SHORT NATURE-STUDIES

-D
cO
JD
r*

°

O

m

D

THE NATURAL HISTORY

OF

AQUATIC INSECTS

•
-e-

MACMILLAN AND CO., LIMITED

LONDON • BOMBAY • CALCUTTA
MADRAS ' MELBOURNE

THE MACMILLAX COMPANY

NEW YORK • BOSTON • CHICAGO
DALLAS • SAN FRANCISCO

THE MACMILLAN CO. OF CANADA, LTD.

TORONTO

THE NATURAL HISTORY

OF

AQUATIC INSECTS

BY

PROFESSOR L. C. MI ALL, F.R.S.

WITH ILLUSTRATIONS BY A. R. HAMMOND. F.L.S.

MACMILLAN AND CO., LIMITE
ST. MARTIN'S STREET, LONDON

1922

COPYRIGHT
First Edition 1895; Reprinted 1903, 1912, 1522

PRINTED IN GREAT BRITAIN

PREFACE

I HAVE here attempted to help those naturalists,
especially those young naturalists, who take delight
in observing the structure and habits of living animals.
It has also been my hope that I might do something
to revive an interest in the writings of certain old
zoologists — Swammerdam, Reaumur, Lyonnet, and
De Geer — who are at present unjustly neglected; I
have tried to carry on as well as to popularise their
work. It would be much if I could persuade some
few working naturalists to lay aside their technical
lists and records of parish distribution, and study the
works of Nature with open eyes, seeking above all
things to know more of life in its infinitely varied
forms.

Some passages in this book, if taken alone and read
hastily, may appear to disparage systematic Zoology.
This is far from my intention. No one can study the
great naturalists of the seventeenth and eighteenth
centuries without feeling how seriously their work is
impaired by the defective systems of the time. It is

vi PREFACE

not systematic but aimless work that I deprecate
-work which springs from no real curiosity about
Nature, and attempts to answer no scientific questions.
I have great pleasure in acknowledging the advice
and contributions of several friends, and especially of
Messrs. W. F. Baker, T. H. Taylor, and J. J. Wilkinson.
The drawings made by Mr. Hammond, in most cases
direct from the Insects, are a valuable feature of the
book, and I could say more in their praise if they did
not speak for themselves.

L. C. M.
LEEDS, February 1895.

Advantage is taken of a reissue to correct the
errors which have come to the author's knowledge,
and to add a few Additional Notes (pp. vii-viii).

L. C. M,
December, .T9o2.

Lj LIBRARY

I •^^•^

ADDITIONAL NOTES (1903)

Page 40 (egg-laying of Dytiscus). Dr. Sharp remarks : — " Lyonnet
was probably correct as to the oviposition of Dytiscus, for though D.
marginalis will die rather than oviposit except in vegetable tissue,
D. punctulatus will occasionally scatter its eggs at random."

Page 97. Since this book was written and published Major Ronald
Ross has investigated with remarkable success the relation between
Gnats and malarial fevers. An extensive literature has already grown
up, dealing with the various questions involved. Dr. L. O. Howard's
Mosquitoes, New York, 1901, may be recommended to the reader as a
source of information on this subject.

Page 122. A full account of the structure and life history of
Chironomus has been published in The Plarlequin-Fly, by L. C. Miall
and A. Hammond, Oxford, 1900.

Page 149, line 7. Some Polychcet worms furnish additional examples
of gelatinous egg-masses.

Page 182. The trachea! system of Simulium in all its stages has
lately been fully described by Mr. T. H. Taylor. Trans. Entomological
Soc., 1902.

Page 206, line 5. Mr. Wilkinson has given a fuller account in his
Pharynx of the Eristalis larva. Privately printed. 1901. Two plates
and figs, in text.

Page 2 1 8. The remarkable larva and pupa of Phalacrocera have
been described by L. C. Miall and R. Shelford in Trans. Entomological
Soc., 1897.

Page 219. Among aquatic Orthoptera may also be mentioned a
new Fijian Cricket (Hydropedeticus) described by L. C. Miall and
G. Gilson in Trans. Entomological Soc., 1902.

viii ADDITIONAL NOTES (1903)

Page 235. Dr. Sharp in the Cambridge Natural History, Insects,
Vol. II., p. 425, quotes the following descriptions of the larva of
Acentropus : — " Disque, Ent. Zeit. Stettin, li. 1900, page 59. Cf. also
Rebel, Zool. Jahrb. Syst. xii. 1898, p. 3." I am indebted to Dr.
Sharp's work also for pointing out an error in the figure of the spiracles
of Dytiscus (fig. 12, supra), which has now been corrected. The work
in question is an invaluable repertory of information respecting insects
of all kinds, of which much greater use would have been made, had
it existed when this book was written.

Page 251, near bottom. Further information about the curved spine
is supplied by Prof. G. Gilson, On Segmentally disposed Thoracic
Glands in the Larvce of the Trichoptera. Linn. Journ., Zool., Vol.
xxv. page 407 (1896).

Page 271, footnote 2. This larva, which is that of Oithocladius, a
sub-genus of Chironomus, is described and figured by Mr. T. II.
Taylor in Miall and Hammond's Harlequin-fly, pp. 14-19.

I ought to have mentioned in the original preface that Mr. Hammond
contributed several original drawings from his own preparations, and
that the accounts of Chironomus, Ptychoptera, Simulium and
Stratiomys in particular were much improved thereby.

TABLE OF CONTENTS

INTRODUCTION

Water as a Sphere of Life, p. I. The Invasion of the Waters by Insects,
3. The Dominance of Insects, 8. Degrees of Adaptation to Aquatic
Conditions, II. The Surface-film of Water, 12. Equilibrium of
Aquatic Insects, 15. The Wintering of Aquatic Insects, 18. Some
Cases of Abridged Life-histories, 20. Live Natural History, 24.
The Groups to which Aquatic Insects belong, 28.

CHAPTER I

AQUATIC BEETLES

The Whirligig Beetle (Gyrinus), p. 30. Respiration by Tracheal Gills, 36.
Dytiscus, 39. The Great Water-beetle (Hydrophilus), 61. Life of
Lyonnet, 62. Hydrobius, 87. Donacia, 93.

CHAPTER II

FLIES WITH AQUATIC LARVAL

The Gnat (Culex), p. 97. Experiment on Surface-tension, 100. Corethra,
113. Collecting Tackle, 114. Mochlonyx, 121. Chironomus, 122.
Microscopic Study of Salivary Glands, 127. Haemoglobin in Blood,
130. Development of the Winged Insect within the Larva, 134.
Animals which coat their Eggs with Jelly, 148. Tanypus, 152.
Ceratopogon, 155. Dixa, 157. Dicranota, 164. Pupal Armature
for enabling the Insect to creep to the Surface, 169. Ptychoptera,
170. Simulium, 175. Stratiomys, 189. The Rat-tailed Larva (Eris-
talis), 198. J. J. Wilkinson on the Tail of Eristalis, 201 ; on the
Pharynx ot ditto, 206. Baron Qsten Sacken on the Oxen-born
of the Ancients, 215.

32604

x TABLE OF CONTENTS

CHAPTER III

AQUATIC HYMENOPTERA

Lubbock on Polynema, p. 219. Klapalek on Agriotypus, 223.

CHAPTER IV

. AQUATIC CATERPILLARS

Hydrocampa, p. 226. Paraponyx, 231.
CHAPTER V

CADDIS-WORMS (TRICHOPTERA)

Life of Reaumur, p. 236. T. H. Taylor on Plectrocnemia, 265. Case-
bearing Insects, 270.

CHAPTER VI

The Alder Fly (Sialis), p. 273.
CHAPTER VII

PERLID.^E

Stone Flies (Perla), p. 279.

CHAPTER VIII
MAY-FLIES (EPHEMERID^E)

Life of Swammerdam, p. 285. Swammerdam on Palingenia, 287.
Reaumur on Polymitarcys, 304. De Geer on Ephemera, 318.
Pictet on Larvre of Ephemeridce, 322.

CHAPTER IX

DRAGON-FLIES (ODOXATA), P. 328

TABLE OF CONTENTS xi

CHAPTER X

POND-SKATERS, WATER-SCORPIONS, AND WATER-BOATMEN

(RHYNCHOTA), r. 346

Hydrometra, 349. Gerris, 349. Velia, 351. Nepa and Ranatra, 353.
Notonecta and Corixa, 355.

CHAPTER XI

THE WATER SPRING-TAIL (PODURA)

Life of De Geer, p. 362. Podura, 364. Isotoma, 366. Leaping upon
Water, 367.

CHAPTER XII

INSECTS OF THE SEA-SHORE, P. 370

Diptera, 373. Beetles, 374. W. F. Baker on Tibial Comb of Beetles,
376. Caddis-worms, 379. Hymenoptera and Rhynchota, 380.

CHAPTER XIII

THE CONTRIVANCES OF AQUATIC INSECTS

Modes of Locomotion, p. 382. Modes of Feeding, 385. Modes of
Respiration, 385. Attack and Defence, 386. Egg-laying, 387.

(LIBRARY

THE NATURAL HISTORY

OF

AQUATIC INSECTS

INTRODUCTION

WATER AS A SPHERE OF LIFE

WATER offers many advantages to the simpler
animals and plants. The vast extent of the seas,
lakes, and rivers is in itself a sufficient motive for
their occupation by living things, considering the
crowded state of the land. But more special induce-
ments are not wanting. The density of water supports
the weight of the most fragile organisms, and enables
them to move from place to place by such easily
contrived means as cilia, which may be mere drawn-
out threads of protoplasm, or fins, which may be
shaped out of a fold of skin or a row of bristles.
Some of the very simplest aquatic animals, such as
Amoeba, move along in the water by an act which
can only be described as flowing, the stream of living
protoplasm gathering itself up continually in the,
£ pi

2 NATURAL HISTORY OF AQUATIC INSECTS

rear, and rolling forward to the front. All the very
simplest animals and plants are aquatic, and a fair
degree of complexity of structure is implied in the
mere fact of residence out of water.

Shallow waters are easily penetrated by the sun's
rays, and are therefore often occupied by green
plants, whose nutrition depends upon the decom-
position of carbonic acid by sunlight. Upon these
green plants plant-feeding animals establish them-
selves, and many predatory animals come in turn to
devour the vegetable-feeders. The shallow waters
are probably richer in living things of all kinds
than any other part of the earth's surface. But the
organic products of the plants and animals of the
shallow waters may be transported by moving water
to great distances. . The great depths of the ocean
and the bottom of our large lakes have been explored
by the dredge, and are found to support a popula-
tion of their own, which subsists upon a nutritive
sediment originally formed at or near the surface.

The conditions of life in fresh waters are materi-
ally different from those afforded by the sea. The
chemical difference between fresh and salt waters
is of small importance in itself — so small that
many animals have been able to adapt themselves
gradually to a change from one to the other. A
degree of saltness far greater than that of sea-water
is perfectly compatible with animal life, as is shown by
the fact that Crustacea and Insects are often found
quite at home in brine-pits and vats. There are even
a few animals, like the Salmon, which can pass sud-
denly from fresh water to salt, or from salt to fresh.

INTRODUCTION 3

More important from many points of view is the fact
that the seas are of vast size and, for the most part,
continuous ; while the fresh-water areas are relatively
inconsiderable in extent, and broken up into count-
less small areas. Every lake and river-system is an
isolated territory, and most of the plants or animals
found therein can only make their way into another
territory by help of a rare opportunity. This par-
ticular circumstance has governed the course of life
to a material degree, and has brought about some
of the most marked differences between the fluviatile
and the marine populations. It is also worth while to
notice that in the large fresh-water areas, which in
nearly every part of the world are the rivers, the
water is in motion, and flowing steadily towards the
sea. Immersion in the sea is quickly fatal to most
fresh-water animals and plants, so that here is a
special danger to which the smaller and more help-
less fluviatile animals are exposed. We have reason
to believe that this special danger has helped to bring
about one of the most noteworthy peculiarities in the
life-history of fluviatile animals, viz., the suppression
of free-swimming and ciliated larval stages.1

THE INVASION OF THE WATERS BY INSECTS.

