue, increas-
ing in length from I mm. on 3 to 1.3-1.75 on 8; inner margins
rounded and separated by about .5 mm. or somewhat less,
except on 8 and 9, where they are straight, somewhat diver-
gent cephalad, and more widely separated. Seg. 10 entirely
black. PL on 3 only, a small curved offshoot from PD some-
times separated as a minute dot or absent altogether.

Wings hyaline or slightly flavescent in tenerals, costal
veins rather dark reddish brown, pterostigma dark brown,
membranule of hind wing dark smoky brown, fading into pale
greyish at base, not reaching cross-vein of anal triangle, which
is 2-celled. Twocells between A2 and A3 attheir origin. Rs
forking at the level of 2-5 (fore wing) 3-6 (hind wing) post-
nodal cells before the stigma; 3 (rarely 2 or 4) rows of cells
between the forks at the level of the distal end of stigma and
3-6 cells between the forks at the wing-margin. Three or 4 rows
of cells between Rs and Rspl, the latter not strongly sinuate;
Mia arising at a point between the level of the middle and
proximal ends of the stigma (almost always near the latter).
Three (occasionally 2 or 4) cross-veins under stigma.

Antenodals + oes postnodals py CuCr sath Spt £4.

11-16 8 5-6 2-3

oh

156 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

Female—Abdomen scarcely five times as long as thorax,
slightly widened at apex of 8 and base of 9, the parts of the
genitalia larger than in any other North American species of
Aeshna except Ae. constricta. Genital valves as long as dor-
sum of 9, or barely longer, in profile arcuate toward the base,
the apices slightly elevated; lateral carinae prominent, per-
current, in profile sinuate, in ventral view nearly straight, the
enclosed space narrowing slightly caudad ; its middle breadth
about one-fourth the length; apices rather broad and truncate.
Ventral surfaces of valves distinctly declivent, especially at
the middle, feebly sulcate. Basal plate of ovipositor of
moderate size, hind margin straight; lateral plates distinct.
Spines on ventral surface of 10 few and scattered, rather
coarse. Styli 1.5 mm. long, or about as long as the dorsum of
10. Appendages as long as, or slightly longer than, 8+9, ex-
panding rapidly from the base to one-fourth or one-third the
length, the breadth there being slightly more than one-fifth
of the length, tapering in the distal half to the acute or
roundly angulate apices, which sometimes terminate in a
minute spine. The curve of the margins is generally stronger
towards the base than the apex and that of the inner margin
stronger and more regular than that of the outer.

The colour-pattern shows the usual marks of dis-
tinction from that of the male: the dorsal thoracic
bands narrower (.5 mm.), thelateral abdominal spots somewhat
larger and PD slightly smaller and on 9 more widely separated.
Ill-defined traces of PD are sometimes present on 10.
PL, as in the male, is represented only on 2 and 3, though a
little larger, and confluent on 2 with AML.

In the single female in which the colours were seen while
still fresh they were purely homceochromatic (pl. 26, fig. 2).
None of the dried specimens are well enough preserved to
determine the natural colours, though in some the PD spots
were apparently blue in life, while in one from Hampden,
Mass., these spots on the posterior segments are yellow and
were probably so in life, so that Ae. tuberculifera is probably
more or less dichromatic, like related species.

The wings are sometimes slightly flavescent, especially at
the bases and along the costal borders, but I have seen no

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 157

specimens with deeply coloured wings. The pterostigma is,
as usual, generally longer than in the male.

Measurements—Thor. o& 9.5-10.5, @ 10-11; abd. & 47.5-
52.5, 9 45-54. 5; seg. 3 of 8-9.66, 9 7.5-9; depth seg. 2 of 5.3-
5-75, % 5-75-6.25; depth seg. 3 9 2.33-2.9; apps. @* 4.8-5.75,
? 7.-33-8.33; gen. v. 3.1-3.5; h.w. o& 45.5-49, 2 44-50.5;
width h.w. o 13-14.6, 9 12.5-15; pter. co 3-3.6, 2 3.5-4.

Types—o 9 , Acad. Nat. Sc. Philadelphia—Isleboro, Me.,
taken 2n copula.

Nymph—Unknown.

Material examined—14 o' 11 9. Marne: Isleboro (Dr. S.G. Dixon, Ac.
Nat. Sc. Phil., o 9 in cop.); Manchester, Aug. 26, 1895, July 20, Aug. 8,10, 1910
(Miss Wadsworth, 7 o' 3 9). Massacnusetts: (A. N. Caudell, U.S.N.M.,
I o); Walpole, Aug. 30, 1896 (F. H. Sprague, M.C.Z., 1 o); Essex Co. (Hagen,
M.C.Z., 1 9); Provincetown, Aug. 5, 6, 1899 (J. E. Benedict, jr.. U.S.N.M, R
C. Osburn, 3 o' 2 9) ; Hampden, Aug. 7, 1902 (Needham, 1 2); Westborn,
Aug. 30, 1894 (Needham, 19). Ontario: East Toronto, July. 3, 1907 (Hunts-
man, I 0’); Etobicoke Creek, near Toronto, Sept. 30, 1907 (Walker, 1 9); Go
Home Bay, Georgian Bay, Aug. 26, 1907 (Huntsman 1 go’). WISCONSIN: Oost-
burg, July 16, 1908 (E. Was, coll. Muttkowski, 1 9).

Distribution — Alleghanian; New England, Ontario
and Wisconsin.

Aeshna palmata Hagen.

(Pl. 1, fig. 4; pl. 14, fig. 1; pl. 20, figs. 4, 4a; pl. 26, figs. 3, 4.)

Aeschna palmata, Hagen, Stett. Ent. Zeit., XVII, p. 369 (1856) ; Selys, Ann.
Soc. Ent. Belg., XV, p. 34 (1872); Martin, Gen. Ins., CXV, p. 12 (1911).

Aeshna palmata, Walker, Can. Ent., XL, pp. 379, 388, 450 (1908); Mutt-
kowski, Cat. Od. N.A., p. 113 (1910); Bull. Wis., N.H. Soc., VIII, p. 174 (1910).

Aeschna constricta, Hagen, Pr. B.S. N.H., XVIII, p. 34 (1875); Kirby, Syn.
Cat. Od., p. 88 (1890) [in part]; Calvert, Trans. Am. Ent. Soc., XX, p. 249 (1893)
{in part]; Currie, Pr. Wash. Ac. Sc., III, p. 220 (1901); Pr. Ent. Soc. Wash., VII,
p. 18 (1905); Osburn, Ent. News, XVI, pp. 185, 190 (1905).

Aeshna constricta, Calvert, Od. B.C.A., p. 183 (1905) [in part].

Male—A strongly built species of medium or rather large
size. Occiput pale yellow or yellowish green, one-half to
three-fifths as long as the line of contact of the eyes. Yellow
area of frontal vescicle extending laterally to the lateral ocelli.
Eyes blue above, with a dark transverse stripe, pale greyish
brown below, hind margin whitish. Preocular band varying
in width, like the other dark markings of the head, according
to locality, but almost always two or three times as broad at
base of antennz as one of the lateral ocelli, narrowed to a
heavy black line on the sides of the frons and nasus. T-spot
heavy but exhibiting great variability, generally 2.5-3 mm.,

158 WALKER: NORTH AMERICAN SPECIES OF AESHNA

strongly arcuate in front; stem short, .6-.75 mm. broad, with
parallel sides. In specimens from Banff this spot is inclined
to be a little under the average size, while in examples from
Vancouver Island it reaches its greatest development, the
cross-bar measuring 3-3.33 mm., and proportionately wider,
so that the stem is shortened, and in one example from Depart-
ure Bay has virtually disappeared, the cross-bar being di- .
rectly connected with the preocular band. Frons, nasus and
labrum pale green or greenish yellow, a black line on the
fronto-nasal suture, varying in width according to locality.
In one unusually large and pale example from City Creek
Cafion, the line is brown and extremely narrow. Lateral
lobes of nasus rotundo-obtusangulate, scarcely or not at all
flaring. Labrum bordered above and below by a black line.
Rhinarium brownish or plumbeous. Labium pale dull yel-
low, more or less obscured by reddish brown. Rear of head
black.

Thorax robust, dark reddish brown, paler beneath.
Dorsal bands yellowish green, .6-.8 mm broad, tapering below,
expanded suddenly at the extreme upper end, this expanded
portion being frequently separated as a distinct spot or absent
altogether, the band in such cases tapering at both ends. No
humeral spots or bands. Lateral bands bright greenish
yellow, often bluish above, nearly straight, more or less dis-
tinctly surrounded by a diffuse dark brown margin; first
band broadest near the rounded lower end (1 mm. or a little
more), straight or feebly sinuate about the middle, somewhat
narrowed above, not expanded at the upper end and without
a posterior offshoot. Second band of about the same width
below as the first band, straight, slightly expanded at upper
end. There are often two very small greenish spots between
the two lateral bands, one at the metastigma, the other above
it, but these are not always present.

Legs black, femora more or less reddish brown above in
alcoholic specimens. Anterior trochanters and basal fourth
or more of posterior surface of anterior femora pale yellowish.

Abdomen about four and one-half times as long as thorax,
rather robust, strongly constricted at 3. Seg. 1 without a
distinct ventral tubercle. Auricles with 4-5 teeth. Tergal

WALKER: NorTH AMERICAN SPECIES OF AESHNA 159

margins of 2 approximated opposite sheath of penis. Spines
of anterior lamina strongly developed, projecting caudad about
as far as the hamular processes, moderately and evenly curved
dorsad. Hamular processes broad and thick, their antero-
lateral surfaces concave and bounded behind by an arcuate
ridge which runs from base to antero-internal angle; apical
Margin straight and transverse; inner margins approximate in
front, divergent behind, where they are in unbroken continuity
with those of the hamular folds. Hamular folds nearly
meeting behind, sinuously bent, but not produced into ear-
like processes. Seg. 3 one-fifth to one-fourth longer than 1
+2 or 4; lateral carinae of 7 and 8 feebly sinuate in ventral
view ; dorsum of 10 smooth, without a vestige of a tooth.

Superior appendages about as long as 9+10, their bases
comparatively stout and as broad as the space between them,
expanding rapidly by the curving of the inner margin, so
that the greatest breadth, at about three-fifths of the length,
is equal to about three-tenths of the length. Outer margin
curled inwards over the dorsal surface, feebly convex in the
proximal half or two-thirds, straight or feebly concave be-
yond, passing distally into the broad rounded apical margin.
Inner margin strongly concave in the basal half, very strongly
convex beyond and emarginate before the apex, where it
terminates in a slender pointed spine, directed downwards
and backwards and surmounted by a stout inwardly directed
tubercle, which bears a tuft of brown hairs. Dorsal surface .
concave, facing rather inwards than upwards in the natural
position. Superior carina indicated by a low rounded median
ridge.

Inferior appendage about half as long as the superior
pair, the basal breadth about half the length; lateral margins
feebly convex next the base, straight beyond and meeting in
the rather broadly rounded or truncated apex, which is armed
above with a pair of minute recurved black spines. Profile
view: gently curved upwards, the curve of the lower margin
being a little stronger than that of the upper.

Colour-pattern of abdomen—Seg. 1 fuscous, dorsal spot
blue, lateral spot greenish yellow, elongate-triangular, of
variable size, 1.5-3mm. long (transverse), I mm., or less, broad.

160 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Seg. 2 fuscous. AD small, triangular, pale yellowish.
AML greenish yellow, about 1.5 mm. broad at middle, giving
- off ventrad a narrow streak along the front margin and con-
fluent above with MD. PL greenish yellow, partly separated
by a deep notch from PD, which is a blue band, from the
front margin of which a narrow medio-dorsal streak runs for-
ward to connect with AD.

Segs. 3-10 black, the blue spots varying somewhat in size
according to locality, but never much reduced. AD on 3-7
or -8,small yellowish triangular spots, rudimentary caudad.
MD well developed, yellowish, triangular on 3-7, and gener-
ally represented by a pair of dots on 8. PD well developed,
generally rather narrowly separated, increasing in size caudad;
on 8 usually nearly half as long as the segment, the inner
margins straight and subparallel and separated by less than
I mm.; on 9 three-fifths as long as the segment, broadly con-
nate or separated by a narrow line; on Io rather narrowly
connate. In specimens from Vancouver Island PD, as well
as the other blue abdominal spots, is slightly smaller than
usual, and more widely separate from its fellow (except on
10, where they are generally narrowly connate). AL on
3-8, blue or frequently more or less green on 3, connected on
most of the segments with a narrow basal ring or narrowly
separated from it; upper margins straight and horizontal.
ML on 3-8, subquadrate becoming larger and more rounded
caudad, not confluent with MD. PL on 3-5, -6 or -7, variable,
decreasing. .

Wings hyaline, costal veins brownish yellow, pterostigma
dark smoky brown, paler beneath, membranule of hind wing
whitish in the proximal fifth, smoky brown distally, reaching
nearly to the cross-vein of anal triangle, which is 3-celled.
One‘cell (rarely 2) between A2 and A3 at their margin. Rs
in front wing forking opposite the first or second postnodal
cell before the stigma, in the hind wing opposite the first to
third cell; 3 or 4 rows of cells between the forks ‘at level of
distal end of stigma and 3-6 cells between the forks at the
wing-margin. Three or 4 rows of cells between Rsand Rspl. Mia
arising beyond the middle of stigma or beyond stigma al-
together. .

WALKER: NORTH AMERICAN SPECIES OF AESHNA I6I

Antenodals ily postnodals ak As Ge 56. Spt es.
11-14 4-5 I-2

Female—Varying considerably in form according to lo-
cality (vide postea); abdomen but little or no longer than the
hind wing. |

Genital valves rather large, reaching slightly beyond the
posterior margin of 9, in profile considerably arcuate, apices
elevated ; lateral carinae percurrent, prominent, in lateral view
somewhat sinuate, in ventral view almost straight, the space
between them broadest at about one-third of its length, the
breadth here being equal to about one-fourth of the length,
narrowing caudad very slightly to the broad truncated apices.
Ventral surface of valves declivent in the proximal third, dis-
tally rather shallowly sulcate. Styli about I mm. long,
shorter than dorsum of 10. Appendages about as long as 9
+10, tapering somewhat more gradually towards the. base
than the apex; greatest breadth about the middle and equal
to one-fifth, or a little more, of the length; curve of inner
margin somewhat stronger than that of outer margin; apices
rounded, sometimes terminating in a more or less obscure
point.

Dorsal . thoracic : bands much -reduced or obsolete,
sometimes divided into a superior and inferior spot.
Abdominal spots differing from those of the male in the usual
way, i.e., PD is smaller, the lateral spots larger, especially
PL, which is sometimes confluent with PD and on 2 with
AML. In the specimen from Twin Lakes, Col., which is
figured on pl. 26, the spots are decidedly larger than usual
and the difference between the two sexes asshownon the plate
is greater than it is in specimens from the same locality. In
the female from Banff, as in the Twin Lakes specimen, PL
is confluent with AML on 2, but elsewhere it is isolated and
the basal rings are mere lateral extensions of AD and do not
reach AL. In individuals from Baker City, Oregon, and
Kaslo, B.C., the spotsarestill smaller and PL is isolated on all
the segments on which it is present. In the Kaslo specimens
the basal rings have apparently disappeared altogether.

162 WALKER: NORTH AMERICAN SPECIES OF AESHNA

The range of colour variation is as great in this as in any
of our species of Aeshna. Mr. C. H Kennedy has sent me a
coloured drawing of a purely homceochromatic female, in
which the abdomen is blackish brown and all the spots on
segs. 3-10 are blue. The upper part of the eyes is also blue.
On the other hand the Twin Lakes specimen is reddish brown,
somewhat darkened about the margins and elevated parts and
the pale markings where well preserved are all yellow with a
greenish tinge. Most of the other specimens seen are not
well enough preserved to reveal the natural colours. One
from Banff was apparently of the intermediate type. The
wings are hyaline in all the specimens except the hetero-
chromatic one from Twin Lakes, in which they are flaves-
cent. The stigma is pale brown.

Measurements—Thor. o& 10.75-11.5, 9 10.5-10.75; abd.
& 47-52, 2 42-47 ; seg. 3 o 7-9, 2 4.33-7.5; depth seg. 2
o& §.25-6, 2 5.5-6 ; apps. &' 2 5.5-6; gen. v. 3-3.5; hw. of
42-46.5, 9 41-42.5 ; width h.w. o 12.66-15, 9 13-14; pter.
O' '2:33-3, “9 + 2.6-3-5:

Type— oo, Museum Comp. Zoology—Kamtchatka.

Nymph (pl. 7, fig. 1; pl. 9, fig. 1; pl. 11, fig. 1)—Very
similar to that of umbrosa, differing chiefly in the stouter
form, broader labium and larger genitalia of the female.

Eyes as in umbrosa, postocular part of head also similar,
except that the postero-lateral corners are somewhat less
broadly rounded. Mentum of labium decidedly broader than in
umbrosa, but strongly narrowed at base, the basal breadth
being less than half theapical breadth, which is equal to about
four-fifths of the length. Proximal four-sevenths expanding
gradually but more rapidly than in wmbrosa, the sides straight;
distal three-sevenths much dilated, the sides strongly arcuate.
Lateral lobes not narrowed beyond the base of the movable
hook, squarely truncate, angles scarcely rounded, the inner
one without a terminal tooth. Supracoxal processes some-
what larger than in wmbrosa, equal, slightly outcurved, acute ;
intervening space rectangular or nearly so.

Abdomen distinctly stouter than in umbrosa, having
about the same form as interrupta and eremita, broadest at
7 or 8; lateral spines present on 6 to 9, a trifle longer than in

WALKER: NORTH AMERICAN SPECIES OF AESHNA 163

umbrosa, those of 8 extending slightly beyond the posterior
margin of the segment, and those of 9 to about three-fifths
or two-thirds the length of 10. Abdominal appendages as
in umbrosa. Genitalia of female extending over the basal
third of 10, the valves slightly declivent.

In the female specimen the pale markings of the
abdomen are obscure, but in the male they are as dis-
tinct as in average specimens of wmbrosa ; and the pattern
is nearly the same in the two species. The pale lateral mar-
gin of the head is less distinct than in palmata, but the an-
nulations of the legs are quite as in umbrosa. The pale
dorsal area of the abdomen in the male specimen is divided
anteriorly into two bands by a dark median band, which
' broadens and becomes diffuse in the more posterior segments.
Just external to the pale bands, in line with the dorso-lateral
punctae, is a series of narrow longitudinal pale streaks. The
lateral scars are outlined in brown and surrounded by diffuse
pale areas. The other markings described for umbrosa are
less distinct than in that species except the punctae, of which
the dorsal as well as the lateral and dorso-lateral are marked
with deep brown.

Measurements—Length of ‘body 40.5-41; mentum of
labium 6.8; h.f. 7-7.2; h.w. 8.5-9; inf. apps. 4.25-4.5; width of
head 8; of abdomen 7.75-8.

