Article I.

—

The Dragon-Jlips ((Moiiafa) of lUinois. iriffi

Descriptions; of the Iiininifhit Stages. Pint 1. Petaliiritlie, .Kscli-

ni(he, and (ioiiipJiiihe. By James G. Needham and Charles A.

Hart.

Introduction.

The present paper is the combined product of studies made
by the authors upon the collections of the Illinois State Labo-

ratory of Natural History, including those of the Biological

Station at Havana, and by Professor Needham on his private

collection. The state collections comprise the results of many
years' w^ork in all parts of Illinois, including five successive

years of field observation and collection at the Illinois Biolog-

ical Station on the Illinois Eiver at Havana ; specimens ol)-

tained by Professor H. Garman from the Mississippi and related

waters at Quincy, 111. (Garman, '90); a large series from the

small lakes of northeastern Illinois ; and collections made by

Professor Forbes in Yellowstone Park and its vicinity (Forbes,

'93) and from lakes Geneva and Delavan, Wis., under the au-

spices of the IT. S. Fish Commission. Professor Needhaui's pri-

vate collection used in the preparation of this paper contains

many Illinois and New York species, reared by him to the

imago, and others contributed by correspondents, particularly

by Mr. F. G. Schaupp in Texas, Mr. Adolph Hempel in Florida,

and Mr. F. C. Willard in Arizona. The preparation of the in-

troductory matter, biological discussions, and keys to the

nymphs has devolved upon Mr. Hart, while the descriptive

matter, with a few exceptions, was prepared by Professor

Needham.
We are indebted to Mr. C. C. Adams for abundant data on

the occuiTence of the imagos in Illinois derived from his large



private collection, and also for work on the keys to the genera

of ima^os. We are under special obligation to Dr. P. P. Calvert,

of Philadelphia, for aid given Mr. Adams in the revision of

keys and in the determination of doubtful species, and for notes

on dragon-flies in Illinois made by Mr. Harry Walker at Belvi-

dere, Mr. Shafer at Mt. Pulaski, and Mr. E. J. Kuegeman at

Ravenswood and Edgewater. Data relating to collections in

Illinois have been furnished us by Mr. Maurice Ricker, of Bur-

lington, la., and by W. E. Longley, and others, of the Entomo-
logical Society of Chicago.

Of the twenty-eight recognized Illinois species of the fami-

lies of Anisoptera herein treated, we have here described the

nymphs of twenty-four (six of them for the first time), repre-

senting all our eleven genera. To these have been added by

Professor Needham descriptions of ten nymphs of extralimital

species. Thirty-four nymphs are thus described in all, fourteen

of them for the first time.

Much careful study has been given to the preparation of

the keys. In all cases the linear arrangement is according to

the principles suggested by Comstock, the more generalized

group or species being followed by the divergent ones in the

order of the direction and amount of specialization. The de-

scriptions of the nymphs are drawn up from full-grown exam-

ples unless otherwise stated. The newly hatched insect quickly

takes on the form and structure of the full-grown nymph, and

may usually be recognized while still very young by the char-

acteristic sculpture and armament of its species.



ORDER ODONATA.
The common dragon-flies of the suborder Anisoptent are

familiar to every one, but the damsel-flies, constituting the sub-

order Zi/ffoptera. might not be recognized as also belonging to

this order. These damsel-flies are small narrow-winged forms,

which, like the butterflies, hold their wings back to back while

at rest. The Odonntd have no quiescent i^upal state; the

immature stages after the egg are collectively designated as

the nymph. The latter is always aquatic. It has highly

developed thoracic legs but no abdominal ones. Wing-pads

appear at the third or fourth molt. There are no external gill

structures except the three terminal appendages of the Zi/(/(ip-

tera. The abdomen of the nymph is slender among the

Zjiijoptera, but in the ordinary dragon-flies [A)ii.sopferii) it is

rather short and broad. A very distinctive feature is the large

and elongate labium, folded beneath the body like an arm, the

"hand" of which, ending in a pair of claspers, covers the

mouth or the entire face. The nymphs crawl rather slowly,

often clumsily, but can dart some distance through the water

like the crawfish, being propelled by the sudden ejection of the

water in the rectal respiratory cavity.

The Odonatii and their near relatives the May-flies

[Epheinerlda), isolated remnants of former insect life, in gen-

eral of primitive character although highly specialized along

some lines, are probably among the oldest orders of winged

insects. In younger groups the branching of the " family tree"

of development may often be traced with some degree of satis-

faction by a study of the primitive characters retained hj still

existing forms ; but in the lapse of ages so many of the earlier

lines of descent have been obliterated from the earth that in

the Odonafa only the upper parts of a few separated branches

are traceable, their points of origin being involved in obscurity.

In Illinois the branch nearest the primitive stock is probably

that of the "black-wings," or Calopferyx, of our smaller streams,

representing the CaIopferi/i/id(e. Another branch [Ayrionid(i'),

including the more common damsel-flies, Aijrion, Lcsfes, etc.,



4

has a number of characters in common with the preced-

ing, and the two constitute the suborder Ziiyoptera. In

forms related to Calopferijx the three tracheated caudal setae of

the May-flies seem to be represented by three shorter and
thicker respiratory appendages, the laterals especially being

more or less spike-like and triquetral. In the Agriouiche these

appendages are vertically flattened thin gill-plates. Quite a

different development of the original type is found in the

nymphs of the ordinary dragon-flies, the suborder Anisoptera.

In these respiration is principally performed by internal rectal

gills, and the three appendages, very short and subtriangular,

are supplemented by two others, the flve converging to protect

the opening of the respiratory cavity.

The five families constituting the Anisoptera—Petaluridu',

JEschnidce, GompJndm, Corditlegasteridce, and LiheUulidce—are

variously connected by common characters. It is worth not-

ing that the obscure color, rough sculpture, and general habits

of the Calopteri/x nymph ai'e also found in the more primitive

forms of these families. The Petaluridce is probably the

oldest of ths five. It has not as yet been found in Illinois.

In this family and the j-Esclinid(e the ovipositor is well devel-

oped and oviposition is endophytic, as in all the Zijgopiera.

The Petaluridce, judging from Mr. Williamson's observations

(Williamson, '01), breed in boggy or swampy spots, where
the imagos may be found ovipositing. The nymph is re-

markably synthetic in its characters and supplies a hitherto

missing link in the evolution of the labium, being interme-

diate in this respect between ^Eschnidce and Conhilegaster-

id(e, thus indicating a very primitive' origin. In its form and
antennal structure, on the other hand, it strongly approaches

the GoinpJiid(e. In general, however, it is clearly most closely

related to the JEschnidw. The nymphs of the jEsclniidce

are somewhat elongate, and climb on submerged branches,

driftwood, roots, and smaller vegetation. The imagos are

unusually large and high-flying, the eyes large, the wings

much specialized. The nymphs of the third family, Gotu-



phidce, nearly all IniiTOW in mad and sand. The images are

usually of medium size, barred with green or yellow, the

apical part of the abdomen often expanded to a greater or less

extent. They are usually seen near water, but are not numer-

ous as a rule. The ovipositor is rudimentary or wanting and

ovipositiou is exophytic, as in the remaining families. The
eyes are small and widely separated, in which respect the fam-

ily approaches the Zi/f/djifcrd. The nymphs of the .fjsclnnda'

and Goiiqt/iidd' differ but slightly in the structure of the labium

and gizzard, and both these organs are quite similar to the

type found in the Zi/i/iqjfcriL The next family, Cordidegaster-

i(hi\ is of special interest. It lias ])ut few species, and examples

are rarely seen in Illinois. While evidently related to the pre-

ceding families in diiferent ways, the nymphs are like those of

the Lihcl/iilidfr in having a spoon-shaped labium and a sym-
metrically 4-toothed larval gizzard. The last family, Libelln-

lidce, comprises most of the commoner species. They are

diversified in structure and habits, and may be grouped in three

subfamilies. The first two of these ( S'//////u'/;///»r/' and Cordii-

/iiiKf) have a number of common features and have hitherto

been classed as one, Init the wing structure and nymphal
characters of the Sijidliviniime, represented in Illinois by Ma-
croinia and Didipnops, seem to warrant its separation from the

Cord III iiiKt'. The greater part of the family belongs to the Li-

belln/uKt', which includes three general types, exemplified by

Si/nqjetridii (Dipldx ), Lib(dli(l(i, and Traincfi. The grouping of

these is based by Mr. Adams on the extent of development of

the ovipositor.

LITERATURE.

A list of the more important papers, to which references

are made, is appended to this paper. A brief summary of their

contents may be given here.

On the nymphs, the most important papers are by Hagen
and Cabot, describing and figuring the nymphs of the Museum
of Comparative Zoology, at Cambridge. Cabot's first paper [ '72

)



covered the Goniphiche and Cordulegasfer/ihe, his second ('81)

the jEschuida', and his third ('90) the Synthemiina' and Cor-

dit/iinie and also Tramea and Pantala. Hagen ('85) gave a greatly

extended revision of the two families previously treated by

Cabot ('72). The large and abundant family LihelUilidw was
not treated by them. Karsch ('98) gives descriptions, figures, and
keys to West African nymphs. The characters assigned to the

major groups in his keys are, as Dr. Calvert has pointed out,

only locally or partially true. Lucas ('97, '97a) has described

and figured a few English nymphs, and Roster ('86, '88), a num-
ber of Italian Zijgoptera. Beutenmiiller ('90c) has given a

bibliography of nymphal descriptions, excepting those in Cabot's

'90. Needham ('97a, '99) gives suggestions ou collecting and
rearing nymphs.

Among the general works, Dr. P. P. Calvert's admirable

paper ('93) deserves first mention. It includes a complete and
accurate summary of all the more important biological and
anatomical details concerning dragon-flies and their nymphs,
with generic keys for both. Descriptions are given of the

species of the region about Philadelphia, among which are

most of our commoner forms. The phylogeny of the dragon-

flies in general, with especial,reference to the structure of the

gizzard, is discussed in a scholarly article by Dr. F. Ris ('97).

The principal monograph of American species is in Hagen's

well known " Neuroptera of North America" ('61). A num-
ber of monographic papers have been published by Dr. Selys-

Longschamps in Belgium. The best catalogue of species is

that of Kirby ('90). It contains several surprising changes in

nomenclature, one of which—the use of ^Eshiia in place of

Gomphas—is fortunately not well founded. Banks ('92) has

listed the American species, and given a key to the genera.

From an economic point of view, especially as to their utility

as destroyers of obnoxious Dipfeni, dragon-flies are treated in

Lamborn's collection of prize essays ('90).

The numerous published lists of Odonata in this country

contain much valuable information as to the life history of our



species, dates of occun-euce, and the like, and have been exten-

sively used in the preparation of this paper. Such are those of

the late Dr. Kellicott ("95. "96. "aVa) : of Miss Wadsworth ("90)

and Mr. Harvey (91a) for Maine ; of Calvert, Banks, and Van
Duzee for New York ; of Calvert for Virginia and the vicinity of

Philadelphia : of Williamson ( '00 ) for Indiana : and of Elrod for

Iowa, and Banks for Kansas. The above localities are not far

from our own latitude, and the data given ai'e presumably quite

close to the normal for Illinois.

Among Illinois contributions to this subject we ma-y men-
tion the elaborate papers ("(i"2. '(VS) of our pioneer entomologist.

Mr. B. D. Walsh, who listed sixty species occurring about Rock
Island, describing a number as new : a li.^t of dragon-flies

taken at Kensington. 111., by Mr. Jas. E. McDade (.'92) : and the

treatment of the OdotK/ta in Mr. H. Garman's "Animals of the

Waters of the Mississippi Bottoms near Quincy. 111." ('90). Eight

nymphs are listed in the latter paper. The original specimens

have been examined and determined as follows: His "4" is

Gomphus externus; "8. Epiceschna heros" is NasUfschita penia-

cauthii of this paper: "9" is Trained laccrata: "10", EpicorduUa

princeps: and "11". Perithemis domitia. "'12" is immature and
not positively determined as yet. The others are con-ectly

identified by him.

LIFE HISTORY.

Dragon-flies hibernate commonly as nymphs, less com-
monly in the egg stage, and in one species as imago. Xymphs
of the A/ilsoptera mostly transfoi-m to the imago in the early

part of the season, especially from the latter part of May to the

flrst of July in our latitude, the images scattering and becom-
ing reduced in numbers by midsummer. Some species, particu-

larly those of the (TO)iiplii(l(i', have a very short period of emer-

gence, large numbers transforming within a few days. Anux.
and perhaps some species of yEsclma, has apparently two broods

in a year. The adults appear very early in the season, and
the young nymphs, feeding voraciously and actively, grow rap-



8

idly, and have been known to produce imagos before the close

of the same season. Tnniipo also seems to be two-brooded,

emerging numerously both in spring and fall. On the other

hand, the nymphs of the Goniphkhe probably require more
than a year in which to mature, emerging in the second or

third season after hatching. The one-year life cycle is, how-

ever, the rule among the LibeUiiIidxe, which contains most of

our commoner Aiiisopto'a, the eggs laid during midsummer
hatching in late summer, and thenymphsmaturingintimeforthe
general emergence the following season. In late August there

is a surprising number of very small nymphs,—tiny spider-like

youngsters,—and even in September and October the prepon-

derance of young nymphs is still manifest. As the period of

maximum emergence of their species approaches, usually in early

summer, they mostly attain full size, and are at this time most

readily seen and captured and apparently more abundant than

in the fall, when they were small and easily overlooked. With
regard to the Zygoptera, it is highly probable that there are a

number of broods in a season, the processes of transformation and

oviposition beginning as soon as the weather permits and con-

tinuing industriously to the close of the season.

In the species whose life cycle is apparently more than a

year the nymphs are of two or three distinct sizes, the largest

presumably being of the next brood to emerge. In species

whose life cycle is completed within a year, the nymphs are

fairly uniform in size, but there is, nevertheless, a sufficient

extent of variation to cause a considerable number of strag-

gling emergences during the season ; and it therefore follows

that, in general, nymphs of nearly all species of Odoiiafii may
be -found throughout the entire season. Because of this fact it

has not seemed worth while to give dates of the occurrence of

nymphs. The accessible data concerning the imago period, on

the other hand, is very fully given, thus indicating the limits

of the nymphal period also.

Like many other immature aquatic forms, the grbund color

of the nymphs darkens greatly up to the close of each molting



9

period. Immediately after molting they are very light green-

ish or grayish and their characteristic color-pattern is beauti-

fully distinct, but they gradually darken and the coloration

becomes more and more obscure until, as the time for the next

molt approaches, it is almost entirely lost and the nymph be-

comes uniformly dark and dingy.

When grown, the nymphs seek the shore or some floating

object and clamber up a little way on standing vegetation, logs,

tree-trunks, sticks, bridge-piling, the sides of boats, or the like,

and, fixing their feet firmly, proceed to transform to the im-

ago stage. Transformation mostly takes place very early in the

day and is largely over with by nine o'clock, although scatter-

ing emergences may occur at any time. The usual process will be

fully described under Traniea Jacer<ita. The adults scatter con-

siderably, but a large number remain in the original vicinity,

busily ovipositing for a new brood of nymphs. A short but

undetermined period elapses liefore egg-laying begins.

Oviposition is of two kinds, endophytic and exophytic.

That of the groups with more slender nymphs, Zti(joptera and
y£.srluiid(i\ is endophytic. They have an elongated egg, which

is inserted by means of an ovipositor into living or dead vege-

table substances, either resting in water or at least moist.

The female immerses the tip of her abdomen or enters the

water completely. She usually succeeds in escaping safely

from it. but is sometimes rescued by the male (Todd, '85). The
oviposition of the groups with broader nymphs, Gomphidw,

ConJiiJegaster'nhv, and Lib(dlii/i(la', is exophytic. Their eggs

are shorter and oval, and are extruded in a gelatinous matrix.

The female dips her abdomen in the water, usually during

flight, releasing at each dip a numl)er of eggs, which sink to

the liottom or lodge on the vegetation. Sometimes, when too

hotly pursued by males, she will alight on water moss or drift-

wood and cast her eggs loose there. In the case of LeiicorJiliila

this is apparently the usual method of oviposition. In some

cases the eggs may be deposited on moist mud (Diphi.r. Soiiiat-

och/ura) or affixed to the bank or to water plants. The female



10

of Cel'dhemis is usually accompanied by the male, who helps

her to escape when menaced by the open mouth of a hungry

fish. Several hundred eggs are often laid by a single female.

The period of incubation varies with the season and also

independently of it. In midsummer, eggs of some species hatch

in from six to ten days, while others, laid in autumn, do not

hatch until the following spring. In the same lot of eggs the

period of incubation may vary, even in midsummer, from a

week to more than a month.

The apparent abundance of nymph and imago is far from

corresponding, the difference in some cases being quite surpris-

ing. Of the nymph of Celithemis eponina, we have secured only

a few examples; yet the imago is a familiar sight everywhere

about the Station. The species of Syiiipefniin are common and

familiar dragon-flies; but we have obtained only a few of the

nymphs. This genus probably breeds in swampy places, where
the vegetation is so dense, the water so shallow, and the mud
so deep as to make collecting very difficult. On the other hand,

Epicordulia princeps is abundant and widely distributed as a

nymph, but the imago is not commonly taken. Nymphs of

Macroiiiia, Proyomphiis, and Hafjenins are not at all rare in

sti'eams ; the imagos are considered very rare or almost

unobtainable. In the Gomphid/e are numerous similar exam-
ples. This discrepancy may be due either to the swift, high, or

prolonged flight of the imago, or to the shortness of its life
;

but in some cases it is almost inconceivable how the imagos

can vanish so completely as they do.

Walsh ('63, p. 239) makes some remarkable statements as

to the relative proportion of the sexes in Go inph us. In some
species he found four males to every female, and in another

two or three females to each male, and he asserts that this is

the case in freshly emerged material. Mr. Needham is of the

opinion that in nymphs generally there is no notable excess of

either sex, but that in the imago an excess of males may occur

because of the destruction of the females by fishes in species the

females of which oviposit unattended by the males. It is prob-



11

able that Walsh's statements were based ou insufficient data.

The imagos usually keep near their nymphal home. The
Zjiijoptera both fly and alight low, usually over water; the

stronger ^Ksclnii<l(f fly high and scatter more; while the other

imagos are diversified in these respects, many species being

taken both along streams and along borders of roads and fields

at some distance from water.

All dragon-flies are most active in hot, quiet, sunshiny

weather. They cease flying by sunset, except a few of the

^schnidd', which may 1)6 seen in hot pursuit of small Dipiera

during the evening twilight.

The history of the order as a whole in the vicinity of

Havana, while under our observation, is interesting but some-

what puzzling. In 1894 there was a great abundance of indi-

viduals, m all stages. In 1895 the season's collecting opened up

well. Extremely low water prevailed, however, and in June the

shallow waters became greatly heated by the sun, their recorded

temperatures reaching as high as 9(r Fahr. On one occasion,

when the heat of the water was especially noticeable, a num-
ber of dead nymphs were found floating about, along the west

shore of Quiver Lake. Not many nymphs were seen in the usual

situations at the close of this season, and from that time on

through 1S96, they were reduced to insignificant numbers. The

low water of 1895, however, encouraged the development of spe-

cies liking swampy situatious, and there was consequently in

1896 an excessive development of I'aclujdipldr luiiijipciuil.s and

species of Si/iHpctnnn. A slight rise took place in the spring of

1896 and the river became very foul. At the close of the year the

drouth w^as broken, with a more extensive rise of the river, and

the season of 1897 showed a-n increase in the number of nymphs.

The winter and spring following brought much high water in

the river, but the nymphs did not greatly increase in numljers.

Kellicott has noted ('97a ; 97, p. 69) that in Ohio the warm
rainy weather of April, 1896, following the drouth of 1894-95,

caused emergence earlier than usual, and the number appear-

ing did not seem noticeal)ly reduced by the drouth.



12

WATERS FROM WHICH COLLECTIONS WERE MADE.

The nymphs in the State Laboratory collections were de-

rived from a large variety of waters : from the broad Mississippi

and Ohio, slow and majestic, but flowing more hurriedly at a

few points—as at Rock Island and Golconda ; from the bottom-

land lakes of these streams, muddy and without much vegeta-

tion ; from the lower course of the Illinois, broad, shallow, slow-

flowing, with the bottoms more of mud than sand, the shores

often margined with water vegetation ; from the broad and
shallow bottom-laud lakes along this river, some slightly deeper

than others, but abounding in floating vegetation,—as Quiver,

Thompson's, and Dogfish lakes near Havana,—some shallower

and swampy, as the great expanse of Flag Lake, and others

changeable and temporary in character, as Phelps Lake ; from
the smaller affluents of the rivers, usually flowing with an oc-

casional descent over beds of mud, sand, and gravel ; from the

small lakes which rest in the ancient glacial hollows to the west

of Lake Michigan in Illinois and Wisconsin, quite deep in places,

permanent in character, usually well supplied with aquatic life

of all kinds ; from the shallow ponds of similar origin scattered

over the state, of all degrees of permanency ; and, lastly, from
the rapid flowing rocky streams of Yellowstone Park. These

waters present a great variety of situations, each with its char-

acteristic forms, and it is interesting to note how each of these

situations is occupied by a definite series of nymphs, in accord-

ance with its particular character.

NYMPHS INHABITING THE VARIOUS SITUATIONS.