The waters, both fresh and salt, have been success-
fully invaded by Insects, and this not in one or two
cases only, but many times over, and by the most
diverse kinds of Insects. The aquatic Insects do not

1 This has been pointed out by Professor Sollas (Proc. Roy.
Dublin Society, 1884).

B 2

4 NATURAL HISTORY OF AQUATIC INSECTS

form an order or a family, but comprise a miscel-
laneous selection of genera and species from many
families and several orders. Let us consider what may
be the significance of this remark. A family or an order
of Insects is a group which the naturalist supposes
to have descended from a common stock. An order
may include many families, and the common stock of
the order is therefore presumed to be much more
remote than that of any of its component families.
Now if all Insects were at one time terrestrial (and of
this we have no sort of doubt, for reasons which will
shortly be given), and if we find aquatic Insects in two
different orders, which cannot be supposed to be
derived one from the other, we believe that the
adaptation to aquatic life must have been effected
independently. For the common origin of both
orders is a stock unknown to us by observation, but

A

which we have good reason to suppose was not
aquatic. When we have a smaller group, such as an
aquatic family, to consider, we must ask whether it is
more likely that the common stock from which it is
descended was aquatic or not. In a few cases there
is reason to believe that two or more aquatic families
have descended from a common stock which was
aquatic also, but this is unusual. The aquatic families
of Insects are seldom wholly aquatic, they generally
include some terrestrial forms, and these axe, prima facie
the more primitive, or less altered in the matter of
habitat. \Ye do find some cases of families all of which
are aquatic, and so related to allied families which
are in the same case, that it is most likely that the
common stock in which all originated was aquatic

INTRODUCTION 5

too. But this is not very frequent, and the natur-
alist, exercised in cases of affinity, who studies a
table of aquatic Insects, is likely to conclude that
different families, and different genera within one
family, and different species within one genus have
betaken themselves independently to an aquatic life.
I think we can say with a considerable degree of
probability that this change of habitat from terres-
trial to aquatic has taken place in the class of Insects
at least a hundred times quite independently, and
the number may be very much higher than a
hundred.

I have spoken with confidence of the original
terrestrial habitat of Insects, and this requires some
further discussion. Writers have been found to main-
tain the very high antiquity of the aquatic habitat,
and to suppose that many, if not all, of existing
Insects had aquatic ancestors. What can be found
out on this point ? It is of no use to appeal to the
geological history of the earth, for the fossil remains
of Insects are too few and scattered to justify any
conclusion as to the mode of life of the first Insects
which appeared on the globe. But we are not alto-
gether without the means of forming an opinion. One
of the most universal features of Insects in all stages
of growth is the aeration of the tissues by air-tubes
or tracheae, which ordinarily open on the outside of
the body by special valvular inlets, the spiracles. Even
in aquatic Insects the spiracles are commonly present
and functional, and we can hardly name an Insect
which has not tracheal tubes, though they may in rare
instances be greatly reduced. This is pretty good

6 NATURAL HISTORY OF AQUATIC INSECTS

proof that Insects were primarily air-breathing
animals, but the distribution throughout the different
orders and families of the species which breathe gas-
eous air and of those which breathe only air dissolved
in water, leaves no doubt that the last-named species
are those which deviate from the general and primitive
rule. I think that every entomologist would agree
that the Insects which exhibit the most generalised
and presumably the most primitive structure are
terrestrial rather than aquatic.

Insects are just the sort of animals that might have
been supposed likely to change from one medium to
another with great ease and advantage. They are
active, hardy and ingenious. The plan of their bodily
structure has lent itself to a multitude of special
adaptations of other kinds. Notice the various shapes
of Insects. We find among Beetles, for instance, some
so round and fat that they can hardly creep, some
long-legged and nimble, one as flat as a wafer for
convenience of creeping under the adherent bark of
fallen trees, some with long wings, others with short
wings, others with no wings at all. Then among
caterpillars what a range of shapes ! They may be
long and narrow, or egg-shaped, hard or soft, with
many feet or with none, hairy or smooth, and in
special cases with the oddest horns, warts, brushes,
tentacles and hooks. Notice too, the extraordinary
variety in their food. Insects can be found which live
upon wood, upon earth and the things contained in
earth, upon fungi, upon decaying seaweed, upon paper,
upon fur and wool, upon leather, upon argol (crude
potassium tartrate), in addition to the animal and

INTRODUCTION 7

vegetable products which are the ordinary food of the
higher animals.

Equally remarkable is the facility with which Insects
adapt themselves to various conditions. The most
primitive Insects are terrestrial, running about on the
surface of the earth to seek their food. But many
have learnt to burrow into the earth, and this in a
great variety of ways. A large proportion have
learnt to fly. Some run their galleries in the wood of
trees ; others eat out slender mines in the thickness
of a leaf; others live in the bodies of dead animals ;
and there are not a few which burrow in the bodies
of living animals — a painful example of the cruelty of
Nature ! The adaptations of Insects to aquatic life,
varied and ingenious as we shall find them to be, are
only fresh examples of the versatility which character-
ises the whole class.

Even if we confine our attention to the Insects
which pass part or the w7hole of their lives in water,
we shall find a great diversity in their habitats. Still
water abounding in vegetation is suitable to many ;
others like running water, and some few prefer torrents
or cascades, though in this case they are seldom
exposed to the full force of the current. The most
rapid parts of an English mountain brook will, how-
ever, often be found to harbour certain kinds of
Caddis-worm and the larvae of Simulium. Water
loaded with dead organic matter is congenial to other
Insects, and the larva of Eristalis tenax prefers water
which is putrid and stinking. Salt water is selected by
others. The sea nourishes a considerable number,
and brine-pits, notwithstanding the bitterness and

8 NATURAL HISTORY OF AQUATIC INSECTS

density of the water with which they are filled, have
an Insect population of their own. Hot springs are
preferred by some, arctic snows by others ;l some few
Dipterous larvae live immersed in the juices which
flow from wounded trees and many in the fluids of
living animals. Some attach themselves to the roots
of submerged plants, and several bury themselves in
the substance of the freshwater sponge.

THE DOMINANCE OF INSECTS.

The countless adaptations of Insects to the most
various conditions not only illustrate the extraordinary
elasticity of their plan of structure, but also their
power in competition. The profuse variations of the
Insect type would never have been called forth, had
not the type been eminently successful in the struggle
for existence.

Such success may be shortly named dominance, and
Insects are an excellent example of a dominant class.
They are able to drive out their rivals, able to main-
tain and even to enlarge their territory in spite of all
competition. \Ye cannot ascertain such qualities by
direct observation of the struggles which are going on
everywhere around us. Human experience is too
short and too blind for direct measurement of these
secular changes. But though we cannot see for our-
selves the turns of the slow and infinitely complex
struggles of the races of animals and plants, we have
before us the results of ages of such strife. The com-

1 Examples are given by Kirby and Spence, Introduction to

Entomology, Vol. II., p. 231 (1826).

INTRODUCTION 9

monest animals of to-day are those which have so far
proved victorious, and numerical strength is one mark
of dominance, a sure one, when it can be clearly
ascertained.

Any group of animals or plants which is numeri-
cally very strong will be found to include a large
number of species, and these will be finely graded,
one species being with difficulty distinguished from
those which most nearly approach it. Dominant
species are usually combined into what we may call
continuous groups. It is only when a group is domi-
nant, or so to speak in easy circumstances, that many
nearly allied forms can maintain themselves side by
side within it. For declining groups there is a
rigorous selection. The principle applies to other
things beside natural species. It has become a proverb
that even bad workmen find employment when trade
is good. Growing industries, such as lithography or
paper-making, turn out an immense variety of forms
and qualities ; but articles which are going out of use,
say tallow candles or sand-glasses, can only be bought
in one or two patterns. Sealing-wax thirty years ago
was sold in an endless variety of colours, but the com-
petition of adhesive envelopes has now reduced the
number to one or two. I am told that while you can
get ample choice of cloth-covered buttons, you can
hardly buy more than one kind of polished steel
buttons, which have now gone out of fashion.

All prosperous species, families, orders and classes
at least hold their ground. If they can be successfully
invaded, they are no longer dominant. Dominant
groups are therefore geographically continuous ; they

io NATURAL HISTORY OF AQUATIC INSECTS

lie within a ring-fence, and the facts of distribution
will soon tell us whether a group is dominant or not.
It is especially characteristic of the most dominant
groups that they can hold their own in the wider
areas, where the struggle for existence is most severe.
Groups which can maintain themselves in the wide
continents of the northern hemisphere, will easily
prevail over the populations of the southern islands ;
groups which can maintain themselves in the vast
fresh-water lakes of North America, will easily occupy
the small fresh-water lakes and ponds of England, if
ever they are brought into competition with the in-
digenous population.

All the various marks of dominance are united in
Insects, and are there exhibited in a striking degree-
Numerical abundance, zoological continuity, or the
co-existence of many allied forms, geographical con-
tinuity, and predominance in the widest areas of
competition distinguish the whole class. There are,
of course, degrees of dominance among them. The
Diptera are eminently dominant among Insects, the
Muscidae dominant among Diptera, and so on. The
extraordinary variety in form, size, and habitat which
Insects present is in keeping with all that we know
of their range and frequency.

What are the qualities which confer dominance
upon Insects or any other group of animals ? The
question is too hard for us. It may be one quality
here, and another there. But though we must not
speak with confidence on so difficult a question, there
is some likelihood in the supposition that adaptability
to new conditions would be one ground of success. If

INTRODUCTION u

two groups of animals, both terrestrial, were to have a
supply of food opened up to them in the water, it
seems probable that the group which had proved
itself the stronger by land would be more likely to
appropriate it. The past success of the European in
competition with the Hottentot gives us some reason
to believe that Europeans will occupy any country at
present uninhabited before the Hottentots will. But
we soon get out of our depth in such discussions as
this.

DEGREES OF ADAPTATION TO AQUATIC

CONDITIONS.

How did Insects ever come to seek the water, seeing
that their mode of respiration is primarily adapted to
another element ? We can see almost all the steps
of the adaptation on the shores of our rivers, lakes
and seas. We can see Dipterous larvae which, like
the "leather jacket ': (the larva of the Daddy-long-
legs), burrow in the ground for their vegetable food,
and devour the roots of grasses. Other larvae of the
same family (Tipulidre) prefer moist earth in the
neighbourhood of streams. Others again live im-
mersed in water, or mud saturated with water, though
they come to the surface at times and push their
tails, which carry the spiracles, into the air. Some few
have become so completely aquatic that they seldom,
if ever, come to the surface, and all their supply of
oxygen is obtained from the water. Again we find
on any sandy sea- shore little Beetles, which hide in
the decaying seaweed. Some of the species only

12 NATURAL HISTORY OF AQUATIC INSECTS

venture a little way below the level of the very highest
spring-tides ; others are found a little below the
ordinary high-water mark, and these are immersed
for a few minutes twice a day. But others withstand
prolonged immersion, and cannot be drowned in salt
water. There is vegetable and animal food in the
water, and some creatures will find it and devour it.
Insects, nimble, enterprising, and ingenious, have ven-
tured in and have succeeded in wresting a share
from the slow Mollusks to whom it might seem more
properly to belong.