The two exuviae from which the above description was
taken were found by Dr. Huntsman at the edge of a pond
near Robson, B.C. Aeshna palmata was flying here and was
the only species seen except a single female of Ae. californica.
These exuviae are very close to those of Ae. umbrosa but are
distinct and can belong to no other known species than pal-
mata. As in the imagoes of these two species the body is
somewhat stouter and the female genitalia a little larger than
in umbrosa. The difference in the abdomen is in reality
greater than appears on the plate, as the drawing of palmata
was made from an exuvia, that of wmbrosa from the nymph
itself.

Material determined —50 of 15 9. KamtcHaTKa: (M.C.Z., 1 o)-
AvasKA: Kodiak, July 20, 21, 1899 (T. Kincaid, U.S.N.M.,1 o2 9) ; Cook’s
Inlet, July 20, 1899 (T. Kincaid, U.S.N.M., 1 o), British CotumBia: (Crotch,
M.C.Z., 1 &@ 19); Departure Bay, Vancouver Is. (Huntsman, 7 0); Victoria

164 .WALKER: NorRTH AMERICAN SPECIES OF AESHNA

(M.C.Z., 1 9); Shawnigan Lake, Vance. Is. (Currie, U.S.N.M., 1 o); Glacier
(Mrs. C. Schaeffer, Ac. N.S. Phil., 1 9); Peachland, Aug. 18, 21, 1909 (Wallis,
3 &); Penticton (Wallis, 2 @); South Fork Creek, Aug. 11, 1903 (Currie
U.S.N.M., 1 2); Kaslo (Currie, U.S.N.M., 1 9). ALBERTA: near Waterton
Lake, 4100 ft. (E. V. Cowdry, coll. Walker, 3 o) ; Banff, July 10-Aug 6, 1908
(Sanson, 2 o' 1 9); Laggan, July 22, rgor1 (Osburn, 1 co). WASHINGTON:
Yakima, Sept. 1894 (Calvert, 1 o&) ; Loon Lake, Colville Valley, July 25, 1882
(Henshaw, M.C.Z., 1 9). OREGON: Baker City and vicinity, July 31, 1908,
Aug. 6-Sept. 18, 1909 (C. H. Kennedy, 16 o& 1 9); Corvallis (Needham, 1 o).
CoLorapbo: Twin Lakes, Aug. 13-31, 1903 (Calvert, coll. Calvert and Williamson,
3 o& 19); Fort Hill, 1873 (M.C.Z., 1 o); Summit Cafion, Aug. 8, 1875
\M.C.Z., 1 oc); ‘‘South-western part of Colorado,”’ 11000-12000 ft. (T.S. Gillin,
Ac. Nat. Sc. Phil., 1 oc‘). YELLOWSTONE: Aug. 1872 (Hayden, M.C.Z., 2 &
19). UtTan: Mountains east of Ogden, Aug. 22, 1897 (Needham, 1 9); City
Creek Cajfion, July 5, 1899 (Calvert, 1 o). NEvapDA: Reno, Aug. 11, 1890 (F.
H Hullman, Ac. N.S. Phil.,1 9). Alsor @ without data.
Nymphs—Exuviae, 1c’ 19, Robson, B.C., July 21, 1907 (Huntsman).

Identity—The Alaska specimens, as far as can be judged
from their discoloured condition, are identical with the type .
specimen from Kamtchatka. In the latter the abdominal
spots are smaller than in any of the American specimens,
but there is much variation in this character, according to
locality.

In the great majority of cases the black line across the
face will serve to distinguish this species from umbrosa and
constricta but occasional individuals occur in Colorado and
Utah in which this line is very narrow or practically absent.
Such specimens may be known by the other characters given
in the key. ;

_ Distribution—Boreal ; Kamtchatka and Alaska to Col-
orado and Utah.

Geographical variations—A considerable range of colour
variation is exhibited by this species in the different parts of
the territory that it inhabits. The extent of the black
markings of the head is greatest in the specimens from Van-
couver Island and least in those from Alberta, Colorado and
Utah. The reverse is the case with the spots of the abdomen;
e.g., the PD spots on seg. 9 of the male are separate in all the
Vancouver Island specimens but are broadly connate in
nearly all of those from the other localities.

In my key of 1908 I gave Baja California as the southern
limit of this species’ range. The specimens from this lo-
cality which I referred to Ae. palmata were part of the material
recorded by Calvert (’95) as Ae. constricta. They differ from

WaLKER: NorTH AMERICAN SPECIES OF AESHNA 165

typical palmata in the broader superior appendages with
shorter preapical spines, in the smaller anterior hamuli,
the somewhat more complex venation (there being, e. g.,
two cells between A2 and A3 at their origin) and in the
slight reduction of the abdominal spots. AL is quite in-
significant on 7 and 8; ML present only to 5 or 6 and very
small, while PLis represented only to 4 or 5, though connected
with PD where present. PD is but little reduced.

As I have seen no females from this locality I prefer to
leave undecided the question of the status of the Baja Cali-
fornia variety.

Aeshna umbrosa Walker.

Aeschna constricta, Scudder, Pr. Bost. Soc. N.H., X, p. 212 (1866) ; Hagen,
Pr. Bost. Soc. N.H., XV, p. 376 (1873) ; Provancher, Nat. Canad., IX, p. 42
(1877); Cabot, Mem. M.C.Z.,VIII,1, p. 24, pl. 3, fig.1 (nymph) (1881); Kirby,Syn.
Cat. Od. N.A., p. 88 (1890) [in part] ; Calvert, Trans. Am. Ent. Soc., XX, p. 249
(1893) [in part]; Can. Ent., XXVI, p. 317, 318, (1894); Ent. News, X, p. 42 (1899)
Williamson, Drag. Ind.,pl. 4. fig. 10 (nymph); Howard,Ins. Bk.,pl. 41,fig. 4 (1901);
Needham, Bull. 47,N.Y. St. Mus., p. 470 (nymph),pl. 17, fig. 1 (1901); Needham
and Hart, Bull. Ill. St. Lab., VI, p. 42 (1901); Williamson, Ent. News, XIII, p.
110 (1902); l.c. XIV, p. 227 (1903); Needham, Bull. 68, N.Y. State Mus., pp.
205, 212, 275, pl. 5 (1903); Walker, Ann. Rep. Ent. Soc., XV, p. 69 (1906) ; Can.
Ent., XXXVIII, p. 149 (1906) [in part]; Martin, Cat. Coll. Zool. Selys, XVIII,
p. 48, fig. 44 (1908). ;

Aeshna constricta, Calvert, Od. B.C.A., p. 185 (1905) [in part]; Occ. Pap.
oa N.H., VII, 6, p. 24 (1905) [in part]; Ent. News, XVII, p. 148, pl. VII
1906).

Aeshna ‘‘z’’, Williamson, Ohio Nat., VII, p. 145, 146 (1907); Walker, Ott.
Nat., XXII, p. 54 (1908).

Aeshna umbrosa, Walker, Can. Ent., XL, pp. 380, 390, 450 (1908); Mutt-
kowski, Bull. Wis. N.H. Soc., VIII, p. 57 (1910); Cat..Od. N.A., p. 114 (1910). >"

Aeschna umbrosa, Muttkowski, Bull. Wis. N.H. Soc., VI, p. 167 (1908) ;.
Smith, Ins. N. J., p. 78 (1909).

This widespread species varies considerably in the differ-
‘ent parts of its geographical range. Two principal varieties
may be distinguished, the one occurring in the eastern half of
the continent, the other from the Rocky Mountains west-
ward. These may be separated as follows:

PD much reduced, greenon segs. 5-10 or on all the seg-
ments; seg. 3 variable, but generally rather long and slender;

abe MM un eee aN lication grog 0) ob ENS fh ih Ae. u. umbrosa.
PD of the usual size, blue except on 10; seg. 3 of ordinary
ese Pea ol Ve CSS cor cae RU a GE Ae. u. occidentalts.

Aeshna umbrosa umbrosa Walker.
(Pl. 14; fig.2; pl. 17, figs. 2, 2a; pl. 20, figs. 3, 3a;-pl. 26, figs. 5, 6.)

166 WALKER: NorTH AMERICAN SPECIES OF AESHNA

Male—Occiput lemon-yellow, two-fifths to one-half as
long as the line of contact of the eyes. Frontal vesicle
greenish yellow, this colour extending laterally to the lateral
ocelli. Eyes above rather dark olivaceous green with a
transversely elongate black spot, partly surrounded by a
diffuse violet margin; paler below; edged behind with bright
yellow or greenish yellow. Preocular band, at base of anten-
nz, rarely more than twice as broad as one of the lateral
ocelli; reduced on the sides of the frons to a narrow line,
which disappears at the nasus. T-spot of moderate size,
2-2.33 mm., stem very variable, frequently a mere line, .25
mm. broad, sometimes as narrow as this in front but expand-
ing to .5 or .7 mm. at base; more rarely equal throughout.
Frons and nasus varying from pale green to comparatively
dark olivaceous; pale yellowish laterally and on dorsal surface
of frons. Fronto-nasal suture often with a very narrow brown
line. Rhinarium generally darker than the other pale por-
tions of the face, more or less plumbeous. Labrum pale to
moderately dark dull greenish, margined above and below
by a very narrow black line. Labium drab or yellowish, more
or less obscured peripherally by reddish brown. Rear of
head black centrally and above, pale yellowish green (brown-
ish in dried specimens) laterally.

Thorax rather slender, dark rich brown with a reddish
or coppery tinge. Dorsal bands yellowish green, 3.5-4 mm.
long, separated by about 4 mm. atthe pointed lower ends,
converging upwards to the antealar sinus, their breadth in-
creasing to about I mm. at the middle and 1.5 to 2 mm. at |
upper end, where they are separated by I-1.5mm.; outer
margin more or less strongly curved outwards at the upper
ends. A transverse green spot in each antealar sinus, and
generally also a darker green humeral streak. Lateral bands
bright yellow or greenish yellow, often blue at upper end,
rather narrow, straight and surrounded by a diffuse black
margin. First band rounded at lower end, where it is 1.25-
1.5 mm. broad, narrowing to .6 or .8 mm. at the middle and
generally giving off caudad at upper end a small triangular
offshoot. Second band .75-1.25 mm. broad, equal or slightly
constricted by a slight excavation of the hind margin; upper

WALKER: NorTH AMERICAN SPECIES OF AESHNA_ 167

end often with a narrow cross-bar. There are also frequently
one or two small greenish spots at the metastigma and another
farther dorsad. Spots at bases of wings above yellowish
green, interalar spots pale blue and greenish blue.

Abdomen about five times as long as thorax, strongly
constricted at 3, which is generally longer than in most species,
elsewhere fairly broad though not deep. Seg. 1 without a
distinct ventral tubercle. Auricles with 4 (rarely 3) teeth.
Tergal areas on each side of penis moderately elevated with
numerous spinules, their margins somewhat approximated but
less so than in Ae. palmata. Spines of anterior lamina well-
developed, projecting caudad a little beyond the hamular
processes, their apices curved somewhat more abruptly dorsad
than in palmata and constricta. Hamular processes smaller
than in these two species, their antero-lateral surfaces concave
and separated from the rounded inferior surface by an in-
distinct arcuate ridge; apical margins truncate; inner margins
approximate in front, divergent behind, where they are in
unbroken continuity with those of the hamular folds. Hamu-
lar folds lying directly dorsal to the processes and projecting
scarcely any farther caudad, similar to those of Ae. palmata.
Seg. 3 nearly or quite one-third longer than 1+2, or 4.
Lateral carinae of 7 and 8 strongly approximated in the basal
third, being separated there by a space about one-half as
broad as at apex and appearing strongly sinuate when viewed
ventrad. Dorsum of 10 smooth, without a tooth.

Superior appendages as long as 9+10, dull brown,
shaped as in Ae. palmata, the only differences being the paler
colour and slightly shorter preapical spine. Inferior appen-
dage pale horn colour, darkened at apex and along lateral
margins, otherwise as in palmaia.

Colour-pattern of abdomen—Seg. 1 fuscous; dorsal spot
blue, narrow; lateral spot absent.

Seg. 2 fuscous, spots reduced. AD obsolete ; AML yel-
low or greenish yellow, rhomboid, about 1.2 mm. in diameter,
the upper posterior angle confluent above with MD. PLand
PD separate, the former greenish yellow, the latter blue,
forming a band half as broad as the posterior part of the
segment, with a concave front margin, from which a narrow

168 WALKER: NorTH AMERICAN SPECIES OF AESHNA

green streak passes forward in the middle line, sometimes ex-
panding to form a large diffuse green spot. -

Segs. 3-10 dark brown with a greenish tinge, most of ch
spots green or greenish blue, and so reduced as to give the
whole insect as unusually dark appearance. AD a minute
elongate greenish spot connected on 4-7 with a narrow pale
green basal ring. MD on 3-7, well-developed triangular
greenish spots, separate from ML, except sometimes on 3,
often represented on 8 by a pair of minute dots. PD much
reduced but variable, .8-1.25 mm. long on 3, decreasing and
narrowly separated mesially to 7, where they measure .4-1
mm., somewhat larger and more widely separated on 8 and 9,
diffuse and of variable size on 10. Frequently they are grass-
green or yellowish green throughout, but often the first one
or two pairs are pale blue, the remainder green. AL on 3-6
or -7; on3 reduced to a triangular bluish green or blue spot,
occupying the antero-ventral angle of the segment; on the
others a small rounded pale greenish spot, unconnected with
the basal.ring and becoming a mere dot on 6 or 7.. ML on
3- or 4-8, pale green or greenish blue, small and subquadrate on
3 and 4, then rapidly enlarging to 7, those on 7 and 8 being
particularly large and more rounded than the others. PL on
3-6 or -7, small pale greenish elongate spots, becoming minute
caudad, widely separate from PD, except sometimes on 3,
where they may be narrowly confluent.

There is also a pair of large pale bluish basal spots on the
ventral surface of 4-6.

Wings with a faint ieeterntiel tinge, distinctly shorter
than the abdomen, costal veins. brownish olivaceous; pter-
ostigma dark smoky brown, paler beneath; membranule
whitish in proximal half, dark brown distally; anal triangle
with 3 cells; generally 2 cells between A2 and A3 at their
origin; Rs in fore wing forking opposite the second to the fourth
postnodal cells before the stigma, in hind wing opposite the
second to fourth cell; 3 or 4 rows of cells between the forks at the
level of the distal end of the stigma and 4: to 7 cells between
the forks at the wing-margin. Four to 5 rowsof cells between Rs

WALKER: NorRTH AMERICAN SPECIES OF AESHNA_ 169

and Rspl, where most widely separated. Mta arising before the
middle of the stigma (occasionally at the middle).

Antenodals 2°, postnodals 12 Gur 2 Spe aes ay
II -16 13-20 4-6 2-3 (rarely 1)

Female—Abdomen a little less than five times as long as
thorax; dorsum of 9 distinctly shorter than that of 8 and about
half as long as that of 7.

Genital valves extending slightly beyond the posterior
margin of 9, in profile moderately arcuate, the apices very
slightly elevated. Lateral carinae percurrent, prominent, in
profile nearly straight, in ventral view arcuate, the interven-
ing space lanceolate, broadest at two-fifths of the length, the
breadth here being equal to about two-fifths the length,
tapering distad to the narrow rounded apices. Ventral sur-
face of valves declivent and flat in the basal fifth, sulcate
distally, becoming deeply so at the apices. Styli .75-1 mm.
long, distinctly shorter than dorsum of 10. Free margin of
basal plate of ovipositor slightly arcuate. Lateral plates
distinct.

Appendages thin and delicate, generally broken off except
in tenerals, about as long as 8+9, tapering much more grad-
ually towards the base than the apex; greatest breadth be-
yond the middle and equal to about one-fifth the length;
curve of inner margin somewhat stronger than that of the
outer margin; apices rounded, sometimes terminating in a
more or less obscure point.

The colour-pattern corresponds closely to that of the
male and there is apparently no tendency to dimorphism.
The face tends to be somewhat darker in the male; the dorsal
thoracic bands are of a duller green and often more or less
obscured with brown, being frequently invisible in dried
specimens. They are also narrower and generally divided
into an upper and lower spot; the former small, the latter
more elongate. The lateral bands are often obscured at the
extreme upper ends.

Abdomen somewhat paler brown than in the male, the
dorsal spot on 1 and PD on all the segments green, the latter
still more reduced than in the male, except on the apical three

I2

170 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

or four segments. AL and ML pale greenish or nearly white,
similar in size to those of male. PL larger, often confluent with
PD on 2-3 or -4.

Wings often quite deeply flavescent, this colour extending
to the apices; pterostigma pale brown.

Measurements (excl. Anticosti specimens)—Thor. ¢@ 9.3-
II, 9 8.5-9.75; abd. o& 48.5-54.5, 2 44-50.75; seg. 3 co! 8.75-
10.33, 9 6.5-8; seg. 9 9 2.33-2.75; apps. & 5.33-6, 2 5.4-6.9;
gen. Vv. 2.66-3.25; h. w. oc 42-46.5, 9 43.5-46.7 ; width h.w.
J'12.5-14, 9 13.6-14; pter. oc 2.66-3, 2 3.3-3.66.

Measurements of Anticosti specimens—Thor. & 9-9.25,

¢ 9; abd. o& 44-48, 2 44; seg. 3 o'7.8-8.4, 9 6.5; seg. 9
92.5; apps. o& 5-5.2; gen. v. 2.4; h.w. co" 40-41.5, 2 41;
width h.w. of 11.5-12, 2 12; pter. of 2.7-3, 2 3.2.

Types—o' 2, United States National Museum—De
Grassi Point, Ontario.

Nymph (pl. 4, figs 1-3;pl.7, fig.2; pl. 9, fig.2; pl. 11, fig. 2)—
Eyes very prominent, the antero-posterior and transverse
diameters about equal; lateral margins of head nearly straight,
about half longer than the interocular space, or equal to two-
fifths of the length of the hind margin; postero-lateral angles of
head prominent but well rounded; hind margin slightly
emarginate. Mentum of labium strongly narrowed in the
proximal half, but considerably expanded distally, the breadth
at base being much less than half that at apex, which is equal
toabout three-fourths of the length; sidesstraightand slightly
divergent from the base to slightly beyond the middle, thence
strongly arcuate; middle lobe broadly obtusangulate; lateral
lobes broad, equal, squarely truncate; outer apical angle
scarcely rounded, inner angle terminating in a small tooth.

Supracoxal processes well developed, conical, the poster-
ior larger and somewhat stouter than the anterior, and
usually slightly recurved, the intervening space generally a
little greater than a right angle.