The nymphs may be roughly grouped in three divisions,

according to the kind of situation preferred by them. The first

includes groups having the endophytic habit of oviposition,

Zygopfera and y£schnid(e, as well as a few of the more agile

members of the other families, such as Hagenius among the

Gomjjhidif, and Mesothemis, Celifhemis, and Tramea among the

Lihellulkke. The lighter-colored of these clamber among sub-

merged vegetation, while the dark-colored forms {Calopteryg-



13

i(1a\ Uaf/oiiits, and the lower ^^.sr//»/r/r/') inhabit driftwood and
submerged roots and brush. The second division comprises the

heavier nymphs of the Llhc/i/i/ii/(t\ which usually sprawl upon
the bottom or climb over fallen ru])bish ; and the third con-

sists of the remaining Goinpliidfr, the ('or(lii/e(/((sfcr/(I(i', and,

to some extent, LihcJJida and its relatives, which occupy the

mud or sand of the bottom.

The species inhabiting submerged vegetation, constituting

the first group, are of course most abundant in the quiet waters

of slow streams, lakes, and ponds, where such growths have an
opportunity to develop freely. The large and active nymphs
of Trained and Putifitla appear to prefer the more exposed

shores of large lakes and rivers. Mesnthnnis rol/orafa is

adapted for life in the high temperatures of water from the hot

springs of the west. The A(/rioiu(l(i' and Aiia.r are common
everywhere in bodies of water with vegetation, from the small-

est pool to considerable lakes. The other yEschnida' climb on
dead sticks and driftwood where there is little vegetation and
food is not abundant, as in the bottom-land lakes and sloughs

of the larger streams with good current, or [Boi/eria) in the

deeper holes along small streams, where projecting roots catch

the floating driftwood. Tfar/ciiiiis also lives amongst driftwood

and dead leaves in streams. The ('ahipfcrj/f/idd' thrive in little

rapid-flowing sandy rivulets like White Oak Run and Quiver

Creek at Havana.

In the second group, those living on the liottom, depth and
current are important in determining the particular species

likely to be found. EpiconhiHn likes the deeper waters, as of

the Illinois River itself or its deeper lakes, or the bottoms of

the small lakes of northeastern Illinois, finding there the

molluscan diet it enjoys. On the other hand Si/nqxinnii and
Leucorhinia inhabit marshes and reedy shores, as well as shal-

low weedy ponds and wet meadows. The " Pumpkin Patch
''

at Havana, the shores of Fourth Lake in northeastern Illinois,

and the swampy spots of the upland prairie, are situations of

this character. Paclnjdipla.r appears to like the deeper swamps



14

and ponds, such as Flag Lake at Havana. In the smaller

streams, where tlie current is often rapid, such as the Mackinaw,

the upper Sangamon, and the Little Wabash rivers, a number
of very interesting forms find their home, especially the

Synihemiince, Somatochlora and other Corduliime, and the species

related to Lihelbila.

Of the third group, which live in the mud or sand of the

bottom, there may be found in the streams just enumerated

other interesting material, especially Cordulegaster, Progoniphus

and Gomphus spiniccps. Li this group the most definite selec-

tion of situations prevails. Where rapid currents flow over

rocky beds, as in the Mississippi at Rock Island, in the Mack-

inaw at the " Dells," and in the streams of Yellowstone Park,

is the chosen home of Diastatomma {Ophlogoniphiis). Where
slow currents and a muddy bottom occur, as in many prairie

streams and ditches, PlatheniLs and Lihelhda, covered with

dirt, trail over the mud along the shores leaving a well-marked

track, or lie concealed amongst fallen trash, the dorsally pro-

jecting eyes of Plathemis enabling it to remain hidden and yet

keep a sharp lookout. The nymphs of the Cordulegasteridee.

conceal themselves in the sand beneath the eddies of streams

or under the alluvial deposit in marshes, but do not burrow.

Each scratches a hole for itself, and descends into it like a

chicken into a dust bath, and, like the chicken, kicks the sand

over its back, and does not rest until almost hidden, only the

tops of its eyes, the tip of its treacherous labium, and the re-

spiratory aperture at the end of its abdomen reaching the sur-

face. The burrowers of the order are the members of the large

and as a whole abundant genus Goniphus. These burrow shal-

lowly along in the midst of abundant life, with the tip of the

abdomen turned up for respiration. Some are lake ov pond
species, such as gyaslinelliis and pallidus. The former is com-

mon in prairie ponds and in the small lakes near Chicago ; the

latter in similar localities, and in open lakes near Havana, such

as Matanzas and Clear lakes. Others choose the larger rivers,

being most numerous in quiet corners where fine rubbish and



15

animal life accumulate. Such are Hotains aud frdlcniKs;

while spinicpjjs prefers swifter flowing waters like Quiver Creek
at Havana, or the upper Illinois Kiver at Ottawa.

In some species with which we have had but a limited

experience further study maj" show a wider range of normal
situations than that here assigned, but the preceding generali-

zations should have a value to the student and collector, and
may serve as a basis for a more exact account in future.

The Illinois waters which afford the widest range of situa-

tions and are most prolific in variety of dragon-fly nymphs are

the small lakes of Lake county. Their shores are sometimes
exposed, wave-washed, and rocky, and the life here found is

remarkably like that of swift running water. Again, they are

broad and marshy, and afford a home for S if inpetnun. Out
from shore we may find shallow waters with mats of aquatic

vegetation, bare sandy surfaces, or, at great depths, a bottom
of soft mud. As to numerical abundance I have seen nothing
in Illinois that can compare with the multitudes of dragon-flies

that issue in favorable seasons from the broad shallow lakes

along the Illinois River.

ODONATA OF THE VARIOUS WATERS IN ILLINOIS.

Taking the subject from a different point of view, we may
briefly summarize the odonate life of the Illinois waters as fol-

lows.

In the larger rivers, down to the size of the Mackinaw, in

places where the water flows with considerable current over a

rocky bottom, Diasfafoiinixi may l)e looked for ; where mud
or sand bottom and quieter waters prevail, Epifordulld and
some species of Guniplms may be found. Other species of

Gomphus occur in the bare muddy or sandy bottoms of the

sloughs and bottom-land lakes. In tree-shaded waters, where
driftwood and branches have gathered, or along muddy margins,

especially amongst exposed roots, the lower .-Esrhniihr may be

looked for. In bottom-land lakes where vegetation is abundant,

one may find Aiiax, Agrionithe, Mesoflionis, ('c/iflwiins, Tramea,



16

and PantaJa amongst the vegetation, the latter two especially

on more exposed shores; and Tetragoneuria, Lihclhihi, Epironhi-

lia, and Lencorhinia on the bottom underneath. If the situation

is inclined to be marshy, Pachydiplnx, Feriihemis, and Cell-

fheniis will be scattered over the bottom ; and the shallowest

and most temporary waters or wet lands are the especial home
of Sympetnuu.

In the smaller and quicker flowing streams, like the upper

Mackinaw and Sangamon, quite a different series occurs: Ha-

genius, clinging to stones and driftwood and amongst dead

leaves; Boi/cria and other dark Aisrknidce on submerged

branches, roots, and sticks ; Cordulcgaster and the long-legged

Macromia hidden at the bottom in sheltered eddies ; Somafo-

chlora ; and, finally, Pror/omphiis, DroitiogompJiiis, and certain

species of Gomphus burrowing in the sandy bottom. In the

prairie ponds and slow streams and ditches, Anax, Agrioniche,

and Mesofhemis and other Lihellulidce occur amongst vegetation,

and Sijmpefiuon in shallower parts, while Libelhtia and PUithe-

mis will be found where there is more mud and less vegetation,

as in ditches and tile ponds, resting at the lower ends of well-

defined tracks. In streams of rapid flow, but not especially

rocky or shaded, the Ca/opferggirhp are most likely to be found,

the imagos fluttering along the banks.

In the small northern glacial lakes of Illinois, a remarka-

bly varied odonate life can be found. Among the vegetation

and along shore, Agrionidfe, Anax, Trainea, LeiicorJiiiiia, Li-

beUida, Tetragoneuria, Epicordnlia, and Baskeschna may be taken.

Sympetrum has been found along the broader, shallow, reedy

margins. On the sparsely grown flats, in somewhat deeper

water, Gomphus graslinelhis, spicafas, and pallidas occur, and,

especially where it is clayey, the Synthemiince—Didymops and

Macromia—have been taken. On the other hand, Mesofhemis,

Pachydiplax, and Per ifhem is, elsewhere abundant, do not appear

in our collections from these waters.

As to the imagos, they are most likely to be found along

the shores of waters inhabited by their nymphs, though many



17

forms, such as some Agrim/idd', most of the yEsclniiihr, Lihrl-

lula, S11mpetmm, and Mrsotlinitls scatter widely inland. Sipn-

pctnoii will be seen about fields and lawns, and the ALxrlmUJif

in the vicinity of houses.

FOOD RELATIONS.

The nymphs are all predatory in habit. Most species

remain in ambush, aided by coverings iii sand, mud, silt, and
algal growths, and by their own protective coloring, until their

prey wanders within reach. Atxtx jmi ins and a few others

choose their prey. All capture it with a marvellously sudden
extension of the labium, bringing it into the grasp of the for-

midable lateral labial lobes. Ahnost all kinds of small aquatic

animals appear on the bill of fare of the group as a whole.

The Aijrioiildif have a seeming preference for Entoiiiostnirii

and May-fly nymphs. The vegetation-inhabiting species have
the most varied diet, including especially back-swimmers
{Notoiu'cfa) and water-boatmen (f'w/v'.sy/), small crustaceans, such

as Asellus and AUofchestes, thin-shelled mollusks, like Pln/sa,

coleopterous and dipterous larvae, and even the younger or

weaker members of their own order. Aiia.r takes even the

thicker-shelled univalves, like Aiiiniroln. The deep-water Kpi-

ronhi/id feeds principally on small mollusks, such as Amnirolo
and I'lii/sfi, as well as on other life of the l)ottom. The ^Esc/i-

)ii(hr, especially Aiuix, are most omnivorous creatures. The
larger odonate nymphs eat very young fish, and in some cases

appear to have caused a sweeping destruction of large numbers
of them.

On the other hand, the nymphs are apparently eaten prin-

cipally by fishes and by one another ; hence their need for hid

ing places in mud and sand or among matted vegetation,

the course of Professor Forbes's studies of the food of

('8ya, pp. 485, 524) he found odonate nymphs most p'

(twenty-five per cent.) in the food of the grass picl-

veniiinildfui^), and forming ten to thirteen per cent

of the crappie {Pomoxis anHularis), the pirate npv-



18
•

sai/aniis), and the common perch {Perca fiavescens). The latter is

anabundant species in the small northeastern lakes of the state.

Belosioma, Ranafra, Nofonecta, and the like, according to

Mrs. Aaron (Lamborn, '90, p. 50), prey on young nymphs, and

she also mentions a small red mite and a minute dipteron as

parasitic on the eggs of Odonaia. Mr. Needham ('98) has

recorded the finding of numbers of nymphs in the stomachs of

herons, and he once found the intestine of a nymph parasitized

by very large Gregarin'nhe fully 1 mm. long.

Emerging imagos while limp and pale ai'e the easy prey of

even their weakest enemies, and great decimation takes place

among them at this time. Ants, spidei's, robber-flies, frogs, and

birds eat many of them.

The full-fledged dragon-flies devour the small insect life of

the air in vast quantities, especially gnats and mosquitoes, most

of which probably developed in the waters previously inhabited

by the dragon-fly nymphs. These seem to make little or no

attack upon the larval stage of the gnats and mosquitoes, as if

reserving them for their imago life. Dr. Lamborn, by offers of

prizes, started an investigation to determine the practicability

of artificially utilizing dragon-flies for the destruction of mos-

quitoes and flies; but the evidence published (Lamborn '90)

indicates that these pests may be dealt with more effectively

by direct measures, since the the attack of dragon-flies could

not easily be controlled and directed, although in their own way
and time they doubtless reduce the numbers of the pests very

considerably.

Mature dragon-flies do not usually suiSer seriously from
natural enemies. They are frequently found infested by small

^ed mites. Some birds, such as the king-bird, possess sufficient

Hness to capture them as regular articles of their food. It

to be the weaker forms, such as Agyion, that suflier thus

extensively (Needham '98). The females while oviposit-

netimes captured by fishes.

COLLECTING AND REARING.

"'^'"''^ration of the statements and suggestions



19

in this article as to the various haunts and seasons of the dif-

ferent nymphs and imagos. will give one a pretty good idea what
species may be found in the waters near at hand. Nymphs of

species inliabiting vegetation may be secured with a dip-net or

rake. The dip-net is quite an essential article. The kind most
in use at the Biological Station has a D-shaped ring made of

heavy \^^re about a quarter of an inch thick, the two ends joined

at the middle of the curve, the last three or four inches of the

ends bent outward, welded together, and inserted into the

handle, as in a hoe or rake. The net is of bobbinet, with a

mesh about like that of mosquito-netting, or finer. It is cut about

four or five inches larger than the ring all around and gathered

up to form a shallow bag without seams. This bag is edged

with a band of strong cloth slightly larger than the ring and
then sewed by this edge to the ring with good fine twine.

The nymphs sprawling on the I)ottom are secured by vig-

orous movements of the net close to the bottom, the cun-ents

produced sweeping them into the net. The liurrowers are taken

by scraping the surface layer of the liottom into the net or into

a sieve, and then sifting or washing out the mud or sand, thus

leaving only the nymphs and coarser rubbish. In deep-water

collecting a dredge must lie used. The swamp-inhabiting

nymphs are easiest secured in the early morning as they come
up for transformation. A pair of rubber boots is often very

useful.

The smaller imagos are easily taken, Init the larger ones

will often stimulate the activity and alertness of the collector

to a considerable degi'ee. If the air net be of light, strong, clean

material, like bolting cloth, not too fine-meshed, the possibili-

ties of capture will be gi'eatly increased. The best time to col-

lect dragon-fiies is on cool, cloudy, or windy days. On one

memorable occasion on the shore of Cedar Lake, just after a

light rain, a heavier storm impending, the dragon-fiies were

found resting on weeds among the trees on the sloping shore,

so sluggish that they were picked up by hand in large numbers.

Eggs of the families which oviposit fi"ee in the water are



20

easily secured by capturing the ovipositing female and touch-

ing her abdomen repeatedly to some water in a small dish,

holding her only by the fore wings back to back, while those

which oviposit in plants or soft wood may be watched and the

stems or wood examined afterward. Experiments indicate that

imagos will not voluntarily oviposit while in captivity.

Formalin is not a good preservative for any of the larger

insects. The best preservative is strong alcohol carefully

heated in a water-bath. The hot alcohol penetrates more rap-

idly than cold, arresting internal decomposition, and thus re-

tains the beautiful but fugitive colors of the imago. The usual

process of sterilizing and fixing the tissues by heating in water

in a test-tube to the boiling point before transferring to alcohol

is eminently satisfactory for aquatic forms as a rule, but in the

Anisoptera this expands the air in the rectal gill-chamber and

distorts the abdomen somewhat, while in the Ayrionidre, as in

the Epheiiieridiv, the Hat external gill-plates are badly injured

by inflation and gumming together. The slender and brittle

abdomen of the imago breaks off very easily, and a bristle or

fine non-corrosive wii'e should therefore be passed lengthwise

through the Ijody as far as the tip of the abdomen, but not so far

as to project among the terminal appendages. A couple of insect

pins, inserted lengthwise, one at each end of the body, are

used by some. Specimens for the cabinet may be spread like

Lepidoptcra. For shipment or exchange they are usually in-

closed in soft papers folded diagonally.

Rearing the nymphs is not usually difficult. They need

plenty of clean water, something to crawl out on, and room to

transform in. A pail or tub covered with mosquito-netting

answers nicely. It must get plenty of sunshine, but not so much
as to overheat the water. If the breeding-cage can be immersed
in the water of the stream or lake where the nymphs live, suc-

cess is almost assured. They may be fed bits of fresh meat or

fish, insect larvse, flies, or the smaller aquatic Hemrptera. If

meat is fed, it must be kept in motion before them, as they will

refuse anything that does not seem to be alive. Mr. Needham



21

has reared Goitiplnis nymphs, which do not feed so readily as

others, on earthworms. As they are more or less ready to eat

each other, they must not be crowded. Aikix is especially

unprincipled in this respect, and diiiiijihus is fond of libellulids.

Recently emerged living imagos should be transferred, each

with its cast skin (exuvia), to dry boxes for a short time, till

their surfaces and wings gain Hrmness and their colors are

matured. The imago and its exuvia should always Ite kept

together.

If the observer is able to keep a close watch on waters

where nymphs occur, al)0ut the time of their emergence, and
does not mind an early morning visit l)efore breakfast for some
larger species, he will be amply rewarded for his enteri)rise

some day by discovering a large nunilier in the midst of the

processes of transformation, often of some species rarely taken

on the wing and very desiralile for excliange purposes.

In all cases and by all means full notes of habits, food,

flight, manner of oviposition, and the like, should be carefully

recorded.
CHARACTERS USED IN CLASSIFICATION.

The Nifinpli.—The most important recognition characters in

the head of the nymph are derived from the lalnum, the an-

tenna^, and the form and sculpture of the upper surface. The
latter is useful in the separation of species anil genera. If the

tip of the labium, or mask, which covers the mouth like a hand,

be grasped and drawn forward, the labium will straighten out,

and may be kept from springing back again. Like the arm, it

consists of two joints, the sii/uiicitfiiiii and iiiciifinii, the end of

the latter liroad, bearing a pair of broad flattened appendages,

the liifrnil /ohes of the lal)ium. The mentuni itself is called the

nii'diait lohr. Each of the lateral lobes bears outwardly, near its

tip, a strong niorah/e hook. In the Pctohir'nht', ^-Eschiiida', and
Goniphidw, the lal)iuin is flat, and applied to the lower surface

of the head ; the lateral lobes are somewhat mandil)le-like,

bearing a row of fine short teeth on the terminal edge in the

Petaluridce, and on the inner edge in the other two families



22

mentioned. In the Coi-fhilegasfcrhhe and LihelhilicJd' the labium

is spoon-shaped, covering the face, the lobes Ijroad and sub-

triangnlar, fitting closely against the median lobe and to each

other. The teeth of the opposed terminal margins are, in the

former family, large, acute, and interlocking, but in the latter

they are rounded crenations, at most not higher than broad.

These are tipped with short hairs, however, which when wet
make them appear pointed. On the inner surface of each lat-

eral lobe may be a row of several conspicuous sette, the lateral

seta', and a similar row on each side of the middle of the median
lobe— the nienfal sefce. The antennae are usually small and

cylindrical, 6- or 7-jointed, but are broad and flattened in the

Petalurida' and Goniphida', ?ind only 4-jointed in the latter family.

The thorax and its appendages are not extensively used in

the classification of the nymph. On each side of the prothorax,

in the .Exchiiida', are a pair of small tubercles, the siipracoxal

processes, best viewed in profile from above. They are fairly

constant in form for each species. The tarsal joints are usually

three in number, but in the Gomphiehe there are only two joints

in the anterior and middle tarsi. The suture between the first

two tarsal joints is oblique, making the basal joint much
shorter above than below. The comparative length of these

two joints, measured on the dorsal line, is a useful distinction in

the Lihellidida'. As the suture between them is usually marked
in darker color above, it is easily located. The iving-pads vary

in size with the age of the nymph, being at first entirely want-

ing. In the adult nymph four or five abdominal segments usu-

ally remain exposed behind them.

The two main groups of dragon-flies are quite unlike in the

terminal appendages of the nymphal abdomen. In Zijgoptera

it ends in three large leaf-like gills, while in the Aiiisoptera the

last segment—the tenth abdominal—bears five small tapering

appendages, which converge and form a valve closing the rectal

opening. The middle one above is the superior appendage, the

lower pair are the inferior appendages, and on each side of the

superior appendage, above the inferiors, are the lateral append-



23

ncjes. These are of considerable value in classification. The
posterior lateral angles of some of the abdominal segments bear
the hifentl spini's, sometimes very small but distinct, as in Li-

helhild, rarely entirely wanting as in Mcsofliciiiis. There are also

spines or teeth along the median line of the abdomen above,

varying greatly in number and form, collectively known as the
dorsal hooks. These and the lateral spines are extremely use-

ful in separating species and subordinate groups. The dorsal

hooks may be viewed in profile from one side, springing the

abdomen down away from the wing-pads ; but when they are

represented merely by teeth projecting backward on the hind
margin and not elevated above the general level, as in Godi-

pliKs, they are of course best viewed from above. The dorsal

hooks are always absent on the first one or two abdominal seg-

ments, beneath the wing-pads, and as their appearance on the

posterior abdominal segments is of especial importance, it has

usually been necessary to refer only to those on the exposed

segments of the abdomen.
The characters given can usually be applied to nymphs of

any age except the very young ones. In case more or less of

an antenna or leg is broken off during the life of a nymph, it

may be imperfectly replaced, usually with one joint less than

before and the relative size of the joints alinormal.

A slight elevation on the under side of the abdomen, if near

its base, indicates the male nymph ; if towards its tip, the fe-

male. This may be easily observed in the common nymphs of

Atiii.r JiiN/its. In the ^Eschnida', the male nymph is also dis-

tinguished by what seems to be a small median scale resting on

the basal part of the superior appendage.

The Iiii(i(/o.—It is unnecessary here to descrilie in detail

the external anatomy of the adult, Init for the understanding

of the keys a brief statement of the wing venation in the Aui-

soptera is desirable. The fore and hind wings have essentially

the same structure, modified to suit the difference in outline

especially at the anal angle. The large number of adventitious

minor longitudinal veins and cross-veins and cells are extremely



24

variable and unreliable, but the relative positions of the princi-

pal veins and cross-veins are constant, and useful as distinguish-

ing characters.