The life of some aquatic animals is greatly com-
plicated by the peculiar properties of the surface-film
of water. Like other external conditions, the surface-
film may be either a hindrance or an advantage,
according to the way in which it is treated. Those
who have had experience of the extraordinary versa-
tility of living things, and of their power to adapt
themselves to the most various conditions, will be
prepared to find that plants and animals have been
able in various ways to turn to account the properties
of the surface-film.

THE SURFACE-FILM OF WATER.1

I fear that it is not quite safe to presume that the
properties of the surface-film are familiar to every
reader. Let me, at the risk of superfluous explanation,
run over some of the elementary facts.

Take a tumbler and pour water into it until it is

1 A fuller but quite elementary account of this subject will be
found in Prof. Boys' charming little book on Soap Bubbles.

INTRODUCTION 13

full to the brim, i.e. until the water-level is as high as
the edge of the tumbler. Then we say that the
tumbler is full of water. But it will still hold a good
deal more without spilling. Add water by teaspoon-
fuls, or drop coins in, one by one. The level rises
well above the edge, and still if due care is taken, no
water is spilt. The tumbler can be made over-full
of water. It is plain that something restrains the
natural tendency of water to seek the lowest level,
and to escape by flowing over the edge.

Can it be that the water is sensibly viscous, like
glycerine or thick treacle ? This can hardly be the
explanation, for the excess of water does not level
down gradually, but takes up a position of rest in
less time than it costs us to focus the eye upon it.
It is not the want of fluidity in the water which
prevents it from flowing over the edge. There is a
film overspreading the surface, even of the purest
water, which prevents it from escaping until the
pressure becomes considerable.

Dip your ringer into water and lift it out again.
The water does not all run off, but flows down to the
tip of the finger, and there forms a big drop. What
holds up the drop ? It is the surface-film, which does
not differ in its composition from the water within,
but is for the moment in a peculiar physical state.
The film is incredibly thin. It can be renewed in an
instant, fresh water-particles taking its place, while
the old film loses all its peculiar features, and mingles
with the water beneath the surface.

The surface-film can be made evident in another
way, Take a clean needle, poise it horizontally on

14 NATURAL HISTORY OF AQUATIC INSECTS

the prongs of a fork, and lower the fork and needle
steadily upon the surface of the water in a basin. If
the fork is taken away from beneath, the needle will
be left floating on the surface. Steel is many times
heavier than water, but the resistance of the surface-
film prevents the needle from sinking. Once beneath
the surface, the needle sinks rapidly to the bottom.

The surface-film is in a state of tension, i.e. it exerts
a pull. It behaves like a drum-head, or a sheet of
stretched india rubber. This may be made evident
in various ways. One of the simplest is this. Take
two lucifer matches, which are to form opposite sides
of a square. Tie their ends together with pieces of
sewing-cotton, which will form the remaining sides.
Then run a needle into the middle of one of the
matches to make a handle. If this frame is dipped
into soapy water (a strong solution is required) and
then lifted out, a transparent film of soapy water will
overspread the square. This film will be in a state of
tension, as you will see by holding the frame upright
with the two matches horizontal. The threads will
be curved inwards by the pull of the film. Why do
we use soapy water in this experiment ? Because the
film of pure water will not last long enough for us to
examine it. It breaks up instantly into a crowd of
small drops. The film of soapy water, lifted off in
this way, is not a simple surface-film. It consists of
a double film, with an appreciable quantity of water
imprisoned between the two layers.

The two or three experiments just described will,
if actually performed, establish in the reader's mind
some elementary truths, which have a good deal to do

INTRODUCTION 15

with the behaviour of Insects living at the surface of
water. The particles close to the surface are in a
peculiar condition of aggregation, and temporarily
cohere to form a film. The film, as we have seen,
offers resistance to the passage of solid bodies', and
can therefore support a weight or keep a buoyant
object from rising through it. It also exerts a pull,
and we shall find that advantage is taken of this pull
by certain aquatic Insects. The extreme tenuity of
the film, which is thinner than our imagination can
realise, helps us to understand how it is that only
small objects are affected by it. Ships, boats, swim-
ming quadrupeds, and all objects whose weight is
large in comparison with their contour, are practically
uninfluenced, but objects whose dimensions are given
in fractions of an inch may be largely controlled by the
peculiar properties of the surface-film. The larva of the
Gnat turns these properties to account in a peculiarly
interesting way, as do many other aquatic Insects,
Crustacea and floating plants. On the other hand,
minute aquatic animals, approaching the surface of
water without precaution, may find it a death-trap, as
we learn by observation of certain small Crustacea,1

EQUILIBRIUM OF AQUATIC INSECTS.

The density of water is so much greater than that
of air as to require peculiar adjustments in the
equilibrium of aquatic animals. A simple case is pre-
sented by the air-breathing aquatic Insects. These,

1 D. J. Scourfield, Linn. Jotirn., Zool., Vol. XXV., pp. i — 19,

pi. I, 2, 1894.

16 NATURAL HISTORY OF AQUATIC INSECTS

with few exceptions, are lighter than water.1 Hence,
if by loss of respirable air, or from any other cause,
they are disabled, they have only to cease to struggle
in order to come to the surface, and it will in all cases
be found that they float upwards in such a position
that the part of the body by which air is taken in
comes to the top.

Locomotion in any medium by an animal of
complex structure involves some power of orientation,
and especially a rapid and instinctive judgment as to
the horizontal plane and the upward and downward
directions. The pull of gravity may furnish the
necessary indication, but an animal which swims in
water, or flies in air, may be unable to determine the
horizontal plane with facility, especially if its body is
nearly of the same density as the surrounding
medium, since the pressure will be approximately
equal on all sides. The semi-circular canals of
Vertebrates and the so-called auditory organs or
otoliths of many Invertebrates are probably organs
for the perception of the horizontal position. Some
animals, however, such as Insects, possess no organs
of this kind, although they are excellent fliers or
swimmers. How are they enabled to maintain their
course at a fixed distance above the surface of the
earth or below the surface of the water ? Sight may
help, but that sight does not furnish the sole indica-
tion may be inferred from the behaviour of sightless
animals, such as the inhabitants of dark caverns, or
animals temporarily blinded for the purpose of

1 Among the exceptions are Corixa, and the larvae of the,
Gnat, Agabus, and Hydrobius.

INTRODUCTION 17

experiment. These have been found to be capable of
maintaining a horizontal course, or of ascending and
descending according to need. Their perception of the
upward and downward directions is, as experiment
shows, falsified by artificially induced changes in the
specific gravity of the medium, so that it is probable
that they are guided chiefly, if not altogether, by the
action of gravity.

We have seen that air-breathing aquatic animals
are, as a rule, lighter than water, and that when they
come to rest they float upwards in one particular
position. The normal attitude is maintained by the
distribution within the body of two or more sub-
stances of very different density, such as air and
flesh or shell. Whatever the density of the medium,
and whether the body rises or falls, it will tend to
maintain the same position, the lighter tract tending
to rise.

Ground-feeders, and swimmers which do not
breathe gaseous air, maintain the normal attitude in
a different way. Here the body is of nearly uniform
specific gravity, and the normal attitude is secured by
the movements of the limbs and the form of the
external surface.

In ground-feeding aquatic animals, whose bodies
are a little heavier than water, the legs are usually
placed ventrally, and the back is rounded from side
to side. A Lobster is a familiar example. Here the
convex dorsal surface tends to move foremost, and
the action of the limbs treading the water, helps to
maintain the same direction of movement. As a
rule the limbs are turned down, and the body con-

C

iS NATURAL HISTORY OF AQUATIC INSECTS

sequently rises when they are energetically moved,
but the attitude is very easily reversed.

If such an animal ceases to use its limbs, and is
moved through the water by a current or by its own
weight, it will tend to move back foremost. Hence
a dead Lobster or Crayfish thrown into water will
commonly come to rest lying on its back. Swimmers
in water, such as some Ephemera-larvae, find it easy
to invert their bodies. Some Crustacea swim in all
sorts of attitudes, back upwards, back downwards, or
sideways, and change incessantly from one position
to another. This is effected by the bending of the
body. If the body is much curved, the centre of
gravity will lie outside it, in the concavity and well
below the line about which the moments on both sides
form a couple. The convex side will tend to sink.
If the ventral side is made convex it will tend to
turn downwards, and so with any other side.1

THE WINTERING OF AQUATIC INSECTS.

»

Winter, of course, brings many hardships upon
aquatic Insects, as the great reduction in their num-
bers proves. The enormous number of eggs laid by
so many of them is doubtless connected with the
heavy risks to which they are exposed during half the
year. I find comparatively little information recorded
even upon so fundamental a point as the stage in
which particular Insects pass the winter. It is much

I have derived useful notions on this subject from a short
paper by Bef.he (" Erhaltung des Gleichgewichts," Biol. Ccntral-
blatt, 1894), which contains, however, some serious mistakes.

INTRODUCTION 19

to be desired that observing persons would attend to
this and like subjects instead of piling up lists of
species, which tell us so little that signifies.

The common rule is, I think, that aquatic Insects
winter as larvae. Nearly all aquatic Diptera, Dragon-
flies, May-flies, Stone-flies and Caddis-flies do this.
Occasionally, however, the winged individuals hiber-
nate. Examples are furnished by the Gnat, and one
of the Dragon-flies (Lestes).

Fully armoured aquatic Insects, such as Beetles and
Bugs, commonly pass the winter in the winged state,
burying themselves in the mud or in the earth during
unusually severe weather.

Aquatic Insects which have wintered as larvae
usually undergo transformation and lay their eggs in
the following spring. From these eggs a summer
generation proceeds, which becomes ready for egg-
laying in September, and so the cycle comes round.
This is the usual course with aquatic Diptera of small
size. There may be more than one summer-brood,
and sometimes the new generations appear so fre-
quently that larvae, pupae and flies can always be
found during the summer half-year. This is the case
with Chironomus, for instance. Large Insects, and
the great majority of the aquatic Beetles, produce
at most one brood in the year.1 Egg-laying takes
place in spring or early summer, and the transforma-
tion takes place in the height of summer, though not
always in the summer which immediately follows the
hatching-out of the eggs. Upwards of a year may be
spent in the larval stage.

1 Hydrobius produces a second brood in autumn.

C 2

20 NATURAL HISTORY OF AQUATIC INSECTS

SOME CASES OF ABRIDGED LIFE-HISTORIES

In the following pages many instances will be
givren of Insect larvae which have become completely
adapted to aquatic conditions. The spiracles or
inlets of the air-tubes are closed ; some kind of gill
takes their place ; fins useful only in water are
developed. The pupa, which succeeds to the larva,
is occasionally wholly aquatic too, though it is much
more common for the Insect when the time of
pupation is at hand to creep out of the water and to
begin to breathe gaseous air. Still there are a very
few cases of pupae whose respiration is adapted to a
completely submerged life. The pupae of Chironomus
and Simulium obtain their whole supply of oxygen
from the water. But there is as yet no one instance
of an imago, or adult Insect, which breathes the air
dissolved in water. All, when they reach their final
stage, become air-breathers, drawing in gaseous air by
open spiracles. This is true of the aquatic equally
with the terrestrial imago. A Dytiscus, a Hydro-
philus, a Water-scorpion or a Water-boatman must
be able occasionally to reach the air with some part
of its body, and will perish if prevented from doing
so. Some fully developed insects can, it is true,
remain submerged for many hours together, but all
require to take in gaseous air at intervals. Why, we
may ask, could not that adaptive power of Nature
which has enabled so many larvae and some few pupae
to subsist upon dissolved air, have furnished us with
one or two examples of an adult Insect capable of
doing the same ?