Abdomen comparatively slender, broadest about seg. 6
or 7; lateral spines present on 6-9, those on 6 extending one-
third to one-half of the distance to the hind margin of the
segment; on 7 to the margin or nearly so; on 8 to the margin or
slightly beyond; on g to the middle of 10. Lateral append-

WALKER: NorTH AMERICAN SPECIES OF AESHNA I7I

ages about three-fifths as long as the inferior pair, very slender
in the female but somewhat stouter in the male than in jun-
cea and the clepsydra group, and tapering somewhat suddenly
in the distal third or fourth, the apices very slender and fine-
pointed. Basal part of superior appendage of male about one-
fourth longer than its basal breadth and one-half shorter
than the lateral appendages; sides somewhat concave, apex
bluntly rounded. Genitalia of female extending barely to the
posterior margin of 9; genital valves rather steeply declivent,
about three times as long as broad.

Colour dull dark brown, more or less distinctly spot-
ted with pale yellowish, the degree to which the pale
markings are obscured varying greatly. Head with a pale
lateral marginal band and a submarginal pale spot just behind
each eye. The marginal bands are continued caudad across
the lateral portions of the pronotum to the meso- and meta-
thorax, where they become much broader, but diffuse and often
obscure, covering the whole thoracic dorsum, or sometimes
broken into spots; sides of thorax dark brown. Legs dark
brown, femora and tibiae each with three pale rings, the form-
er having sub-basal, median and preapical rings. Tarsal
joints pale, darkened distally. Abdomen rather dark brown,
varied with pale yellowish and brownish, the pale markings
often almost entirely obscured. A broad medio-dorsal pale
band with slightly undulate margins extends the whole
length. Anteriorly it is well defined and about 4-5 mm.
broad but posteriorly it becomes gradually narrower and less
distinct. Fhis band is more or less darkened mesially by a
series of pairs of dark basal blotches, which run together
on the more proximal segments and sometimes throughout
the whole length of the abdomen to form an irregular median
double dark band. Lateral scars pale, distinctly outlined
with dark brown; in front of each, on the middle segments, is
a basal transverse series of alternate pale and dark spots,
about three of each, the outermost of the former occupying
the antero-lateral angle of the segment. Spines pale, tipped
with black. Lateral and dorso-lateral punctae nearly black;
dorsal punctae no darker than the ground colour.

172 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

Measuremenis—Length of body 38.5-43.5; mentum of
lakium 6-7; h.f. 6.3-7.5; h.w. 8-11; inf. apps. 4-4.5; gen. 9
2.3-2.6 width of head 7.4-8.8; width of abdomen 7-7.5.

Materi:! deiermined—295 3, 83 9. Nova Scotia: (M.C.Z., 1 o);
Pictou, Aug. 22, 1889 (W. Sheraton, Ac. N.S. Phil., 1c’). QuEBEc: Anticosti,
Sep . 12, 1910 (E. V. Cowdry, 10 o& 1 9); Quebec (Provancher, M.C.Z., 2 @ );
Hull. Sep. 14, 1907 (Letourneaux, Fyles, 2 &). Ontario: Grimsby, (J.
Pettit, M.C.Z., 1 o&) ; Toronto, Aug. 21, Sept. 11, 1907, Oct. 10, 1908 (Walker,
6 & 3 ¢, incl. 4 :eared) ; East of Scarboro, emerged July 22, 1908 (Walker,
1 co’); De Grassi Pt., Lake Simcoe, July 2, Sept. 6, 1904 (Walker, 8 o& 7 9):
Go Home Bay, Georgian Bay, Ont., Aug-Sept. 4, 1910 (Huntsman, T. R. Hanley,
30° 2 9); Ottawa, Aug. 7 1907 (Fletcher, t 9); Rostrevor, Muskoka,
Sept. 15, 1907 (Gibson, 1 o’) Port Sidney, Muskoka (Brodie, 1 &); North River
and Little Jo Creek, Algonquin Park, Aug. 29, 30, 1902 (Walker, 1 @ 1 9);
Temagami District, Sept. 1-13 1908 (Walker,13 &@ 5 G, pair in cop., Sept.
13); Heyden, Algoma, July 30-Aug. 4, 1906 (Williamson, 17 o 2 @) ; Search-
mont, Algoma, Aug. 6, 1906, Aug. 23, 1907 (Williamson 3 o ); Root River
Algoma, Aug. 4-7, 1907 (Donaldson, 3 o; Sault Ste. Marie, Sept. 6-Oct. 5,
1907 (Donaldson, 27 & 5 2); Nipigon, Aug. 28-30, 1907, Sept. 5, 1910
(Walker, 3 o&\ 1 9); Fort William, Aug. 26, 27, 1907 (Walker, 12 o);
Rainy River District (H. Cook, 1907, 2 o&). MANITOBA: Winnipeg Beach,
Sept. 6, 1909 (J. B. Wallis, 1 co); Hilton and Treesbank, July 28, 1910 (J. B.
Wallis, 2 co). Marne: Manchester, July 6-Oct. 6, 1888-1910 (Miss Wads-
worth, 33 o& 8 @., in cop., July 6 and 7, Sept. 27, 1906); Augusta, Sept. 16,
1910, (Miss Wadsworth, 1 9); Norway, (S. J. Smith, M.C.Z., 2 @&,2 9);
Gorham (M.C.Z.,1 o) ; Millinocket, Aug. 9-Sept. 7, 1903 (F. L. Harvey, coll.
Williamson, 6 o 4 @ ); Russell Stream, N.E. Corry, Aug. 28, 1899 (F. L.
Harvey, 5 0”); Bradley, Aug. 2, 1899 (Harvey, 1 o); Greenfield, Sept. 8, 1897
(Harvey, U.S.N.M., 1 o”); Six Ponds,Pisc. Co., Sept. 14, 1898,(Harvey, U.S.N.M.,
3 o'); New HAmpsHirE: Franconia (Mrs. Slosson, U.S.N.M., 3 o&);
White Mountains, Aug. 15, 1896 (F. H. Sprague, M.C.Z., 2 co); id., Aug. 7
(M.C.Z., 1 o& 2 9); Hermit Lake, White Mountains, Aug. 29, 1862 (Scudder,
M.C.Z., 10”) Centre Harbour, July (M.C.Z., 19). Massacnusetts: (M.C.Z., 2
o'2 2); Boston, Sept. 6 (B.S.N.H., 1 o&*) ; Cambridge, Sept. (M.C.Z.,2 9);
Amherst, Aug. 5, 1902 (Needham, 1 o’); Wilbrahim (J. O. Martin, coll.
Needham, 1 o'); Auburndale, Sept. 11 (B.S.N.H., 1 co). Gor Head, Martha’s
Vineyard, Aug. 25, 1904 (B.S.N.H., 1 o').. NEw York: Caro, Aug. 19, 1897
(Calvert, 1 9); Catskill Mts., Aug. 28 (E. M. Aaron, Ac. Nat. Sc. Phil., 1 &);
Lake St. Regis, Franklin Co., Sept. 22, 1890 (Ac. Nat. Sc., 1 9); Clyde, Aug. 30,
1902 (Nelson R. Wood, U.S.N.M., 1 o*); Oneida, Aug. 1, 1902 (Mrs. S. A. Maxon,
U.S.N.M., 1 o*): Ellenville, Aug. 20, 1907 (Needham, 1 co’); West Point, Sept.
1874 (Uhler, M.C.Z., 1 &); Dobb’s Ferry, Aug. 24, 1888 (Calvert, 1 @); New
York (Riley, U.S.N.M., 1 o& 19). PENNSYLVANIA: Delaware Co., Sept. 5,
1889, Sept. 10, 24, 1891 (Calvert, 3 o& I 9): id., Oct. 1, 1893 (Calvert, pair 1m
cop.); Alleghany Co., Oct. 9, 1898, Oct. 6, 1901, Aug. 6-Oct. 13, 1903, Sept. 25,
1904, Sept. 20, Oct. 6, 1908 (J. L. Graf, 31 ot 1 9); id., June 18, 1899 (Graf., 1 9
teneral); Morton, Oct. 3, 1889 (Ac. N.S. Phil. 1 9); Ohio Pyle, Oct. 1, 1905
(Williamson, 2 o); Sunbury (D.C. Heim, coll. Calvert, 1 o&); Winthem,
(Hagen, M.C.Z. 18); District oF CoLumBiA: Takoma, Sept. 16, 1899 (J. A,
Nilliams, U.S.N.M.,1 9). MaryvLanp: (Uhler,1858, M.C.Z., 1 co") ; Baltimore
(M.C.Z., 1 oc). Vircinta: Mountain Lake, Aug. 25, 1899 (W. R. Maxon and
C. C. Pollard, U.S.N.M., 3 co’ 19); Rosslyn (N. Banks, U.S.N.M., 1 o@).
TENNESSEE: (L. Thomas, Ac. N.S. Phil., 1 9); Roan Station, Sept. 3, 1901, (Ac.
N.S. Phil.,1 9). Onto: (U.S.N.M., 1 o*): Clarkston, Oct. 13, 1901 (Williamson,
2 o'); Medina, Aug. 20, 1897 (J. S. Hine, U.S.N.M., 1 co). INDIANA: Ft.
Wayne, Sept. 18, 1906, Sept. 15, 1907, Aug. 23, Sept. 13, 1908 (Williamson, 24
o 2 ¢); Bluffton, Sept. 18, 1904; Aug. 25, Sept. 13, 1905, Aug. 26, 1906

WaLKER: NorTH AMERICAN SPECIES OF AESHNA 173

Sept. 15 1907 (Williamson, 4 @ 2 9 ); id., Aug. 30, 1908, (1 @ teneral);
Whitby Co., Sept. 2, 1897 (Williamson, 1c¢’ 19). MICHIGAN: Oden, Aug. 8- 15,
1906 (Williamson, 2 co’ 2 9); Minnehaha Falls, Emmet Co., Aug. 23, 1907
(Williamson, I co’); Wisconsin: Milwaukee Co., Aug. 24, 1902, Aug. I, 1907
(C. G. Brown, Muttkowski, I o& 1 @). NEBRASKA: Loretto, Sept. 1, 1906
(coll. W. J. Fraser, 1 9). Also 22 o’9 9 without data.

Nymphs—Nipigon, Ont., Aug. 7, 1910, 1¢’ exuvia; Go Home Bay, Aug. 8,
1908, 1 spec. (St. B); Aug. 7, 1908, 1 9 exuvia from reared imago ; also exuviae
16°19. De Grassi Point, Ont., July 15-20, 1910, exuviae 2o’ 292. Toronto,
Ont., May, 1908, 167 32 (St. H—St. F also common) ; July 21-Sept. 1, 1908,
exuviae from reared nymphs 2072 9 ; also exuviae 5c’ 59. Ithaca, N.Y. (cold
brook) 1 o@ (St. F) 2 9 (St. F) (Needham). Also the following without data
(Ontario and New York State)—30 29 (St. F), 4c’ (St. E), 367 39 (St. D),
3 spec. (St. C.)

Distribution—Carolinian to Canadian Zones, probably
ranging also into the Hudsonian Zone, as the record of Ae.
constricta from Labrador (Hagen ’61) doubtless belongs to
this species. The race uwmbrosa probably grades into occt-
dentalis in the west. This is the commonest species of
Aeshna in North American collections.

Geographical variations—This species is subject to the
usual variations in form depending upon climatic conditions
(vide p. 26). It varies also slightly in size in different parts
of its range. The largest average size is exhibited by Ohio
and Indiana specimens, the smallest by those from Anti-
costi. Northern examples are most like the race occidentalis
in coloration, the face being generally pale, the lateral thoracic
bands of the male bluishat their upper ends, and the PD spots
less reduced than in the more southern specimens and blue on
the basal three or four segments in the male. In specimens
from Lake Simcoe, Ont., and southward there are usually
no traces of blue, except on segs. I and 2, and the PD spots
are greatly reduced.

Habitts—This common dragonfly differs considerably
from our other species of Aeshna in its habits. It does not
frequent open marshes but flies along woodland streams,
ditches, or small pools on the edges of woods. It is also
common in more or less shady glades and paths in woods and
is most often seen in flight late in the afternoon and at dusk.

Adults begin to appear about the end of June but the
period of emergence seems to be irregular; so some individuals
do not come out until nearly the end of August. It is the
last species to remain on the wing in the autumn, sometimes

174 WALKER: NORTH AMERICAN SPECIES OF AESHNA

flying, in the vicinity of Toronto, as late as the second week
in October, and probably considerably later farther south.

The nymphs are common in woodland streams, small
pools and ditches, and sometimes occur in considerable
numbers. They are very like those of the European Ae.
cyanea in both appearance and habits. They appear to be
the only Aeshna nymph, that is at all easily obtainable in the
full-grown state, in eastern North America.

Aeshna umbrosa occidentalis subsp. nov.

(Pl. 26, fig. 5.)

Differs from race uwmbrosa mainly in the blue colour of
nearly all the abdominal spots, and the much larger size of
PD, which is somewhat reduced only on g and 10, being
elsewhere as large as in most other species of Aeshna. The
abdomen is somewhat shorter and stouter at seg. 3 than in
typical specimens of umbrosa, generally resembling that of
Anticosti specimens, though varying considerably with lo-
cality.

Occiput lemon-yellow; frontal vesicle greenish yellow;
T-spot 2-2.3 mm., stem .25-.33 mm. broad in front, sometimes
expanding a little caudad. Face pale green, generally of a
bluish shade; avery fine brown line often present on the fronto-
nasal suture; rhinarium plumbeous; labrum pale green;
labium dull yellow, brownish or drab, the lateral lobes some-
times slightly plumbeous. Rear of head black above, pale
brown in lower half.

Thorax a little stouter as a rule than in race umbrosa,
marked as in northern examples of that race,i.e., colour red-
dish brown, dorsal bands yellowish green, antealar sinus
pale green, lateral bands yellow passing into pale blue at the
upper ends, surrounded by a nearly black margin.

Abdomen brownish black, the spots all blue except MD
(dull greenish), PD on 10 (dull greenish) and sometimes AML
on 2, which may be blue or green.

The spots are similar in form to those of race umbrosa,
except PL and PD, the former being considerably, the latter
much, larger. PD is in fact as large as in constricta and many
specimens of palmata, except on 9 and 10, where it is more or

WALKER: NORTH AMERICAN SPECIES OF AESHNA 175

less reduced. On 1-6 they are separated only by the median
carina, on 7 a little more widely, while on 8 and 9g the inter-
vening space is nearly 1 mm. broad behind, the inner margins
strongly divergent cephalad. On 10 they may be connate
behind, but are always more or less indistinct and apparently
sometimes absent altogether. On the ventral surface of 4,
5 and 6 there is a pair of large basal spots as in race umbrosa,
but they are blue instead of green.

The female differs but very slightly from that of um-
brosa, thelargersizeof PD and PL and the usually paler face and
slightly shorter average length of the abdomen being the
only characters by which they can be separated. The
wings vary from hyaline to distinctly flavescent.

Measurements — Thor. o 9.7-10.5, 2 9-10.25; apps.
m. 5-5.5, 9 5-6; h.w. 9 41-46.5, 2 38-45; width hw. of
12.3-13.5, 9 12.5-14.

Types-o' 9, Bluffton Public Museum—New Bridge, Ore-
gon.

Material determined —51 o& 11 @. BriTISH CoLUMBIA: Greenwood,
Oct. 1, 1906 (W. J. Alexander, call Walker, I o'); Bear Lake, July 21, 1903
(Currie, U.S.N.M., 19 teneral); Wellington, Vancouver Is., Sept. 2, 1903 (H.G
Dyar, U.S.N.M., 1 o); Langford Lake, Victoria, July 20, 1902 (coll. Williamson,
1 o'). WasHINGTON: Colfax (L. O. Howard, U.S.N.M., 1 o); Almota, Aug.
1894 (coll. Calvert, 1 9); Skokomish River, (T. Kincaid, Ac. N.S. Phil., 1 9)
Seattle, July 6, 1892 (Ac. N.S. Phil., 1 co). OrecGon: (Ac. N.S. Phil., 1 );
New Bridge, Sept. 8-18, 1909 (C. H. Kennedy, 29 o& 2 9); Le Grande Val-
ley, Sept. 30, 1909, (Kennedy 2 co’); Pine Valley, Sept. 23 (Kennedy, 13 o);
Carson, Sept. 22, 1902 (Kennedy, 4 9 9). NEVADA: Reno, (Morrison, M.C.Z.,
Io'1I 9). Cavirornia : (Calvert, 1 <’).

Nymph—Dr. Ris has sent me a full-grown female
nymph of Ae. umbrosa from the Yosemite Valley, Cal., dated
July 26, 1907. It does not differ in any way from the nymph
of the race umbrosa, but on account of the locality it doubtless
belongs to occidentalis.

Distribution—British Columbia to California and Ne-
vada, probably grading eastward into the race umbrosa.

Habits—Mr. Kennedy has taken this form in Oregon,
flying along ditches. Its habits are probably not different
from those of the eastern race.

176 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

Aeshna constricta Say.

(Pl. 2, figs. 1-4; pl. 3, figs. 5, 6; pl. 14, fig. 3; pl. 21, fig. 1;
pl. 27, figs. 2-4.)

N.B.—Undoubted references to the species described here as constricta
are marked with an asterisk.

Aeshna constricta, Say, Journ. Acad. Phil., VIII, p. 11 (1839); Calvert, Od.
B.C.A., p. 180, 285 (1905); Occ. Pap. Bost. Soc. N.H., VII, 6, p. 24 (1905) ;
Williamson, Ohio Nat., VII, p. 150 (1907) ; Walker, Ott. Nat., XXII, p. 54
(1908) ;* Wilson, Pr. U.S.N.M., XXXVI, p. 659 (1909) ; Muttkowski, Cat. Od.
N.A., p. 109 (1910).*

Aeschna constricta, Hagen, Syn. Neur. N.A., p. 123 (1861);* Pr. Bost. Soc.
N.H., XV, p. 271 (1873); Rep. U.S. Geol. Surv. Terr., 1872, p. 727 (1873); l.c.,
1873, p. 591 (1874); Pr. Bost. Soc. N.H., XVII, p. 34 (1875); Rep. U.S. Geog.
Geol. Surv., p. 919 (1875); Kirby, Syn. Cat. Neur. Od., p. 88 (1890); Wads-
worth, Ent. News, I, p. 37 (1890); Beutenmuller, Prel. Cat. Od. N.Y., p. 163
(1890); Harvey, Ent. News, II, p. 73 (1891); Banks, Trans. Am. Ent. Soc., XIX,
p. 353 (1892); Calvert, Trans. Am. Ent. Soc., XX, p. 249 (1893); Ent. News,
V, p. 243 (1894); Journ. N.Y. Ent. Soc., III, p. 46 (1895); Kellicott, Journ. Cinc.
Soc. N.H., XVII, p. 272 (1895); L.c., XVIII, p. 114 (1896); Calvert, Journ. N.Y.
Ent. Soc., V, p. 93 (1897); Van Duzee, Ent. News, VIII, p. 89 (1897); William-
son, Rep. State Geol. Ind., pp. 404 (1897); Davis, Journ. N.Y. Ent. Soc., VI,
p. 197 (1898); Kellicott, Od. Ohio, p. 83 (1899);* Calvert, Ins. N. J., Od., p. 71
(1900); Williamson, Drag. Ind., p. 305 (1900); Needham, Bull. 47, N.Y. State
Mus., p. 469 (1901); Elrod, Bull. Univ. Mont., X, p. 150 (1902); Comstock, Ent.
News, XIV, p. 200 (1903); Williamson, Fnt. News, XIV, p. 369 (1903); Brimley,
Ent. News, XVII, p. 91 (1906); Walker, Can. Ent., XXXVIII, p. 149 (1906);
Muttkowski, Bull. Wis. N.H. Soc., VI, p. 96 (1908);* l.c., VI, p. 167 (1908) ;*
Martin, Cat. Coll. Zool. Selys, XVIII, p. 46 (1908); Cockerell, Ent. News, XIX,
p. 457-459 (1908); Martin, Gen. Ins., CXV, p. 11 (1911).