On the front margin of the wing, near its apex, is a con-

spicuous opaque cell, the sf it/ma. The strong vein which runs

along its inner side, extending the whole length of the wing,

is the rail ins. This is intersected about midway of the wing by

a vein (the nodol scrtor) which starts at the margin in a notch

of the costa called the iioihis. Between the costal margin and

the basal half of the radius is the suhcosta, extending as far as

the nodal sector. On either side of the subcosta is a row of

cells separated by short cross-veins, the antecubifal cells and

cross-veins. On either side of the radius, between the nodal

sector and the apex of the wing, are similarly the postcuhital

cells and cross-veins. Behind the base of the radius is a large

cell, the haal/ar spare, bounded posteriorly by the riihitns and
outwardly by a conspicuous cross-vein, the arcnliis. Near the

middle of the arculus arise, jointly or separately, two longitudi-

nal veins, the upper and lower sectors of the arculus. The up-

per sector is the main stem of the median vein, the lower is its

posterior branch. The bases of the media and the radius form

one vein as far as the arculus. The anterior branch of the up-

per sector is the principal sector. It also is intersected by the

nodal sector. The next apparent branch of the upper sector,

running parallel to and just behind the nodal sector, is really

a branch of the radius, and should be called the radial sector.

Two adventitious longitudinal veins, formed by the stringing

together of cross-veins, are the apical sector, just behind the

tip of the radius, and the siipplcniciifari/ sector, behind the radial

sector. A little beyond the arculus, the cubitus leads to the in-

ner angle of a conspicuous triangular cell, or group of cells,

known as the triangle. The elongate cell (sometimes subdivided

by minor cross-veins) above the triangle is the supratrianyti-

lar space. The next and last principal longitudinal vein, behind

the cubitus, is the anal rein. Of the numerous apparent branches

that it sends back toward the hind margin of the wing, three



25

principal ones may usually be ol)served in the hind wing. The

first hf(i)icli of the anal vein starts near or at the posterior cor-

ner of the triangle, and the other two hraiirlies are between
that and the base of the wing. The first and second branches,

in the hind wing, are connected by a strong cross-vein, thus

inclosing the mia/ loop. This and the triangle are particularly

useful in classification.

The male imago may be recognized by the complex struct-

ure on the under side of the abdomen near its base, forming a

conspicuous projection. In the families treated in this article

the male has the anal angle of the hind wing sharply rectangu-

lar, not rounded as in the female. In the LihcUiiUdiv it may
sometimes lie known by the a]ipearance of milky spots on the

wings, in addition to the usual black spots, or by a difference in

wing-markings. The terminal appendages of the male are un-

like those of the female, lieing usually larger and more ii-regu-

larly formed.

KEY TO THE NORTH AMERICAN FAMILIES OF ODONATA.

Nl/IHp/lS.

u(i. Last a])dominal segment bearing three leaf-like tracheal

gills. (Suborder Zi/i/ojiffrn.)

1)1). Basal segment of the antennge extremely elongate.
( 'tf/opfi'ii/j/idir.

b. Basal segment of the antenme short, subrotund.

A(iri<))ii(hv.

(I. Last abdominal segment terminating in five converging,

short, spine-like appendages. (Suborder Anisopfcnt.)

cc. Labium fiat, not concealing the face, with mandible-like

or oblong lateral lobes.

(III. Antenme 6- or 7-jointed, superior appendage usually

notched at apex.

cc. Lateral labial lobes toothed on terminal margin,

antenna^ broad, fiattened. Pcfaliiridif.

e. Lateral labial lobes toothed only on the inner edge,

antennoj slender. ^Escliitidw.



26

(1. Antennae 4-jointed, broad and flat, superior appendage
not notched at apex. Goiiiplilrhf.

c. Labium spoon-shaped, covering most of the face, with

subtriangular close-fitting lobes.

ff'.
Teeth on the opposed edges of the lateral labial lobes

acute, very large and irregular, interlocking; two
stout teeth with a cleft between them at the apex

of the median lobe. Cordulegasferidxf.

f. Opposed edges of the lateral labial lol)es crenate or with

rounded teeth ; at most but a single median
tooth on median lobe. LiheUidhlu'-

l)tl(((JOS.

act. Front and hind wings similar, or nearly so, in outline, usu-

ally elevated when at rest ; males with two inferior

abdominal appendages {Zy<jopiera). Head transversely

elongated, eyes widely separated. Females with genital

valves.

hh. Not less than five antecubital cross-veins in the an-

terior series. Ca/opferyc/idce.

b. Not more than three antecubitals, usually two.

Agrionidce.

a. Front and hind wings dissimilar, hind wings usually much
broader at the base, horizontal when at rest ; males with

one inferior appendage, or none. {Anisoptera.)

cc. Triangles of front and hind wings of similar shape

;

antecubitals of first and second series not coinci-

dent, except the first and another thick one ; second

series of postcubitals complete.

(Id. Abdomen with lateral carinse ; female with genital

valves ; head globose, eyes meeting above along the

middle line of the head. jEschnida'.

d. Abdomen without lateral carinae.

ff. Eyes distinctly separated, head transversely elongated.

(jg. Median labial lobe bifid ; female with genital

valves. Petaluridce.



27

//. Median labial lobe entire ; female without genital

valves. Gonip/i/ihr.

f. Eyes ])ut little separated, or meeting at a single point

dorsally: median labial l()l)e bifid.

Cord II /I'l/asfi'fidee,

r. Triangle of front wings with its long axis at right angles

to the length of the wing ; triangle of hind wings with
its long axis coincident with that of the wing ; ante-

cu])itals of the first and second series mostly coinci-

dent ; second series of postcubitals incomplete at in-

ner end. Libellididif

Family PETALURTD^.

About all wo know of tlie liiology of this interesting fam-

ily is contained in two late papei's by Williamson ('00a, '01) on
the recently discovered nymph and habits of the adult of our

single North American species, 'rdc/iDjdcri/.r f/iorcj/i. This is

eastern and southern in its range, and has been found in an

adjoining state. Kentucky. While the adult has usually been

grouped with the (loiiipliidir. the nymph is nearest to the

^i5'.srAy//r/r/', although it approaches nearer to ])oth the Cordu-

lecjasti'iidif and Goinp/iidie than any other known form, possess-

ing a remarkably synthetic combination of characters. The
adults were observed, according to ^^'illiamson. ovipositing in

boggy ground along a small hill-stream, where a single nymph
w^as found emerging.

Family ^SCHNID.E.

The elongate nymphs of this family have a flat labium very

much like that of the (ionipliidif, consisting of a long mentum
l)earing at its extremity a pair of curved appendages {t\\e lat-

eral lolies) like the mandibles of a l)eetle, more or less toothed

on the inner edge and armed outwui-dly. toward the apex, with

a strong movalile hook. The apical tooth of these appendages

is always present, not terminal and curved inward as in Hom-

phkhv, but rather on the inner side of the squarely truncate or



28

obtusely rounded apex, and directed inwardly. The antennae

are six- or seven-jointed, slender, inconspicuous, and easib

broken, quite unlike the heavy, thick, four-jointed antennae of

the Gomphida'. Dorsal hooks are usually absent. The lateral

spines are small, but their number is of value in generic sepai'a-

tion. The head is widest across the eyes, its hind margin more
or less concave. The structure of the hind angles furnishes

reliable systematic characters. There is on either side of the

prothorax above the front coxa a pair of variously shaped proc-

esses, designated in the descriptions which follow as thesupra-

coxal processes. The abdomen is unusually slender and pliant

at the base, thence indented as far as the beginning of the very

large rectal respiratory chamber. The appendages are rather

long and slender, resembling those of the Curdii/cgasterida',

while the Gomphidce resemble the Libelliilidic in the form of

these appendages. In the young nymphs the laterals are rel-

atively much shorter than in the full-grown ones, and the

superior somewhat so in comparison with the inferiors. This

character is therefore applicable only to nymphs about full

grown. In the male there is apparently an overlapping scale

at the base of the superior appendage.

Like their imagos, these nymphs are among the largest

and most powerful in the order. While the Goiiiphidie burrow

and crawl on the bottom, these are great climbers on submerged

driftwood, branches, roots, and vegetation of all sorts.

The two extremes of this family are curiously unlike in

aspect. Nymphs of the lower genera—which are placed first

in the arrangement of the species which follows—climb upon

dead sticks and driftwood, and have developed a j)rotective

resemblance to these objects. They are rough, blackish, and

inactive creatures, seen but rarely. Those of the last genus

{Anax) clamber actively and abundantly among water vege-

tation, and are bright green or brownish in color, preying vora-

ciously on the smaller life they find there. Nymphs of this

type have paler markings between and behind the eyes and

along the sides of the thorax superiorly, three jmle rings on



29

each femur and tibia, a pale mid-dorsal alidomiual line, and the

apical margins and hind angles of abdominal segments and the

tips of abdominal appendages also pale.

The images are like the Lihi'l/n/h/it' in the great develop-

ment of the eyes, which touch for some distance along the

middle of the head, and in the presence of a roughened carina

on the sides of the abdominal segments. The nearness of the

lowest forms of this family and the Libi'lhiJi(hi- in wing vena-

tion and nymphal sculpture is evident; l)ut the .Esvlniida' as a

whole are much nearer the primitive stock by reason of several

characters, especially the structure of the labium and gizzard, the

general form of the nymph, the form of the egg. and the manner
of oviposition. The wing triangles are elongate lengthwise of

the wing, as in the Gonipliidw, usually with several cro.ss-veins;

most of the antecubitals not coincident. The female is provided

with an ovipositor, by means of which the elongate cylindrical

eggs are inserted into various plants and substances in the water

instead of being washed off by dipping the abdomen during

Hight as are the more or less oval eggs of the other Anhoptcrd.

Even in the nymphs the developing ovipositor, on the eighth

ventral al)dominal segment, is well marked and recognizable in

all but very young individuals. In the males the corresponding

parts are more ditticult to see. but this sex possesses a very

evident scale-like piece overlying the base of the superior ter-

minal appendage.

Mr. Needham has observed that the eggs are deposited in

the stems of plants, in tloating timliers, in piers, etc.. at or very

near the surface of the water, either above or below it, but

always in moist tissue. He has floated pieces of decaying wood
upon a pond as a means of olitaining a^schnid eggs, and these

proved very attractive to the ovipositing females.

The usefulness of the imagos, especially Aikix Jmi ins, on

account of the enormous quantities of pestiferous gnats and
mosquitoes which they destroy, puts them among the particular

friends of mankind.



30

KEY TO THE GENERA OF ^SCHNID^E OF THE UNITED STATES.

Imagos.

AA. Upper part of arculus equal to or longer than its lower

part, its upper sector arising near the lower sector and
far distant from the median vein above ; anal loop evi-

dently 5-7-sided.

an. Supplementary sector, just back of the usually forked

radial (subnodal) sector, nearly straight and sub-

parallel to it, separated from it by one or two (rarely

three) rows of cells.

hh. Radial sector forked near the middle of its length

;

first anal vein leaving the border of the anal loop

at its first angulation, border continued by a cross-

vein which usually meets an apparent fork of the

second anal ; bottom of the loop accordingly 3-sided,

the part formed by the first anal short, not longer

than the remaining distance to the stem of the

second anal.

cc. Anal loop sub-oval, much as in Macramia, wide but

not deep, with a row of, usually, three cells bor-

dering the anal vein, and a second irregular row
along the bottom of the loop ; triangles 3-celled

(rarely 4-celled), inner side not receiving a cross-

vein ; supratriangular spaces with 2 or 3 cross-

veins. I. Nasia'srhna.

c. Anal loop deep, transversely oblong, with two (rarely

three) vertical rows of cells, ending in two or

three cells on the anal vein ; anterior wings with

discoidal triangles 5-7-celled, the inner side

receiving a cross-vein, and the supratriangular

space with 4 or 5 cross-veins. II. Epia'schna.

b. Radial sector not forked ; bottom of anal loop normally

2-sided, one of these sides much longer than the

other, formed by the base of the first anal, the other



31

1\V a cross-vein connecting it with the stem of the

second anal without an evident intervening fork.

(Id. Basilar space (between the arculus and Jiase of

wing) crossed by 2-6 cross-veins ; supratrian-

gular space with cross-veins; male anal triangle

3-5-celled. III. Boi/n-ia.

d. Basilar space free or with a single cross-vein.

ee. Supratriangular space with cross-veins; 2 or 3

rows of cells between the radial sector and
the supplementary sector below it ; male anal

triangle 2-celled. IV. Ba.sitrsrinia.

('. Supratriangular space free from cross-veins, one
row of cells between the radial and supple-

mentary sectors, male anal triangle 1- or 2-

celled. Gomphccsclina.

(I. Radial sector forked or liranched lieyond its middle,

supplementary sector diverging from it until three

or more rows of cells intervene between the supple-

mentary and the basal part of the posterior branch,

lieyond which the supplementary curves rather rap-

idly upward toward the liranch ; anal loop much as

in h/i. the first anal reaching more than half way
(usually about two thirds) across the bottom of the

loop, and connected with the second anal by a short

oblique cross-vein, so that the anal loop appears

4-sided with a truncated inner hind angle ; triangles

of both wings 4- or 5-celled. inner side almost always

receiving a cross-vein. V. ^Escluia.

.. Upper part of arculus much shorter than its lower pari, its

upper sector arising about midway l)etween the lower

sector and the median vein aliove ; radial sector emitting

a number of oblique liranchlets on its posterior side, but

not regularly forked ; anal loop apparently 4-sided, the

first anal extending nearly across the bottom of the loop,

the connecting cross-vein very small and inconspicuous,

usually meeting the second anal at about a right angle.

VI. Ainu:



32

Nymphs.

AA. Four or more pairs of lateral spines (sometimes an imper-

fect additional pair anterior to those here counted as

the first pair) ; eyes occupying less than two thirds of

the side margin of the head ; scale of male on base of

superior terminal appendage triangular, not emarginate

at apex.

aa. Five pairs of fully-developed lateral spines.

hh. Antennae 6-jointed, the last two segments of the other

genera being represented by a single long joint

sometimes divided by an indistinct suture in full-

grown nymphs ; eyes small, occupying only one

third the lateral margin of the head ; apex of supe-

rior appendage truncate or feeljly emarginate.

cc. Abdomen carinate posteriorly, forming low dorsal

hooks. I. Naskesrhna.

c. Abdomen without trace of dorsal hooks.

II. Ep'KesrhHO.

b. Antennae 7-jointed ; eyes occupying about half the

lateral head margin ; apex of superior appendage

deeply emarginate.

del. Lateral labial lobes scarcely tapering, obtuse or

subtruncate at tip, median lobe minute!;;

toothed each side of cleft ; superior appendaj:,e

not less than five sixths as long as the inferiors.

III. Boi/eria.

(I. Lateral labial lobes tapering to the incurved

apical tooth, median lobe with a chitinous

callosity on each side of cleft ; superior append-

age about three fifths as long as inferiors.

IV. Basiceschna.

a. Four pairs of lateral spines, the anterior pair much
smaller than the others but distinctly spiniform

;

apex of superior appendage broadly but not very

deeply notched ; median lobe of labium not toothed.

V. jEschna.



33

A. Three pairs of lateral si)iiies, all well-develoiied ; eyes

occupying at least two thirds of the side margin of the

head ; scale of male on l^ase of superior appendage short

and broad, emarginate at apex ; apical eniargination of

superior appendage about as in ^Esc/uki. VI. An((.r.

I. Nasi^eschna Selys.

The single species which follows has until recently re-

mained in the genus .EscIiiki, but it is very evidently geuer-

ically distinct from all other American .Esvhnida', and has. in

fact, lieen lately so recognized by Selys-Longchamps and Fors-

ter under the name of Xasia'scliiid* It has been referred by

Hagen (77, p. 37) to Bnir/ii/froii Evans, but it is intermediate be-

tween that and Epifeschna Hagen. Aside from secondary sexual

characters, it is nearer the latter genus. The following char-

acters sufficiently distinguish the imagt) from related genera:

Face concave ; frous narrow, produced above into a shelf-

like prominence, with a In-oad longitudinal superior furrow, and

without "T" spot. Radial (sulniodal) sector forked midway
between the nodus and the middle of the stigma, the branches

of the fork diverging all the way to the wing-margin ; first and

second l)ranches of media (principal and nodal sectors) not ap-

proximated beyond the fork of the radial sector ; apical sector

originating near the anterior end of the stigma ; anal loop ovab

lying nearly lengthwise of the wing, with about two longitudi-

nal rows of cells, usually one cell between bases of the 2d and

3d anals ; supratriangular space with two or three cross-veins,

triangles 3- or 4-celled, inner side not receiving a cross-vein.

Inferior abdominal appendage of male hardly shorter than

superiors.

EpIa'srIiiHi (Icbifls Karsch (Ent. Nachr., Vol. XVII.. p. 2S6)

perhaps belongs to this genus.

The nymph of NasirrscJuKi differs from all other a?schnids

in the possession of dorsal hooks. Our species represents a dis-

tinct and primary step in the evolution of the ^Eschnichr, and

its claim to generic distinction is beyond question.

See "Ent. News," Vol. XI., p. 546.



34

Nasiffsrhnn pentacnnfha Ramb.

The nymph of this species was doubtfully referred by Gar-

man ('90, p. 178) to Epiceschna heros. It differs from all Amer-
ican seschnids hitherto described in the possession of dorsal

hooks on the 8th and 9th abdominal segments, having the form

of a median carina terminating in a posteriorly projecting tooth.

In aspect and habits this interesting nymph is very peculiar. It

is rough, blackish, and very sluggish, and clings to dead branches,

roots, or driftwood in quiet water. Doubtless it depends, as

Mr. Grarman suggests, on its obscure color and slow movements
for escaping detection. Examples were taken by Mr. Hart in

April and May on branches in dead water and along the banks

of a slow-flowing stream near Urbana, and also a very young
one in August in a muddy remnant of a temporary stream in

Perry county, in southern Illinois. Mr. Garman's specimens

were from a muddy slough of the Mississippi near Quincy. Out-

side of Illinois the imago has hitherto been reported only from
New York (June), Louisiana, and Texas. Mr. Needham saw
one taken at Wellesley, Mass. In Illinois it was taken by Walsh
at Rock Island ; and we have specimens from Urbana, and also

from Cobden in southern Illinois. An imago was found float-

ing May 20 in a narrow passage where nymphs were at the time

quite numerous on dead branches lying in the water. From
these nymphs adults were obtained up to June 10. Our other

imagos were taken June 11 and 16.

In addition to the presence of dorsal hooks, as already stated,

the nymph differs further from E. heros, as described by Cabot

('81, p. 30, PL I., Fig. 3), in the presence of two pairs of tuber-

cles upon the head, in the absence of lateral spines on the 4th

abdominal segment, and in that the two processes above each

front coxa are of equal length.

The nymph measures 48 mm.: abdomen, 34 mm.; hind

femur, 6 mm.; width of abdomen, 8.5 mm., of head, 9 mm.
The median labial lobe is prominent, cleft, without teeth

on each side of the cleft. The lateral lobes are squarely trun-

cate, with the outer angle rounded, their sides parallel beyond



35

the base of the hook, the lobes terminating in an internal tooth

as long as half their apical width.

The eyes are small but very prominent, about as in Bnicln/-

fniii. Behind them the lateral margins of the head extend in

a ridge to the rather acute hind angles. Back of the middle of

this ridge is a prominent tuliercle. There is another pair of

tubercles, close together on top of the head. The hind margin
is broadly concave.

Abdomen with lateral spines on segments 5 to 9. increas-

ing in length posteriorly, those of 9 about half as long as the

10th segment.

Superior and inferior appendages equal, obtusely pointed,

male scale equilaterally triangular, its apex shining l)lack,

rounded. Dorsal hooks on segments S and 9 and sometimes on

7 ; represented on the other segments by median ridges.

II. Epi^schna.

This genus is intermediate in structure between Xasia'srluia

and ^srhita, the three genera which follow between it and
yEschnu representing a branch line of development. The
nymph reseml)les that of Nas'ui'schnd and proliably has similar

habits, but it is without dorsal hooks. The adult n^ay l)e recog-

nized by the structure of the anal loop and other details of

wing venation. We have but one species in this country.

Epueschna lid'os Fabr.

This, the largest of our dragon-flies, is not common in Illi-

nois, though Mr. A. H. Muudt ('82) has recorded the passage of

a swarm of them towards the southwest over Fairbury, Living-

ston county. The air, he says, was literally alive with them

;

few alighted, and on the following day only a few stragglers

remained. Aiuix Junius was, and continued to be, the common
local aeschnid species. Heros is found from Quebec to Mexico;

thence west to the Mississippi River. The Illinois localities are

Rock Lsland, Quincy, Bloomington,and Ur])ana, and the extreme

dates April 23 and September 1. Williams says it often enters



36

houses before storms, and is commonest in May and early sum-
mer, being found along roads, in open woods, or over fields or

water—wherever it may capture flying insects.

We have not seen the nymph, but Mr. Needham has com-
piled from Cabot's description and figure ('81, p. 30, PI. I., Fig. 8)

the following characters.

The nymph is 40 mm. long.

The head is deeply concave behind between the rounded

hind angles Of the acute and approximated supracoxal proc-

esses, the anterior process is the longer. Median lobe of . labi-

um slightly notched at middle, the borders of the notch varia-

ble. Lateral lobes truncate on the end, denticulate within.

Lateral spines on abdominal segments 5 to 9. Appendages
equaling segments 9+10; laterals slightly more than half as

long as the inferiors, which are a little longer than the supe-

rior. The tip of the latter is blunt.

in. BoYERiA McLachl.

In appearance, structure, and habits this is much like the

preceding genus. The iniagos of the single North American
species are not very large. They have a small brown spot

at the base of each wing, reddish veins, and a yellow ptero-

stigma.

Boyeria vinosa Say.