INTRODUCTION 21

I am not bold enough to say that the case is theoreti-
cally impossible, but if it is ever realised by the dis-
covery of actual examples, we may expect to find that
the whole advantage of the transformation has been
sacrificed. For transformation is merely a step to
the acquisition of wings, and wings would be useless
except in air. The power of flight implies great
muscular activity, keen senses, food which can be
quickly absorbed and highly nutritious in proportion
to its bulk, mouth-parts adapted to such food. Hence
the striking difference which so often arises between
the slow, heavy and voracious larva and the nimble,
honey-sucking imago. There is a division of labour
between the two stages, and each becomes specialised
for its own purposes until the extremes can only be
reconciled by a prolonged resting-stage. The pupa
and all that relates to it are subordinate to the
acquisition of wings.

The acquisition of wings is not however an ultimate
purpose, but only a means to an end. That end is,
primarily, the dispersal of the eggs in fresh sites.
Where the food consists of the foliage of a particular
thinly distributed plant, all the eggs must not be laid
together, lest the larvae, being over-crowded, should
consume their whole supply of food and perish of
starvation, leaving other plants of the same species
untouched. If the female were endowed with the
power of flight, able to absorb some highly nutritious
food which would neither require much time to col-
lect nor over-load her body, and gifted with adequate
powers of perception, she might find out a number of
scattered plants, and lay her eggs upon them a few at

22 NATURAL HISTORY OF AQUATIC INSECTS

a time. Again, the many Insects which frequent
small pools, liable to disappear in summer drought,
will enjoy a great advantage if they are able to fly
abroad, and lay eggs in places which offer a supply
of food and yet are not permanently habitable.
Where these considerations do not apply, as for
instance in the case of Moths whose food-plants
(grass, heather, &c.) are social and not easily ex-
hausted, or of Moths whose larvae are indiscriminate
in their food (the Vapourer is a good example) we
often find that the female is wingless.1

The mating of the male with the female is another
motive for the acquisition of wings. Were the fully
developed Insect restricted, like the larva, to one pool
or to one food-plant, choice of mate and mixture of
families would be hard to obtain. Hence even in
those cases (wingless Moths, Scale-insects, Strepsip-
tera) where abundance of food or parasitic life has
led to the suppression of the wings of the female, and
where the eggs are not widely dispersed, the male
Insect is nearly always capable of flight.

Though no Insect is so completely aquatic in its
final stage as to be unable to live in the air, we do
find, as a rare exception, that the final stage is sup-
pressed, and that the pupa becomes capable of laying
fertile, though unfertilised eggs. This is the case
with Chironomus. Grirnm 2 found in his aquarium a
small species which in spring and summer laid un-

1 Weismann, Biological Memoirs. Essay on the "Duration
of Life."

2 " Die ungeschlechtl. Fortplantzung einer Chironomus-
Art." MJm. Ac. St. Petersb. VII. Ser., Tom. XV. (1870).

INTRODUCTION 23

fertilised eggs while still a pupa. These eggs pro-
duced larvae. The egg-laying pupa had on the under
side of the last abdominal segment a pair of special
orifices for the passage of the eggs. Occasionally,
after laying a few eggs, the pupa was transformed
into a fly. In autumn the normal stages were gone
through, and eggs were laid by winged females after
being fertilised by the males. Many other cases are
known of the production of fertile though unfertilised
eggs (Parthenogenesis) among Insects and other
animals. Reproduction by immature animals (Paedo-
.genesis), as in Grimm's Chironomus, is rare but not
unparalleled. The larva of a minute fly (Miastor
metroloas) found under the bark of the poplar, willow
and beech, was observed to produce viviparously
small larvae, which escaped by tearing open the body
of their parent and in turn produced other larvae after
the same fashion. This went on throughout the
greater part of the year, but in summer winged males
and females appeared, which laid a very few eggs of
large size. When these facts weie first published by
Wagner l they were received with a good deal of
incredulity, which has been completely removed by
the independent verification of the facts, and by the
discovery of two other allied species of similar life-
history. Parthenogenesis, Paedogenesis and viviparous
reproduction are all methods of abridging the tedious
operations by which in Insects a new generation is
normally produced. Where food is plentiful and
time precious, the mating of the sexes, the full

1 " Beitr. z. Lehre v. d. Fortpflanzung der Insectenlarven,'5
Zeits.f. wiss. Zool. Bd. XIII. p. 513 (1863).

24 NATURAL HISTORY OF AQUATIC INSECTS

development of the winged imago, and the develop-
ment of the eggs after mating may all be omitted or
shortened. But the normal process is never wholly
suspended. Fertilised eggs are produced at regular
intervals by fully developed males and females, and
most commonly towards the end of summer, as if the
approaching season of scarcity and cold could only be
faced by fertilised eggs or their product. We cannot
but speculate on these curious deviations from the
familiar course of nature, but let us hold our specula-
tions cheap. Before long we may expect some im-
practicable fact to start up, as if maliciously, and
confound our theories.

LIVE NATURAL HISTORY.

I hope to persuade some of my readers to take up
the study of aquatic Insects in a practical way, and it
will be desirable to tell them what they may expect
if they do. Aquatic Insects are not a definite group
of animals from the point of view of zoologists, but
a chance lot. The grave labours of the anatomist
or systematist would be better bestowed upon a
special order or family. Aquatic Insects, however,
make a capital study in Natural History. While you
are looking for one kind you will come across another.
The same methods and the same tackle will do for
all. If a young student wants to observe the ways of
living creatures, we may recommend aquatic Insects
to him as an accessible and very imperfectly explored
field. He will find plenty of undescribed forms and

INTRODUCTION 25

plenty of beautiful contrivances which no one has
ever taken the trouble to observe. But to make
out the way in which the exquisite machinery of
nature is meant to work is no childish pursuit. The
very attempt will lead the naturalist to acquaint
himself with scientific laws which seem altogether
foreign to Natural History; it will exercise his in-
dustry and sagacity ; it will extend his knowledge of
the possibilities of life.

This is a study, I mean the study of living animals,
which is not very seriously prosecuted in our time.
The naturalists of old, Swammerdam and Reaumur,
and De Geer, made this their great interest. The
needful work of classification has since drawn off
many students, and exciting questions concerning
the descent of animals have drawn off others. But
Natural History, which I use here to denote especially
the study of living animals and plants, will surely
revive, and charm mankind once more.

The zealous naturalist, setting out upon a voyage
of discovery, will generally be too ambitious at first.
He will try to cover too much ground, and to do
too much with his own hands. Let us consider what
he may fairly hope to do. The first aim of the
beginner is usually to collect and name. He likes to
see a growing store of properly classified Insects.
He likes to give the right name to any one that is
shown him. But this feeling must be checked if he
is to do much work upon habits or structure. The
man who can name many Insects can seldom do
anything but name them (there are some conspicuous
exceptions !). Useful it is to all of us to have such

26 NATURAL HISTORY OF AQUATIC INSECTS

men, and to benefit by their knowledge. They give
us the key to what has been found out and set down
in books by many generations of observers. But the
student of Natural History or Anatomy will not, as
a rule, be even moderately good in naming. He will,
it is to be hoped, be able to place an Insect approxi-
mately in its right place, but he will not carry generic
and specific characters in his head. If it is the
Natural History of Insects which attracts him, he
will do well not to seek to add all the attainments of
the Entomologist. The collection that he should

* t

possess will not be extensive or showy. He will use
up his specimens to see how they are made ; he
will have no time for the niceties of the professed
collector.

I hope it will not be too discouraging if I add that
if he is bent upon increasing knowledge he will work
slowly, and will cover little ground. It takes a man
who has other occupations to mind about a year to
get a fair notion of the structure and mode of life of
a new Insect. But to verify the descriptions of good
observers, and so to be guided over the ground, is of
course a far more rapid business, and this is enough
for most of us. Still we do not taste the full delight
of Natural History unless we attempt to walk where
no one has walked before.

It is necessary to warn the beginner that unless he
has that sure perception of truth and error which
belongs to very few among men, he will continually
make mistakes, and very humiliating these mistakes
are, especially when they have been published with
full confidence. There is no department of human

INTRODUCTION 27

enquiry with which I am even slightly acquainted
which presents so many pitfalls as the interpretation
of natural contrivances. I have learnt by sad ex-
perience to believe that my own first attempts to
interpret a new structure are nearly always wrong,
and it would not be hard to show that some very
eminent naturalists have had an experience not
altogether unlike mine. The only remedy is to be
patient, and to go on working until your explanation
has stood the test of time and wider experience.
Above all, do not print anything the moment you
have come to believe in it. When you have done
your utmost, you have not taken away the possibility
of being seriously wrong, but to print in haste is to
invite disaster.

This is a discouraging reflection, and may lead
some to prefer studies where surer if less interesting
results are to be won. But it will never do to turn
away from the interpretation of natural contrivances
merely because the task is difficult and risky. This
is one of the permanent interests of mankind ; to
know how the wonderful adaptations of plants and
animals came into existence, and what purposes they
serve, is the great aim of the study of Nature. And
experience shows that though many investigators fail
at times, and though some fail altogether, substantial
progress is made in nearly every generation. We now
know much more about natural contrivances than
was known before Charles Darwin, and almost in-
finitely more than was known to Swammerdam 01
Leeuwenhoek.

28 NATURAL HISTORY OF AQUATIC INSECTS

THE GROUPS TO WHICH AQUATIC INSECTS

BELONG.

The aquatic Insects are here described in the
following order :-

1. Coleoptera (Beetles).

2. Diptera (Two-winged Flies).

3. Hymenoptera (Ichneumons, &c.).

4. Lepidoptera (Moths).

5. Trichoptera (Caddis-flies).

6. Sialidae (Alder-flies).

7. Perlidae (Stone-flies).

8. Ephemeridae (May-flies).

9. Odonata (Dragon-flies).

10. Rhynchota (Water-scorpions, &c.).

11. Thysanura (Spring-tails).

The names employed by anglers should be noted,
as much information respecting the habits of aquatic
Insects can be extracted from anglers by those who
speak their language. Ronalds' Fly-fishers' Entomo-
logy and other books give the zoological names of
every species imitated by the angler. In the following
short list each is merely referred to its proper
order.

2. Diptera. — Golden Dun Midge.

5. Trichoptera. — Blue Dun, Little Red Spinner,
Sand Fly, Grannom, Turkey Brown, Dark Spinner,
Silver Horns, Cinnamon Fly.

6. Sialidae. — Alder or Orl Fly.

7. Perlidae.— Red Fly (Old Joan), Stone Fly,
Willow Fly (Shamrock Fly).

INTRODUCTION 29

8. Ephemeridae. — March Brown (Dun Drake), Great
Red Spinner, Yellow Dun, Iron Blue Dun, Jenny
Spinner, Little Yellow May Dun (Silk Fly), Sky Blue,
Green Drake (May Fly, Cardew), Grey Drake, Orange
Dun, Black Drake, Dark Mackerel, Pale Evening
Dun, July Dun, August Dun, Whirling Blue Dun,
Little Pale Blue Dun (Willow Fly).

CHAPTER l

AQUATIC BEETLES

THE WHIRLIGIG BEETLE (GvRiNus)1

GYRINUS, also called the Whirligig Beetle or Steel-
coat, is a small blue-black Beetle, about a quarter of
an inch long, which abounds on the surface of still
streams. It is sometimes seen disporting itself on
a clear pond or small puddle, but it prefers gently
running water. During the summer months, the
Whirligigs congregate in swarms, darting swiftly to
and fro like Swallows or Bats with movements so
rapid that the eye can scarcely follow them. Now
and then, especially when alarmed, they dive to the
bottom of the water, or swim about beneath the
surface. When they descend, they carry with them
a bubble of air attached to the hinder end of the
body, which glitters like quicksilver. They may be
often seen clinging to submerged weeds, but as soon
as all is still again they let go their hold, rise without
effort to the surface, and resume their gyrations. It

1 I have to thank Mr. W. F. Baker for some useful observa-
tions on this Insect.

CM. I

AQUATIC BEETLES

is particularly easy to observe them from a low
bridge crossing a still stream.