Aeschna constrictor, Burnham, Pr. Manch. Inst. Arts Sci., I, p. 32 (1900).

A species of average size and rather broad abdomen, the
spots of which are of large size and blue in the male.

Male—Occiput lemon-yellow, usually a little larger than
in the two preceding species, but varying in length from two-
fifths to two-thirds that of the line of contact of the eyes.
Frontal vesicle greenish yellow, this colour extending laterad
to the lateral ocelli. Eyes green and grey-green with a violet
area above, bounded behind by a dark brown transverse
stripe; paler below; edged behind with bright yellow. Preoc-
ular band little or no broader at base of antenne than one
of the lateral ocelli, reduced on sides of frons to a narrow
black line which fails to reach the fronto-nasal suture. T-spot
of moderate size, 2-2.6 mm., stem longer than in palmata
and umbrosa, .4-.7 mm. broad in front, expanding to I mm.,
or a little more, behind. Face pale green, the frons often
more yellowish than the nasus and projecting alittle farther
in advance of the eyes than in umbrosa and palmata; a fine,

WALKER: NORTH AMERICAN SPECIES OF AESHNA 177

pale brownish line on the fronto-nasal suture; lateral lobes of
nasus obtusangulate, not flaring. Rhinarium plumbeous or
brownish. Labrum pale green, very narrowly edged above and
below with black. Labium dull greyish to yellowish, more
or less obscured peripherally with reddish brown. Rear of
head black.

Thorax moderately robust, reddish brown, paler than in
most North American species of the genus. Dorsal bands
yellowish yreen, nearly straight., 4 mm. long, separated by
about 4 mm. at their pointed lower ends, from which they
converge upwards to the antealar sinus; their breadth in-
creasing to about I mm. at middle and 1.5 mm. at the upper
end, where they are separated by about I-1.5 mm. A trans-
verse spot occupies each antealar sinus, but there is no humeral
streak or spot. Lateral bands greenish yellow below, blue
above, expanding above the middle, not surrounded by a
black margin, though edged behind with black in the lower
half. First band about 1.3 mm. broad in the lower half, nar-
rowing a little above the middle tosomewhat less than I mm.,
then expanding again and giving off a small posterior offshoot
at the upperend; anterior edge distinctly sinuate about the
middle, posterior edge gently curved or slightly sinuate.
Second band subequal in the lower half, the breadth at middle
about Imm., expanding at upper end by the outward curving
of both margins to3or4mm. __Interalar spots greenish blue.

Legs black, upper surfaces of femora dark reddish brown;
posterior surface of fore femora pale greyish in their proximal
half or four-fifths.

Abdomen about 4.5 times as long as thorax, strongly
constricted at 3; rather broad, the greatest breadth at about the
apex of 5; somewhat depressed. Seg. I without a distinct
ventral tubercle. Auricles typically with 4 teeth (3-5). Tergal
areas on each side of penis moderately elevated, with numerous
spinules, the margins moderately approximated. Spines of
anterior lamina large and strong, projecting caudad as far as
the posterior margins of the hamular processes or a little farth-
er. Hamular processes larger than in palmata and umbrosa,
ventral surfaces somewhat convex, separated from the strong-
ly concave antero-lateral surface by an arcuate ridge which

178 WAaALKER: NorTH AMERICAN SPECIES OF AESHNA

runs from the base to the inner angle of the truncated apices ;
inner margins closely approximated in front, divergent be-
hind, where they are continuous with those of the hamular
folds; the latter similar to those of palmata but better develop-
ed, closely approximate in the middle line, the margins pro-
duced into a rounded lobe just beyond the hamular processes.
Seg. 3 one-sixth to one-fifth longer than 1+2, and two-fifths te
one-half longer than 4. Lateral carinae of 7 nearly straight,
or feebly sinuate, those of 8 generally slightly sinuate proxi-
mally. Dorsum of 10 smooth, without a tooth.

Superior appendages dark brown, very similar to those
of palmata and umbrosa. The slender basal part is slightly
longer and more arcuate, the outer margin in the basal half
appearing in dorsal view very slightly convex, instead of
straight as in the other species; the greatest breadth, which is
at about seven-tenths of the length, is equal to three-elevenths
of the latter; the ventro-apical spine is distinctly longer than
in umbrosa, less sharply pointed than in palmata, the rounded
notch between its base and the apical margin generally
broader and the preapical tubercle somewhat lower than in the
other two species. Inferior appendages as in palmata.

Colour-pattern of abdomen—Seg. 1 reddish brown; dor-
sal spot blue; lateral spot represented by a. very narrow pale
green line along the posterior margin, sometimes also an ill-
defined pale blotch present.

Seg. 2 reddish brown; AD absent. AML pale yellowish
green below and behind, more or less bluish above, 2-2.5 mm.
broad in the middle, extending ventrad along the front mar-
gin as a narrow streak and confluent above with MD. PL
and PD united, forming a broad blue band, often greenish be-
low; the front margin indented laterally and on each side of
the mid-dorsal line, where it is connected with an ill-defined
median blue spot. An additional blue or greenish blue spot
occupies the postero-ventral angles of the segment.

Segs. 3-10 black, the blue spots all well developed. AL
on 3-8; on 3 blue or green, the upper margins horizontal and
separated by about 1.3 mm.; on the other segments blue, the
upper margins straight and somewhat oblique. Basal rings
absent. ML on 3-8, enlarging to 7, front margin not indented,

WALKER: NorTH AMERICAN SPECIES OF AESHNA 179

hind margin rounded except on 3, where it is acutangulate
or narrowly confluent with PL. MD on 3-8,small narrow
pale green triangles, decreasing in size caudad, smaller and
more widely separated than in palmata and umbrosa, separate
from ML, except frequently on 3. PD well developed, pale
blue to azure blue, enlarging caudad, being about 1 mm. long
on 3 and 1.8 or 2 mm.ongQ, connate behind on 10 in about 66
per cent. of cases, well separated on theother segments, the inner
margins straight and subparallel from 5 or6tog9. PL on 3-4,
-5 or -6; on 3a broad truncated offshoot from PD, on the other
segments small isolated spots diminishing caudad. Ventral
surface of abdomen wholly black.

Wings a little shorter than abdomen, costal margin dull
ochraceous; pterostigma dark brown; membranule of hind
wing whitish in the basal half or three-fifths, smoky brown
distally, extending beyond the cross-vein of the anal tri-
angle, which is 3-celled. Rs forking at the level of 1
or 2 postnodal cells before the stigma; with 3 rows of
cells between the forks opposite the distal end of the stigma,
and 3-5 rows of cells between Rs and Rspl, where most widely
separated. M 1a arising under distal half of stigma (opposite
middle in 5 per cent. of rire

an CucrS -6 Spt = 3 (rarely 4)
10-1 I-2

ke Be barely aay ie ne hind wing, con-
siderably inflated at base and constricted at 3, widening con-
siderably on 8 and 9g, the latter segment unusually long,
being as long as 8 and three-fourths as long as 7.

Genitalia much larger than usual in the genus; genital
valves extending well beyond the apex of 9, in profile moder-
ately arcuate, the apices strongly elevated; lateral carinae
percurrent, very prominent, in lateral view nearly straight
and subparallel; the space between them a little broader at
the basal third than elsewhere, the breadth here being one-
fourth of the length, narrowing very slightly caudad to the
rather broad, rectangular, and oftenslightly divergent, apices;
ventral surface of valves sulcate throughout their entire
length. Styli 2 mm. long, and longer than the dorsum of 10.

pee nodals « 5 he postnodals

180 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Appendages unusually large and strong, rarely broken off
in the living insect, a little shorter than 8+9, greatest width
a little before the middle and equal to about two-sevenths of
the length; margins curved somewhat more strongly at base
than apex and the inner slightly more than the outer; apices
acute.

The markings show the usual differences charac-
teristic of the sex. The dorsal thoracic bands are re-
duced in width to somewhat less than I mm. and are but
little, or not at all, expanded at their upper ends. They are
usually as conspicuously coloured as in the male but are
often obscured by brownish. The lateral bands have the
same form as in the male. Of the abdominal markings AL
is connected with narrow basal rings; PD is somewhat smaller
than in the male on all the segments except 9, where it is
considerably enlarged laterally, extending to the tergal mar-
gins; on 8 and g it is more widely separated from its fellow,
and on 10it isobscure or absent. PL is large on 2, reaching
the transverse carina; on 4, as-well as 3, it is generally con-
fluent with PD.

In colour great variation is met with and a marked ten-
dency to dimorphism is present. In extreme heterochro-
matic individuals all the pale markings are bright yellow
without a trace of green, the eyes are yellowish olivaceous
with a bright yellow hind margin, the wings are strongly
flavescent and the whole insect has a decidedly yellowish
cast. Such individuals are by no means rare but are less
frequently met with than the pure homeochromatic type
and the intermediate forms, which exhibit every gradation
in colour between the two extremes.

A common intermediate type of coloration may be il-
lustrated by the following description taken from a fresh
specimen captured at De Grassi Pt., Ontario: Occiput
lemon-yellow, eyes above grey-green with a deep blue trans-
verse stripe, paler and more brownish green beneath; hind
margin lemon-yellow. Frontal vesicle, frons and nasus
ochre-yellow, the latter two with a greenish tinge; rhinarium
pale brown; labrum dull clay-yellow; labium pale drab with
darker brownish margins. Thorax chocolate-brown with a '

WALKER: NorTH AMERICAN SPECIES OF AESHNA I8I1

slight greyish bloom; interalar spots pale yellowish green.
Wings strongly flavescent, stigma rather dark brown. Fem-
ora dark brownish red, the first pair pale yellowish beneath;
tibiae dark reddish brown, rarely black beneath. Tibiae,
spines and tarsi black. Abdomen reddish brown, D on 1
and PD on 2 bluish; AML on 2 bright yellowish green;
MD bright yellow; AL, ML and PD on 3-10 very pale green
with a bluish tinge; PL pale green; basal rings pale yellowish
green.

Measurements—Thor. o& 10.5-11, 9 10-10.5; abd. &
47-5-50.75, 2 44.7-46.5; seg. 3 co’ 8-9.2, 9 7-7.25; seg. 9
9 3.6-4; apps. & 5.75-6.25, 9 6.25-7.25; gen. v. 4-4.5; h.w.
SO 43-47, 2 43-46; width h.w. o& 13.6-14.9, 9 13-5-15.2;
pter. o& 2.75-3.66, 2 3.5-4.

Types—Say's types are lost. Neotype (Hagen '61)
o, Museum of Comparative Zoology—Rhode Island, Ill.

Nymph (pl. 6, fig. 5; pl.9, fig. 3;pl.11, fig. 3)—Eyes less
prominent than those of Ae. umbrosa and palmata, but
slightly larger, appearing somewhat deeper when viewed
from the front. Lateral margins of head straight or slightly
convex; postero-lateral angles of head not prominent, well
rounded off; posterior margin of head feebly emarginate
about the middle. Mentum of labium about half as broad at
base as at apex, the latter breadth being equal to about six-
sevenths of the length; proximal four-sevenths expanding
considerably, the sides straight; distal three-sevenths more
rapidly widening, the sides moderately arcuate. Middle’
lobe broadly obtusangulate; lateral lobes, beyond the base
of the movable hook, tapering to a slender, slightly incurved
point.

Supracoxal processes of moderate size, subequal, bluntly
conical ; intervening space subrectanguate. ©

Abdomen considerably stouter than in Ae. umbrosa
and somewhat shorter; broadest at 6. Lateral spines present
on 6-9; on 6 extending one-fourth or less, on 7 one-half to
two-thirds, of the distance to the hind margin of the segment;
on 8 two-thirds of the distance, or quite to the margin; on
g as far as the proximal two-fifths of 10, Length of venter
of 9 about equal to half of the basal breadth, sides slightly

182 \VALKER: NORTH AMERICAN SPECIES OF AESHNA

arcuate, strongly so at base. Lateral appendages two-thirds
as long as the inferior pair; in the male fairly stout and taper-
ing abruptly to the slender, sharp-pointed. apices; in the fe-
male slender, more gradually tapering. Basal part of su-
perior appendages of male scarcely two-thirds as long as the
lateral appendages, about one-fourth longer than broad, sides
slightly concave, apex rounded. Genitalia of female re-
markably large, the regularly arcuate ovipositor reaching
back nearly or quite to the posterior margin of 10, or even
a little beyond it; genital valves steeply declivent.

Colour light brown; head, thorax and legs concolorous.
Abdomen with a fine, pale mid-dorsal line between two nar- °
rower dark streaks. Punctae all marked with dark brown,
the dorsal and lateral rather heavily. The lateral scars are
not distinctly outlined and there are no other dark markings,
except faint indications of two lateral streaks on each side.

Measurements—Length of body 36-38; mentum of labi-
um 5.7-6.5; h.f. 6-6.25; h.w. 8.75-9; inf. apps. 3.8-4; gen. 9
4.3-4.75; width of head 7.8-8; width of abdomen 7.3-7.5.

I refer these nymphs without hesitation to Ae. con-
stricta on account of the very large genitalia of the female,
which correspond in size with those of the imago of this
species. The only other species in which the genitalia are
comparable in size is Ae. tuberculifera, which does not occur
at Whitefish Creek nor any other part of Lake Simcoe, so
far as 1 am aware. Constricta on the other hand is abundant
and is the common species at the creek. The only other
species I have ever seen there are Ae. umbrosa and occasional
individuals of canadensis, both of which have been reared.

Material determined — 108 o&' 56 9. Nova Scotia: (British Museum
3 0° 3 9). Marne: Manchester, Aug. 1 - Sept. 10, 1889-1910, Sept. 1909
(Miss Wadsworth, 9 o& 7 @). MASSACHUSETTS: Hampden, Aug. 7, I902
Needham, 1 9). QuEBEC: Chateauguay (J.G. Jack, M.C.Z., 2 9); South
branch of Nation River, Dundas Co. (A. R. Cooper, 3 co’). NEw York: Hotel
Champlain, Watkin’s Glen, Aug. 21, 1890 (Calvert, 1 co); Clyde, 1902 (Nelson
R. Wood, U.S.N.M., 1c). ONTARIO: Ottawa, July 26, 1900 (Fletcher, 1 9); id.
(Harrington, 1 9); Toronto, Aug. 26, 1908, Sept. 12-27, 1906, 1908 (Huntsman,
Walker, 8 o& 2 9); De Grassi Pt., Lake Simcoe, Aug. 7- Sept. 8, 1906-1910
(Walker, 16 o& 10 9). Onto: Medina, Aug. 20, 1897 (J. S. Hine, U.S.N.M.,
1 9). Ittrnors: Lake Forest, Aug. 2, 1904, Sept. 2, 1902 (Needham, 2 9);
Rhode Island, 1860 (Walsh, M.C.Z., 1c 19). INDIANA: Elkhart (Ac. N.S.
Phil., 1 9); id.,Oct. 12, 1899 (R. J. Weith, coll. Williamson, 1 9); Bluffton,
Sept. 5-Oct. 2, 1904, Aug. 20, Sept. 24, 1905, Aug. 11-Sept. 22, 1907 (Williamson,
36 o& 10 9); Fort Wayne, Sept. 7, 1902, Sept. 16-18, 1906, Aug. 25-Sept.

WaLKER: NorTH AMERICAN SPECIES OF AESHNA 183

22, 1907, Aug. 30-Sept. 13, 1908 (Williamson, 17 o& 3 9); Crooked Lake,
Steuben Co., Sept. 1, 1905 (Williamson, 1 9). Wisconsin: Lone Rock, Aug. 18,
1906 (J. D. Hood, coll. Williamson, 1 o"); Milwaukee Co., July 17-Aug. 24, Sept.
6, 1902, Sept. 5, 1904, July 10-27, 1905, July 16, 1907, Sept. 9, 1908 (Muttkow-
ski, C. E. Brown, Mrs. P. Laur, Milw. Pub. Mus., 6 o& 3 2); Coney Island, Mil-
waukee River, July 17, 1902 (Val. Fernke, Milw. Pub. Mus., t 9). Sout
DAKOTA: Volga, Aug. 29, 1889 (P.C. Trueman, Ac. N.S. Phil. 1 9). Nort
Daxota, Fargo, July 30, 1901 (R. C. Osburn, 1 oc’ 1 9). MANITOBA: West-
Seis Aug. 26, 29, 1908 (Wallis, 2 o& 1 9). British CoLtumBia: (Brit. Mus.,
Io’). :
Nymphs—Whitefish Creek, Lake Simcoe, Ont., «9 (St. D), taken about
July 15, 1910, moulted twice; male died after second moult, Aug, 18, 1910 ;
female lived until Dec. 24, 1910. Both acquired fully developed wing-pads at
second moult and the female full-sized genitalia (vide p. 48). Also ¢@ 9 (St. D),
o' 9 (St. E) and exuviae of full-grown nymphs, 2c 49, Aug. 10, 1910. aise

Identity—Say’s type came from Indiana and’ therefore
must have belonged either to the species described here or
to Ae. umbrosa umbrosa. Hagen’s neotype belongs to the
present species, and, apart from this, there is one point in
Say’s description which applies to this species better than to
umbrosa. This is the statement that ‘‘the posterior inter-
rupted bands (PD) might be called rounded or quadrate
spots and are largest and more glaucous on the posterior seg-
ments.’’ (The italics are mine.)

Ae. constricta is more closely related to Ae. palmata than
to Ae. umbrosa, but is very distinct from both, especially
in the female genitalia. The taper-pointed lateral lobes of the
labium is a remarkable and distinctive feature of the nymph.

Distribution—Upper Austral and Transition Zones from
the Atlantic Ocean to Manitoba and the Dakotas, ranging
into the Canadian Zone in Manitoba. The British Columbia
record needs corroboration.