Fonscolotnbia vinosa, Calvert, '93, p. 247.

Nenrcesch7ia furcillata, Cabot, '81, p. 29, PI. II., Fig. 3 (male nymph).
Gomphcpschtia furcillata, Cabot, '81, p. 28, PI. II., Fig. 4 (female nymph).

A single exuvia was found by Mr. Needham attached to

the under side of a plank, several feet above the water, at

McHarry's mill-dam on Quiver Creek, in June. He has col-

lected nymphs since in the rapid streams about Ithaca, N Y-

Another nymph was taken by Mr. Hart in April, clinging to

floating driftwood in Quiver Creek below the mill-dam. Like

iV. pentacantha, it is dark colored and sluggish, and appears to have
similar habits. Kellicott ('96, p. Ill) found the imagos quite



37

abundant in Ohio in late summer ; his notes on their habits are

worth repeating here, as thej- tend to confirm the preceding

statements of the habits of the nj'mph. " It prefers seclusion,

hence should be looked for along ponds and streams with well-

wooded banks, where branches overhang the water, and where
half-submerged logs and rubbish al)ound. The males may be

seen from early in the forenoon until dark, on warm days, ex"

ploring every corner among the obstructions at the water's edge.

An interesting ha])it noted was that it would often fly out and
carefully examine a passing skiff. The females, when not ovipos-

iting, are suspended from some overhanging twig." Van Duzee
('97) makes a similar observation. The imago is widely dis-

tributed, lieing reported from Canada and the eastern United

States as far south as the Oarolinas, and west to Arkansas and
Illinois. Professor Frank Smith took examples at Macatawa,
Mich., where he says it is fairly common. In Illinois it has
been taken only at Havana, Bloomington, and Url)ana, a single

specimen in each case. The one taken at Havana was flying

along the river front. The extreme dates on record are June
20 and October 15, most of the occurrences being in July.

The nymph measures 35 mm.; al>domen, 24 mm.; hind

femur, 6 mm.; width of abdomen, 7.5 mm., of head, 7 mm.
Body elongate ; eyes large and very prominent, the sides of

the head extending backward from the middle of the hind

border of the eyes to very prominent slightly elevated hind

angles, between which the hind margin i.s regularly concave.

Lal^ium extending lietween middle legs. Median lobe prom-
inent, cleft, a conspicuous tooth on each side of cleft, at some
distance from it, on front margin.

Lateral lobes truncate apically, with a stout tooth at the

inner apical angle and the convex inner margin denticulate.

Hook stout, arcuate, and rather sharply incurved at tip.

Prothorax a little narrower than hind margin of head.

Alidomen widest across segment 7, wing-cases reaching only the

middle of 4, lateral spines well marked on 5, larger and gradu-

ally increasing in length on succeeding segments. Inferior ab-



38

doininal appendages incurved beyond the end of the superior,

which is one foui-th to one third shorter, in the male narrowly
cleft at tip, in the female pointed. Lateral appendages in the

male one third, in the female one fifth, the length of the in-

feriors.

IV. Basi^schna Selys.

This genus contains but the one species following.

Basi(eschna Janata Say.

The nymph of this species resembles that of ^Eschna, but

may be recognized by the greater number of lateral spines,

the deep notch of the superior appendage, and the unusually

narrow and pointed lateral lobes of the labium. We have
taken the nymphs in Sand Lake, northeastern Illinois, on
Cham in shallow water, and upon rotten logs along the shore of

Delavan Lake, Wisconsin. Mi*. Needham has studied the spe-

cies at Ithaca, and made notes on it as follows: "This is a

stream-loving species. It flies from the first of May until mid-

summer. It is very common in the small turbulent creeks

about Ithaca, where I bred a good many in 1897, and collected

a large number of nymphs." He has also a large number of

specimens, in all stages except the youngest, from Mr. F. G.

Schaupp, of Texas.

The imago is on record from several Atlantic states, and is

reported by Williamson in Indiana, and by Kellicott as not un-

common May 1 at Columbus, Ohio. It has been taken once in

Illinois, at Thatcher's Park, Chicago. The dates given run from

the last of April to September. Kellicott ('97) collected it in

April, under unusual weather conditions.

The nymph measures 37 mm.; abdomen, 27 mm.; hind fe-

mur, 6.5 mm.; width of head, 7 mm., of abdomen, 7 mm.
Generally dark colored, the mid-dorsal pale line on the ab-

domen often breaking down over the sides in pretty mottlings.

The usual three yellow rings on the femora usually very dis-

tinct.



39

The head is evenly narrowed behind the eyes to the sharply

angular hind angles.* The labium is abruptly widened in its

apical third ; its median lobe is without teeth, but with the

usual fringe of hairs and closed median cleft. The lateral

labial lobe is slender beyond the base of the hook, and is grad-

ually narrowed to a slightly incurved point, with indistinct

denticulation along its inner margin.

The abdomen bears distinct lateral spines on segments 5 to

9, those of 6 to 9 hardly increasing in size posteriorly, those of

5 somewhat smaller. The abdominal appendages are a little

declined at their sharp and slender apices ; the superior is less

than two thirds the length of the inferiors, deeply notched at

the tip ; the laterals are about half as long as the superior.

V. iEscHNA Fabr.

The nymphs of this genus resemble those of Auax, but

have four or five pairs of lateral spines instead of three. Their

habits are similar to those of Aiuu; but they appear to be more
retiring and inactive, being often found clinging to submerged

roots and twigs. A single example was found in the Illinois

River at Havana upon a submerged branch. It was clearly an

j-Esclni(( nymph, but escaped before it could be further studied.

Dr. R. H. VVolcott had under observation a nymph of this genus

in Michigan. It liked to cling to the under side of floating ob-

jects. It was once seen eating an Aijrion nymph, but was usu-

ally fed on flies, of which it ate two to four a day, coming to the

surface for them each morning, and refusing food in the after-

noon even when, as an experiment, the morning meal was
purposely omitted.

The imagos are of lai'ge size, and are dark brown and blue,

instead of green as in A mix, their wings clear, with black veins.

Like Epiu'scltiKi licros they seem to have a fondness for the

vicinity of dwellings.

*In the younger of Mr. Needham's Texas specimens these angles bear distinct

tubercles.



40

The species are numerous and difficult of separation. The
nymphs are not usually abundant, and as few of even the com-
moner American species have been bred, our knowledge of the

immature stages is as yet unusually imperfect. The key which
follows indicates the salient characters of the nymphs here

treated. In the young nymphs of ^Ese/nia sp. h a tendency to

abdominal cross-bands was noted, and the sutures beneath

were broadly bordered with paler color, except those between
3 and 4, 6 and 7, and 8 and 9.

KEY TO THE NYMPHS OP ^SCHNA HERE DESCRIBED.

aa. Lateral spines of sixth abdominal well developed, reaching

at least about half way to the hind margin of the seg-

ment; those of seventh attaining about to the hind

margin; lateral labial lobes squarely truncate at tip.

bij. Fifth abdominal with minute lateral spines, lateral

appendages fully two thirds as long as the superior

appendage; supracoxal processes sharp, equal.

1. rh'jjsi/fh'a.

h. Fifth abdominal with at most only a trace of lateral

spines; lateral appendages less than two thirds as

long as the superior; supracoxal processes less sharp,

the posterior larger; abdomen in younger individ-

uals with a medio-dorsal pale line bordered each

side with blackish. 2. consfricfa.

a. Lateral spines of sixth abdominal minute, distant from its

hind margin by several times their own length, those

of seventh reaching about half way to the hind margin.

re. Apices of lateral labial lobes squarely truncate, con-

tiguous when closed; in younger specimens a pale

spot near anterior ocellus and another on clypeus,

and a pale median line enlarged on the middle

of abdomen into an increasing series of pale spots,

the largest and last on segment seven.

f]<J. .Superior appendage sparsely fringed laterally with

pale hairs; posterior part of head with some pale



41

markings ou the darker gi'ound-eolor; outer api-

cal angle of lateral labial lohes not rounded.

cc. Hind angles of head prominent; quite young exam-
ples (7 mm.) with separate spots at the posterior

ocelli, the 1)ody not transversely banded. 3. sp. ( (i).

('. Hind angles of head rounded; young examples (10

mm. or less) with a trilobed spot at the back of

the head above, the metathorax, a V-shaped spot

on the base of the abdomen, and segments (5 and

7, all pale.* 4. sp. {h).

(I. Superior appendage fringed laterally with blackish

hairs; posterior part of head pale, with a dark lateral

stripe, a median pair of dark spots, and the poste-

rior margin dark; outer apical angle of lateral laliial

lobes narrowly rounded; head behind the eyes short.

5. nilifuniico.

r. Apices of lateral labial lobes exteriorly broadly curving

to the acute terminal hook at the inner apical angle,

not contiguous when closed; color dark brown, a fine

median pale line on labrum. a transverse spot at the

anterior ocellus, behind which are two pairs of small

spots; a fine median whitish line ou thorax and abdo-

men, vanishing near the middle of the abdomen.

G. ii'iiicdlis.

1. .Ksi-luKi c/cjisi/dra Say.

.£sc/i/!a fii/iiitica Cabot (nymph).

This wide-spread species has been found in the northern

part of Elurope and Asia, but especially in North America, ex-

tending south as far as the Dakotas. Wisconsin, Indiana, north-

ern Ohio, and Maryland. It has not been taken in Illinois. Walsh
reported it from Rock Island ("()2, p. 397), but Hagen has iden-

tified Walsh's specimens as ccrtintJis. A number of r/rpsi/<Jra

imagos were taken, however, by Mr. Hart at the southwest end

*In the young . I iiaxJunius the anterior pale transverse band includes the last

two thoracic and the first three abdominal segments, and the posterior band covers

segment 8, often also 7, and sometimes even the adjacent part of 6.



42

of Lake Geneva, Wisconsin, within ten miles of the Illinois

line, September 2 and 5, about the summer resort cottages and
along the lake shore. In New York it has been taken from
August 28 to September 22 ; in Nova Scotia, July 26. Cabot

described and figured ('81, p. 23; PI. II., Fig. 2) a nymph
taken at the same time and place as imagos determined as

jE. " ereiiiifica," and assigned it by supposition to that species.

Eremitica is an erroneous writing of prr-'>»/Yf/, which is a syno-

nym of clepsydra.

The nymph of clepsi/dra differs from all other known
JEschna nymphs in having recognizable lateral spines on the

fifth abdominal. It is otherwise similar to constrida, and may
be distinguished especially by the following characters given

by Cabot ('81, p. 37) from full-grown nymphs of both sexes

from Hermit Lake in New Hampshire, mouth of Red River of

the North, and Minnesota: " Hind angles of head oblique, proc-

esses long, equal, sharp, tips bent a little outwards, inclosing

less than a right angle ; lateral appendages two thirds the

length of the middle one ; female valves not quite reaching tip

of segment."

2. jEschna constricfa Say.

This is one of our more common dragon-flies, appearing on

the wing in Illinois after midsummer, and disappearing only

with the autumnal frosts. It ranges over the entire northern

part of the United States, and from Labrador into Siberia,

being apparently less common throughout the Mississippi

valley than on the north Atlantic and north Pacific slopes.

The Illinois specimens are from various localities in the cen-

tx'al and northern parts of the state. Van Duzee ('97) found

it most abundant along the meanderings of small brooks in

hilly country.

Our nymphs were collected in a small ditch at the west

end of Lake Geneva, Wisconsin, in August, and from a little

streamlet in the Mississippi bluffs near Savanna, Illinois. We
also have a half-grown nymph from Pine Lake, near Charle-



43

voix, Michigan, sent liy Dr. Henry B. Ward. Mr. Needham has

also two quite young nymphs of perhaps the same species sent

him by Mr. F. 0. Willard from Tombstone, Arizona. These all

have labial characters similar to Epiu'srlnia heros. The nymph
has been described and figured by Cabot ('81, p. 28; PI. III., Fig.

1), but it does not seem superfluous to present here a detailed

description, drawn up from the Lake Geneva specimens and
from material ])red by Mr. Needham at Ithaca, New York, in

1897. It differs from the remaining species of the genus in the

greater size of the lateral spines of the sixth abdominal, as

described in the key.

A full-grown male nymph from Montana presents some ap-

parently trivial differences which, under the circumstances, it

will perhaps be best to mention. On either side of the open

end of the labial cleft is a minute tooth ; the tooth on the inner

apical angle of the lateral lolies is less acute than in the typ-

ical specimens and not so distinct. The lateral spines of the

fifth segment are identical with those of the typical examples,

but the others are a trifie longer and more divergent. The
notch of the superior appendage is a little deeper. In length

this appendage is exactly intermediate l)etween the inferiors

and laterals, as in the typical specimens, but the laterals are

relatively shorter and the inferiors longer, the laterals being

very little more than half the length of the inferiors.

The nymphs measure 40 mm.; al)domen '2(5 mm.; hind femur
8 mm.; width of abdomen 8 mm., of head 8 mm. Color blackish

brown, paler l)elow on legs, on lateral margins of abdomen, and

on superior and inferior abdominal appendages. Two black

marks upon the costal edge of each wing-case near the base, a

pair of l)lack dots near the tip of the basal male enlargement

of the superior appendage, and a similar dot near the tip of

each lateral.

Head narrowed behind the eyes to the broadly rounded

hind angles, one third wider across the eyes than across the

hind angles ; hind margin moderately concave, with some coarse

pubescence disposed in numerous irregular rows near the hind

angles.



44

The labium extends posteriorly beyond the bases of the

second pair of legs. The median lobe is considerably produced,

and rather deeply cleft in the middle, the angles each side of

the cleft minutely rounded. Latei'al lobe squarely truncate on

the end, with a minute tooth terminating the inner border,

which is finely denticulate. Movable hook long and arcuate.

Supracoxal processes of equal length, a little obtuse, pointing

in slightly diverging directions, the angle between them about

75 degrees, the posterior a little broader.

Abdomen with lateral spines on segments 5 to 9, increasing

in length posteriorly, those of 5 very minute and rudimentary,

of 6 somewhat appressed, of 8 and 9 nearly equal, of 9 three

fourths as long as the 10th segment at middle. Appendages
longer than segments 9 and 10, inferiors longest, slightly in-

curved at tip ; superior four fifths as long as inferiors, exactly

intermediate in length between the two pairs, a pair of short

spines at its terminal angles, between which the tip is semicir-

cularly notched, sides parallel and fringed with fine hairs, its

basal male enlargement elongate, triangular, scale-like, cover-

ing its basal three sevenths ; laterals three fourths as long as

superior, cylindrical, abruptly acuminate apically.

3. jEschna sp. (a).

A number of small nymphs from Yellowstone Park and
Montana, 15 mm. and less in length, appear to be different

from any others known to us. The lateral spine on the sixth

segment is very small ; on the fifth segment there is merely a

smooth angulation ; the rear angles of the head are very slightly

or not at all rounded. The supracoxal processes are exactly

equal and similar. The very young individuals (7.5 mm.) are

not transversely banded. The nymphs are too young for any

comparison of the lengths of the terminal appendages. They
were collected July 25 to August 19 from bark and rotten logs

in a bayou of Flathead River, near Demersville, Montana, and

in Yellowstone Park from the Gibbon River and a lagoon of

geyser water near Lewis Lake.



45

4. jEsclnid sp. (/;).

This species is represented in our collections by a quantity

of young nymphs, 5 mm. to 24 mm. long, all taken from rushes,

wood, etc., along the weedy margin of a lake in Yellowstone

Park, near (iardiner River, August 80. Unlike the preceding

species, which it closely resembles in structure, the hind angles

of the head are rounded, and the very young nymphs are banded
much as in Anax Junius but not on quite the same segments.

5. yEsvhiut calij\intini Oalv.

AS. californica, Calvert, '95.

This is a species of the far West and the Pacific coast, not

found in Illinois. The nymph is here described for the first

time. A single one was taken from an irrigation ditch near

Tombstone, Arizona, and l)red in May. 1S97, by Mr. F. C. Wil-

lard. The exuvia and imago are now in the Cornell University

collection. Through the kindness of Dr. P. P. Calvert, Mr.

Needham has examined three young nymphs from the collec-

tion of the Academy of Natural Sciences, Philadelphia, which
appear to belong to the same species.

The nymph measures 34 mm. ; alidomen 20 mm.; hind

femur (5 mm.; width of head, 7.5 mm., of a))domen 7.5 mm.
The head is narrowed behind the eyes to the rounded hind

angles, lietween which are two pairs of scars, the smaller pair

on each side of the median line. The median lobe of the

labium is not prominent; its border is without teeth, its median
cleft closed. The lateral lobes are truncate, but their external

angle is rounded off slightly. The supi-acoxal processes are

equal and divergent at almost a right angle.

The lateral spines on the 6th abdominal segment are slen-

der and appressed, those on the succeeding segments increas-

ing regularly in length. The inferior abdominal appendages

are as long as segments 9 + 10, the laterals one half as long as

the inferiors; the superior a little shorter than the latter,

roundly notched as usual at the tip, and fringed witli a row of

stout blackish bristles along either side.



46

6. yEsrhna verfira/is Hag.

^schnajimcea verticalis, Calvert.

This species ranges from Nova Scotia to the District of

Columbia, and has also been taken in Ohio, Illinois, and Cali-

fornia. At the East it is commoner than dcpsjidnt, but not in

Illinois, having been taken only at Rock Island. The published

dates range from July 16 to October 18. Harvey found it com-
mon over meadows, bogs, and rivers in Maine. The nymph
has not been described or bred. We have in the Laboratory

collections, however, two lots of young specimens taken from

the marshy shore of Grass Lake, 111., which are clearly distinct

from the others here treated, especially in the form of the lat-

eral labial lobes, and which may be reasonably assigned to the

above species.

VI. Anax Leach.

The abundant and very uniformly distributed nymphs of

this genus may be separated from any of the others of this

family—which are usually much less common—by the fact

that they have only three pairs of lateral spines. In habit

they are quite unlike the lower seschnid genera, as they thrive

especially—even in small temporary pools and ditches—wher-

ever there is a luxuriant aquatic vegetation, through which

they clamber actively, preying on all sorts of water insects.

The imagos are familar objects, large, bright green, the

wings with yellowish costal mai'gins, circling about in swarms
or singly in search of small insects in the air.

Anax Junius Drury.

King of the dragon-flies, powerful and fearless, our great

Anax is a dominant type of its class ; and its worthy offspring^

omnipresent and omnivorous in the water world, is equaled

only by the large water-tigers {Cyhisfer) in strength, activity,

and ferocity. It has none of the sluggish obscurity of Calop-

teryx, GompJms, Boyeria, or Epicordidia. It clings to water



47

weeds, usually nearer the surface than the bottom, in an attitude

of alertnes.s, with the head poised low and alidoinen slightly

elevated. Its locomotion is relatively rapid, accomplished

either by walking or by ejecting water from the respiratory

chamlier. It is notoriously cannilialistic. In the midst of

abundant and choice food the larger nymphs will eat the smaller

of their own species, and two of equal size may not safely be

kept in close quarters. An(i.r nymphs are easily recognized liy

their having lateral spines on but f/ircc abdominal segments pre-

ceding the last. The young nymphs have wide alternating

transverse bands of brownish and white, but become in later

life a nearly uniform green or bi'ownish green.

To get a definite idea of the food of the nymph, the stom-

achs of a considerable number, taken at different dates and in

various situations, were examined by Mr. Hart. Thirteen of

them contained small quantities of comminuted food, the most

unexpected feature of which was the large quantity of remains

of univalve Mulhisca. This was principally Amnicola, the lin-

gual rilibons and opercula of which were easily recognized. The
molluscan element was estimated as 15 per cent, of the food.

Filamentous alga?, in quantities too large to be accidental, con-

stituted 11 per cent. Crustacea were also 11 per cent., nearly

all a small amphipod species, Al/orclirsfrs (/riitafa. Of the re-

mainder, 56 per cent, was recognized as the remains of in-

sects, including larvte of ('hiroiionius. of Sfrafionii/iida', of Tipula,

of beetles, of A(jrion, and of a small caddice-tiy {Ehii(iniphUa),

smaW Hem ipfcm, ^nd even themo.ss-bug (/'r/wror/.s). the sharp

sting of whose beak often temporarily interrupts the routine of

station field-work. A nymph was observed by us to attack a

crawfish three fourths of an inch long, and devour its abdomen.

Young nymphs in a breeding-cage ate A sellus eagerly.

In the vegetation-filled waters everywhere about Havana
these nymphs developed in great numbers, while in the Missis-

sippi and associated waters about Quincy. in which vegetation

is comparatively scanty. Mr. Uarman found but few individuals.

Our specimens of the nymph were taken from a remarkable



48

variety of waters containing vegetation, and even where other

species of nymphs are largely wanting. Lakes, rivers, ponds,

swamps, transient pools, small ditches, springs, throughout the

state,—all produce a crop of these nymphs. This fact together

with their active habits and voracity, would indicate a rapid

development. Kellicott states that from an excavation for an

artificial lake, which was filled up with water in early May,

large numbers of_y'««/H.simagos emerged in late August, indicat-

ing the occurrence of two broods in a year. This is confirmed

by one of our experiments, half-grown larvae placed by Mr.

Hart in a breeding-cage June 16 reaching the imago stage

August 4. A noticeable reduction in the number of large

nymphs and an increase of the younger ones about July 1 has

been recorded in two different years. Young predominate in

our October collections, and those taken early in spring are

mostly of large size. The imagos attain greatest abundance in

May. On May 19, 1894, the abundance of fresh exuvi^ was con-

sidered worthy of record.