FIG, i. — Gyrinus marinus (G. natator hardly differs) Head, seen from beneath,
showing mouth-organs and lower pair of eyes ; antenna ; third leg ; and tarsal
joints of do. separate and extended.

The Gyrinus appears quite early in the season, but
only in twos and threes. Those which appear in

32 NATURAL HISTORY OF AQUATIC INSECTS CH.

early spring have passed the winter in a torpid
condition as full-grown Insects. A new generation
of larvae is now produced. They grow rapidly, pass
through their transformations, and emerge as perfect
Insects when the height of the summer is passed.
They are most plentiful about the beginning of
September ; after that, their numbers steadily de-
crease, and by the middle of November very few can
be found ; most of them have now retreated to their
winter quarters, though a spell of fine weather, even
in the depth of winter, will tempt a few to the surface.
They pass the winter months buried in mud among
the roots of water-plants, and can be dug out with a
scoop. If kept indoors in an aquarium, they clo not
become torpid, except in the severest weather. Under
such conditions their numbers rapidly decrease, for
food is scarce, and when the Insect is active, its
tissues waste.

The full-grown Gyrinus, like most other adult
Beetles, flies with ease. It seems to be unable to take
flight from the surface of the water, but finds it
necessary first to climb up the stem of some water-
plant, where it rests for a time. Then it slowly opens
its wing-covers, extends its long wings, and takes to
flight. When kept in captivity, they are unwilling to
remain very long at a time upon the surface of the
water. They will climb up the glass sides of an
aquarium with great effort, making use chiefly of the
hindmost pair of legs, and struggling with the
tenacious film of water which they bring up with
them.1 They will creep nearly an inch up the
1 The action is more fully explained on page 162.

I AQUATIC BEETLES 33

perpendicular sides of the glass, and it takes them
about half an hour to rise so high. Failures do not
discourage them, and no matter how long they
have been kept in captivity they are never weary of
endeavouring to creep out of the tank.

When handled, they give off a milky fluid of un-
pleasant odour from nearly all the joints of their
body, but especially from the fore and hind edges of
the thorax. The scent is rather like that given off by
Cockroaches. According to most authors, the perfect
Gyrinus is largely carnivorous. Westwood says that
he has ascertained that its food consists of small dead
floating Insects. Fowler remarks that the broad blunt
mandibles point to a partially vegetable diet, though
the sharp sickle-shaped maxillae favour a different
conclusion. In captivity they feed upon water-plants.1

The body of the adult Beetle is of oval shape,
convex above and flat beneath. The head is sunk in
the thorax. The end of the abdomen projects, and
forms a short tail upon which are carried two small
retractile prominences. Between the tips of the
wing-covers and the extremity of the body rests the
air-bubble which the Beetle carries with it when it
dives.

1 Air. W. F. Baker favours me with the following observa-
tion : " I was once watching some Gyrini, when a gad-fly came
flying round my head. I struck at it, and knocked it into the
water, stunned if not dead. Two or three Gyrini seized it, and
shortly afterwards the whole swarm clustered round. In a
short time nothing was left except the wings of the fly, which
were allowed to drift away. This is the only occasion on which
I have seen anything of the sort. I could never get captive
Gyrini to eat dead flies."

D

34 NATURAL HISTORY OF AQUATIC INSECTS CH.

The compound eyes are each divided into two
separate portions, the socket of the antenna being
interposed. Since one mass of facets is above and
the other beneath, it has been ingeniously conjectured
that one serves for vision in air and the other in water.
This conjecture has never been tested with the care
necessary to confirm or refute it. I once endeavoured
to determine by direct observation whether the lower
.patch is actually submerged or not, but found that
the capillary curves about the head and body render
it very difficult to decide where the water-line comes.
These curves must greatly obscure or at least limit
vision by the lower lenses. The occurrence of divided
eyes in some Dragon -flies, in Adelotopus, a Carabid
beetle, which is found under the bark of trees,1 as
well as in many Dipterous Insects which never
approach the surface of water,2 must be accounted
for in some other way. The antenna has a very
unusual shape. The basal joint is small, the second
larger, the third much bigger, looking like a second
antenna outside the other. The remaining part of
the shaft is club-shaped and indistinctly jointed.
The tip is provided with pits and hairs, and is
probably highly sensitive. In the Dytiscidae the
antennae, as I learn from Dr. Sharp, want the usual
sensory structures and are always wet. The peculiar
form of the Gyrinus-antenna is probably a means of
keeping it dry.3 The Insect-antenna, when highly

1 Dr. Sharp, Ent. Month. Mag. V. p. 52.

2 Osten Sacken, Characters of the three Divisions of Diptera^
p. 447.

3 " In some aquatic Beetles (Gyrinus, Parnus) they [the

I AQUATIC BEETLES 35

specialised, seems to require air for the discharge of
its peculiar functions.

The wings and wing-covers of the adult Insect are
much like those of other Beetles, but the legs are very
peculiar. The third pair are broad and shaped like
paddles. They can deal a powerful backward stroke
upon the water with the broad side, and can be
drawn up again edgewise with greatly less resistance.
The effectiveness of the paddle is enhanced by the
long stiff hairs which fringe it, and especially by the
peculiar form of the tarsal joints. These are
expanded, and so articulated that the fulcra upon
which they turn come near together. The effect is
like that of the ivory tablets used for memoranda,
which are held together by a pin, so that they can
either be opened fanwise or closed in a moment. It
is impossible to study the action of the limb in the
swift darting movements of the Gyrinus, but the
power of instantaneous extension or collapse must
be of great practical service. The middle legs are
also expanded, but in a less degree. The fore legs
are prehensile. In the male the tarsal joints are
dilated, and bear a great number of circular organs,
commonly called suckers, which are, it is believed,
used to grasp the female.

Several Beetles have been observed to make a
squeaking noise, and among the number are Dytiscus
and Gyrinus. The sound is produced by rubbing the

antennas] are furnished with an auricle at their base, which
like the lid of a box, shuts them in when unemployed, and
protects them from the water."- -Kirby and S pence, Vol. III.
p. 516 (1826).

D 2

36 NATURAL HISTORY OF AQUATIC INSECTS CH.

under side of the wing-covers against the end of the
body, and is probably a call to other Beetles of the
same species. Gyrinus always makes this sound
before taking flight. A flying Gyrinus hums like a
Bee, apparently in consequence of the whistling of
the air about its hollow wingr-covers.

o

The female Beetle lays a number of elongate oval
eggs, end to end, upon the leaves of water-plants.
They are found beneath the surface of the water, and
sometimes at a considerable depth.

Schiodte 2 has given an excellent set of figures of
the larva of Gyrinus. It is 14 mm. (4 in.) long.
Behind the head come three thoracic segments, of
which the foremost is protected by a dorsal plate, as
is usually the case with burrowing larvae. The legs
are long and slender. Each of the eight foremost
abdominal segments carries a pair of gills, and the
ninth two pairs. A small tenth abdominal segment
is armed with two pairs of long, sharp, curved hooks,
which are believed to be of use in climbing. The
general appearance of the larva is that of a small
Centipede. The mouth-parts are those of a carni-
vorous larva, the mandibles especially being pointed,
perforated by a slit beneath the apex, and suctorial.
The gills are long, tapering, and fringed with fine
hairs. A tracheal tube traverses each, and gives off
many fine branches to the delicate walls.

It is generally thought to be easy to understand
the principle of such a respiratory organ. The water

1 " De metamorphosi Eleutheratorum, Bidrag. til Insekternes
Udviklings-historie." Pt. I. pi. iii., figs. 1—9. Kroyer. Naturh.
Tikskr. (1862).

AQUATIC BEETLES

37

in which the gills are bathed contains much dissolved
oxygen, and this oxygen is supposed to diffuse through
the delicate wall, while the carbonic acid formed within
the body passes out. Little
attention has been paid to the
circumstance that in a number
of aquatic Insects the tracheal
tubes become thereby filled with
a gas, probably rich in oxygen.
How this is accomplished I for
one cannot understand. I be-
lieve that no physicist would
undertake to set up an apparatus
which would without rise of
temperature or diminution of
pressure remove dissolved oxy-
gen from water, and store it up
in a gaseous form within a closed

o

chamber. To imitate the appa-
ratus of the Insect, there should
be no gas whatever within the
closed chamber to begin with.
In the fresh-hatched larva of
certain aquatic Insects, or in the
young pupa of others, tiny bub-
bles can be seen to form inside
the air-tubes, and gradually ex-
pel the watery fluid, with which
they were previously filled. ^^^IcSST

The separation of a gas from

water or blood is not peculiar to Insects. Fishes
their swim-bladders, the Pearly Nautilus the

38 NATURAL HISTORY OF AQUATIC INSECTS CH.

chambers of its shell, some Zoophytes their floats,
some aquatic Crustacea and Polyzoa their eggs with
air or some other gas. Even the microscopic Proto-
zoon, Arcella, possesses the same faculty.1 Green
plants, under the influence of sunlight, liberate
gaseous oxygen from their tissues. This is probably
not due to the direct action of the sun's heat, for
it can be observed in cold running water, just
escaped from the fissures of a limestone hill.

The passage of dissolved gas through living mem-
branes is equally mysterious. Direct experiment upon
the freshly excised lung of the Frog shows that no
diffusion takes place until the cells are dead.2

It is well known to the makers of microscopic
preparations that tissues will not take up colouring
matter so long as the cells are alive, nor will the
natural colouring matter, e.g. of Beet-root, pass out
into water, until the cells are killed.

Bohr3 finds that the air-bladder of a Fish refills
after puncture and complete emptying, the new gas
containing as much as 80 per cent, of oxygen. But
if the branches of the vagus nerve which supply the

1 In this case the gas is carbon dioxide, whereas in others
it is apparently oxygen or some mixture in which oxygen pre-
dominates. Our information as to the composition of the
gases concerned is unfortunately most imperfect. Biot in 1807
showed that the gas in the air-bladder of Fishes is rich in
oxygen.

2 Tigerstedt und Santesson. Bihang till K. Svenska Vet.
Akad. Handl. Bd. XI. No. 2 (1886). NaCl could not be made
to diffuse through the wall of the lung so long as the epithelial
cells were alive.

3 C. R. CXIV. p. 1560(1892).

I AQUATIC BEETLES 39

bladder are cut, no more gas is formed. Oxygen
could not be made to diffuse into the air-bladder of a
Pike, which was filled with atmospheric air and sur-
rounded by pure oxygen, but when the epithelium
was killed by maceration in distilled water, oxygen
readily passed through by diffusion. Bohr accord-
ingly adopts Moreau's view, viz. that oxygen is not
•passed into the air-bladder by diffusion, but is formed
within it by " specific secretion."

The Gyrinus larva feeds upon water-insects, and
possibly upon other aquatic animals. Failing these,
it will eat the tender parts of submerged plants. It
has been noticed that well-fed larvae produce larger
Beetles than those which have had a poor supply.

The pupa of Gyrinus is so well hidden that few
naturalists have ever seen it. Modeer, quoted by
De Geer (Tom. IV. p. 364), says that about the
beginning of August the larva creeps out of the
water by climbing up the water-plants, and then spins
a grayish cocoon pointed at both ends. Enclosed in
this cocoon it changes to a pupa, and emerges as a
perfect Insect towards the end of the same month.
Modeer adds that the pupa is very liable to the
attacks of Ichneumons.