Habits—This species first appears in Ontario during
the second half of July, becoming numerous about the middle
of August and remaining until about the beginning of Octo-
ber, or possibly later.

At Lake Simcoe, where consiricta is common and in some
seasons very abundant, it flies over open marshes, pastures
and fields, especially near woods, though not often occurring
in the small clearings and glades, which are frequented by
its congener Ae. umbrosa. It ismost often seen on warm,
sunny days, and unlike umbrosa it apparently does not fly
after sundown, although in hot still weather I have seen it

184 WALKER: NORTH AMERICAN SPECIES OF AESHNA

late in the afternoon, flying low over grassy spaces (vide p. 34).
When at rest it hangs from the branches of trees or settles
in bushes, or even weeds close to the ground.

The copulating habits of this species and its mode of
oviposition have been described under “‘ General Life History”’
(pp. 38, 44).

Whitefish Creek where the nymphs were found is a
small sluggish stream, which winds through woods and
pastures in its upper part, where it is a mere brook, but near
its mouth traverses an open swamp supporting a dense growth
of reeds, rushes and sedge. The upper, shadier part of the
stream is inhabited by nymphs of Ae. umbrosa, the lower part
by those of Ae. constricta. The territories occupied by the
two species are not sharply separated but on the whole are
quite distinct.

Where constricta is most at home, the stream is quite
clear of reedy plants except at its immediate margin where
there is a tall dense growth of Acorus calamus, Sparganium sp.
and several species of tall sedge. The bottom shelves rapidly
from the bank and there is no shallow reed-grown area, such
as forms the characteristic habitat of most Aeshna nymphs.
Thus the habitat of constricta as represented by Whitefish
Creek is somewhat intermediate in character between those
of umbrosa and the majority of species of the genus.

The nymphs appear to be difficult to secure and many >
hours of dredging resulted in but few individuals, and none
of these were full-grown. The exuviae, too, are not readily
seen. They cling to the reeds, near the base, and are often
hidden from view.

Aeshna californica Calvert.

(Pl. 14, fig. 4; pl 17; figs. 4, 4a; pl. 21, figs. 2, 2a; pl.
28, figs 2s)

Aeschna californica, Hagen, Pr. Bost. Soc. N.H., XVIII, p. 33 (1875) [no desc];
Calvert, Pr. Cal. Ac. (2) IV, p. 504, pl. XV, figs, 19, 20, 23 (1895) ; Needham
and Hart, Bull. Ill. State Lab., VI, p. 41, 45 (1901); Williamson, Ent. News, IV,
p. 3.7 (1903) ; Needham and Anthony, Jr. N.Y. Ent. Soc., XI, p. 121 (1903) ;
Osburn, Ent. News, XVII, pp. 186,190 (1905); Amer. Nat., XL, p. 396 (1906);
Martin, Cat. Coll. Zool. Selys, XVIII, p. 47, 84, fig. 83 (1908); Cockerell, Ent.
News, XIX, p. 455-457 (1908).

Aeshna californica, Calvert, Od. B.C.A., p. 183 (1905) ; Walker, Can. Ent .
XL, p. 378, 386, 450 (1908); Muttkowski, Cat. Od. N.A., p. 109 (1910).

WALKER: NORTH AMERICAN SPECIES OF AESHNA 185

A species of somewhat less than medium size, with a
rather short broad abdomen.

Male—Occiput pale lemon-yellow, rather large, one-
third to two-fifths as long as the line of contact of the eyes;
frontal vesicle with a large pale yellow reniform spot,or two
smaller spots; eyes pure blue, fading below to bluish grey ;
preocular band two to three times as broad at bases of an-
tennz as one of the lateral ocelli, narrowing on the sides of
the frons to about half that breadth, and widening again at
the fronto-nasal suture; T-spot 2.3-3 mm., stem two or
three times as broad behind as in front, the former breadth
varying from 1.25-2, the latter .4-1.4. Face pale dull blue
or greenish, yellowish on each side of the stem of the T-spot.
A black line on the fronto-nasal suture and another at base of
labrum. Lateral lobes of nasus rotundo-angulate, distinctly
flaring. Labium pale blue to pale yellowish or dull oliva-
ceous. Rear of head black.

Thorax moderately dark brown, appearing somewhat
pale in dried specimens on account of the thinness of the
cuticle and the rather long and dense growth of pale brownish
hairs; sutures very dark brown. Dorsal bandsreduced to a
pair of small elongate pale spots on the anterior half of the
thoracic dorsum, often obscure or indistinguishable in dried
specimens, Lateral bands bluish white (often yellowish
white in dried specimens) more or less distinctly edged be-
hind with blackish; first band nearly straight, about I mm.
broad near the rounded lower end, tapering rapidly dorsad to
one-half or less of this breadth; second band nearly equal,
scarcely 1 mm. broad, curved slightly caudad.

Legs black, posterior surface of.fore trochanters and fe-
mora in their basal half pale yellowish,

Abdomen about four times as long as the thorax, eadicn
stout, strongly constricted at 3 but rapidly expanding again,
so that the posterior margin of 3 is but little narrower than
that of 4. Seg. 1 witha prominent spinulose ventral tubercle.
Auricles with 2 or 3 teeth. Spines of anterior lamina well-
developed, strongly curved, extending back to the hind mar-
gin of the hamular processes or a little beyond; the latter with
the inner margins straight and attingent, the apices sharp-

13 ,

186 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

pointed and directed cephalad; hamular folds large, mouse-
ear-shaped, continuous with the hamular processes, closely
approximated mesially. Segs. 5-8 with rudimentary accessory
lateral carinae, best marked on 6 and 7. Median tooth-like
elevation in dorsum of 10 low and rounded as seen in profile.

Superior appendages about twice as long as Io, slender
and distinctly divergent in the proximal fifth, thence conver-
gent, the breadth increasing to or slightly beyond the middle,
thence decreasing very slightly to the bluntly angulate and
closely approximated apices ; inner margin viewed dorsad
rather strongly sinuate, being strongly concave before the
middle, gently convex beyond; outer margin viewed dorsad
distinctly convex at base, gently so or nearly straight beyond;
in lateral view curved regularly upwards. Superior carina
percurrent, rather strongly elevated and arcuate in the apical
third, not denticulate; infero-internal margin gently curved
forming a low inferior carina: a low sub-basal ventral tuber-
cle at about two-sevenths of the length of the appendage.
Inferior appendage slightly less than one-half as long as the
superior pair, and about three-fourths as broad at base as
long, triangular with slightly convex lateral margins and
bluntly rounded apex, which bears a pair of minute dorsal
teeth; in lateral view moderately curved upwards.

Colour-pattern of abdomen—Abdomen brownish black,
varied with castaneous; all the spots caerulean blue, mostly
well developed.

Seg. 1. Dorsal spot absent, lateral spot indicated by a
pale marginal line less than 2 mm. long.

Seg. 2. Blue spots very large; AML bounded below by the
auricle, upper margin very oblique, confluent with MD;
posterior part of the segment above the level of the auricles
entirely blue except for a pair of oblique black dorsal spots
just behind the transverse carina.

Segs. 3-10. Black ground colour on the dorsum of 3-5
in front of the transverse carina and behind MD varied with
castaneous. AL on 3-7 or -8, very large on 3, covering whole
of lateral surface; of moderate size on 4-6, the upper margin
oblique; minute on 7 and a mere dot when present on 8. MLon
3-8, very large on 6 and 7. MD on 3-8, greenish blue,

WaLKER: NORTH AMERICAN SPECIES OF AESHNA 187

subcrescentic on 3-6, a pair of dots on 7-8. PD on 3-10,
well developed, semi-elliptical, each separated from its fellow
by about 5 mm., except on 8 and 9, where they are somewhat
smaller, more elongate and separated by I mm. or more; on ©
10 as large as on 9, not connate.

Wings hyaline; costal veins pale horn-yellow; pterostigma
dark brown; membranule of hind wings extending nearly to
apex of anal triangle, whitish in proximal half, smoky brown
distally, the two areas generally sharply separated; anal tri-
‘angle 3-celled; one cell between A2 and A3 at their origin,
in the hind wings; outer side of triangle of hind wing fully 1.5
times as long as inner side; Rs forking at the level of 1-2 post-
nodal cells before the pterostigma, base of the fork but slightly
asymmetrical, the two branches equally well developed, not
or but little convergent at the wing-margin; 2-3 rows of cells
between the forks except at the margin,where there are usually
4 or 5 cells; 3 (2-4) cells between Rs and Rspl where most
widely separated; Mtiaarising beyond the level of the distal
end of the stigma (sometimes opposite the distal end).

Antenodals pap postnodals fie Cur om , Spt ai
8-10 g-12 4-5 O-I

Female —The usual differences from the male, in form,
are present. Abdomen not quite four times as long as the
thorax and about equal in length to the hind wings.

Genital valves as long as seg. 9, in profile arcuate in the
proximal three-fifths or three-fourths, the apices straight,
approximated; no distinct lateral carinae nor ventral surface;
styli about one-third the length of the ovipositor (.66 mm.);
basal plate small, more or less furrowed on each side, posterior
margin straight.

The most marked difference in the pattern of the abdom-
inal spots is the smaller size and wider separation of the PD
spots, except on seg. 10. On seg. 2 these spots are frequently
separate at the mid-dorsal line as on the succeeding segments
and are partly or wholly marked off from PL. The lateral
spots are larger and are more or less confluent on 2 and 3 and
sometimes on 4.

Both blue and yellow females occur, according to Mr.

188 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Kennedy, who has observed this species in life. He has
sent me a coloured drawing of a specimen in which all the
pale markings are a pure chrome-yellow on a ground colour
which is of a more yellowish brown than in the male. In the
specimen figured, the only alcoholic one I have seen, the
colours were apparently perfectly preserved at the time the
drawing was made, and the ground colour was noticeably
more yellowish than in the alcoholic male. The face was
greenish yellow, labium pale greenish, eyes grass-green with
a yellow outer margin; thoracic bands very pale yellowish
green; abdominal spots pale green, PD more yellowish than
the others. Many of the dried specimens that I have ex-
amined appear to have been coloured like the males.

Measurements— Thor. @@ 10-11; abd. o& 39.5-43.5,
9 37-40; seg. 3 co’ 6.3-7.5, 9 5.5-5.8; width seg. 2 9 5-5.33;
width seg. 3 @ 3.3-3.5; apps. co 4.25-5, 9 4.5-4.9; gen. Vv.
Q 1.75-2; hw. o& 38-49, 2 37-40; width hw. o@ 12-12.6,
9 11.66-12.25; pter. & 3-3.8, 2 3.5-3.8.

Types—o& 9, Coll. Calvert—Mt. Tamalpais, Cal.

Nymph (pl. 7, figs. 3, 3a; pl. 9, fig. 4; pl. 11, fig. 4)—Eyes
larger but less prominent than in the other species here
treated, except multicolor; postocular part of the head short,
the lateral margins slightly arcuate next the eyes, thence
passing by a stronger curve into the almost straight posterior
margin; mentum of labiumrathershort, about half as broad at
base as at apex, the apical breadth a little shorter than the
length, expanding distally throughout the length, especially
in the distal three-sevenths; lateral margin slightly sinuate,
the curve proximally very feebly convex, more strongly so
towards the distal end; median lobe slightly produced, ob-
tusangulate; lateral lobes slightly narrower distad of movable
hook than in multicolor, the outer apical angles somewhat
rounded off, inner angles with a minute tooth.

Supracoxal processes rectangularly divergent, acute, the
anterior of moderate size, the posterior considerably larger,
recurved.

Abdomen moderately stout, broadest at 6 or 7; in the
exuvia, which is deeply convex, at 5. Lateral spines present
on 6-9, those on 6 distant from the hind margin by 1-2 times

WALKER: NORTH AMERICAN SPECIES OF AESHNA 189

their own length, on 7 reaching back as far as the margin or
but little short of it, on 8 slightly beyond the margin, on 9
to about the middle of 10. Venter of 9 nearly two-fifths as
long as broad. Lateral appendages about three-fifths as long
as the inferior pair, slender, tapering gradually to the fine-
pointed apices. Basal part of superior appendage of male
triangular, one-half shorter than the lateral appendages and
about as long as the basal breadth; sides nearly straight, apex
acute. Genitalia of female distant from the posterior mar-
gin of 9 by one-sixth or one-fifth of their length, genital valves
not steeply declivent.

General colour (exuvia) dull brown, pale markings more
or less obscure or inconspicuous. Head with a group of pale
spots between the eyes, including a median and three or four
spots on each side of it; a roundish spot just behind each eye
and a pale marginal band which is continued across the outer
ends of the pronotum. Sides of thorax with a few small
pale spots. Femora dark with three narrow, ill-defined pale
annuli; tibiae generally also with traces of two or three such
annuli. Abdomen, with two dorsal pale bands, which are
broad and distinct on the first four segments but become
gradually narrower and more obscure caudad, usually being
almost or quite indistinguishable on the last four or five seg-
ments. Lateral scars inconspicuously ringed with brown,
punctae marked with brown, but all inconspicuous. Lateral
appendages pale, with basal and median annuli and the apices
dark.

Measurements—Length of body 33-36.5; mentum of
labium 4.5-5; h.w. 6.8-8; inf. apps. 3.75-4.2; width of head,
7-7.3; of abdomen 7 (exuvia 6-7).

Material determined—41t #19 9. British CotumBIA: (Crotch, M.C.-
Z., 1 o&) Victoria, July 14, 17, 1901 (Currie, U.S.N.M., 3 o& 1 9); Departure
Bay, Vancouver Id., June 20,July 4, 5, 1908 (Huntsman, § o' I ?); Diver Lake,
Vance. Id., July 23, 1908 (Huntsman, 1 o’); Peachland (Wallis, 1 9). Wasu-
INGTON: Sunnyside, April 23-28, 1910 (Kennedy, 2 o' I ¢) id, May 8, I910
(Kennedy, 14 o& 4 @); id. June 12, 1910 (Kennedy, 1 <); Ellenburgh,
April 8, 1897 (Needham, 1 9); Seattle, July 6, 1892 (Ac. N.S. Phil., 1 9); Kent,
June, 9, 1905 (H. E. Burke, U.S.N.M., 1 oc’). OREGON: Portland, July 8, 1905
(Currie, U.S.N.M., 1 9); Corvallis, May, 7, 1878, June 30, 1896 (Needham, 1
o'1 9). CaALiFrorNiA: (M.C.Z.,2 9); Fresno, April 7-23, 1900 (E. A. Schwarz,
U.S.N.M., 1 @ 2 9); Utah Lake, June, 19, 1891 (E. A. Schwarz, U.S.N.M., I
o'); Gulf of Georgia (A. Agassiz, M.C.Z., 4 o); Mendocino (A. Agassiz.

190 WALKER: NORTH AMERICAN SPECIES OF AESHNA

M.C.Z., 1 co’); San Mateo (A. Agassiz, M.C.Z., 1 o'); Sonoma Co., Apr. 27-May
9 (R. Osten Sacken, M. C.Z., 1 o"). NEVADA: Reno, 1878 (Morrison, M.C.Z.,
3 co’). Utan: Ogen (Cyrus Thomas, MICZ. 1)

Nymphs—Nigger Pond, Sunnyside, Wash., May 10, 1910; 19 (about to
transform); 19 (St. F); 19 (St. E); exuviae 10d? 149 (Kennedy).

Distribution—Upper and Lower Sonoran. California
and Arizona to southern British Columbia.

Habits—This species is the earliest North American
Aeshna to appear in the adult state. As the dates of capture
een above show, it may appear as early as the second week

April (California, Washington). Mr. Kennedy found it at
seueaes Washington, on May 8, 1910, flying about Nigger
Pond in considerable numbers. He observed several pairs
tn copula and states that, while thus engaged, they are more
retiring than Ae. multicolor and do not indulge in wild nuptial
flights. A few weeks later they had left the pond and were
ranging freely over the countryside. The nymph was
first described from a specimen taken from an irrigation
ditch at Tombstone, Arizona (Needham and _ Hart
‘or). It has also been found by Osburn (’06) to inhabit
brackish water.

Aeshna multicolor Hagen.

(Ph 14, fig.:53 pl. 17, figs. 5;5a; pl. 25, figs. aaae

Aeschna multicolor, Syn. Neur. N.A., p. 121 (1861) ; Rep. U.S. Geol. Surv.
Terr., 1872, p. 727 (1873); 1. €./ 1873, p: 591 (1874); Pr. B.S.N.H., XVIII, 3
(1875); Kirby, Syn. Cat. Od., p. 88, 1890; Banks, Trans. Am. Ent. "Soc., XI
353 (1892) ; Calvert, Ent. News, III, p. 26 (1892) ; Pr: Cal: Ac’ Se:, (2) IV, p.
508, pl. 15, figs. 25, 26 (1895); Tr. Am. Ent. Soc. bode p. 43 (1902); Williamson,
Ent. News, XIV, p. 7 (1903); Currie, Pr. Ent. Soc. Wash., V, p. 303 (1903);
Ke.\ Vil, ps6 (1905); Osburn, Ent. News, XVI, pp. 186, 190 (1905): Baker, In-
vert. Pac., I, p. 87 (1905) ; Martin, Cat. Coll. Zool. Selys, XVIII, p. 48, fig. 45
(1908); Gen. Ins., CXV, p. 12 (1911).

Aeshna multicolor, Calvert, B.C.A, p. 183 (1905); Williamson, Ent. News
IX, p. 265, 301, text fig. (1908); Walker, Can. Ent., XL., p. 279, 386, 450 (1908) ;
Muttkowski, Cat. Od. N.A., p. 113 (1910.)

Aeschna furcifera, Karsch, Ent. Nachr., XVII, p. 310 (1891).

Of average to rather large size, the thorax robust and the
abdomen somewhat short.

Male—Occiput pale bluish, two-fifths to one-half as
long as the line of contact of the eyes; frontal vesicle pale
blue, the blue area separated from the lateral ocelli by a space
narrower than one of the latter. Eyes pure blue, of the same
shade as the abdominal spots, sometimes pale greenish along
the hind margin. The blue colour usually disappears in
dried specimens but may be retained to some extent in well

Waker: NorTH AMERICAN SPECIES OF AESHNA I9QI

preserved examples. Preocular band scarcely broader, at
base of antenne, than the lateral ocelli, continued ventrad on
each side as a narrow line on the lower half of the frons and
nasus. T-spot 2.8-3.2 mm., stem 1.45-1.66 mm. broad be-
hind, .55-.75 mm. in front, sides straight or slightly concave.
Face slightly broader than in Ae. mutata. Frons and nasus
pale blue or grey-blue with a narrow yellowish white submar-
ginal area; fronto-nasal suture ochraceous; lateral lobes of
nasus rounded, their margins but little flaring. Rhinarium
and labrum greenish, the latter narrowly margined above and
below with dark brown. Labium blue, plumbeous or oliva-
ceous in dried specimens. Rear of head black.