According to Kellicott this is the first dragon-fly abroad in

early spring, remaining until the middle of October. Several

pairs were seen by him flying in union March 21, the female

ovipositing. In early spring, according to Mr. Needham, the

eggs are deposited in the water-soaked stems of reeds, in float-

ing sticks, pieces of board, etc.; while later in the season they

are placed in the green and growing stems of aquatic plants.

He states that the females are usually held by the males during

oviposition, and often descend into the water for this purpose.

The imagos wage a ceaseless warfare on gnats, mosquitoes,

and other small winged insects. In August, writes Mr. Need-

ham, I saw a small swarm (more than a score) moving together

through an orchard in Cass county. 111., miles from open water.

At the same place a few days later, during three successive

days of high wind, Anax jimius and Tirnnm lacerafa could

always be seen hovering in the lee of orchard trees, grape

arbors, stacks of grain, etc., circling swiftly several rods away
from shelter then beating slowly toward it again, head to wind,



49

evidently watching for game, and making short dashes forward

betimes, presumably to seize some small insect driven from
shelter by the wind.

Mr. Adams and Mr. Hart observed a large swarm which
gathered in the lee of the field laboratory boat August 8, at

twilight, and after half an hour or more of rapid circling about

scattered abruptly at about 7:40. An adult Aanx was once

seen to devour an agrionid imago.

The species covers the continent from Alaska to the West
Indies, and is found in eastern Asia.

Cabot ('81, p. 15; PI. I., Fig. 2) has figured the nymph, and
many different reproductions of his figure are current. The
following brief description covers the "more important char-

acters.

The nymph measures 45 mm.; abdomen, 31 mm.; hind

femur, 8 mm.; width of abdomen 8 mm., of head 8 mm.
The head is flat and broad, widest across the posterior

third of the eyes. Antennae minute, slender. Eyes broad, lat-

erally prominent, produced well toward the dorsal median line

at their posterior internal angles. Labium very long and fiat,

extending posterioi-ly beyond bases of middle legs; median lobe

with a narrow median cleft. Lateral lobes oblong, denticulate

along inner margin, an incurved tooth at tip; hook long, strong.

Abdomen widest across segment 7, tapering both ways. No
trace of doi'sal hooks; short lateral spines on segments 7 to 9.

Hind margin of 9 (as seen from above) straight. Superior

appendage a trifle longer than segments 9 and 10. notched at

tip; inferior appendages a little longer, sharp pointed; lateral

appendages half as long.

Young nymphs are more unlike full-grown ones in habitus

than are those of most other species. In the earlier stages the

abdomen appears more attenuated toward the base. The supe-

rior appendage is at first very short and blunt and directed

upward, but after a few molts it becomes elongated and

notched at the tip, though it remains for a time much shorter

than the inferiors. Lateral spines on 7 to 9 appear very early.



50

Nymphs one third to one half grown are transversely banded

with brown across the head, across segments 4, 5, and 6 (some-

times 7), and across 9 and 10. These bands disappear with

subsequent molting, and the full-grown nymphs are of nearly

a greenish color variously mottled with brown.

Family GOMPHIDtE.

There is a marked "family resemblance" among the

nymphs of each dragon-fly family, and this is quite as true of

the present one as of any other. Its nymphs are all evident

gomphids, even Hagenlus, broad and flat as it is. Their most

marked characteristic is the pair of thick, rough, four-jointed

antennae. The flat labium is built on the same plan as that of

the Agrionidce and ^Escknidce—the mentum, nearly truncate in

front, bearing a powerful mandible-like pair of grasping arms,

carrying on the outer side beyond the middle a large movable
hook, the arms usiially toothed within and ending in an

incurved point. The domain of these nymphs, except Hagenius,

is the muddy, sandy, or rocky bottom of various kinds of

bodies of water, according to species. They live amongst
fallen trash and sediment, burrowing shallowly along with the

tip of the abdomen turned up so as to reach the water, thus

enabling them to breathe while foraging in a stratum of great

biological richness. Their colors are similar to the mud and
sand in which they dwell, and are often obscured by a coating

of mud. The flattened body and stout legs are well adapted

for burrowing. The head is broad and more or less wedge-
shaped, the antennae, laid close upon the labrum, forming the

point of the wedge. The third antennal joint is much the

largest; the fourth, a mere rudiment. The legs are stout, the

two anterior pairs directed forward, their tibise armed at tip

externally with more or less well-developed burrowing hooks.

The hind legs are directed backward and used to push the body
forward. The anterior and middle tarsi are two-jointed, not

three-jointed as in other Odonata. The dorsal hooks and lat-

eral spines are rather feebly developed.



51

When the nymphs are readj^ to leave the water to trans-

form they cannot ascend very small stems, such as those of

reeds, owing to the wide divergence and separation of the legs;

but they can readily climb up on a broad surface like the

piers and timbers of a bridge or the side of a floating liarge, or

on large rocks or tangled growths. They do not go far above
the water, not more than a few feet at most, and here they may
be seen transforming at any hour of the day or night, but

most commonly about daybreak, leaving behind them their

mud-incrusted shells, which are often seen in great numbers in

the situations already mentioned. A study of the different

sizes of larvae indicates that the nymphs require two or three

years to attain maturity.

The imago gomphid is in most species a medium-sized

clear-winged dragon-fly, the thorax rather prettily striped with

green or yellow, the abdomen rather slender in comparison
with the thorax but often dilated towards the tip, the eyes not

touching as in the ^Esrliiiidcp and remaining Anisoptera, but

widely separated as in the Ayrionidfe and Pefaliirida'. In wing
venation they approach the Libelhilidce, the triangles being well

differentiated, with few or no cross-veins, but the antecubital

cross-veins are mostly non-coincident, as in the other Ani-

ftoptera. The anal loop is small, inclosing from one to a few
cells. As with the LiheUuUda;, the ovipositor is not developed,

and the more or less oval eggs are transferred free to the water

by successive taps at its surface with the tip of the abdomen
during flight, the females ovipositing alone.* The number of

eggs deposited is very large. Copulation takes place while at

rest on shore. The females frequent the trees or pathways back

of the shore, while the males rest nearer the water. The latter

may be quickly recognized by their curious terminal append-

ages resembling four fingers aliout to grasp something, while

those of the female are short, simple, and inconspicuous. The
period of flight usually occurs rather early in the season, vary-

*The eggs in a Nym.ph<ra leaf figured by Lampert, '99, as eggs of Gomphus
have been demonstrated by our breeding work at Havana to be those of Agrionida-



52

ing from a few days to a few weeks according to species. Dur-

ing this season they are little in evidence. Indeed one may not

see a living adult, although thousands of fresh exuviae be

scattered along the banks.

The structural characters of the nymph appear early, and
the descriptions which follow will apply to all but the very

youngest nymphs. The structures latest to develop are per-

haps the teeth on the inner margin of the lateral labial lobes,

which in some species, at least, increase in number with succes-

sive molts. On account of the short life and restricted range

of flight of the imagos they are not often taken in ordinary col-

lecting. If the shores, boats, and bridges near the larval

haunts are examined early each day in the latter part of June
and early in July, the collector may be rewarded some morn-

ing by finding large numbers in the act of emerging ; and if

these are kept alive and uninjured in a roomy box or cage their

colors and texture will mature in a few days.

KEY TO THE GENERA OF ADULT GOMPHID^ OF THE EASTERN
UNITED STATES.

aa. Both triangles with a cross-vein. II. Progomphiis.

a. Both triangles without a cross-vein, except the discoidal tri-

angle in Hagenius.
bb. First and second anal veins distinctly angulated toward

each other at the cross-vein of the anal loop, which
contains 3 to 5 cells.

cc. Discoidal triangle of usual shape, without cross-vein
;

3d femora not reaching base of 2 ; anal loop 3-celled.

IV. Diastatomma.

c. Discoidal triangle distinctly 4-sided, with cross-vein
;

3d femora very long, reaching base of 3; anal loop

with 3 to 5 cells, usually 4. V. Hagenius.

h. First and second anal veins nearly parallel, or only the

first angulated ; anal loop with 1 or 2 cells.

dd. Part of arculus of hind wing above attachment of sec-

tors much shorter than that below them, the lower

sector joining it at or above its middle; length of

body 40 mm. or less. VI. Lanthus.



53

d. Parts of arculus of hind wing above and below attach-

ment of sectors nearly equal, the lower sector join-

ing it distinctly below the middle.

ee. Hind femora long, reaching base of 3, with 5-7 long

spines besides the usual spinules.

VII. Divitiof/oinphus.

e. Hind femora usually moderate, with spinules but

no long spines. VIII. Gomphus.

KEY TO THE GENERA OF THE NORTH AMERICAN NYMPHS OF

GOMPHIDiE.

AA. Tenth (last) abdominal very long and slender, many times

longer than broad, nearly half the length of the alido-

men. I. Aplii/lla.

A. Tenth abdominal usually broader than long, never more
than twice as long as broad.

aa. Middle legs approximated at base; inner pair of wing-cases

strongly divergent; ventral longitudinal sutures of

abdomen diverging to near apical angles of 8, segment
9 similar to 10; lateral labial lobes entire, their apices

rounded; 8d and 4th antennal joints subcylindric,

more than twice as long as thick, the latter recurved;

five or more pairs of lateral spines; dorsal hooks pres-

ent but not prominent. II. Proyonipliiis.

a. Middle legs at least as far apart as fore legs; ventral lon-

gitudinal sutures of abdomen diverging to apical

angles of 9, which is unlike ID ; 4th antennal joint very

short and small ; not over 4 pairs of lateral spines, ex-

cept in Hayenius.

hh. Inner pair of wing-cases strongly divergent ; 3d anten-

nal joint oblong, at least twice as long as bi'oad ; lat-

eral labial lobes dentate, without apical hook ; tenth

abdominal narrowing posteriorly, following the taper-

ing outline of the apex of the al)domen.

III. DiasfdfoiiuiKi. IV. IIcrpcfoyoDiphiis.

b. Inner pair of wing-cases parallel, lying close together



54

or contiguous, except sometimes in exuviae; sides of

10 subparallel, largely embraced by the lateral spines

of the rapidly narrowing 9th segment.

cc. Third antennal joint flat, subcircular or broad oval,

less than twice as long as broad.

(Id. Body very broad, 3d antenna! joint subcircular,

abdomen much flattened and subcircular, with

distinct dorsal hooks; middle legs more distant

at base than fore legs; lateral labial lobes ob-

scurely denticulate, with broadly rounded apices.

V. Hacjenms.

d. Body of the usual form, 3d antennal joint broad

oval, dorsal hooks entirely wanting; middle and
fore legs about equally distant; lateral labial

lobes dentate, with blunt apices. VI. Lanthus.

c. Third antennal joint subcylindric, more than twice

as long as thick; middle and fore legs about equally

distant; lateral labial lobes toothed.

ee. Dorsal hooks with short but acute spiny tips;

median labial lobe slightly concave; lateral lobes

with prominent apical hook.

VII. DromoffotiqjJn/s.

e. Dorsal hooks sometimes present, but obtusely

pointed, usually absent except for a median tooth

in the hind margin of 9; median labial lobe more
or less convex, in some species nearly straight;

lateral lobes with a more or less evident apical

hook. VIII. GoiiqjJius.

I. Aphylla Selys.

Aj)hylla producta Selys.

Two curious nymphs from Florida are described by Hagen
as probably belonging to this tropical sj)ecies, common in

Cuba. The unusual length of the last segment will at once

identify it.



55

II. Progomphus Selys.

ProgompliHs dbscunis Kamb.

The nymph of this species has divergent wing-pads, like

DIatifdtoiiiiiui, but differs from all our other gomphids in the

proximity of the middle legs. It represents an extreme of

specialization for life as a burrower in the bed of running
streams. It has not been found in the Illinois nor in its adja-

cent lakes, l)ut seems to be common in the sand of the smaller

rivers, such as the vSpoon Kiver and upper Sangamon. The
State Laboratory has a large number of examples from the

latter river, taken in Septemlier and October. One specimen
was found at the lower edge of a sand-bar in Spoon River, near

Havana, in July. The published dates for the nymph are April

(Fla.), May, and June. The adults are rare in collections,

although the species seems widely distributed. Perhaps the

peculiar habits of the nymph may be correlated with a rela-

tively long period of nymphal life and short period of life as an
imago. The adult has been taken in Indiana as early as June
26, and in Illinois June 29 and July 18. It is recorded from
Massachusetts, Georgia, Ohio, Indiana, Florida, Texas, Cali-

fornia, Oregon, and Mexico.

The nymph was described and 1)y supposition connected

with this species by Hagen {'Hib, p. 2-17), and the supposition was
verified by Mr. Needham (97, p. 184).

The nymph measures 31 mm.; alxlomen, 20 mm.; hind femur,

4 mm.; width of abdomen, 6 mm., of head 5 mm.
Body depressed anteriorly and tapering posteriorly ; lateral

margins with long hairs.

Head depressed, sloping anteriorly, cordate, broadly notched

behind ; hind angles rounded. Anteume inserted into cylin-

droid elevations on the front, depressed and incurved so as to

almost surround the pilot-shaped labrum ; two basal joints very

short; third, twice as long as the two basal combined, slightly

flattened and upcurved at the tip ; fourth joint small, one third

to one fifth as long as the third, slender antl strongly recurved.



56

Labium rather small, reaching, when folded, to the bases

of the middle legs ; submentum shortened ; mentum narrowed
at its proximal end, its median lobe prominent, rounded, fringed

with a row of flabellate scales whose bases are overlaid and
supported by another series of shorter semicylindrical scales

;

beneath this fringe, the margin cut into a series of obscure

rectangular teeth ; lateral lobes short, nearly straight, unarmed,
rounded at apex ; movable hook stout, moderately incurved,

and tapering.

Thorax sloping to the head and to the bases of the legs
;

prothorax of unusual dimensions on the dorsal side, its hind

margin on a line with the bases of the hind legs, being extended

back upon the other thoracic segments, shield shaped, with a

short collar close behind the head. Wing-cases strongly di-

vergent. Legs conspicuously fossorial, fore legs approximate
to the sides of the head, bearing shields of stiff hairs behind

which the middle legs may be brought forward. Middle legs

approximated on the venter, rotated downward and extended

horizontally close under the fore legs. Hind legs longer, more
nearly normal, directed posteriorly. Pore tarsi with soles fac-

ing laterally ; middle tarsi rotated on tibiae so as to point back-

ward ; hind tarsi elongate, the third segment about as long as

both basal segments, its claws sharp and long ; claws of fore

and middle tarsi short and blunt. Each femur with a distal

anterior process which rests against and supports the tibia

when moved backward.

Abdomen spindle-shaped, segments about equal, the 9th a

little longer than the others; dorsal hooks variable, rudimen-

tary, more or less well represented on segments 2 to 9, small-

est on middle segments. Lateral spines on 5 to 9, on 5 rather

minute. Appendages slender, tapering; superior and inferiors

equal, about one third longer than segment 10, laterals about

half as long as the others.



57

ITT. DiASTATOMMA Burm.

Ojiliioi/diiiplnn^ Selys.

IV. Herpetogomphus Selys.

Diasfatoniiiia and Ho-pcfof/oiHjt/iiis are very closely related,

and no good character has jet l)een found to separate them
when in the nyniphal stage. They are then similar in form to

the common gomphid larv* of the genera which follow, but
may be easily distinguished from them all Ijy the widely diverg-

ing wing-pads, the inner margins of which separate at an angle

of (50" or more instead of being approximately parallel as in

the genera which follow. In exuvi* of the latter forms, how-
ever, the wing-pads are often more or less separated. Her-

pefcH/oiHplnis has not been found east of the Mississippi. Dias-

tittonnnu is widely distributed in North America, but both the

nymphs and imagos are quite rare in Illinois, owing prob-

ably to the small extent of rocky stream-beds and rapid cur-

rents. The rivers near Rock Island and Golconda, in which
examples have ])een found, afford favoral)le situations of this

kind. Three species of the nymphs are in our collections,—one
found in Illinois,—readily separable by the number of lateral

spines, the length of the lateral appendages, etc., as stated in

the key.

Nymph stout, little flattened. Head abruptly sloping for-

ward from the ocelli. Labrum pilot-shaped. Antennae with

the twt) Itasal segments glojjular, third segment twice as long

as both basal, much flattened and laid close beside the labrum.

Fourth joint a minute rudiment. Median lol)e of mentum
rounded, with border of short blunt teeth and a doulile series

of fringing scales. Lateral lobes nearly straight, not terminat-

ing in an end hook and minutely denticulated within ; movable
hook short, arcuate.

Legs rather short. Fore and middle ti])iiB with external

hooks, wing-cases divaricate, strongly sloping downward toward

the sides. Dorsal hooks on abdominal segments on 2 or 3 to 9.

Tenth segment not inclosed by the 9th but triquetral, exceeding



58

the lateral spines, its own lateral margin forming a part of the

margin of the abdomen.

KEY TO THE NYMPHS OP DIASTATOMMA HEREIN DESCRIBED.

Lateral spines on abdominal segments 7-9.

Dorsal hooks distinct, narrowly prominent, ending in a pointed

tooth ; abdominal granulations fine and much less con-

spicuous than in the next species; male superior append-

age scarcely tubercled, lateral appendages three fourths

to four fifths as long. 1. severus.

Dorsal hooks represented by a broad elevation, ending poste-

riorly on each segment in a rounded tooth; abdomen
evenly dotted with rather coarse blackish granulations;

superior appendage of male notched back of middle, in

front of which is a pair of tubercles, laterals about three

fifths as long as the superior one. 3. carolus.

Lateral spines on abdominal segments 6-9; dorsal hooks as in

carolus but much more prominent; male superior append-

age scarcely tubercled, laterals nine tenths as long as

the superior. 4. sp. («).

1. Diastatomma severiis Hagen.
Ophiogomphus severiis Hagen.

The nymphs here described were collected by Dr. Forbes

in large numbers from streams in Yellowstone Park, such as

the Fire Hole River, Nez Perce Creek, and Gibbon and Good-

win rivers, during the latter half of August. They were found

on sand under stones in shallow rapids, and on weedy and

grassy bottoms. They answer well to the description given by

Hagen ('85, p. 259) for severua, which is by far the most abun-

dant gomphid in that region.

The nymph, apparently grown, measures 25 mm.; the

abdomen, 16 mm.; hind femur, 5 mm.; width of abdomen 7 mm.,

of head 5 mm.
Color (alcoholic specimens) fulvous yellowish beneath and

on the sutures, and on the legs beyond the middle of the fem-



59

ora. Yellow also on ocelli, in a broad band between the eyes

behind the suture, another band on the mid-dorsal line of the

prothorax, and sometimes still other bands across the apical half

of abdominal segments 7 and 8; also variable spots arranged

along the sides of the al)domen. Abdominal appendages yel-

low, brownish on the margins.

Body but little hairy. Lateral spines on abdominal seg-

ments 7-9, all equal. Dorsal hooks on 2 to 9, on 2 slender, on
3-9 subcultriform, on 9 low and directed posteriorly. Other-

wise much like nymphs of the other species of Diustatomma.

2. DIasifafoiiuiui riipinsidensis Walsh.

This species was first described from Rock Island, 111. The
only other instance known to me of its capture in this state

was on June 29 at a point on the Mackinaw River known as

Mackinaw Dells, where favorable conditions for the nymph
exist. Kellicott says the imago was first taken in Ohio May 5,

near Columbus, and was common from the middle of May to

the middle of June, flying above the swiftest currents or rest-

ing near rapids. It is recorded from Canada, and from a num-
ber of the northern states from Maine to Illinois. The nymph
is as yet unknown.

3. Didstatomma carolus Needh.
Ophiogoinphtts caro/us Needh., 'q7.

This species flies in May at Ithaca, N. Y., where Mr. Need-

ham bred it by scores in 1S97 and picked up hundreds of exuviae

by the banks of streams. Nymphs taken in October emerged

in March. In April they were abundant in the trash-filled

eddies of the swiftest streams, and after their emergence the

banks were in places fairly covered with exuvise. Notwith-

standing all these evidences of al)undance, he was able to cap-

ture but a single imago there, and saw but two at large, during

a whole season of active collecting. It would be very inter-

esting to know what the imagos do with themselves.



60

The nymph measures 24 mm.; the abdomen, 14 mm.; the

hind femur, 4 mm.; width of head 5 mm., of abdomen 7.5 mm.
Body sparsely covered with clavate hairs and minutely

pointed blackish granulations; 3d joint of antennae very flat,

oval. Labium short, meeting in front a posteriorly directed

fringe of haii'S growing under the edge of the labrum. Lateral

spines on 7 to 9, those of 9 half as long as segment 10 at mid-

dle. Dorsal hooks on 3-9, very low and broad with decurved

apices. Latex'al abdominal appendages al)out three fifths as

long as the subequat superior and inferiors,

3. Diasfafomma sp. [a).

Nymphs taken by Mr. Hart from gravelly shallows of the

Ohio River at Golconda, in southern Illinois, October 27, differ

from those of caivlus in having lateral spines on the 6th

segment of the abdomen.
The nymph measures 25 mm.; abdomen, 15.5 mm.; hind

femur, 4 mm.; width of abdomen 8 mm., of head 5.2 mm. Color

(in alcohol) yellowish with black points at top of thoracic

sutures and a pair each side of each dorsal hook ; scars brown-

ish. Body moderately hairy, lateral spines on abdominal seg-

ments 6 to 9. Dorsal hooks on 2 to 9, large and subcultriform

in front, regularly decreasing to a rudiment on 9 ; superior and

inferior abdominal appendages long, laterals very little shorter.