DYTISCUS.1

No one can hunt long for water-insects without
coming across the rapacious Dytiscus. There are

1 Dytiscus is the name originally given to this insect by
Linnaeus, and used in most systematic books, but Dyticus,
which has the same meaning (fond of diving), has also been
applied to it by some authors.

40 NATURAL HISTORY OF AQUATIC INSECTS CH.

many species and several genera in the family of
Dytiscidae, but it will be best to devote our space
chiefly to the' life-history of D. marginalis, one of the
biggest and the commonest of them all.

Both the larva and the imago are aquatic ; the
pupal stage is however passed in the earth. This
is the case with other aquatic Coleoptera too. The
explanation is not hard to discover. These Insects
are air-breathers throughout their whole life. As
larvae or adults they can come to the surface at
pleasure, and renew their supply of air. But the
pupa is almost motionless. Its limbs are being
reconstructed, and cannot effect more than a slight
change of attitude. Under these circumstances, the
Insect passes its pupal stage where air can be
procured at all times. For the sake of protection it
buries itself in the earth, but so near to the surface
that air can easily reach it.

Lyonnet observed that captive females of Dytiscus
laid their eggs at random in the water.1 For several
years in succession he watched the eggs so laid, and
none were ever hatched. At last one egg produced a
larva, but this soon died. Lyonnet was not aware that
the eggs of his captive Dytisci had been laid under
unnatural conditions. The female, when at liberty to
obey her own instincts, lays her eggs in rushes or
other aquatic plants beneath the surface of the water.
She makes an incision in the stem, which reaches the

1 This would agree better with Acilius than with Dytiscus.
Acilius lets the eggs drop upon the mud while swimming about.
Colymbetes arranges its eggs upon leaves, £c., to which they
adhere strongly (Regimbart).

I AQUATIC BEETLES 41

pith, and then lays a single egg in the cleft. The
ovipositor, which projects from the hinder end of the
body, is armed with two small, hard, pointed plates,
which enclose the oviduct, and it is with these that
the incision is made.1 March or April is usually the
time. From eggs so laid minute larvae proceed, hatch-

«

ing out in about three weeks. They grow very fast

A

FIG. 3.— A. Female Dytiscus, laying eggs. B. The ovipositor, protruded. C. Eggs
of Notonecta, attached to stem of rush. D. Egg of Dytiscus, laid in excavation
in rush. From Regimbart.

when well supplied with food, and attain their full
larval development in four or five weeks, changing

1 Regimbart, Ann. Soc. Ent. France, 1875. The eggs are
very elongate. Dr. Sharp informs me that they are often
accompanied by a white maggot, exactly like them, but rather
shorter. In Yorkshire the eggs are more commonly laid in the
midribs of Pond-weed (Potamogeton natans).

42 NATURAL HISTORY OF AQUATIC INSECTS CH.

their skins from time to time, like other Insects. The
larva when full-grown is more than two inches long.
It is yellow-brown in colour, and of cylindrical shape,

tapering towards the head
and tail.

The head is nearly circu-
lar, flat, and joined to the
thorax by a distinct neck.
The mouth-parts are those
of a carnivorous Insect. The
mandibles in particular are
slender, curved, and pointed.
On the inner or concave side
of each is a longitudinal
slit, which was long sup-
posed to possess a very
special importance as being
the only passage leading into
the mouth.

Swammerdam says of the
Dytiscus larva, and his ac-
count is, I believe, exactly
true : " When it feeds, it
seizes its prey with the
mandibles, and pierces it

FIG. 4.-Larva ^Dytiscus, dorsal ^^ ^^ g^arp, Curved

points. Now the tips of the

mandibles are hollow, and through them the larva
sucks the blood in a singular fashion into the mouth,
into which the channels of the mandibles open. Since
the larva and its jaws are in some degree transparent,
we can see the blood flow from the mouth into the

AQUATIC BEETLES

43

stomach, especially if the victim has red blood. . . .
When the mandible is examined by the microscope,
we see that it runs out into a sharp point and is some-
what curved. In cross section it appears sharp-edged,
flattened on both sides, and convex along its outer
border. The opening by which fluids are drawn in is

FIG. 5. — Head of larva of Dytiscus, from beneath.

near the tip, and forms an elongate slit, fringed by fine
hairs." l

De Geer tells us that on examining the mandible
by the microscope he found the slit described by
Swammerdam. " But," he continues, " is there no
other mouth ? I believe that there is, and that this
mouth lies between the lips (labrum and labium). As
proof of this I have seen a Dytiscus larva not only
suck a Crustacean,2 but little by little devour all its

1 Bibha Nat urcz, Vol. I. p. 325.

2 Cloporte ; the word is elsewhere applied to an Asellus.

44 NATURAL HISTORY OF AQUATIC INSECTS CH.

solid parts, which assuredly could not pass through
the fine openings of the mandibles." l

Other examples of perforated mandibles were also
described. Reaumur,2 in his account of the Ant-lion
(Myrmeleon), affirmed that this predatory larva had
no true mouth, and that the juices of its victims
could be absorbed only through the hollow mandibles.
When he comes to the lace-winged flies (Hemerobidae)3
he says of them that they have mandibles like those
of the Ant-lion. This statement has been confirmed
by Ratzeburg, who compares the mandible of the
Hemerobius-larva to the poison-fang of a serpent,
and says that he saw fluids pass along the internal
cavity.4 The larva of Gyrinus furnishes us with
another instance.

Siebold 5 at length combined these statements in
the following sentence : " The mouth of the larvae
of Myrmeleontidae, Hemerobidae, and Dytiscidse is of
very singular construction. There is no oral opening,
and the mandibles and maxillae are altogether unfit
for mastication, the mandibles being converted into
two curved hooks, which are hollow and provided
with a narrow fissure at their tips. The larvae bury
these hooks in the Insects which they seize, and
through the cavities of the organs, which communicate
at the base with the oesophagus, suck the blood."
Siebold's authority gave fresh currency to this state-

1 Me moires, Tom. IV. p. 386.

2 Tom. VI. p. 340 (1742). 3 Tom. VI. p. 382 (1742).

4 Ratzeburg, Forstinsekten., Bd. Ill, p. 243 (1844). He does
not say that in this larva the mouth is closed.

5 Lehrb. d. vergl. Anat.

I AQUATIC BEETLES 45

ment, which has been repeated many times in text-
books and popular treatises both in old times and
very recently.

Meinert l a few years ago was led to investigate all
these larvae afresh. In Myrmeleon and Hemerobius
he found that the mandibles are not truly perforated,
but only grooved along the inner edge. The groove
is converted into a tube by the maxilla. Sections
prove that there is a narrow cleft in the usual position
of the mouth. In the Dytiscus-larva too a vertical
section from back to front reveals the existence of a
mouth-passage, and Meinert showed that pressure on
the head not only causes the contents of the alimen-
tary canal to flow out drop by drop from the orifices
in the mandibles, but also causes an exudation from
the true mouth. He passed a fine hair into the slit
at the tip of the mandible, and found that it came
out at the base, where Westwood had long ago figured
a second orifice. In the Dytiscus larva the mandible
is traversed by a complete tube, into which the maxilla
does not enter at all.

Meinert's description and explanation were fiercely
disputed by Schiodte, but confirmed by Dewitz and
Redtenbacher. Burgess 2 adds some curious details
of structure. He confirms the existence of a narrow
passage between the upper and under lips, but finds
that this is ordinarily closed by a " mouth-lock ); -a
grooved joint such as is used for dust-proof doors.

1 " Om Mundens Bygning hos Larverne af Myrmeleontiderne,
Hemerobierne og Dytiscerne." Vid. Medd. Nat. Foren., 1879,
p. 69.

2 Proc. Boston Soc. Nat. Hist., Vol. XXI. p. 223 (1881).

46 NATURAL HISTORY OF AQUATIC INSECTS CH.

When the mandible is extended, its basal opening is
outside the mouth, but when it is closed the basal
opening is brought into the corner of the mouth, and
then the larva sucks up the blood of its victim as a

FIG. 6. — Structure of mouth of Dytiscus larva, i. The mandibles, one extended,
the other closed. 2. Section of mouth-lock ; ;;*, mouth. 3. Section through
mouth and pharynx ; Ibm, labium ; nil, mouth-lock ; ;«, mouth ; ph, pharynx ;
fin, put, muscles. From Burgess.

man might suck up water through a straw. It is
easy, he adds, to open the mouth with a pair of for-
ceps. The fitting must be extremely good. Mr. W.
F. Baker tells me that when a Dytiscus larva pre-
served with alcohol is placed in strongly heated

I AQUATIC BEETLES 47

paraffin, the vapour of the alcohol escapes by the tips
of the mandibles and by every other opening, as trains
of minute bubbles, but none appear at the mouth.

I have lately verified by dissection the chief points
of Burgess's description. While doing so it occurred
to me to try whether the mouth-lock acted auto-
matically, opening when the mandibles opened, and
closing when they close. Accordingly two larvae
were taken, one having the mandibles closed, while
in the other they were forcibly extended by a piece
of pith. Longitudinal median sections of each were
cut and mounted, when it appeared that in the larva
with closed mandibles the mouth was completely
closed, while in the larva with open mandibles a
passage about '05 mm. deep in the narrowest part
was immediately apparent. There has not been time
to repeat the trial, but I believe that reliance may
be placed upon this test case.

It seems that the mouth-apparatus of the Dytiscus-
larva is employed in the following way. When a
living victim has been seized by the mandibles, they
close upon it and pierce it, the base of the mandible
with the hinder orifice of the canal being at the same
time brought into the corner of the mouth. The
mouth-lock is by the same action closed, and then the
pharyngeal pump can be employed to suck the blood.
But if the larva should require to swallow solid
morsels, as it occasionally, though rarely, seems to
do, the separation of the mandibles would immediately
relax the mouth-lock and give them passage into the
stomach.

Almost all kinds of aquatic animals, Snails, Worms,

48 NATURAL HISTORY OF AQUATIC INSECTS CH.

Insects, Tadpoles, and Fishes are devoured by the
insatiable Dytiscus larvae. Many a raw naturalist has
put them into his collecting-bottle or aquarium, to
find after a few hours that they have destroyed or
mutilated almost his whole live stock.

The legs are long, and fringed with hairs, so that
they form efficient oars. When the larva swims about
in a leisurely way, the legs are the chief means of
propulsion. But it can also make a sudden spring,
by throwing its body into serpentine curves. It may
also be seen to creep on submerged leaves, and to
cling to them when resting or lying in ambush. The
legs end in double claws, which are probably useful in
seizing the prey as well as in supporting the body.
Some other predatory Coleopterous larvae (Carabidae
and Gyrinidae) have them too, but they are not
common in larvae of this order.

The last segment but one is much narrowed, and
the last still more so. The tip of the tail carries two
small appendages. These, as well as the two last
segments of the abdomen, are fringed with hairs,
which no doubt increase the effect of a stroke given
to the water. But these appendages are chiefly used
to buoy up the tail, when the larva requires to
breathe. The larva is lighter than water, and can
only remain below by grasping solid objects. When
it lets go, it slowly ascends to the surface, and pushes
the tip of its tail into the air. The water rolls off the
hairy appendages in a moment, and then the larva is
suspended by its tail from the surface-film. The
success of the operation depends upon the circum-
stance that the particles of water hold more power-

AQUATIC BEETLES

49

fully to one another than to the chitinous hairs.
Chitin, the hard substance of which Insect integuments
are formed, is not, as a rule, easily wetted by water.
A vigorous effort is required when the larva seeks to
draw its tail under water again. While it remains
suspended, air can be taken in by two spiracles which
occupy the extreme tip of the tail, and lead to the
great longitudinal air-tubes. There are seven other
pairs of spiracles on the sides of the body, but these
are rudimentary and closed. It is only at the time of
pupation that they become functional.