Thorax robust, moderately dark brown with a coppery
tinge. Dorsal bands light blue, 4 mm. long, .65-.75 mm.
broad, sometimes slightly broader at the upper ends, where
they are approximated, and often giving off mesad an offshoot
from the lower ends. Lateral bands light blue, sometimes
greenish below, straight, very oblique; first band nearly
equal, about 1 mm. broad, rounded below; second band
generally a little narrower, tapering to the lower end and some-
times expanding slightly towards the upper end. Interalar
spots blue.

Legs black, anterior femora with a pale streak on the
proximal half of the posterior surface.

Abdomen somewhat less than four times as long as the
thorax, moderately stout, considerably constricted before the
middle of 3, then rapidly expanding again to the apex of 4.

Seg. 1 with a prominent ventral tubercle, bearing a few
inconspicuous spines and a tuft of long brownish hairs. Au-
ricles with 2-3 teeth. Tergal margins of seg. 2, bounding
the genital fossa, strongly sinuate, considerably elevated in
the posterior third. Spines of the anterior lamina well-
developed, reaching caudad to the posterior margins of the
hamular processes, moderately curved dorsad, sharply pointed.
Hamular processes rather broad and thick, mesially attin-
gent, antero-lateral surfaces concave, bounded behind by an
arcuate ridge running from the base to the antero-internal
angle, which is somewhat acute. Hamular folds continuous
with the hamular processes, rather long and contracted,

192 WALKER: NORTH AMERICAN SPECIES OF AESHNA

mouse-ear-shaped, closely approximated mesially, the apices
strongly acutangulate, outer margins slightly divergent.
Seg. 3 as long as I+2, one-third to two-fifths longer than
4, and one-seventh to one-sixth longer than 6. Dorsum of Io
with a median and two pairs of sub-median tooth-like eleva-
tions, the former relatively small and rounded in profile view.

Superior appendages about as long as 8+49, or a trifle
longer, slender at base, expanding on the inner margin to
about five-twelfths of the length, where the breadth is about
twice that at base and equal to about one-seventh of the total
length; thence remaining equal as far as the distal fourth or
fifth, whence curving strongly downwards and slightly out-
wards they taper to the long slender pointed apices. In
dorsal view they appear very slender, the outer margin
gently arcuate, the inner very slightly sinuate. Superior
carina subobsolete in the basal half, suddenly elevated in the
distal half into a very prominent angular crest, the apex of
which is a little more than one-third of the length of the ap-
pendage from the apex of the latter. In profile view a low
sub-basal ventral tubercle appears at one-fifth to one-fourth
of the length; the outer margin is curved slightly upwards
except apically, and above it the superior carina is elevated
to a height about equal to the depth of the inner portion of
the appendage directly beneath it, this part being bent down-
wards to form an inferior carina, which, a little beyond the
apex of the superior carina, is produced into a strong spine,
directed ventrad and caudad, its apex recurved. The dis-
tance between the apex of this spine and that of the appendage
is equal to about one-fifth of the length of the latter. In-
ferior appendage elongate-triangular, feebly acuminate, slight-
ly more than one-half as long as the superior pair, the breadth
at base three-eighths of the length, upper surface with a dis-
tinct median carina, apex rounded with a pair of small re-
curved dorsad teeth; in profile view moderately curved, sub-
equal and tapering but little apically.

Colour-pattern of abdomen — Seg. 1 fuscous ; dorsal
and lateral spots blue, or the latter sometimes narrowed
to a mere marginal line.

Seg. 2 fuscous; AML blue, extending 1-2 mm. above the

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 193

auricle and a variable distance below it, its upper margin
very oblique and confluent with MD. PL+PD covering
the posterior division of the segment except part of the ven-
tral surface and a narrow brown space on each side behind
the transverse carina, which is continued mesad into adark
brown oblique streak.

Segs. 3-10. MD on all the segments and sometimes PD
on 10 greenish, the other spots pale pure blue. AL on 3-7 or
-8, the first three orfour pairs connected by a narrow basal
ring; large on 3, the upper margins straight and separated by
a brown band of about I mm. breadth; of moderate size, de-
creasing caudad on 4-7, the upper margins more or less oblique.
ML on 3-8, of moderate to large size, irregularly quadrate or
roundish, separated from MD, anterior edges not indented.
MD on 3-8, rudimentary on 8, elsewhere relatively large,
triangular. PD on 3-10, irregularly semi-elliptical, varying
in length from about 1.3 to nearly 2 mm., separated mesially
by spaces of varying width but always much wider on 8-10
than elsewhere, especially on 9; on 10 they are paler than on
the other segments and are larger and better defined than in
mutata. PL on 3-5 or -6; short, broad, curved offshoots of
PD, with which they are broadly confluent.

Wings hyaline; costal veins pale horn-yellow; pterostigma
dark brown above, ochraceous beneath ; membranule of hind
wing extending nearly to the anal angle, whitish in the basal
third or fourth, dark smoky brown distally. Anal triangle 3-
celled; A2 in hind wing arising opposite or distal to the last
cubito-anal cross-vein before the sub-triangle; outer side of
anal loop longer than inner side of triangle. Rs forking op-
posite 1-2 (fore wings) 2-3 (hind wings) postnodal cells be-
fore the stigma, the fork almost symmetrical at base, and both
branches equally well developed, generally not converging
towards the wing-margin. Three rows of cells between the forks
opposite the distal end of the stigma. Supplements strongly
curved; 3 or 4 rows of cells between Rspl and Rs between the
points of widest divergence. M 1a arising beyond the stigma.
Usually a single row of cells between M3 and Mq in the hind
wing, for a short distance just before the marginal cells.

194 WALKER: NORTH AMERICAN SPECIES OF AESHNA

Antenodals eds postnodals bea CuCr 5-6 , Spt es)
g-I2 8-13 3-6 I-2

Female—The abdomen is about the same in length as in
the male and is not usually very stout. Abdominal seg. 3
is shorter, seg. 4 longer, than in the male, there being but
little difference in length between these segments.

Genital valves as long as seg. 9, in profile slightly arcu-
ate with the apices a little elevated, lateral carinae distinct
only apically, elsewhere broadly rounded; ventral surfaces
somewhat declivent, not distinctly sulcate. Basal plate
with the hind margin straight; lateral plates distinct. Styli
scarcely half as long as the dorsum of 10 (.75 mm.). Ap-
pendages nearly as long to a little longer than 8+9, tapering
nearly equally at apex and base, curve of the inner margin a
little stronger than that of the outer; greatest breadthabout
the middle, equal to about one-fifth of the length; apices acute
or narrowly rounded.

The dorsal thoracic bands are narrower than in the male
and are either divided into a superior and an inferior spot or
represented by the latter alone. They are frequently in-
distinguishable in dried specimens. Lateral spots of seg. 2
forming a continuous broad band; brown area behind the
median suture broader than in the male, owing to the smaller
size of the PD spots which are also somewhat reduced on the
other segments. They are not confluent on any of the seg-
ments. PL on 3-5 or -6, larger than in the male, often separ-
ated from PD posteriorly, connected with ML on 3. I have
seen no purely homeechromatic specimens though they prob-
ably occur. Intermediate and markedly heterochromatic
examples are present in the material studied. A fairly well
preserved specimen from Portland, Ore.,and two from Sunny-
side, Wash., belong to the latter typeof coloration, while two
from Departure Bay, Vancouver Is., are of the intermediate
type. Mr. Kennedy makes the following notes on one of
the Sunnyside specimens: ‘‘Eyes brown, edged behind with
grey-blue, markings of thorax and abdomen yellowish with
greenish hue.’’ In these specimens the wings are also some-
what flavescent at base and along the costal edge. In the
British Columbia females the head is coloured as in the male,

WALKER: NORTH AMERICAN SPECIES OF AESHNA_ 195

dorsal thoracic bands greenish, lateral bands yellowish
green below, blue above. Spots of abdomen pale green, ex-
cept those on segs. 2 and 3, which are largely blue. The
ground colour of the abdomen is brown shading into black
around the light areas.

Measurements—Thor. @& II-12, 9 I1.3-11.5; abd. ¢@
43-45, 2 43-46; depth seg. 2 o& 5.4-6, 95.8-6; seg. 3 ot
6.5-7.2, 9 6-6.25; seg. 4 co 5-5.25,2 5.4-5.7; apps. co” 6-6.5,
Q 5.5-6.2; gen. v. 2.25; h.w. o 42.5-45, 2 44-45; width h.w.
oS 13-14, 9 13.5-14.5; pter. oO 2.8-3.6, 9 3.5-3.8.

Types—o' 2, Mus. Comp. Zoology.

Nymph (pl. 7, figs. 4,4a;pl.9, fig. 5; pl.11, fig. 5)—General
surface smoother than usual, more than ordinarily transpar-
ent. Eyes large but less prominent than in species of the
juncea, clepsydra and cyanea groups; lateral margins of head
short, curving evenly from the eye to the straight posterior
margin. Mentum of labium about twice as broad at apex as
at base, the apical breadth equal to about eight-ninths of the

length, expanding distad throughout its length, especially
in the distal three-sevenths, though this part is less dilated
than in most of our species of the genus; lateral margins sinu-
ate, the curve being feebly convex or almost straight in the
proximal part, more strongly so towards the distal end.
Middle lobe very little produced, subangulate; lateral lobes,
distad of movable hook, broad, parallel, squarely truncate;
outer angles but little rounded off, inner angles with a small
tooth.

Supracoxal processes short conical, subequal; interval
rather greater than a right angle.

Abdomen rather short and broad, strongly convex above
in the exuvia, in which the greatest breadth is reached at seg.
5; segs. 6 and 7 being of about the same breadth. Lateral
spines present on 6-9; those on 6 mere rudiments, on 7 ex-
tending two-thirds to four-fifths of the distance to the pos-
terior margin of the segment, on 8 to the margin or slightly
beyond, on 9g to the middle of 10 or just beyond. Venter of
9 nearly three times as broad at base as long. Lateral ap-
pendages three-fifths as long as the inferior pair, rather
slender in both sexes, tapering gradually to the slender fine-

196 WALKER: NORTH AMERICAN SPECIES OF AESHNA

pointed apices; distal third in the male somewhat excavated
below. Breadth of basal part of superior appendage of
male equal to about three-fourths of the length of
one of the sides, which are somewhat concave ; apex
small, rounded. Genitalia of female small, about one-
third of their length shorter than segment 9; genital valves
not steeply declivent.

Colour somewhat dark brown, more or less spotted and
mottled with pale yellowish. Head above with a group of
pale spots between the eyes, including a larger central spot
and three or four smaller ones on each side; a pair of cres-
centic spots bounding the scars externally, a spot immediately
behind each eye and a lateral marginal band which is con-
tinued across the outer ends of the pronotum. Thorax with
a number of small scattered pale spots; wings with the veins
heavily marked with dark brown, the interspaces pale. Legs
dark brown, the femora and tibiae each with three pale an-
nuli. Abdomen variable as to distinctness of the pale mark-
ings. In well-marked specimens the following can be made
out: (1) A series of pairs of subtriangular spots, each pair
at the base of one of the segments. Anteriorly the spots are
largest and coalesce to form two irregular bands; posteriorly
they generally decrease in size and often become subobsolete
on the last two or three segments. (2) A series of subcres-
centic spots immediately laterad of. the former series be-
tween the dorso-lateral and lateral punctae. (3) Aseries of
pale blotches surrounding the lateral scars which are out-
lined in dark brown. (4) A mid-dorsal series of sublanceolate
spots on some of the segments, each spot surrounded by the
corresponding dorsal punctae. All the punctae are distinct-
ly marked with dark brown.

Measurements—Length of body 35-40; mentum of
labium 5-5.75; h.f. 6-6.9; h.w. 8-9; inf. apps. 4-4.5; width of
head 7.7-8.5; width of abdomen 8.3-9 (exuvia 7-8.5).

The nymphs and exuviae upon which the above descrip-
tion is based were all taken by Mr. Kennedy at Nigger
Pond, Sunnyside, Wash., where they were associated with
those of Ae. californica, but occurred inlargernumbers. None
were reared, but there can be no doubt that they belong to

WALKER: NorTH AMERICAN SPECIES OF AESHNA 197

multicolor, for many of the exuviae were found on May 8,
a date that is much too early for any other species of Aeshna
in this locality. Moreover the resemblance to the nymph of
californica is so close that it could scarcely belong to any
species except multicolor, the only near relative of californica
in this region; and, indeed, some of the differential characters
of the two nymphs closely correspond to those of the adults
of these species. Thus (1) the triangular basal part of the
superior appendage of the male is more elongate in the
multicolor nymph, this structure being the rudiment of the
inferior appendage of the adult male, which is likewise more
elongate than in californica (pl. 7, figs. 3a, 4a; pl. 17, figs. 4, 5);
(2) the number of antenodal veins is greater in both nymph
and adult of multicolor in the great majority of cases ; (3)
the size of the multicolor nymph is distinctly larger.

Only californica was seen about the pond on May 8,
when the first lotof exuviae wascollected, but on July 17, when
the second lot was taken, only multicolor was present.
The first lot of exuviae consisted of both species, the second
of multicolor only.

Material determined — 45 o& 17 @Q. BritTISH COLUMBIA: Loon Lake,
Ainsworth (Kootenay), July 11, 1903 (Currie, U.S.N.M., 2 o&); Victoria, July
(Crotch, M.C.Z., 1 o&); Departure Bay, Vancouver Is., June 20, July 4-23, 1908
(Huntsman, 12 &' 3 9); Diver Lake, Vance. Is., July 23, 1908 (Huntsman,
I 6’). WASHINGTON: Sunnyside, June 13, July 9, I910 (Kennedy, 7 & 3 9).
OreEGon: Portland, July 12, 1905 (Currie, U.'S.N.M.,1 9). IpaHo: Moscow Mt.
July 27 (H.E. Burke, U.S.N.M., 1 9). CoLorapo: Fort Lupton, July 22, 1900
(Osburn, coll. Williamson, 1 o, fragments); Denver (E. V. Beales, coll. Calvert,
I 9). Uran: Red Butte Cafion, Salt Lake Co., June 18, 1899 (Browning, coll.
Calvert, 1 9). CatiForntiA: (M.C.Z., B. Gerhard, coll. Williamson, 4 oI 9);
San Francisco (R. Osten Sacken, M.C.Z., 1 2); Savoclito, June 1 (M.C.Z., 107);
Claremont (Baker, coll. Calvert, 1 9); Keeler; July 6 (Wickham, coll. Calvert,
1 9); Ontario, July 13, 1907 (Williamson, 3 oo); Los Angeles (U.S.N.M.,-A.
Davidson, coll. Calvert, 1 co 1 9); Pasadena, Sept. 5-13, 1900 (Fordyce Grinnell,
jr., coll. Williamson and Calvert, 3 o’); San Diego, April (Crotch, M.C.Z., 1%).
ArRIzonaA: Winslow, July 31, 1901 (Barber and Schwarz, U.S.N.M., 1”); Williams,
July (Barber and Schwarz, U.S.N.M., 1c7).. New Mexico: Beulah, Aug. I901
(Skinner, Ac. N.S. Phil., 1 o&); La Cruces (U.S:N.M., 19). TExas: Pecos
River (M.C.Z., 1% 2 9). Mexico: San José del Cabo, Baja California, Oct.
1893 (G. Eisen, coll. Calvert, 1 oc’); Tacubaya, D.F., April, 1899 (O. W. Barrett,
coll. Calvert, 1 o&); Santa Maria, Puebla, June (O.W. Barrett, coll. Calvert, 1%).

Nymphs—Nigger Pond, Sunnyside, Wash., May 8, 1910, 86°32 (St. H);
15¢' 109 (St. F); exuviae 13c°11 9. Id., July 17, 1910, exuviae 1207 21 9,

Distributton—Upper and Lower Sonoran, Mexico (Dis-
trito Federal and Puebla) to Texas, Colorado and southern
British Columbia. Of the material recorded by Professor

198 WALKER: NoRTH AMERICAN SPECIES OF AESHNA

Calvert as Ae: multicolor from Costa Rica and Panama, I
have seen 1 9 from Irazu (Costa Rica) and 1 o& from the
Volcan de Chiriqui (Panama). These belong to Ae. jala-
pensis Williamson, as does also a @ from Amula, Guerrero.

Aeshna mutata Hagen.
(Pl. 14, fig. 6; pl. 17, figs. 6, 6a; pl. 21, figs. 4, 4a; pl. 22,

fig. 5; pl. 28, fig. 3.)

Aeschna mutata, Hagen, Syn. Neur. N.A., p. 124 (1861); Kirby, Syn. Cat. Od.
N.A., p. 89 (1890); Banks, Trans. Am. Ent. Soc., XIX, p. 353 (1892).

Aeshna mutata, Williamson, Ent. News, XIX, pp. 264, 301, text fig. (1908);
Walker, Can. Ent., XL, pp. 379, 386, 450 (1908); Muttkowski, Cat. Od. N.A., p.
113 (1910); Skinner, Ent. News, XXII, p. 336 (1911).

Aeschna multicolor, Weith, Ent. News, XI, p. 641 (1900) ; Williamson, Pr.
Ind. Ac. Sci., p. 173, 176 (1900).

Aeschna verticalis, Osburn and Hine, Ohio Nat., I, p. 14 (1900).

A species of average size and build, with a rather short
abdomen, and somewhat long and narrow wings.

Male—Occiput dull yellowish (bluish in life?), edged
laterally with black, one-third to two-fifths as long as the line
of contact of the eyes. Frontal vesicle pale blue above, the
blue area separated from the lateral ocelli by a space fully as
broad as one of the latter. Eyes in life pale blue, in dried
specimens dark olivaceous with a bluish cast. Preocular
band about twice as broad at base of antenne as one of the
lateral ocelli; narrowed toa hair-line on the lower half of the
frons and nasus. T-spot 2.3-2.8 mm., stem about 1.75 mm.
broad behind, .75 mm. broad in front, the sides distinctly
sinuate, being straight or concave in front, convex behind.
Frons and nasus pale blue or grey-blue, pale yellowish next
to the ocular margin; lateral lobes of nasus rounded obtus-
angulate, slightly flaring. Labrum in dried specimens brown-
ish with a very narrow basal and a broader apical dark brown
margin. Labium plumbeous.

Thorax slenderer than in Ae. multicolor, moderately
dark brown with a coppery tinge in dried specimens. Dorsal
thoracic bands light blue, nearly 4 mm. long and about I mm.
broad, except at the upper ends, where they nearly meet in
the middle line justin front of the antealar sinus, and where
the breadth is increased to2 mm. Sometimes the lower ends
are also expanded. Lateral bands pale blue, nearly straight,
surrounded by a diffuse dark brown margin; the first band
slightly sinuate just below the middle, about 1 mm. broad at

WALKER: NoRTH AMERICAN SPECIES OF AESHNA_ 199

the lower end, expanding and becoming diffuse at the upper
end, where the breadth is 2.5 to 3 mm.; second band straight
and equal except at the expanded upper end, breadth below
.5--75 mm., above 1.5 to 2 mm.