V. Hagenius Selys.

There is but one North American species of this interest-

ing genus, the flat nymph of which is conspicuously unlike that

of any other dragon-fly in form, as described below. It is found

clinging to the flat surfaces of driftwood and stones or amongst
dead leaves. The EjjJiemeridce inhabiting such surfaces are also

conspicuously flattened. They are therefore not easily picked

off by fish nor dislodged by a rush of water, and are able to slip

away through narrow crevices. Perhaps Hagenius has this

shape for similar reasons. Its blackish color is evidently pro-

tective.



61

The imago is one of the largest gomphids. It is black, with
clear wings and yellow markings, and flies along the small

rapid streams which are inhaliited l>y the nymph.

Hageuitis birristi/liis Selys.

This grotesque nymph looks more like a colossal bedbug
than a dragon-fly nymph, being extremely flat and broad, with

a disk-shaped subcircular abdomen. It is unique in habits as

well as in form. It seems to prefer rapid-flowing streams, and
does not burrow as do other gomphid nymphs, but clings to

stones or to tethered drift-stuff in the current, or hides among
loose drift and dead leaves. Mr. Needham found nymphs of

three distinct sizes in a stream near Ithaca. The imagos
emerged all at once, and a considerable number of exuvia:> were
found. This would indicate a three-year life-period, the nymphs
not maturing till the third season after the one which brought

them into existence. We have examples taken from the upper

Sangamon River in Piatt county in September and October,

and at Chicago and Quincy. This nymph seems to be quite in-

frequent in Illinois. The imago occurs all over the country

from Maine to Texas and Kansas, and is reported ( Kellicott '95)

as abundant in Michigan, but 1 am not aware that it has ever

been taken in this state. It flies through June, July, and the

gi'eater part of ^Vugust.

In view of the excellent figures and description by Cabot

('72, p. 9) and the very minute description by Dr. Hagen ("S5,

p. 279) it will be sufficient here to give only the more promi-

nent characters.

The nymph measures in length 36 to -lO mm.: hind femur,

10 to 12 mm.; abdomen, 22 to 23 mm.; width of abdomen 20 mm.,
of head 8 mm.; vertical thickness of abdomen about 4 mm.

Body exceedingly flat, ovate in outline, nearly destitute of

hairs.

Head cut off squarely behind: eyes a little projecting on

the sides, front sloping to base of anteuute. Top of head with

a pair of conic tubercles behind the lateral ocelli and a larger



62

pair of flattened ones behind the eyes. Antennae with the

basal joint globular, the second smaller; third joint very flat,

nearly circular, but with the inner edge straight. Below the

eye a blunt process extends downward and forward to the sides

of the labium, obviously for lateral support of that organ.

Labium short and thick; mentum slightly wider than

long, contracted at basal fourth; median lobe of mentum occu-

pying hardly a third of its width, slightly rounded, its thick-

ened edge obscurely cut into about ten blunt teeth, and bear-

ing a fringe of flat scales, with a few more elongate and
bristle-like at the ends of the fringe; lateral lobes arcuate,

broad and strong, with ill-defined blunt denticulation all around

the rounded tip and down the inside; movable hook short,

feebly arcuate, tapering, with an incurved tip.

Prothoracic dorsum elevated at sides into prominent com-
pressed lateral ridges, between which it is excavated. Femora
inclined to be sharp-edged posteriorly and triangular in cross-

section.

Abdomen with dorsal hooks beginning on 2, at first narrow
and acute, highest on 3, becoming gradually more elongate

and blunt, variably reduced on the last three or more segments

to low median ridges; lateral spines of 2 acute, those of 3-9

appearing as broad triangular projections of the latero-posterior

angle, those of 9 inclosing segment 10. Appendages short and

thick, triangular-pyramidal, longer than the very short 10th

segment. Superior slightly shorter than the inferiors, laterals

one third as long. Longitudinal ventral impressions of abdo-

men separated by only about one fifth of the width of the

abdomen, nearly parallel, becoming strongly divergent on 2

and 9. Young nymphs have the peculiarities of the mature

ones even more .strongly marked, especially in the form of the

abdomen, which is more nearly circular.

VL Lanthus Needh.

The nymph of the single eastern species for which this

genus was established, differs from the ordinary Gomphus nymph



63

in several particulars, and is related to Hagenius by its anten-

nal structure. It was found burrowing in sandy stream beds.

Laiithiis 2)arn(Iiis Selys.

Gomphu: parvuhis Selys.

Uropelala thoreyi / Hag. (nymph).

This dainty little gomphid is not yet known to be an Illi-

nois species. The imago has been found only from Nova Scotia

to Pennsylvania, but Dr. Hagen ('85, p. 281 ) has described the

nymph from specimens received from Kentucky October 20;

doubtfully referring them to TarJioptrrifj- thoreyi. Nymphs
answering to his careful description were found sparingly by

Mr. Needham in the sandy beds of streams about Ithaca, N. Y.,

where he collected and reared half a dozen. The nymph is

notable for its small size and the flat subcircular form of the

third antennal joint. It is an active little burrower, and has

the habit of feigning death when withdrawn from the water.

For these reasons it is more ditticult to detect than are other

gomphids. Nymphs kept under nearly normal conditions trans-

formed in May. The imagos were not seen by him at large.

They have been taken in Maine by Miss Wadsworth June 19

and 25.

The nymph measures 21 mm.: abdomen, 12 mm.; hind

femur. 3.5 mm.: width of head 4 mm., of abdomen 6 mm.
Body moderately depressed, a little hairy on sides of thorax

and on legs.

Head compact, with obtuse hind angles between which the

hind margin is concave. Antenna? with the two basal joints

short and thick, the third very flat, broadly oval, almost circu-

lar, the fourth very rudimentary.

Labium short : mentum with sides parallel except in the

abruptly narrowed basal fourth: median lobe almost straight,

with four blunt chitinous teeth and a fringe of long scales;

lateral lobes short and stout, rounded off on the outer angle so

that the inner border is not arcuate ; teeth of the inner border

prominent, acute, recurved, somewhat divergent ; movable hook
short and stout.



64

Legs somewhat hairy ; tibial hooks on the anterior and
middle pairs well developed. Wing-cases reaching the middle

of the 4th abdominal segment. •

Abdomen ovate, widest across the 7th segment and nar-

rowed rather suddenly upon the 9th. No dorsal hooks at all,

but a faint impressed median line on the anterior segments;

lateral spines on 8 and 9 small, angular ; 9th abdominal seg-

ment longer than the 8th and three times as long as the 10th.

Appendages longer than the 10th segment, stout pyramidal,

laterals one half to one third shorter than the others.

VII. Dromogomphus Selys.

One of the three species of this American genus has been
found in Illinois. The nymph may be known by the median
ridge on the 9th abdominal segment, ending behind in a sharp

spine. The imagos resemble those of Gomphus.

Dromogomphus spinosus Selys.

Gomphus sp., Hagen, '85, p. 265, No. ig (nymph).

This appears to be quite rare in Illinois, and we do not

know that any one has ever taken the nymph within our boun-

daries. Walsh records the imago from the Des Plaines River,

near Chicago; and Mr. Adams, from Kankakee, July 6. It has

been taken in a number of states between Maine and Florida

on the east and Texas and Illinois on the west. Kellicott

found it common in Michigan July 15 to August 1, and gives

the following interesting notes: " The female was several times

found ovipositing in a manner similar to Macromin illhioiensis,

that is, by skimming the water and every few feet or rods

touching the water with the abdominal tip, scarcely checking

her speed; at other times I have seen them drop down from

an overhanging tree and repeatedly tap the water, remaining

in one place after the manner of Lihcllidci. Pairs were noticed

to fly up into tree-tops, and remain in union for a considerable

time." The earliest date is June 5, given by Williamson.

Miss Wadsworth observed an individual devouring a Calopteryx

maculata.



65

The nymph has lieen bred by Calvert, and images, nymphs,
and exuviaB were collected by Mr. Needham—all three at the

same time and place—near Ithaca, N. Y., in June. Hagen's
nymph No. 19, from Michigan, is probably this species.

The nymph measures 34 mm.; abdomen, 22 mm.; width of

head 5.5 mm., of abdomen 7 mm.
Body elongate, nearly smooth. Antennae a little more

approximated than in related genera, the 3d segment more
upcurved at the tip. Lal)ium moderate, mentum slightly

widened all the way to its apex; median lobe a little concave,

with minute median tooth and a sparse fringe of flat hairs

either side of the tooth; lateral lobes short and stout, the

apex incurved in a stout hook which considerably exceeds the

8 truncate teeth of the inner margin. Tibial burrowing hooks
strong.

Abdomen cylindric, becoming sharply triquetral in cross-

section posteriorly. Dorsal hooks on segments 2 to 9, on 2

rudimentary, but becoming well developed posteriorly, and on 9

straight, almost equaling the lateral spines and confluent ante-

riorly with the sharp mid-dorsal ridge. Lateral spines on seg-

ments 6 to 9 increasing in size posteriorly. Appendages a little

longer than the 10th segment, their apices bent a little so as

to be divergent; laterals a little shorter than the others.

VIII. GrOMPHUs Leach.

The nymphs of this large and difficult genus represent the

typical and most almndant form of the family. The wing-pads

are not divergent, as in FriMjoinphii^i, HerpctoyomiiliKS, and Dins-

tatonniKi, nor the third antennal joint subcircular, as in Hage-

nius and LaiifJiKs; nor are the abdominal segments armed
with spinose dorsal hooks, as in DroiHououqjhiis. Their habits

are very similar to those of the rest of the family, with the

exception of Huyenius, as described fully under the family

heading. Observations by Mr. Needham at Ithaca and by Mr.

Hart at Havana indicate a nymphal life of more than one



66

year; iu the case of G. villosipes, of two years. The imagos are

all much alike, of medium size, with clear wings, the triangles

normally without cross-veins, the thorax with green or yel-

lowish stripes. They disappear rapidly after transformation,

and are not usually common.

KEY TO THE KNOWN NORTH AMERICAN NYMPHS OF G0MPHU8.

AAA. Four pairs of lateral spines ; lateral labial lobes rather

finely toothed, with 6 to 12 teeth ; side margin of 9

spinulose or denticulate, 10 (last) wider than long.

{Gomphus Needh.)

aa. Dorsal hooks represented more or less on 2 or 3 to 9.

bb. Dorsal hooks distinct, each produced at the hinder mar-
gin of its segment in a pointed median tooth ; side

margins of 9 spinulose.

cc. Lateral spines of 9 about as long as segment 10 ; 9

thrice as long as 10. 1. sp. {a).

c. Lateral spines of 9 reaching about half way to the api-

cal angles of 10 ; 9 twice as long as 10.

2. (irasUiiellus.

b. Dorsal hooks feebly elevated, only the last two or three

forming median teeth in the hind margin of the

segment ; side margins of 9 obsoletely denticulate.

3. descriptus-

a. Dorsal hooks represented only by median teeth in the pos-

terior margin of 8 and 9 ; a fine smooth median line on

the preceding segments ; 9 spinulose laterally, 10 very

short.

dd. Lateral spines of 9 distinctly shorter than the distance

to the median tooth of the segment; labial apical

hook sharp, projecting beyond the teeth.

ee. Lateral spines of 9 broad, scarcely longer than those of

8 ; length of ventral surface of 9 at middle a little

more than one third its basal width ; abdomen less

than twice as long as broad. 5. abbreviatus.



67

t'. Lateral spines of 9 about twice as long as those of S
;

length of ventral surface of 9 at middle about one
half its basal width ; abdomen more than thrice as

long as broad. 6. vastus,

d. Lateral spines of 9 slender, elongate-acuminate, incurved
apically, nearly or quite as long as the distance to the

median tooth of the segment ; labial apical hook not
projecting beyond the teeth. 9. extenms.

A A . Three pairs of lateral spines, the first pair very small

;

teeth of lateral labial lobes mostly distinctly separated

by rather deep incisions, and obliquely truncate; medi-

an lobe rather prominent, a small projecting tooth at

middle ; side margins of 9 entire, more or less fringed

with soft hairs, 10 (except in sptcdfus) as long as, or

longer than, wide ; only a trace of a median posterior

tooth on 9. {Arigompkus Needh.)

ff. Last three segments evenly tapering, length of ventral

surface of 9 at middle about three fourths of its basal

width, 10 wider than long ; traces of dorsal hooks

;

labial apical hook prominent, toothed on its inner

side l)ut not appearing ])ifid. 10. splrnfiis.

/'. Last three segments concavely tapering, length of ventral

surface of 9 at middle about equaling its basal width,

10 longer than wide ; a low scurfy median abdominal
ridge.

gff.
Labial apical hook as in sjiicafKs ; lateral spines of 9

not over one fifth as long as the inferior append-

ages. 11. (I lintrails.

g. Labial apical hook and outermost tooth united to form

a prominent bifid apical tooth.

////. Lateral spines of 9 not over one fourth as long

as posterior appendages. 12. villosipes.

h. Lateral spines of 9 about as long as inferior append-

ages. 13. pa 11id IIS.

A. Pour pairs of lateral spines (minute in .'ipiiiireps); labial api-

cal hook large and strongly incurved ; teeth few, 0-5,



68

usually 2 or 3, broad and subtruncate, separated by

deep narrow incisions ; side margins of 9 entire, often

fringed with soft hairs, as in A A, 10 not longer than wide.

a. Dorsal hooks very small, pointed on 6 to 9; lateral labial

lobes 4-toothed. 14. exilis.

i. Dorsal hooks wanting, except usually a distinct median
tooth in the hind margin of 9; a fine smooth median
abdominal line; labial lobes rarely more than 3-toothed.

JJ. Length of ventral surface of 9 along median line not

more than half its width at base.

kk. Dorsal hooks entirely wanting; lateral spines of 9

as long as segment 10. • 16. sp. {b).

k. A broad triangular flat dorsal median tooth in the

hind margin of 9, as long as segment 10; lateral

spines of 9 twice as long as 10. 17. sp. (c).

J. Length of ventral surface of 9 more than half its width

at base; apical hook of anterior tibiae very small.

{Stylurus Needh.

)

III. Ventral surface of 9 distinctly shorter than its basal

width ; lateral spines of 9 about one third as long

as their distance from the median dorsal tooth of

the segment; median tooth minute. 18. amnicola.

II. Ventral surface of 9 about as long as its basal width

;

lateral spines of 9 at least two thirds as long as

the distance to the median tooth. 19. phigiatus.

I. Ventral surface of 9 twice as long as its basal width.

21. sphiiceps.

1. Gomphus sp. {a).

Gomphiis sp., Hagen, '85, p. 264, No. 17 (nymph).

This species is known only from a single nymph described

by Dr. Hagen from Indiana. It may be distinguished by the

following characters:

Length 30 mm. Lateral lobes of labium with a sharp bent

apical hook and ten strong teeth ; median lobe straight ; seg-

ment 9 longer than 8, three times as long as 10 ; lateral spines



69

on 6 to 9, sharp, the last pair as long as 10 ; lateral margins of

7 to 9 serrate : dorsal hooks well marked on 2 to 9, shai'ply

pointed ; appendages as long as last segment.

2. Gohqjliiis (jraslhwllns Walsh.

Imagos of this species have heen reported from the states

of Washington (Hagen '85. p. 264), Indiana, and Ohio. The
State Laboratory has a number of the nymphs fi-om the shal-

lows of Cedar and Sand lakes, in northeastern Illinois, taken in

June, August, and October, many young occurring in the latter

month. It is common in ponded waters near Urbana, but appar-

ently does not occur in the liottom-land lakes and streams about

Havana. Mr. Needham studied the species at Puriugton Lake,

near Galesburg, III., in 1S95, collecting and rearing a large num-
ber of nymphs. His notes are as follows: "Very early in the

spring I ol)tained from steep clayey lianks in the lake nymphs
lacking one and two molts of maturity. Placed in my aquaria

these nymphs quickly descended into the mud of the bottom un-

til only the upturned tip of the abdomen remained exposed. then

burrowed along parallel to the surface, leaving a shallow groove

to mark their course. When out of the mud they crawl stiffly

and very slowly, but swim fairly well under compulsion, by ex-

pulsions of water in the usual manner. Transformation takes

place commonly between dayltreak and sunrise. The nymph
usually crawls but a few inches from the water, and appears to

transform oftenest while lying Hat upon the bare earth or upon
tangled mats of dwarf clul>rush and other semiaquatics. Ex-

uviae picked up in such places are usually incrusted with mud.
The species appears on the wing in Illinois about the 20th of

May. and Mies for about a month. Imagos were oftenest seen

when Hushed from the grass or from the bare paths which ter-

raced a steep hillside beside the lake. Females were little in

evidence. I took one ovipositing and obtained in a tumbler of

water an immense number of eggs."

Other Illinois localities for the imago are Coal Valley Creek

(Rock Island county), Des Plaines and Chicago Rivers; Chicago,

Bloomington, and L^rbaua,—the dates ranging from late in May



70

to the early part of July. They were observed emerging

numerously at Urbana on May 27.

The nymph was described by Hagen ('85, p. 264) and con-,

nected with this species by supposition.

The nymph measures in length 32 mm.; abdomen, 21 mm.;

hind femur, 6.5 mm.; width of abdomen 8 mm., of head 6 mm.
Body depressed. Color yellowish brown with darker mark-

ings on sides of thorax, on wing-cases, and on median dorsal

area of abdomen. Lateral margins of body and appendages

fringed with luteous hairs, except on tapering posterior end of

abdomen. Scars prominent.

Head compact, notched behind; hind angles broad.

Labium stout, short, not extending posteriorly beyond the

bases of the fore legs. Mentum flat, broad, slightly narrowed

at base; median lobe very slightly rounded, margined with a

dense comb of flat bristling hairs, lateral lobes arcuate, nar-

rowed to an incurved tip within which are about nine teeth on

the inner margin. Movable hook long, stout, incurved, and

rather bluntly pointed.

Abdomen widest across segment 6, acutely narrowed to the

tip; segments 3 to 8 of about equal length, 9 one half longer

than 8; 10 cylindric, less than half the length of 9; short lat-

eral spines on 6 to 9, increasing in length posteriorly, those of

9 reaching slightly beyond the middle of 10; small dorsal hooks

on segments 3 to 9, best developed on 4 to 6. Superior and

inferior abdominal appendages longer than segment 10; laterals

shorter than the others.

Younger nymphs lacking two or three molts of maturity

do not differ in any essential character. A quite young nymph,

8 mm. long, has dorsal hooks on 8 and 9, well marked only on

9, with merest rudiments on the middle abdominal segments.

The four pairs of lateral spines are all present.

3. Gomphiis descriptus Banks.
G. descriphis Banks, '96.

This was recently described from specimens taken in New
York May 15-21. Williamson's statement that it occurs in



71

Illinois is based on an erroneous determination. The only

nymphs of this species known to us were found by Mr. Needham
in New York. It abounds in all the streams about Ithaca, trans-

forms in May by thousands, and scatters to the woods. Many
imagos were seen l)y liim foraging about the borders of upland

woods and going to rest at night among the branches of trees,

but he did not observe their breeding habits.

Transformation takes place generally before daylight and

within a few inches, or at most a few feet, of the water's edge.

The nymph measures 29 mm.; aljdomen, 18 mm.; hind

femur, 7 mm.; width of head 5.5 mm., of abdomen 8 mm. Body

very flat and hairy; legs and antennae and lateral margins of

head very hairy. Head depressed and wedge shaped.

Labium short; mentum squarish beyond the narrowed

basal third; median lobe nearly straight; lateral lobes very

arcuate, ending in a distinct hook on the inside of which is a

very broadly pyramidal tooth; proximo lly the hiteral lobe l)eai-s

on its inner side a series of 7 to 1) short, broad, slightly recurved

teeth; movable hook long and strong.

Abdomen lancet shaped; scars above and ganglion pockets

beneath very plainly marked; segments 2 to 8 about equal in

length, 9 one half h)nger, 10 half as long as 9; dorsal hooks

reduced to broad median swellings on segments 2-9, not hooked

nor pointed; lateral spines on 6 to 9, those of 9 about one half

as long as 10. Superior and inferior appendages one third

longer than 10, laterals a little shorter.

4. GoiHphiis ijiKuh'Ifuloi'Waliih.

This imago is seldom captured in Illinois. The nymph is

unknown. Walsh's examples were taken at Eock Island. Mr.

Adams took the imago at Bloomington June 23. Kellicott

made three captures in Ohio, May 20 to June 15; Williamson

found one June 2(5 ; and it is also on record from Indiana, Michi-

gan, and Massachusetts. It rests, according to Kellicott, on

rocks projecting from rapids, or on the banks near by the

most rapid parts of large streams.



72

5. Gomphus ahhreviatus Hag.

The nymph was collected and bred by Mr. Needham at

Ithaca, N. Y., and is here described for the first time. It is not

an Illinois species, and seems quite rare and limited in range,

imagos being recorded only from Maine, Massachusetts, and
(with doubt) Pennsylvania. The only date found recorded for

the imago is June 10 (Harvey).

The nymph measures 24 mm.; abdomen, 15 mm.;- hind fe-

mur, 5.5 mm.; width of head 5.5 mm., of abdomen 8 mm.
Body very flat, very scantily hairy on the margins. An-

tennae with the third segment flattened and widened toward
the tip. Labium short ; mentum narrowed in its basal foui-th,

beyond which the sides are parallel, median lobe a little con-

vex and with a median marginal tooth in the midst of the

fringing flattened hairs ; lateral lobes regularly narrowed and
incurved to a point which is a little longer than the teeth on

the inner border ; teeth about eight, quadrant-shaped, pointed

at the lower apical angle, successively decreasing in size prox-

imally.

Abdomen flat oval. Conspicuous lateral spines on seg-

ments 6 to 9, increasing in stoutness posteriorly, the last about

as long as segment 10. Dorsal hooks represented by minute
rudiments on 8 and 9 ; a smooth median line anteriorly to this.