At length the larva ceases to feed, creeps into
moist earth near the edge of the
water, makes a roundish cell there,
and changes to a pupa. It is then
soft, yellowish in colour, and much
shorter and broader than before. In
summer the pupal stage lasts only
a fortnight or so. In cold weather
advanced larvae bury themselves in
the banks, probably pupating only
when spring is at hand. There is
however some difficulty in getting
full information respecting the pro-
cess of pupation in the winter or
early spring. The full-grown Beetles
live through more than one winter,
and except in very cold weather
they swim about and feed. In this
Insect, as in most others, the parts
of the imago are all present in the pupa, and are at
least externally perfect, though still shrouded in a

E

FIG. 7. — Pupa of Dy-

tiscus, dorsal view.

From Schiodte.

50 NATURAL HISTORY OF AQUATIC INSECTS CH.

temporary pupa-skin. The head is bent forwards
upon the thorax, while the legs and wings are dis-
posed symmetrically on each side of it. There are
spines on the prothorax of the pupa, resembling
those of the Hydrophilus pupa, but far smaller.

At length the full-formed Beetle creeps out from
its dark cell. Though its ultimate form is now
attained, the new skin is at first soft and pale-
coloured. A week passes before it completely
assumes the dark chestnut colour, the polished
surface, and the firm texture which it is destined to

FIG. 8. — Dytiscus marginalis (male), showing suckers on fore legs and smooth

elytra.

acquire. The outline of the body as seen from above
is regularly oval. The back is gently rounded, the
under side angulated along the middle line. The

I AQUATIC BEETLES 51

head is deeply sunk into the thorax, and does not
break the outline. A body so shaped is obviously
well fitted for rapid movement through the water,

FIG. 9. — Dytiscus marginalis, female.

as well as for easy entrance into narrow spaces.
Dytiscus marginalis takes its name from the yellow
stripe, which runs along the border of the thorax, the
outer edge of the wing-cover, and the front of the
head. Many other species of the genus Dytiscus are
ornamented in a somewhat similar way.

The mouth-parts are adapted to a predatory life,
for the full-grown Beetle is not less voracious than
the larva. The mandibles are no longer slender
and perforated, but stout and double-toothed. The
antennae are rather long and slender, and the
compound eyes large.

E 2

52 NATURAL HISTORY OF AQUATIC INSECTS CH.

On the under side of the first ring of the thorax the
keel-like prominence is formed by a backward-
directed spine, which fits into a cavity in the second
ring. This spine contributes to the rigidity of the
body, like the deeply sunk head, the peculiar fixation
of the thighs of the hind limbs, and the fusion of the
first and second rings of the abdomen (see Fig. 11,
p. 60.)

The wing-covers are very stout, smooth in the male,
and often furrowed along their length in the female.
Beneath them lie crumpled up the ample wings,
which can upon occasion easily bear the body through
the air. When food grows scarce, the Dytiscus takes
to flight, choosing twilight or dark for its journey, and
seeks out a fresh pond.

The hind legs are long and specially adapted for
swimming. The tibia and five-jointed tarsus are
flattened, and furnished with a fringe of long and
stiff hairs along one edge. These spread out during
the stroke, and are depressed during the return of
the leg through the water. The tarsus is so articu-
lated that it rotates a little upon its own axis when
moved through the water, presenting its broad face
during the stroke and its edge during the return ; " in
other words, what rowers call feathering the oar is
performed by the tarsus of the Dytiscidae in a most
perfect manner."1 In swimming the hind legs are
chiefly used for propulsion. They strike the water
together. The middle legs aid the movement, and
appear also to guide the animal's course. There is no

1 Dr. Sharp, on " Aquatic Carnivorous Coleoptera or Dytis-
cidae," Trans. Roy. Dublin Society, Ser. II. Vol. II. p. 254.

I AQUATIC BEETLES 53

sole to the hind foot, and the usual pair of stout clawb
become rudimentary.1 The fore and middle legs are
shorter and stouter, the fringe of hairs less conspicu-
ous, and the terminal claws usually well developed.

In Dytiscus marginalis, as in most other species
of the family, we find a curious modification of the
tarsus of the fore legs of the male, which forms a
circular sucker. A similar structure is visible, though
much less prominent, in the middle leg of the male.
The restriction of these suckers to the male Insect
at once suggests that they are accessory reproduc-
tive organs, to be distinguished from the many cases
of suckers or like organs merely used in creeping
on inclined or vertical smooth surfaces. These last
occur in Insects of many kinds, in both sexes, and on
all the six legs. The clasping suckers, on the other
hand, are restricted to the male, are borne only on the
fore, or fore and middle legs, and are peculiar to the
Coleoptera.

In the fore leg of the male Dytiscus the three first
joints of the tarsus are enlarged to form a circular
disc ; beyond these are two small and ordinary joints,
of which the last bears the usual pair of claws. The
same three joints are somewhat enlarged in the
middle leg also, and can be used as suckers. They
are not however formed into a disc, nor does their
peculiar structure catch the eye at the first glance.
When the circular disc of the fore leg is carefully
examined, we find that its under surface is roughened
by a number of minute stalked bodies, which are

1 In some other Dytiscidas, e.g. in Cybister, the swimming
legs are much more powerful than in Dytiscus marginalis.

54 NATURAL HISTORY OF AQUATIC INSECTS CH.

outgrowths of the chitinous integument (cupules, or
tenent hairs). The great majority of these are small
and close-set, but on the under side of the first tarsal

FIG. 10. — Fore leg of male Dytiscus, under side, with sucker, formed of three
enlarged tarsal joints. A small cupule, highly magnified.

joint they are replaced by two much larger structures
of the same nature (see Fig. 10). The whole sucker,
consisting of the three enlarged joints, is fringed

i AQUATIC BEETLES 55

by stiff pointed hairs. The two large cupules are of
unequal size, and carried on stalks which are deeply
sunk into the integument. Each ends in an umbrella
stiffened by many radiating ribs, which project
beyond the edge. The umbrella is elastic, and can
be pressed out flat without injury to its shape. The
smaller and more numerous cupules exhibit the same
structure, but the stalk is longer in proportion, and
the umbrella much smaller ; the ribs minute, and
enclosed in the web. The enlarged joints of the
middle leg of the male are also furnished with many
small cupules. Simmermacher l says that the male
of Dytiscus marginalis has 170 sucking hairs on each
fore leg, and 1590 on each middle leg.

The adhesive power of the suckers is very great.
Plateau2 found by actual experiment that a Dytiscus
marginalis, fresh killed by ether, when its suckers
were applied to a cylinder of smooth glass, could
support more than thirteen times its own weight.

Mr. Lowne 3 has pointed out several facts respect-
ing the suckers of Dytiscus which have escaped the
notice of other observers. The cavity of the tarsus,
he says, contains a large sac, which is filled with a
gelatinous fluid. The bases of the cupules open into
this sac, and pressure upon their free ends causes the
fluid to exude. The stalks of the cupules are hollow,
and serve as channels for the passage of the fluid.

1 Zcits.f. wiss. Zool. Bd. XI, p. 493. This memoir contains
an excellent account of the suckers of Insects in general, and
especially of Dytiscus.

2 Ann. Entom. Soc. Be/g., Vol. XV. p. 205 (1871-2).

3 Month. Micr.Journ., Vol. V. p. 267 (1871).

56 NATURAL HISTORY OF AQUATIC INSECTS CH.

When one of the suckers of the fore or mid leg is
pressed upon a clean slip of glass, a rather copious
deposit of liquid is left behind. It is coagulable and
insoluble in water ; it solidifies under water, and is
not wetted by water. To this account I may add
that it becomes opalescent on exposure to air. Mr.
Lowne points out that the adhesion of the suckers
suffices to support the weight of the Beetle in vacuo,
a proof that atmospheric pressure is not the cause of
the adhesion of the sucker. Nor can it be the surface-
tension of a thin film of water, for the suckers act
perfectly well when applied to a glass plate under
water. Since they act in water, in air, and in a
vacuum, we can hardly doubt that they act by means
of the tenacity of the coagulable secretion. I have
tried to exhaust the supply of the secretion by attach-
ing and detaching the suckers many times in succes-
sion. After thirty such repetitions, which occupied
about half an hour, the adhesive power of the suckers
was greatly diminished ; when wetted with water,
they recovered some of their efficiency, but were
much weaker than at first ; after thirty more trials
they adhered very feebly indeed.

The cupules discharge the viscid secretion when
pressed against a smooth surface, and it becomes a
question of some interest how the secretion is dis-
charged at the right moment. The act must be
mechanical, not depending upon any nervous stimulus,
for the fluid is discharged as freely in a freshly killed
Dytiscus as in a living one. It is to be observed
that each of the smaller cupules consists of a cylin-
drical shaft, perforated by a duct, and surmounted by

i AQUATIC BEETLES 57

a nearly hemispherical umbrella, in the centre of
which the duct opens. We can roughly imitate its
form by taking half of a hollow india-rubber ball and
applying to the centre of its convex surface the end
of a glass tube. Pass the tip of a pen-knife through
the india-rubber to represent the orifice of the duct.
If this rough model is placed on the table, with the
edge of the hemisphere downwards, the glass tube
may be pressed against the centre of the convex
surface until the hemisphere is deeply sunk in the
middle. It will then be found that the slit made by
the pen-knife gapes, and I believe that this con-
trivance is employed to discharge the contents of the
gland, though direct observation of the action of the
natural organ is hardly practicable.

Mr. Lowne finds that the cupules are occasionally
torn off by the cementing action of the secretion. On
examining with a lens four females taken at hazard,
I found that one of them had a number of cupules
firmly adhering to the prothorax.1

The suckers are no doubt chiefly employed to
retain the female, as their restriction to the male sex
indicates. They are, however, occasionally used to
hold the prey, though this is generally grasped by the
middle legs, and torn to pieces by the fore legs. Mr.
W. F. Baker tells me that he has seen a male Dytiscus

1 In the Cypris stage of all Cirripedes the antennule bears an
organ which has something in common with the cupules of the
male Dytiscus. It takes the form of a cup, and adheres by
means of a coagulable secretion, which attaches the animal to
its future support. In Cirripedes, however, the cup is a modi-
fied joint, whereas the cupules of Dytiscus are modified hairs.

58 NATURAL HISTORY OF AQUATIC INSECTS CH.
hold a Gvrinus with the suckers of the fore tarsi

*

until it was devoured. It may be doubted, however,
whether the peculiar suctorial structure is useful in
this connection ; it is certainly not necessary, for the
females have no suckers at all. \Yhen grasping the
female, the fore suckers are applied to the smooth
prothorax, and the middle suckers to the unfurrowed
margin of the elvtra.

o J

The males of many land Beetles (Carabidae, Cicin-
delidae, Silphidae, and Meloidae) possess suckers, as
well as male Dytiscidae and Hydrophilidae, which are
aquatic.