Legs black, proximal half of anterior femora with a pale
streak on the posterior surface.

Abdomen four to four-and-a-half times as long as the
thorax, moderately slender, somewhat less inflated than
usual at base, moderately constricted at 3, then rapidly
expanding to the apex of 4. Seg. 1 with a very low
ventral tubercle, bearing a few inconspicuous spinules
and a tuft of long pale brown hairs. Auricles_ with
2-3 teeth; ventral surface of 2 with a few minute
spinules, especially nearly the posterior margin. Tergal mar-
gins bounding the genital fossa strongly sinuate, consider-
ably elevated in the posterior third. Spine of the anterior
lamina well-developed, moderately curved dorsad, sharply
pointed, extending caudad to the posterior margins ofthe
hamular processes; the latter rather broad and thick, meeting
along the middle line, antero-inferior surfaces concave,
bounded behind by an arcuate ridge, which runs from the
base to the antero-internal angle, posterior margins arcuate.
Hamular folds continuous with the processes, elongate,
acutangulate behind, deeply concave and closely approxi-
mated mesially, their outer margins subparallel. Seg. 3
one-seventh longer than 1+2, one-third longer than 4.
Segs 5-7 with traces of accessory lateral carinae about the
middle. Median dorsal tooth on seg. 10 appearing very low
and elongate in profile view.

Superior appendages slightly longer than 9+10, slender
in the basal fifth, then expanding on the inner margin to
about five-twelfths of the length,where the breadth is a little
more than twice that at base and equal to about one-fifth of
the length; thence remaining equal as far as the distal seventh,
whence, curving strongly downwards and slightly outwards,
they taper to the sharply pointed apices. In dorsal view
they appear slender, the outer margin gently and somewhat
irregularly arcuate, the inner margin slightly sinuate.

Superior carina percurrent, low in the proximal two-

200 WALKER: NORTH AMERICAN SPECIES OF AESHNA

thirds then suddenly elevated with an angulate crest, the
apex of which is about one-third of the length of the appen-
dages from the apex of the latter. In profile a low sub-
basal tubercle at one-seventh to one-sixth of thelength appears;
the outer margin is nearly straight except towards the apex, the
superior carina raised above it to a height usually much less
than the depth of the inner part of the appendages directly
beneath it, this part being bent downwards to form an in-
ferior carina, which is produced into a strong spine, directed -
ventrad and caudad, its apex recurved. The distance be-
tween its apex and that of the appendage is equal to about
one-sixth of the length of the latter. Inferior appendage elon-
gate-triangular, slightly acuminate, seven-twelfths as long as
the superior pair, the basal breadth three-eighths of the length,
upper surface without a median carina, apex bluntly rounded;
in profile view rather strongly curved proximally, more gently
distally, tapering more rapidly caudad than in multicolor.

Seg. 1 fuscous, dorsal and lateral spots blue, the latter
3-3.5 mm. long, somewhat less than 1 mm. broad.

Seg. 2 fuscous, spots all blue. AML with the anterior
margin extending I-2 mm. above the auricle and a variable
distance below it; upper margin very oblique, continuous with
MD; posterior part of thesegment above the ventral surface
wholly blue, except a narrow brown band on each side, passing
obliquely mesad from the transverse carina, and a dark brown
spot behind the auricles. A few irregular blue blotches on
the ventral surface also.

Segs. 3-10dark brown. MD on all the segments and PD
on 10 pale yellowish (green in life?), the other spots blue.
AL on 3-7 or -8, each pair except the last connected by a nar-
row basal ring; large on 3, the upper margins straight and
separated by about 1.5 mm.; of fairly large size also on 4-7,
the upper margins nearly straight. ML on 3-8, of moderate
size, irregularly quadrate, the anterior edge of all but the last
pair angularly indented, those on 3-5 or -6 narrowly confluent
with MD. MD on 3-8, the last pair rudimentary, the others
well-developed, triangular with the hind edges concave.
PD on 3-10, pairs of semi-elliptical spots about 1.5 mm. long
on most of the segments, narrowly separated mesially except

WaLKER: NORTH AMERICAN SPECIES OF AESHNA 201

on 8-10, where the intervals are much wider, especially ‘on
10; on which the spots are small and ill-defined. PL on 3-5,
generally rudimentary on 5, elsewheré a narrow curved off-
shoot from PD.

Wings rather narrow, more or less flavescent, the hind
pair about as long as the abdomen; costal veins dull brownish
yellow; pterostigma dark brown, ochraceous beneath, mem-
branule of hind wings smoky brown, with the base more or
less whitish or grey, reaching nearly to the apex of the anal
triangle, which is 3-celled. Az2 in hind wing arising basal
to the last .cubito-anal cross-vein before the subtriangle;
outer side of anal loop about as long as inner side of triangle.
Rs forking opposite the third (fore wing) or the third to fifth
(hind wing) postnodal cell before the stigma, the fork nearly
symmetrical at base ; 3 or 4 rows of cells between Rspl and Rs
between the points of widest divergence. Mua generally
arising opposite the distal end of the stigma but often a little
beyond or before the end. Two rows of cells between M3
and Mg in the hind wing, from the point where M4 appears
forked to the margin of the wing.

Antenodals 7-21, ale as 2? CuCr 22 > Spt 24 its
-14 10-13 5-6 2-3,

Aiden as long or a little longer than i in the
stale, rather slender.

Genital valves barely as long as seg. 9, in profile very.
slightly arcuate, the.apices a little elevated.; lateral carinae
rounded; the ventral surfaces not distinctly sulcate; posterior
margin of basal plate straight; styli not quite 1 mm. long.
Appendages a little longer than 8+9, slender at base, outer
margin straight, inner margin strongly arcuate, the curve at
apex a little more marked than at base; greatest breadth
about the middle, equal to about one-fifth of the length;
apices rotundo-angulate.

Colour-pattern—[Colours mainly as described by Wil-
liamson, (’08) from a single specimen]. Occiput greenish
yellow; eyes dark greenish brown with a narrow green poster-
ior border, paler below. Face yellowish green, obscured with
brown, margined with yellowish; frons above dull bluish, a

14

202 WALKER: NORTH AMERICAN SPECIES OF AESHNA

pale brown line on the fronto-nasal suture; rhinarium plum-
beous; labrum pale olivaceous; labium olive-green. Thorax
olivaceous brown; dorsal bands divided each into a superior
and an inferior greenish spot; lateral bands green, yellowish
above. Abdomen castaneous, shading into black around the ~
pale areas, which are green; on the first three or four segments
margined with yellowish, in the posterior segments obscured
with drabbish. On seg. 2 the lateral spots are connected to
form a single broad band, and on 3 and4 ML is confluent with
PL. PD somewhat smaller, PL larger, than in the male.
Wings strongly flavescent, the stigma pale brown.

In the females from Angola, Ind.,and Wilbrahim, Mass.,
both of which are teneral, the lateral thoracic bands are
yellowish white,in part bluish, a possible indication of the
homeceochromatic type of coloration.

Measurements—Thor. o& @ 10-11; abd. & 45-47.75,
9 48-51; depth seg.2 co 4.9-5.2, 2 5.33-5-5; seg. 3 07.3-8,
97-7.5; seg. 4 oc 3.6-5.9, 2 6.5-6.75; apps. o& 5.8-6.3, 9
6.5-7.5; gen. v. 2-2.2; h.w. o& 44-47; 2 47-75-51; width
h.w. co 13-14, 9 14-14.33; pter. o& 3.3-3.75, 2 3-8-4.2.

Type— 9, Mus. Vienna—‘‘ North America.”

Nymph—Unknown.

Material determined—14 o& 49. MAassacHusetts: Wilbrahim, June 5,
1902 (Needham, 2 o 1 9, teneral). Onto: Stewart’s Lake, Kent, June 22,
I (J. S. Hine, coll. Osburn, ). INDIANA: Bluffton, June 30, July 2-13, 1907

illiamson, 12 o& 1 @); Angola, June, 1908 (Mary Shafer, coll. Williamson,
1 Q,teneral). ‘‘ NortH America”’ (M.C.Z.,1 9). ;

Identity—This species is closely related to Ae. multt-
color, but I have seen no individuals that are in any way in-
termediate between the two forms. In fact multicolor shows
no tendency anywhere to vary towards mutata, exhibiting
indeed a marked uniformity of type throughout its entire
area of distribution, as far as is indicated by the material I
have studied.

_The Mexican species Ae. jalapensis Williamson is still
more closely related to Ae. mutata.

Distribution—Carolinian; Indiana, Ohio, Massachusetts,
Pennsylvania.

Habits—All that is known of the habits of this species
has been given by Williamson (Joc. cit.).

WALKER: NoRTH AMERICAN SPECIES OF AESHNA 203

LITERATURE CITED.

BAKER, C. F.
1905. Notes on the Neuropteroid Insects of the Pacific
' Coast of North America, with Descriptions of New
Species by Nathan Banks. Invertebrata Paci-
fica, I, pp. 85-92.
BALFOUR-BROWNE, FRANK.
1909. The Life-history of the Agrionid Dragonfly.
Proc. Zool. Soc. London, pp. 253-258, pls.
33-34-
BANKS, NATHAN.
1892. A Synopsis, Catalogue, and Bibliography of the
Neuropteroid Insects of Temperate North Amer-
ica. Trans Am. Ent. Soc., XIX, pp. 327-374.
1894. The Odonata of Ithaca, N.Y., Can. Ent., XXVI,
pp. 76-78.
BERGROTH, E.
1881. Zur geographischen Verbreitung einiger Odona-
ten. Entom. Nachr., VII, pp. 85-88.
BEUTENMULLER, WILLIAM.
1890. Preliminary Catalogue of the Odonata found in
the State of New York. Dragonflies vs. Mos-
quitoes, pp. 163-164.
BRIMLEY, C. S.
1906. North Carolina Records of Odonata in 1904 and
1905, with Corrections of some Previous Records.
Ent. News, XVII, pp. 91-92.
Brown, J. J.
1891. Flights of Dragonflies. Insect Life, III, pp. 413-
414.
BuRMEISTER, H.
1839. Handbuch der Entomologie, Vol. II. Odonata,
pp. 805-862.
BURNHAM, EDWARD J.
1900. Preliminary Catalogue of the Anisoptera of Man-
chester. Proc. Manch. Inst. Arts and Sciences,

I, pp. 27-38.

204 WALKER: NORTH AMERICAN SPECIES OF AESHNA

BuTLER, HORTENSE. |

1904. The Labium of the Gdosate: Trans. Am. Ent.

Soc., XXX, pp. 111-134, pls. II-VI.
CasBoT, Louls.

1881. The Immature State of the Odonata. Part II,
Subfamily AXschnina. Mem. Mus. Comp. Zool.,
VIII, No. 1, pp. 1-40, pls. I-V.

CALVERT, PHILIP P.

1893. Catalogue of the Odonata (Dragonflies) of the
Vicinity of Philadelphia with an Introduction to
the Study of this Group of Insects. Trans. Am.
Ent. Soc., XX, pp. 152a-272, pls. II-III.

1894a. On theSpecific Identity of Aeschna clepsydra Say
and Ae. crenata Hagen (eremita Scudder.) Ent.
News, V, pp. 9-13, figs. I-5.

1894b. Data on the Distribution of Dragonflies. Ent.
News, V, pp. 242-244.

1894c. Notes on Nova Scotian Dragonflies. Can. Ent.,
XXVI, pp. 317-320.

1895a. Odonata of New York State. Journ. N.Y. Ent.
Soc., III, pp. 39-48.

1895b. The Odonata of Baja California, Mexico. Proc.
Cal. Acad. Science, Second Series, IV, pp. 463-
558, pls. XV-XVII.

1897. Additons to the Odonata of the State of New York.

~ Journ. N.Y. Ent. Soc., V, pp. 91-95.

1898. On Burmeister’s Types of Odonata. Trans.
Am. Ent. Soc., XXV, pp. 27-104.

1899. Foot-note on ‘“‘A Note on Copulation among
Odonata”’ (Williamson). Ent. News, X, p. 42.

1900a. Insects of New Jersey—Odonata. 27th Ann. Rep.
N. J. State Board Agric., pp. 66-75.

1900b. Note on ‘‘Kingbirds eating Dragonflies’’ (J. P.
Moore). Ent. News, XI, p. 340.

1903. A List of the Insects of Beulah, New Mexico
—Neuroptera Odonata. Trans. Am. Ent. Soc.,
XXIX, pp. 42-45.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 205

1905a. Fauna of New England, 6. List of the Odonata.
Occasional Papers Bost. Soc. Nat. Hist., VII, pp.
I-43.
1905b. Biologia Centrali-Americana—Neuroptera Odon-
ata. Pp. 17-420, pls. I-X. London (1901-1908).
1906. Copulation of Odonata, II. Ent. News, XVII,
pp. 148-150, pl. VII.
1909. The Composition and Ecological Relations of the
Odonata Fauna of Mexico and Central America.
Proc. Ac. Nat. Sci. Phil., 1908, pp. 460-491, with
Map, pl. XXVI.
1909. The Insects of New Jersey—Odonata.
CAMPBELL, F. M.,
1885. Dragonfly Migration. Ent. Month. Mag., XXI,
p. 192.
CHARPENTIER, TOUSSAINT DE.
1840. Libellulinae Europaeae, pp. 1-180, pls. I-XLVIII.
Leipzig.
COCKERELL, -F. .D:) A.
1908. A Dragon-fly Puzzle and its Solution. Ent.
News, XIX, pp. 455-459.
Comstock, G. F.
1903. A List of Lepidoptera found in the Adirondack
Mountains. Ent. News, XIV, pp. 197-200.
Odonata p. 200.
CurRIE, ROLLA P.
1901. Papers from the Harriman Alaska Expedition,
XXII. Entomological Results (14): The Odon-
ata. Proc. Wash. Acad. Sci., III, pp. 217-223.
1903. The Odonata collected by Messrs. Schwarz and
Barber in Arizona and New Mexico. Proc. Ent.
Soc. Wash., V, No. 4, pp. 298-303.
1905. Dragonflies from the Kootenay District of British
Columbia. Proc. Ent. Soc. Wash., VII, No. 1,
pp. 16-20.
Davis, Wm. T.
1898. Preliminary List of the Dragonflies of Staten
Island, with Notes and Dates of Capture.
Journ. N. Y. Ent. Soc., VI, pp. 195-198.

206 WALKER: NORTH AMERICAN SPECIES OF AESHNA

ELrRop, Morton J.
1902. A Biological Reconnaissance in the Vicinity of
Flathead Lake. Bull. Univ. Montana, No. 10,
Biological Series No. 3, pp. 91-182.

Evans, W. F.
1845. British Libellulinae, pp. 1-28. 21 coloured figs.
London.

FLETCHER, JAMES.
1906. Entomological Record, 1905. 36th Ann. Rep.
Ent. Soc. Ont., p. 104.
FLETCHER JAMES, AND GIBSON, ARTHUR.
1908. Entomological Record, 1907. 38th Ann. Rep.
Ent. Soc. Ont., pp. 113-141.
1909. Entomological Record, 1908. 39th Ann. Rep.
Ent. Soc. Ont., pp. 99-120.
FyLes, THomas W.
1901. The Dragonflies of the Province of Quebec.
31st Ann. Rep. Ent. Soc. Ont., pp. 52-55.
GIBSON, ARTHUR.
1910. The Entomological Record, 1909. goth Ann. Rep.
Ent. Soc. Ont., p. 110-128.
1911. The Entomological Record, 1910. 41st Ann.
_ Rep. Ent. Soc. Ont., pp. 101-120.
HAGEN, HERMANN AUGUST.
1840. Synonymia Libellularum Europaearum, pp. 1-84.
Koenigsberg, Prussia.
1856. Die Odonaten-Fauna des russischen Reichs. En-
tom. Zeitung Stettin, XVII, pp. 363-381.
1861. Synopsis of the Neuroptera of North America,
with a List of the South American Species.
Smithson. Misc. Coll., 1V. Odonata, pp. 55-187.
1873a. Report on the Pseudoneuroptera and Neurop-
tera of North America in the Collection of the
late Th. W. Harris. Proc. Bost. Soc. Nat. Hist.,
XV, pp. 263-301.
1873b. Report on Mr. S. H. Scudder’s Odonata from the
White Mountains, after an Examination of the
Types... Lic; XV; pp: 376-377.

WALKER: NORTH AMERICAN SPECIES OF AESHNA 207

1873c. Odonata from the Yellowstone. Sixth Ann.
Rep. U. S. Geol. Surv. Terr., 1872 (Hayden),
PP. 727-729.

1874. Report on the Pseudo-neuroptera and Neurop-
tera collected by Lieut. W. L. Carpenter in 1873
in Colorado. Ann. Rep. U.S. Geol. and Geog.
Surv. Terr., 1873 (Hayden), pp. 571-606.

1875a. Report upon the Collections of Neuroptera and
Pseudo-neuroptera made in portions of Colorado,
New Mexico and Arizona, during the years 1872,
1873 and 1874. Rep. Geog. Geol. Expl. Surv.
West rooth Merid. Engineer Dept. U.S. Army,
pp. 9II-922.

1875b. Synopsis of the Odonata of North America,
Proc. Bost. Soc. Nat. Hist., XVIII, pp. 20-96,

1890. TwoSpeciesof Aeschna. Psyche, V, pp. 353-355.

HARVEY, FRANCES LEROY.

1891. A Contribution to the Odonata of Maine. Ent.
News, II, pp. 73-75.

1902. A-Catalogue and Bibliography of the Odonata of
Maine with an Annotated List of their Collectors.
University of Maine Studies, No. 4.

HEBARD, MORGAN.
1910. A few Records from Northern Michigan in the
Order Odonata. Ent. News, XXI, pp. 134-135.
Howaprp, L. O.
1901. The Insect Book. Odonata, pp. 363-376, pls.
XL-XLVIII. New York.
KAMMERER, PAUL.

1907. Symbiose zwischen Libellenlarve und Fadenalge.

Arch. fiir Entw. der Org., X XV, pp. 53-81.
KARSCH, FERDINAND.

1891a. Kritik des Systems der Aeschniden. Ent. Nachr.,
XVII, pp. 273-290.

1891b. Acht neue Aeschniden, I.c., XVII, pp. 305-313.

KELLICOTT, DAVID S.

1895. Catalogue of the Odonata of Ohio, Part I.

Journ. Cinc. Soc. Nat. Hist., XVII, pp. 195-216.