Appendages yellow, almost twice the length of the 10th seg-

ment, the laterals a little shorter than the others, the superior

a little bent upward at the extreme tip.

6. Gomphus vastus Walsh.

Walsh found this species emerging in company with no-

tahis Sindfi-aternus "in considerable numbers" at Rock Island,

on the Mississippi, and described the imago as a new species

('62, p. 391). Riley (Hagen, '85, p. 265) also reared it from the

Mississippi River at Hannibal, Mo. (not "111."). The nymph is

in Dr. Kofoid's private collection, taken in deep water in De-

troit River. The imago is reported from Illinois, Iowa, Indiana,

Michigan, Ohio (common in July), Pennsylvania, and several



73

Atlantic coast states, but seems less abundant iu Illinois than

some other species. Walsh notes an imago taken June 16 eat-

ing one of the larger caddice-flies. Macronema zehratiim. Kelli-

cott (^"96, p. Ill) describes the habits of the imago as similar to

those of fyaternus, and says they copulate at rest in trees or

shrubs. The time of flight seems to be from the latter part of

May through July. Our Illinois imagos are all fi-om the vicin-

ity of the Mississippi along the northwestern border of the

state. Williamson says they fi-equent the larger rivers and lakes,

the males being more often found near the water's edge, the

females in fields or along roadsides.

The nymph has lieen described and figured by Cabot (^72,

p. 3, PL II.. Fig. 4 I and again by Hagen ('85, p. 265). The fol-

lowing brief de.scription is condensed from that of Dr. Hagen.

The nymph measures 31 mm. in length.

The mentum of the labium is one third longer than broad;

the front border of the median lobe is straight, fringed with long

flat scales : the lateral lobes end iu a long sharply incurved

point, inside which the toothed inner margin is concave. Ab-
domen slowly tapering posteriorly, in outline rather bluntly

pointed. Ninth abdominal segment three times as long as the

10th, and a little longer than the Sth. Lateral spines on the

6th to the 9th segments, those of the 9th as long as the 10th

segment. Dorsal hooks on the 8th and 9th segments. Ap-

pendages nearly twice as long as the 10th segment.

7. (toiiijiIiks rcntricosus Walsh.

A widely distributed but rare species in collections. It was
originally described from Rock Island. 111., and has since been

listed from Michigan. Massachusetts, and Virginia. The nymph
is unknown.

8. GoiHjjhus fraternus Say.

This species and the following (fj7('/-«'/«) are closely related

but specifically distinct. The nymphs of frutenius, if any are

iu our collections, have not been distinguished from those of

extenius, which is a common species at Havana. Dr. Hagen's



74

No. 17 ('85, p. 264) which he thought might he fntferniis is

probably not. Walsh states that he found this with other

species emerging in considerable numbers on the Mississippi

at Rock Island. Kellicott says that it is common in Ohio from
May to the first part of July, and flies about the swiftest and
most turbulent parts of the river ovipositing in rapids and
rough waters. The State Laboratory has a single imago taken

near the upper Sangamon, at White Heath, May 18. Fraternus

is chiefly eastern in its distribution; externus, western. In ad-

dition to the localities already given, the present species is listed

from New York, New Hampshire, and Virginia ; Michigan, In-

diana, and Arkansas.*

9. GompJiiis exicrnus Selys.

G. consobrinus Walsh.

G. adelphiis (suppos.) Hagen (nymph).

The nymph of externa is readily recognized by the breadth,

form, and large latei'al spines of the ninth segment, and by the

entire absence of dorsal elevations or hooks except a rudimen-

tary posterior tooth on the eighth and ninth segments. This

species and G. nofafiis are the most abundant gomphids in the

field of the Biological Station at Havana. They are usually

found in company, burrowing, mole-like, in the soft mud bot-

tom wherever a sufficient current is maintained at all stages of

water, and becoming especially numerous in sheltered areas

where there is an accumulation of fine trash on the bottom.

They are quite common throughout the year in muddy flats

about the point where the waters of Quiver Creek become lost

in Quiver Lake, and along the line of the channels of the

Illinois and Spoon rivers, especially in the narrower part of the

Illinois down from the mouth of the Spoon. They are seldom

seen at any of the lake stations. Externus is also represented

in the State Laboratory collections from the vicinity of the

Mississippi River at Quincy, where it was common in the muddy
side passages and the adjoining narrow slough-lakes; from the

*The images described as C/,i-a/^/-««i waAV/zV by Kellicott ('99) have been

determined by Calvert ('01) to be crassus. See also foot-note on next page.



75

Rock Rivei', iu novtheru Illiuois: from the Sangamou. in the

central part of the state: and fr'oiu the Big Muddy at the south.

Transformation was first observed May 21. Mr. Xeedham's
notes of his observations on this species wliile at Havana are

as follows: " Transformation takes place in the night or very

early morning. The nymph crawls but a little distance (one to

two feet I out of the water liefore fixing itself for emergence. I

have found exuvi* on liridge piers and on willow stumps, and
have taken imagos emerging iu such places at aliout six o'clock

on several mornings."

The strong-flying imago was not uncommon at Havana
during July, 1896. The males were frequently seen chasing

each other over the open river, or sitting at rest on the sand at

the bank with the alxlomen elevated and the wings declined

until their tips touched the sand, in a position of great alert-

ness. The females fly less openly. One female captured in the

weeds at the bank, deposited for me in a watch-glass of water

in a few minutes' time altout 5.2n() eggs. This number is an
estimate from a partial count.

An observation on the food of the imago was made l)y Mr.

Needham, who found on shore a female fniteniiis engaged in

eating a teneral imago of Mrsofhemis siniji/iricallis.

Examples of the nymphs collected in June were placed in

breeding-cages immersed in the water of Quiver Lake beside

the field laboratory of the Biological Station. They remained
without transforming, and at the close of the season's work,

September 28, seventeen nymphs, nearly the original number,
were still alive in the cage.

About the middle of July, after a severe rain storm, sev-

eral dead males of this species were picked up from the guards

of cabin-boats along the Havana river-front.

The imago is reported from Illinois. Nebraska. Kansas,

Texas, and New Mexico. In Illiuois it has been taken in Hen-
derson county, as well as at Havana and Rock Island.*

According to Calvert, the imagos described as exUrnushy Kellicott I'gg) and
Williamson foo) are G. crassits Hagen, a species not yet found in Illinois, which is

quite unlike extemus in the form of the terminal appendages of the male.



76

The nymph measures in length 32 mm.; abdomen, 20 mm.;
hind femur, 6 mm.; width of head 5.5 mm., of abdomen 8 mm.

Body stout, only moderately depressed. Lateral inargins

of body and appendages hairy. Ti])ial hooks very prominent.

Labium short and stout ; mentum a little longer than broad

and narrowed at basal third ; median lobe very faintly rounded;

lateral lobes short, thick, and not strongly arcuate, ending in

two teeth which are hardly distinguishable from the five to ten

other teeth which extend in a diminishing series down the in-

ner margin; movable hook short, stout, tapering, and regularly

curved to its tip.

Abdominal segments 3-8 about equal, 9 one half longer, 10

very short, one third as long as 9 ; a smooth median dorsal line

ending on 7 ; rudimentary dorsal hooks on 8 and 9 ; lateral

spines on 6 to 9, incurved at tip, those of 9 about twice the

length of segment 10. Superior and inferior abdominal ap-

pendages twice as long as segment 10, laterals a little shorter

than the others.

Younger nymphs dredged from the bed of the stream differ

only in size and in the shortness of the wing-cases.

This nymph agrees in every point with the very careful

description given by Hagen ('85, p. 262) for " Gomphns adelpJuis

(supposition)."

10. Gomphus spicatus Selys.

In the collections of the State Laboratory are many
examples of a nymph resembling graslinellns, all taken from
shallow waters in Sand and Cedar lakes, northeastern Illinois,

associated with that species, during June, August, and October.

Dr. Ward has sent me another of the same species from a lake

near Charlevoix, Mich. These nymphs we can properly assign

to spicatus, imagos of which were collected at the same place,

and which is one of the very few Illinois gomphids whose
nymphs yet remain unknown.* The nymphs from Cedar Lake

*Mr. Needham has since verified this supposition by breeding.



77

referred by Hagen to jxil/ii/iis ('85, p. 267) are of this species.

Iinagos of spirafiis were taken by Mr. Hart at Cedar. Sand, and
Fox lakes, June 15-22, 1892. This is its first record from
Illinois, it having been previously listed only from Canada,
Massachusetts, New York, and Ohio. In the latter state it was
common in the northern part, flying in June, the latest date,

being July 1. Van Duzee found the imago in a swamp. Kelli-

cott found it along wave-washed shores, the males flying out

over the water, the females, when not ovipositing, remaining
on herbage or trees on shore.

The nymph, not quite fully grown, measures 28 mm.; ab-

domen, 18 mm.; hind femur, 5 mm.; width of abdomen, 6 mm.>
of head 5.5 mm.

Body flat, hairy on legs and lateral margins. Color yellow
ish, eyes black; fuscous mottlings at base of wing-cases and on
sides of abdominal segments between the yellow mid-dorsal

line and the scars, sometimes forming oblique streaks.

Lalnum moderate; front border of median lobe nearly

straight; lateral lobes short arcuate, with a sharply incurved

end hook, on the inner margin 6 to 9 short rectangular back-

wardly directed teeth.

Wing-cases reach the 5th segment.

Abdomen nearly 3 times as long as wide, lateral spines on
segments 7 to 9 increasing in size posteriorly, those of 9 half

as long as segment 10. No dorsal hooks; traces of mid-dorsal

smooth line apparent on middle segments. Lateral margins of

8 and 9 subentire ; segment 10 half as long as 9. Appendages
longer than segment 10, the laterals shorter than the others.

11. Gonipliiis mistrails Needh.

This is a Florida species, described as new by Mr. Needham
('97, p. 184). Nymphs supposed by him to belong to (nisfralis

were collected near Gotha, Fla., in December, 1896, and in

January, 1897, by Mr. Adolph Hempel, who took a single male

imago at the same place a little later. Morphological charac-

ters entirely justify their reference to this species.



78

The nymph (supposition) measures 30 mm.; abdomen, 20

mm.; hind femur, 5 mm.; width of head 5 mm., of abdomen
5.5 mm. Olivaceous, yellowish below and on sutures.

Body slender, pointed posteriorly, moderately hairy on
margins and on all appendages.

Head compact, hind angles prominent, rounded, a pair of

small black tubercles on hind margin between these angles and
the median line. Antennae long, considerably exceeding the

labrum. Eyes large, black; ocelli yellow.

Labium moderate; mentum narrowed for its posterior two
fifths; median lobe very slightly curved, fringe of flat scales

short, in the middle of it a single simple or bifid tooth or a

double one, which is shorter than the scales; lateral lobes stout,

suddenly narrowed and sharply incurved to form the pi'ominent

end hook, which bears usually a minute denticle on its inner

side. Inside the end hook there are about eight obliquely

truncate, short, quadrangular, posteriorly directed teeth.

Hook long, strong, arcuate.

Legs short ; tibial hooks of fore and middle legs well de-

veloped.

Abdomen somewhat depressed, attenuate posteriorly and

upcurved, widest on segment 4 ; segments 3 to 7 and 10 all sub-

equal in length, 8 a little longer, 9 one half longer than the

others, margins of 9 yellowish. No dorsal hooks, or on 9 the

merest rudiment ; lateral spines on 7-9 short, appressed, minute

on 7 and 8, on 9 only about one eighth the length of 10. In-

ferior appendages about as long as segment 10 ; superior a very

little shorter, its basal half thickened; laterals a little shorter

than superior and much more sharply pointed

A young nymph one fourth grown differs from the mature

ones only in size and in having but 6 or 7 teeth on the lateral

lobe of the labium.

12. Gomphiis villosipes Selys.

This elusive gomphid was bred by Mr. Needham in large

numbers at Ithaca, N. Y., in May, 1897. The imago has never

been taken in Illinois, but there is a well-marked nymph of



79

this species in the State Laboratory collection from Swan
Pond, near the Waliash River in southeastern Illinois. The
nymphs were exceedingly abundant at Ithaca. Two distinct

.sizes were observed, a fact which, taken in connection with

the extremely short period of emergence, strongly indicates

that the species takes two years to complete its life cycle,

the nymph hatching in one season not transforming until

the second season thereafter. Two males were all the images
seen at large, yet exuvite m countless numbers lined the banks
of all streams. The perching habits of these two males were
about as described further on for G. pal/iihis. The imago has

been taken as far west as Michigan and Ohio. It flies in early

summer, the recorded period being May 30 to June 26.

Kellicott says of villusijjcs, cxilis, and fuirifer that they

"frequent quiet waters of smaller ponds, or even ditches where
there are floating algae or lily-pads on which to rest. The
males explore the borders and watch from the muddy shore or

floating plants; the females at intervals drop down from their

coverts to oviposit among the plants, dipping into the water in

a manner similar to that of Li/wllu/a."

The nymph measures 87 mm.; abdomen, 24 mm.; hind

femur, 6 mm.; width of head, 5.5 mm., of abdomen 8 mm.
Form identical with that of G. pallidHS. Lateral margins

of apical segments of abdomen marked rather showily with
yellowish, especially beneath. Labium as in G. pallidus, but

with the fringe of hairs bordering the median lobe shorter, the

lateral lolie arcuately incurved at the apex, its extreme point

shorter than the first of the six stout posteriorly-directed teeth

upon the inner margin. Lateral spines well developed only on
the 9th segment. Dorsal hooks almost entirely wanting, even
on the 9th segment. Appendages shorter than the 10th seg-

ment, the laterals a little shorter than the others.

13. Gomphiis pallid/is Ramb.

Mr. Needham studied this species at Purington Lake, near
Galesburg, where it was found in company with G. grasUnellus.



80

In his notes he says it corresponds with the latter species in

habitat, season of flight, habits of oviposition, and nymphal
habits, but is, he observes, a rather more vigorous and active

species. Male imagos have a habit of resting flat upon the

bare earth of a path or sloping bank at the vpater's edge and
making short sallies forth across the v^^ater, vphether for prey

or for sport or for outlook he w^as unable to determine. The
nymph may be knov^rn by its peculiar form and the obtuse

ridge along the middle of the abdomen above. Its preference

for still water is plainly indicated in the Biological Station

field, as it was found only in bare mud and sand on the bottom
of Clear Lake and Matanzas Lake. In the latter it was quite

common in the level lake-bottom, under about six feet of

water. In the Laboratory collections are also specimens from
Mississippi River lakes and sloughs near Quincy, and some
from the Saline River near Shawneetown. It did not occur in

our collections from the Sangamon River. Hagen reports

nymphs from Michigan, Massachusetts, South Carolina, and

Texas.

Dr. Forbes took an imago, just emerged, at Peoria in June,

1878. At Matanzas Lake, near the Biological Station, fresh ex-

uviae were noted July 6 on a log near the water, and imagos

were taken near by. A single example was taken May 27 at

Urbana. The season of flight at Purington Lake was consid-

erably later than that of graslinellus, according to Mr. Need-

ham. The first adult was seen May 28 ; the maximum abun-

dance was attained July 4 and early in August the last ones

disappeared—a period of about two months, which is about

twice as long as that of the commoner species graslinellus.

Although the latter imago is fairly common, that of palliclMS is

singularly rare in collections—almost unknown, in fact. It is

recorded only from Indiana, G-eorgia, and Louisiana. The

Odonafa offer several remarkable instances of this kind, which

should put us on our guard against wrong judgments of dis-

tribution and relative abundance based on the finding of ima-

gos alone.



81

In view of existing; descriptions and tignres, (Cabot, 72, p. 2,

No. 2, PI. I., Fig. 3 ; Hagen, '85, p. 266 1 it will be necessary here

to give only those nymphal characters which are distinctive.

The nymph measures in length 38 mm.; alKlomen, 2f5 mm.;
hind femur, 7 mm.; width of abdouien 8 mm., of head 6 mm.

Body flat, elongate, gracefully tapering posteriorly; surface

but little hairy.

Ijabiuni with median lobe of mentum rather prominent, a

median apical tooth on its front margin, in the middle of the

usual comb of flat spinous scales; lateral lol)e with a blunt-

pointed and incurved apical hook, and on the inner margin

seven or eight long irregular teeth, widening apically, and
obliquely truncate so as to appear to be directed backward.

Abdomen lanceolate, attenuate to apex, slowly tapering for

half its length, to the cylindrical lOth segment; a dorsal hook
on segment 9, represented on other segments by obtuse ridge-

like elevations ; lateral spines on segments 7 to i), on 7 minute,

on 8 short, on 9 long, closely appressed, nearly equaling the

length of segment 10 ; superior and inferior appendages sub-

equal, shorter than segment 10, laterals a little shorter than

the others.

Half-grown nymphs show clearly all these characters.

14. (toiiipints c.fi/is Selys.

In the Atlantic coast states from Maine to Maryland this

species is often abundant. In Illinois the imago has been taken

at Crystal Lake, about 40 miles northwest of Chicago, l)y Mr.

Longley, and at Kensington by Mr. McDade. Williamson

records a few examples from Indiana, and according to Kelli-

cott it is common throughout Ohio, especially about canals and
ponds. Hagen has described the nymph ('85, p. 263), and
doubtfully referred to the same species some examples sent him
from Kentucky. Dr. Calvert records it as the first gomphid to

appear in spring (May 24), its season of flight lasting until the

early part of July. The extreme recorded dates are May 9

(Kellicott) and July 8 (Williamson).



82

The following brief statement of the characteristic features

of the nymph is selected from Dr. Hagen's description.

Length, 19 to 24 mm. Body rather slender; median labial

lobe very faintly rounded, lateral lobes ending in a strongly

bent and pointed hook, with four strong teeth inside; abdomen
tapering slowly after segment 6, segments as far as 8 of equal

length, 9 a little longer than 8, 10 half as long as 9, cylindrical;

lateral spines on 6 to 9, the last one third as long as 10; dorsal

hooks very small, pointed on 6 to 9; appendages as long as last

segment; third joint of hind tarsi scarcely longer than second.

15. Gomphus cornutus Tough.

G. corniilus Tough, ('oo).

This recently published species, readily recognized by the

unusual form of the male abdominal appendages, was described

from two males from DuPage county, taken May 30 and June 14.

The Bolter Collection contains additional male and female ex-

amples, but without data.

16. Gomphus s]). {b).

Gomphus sp., Hagen, '85, p. 262, No. 14 (nymph).

This interesting species, described by Hagen from three

nymphs collected by Prof. Forbes at Cairo, 111., July 26, is quite

distinct from any other known to us. The following are the

principal distinctions:

Length, 18 to 23 mm. Body flat; lateral labial lobes with

only two incisions near the strong and sharp apical hook,

median lobe slightly rounded; side margins of abdomen very

hairy, segment 9 as long as 8, 10 very short; lateral spines on 6

to 9, the last less sharp, as long as 10; no trace of dorsal hooks;

appendages as long as 9, broad; anterior tibial hooks short.

17. Gomphus STp. {c).

Gomphus sp., Hagen, '85, p. 263, No. 15 (nymph).

A single nymph received from Newport, Ky., by Dr. Hagen
is similar to the preceding, but was believed by him to be a

distinct species, perhaps G. mimdtis.



83

Length 23 mm. Body flat : lateral labial lobes with a sharp

bent apical hook, and but three strong teeth near tip ; abdo-

men tapering slowlj- from 7th segment. 9 one thii-d longer than

8, 10 very short, one sixth of 9 : lateral spines on 6 to 9, sharp,

the last pair twice the length of 10: no dorsal hooks, except

that 9 bears a liroad. triangular. Hat. median apical projection

as long as 10 : appendages twice as long as 10,

is. Goiiqjinis miniico/a V^ alsh.

? G. o/i7ia<:eus (suppos.), Hagen. '85, p. 271, No. 27 (nymph).

This and the remaining species here treated belong to the

subgenus Sfi/Jiirifs. Williamson has made a thorough .study of

this troublesome group of species. A copy of his revision of

the subgenus (Williamson '(illn has been received from him

iu.st as these pages are being set up l)y the printer, enalding us

to incorporate his data concerning the geographical and sea-

sonal distribution of the species, derived from a large number of

specimens examined by him,

A single nymph in the State Laboratorj" collection, taken

at Colona, Henry Co., 111., Augu.st 20, is alnio.-<t identical with

that described bj' Hagen from Itah and connected bj' supposi-

tion with G. ()Jir(nciif<. It stands in about the same relation to

Gonijjhus )i()t(itiis that the imago (iniiiln>l<i does, and may be

assigned to this species.

Mr, Needham is of the opinion that the affinities of both

the present species and G. snuhh-ri are with the ili/dfufiis. rather

than the 2il(((/i((tiis [Sfi//iinif< Xeedh.j, group of species, in which
case the nymph here assigned to aninicohi may be wronglj'

placed. It is earnestly to be hoped that the nymphs of these

species may soon be discovered, as this would doubtless throw
much needed light on their specific relatiou-ships.

The imago of this species has been taken in Illinois only

along the ]\lississippi in the nortliwe.><tern part of the state, and
by Mr. Longley at Riverside, near Chicago. It was found at

Rock Island and described as new by Walsh. Our specimens

were captured at Savanna, August 1, by Mr. F. M. McElfresh.



84

Single examples are also listed from Galena, 111., and from Iowa
(July), Kansas, and New York. It was taken by Mr. Longley
in June.

The principal characteristics of the nymph are as follows

:

Nymph (supposition) very similar to G. «ofo^»s ; median
labial lobe rounded but short, lateral lobes similar to those of

notatus ; abdomen broader, more lance-shaped ; 9 a little longer

than 8, with a very short apical spine above ; lateral spines on
6 to 9, the last one blunt, half as long as 10, which is short, cy-

lindrical
; appendages and legs similar to to those of notatus.