208 WALKER: NORTH AMERICAN SPECIES OF AESHNA

1896, Catalogue of the Odonata of Ohio, Part II, L.c.,
XVIII, pp. 105-114.
1899. The Odonata of Ohio. Special Papers, Ohio
Acad. Sci., No. 2. viii + 116 pp., pls. 1-4,
Kirsy, W, F.
1890. A Synonymic Catalogue of Neuroptera Odonata
or Dragonflies, with an Appendix of Fossil
Species. x+202 pp. London.
KOLENATI, FREDERICO A.
1846. Meletemata entomologica, V—Neuroptera, pp.
113-120. St. Petersburg.
LINNE, CARL VON. |
1758. Systema Naturae, [—Neuroptera, pp. 543- 552.
Lueas,:W, J:
- 1900. British ‘Dragonflies (Odonata). xiv + 356 PP. 27
Be Ps pls.,.57 text figs. - London.
MACLAUGHLIN, T. i
1887, Ottawa Dragonflies. Trans. Ottawa Field. Nat.
Club., II, pp. 329-342.
MartTIN, RENE.
1908. Collections Zoologiquesdu Baron Edm. de Sale
Longchamps. Catalogue systematique et des-
criptif. Fasc. XVIII, Aeschnines, pp. 1-84, pls.

I-II.
1909. Ibid. Fasc. XIX, Aeschnines, pp. 85-156, pls.
; ITI-IV. ,
1910. Ibid Fasc. XX, Aeschnines, pp. 157-223,
V-VI. |

1911. Genera Insectorum, Fasc. 115. Odonata, Fam.
Aeschnidae, Subfam. Aeschninae, pp. 1-34, pls.
_ I-VI.
MooreE, J. PERCY.
1900. Kingbirds eating Dragonflies. Ent. News, XI,
Pp. 340.
MUELLER, F. O.
1867. Enumeratio ac Descriptio Libellularum agri
Friedrichsdalensis. Nova Acta Acad, Caes.
Leop. Car., III, pp. 122-131.

WALKER: NoRTH AMERICAN: SPECIES OF AESHNA | 209

Muttkowskli, RICHARD A.
1908a. Review of the Dragonflies of MWaeeorsin. Bull.
Wis. Nat. Hist: Soe.,. VI, pp.'57-127..- . -
1908b. A. Summer’s Insect. Collecting—-Odonata. Dict!
VI, pp. 166-168.
I910a. New Records of Wisconsin Dragonflies. L.c.,
VIII, pp. 53-59.
1910b. Catalogue of the Odonata of North America.
Bull. Publ. Mus. Milw., I, art. 1, pp. 1-207.
1910c. Miscellaneous Notes and Records of Dragonflies
(Odonata). Bull. Wis. Nat. Hist. Soc., VIII, pp.
AGO Ge BOaT 7G,
Igtt. New Records of Wisconsin Dragonflies, II. Lc,
IX, pp. 28-41, pl. IV.
NEEDHAM, JAMES G.
1900. The Fruiting of the Blue ile amen Nat.,
XXXIV, No. 401, pp. 361-386.
1901. Aquatic Insects in the Adirondacks. Bull. 47,
N.Y. State Mus., pp. 383-612, pls. I-XXXVI,
42 text figs.
¥903a.-Aquatic Insects in New York State. Bull. 68,
- = ....N.Y.State Mus., Ent. 18,-parts 2 and 3, Pp. 204-
1903b. A Cenralowie Saitly, of Dragonfly Venation.
Proc. U.S. Nat. Mus., XXVI, pp. 703-764, pls.
I-XXIV, 44 text figs.
1904. New Dragonfly Nymphs in the United States
National Museum. L.c., XXVII, pp. 685-720.
1908. Notes on the Aquatic Insects of Walnut Lake.
‘Rep. Board Geol. Surv. Mich. 1907 on the
Biology of Walnut Lake. App. III, pp. 252-271.
1910. Report of the Entomologic Field Station con-
ducted at Old Forge, N.Y., in the Summer of
1905. Bull. 124, N.Y. State Mus., App. C., pp.
156-284.
OsSBURN, RAYMOND C.
1905. The Odonata of British Columbia. Ent. News,
XVI, pp. 184-196.

2tT0 WALKER: NORTH AMERICAN SPECIES OF AESHNA

1906. Some Experiments with Dragonflies in Brackish
Water. Amer. Nat., XL, pp. 395-399.
OssBurN, R. C., AND HINE, J. S.
1900. Dragonfliestaken ina Week. Ohio Nat., I, p. 14.
PROVANCHER, L’ABBE.

1897. Petite Faune entomologique du Canada. Neév-
roptéres. Le Naturaliste Canadien, IX (con-
tinued from vol. VIII), pp. 38-43.

Ris, F.

1885. Neuroptera Helvetiae, analytisch bearbeitet als
Grundlage einer Neuropteren-fauna der Schweiz
(Gustav Schoch). Odonata (Die schweizeri-
schen Libellen), pp. 35-85.

RUTHVEN, ALEXANDER G.

1906. Spiders and Insects from the Porcupine Moun-
tains and Isle Royale, Michigan. From ‘An
Ecological Survey of Northern Michigan”’ (C. C.
Adams). Ann. Rep. State Board Geol. Surv.,
1905, pp. 100-106.

Say, THOMAS.

1839. Descriptions of New North American Neurop-
terous Insects, and Observations on some already
described by (the late) Th. Say. Journ. Acad.
Nat:Se.) Phil., VILL, Pel, pp.)\9-46.

SCUDDER, SAMUEL H.

1866. Notes on Some Odonata from the White Moun-
tains of New Hampshire. Proc. Bost. Soc. Nat.
Hist., X, pp. 211-222.

SELYS-LONGCHAMPS, EDMOND DE.

1840. Monographie des Libellulidées d’Europe, pp.
1-220, pls. 1-4. Paris and Brussels.

1850. Revues des Odonates ou Libellules d’Europe,
pp. 1-408, pl. I-II. Brussels and Leipzig.

1875. Notes on Odonata from Newfoundland, collected
in 1874 by Mr. John Milne. Ent. Month. Mag.,
II, pp. 241-243.

SELYS-LONGCHAMPS, EDMOND DE, and MACLACHLAN, ROBERT.

1872. Matériaux pour une faune névroptérologique de
l’Asie septentrionale. Ann. Soc. Ent. Belg.,

WALKER: NORTH AMERICAN SPECIES OF AESHNA 2II

XV, pp. 25-77 (Odonata by Selys-Longchamps,
Pp. 25-45).
SKINNER, HENRY M.

1902. ‘Doings of Societies.” Ent. News, XIII, p. 26.

1963... -Ibid., 1.c., XV, ‘p:. 288:

Pout Paid. hc: XXL) ‘p:'336;

SMITH, JOHN B.

1909. Insects of New Jersey. Rep. N. J. State Museum.

Odonata, pp. 73-82.
STEFANELLI, P.

1882. Osservazioni sui costumi e sullo svilluppo dell’
Aeschna cyanea Miill. Bull. Soc. Ent. Ital.,
1882, pp. 236-238.

STEPHENS, JAMES FRANCIS.

1835. Illustrations of British Entomology, or a Synop-
sis of Indigenous Insects, etc., VI—Mandibulata.
Odonata, pp. 70-96. London.

TILLYARD, R. J.

1909. Studies in the Life-histories of Australian Odon-
ata. I.The Life History of Petalura gigantea
beach: Proc. Lin Soe ins tS. W.. Koen
part 2, pp. 256-267, pl. XXIV.

Van DuzeEE, E. P.

1897. List of the Dragonflies taken near Buffalo, N.Y.

Journ. N. Y. Ent. Soc., V, pp. 87-91.
WapbswortTH, Miss MATTIE.

1890. List of the Dragonflies (Odonata) taken at Man-
chester, Kennebec Co., Me., in 1888 and 1889.
Ent. News, I, pp. 36-37, 55-57-

WALKER, EpMUND M.

1905. Orthoptera and Odonata from Algonquin Park,
Ontario. 36th. Ann. Rep. Ent. Soc. Ont., pp.
64-70.

1906. A First List of Ontario Odonata. Can. Ent.,
XXXVIII, pp. 149-154.

1908a. The Dragonflies of the Ottawa District. Ott.
Nat., XXII, pp. 49-63.

1908b. A Key to the North American Species of Aeshna.
Can. Ent., XL, pp. 377-391, 450-451.

212 WALKER: NORTH AMERICAN SPECIES OF AESHNA

WALLENGREN, H. D. J.

1894.

WEITH, R.

1900.

Ofversikt af Skandinaviens’ Pseudoneuroptera.
Ent. Tidskrift, IV, pp. 235-270.

J.
Indiana Odonata. Ent. News, XI, p. 641.

WILLIAMSON, EDWARD B.

1897.

1899.

1900a.

~- 1900b.

19OI.

1902a.

1902b.

1903a.
1903b.

1906a.

1907a.

September Dragonflies of Round and Shriner
Lakes, Whitley Co., Indiana. Rep. State Geol.
Ind., pp. 403-405.

A note on Copulation Among Odonata. Ent.
News, X, pp. 42-43.

The Dragonflies of Indiana. 24th Ann. Rep.
Dept. Geol. Nat. Res. Ind., pp. 229-233.
Additions to the Indiana List of Dragonflies, with
a Few Notes, No. I. Proc. Ind. Ac. Sci., pp.
173-178.

Additions to the Indiana List of Dragonflies, with
a Few Notes, No. II. L.c., pp. 119-127.

A list of the Dragonflies observed in Western
Pennsylvania. Ent. News, XIII, pp. 65-71,

108-113.

Dragonflies (Odonata) from the Magdalen Islands.
Ent. News, XIII, pp. 144-146.

A proposed new genus of Odonata (Dragonflies)
of the Subfamily Aeschninae, Group Aeschna.
Ent. News, XIV; pp. 2-9, pl. II.

The Dragonflies (Odonata) of Tennessee, with a
Few Records for Virginia and Alabama. Ent.
News, XIV, pp. 221-229.

Dragonflies (Odonata) collected by Dr. D. A.
Atkinson in Newfoundland, with Notes on some
Species of Somatochlora. Ent. News, XVII, pp.
133-139, pls. V and VI.

A Collecting Trip North of Sault Ste. Marie,
Ontario. Ohio Nat., VII, pp. 129-148.

WALKER: NortTH AMERICAN SPECIES OF AESHNA 213

1907b. List of Dragonflies of Canada. L.c., VII, pp.
148-150.

1908. Three related Species of Aeshna. Ent. News,
XIX, pp. 264-271, 301-308.

WILson, CHARLES BRANCH.

1909. Dragonflies of the Mississippi Valley collected
during the Pearl Mussel Investigations on the
Mississippi River, July and August, 1907. Proc.
U.S. Nat. Mus., XXXVI, pp. 653-671.

ZETTERSTEDT, JOHANNE WILHELMO.
1840. Insecta Lapponica. Neuroptera, pp. 1015-1080.
Leipzig.

1) ;
i

ed
i vis

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES

No. 11. PLATE
g/l
la oe
1b 2b
le 2c
a4 4a
3b 4b
3c 4c
an
oie ie ope
Syl <TA <4 ——
5a 6a 7a
| a Le
, Lo —
5b 6b 7b
a Sete ber
t 2&
5c 6c 7c

CLIMATIC VARIATIONS

UNIV. OF TORONTO ‘STUDIES, BIOLOGICAL SERIES No. 11. PLATE 2

PF. - Ad,

« M. W. DE a
—E. M v L HELIOTYPE CO. BOSTON
A. J. READING PHOTO.

COPULATORY POSITION

7 GOMPHUS SPICATUS Hagen.

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11

ad @]

3 4 6

E. M@. W. DEL.

OVIPOSITION

1-4 AESHNA EREMITA Scudd. 5,6 AESHNA CONSTRICTA Say.

HELIOTYPE CO. BOSTON

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 4,

AESHNA UMBROSA Walk

INSTARS B-G

PLATE 5.

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11.

HELIOTYPE CO. BOSTON

Walk,

AESHNA CANADENSIS

EMERGENCE OF IMAGO

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 6.

—E. M. W. DEL. HELIOTYPE CO. BOSTON

NYMPHS OF AESHNA

UNIV. OF TORONTO STUDIES, BIOLOGICAL Series No. 11. PORATED ve

Ven Orsi!

i om a oe Be a

4

rs
~
*

ay

ame |

4a

HELIOTYPE CO. BOSTON

E, M. W. DEL.

NYMPHS OF AESHNA

UNIV. OF TORONTO STUDIES, BIOLOGICAL Series No. 11. PLATE 8.

HELIOTYPE CO., BOSTON

LABIA OF NYMPHS

+)

i

UNIV, OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 9,

HELIOTYPE CO., BOSTON

LABIA OF NYMPHS

a

=

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11.

TERMINAL ABDOMINAL

GENITALIA OF FEMALE

SEGMENTS

At

NYMPHS

1D

LIOTYPE CO., BOSTOD

UNIV. OF TORONTO STuDIES, BIOLOGICAL SERIES No. 11. PLATE 11,

TERMINAL ABDOMINAL SEGMENTS AND

GENITALIA OF FEMALE NYMPHS

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 12.

E, M. W., CAM. LUC, ET DEL. HELIOTYPE CO., BOSTON

ANTERIOR HAMULI AND ANTERIOR LAMINA

UNIV, OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11.

« M. W., CAM. LUC. ET DEL.

A

ANTERIOR HAMULI AND ANTERIOR LAMINA

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11

ANTERIOR HAMULI AND ANTERIOR LAMINA

HELIOTYPE CO.,

BOSTON

ee ee
re el ay re ru oy
+ write eg

He

UNIV, OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11, PLATE TS:

E. M. W., CAM. LUC, ET DEL. HELIOTYPE CO., BOSTON

ABDOMINAL APPENDAGES OF MALES

ss

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLAT

E. M. W., CAM. LUC. ET DEL. HELIOTYPE CO., BOSTON

ABDOMINAL APPENDAGES OF MALES

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11, PLATE 17.

6a

E, M. W., CAM, LUC. ET DEL. HELIOTYPE CO., BOSTON

ABDOMINAL APPENDAGES OF MALES

s\

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIaS No. 11. PLATE 18.

E. M. W., CAM. LUC. ET DEL, HELIOTYPE CO. BOSTON

FEMALE GENITALIA AND APPENDAGES

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 19.

E, M. W., CAM. LUC. ET DEL. HELIOTYPE CO. BOSTON

FEMALE GENITALIA AND APPENDAGES

nia

20.

PLATE

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11.

HELIOTYPE CO. BOSTON

LUC. ET DEL.

E. M. W., CAM.

FEMALE GENITALIA AND APPENDAGES

4

oa

UNIV. OF TORONTO STUDIES BIOLOGICAL SERIES NO. 11. PLATE 21.

E. M. W., CAM. LUC. ET DEL. HELIOTYPE CO. BOSTON

FEMALE GENITALIA AND APPENDAGES

BA)

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No, 11.

14
-

ATE 22,

beh a =

1-3 AESHNA CAERULEA SEPTENTRIONALIS

4 AESHNA SITCHENSIS
5 6©AAESHNA MUTATA

6-11 AESHNA INTERRUPTA

UNIV, OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11. PLATE 23.

VP

» PS
2 7

—E. M. W. DEL. HELIOTYPE CO. BOSTON

1 AESHNA SITCHENSIS ¢% 3 AESHNA JUNCEA ¢

2 ” ’ e) 4 va 3

AESHNA SUBARCTICA ¢&

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No. 14

O. BOSTON

AESHNA EREMITA & ; SHNA INTERRUPTA

INTERRUPTA +
2 ” rm)

Sie

as

Bo. ot. 24 5)

PLATE 25.

Figs. 1. Aeshna Joni i o&', Go Home Bay, Ont. (Fresh.)

2.

‘““ 9, De Grassi Point, Ont. (Fresh.)

canadensis 3, Go Home sais oo oe)

ait at
bie eae

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES NO. 11 PLATE 25

E. M. W. DEL. HELIOTYPE CO, BOSTON

AESHNA CLEPSYDRA ¢ 3 AESHNA CANADENSIS ¢

” ” ©, 4 ” ” ” ©

Ne)

5 AESHNA VERTICALIS &

6 ” ”) ) g

be

oor

UNIV, OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11

m
=
n

1 A&SHNA TUBERCULIFERA ¢

< }

AESHNA UMBROSA

” ”

4

UMBRO

A

4

LATE 26
BOSTON

aN

UNIV.

OF TORONTO STUDIES, BIOLOGICAL SERIES No. 11

1 AESHNA UMBROSA OCCIDENTALIS ¢

2 y exo SCTRI( Ad

3 yy Wy HOMCEQCHROMATI(
4 yy rT

2

wand Rs!
: wai

“an 4, 7
; £
Ss eS de a

i - [<a s
ey ro

UNIV. OF TORONTO STUDIES, BIOLOGICAL SERIES No, 11 PLATE 28

E. M. W. DEL. HELIOTYPE CO. BOSTON

1 AESHNA CALIFORNICA & 4 AESHNA MULTICOLOR ¢&
2 yy p) 2 5 ) ” °)

3 AESHNA MUTATA &

“i

&

ee
”) f sve '
LOY

ay * 6 5 ;
: f / Na

Laman

mt ALN sl
1 alle a had h Ai we

ae ae bir me oun ew
ah Pee YAP i iy yi

i , Ca 7 - peat ; Rai iif vee be oY

way ; chy i
f 32 Cite” | STS GN
‘2 y
wo 4

iy
Ma,
h

’
‘

RINDING ''47 cf
fone G * | bee i | yw

mi University of Toronto. Biological
T68 series
no, 10-11

Biological]
& Medica]

Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

pepe? 20808 We Om nate egw imme onl

: raSosmenten ea ie se Te arcsec Peg ae apis pt a i Ba ee a lai em ania 3. Sed gh Oa ear anard ba LN CEL eae na

en,
ri hed is er Gukepes Seta tsb tH ah th Ohno) Se 3 =~ Set a a ene a cae Ere, (ores “Se Sores ~- o Cs
ytrbetets sete be beteme as riers se teense tea tNe ROSE cack pees SOUT siy Wea. Fn P91 bee Be STIS keg er ti A TN da fons eee a Nh oe ~ 2 A OAD ORO LEIBA LI OOO rg ay FERRARA EP RMD SO A A ARE

ee es
et Reh a RA Oe ey te OME EAE A th Se
Ae Re A LO

itor

eee

<r ee

a he

RE eae eta helt Satine an oat in OE Oa IO
~ — amet ey eannenme é 2 renee -- . - ee rear ee ar ene a TE ena
LAA Ae LOE PRES SOIE ON II ET OE a a aetna mt OS PE Ariana nnge ee EEE aT A

ate:
ae

wet Soh Se
rete eh
> i aahaeed

ae 4 renege et
me ezacsbone eye Wingactss bene aoe
: . sbtohetetotettiersh

Ceogerorer

’
-
i eegi-ehaher OT Ta Leakelas chahaneneaeeetst eee e
* -