19. GomphuH plagiatus Selys.

G.fluvialis, Walsh (female).

G. notatus, Hagen, '85, p. 270, No. 25 (nymph).

G. plagiatus, Hagen, '85, p. 269, No. 24 (nymph).

This is probably the most abundant gomphid nymph in the

larger rivers of Illinois. It may be distinguished from any
others herein described by the penultimate (ninth) abdominal
segment, with its lateral and basal margins about equal and a

rudimentary posterior tooth on the median dorsal line, the

other segments being without trace of dorsal hooks or eleva-

tions. As already stated under externus, it is found about

Havana in muddy bottoms where there is current, but not so

exclusively in these situations as externus. Mr. Garman re-

ported it as the commonest species in waters associated with

the Mississippi at Quincy, and states that it does not breed ex-

clusively in running water, presumably basing this statement

on the fact that he found it in slough-lakes. These lakes form
channels of flow during higher stages of the river, but are more
or less completely isolated during low water. Unlike similar

waters about Havana, these lakes are but scantily supplied

with aquatic vegetation. In our general collection are nymphs
from Cairo, 111., where they are probably common, judging from
the number taken ; from the Rock River at Colona and Milan ;

from the Ohio at Golconda ; and from the Illinois at Pekin.

The species did not appear in extensive collections from the

Sangamon in Champaign county. The nymphs were not more



85

common than spiniceps or exfernus at the mouth of Quiver

Creek, but in the main channel of the Illinois below Spoon
River they were the most abundant form. Nymphs have also

been taken in Quiver and Matanzas lakes, in the Illinois River

at Meredosia, and in Spoon River near Bernadotte, 111. Ex-

amples of this species and I'.rffnius taken in June and placed in

breeding-cages immersed in the water along the Quiver Lake
shore, remained alive without transformation until the season's

work closed at the end of September. This species is very close

to G. )i(>t(iti(x, which is comparatively rai"e in Illinois, and the

nymph of which has not been satisfactorily distinguished from
that of plagiatHs.

The first transforming imago was noted May 21. On July

2, 1894, large quantities of nymphs transformed in the early

morning upon the piers of the river bridges. Upon arriving in

Havana in the latter part of June, 1897, Mr. Hart found the

imagos emerging numerously up to about the end of the

month. In 1898 the largest number emerged June 21, and
thereafter occasionally up to June 30. Mr. Needham's notes,

made while at Havana in July, 1896, are as follows:
" Transformation takes place mostly at night, but not un-

commonly late in the afternoon. The nymph crawls a little

way (8 to 20 inches) out of the water upon any flat surface. I

have found exuvias sticking to bridge piers and to the sides of

barges and fishing boats, to willow stumps, and to bai'e mud
banks. I obtained hundreds of nymphs and exuviee, and yet

during a month spent upon the river collecting 1 did not see a

single imago of this species on the wing. I captured one newly
emerged, at dusk, resting in the grass at the bank ; no others

were obtained except by rearing them. Consequently, no op-

portunity was found for studying breeding habits and ovipos-

ition."

Kellicott also notes the emergence on piling and walls in

deep water, and found the imago " resting on coarse grasses

during July. None were seen ovipositing, or flying except to

escape from danger."



86

The images are reported from Illinois, Texas, Michigan,
Ohio, Tennessee, and the Atlantic coast states from New Jersey

to Florida, the dates of collection ranging from May 25 to

September 30. The period of flight usually ends about August
1, but a single example was taken as late as August 25 at

Havana. In Illinois adults have also been listed from Bloom-
ington, Moline, and from Winchester, in Scott county.

The nymph measures in length 35-36 mm.; abdomen, 25

mm.; hind femur, 4.5 mm.; width of abdomen 6 7 mm., of

head 5 mm.
Body narrow and elongate, a little hairy on lateral margins

and appendages. Head cordate, arcuately notched behind, with
broadly rounded but prominent hind angles. External tibial

hooks minute.

Labium elongate, mentum one half longer than broad,

widened apically; median lolje slightly rounded in front, its

comb of scales short and sparse; lateral lobes arcuate and ter-

minating in a sharply incurved hook. Teeth on inner margin
varying from none at all to four or five, increasing in size and
distinctness proximally. Movable hook long, strong, and regu-

larly incurved.

Abdomen widest on 6, median dorsal smooth line on seg-

ments 2 to 7 or 2 to 8; a flattened rudiment of a dorsal hook
on 9; segments 3 to 8 of about equal length, 9 one third longer,

10 very short; lateral spines on 6 to 9 appressed, those on 9 half

as long as the inferior appendages;- appendages about equal

and exceeding the length of 10.

20. Gomphus notatus Selys.

G.Jluvialis Walsh (in part).

The true notatus is represented in our Illinois collections by
only a single female, taken on the bank of the Mississippi near

Savanna August 1. In its wing venation and external sexual

structures it seems quite distinct from ijUnjiatus. The nymphs
described by Hagen as notatus and plarjiatus are probably all

pla^iatus. His description of notatus agrees exactly with un-



87

questionable nyinphs of j)la(/i(ifiis. Walsh apparentl.y included

lioth species in his Jiiiria/is, described from specimens collected

about Rock Island, but his statements of the habits ot ffiirialis,

which are quite unlike those we have recorded for plagiatus,

very likely apply to iiofa/i/s. and may account for its apparent

rarity. He says that while frafcrnns and rasfi(s haunt the land

and are easy to capture, /?»r/ff//.s "makes long excursions to and
fro on the surface of the river, scarcely ever approaching the

land except for a second."

Williamson ('Olb) lists the species from Illinois, Michigan,

Ohio, Tennessee, and Quebec, Canada ; June 1 and 21 in Ohio

and Michigan, and September 30 in Tennessee.

21. (T<iiiiji/nis spiii/i-cji.s Walsh.

Macrogotiiphu!:? spinkeps Walsh, "62, p. 389 (female).

Goinphus spiniccps, Kellicott, '95, p. 209 (male).

Goinphus segregans, Needham, "97, p. 185 (male).

This seems to jirefer somewhat rapid currents. Excepting

a single example from the Illinois Kiver, the nymphs appear in

the Lalioratory and Biological Station collections only from

Quiver Creek and Quiver Lake near the mouth of the creek.*

This creek is a peculiar and lieautiful stream, shallow and
sandy, fed liy springs of soft water flowing out of the sand

beds, and hence quite uniform and constant in its flow. At
the mouth of this stream spinirq)i<: is associated with externum

and ji/(((//(ifiis. The nymph agrees with that of pJafjintus in

having no dorsal elevation or hooks except a single rudimentary
posterior tooth on the penultimate segment, but differs from it

conspicuously in the form of that segment, which is long and

narrow, the lateral margin fully twice the basal width. In this

character it resembles Goiiipliii^pal/idus, but has not the elevated

dorsal ridge nor the long tenth segment of that species. As
in the case oi plagiatus and erfeniiift, nymphs placed in breed-

*Numerous exuviee and emerging individuals have since been taken on the

rocky walls of the shallow gorge of the Illinois River at Ottawa, July 21 and

August 19, the larger number of images being secured on the latter date.



88

ing-cages in June failed to transform, one being still alive on
September 28. In 1897 two full-grown nymphs were taken

June 22 along shore in the quiet weedy waters of Quiver Lake,

not far from the mouth of Quiver Creek. The only imago seen,

emerged June 23 from a nymph which we found that day
crawling up some blades of bur-reed {Sparganiam eurycarjjum)

beside the stream. A number of exuvise have been observed at

various times during June and July on Sagiitaria stems, brush-

wood, and the bases of tree trunks along the margin of the

creek, from near its mouth to McHarry's Mill, a few miles above.

Cabot (72, p. 5, PI. II., Fig. 1) described the nymph from
examples taken in Massachusetts July 4, one of them partly

transformed, and Hagen ('85, No. 26, p. 270) redescribed it.

Kellicott ( '95, p. 209) observed the imagos in Ohio Sep-

tember 4 flying about and ovipositing "in a manner similar to

that of the LibelJulas" in a small brook that was rippling over
pebbles, and he later ('96, p. Ill) gives the time of flight as

August and September. Williamson found the adults trans-

forming abundantly in western Pennsylvania July 4, and also

records the species from Tennessee (August) and New York
(September). The imago was originally described from Rock
Island. Other Illinois localities for the imago are Bloomington
and the Mackinaw River at the " Dells."

The nymph measures in length 40 mm.; abdomen, 29 mm.;
hind femur, 5 mm.; width of abdomen 7 mm., of head 5 mm.

Body very elongate and posteriorly attenuate, sparsely

hairy on lateral margins and on appendages. Tibial hooks
minute.

Labium elongate ; mentum widened toward the apex

;

median lobe with a very straight front border; lateral lobes

arcuate, with a long-pointed sharply incurved end hook, and on
the inner side one to three teeth, though sometimes only the

middle one is distinct. These teeth are widened distally, and
obliquely truncate so that they appear directed backward.
Movable hook long, strong, and more strongly curved toward
the tip.



89

Abdomen veiy elongate, acuminate toward the end,

scarcely depressed ; segments slightly increasing in length

posteriorly to the yth ; 9 about twice as long as S, its sides

nearly parallel on apical half ; 10 very short; smooth mid-dorsal

line on 2 to 7, a flattened rudiment of a dorsal hook on apex of

9 ; short lateral spines on 6 to 9. those of 9 about half as long

as the superior appendage ; appendages subequal.

Young nymphs in the station collection, dredged from the

bed of the creek, some not more than one third grown, show
these same characters very clearly marked.



LITERATURE.

Banks, Nathan.
'92. A Synopsis, Catalogue, and Bibliography of the Neuropter-

oid Insects of Temperate North America. Trans. Am. Ent.

Soc, Vol. XIX., pp. 327 374.

'96. A New Species of Gomphus. Journ. N. Y. Ent. Soc, Vol.

IV., pp. 193-195.

Beutenmiiller, Wm.
'90. Mode of Oviposition of certain Species of Odonata. Ent.

Amer., Vol. VI., pp. 165-166.

'90a. Essay on the Destruction of the Mosquito and House Fly.

Drag-on Flies vs. Mosquitoes (Lamborn, '90), pp. 97-127, PI.

VIII., IX.

'90b. Records of Migration of Dragon Flies, especially Libellula

qiiadrimaailata, from 1868 to 1888. Dragon Flies vs. Mos-

quitoes (Lamborn, '90), pp. 161-162.

'90c. Preliminary Catalogue of the Described Transformations

of the Odonata of the World.* Dragon Flies vs. Mosquitoes

(Lamborn, '90), pp. 165-179.

Cabot, Louis.

'72. The Immature State of the Odonata. Part I. Subfamily

Gomphiua. lUustr. Cat. Mus. Comp. Zool., No. V. 17 pp., 3 pi.

'81. The Immature State of the Odonata. Part II. Subfamily

^schnina. Mem. Mus. Comp. Zool., Vol. VIII., No. 1. 40

pp., 5 pi.

'90. The Immature State of the Odonata. Part III. Subfamily

Cordulina. Mem. Mus. Comp. Zool., Vol. XVII., No. 1. 52

pp., 6 pi.

Calvert, Philip P.

'93. 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, Vol. XX., pp. 152a-272,

PL II., III.

'95. The Odonata of Baja California, Mexico. Proc. Cal. Acad.

Sci., Second Ser., Vol. IV., pp. 463-558, PL XV.-XVII.
'01. On Gomphus fraternus, externus and crassus (Order Odo-

nata). Ent. News, Vol. XII., pp. 65-73, PL III.

*The descriptions in Cabot, 'go, are not listed in this worl<.

00



91

Comstock, J. H.
'88. An Introduction to Entomology. iv-r234 pp., fig-ures. Ithaca,

N. Y. (Odonata, pp. 70-75.)

Elrod, Morton J.

'98. lowan Odonata. Ent. News, Vol. IX., pp. 7-10.

Forbes, S. A.
'88. Studies of the Food of Fresh-Water Fishes. Bull. 111. State

Lab. Nat. Hist., Vol. II., Art. VII., pp. 433-473.

'88a. On the Food Relations of Fresh-Water Fishes: a Summary
and Discussion. Bull. 111. State Lab. Nat. Hist., Vol. II.,

Art. VIII., pp. 475-538.

'93. A Preliminary Report on the Aquatic Invertebrate Fauna
of the Yellowstone National Park, Wyoming-, and of the

Flathead Region of Montana. Bull. U. S. Fish Comm., 1891,

Art. 6, pp. 207-258, PI. XXXVII.-XLII.
Garman, H.

'90. A Preliminary Report on the Animals of the Mississippi

Bottoms near Ouincy, Illinois, in August, 1SS8. Part I. Bull.

111. State Lab. Nat. Hist., Vol. III., Art. IX,, pp. 123-184.

(Odonata, pp. 175-179.)

Hagen, H. A.

'61. Synopsis of the Neuroptera of North America, with a List

of the South American Species. Smithsonian Miscel. Collec-

tions. xx-i-347 pp. (Odonata, pp. 55-187.)

'77. Synopsis of the Odonata of America. Proc. Best. Soc. Nat.

Hist., Vol. XVIII., pp. 20-96.

'85. Monograph of the Earlier Stages of the Odonata. Sub-
families Gomphiua and Cordulegastrina. Trans. Amer. Ent.

Soc, Vol. XII., pp. 249-291.

Harvey, F. L.

'91. A Contribution to the Odonata of Maine. [I.] Ent. News,
Vol. II., pp. 50-51, 73-75.

'92. A Contribution to the Odonata of Maine. II. Ent. News,
Vol. III., pp. 91-93, 116-117.

'98. Contributions to the Odonata of Maine. III. Ent. News,
Vol. IX., pp. 59-64, S5-SS.

'01. Contributions to the Odonata of Maine. IV. Ent. News,
Vol. XII., pp. 178-179, 196-198.

Karsch, Ferdinand.

'93. Die Insecten der Berglandschaft Adeli in Hinterland von



92

Togo (Westafrika) nach dem von den Herren Hauptmann

Eugen Kling (1888 und 1889) und Dr. Richard Buttaer (l&9i»

und 1891) g-esammelten Materiale bearbeitet, mit einem Vor-

worte von Dr. Richard Biittner und einer Uebersichtskarte

des Togogebietes. I. Abtheilung: Apterygota, Odonata,

Orthoptera Saltatoria, Lepidoptera Rhopalocera. Berl. Eat.

Zeitschr., Vol. XXXVIII., pp. 1-266. 6 pi. and fig. in text.

(Odonata, pp. 41-48, PI. I.-IV.)

Kellicott, D. S.

'95. Catalogue of the Odonata of Ohio. Part I. Journ. Cine.

Soc. Nat. Hist., Vol. XVII., pp. 195-216.

'96. Catalogue of the Odonata of Ohio. Part II. Journ. Cine.

Soc. Nat. Hist., Vol. XVIII., pp. 105-114.

'97. Catalogue of the Odonata of Ohio. Part III. Journ. Cine.

Soc. Nat. Hist., Vol. XIX., pp. 67-71.

'97a. Notes on Dragonflies. Amer. Nat.,Vol. XXXI., pp. 351-354.

'97b. An Odonate Nymph from a Thermal Spring. Journ.

Cine. Soc. Nat. Hist., Vol. XIX., pp. 63-65. Figures.

'99. The Odonata of Ohio. Special Papers of Ohio Academy
of Sciences, No. 2. viii+ 116 pp., 4 pi.

Kirby, W. F.

'90. A Synonymic Catalogue of Neuroptera Odonata, or Dragon-

flies. With an Appendix of Fossil Species. x+202 pp.

London.

Lamborn, Robt. H.
'90. Dragon Flies vs. Mosquitoes. 202 pp., 9 pi. New York.

Lampert, Kurt.

'99. Das Leben der Binnengewasser. xvi+S91 pp., 12 pi., 223

fig. in text. Leipzig. (Odonata, pp. 160-167.)

Lucas, W. J.

'97. Dragonflies in 1896. Ent., Vol. XXX., pp. 29-36. 2 fig.

'97a. Dragonflies in 1897. Ent., Vol. XXX., pp. 277-283, PI. L
'00. British Dragonflies (Odonata). xiv+356 pp.. 27 pl., 57

fig. in text. London.

McDade, Jas. E.

'92. Illinois Neuroptera. Popular Science News, Vol. XXVI.,

p. 150.

Mundt, A. H.
'82. Migration of Dragon-flies— iGschna herosFabr. Can. Ent.,

Vol. XIV., pp. 56-57.



1)3

Needham, J. G.

'97. Preliminary Studies of N. American GomphinEe. Can. Eat.

,

Vol. XXIX., pp. 164 168, 181-186, PI. VII.

'97a. On Rearing Drag-onflies. Can. Eat., Vol. XXIX., p. 94.

Figures.

'97b. Libellula deplaaata of Rambur. Can. Eat., Vol. XXIX.,
pp. 144-146.

'98. Birds vs. Drag-onflies. Osprey, Vol. II., pp. 85 86. Fig-ures.

'99. Directions for coUectiaa: and reariag Drag-on Flies, Stone
Flies, and May Flies. Bull. U. S. Nat. Mus., No. 39, Pt. O.

9 pp., 4 fig-.

'00. Insect Drift on the Shore of Lake Michig-an. Occasional

Mem. Chicago Ent. Soc, Vol. I., pp. 19-26. Fig-ure.

Needham, J. G. and Weith, R.

'01. The Life-history of Nanaothemis bella Uhler. Can. Ent.,

Vol. XXXIII., pp., 252-255. Fig-ure.

Packard, A. S.

'80. Guide to the Study of lasects. i.x+715 pp , fig-ures. New
York. (Odonata, pp. 597-605.)

'84. Pseudo-neuroptera. Stand. Nat. Hist., Vol. II., pp. 140-

154. Fig-ures. (Odoaata, pp. 147 -151.

)

Ris, F.

'97. Untersuchung iiber die Gestalt des Kaumagens bei dea
Libellea uad ihrea Larvea. Zool. Jahrb., Abth. fiir Syst.,

Bd. IX., pp. 596-624. 14 fig.

Roster, Daate Alessandro.

'85. Contributo all' anatomia ed alia biolog-ia degli Odonati.

Bull. Ent. Soc. Ital., Vol. XVII., pp. 256-268, PL III., IV.
'86. Cenno Monographico deg-li Odonati del Gruppo Ischaura.

Bull. Ent. Soc. Ital., Vol. XVIII., pp. 239 258, PI. II.-VI.

'88. Coatributo alio Studio delle forme larvali degli Odonati.

Cenno iconog-rafico delle ninfe dei Caudobranchiati. Bull.

Soc. Ent. Ital., Vol. XX., pp. 159 170, PI. I. -IV.
Sharp, David.

'95. Insecta. Cambr. Nat Hist., Vol. V., pp. 81 565. (Odonata,

pp. 409-428.)

Todd, J. E.
'85. Notes on the Breeding Habits of the Libellulida;. Amer.

Nat., Vol. XIX., pp. 306-307.



94

Toug-h, James.

'00. A New Species of Gomphus. Occasional Mem. Chicag-o

Ent. Soc, Vol. I., pp. 17-18. 1 fig-.

Uhler, P. R.

79. List of Animals observed at Fort Wool, Va. Ches. Zool.

Lab., Scientific Results of the Session of 1878, pp. 17-34.

(Odonata, pp. 32-34.)

Van Duzee, E. P.

'97. List of Drag-onflies taken near Buffalo, N. Y. Journ. N.

Y. Ent. Soc, Vol. V., pp. 87-91.

"Wadsworth, Mattie.

'90. List of the Dragonflies (Odonata) taken at Manchester,

Kennebec Co., Me., in 1888 and 1889. Ent. News, Vol. L, pp.

36-37, 55-57. Additions and corrections, Ent. News, Vol. IL,

pp. 11-12; Vol. Ill
, pp. 8-9; Vol. V., p. 132; Vol. IX., p. 111.

Walsh, B. D.

'62. List of the Pseudoneuroptera of Illinois contained in the

Cabinet of the writer, with Descriptions of over forty new
Species and Notes on their Structural Affinities. Proc. Acad.

Nat. Sci. Phila., 1862, pp. 361-402.

'63. Observations- on certain N. A. Neuroptera, by H. Hagen,
M. D., of Kiinig-sberg, Prussia; translated from the original

French MS., and published by Permission of the Author, with
Notes and Descriptions of about twenty new N. A. species of

Pseudoneuroptera. Proc. Ent. Soc. Phila., Vol. II., pp. 167-272.

Williamson, E. B.

'00. The Dragonflies of Indiana. Twenty-fourth Ann. Rep.

(1899) Dept. Geol. and Nat. Resources Ind., pp. 229-233.

'00a. On the Habits of Tachopteryx Thoreyi. (Order Odonata.)
Ent. News, Vol. XL, pp. 398-399.

'01. On the Manner of Oviposition and on the Nymph of Tachop-
teryx Thoreyi. Ent. News, Vol. XII., pp. 1-3, PI. I.

'Ola. Preservation of Colors in Dragon Flies (Odonata). Ent.
News, Vol. XII., pp. 131-133.

'01b. The Subgenus Stylurus Needham: Selys' Groups VI. and
VII. of the Genus Gomphus (Odonata), and on the Postanal
Cells in the latter. Trans. Am. Ent. Soc , Vol. XXVII., pp.
205-217, PI. VIII., IX.

-
/



Plate I.

I. yasiirsr/iiui priitiuaiillia.

4. lUiislaliuiiiiia

caroliis.

sM'^*^-^

2. Boyeria viiwsa.

3. /'roi;oi/i/i/i!is

obscin-iis.

5. Aiin.v Junius.

Nymphs.