\Ye have next to notice the time-honoured ex-
planation of a contrivance supposed to be accessory
to the action of the suckers of the male Dytiscus.
Kirby and Spence 1 tell us that " the elytra of the
females of many of the larger water beetles (Dytiscus)
are deeply furrowed, while those of the males are
quite smooth and level. . . . Another aquatic Beetle,
Acilius sulcatus, Leach, has not only its elytra
sulcated, but the furrows of these and a transverse
one of the thorax are thickly set with hair ; while
the male is smooth and quite naked. Particular care
seems to have been taken bv the Creator that when

*

all the above inhabitants of the water are paired, the
male should be able to fix himself so firmly, by means
of his remarkable anterior tarsi . . . and these as-
perities, etc., in the upper surface of his mate, as not
to be displaced."

Further inquiry has greatly shaken an intcrpreta-

1 Yol. III. p. 305 (1826). The authors mention the occasional
occurrence of smooth females.

I AQUATIC BEETLES 59

tion which was at its first promulgation likely enough.
Plateau l in particular has pointed out a number of facts
hard to be got over, which tell strongly against the
old view, (i) Furrows on the elytra diminish instead of
increasing the holding power of the suckers. Ground
glass and paper do not on trial act so well as polished
glass. (2) The suckers of the male are not applied
to the furrowed part of the elytra at all, but to
the prothorax and the smooth edges of the elytra.
(3) Females with smooth elytra occur now and then in
England, and the male can hold these as well as the*
common form. Sometimes the male has furrowed
elytra, and it is said that this is apt to be the case in
cold ponds, ill supplied with food. Camerano 2 says
that the sculpturing on the elytra of the female is
not constant, but becomes less marked in south-
ern countries, and Redtenbacher3 adds that around
Vienna the smooth females are as common as the
furrowed ones. Many new explanations have been
offered of the furrowed elytra, but none seem to cany
conviction. It is to be noted that the Carabidae
(carnivorous land-beetles allied to Dytiscidse) com-
monly have furrowed elytra in both sexes.

The bases of the hind legs are very broad, and
soldered to the under side of the body. This increases
the rigidity of the body, and also removes the articu-
lation of the hind legs from the margin to the centre
of the Insect. There is no doubt some mechanical
advantage in this, but I am unable to explain what

1 Ann. Entom. Soc. Belg., Vol. XV. p. 205 (1871-2).

1 Atti. R. Ace. Torino, Vol. XV. Quoted fron> Simmermachcr.

3 Fauna Attstriaca, Die Kdfer.

60 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. ii. — Under side of Dytis-
cus, showing the bases of the
middle and hind legs, c-,
coxa (basal joint) of middle
leg ; c%, coxa of hind leg,
very large and immovably
united to its fellow and to
the metasternum, st.

it is. If the bases of the hind legs were free, water

would be squeezed out between
them and the body with some
force. The progress of the
Insect through the water would
probably be impeded, especially
if the escaping water passed
out forwards.

When the Dytiscus wishes
to come up to breathe, it need
only cease to exert itself in any
way. Its body, which, like that
of the larva, is lighter than
water, then rises slowly and
steadily to the top. The tail

is tilted a little upwards, which shows that this end

of the body is more buoyant than

the rest. A quantity of air is

always lodged beneath the wing-
covers. This clings to the felted

hairs which cover the back of the

abdomen, and forms a large flat-
tened bubble. When the Beetle

reaches the surface, its tail-end

at once pushes through the air.

Then the wing- covers are slightly

raised, arid the air concealed by

them is renewed. The two last

spiracles, which rise above the

surface of the water when the

Beetle is engaged in breathing,

are unusually large. The smaller spiracles in front

FIG. 12. — Abdomen of Dy-
tiscus, seen from above
after removal of wing-
covers and one wing ; the
spiracles are seen on the
risrht border.

I AQUATIC BEETLES 61

of these draw air from the supply which is lodged
beneath the wing-covers, and this they can do at all
times even if the Beetle is submerged.

When a Dytiscus is captured, it often discharges
a milky fluid from the thorax, just behind the head.
The fluid smells like sulphuretted hydrogen ; whether
it has any peculiar taste or not, naturalists have not,
I believe, tried to ascertain. At the tail-end again
are two anal glands which can suddenly discharge
an unpleasant ammoniacal fluid of yellow colour,
and this, there can be little doubt, is a startling
proceeding, which may often cause a bird to drop
the Beetle which it has seized.1 The white fluid is
most abundant in fresh-caught specimens ; the yellow
fluid in captives.

The Dytiscidae are long-lived and have been kept
several years in captivity.

THE GREAT WATER-BEETLE (HvDRorniLus).

The Great Water-beetle, Hydrophilus piceus, has
attracted the attention of spme of the most eminent

1 Many Carabidas and some other Beetles, e.g. Timarcha,
eject a fluid, often of unpleasant odour, from the mouth, when
first captured. Brachinus discharges a pungent and explosive
liquid from the anus. Cantharis, Meloe, Gyrinus. etc., emit a
fluid from the leg-joints. Ilybius fuliginosus emits a strong-
smelling fluid. Ponds containing the Beetle can be distinguished
by the smell, even if the Insect is not visible. Gyrinus, besides
the fluid emitted from its joints, secretes a sticky white paste,
of pungent odour, from its tail. It is probably this which is
used to fasten the eggs to leaves.

62 NATURAL HISTORY OF AQUATIC INSECTS CH.

naturalists of past generations. Chief among them
is the celebrated Lyonnet, who gives an interesting
account of the larva and pupa. In this, as in other
cases where classical memoirs are cited, I think that
a short account of the author may be acceptable to
some readers.

The work on which Lyonnet's fame mainly rests
is his memoir on the larva -of the Goat Moth (Traitc
Anatomique de la Chenille qui rouge le bo is de Saide
1760). No anatomical study shows greater fidelity
and skill. The dissections of the head of the larva
are truly an extraordinary feat, and will never be
surpassed. Modern treatises on Comparative Ana-
tomy continue to reproduce some of these figures,
such as the general view of the viscera, the structure
of the leg, and the digestive tract. Nearly the whole
interest of the volume lies in the plates, for the text
is little more than a voluminous explanation of the
figures.

It is not without surprise that we find that Lyonnet
was an amateur, who had received no regular training
either in anatomy or engraving, and that he had many
pursuits besides the delineation of natural objects.
He was brought up for the Protestant ministry,
turned to the bar, and finally became cipher-secretary
and confidential translator to the United Provinces of
Holland. He is said to have been skilled in eight
languages. His first published work in Natural
History consisted of remarks and drawings contri-
buted to Lesser's Insect Theology (1/42). About
the same time, Trembley was prosecuting at the
Hague his studies on the freshwater Polyp, and

I AQUATIC BEETLES 63

Lyonnet gave him some friendly help in the work.
Those who care to turn to the preface of Trembley's
famous treatise (Memoires pour servir a Vhistoire dcs
Polypes d'eau douce, 1744) will see how warmly
Lyonnet's services are acknowledged. He made all
the drawings, and engraved eight of them himself,
while Trembley is careful to note that he was not
only a skilful draughtsman, but an acute and experi-
enced observer. When the work was begun, Lyonnet
had never seen the operation of engraving a plate.
Wandelaar, struck by the beauty of his drawings,
persuaded him to try what he could do wjth a burin.
His first essay was made upon the figure of a Dragon-
fly, next he engraved three Butterflies, and then,
without longer apprenticeship, he proceeded to
engrave the plates still required to complete the
memoir on Hydra.

Lyonnet tells us that the larva of the Goat Moth
was not quite his earliest attempt in Insect Anatomy.
He began with the Sheep Tick, but suspecting that
the subject would not be popular he made a fresh
choice for his first memoir. Enough interest was
excited by the Traite Anatomique to call for the
fulfilment of a promise made in the preface that the
description of the pupa and imago should follow.
But though Lyonnet continued for some time to fill
his portfolio with drawings and notes, he never pub-
lished again. Failing eyesight was one ground of his
retirement from work. What he had been able to
finish, together with a considerable mass of miscel-
laneous notes, illustrated by fifty-four plates, was
published long after his death in the Memoires du

64 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. 13. — Larva of Hydrophilus
piceus, dorsal view.

Museum (Vols. xvm — xx).
The following account of
Hydrophilus is taken from
that source.

4< A larva, three inches
long," says Lyonnet, " was
discovered in the middle
of August on the grass at
the foot of a tree. Its
body was soot-coloured,
flattened from above down-
wards, tapering behind and
ending in a blunt point.
The head was broad, black,
and polished. The tip of
the tail was provided with
two hooks of unknown use,1
which possibly enclose the
two filaments which the
pupa bears in the same
place. From the sides pro-
jected short and stout
spines, directed obliquely
backwards. The legs were
short and appeared unfit
either for running or for
swimming with rapidity.
The head, armed with

1 It is believed that these
tail appendages are accessory to
respiration, as in the Dytiscus-
larva.

I AQUATIC BEETLES 65

powerful mandibles, seemed to indicate that the
larva was carnivorous, but the slowness of its move-
ments was hardly to be reconciled with this notion.
I offered it Insects and plants, but it refused to touch
them. This made me conclude, taking into account
its uncommon size, that the time for its transforma-
tion had come. I therefore put it upon freshly turned
up soil, on which I scattered some grass. It made a
hole, which it lined with the grass, and remained
within it several days in a curved position, lying on
its back. When touched it moved actively, but soon
resumed its first attitude.

" On the 2nd September, by which time the larva
had greatly diminished in size, the skin split along
the back up to the head, and there came out a white
pupa, which had been wounded on the right side,
from which a brown fluid exuded. This wound, how
caused I do not know, led to the death of the animal,
which happened some weeks afterwards without
further change of form.

" Some naturalists would not hesitate to conclude
from these facts that the larva in question must be
terrestrial, since it was found upon the earth ; further,
that it must be a vegetable feeder, for it was creeping
among the grass and was too heavy to pursue other
animals. It is wise, however, to suspend our judg-
ment until the facts are perfectly ascertained. Both
these conjectures prove to be erroneous. The larva
is not terrestrial but aquatic ; it does not feed upon
grass but upon animals.

' In the beginning of July I had noticed in the
ditches a kind of cocoon which I did not recognise.

66 NATURAL HISTORY OF AQUATIC INSECTS CH.

It was whitish, of the size of the end of the finger,
nearly spherical but rather oval and flattened. The
surface, which looked like tow, was not quite smooth.

A

FIG. 14. — Cocoon of Hydrophilus. A shows the mast; B is opened to expose the

eggs. From Miger.

One of the two ends was flatter than the other, and
furnished with a raised rim. From the space within
this rim projected a sort of little tapering mast, about
as long as the cocoon.

" I opened several of these cocoons, and found in
each about a hundred eggs. They were white and
oblong, regularly arranged side by side with the points
upward, and though provided with a double covering
were so transparent that a day or two before they
were hatched one could see the animal within. The
head appeared bent upon the thorax as in a pupa.

" The larvae, when hatched out, remained one day
enclosed in the cocoon before escaping. Then they
made an oval aperture in the lower part of the flat-
tened end of the cocoon, and escaped through this into
the water.

" One remarkable circumstance was that before

I AQUATIC BEETLES 67

they had taken any food they swelled out to three or
four times the size of the egg from which they had
emerged. The young larva is mouse-coloured, and
the whole of the body transparent except on the
sides.

" I took about thirty larvae from the brood on the
8th of July, and fed them with very small water-
snails, which they devoured in the same way as the
full-grown larvae do. The larva seizes the snail with
its mandibles, then bends its body backwards, and
rests the snail upon the broad back, which serves as
a table. In this position, holding the snail in its
legs, the larva breaks up and devours it. When small
snails were not to be had, the larvae managed very
well with large ones cut into pieces, and also with
tadpoles. If food ran short, they devoured one

FIG. 15. — Larva of Hydrophilus piceus, from Lyonnet, showing characteristic bent

attitude.

another, though at other times they lived quietly
together, so that I have seen them share tadpoles
without disputes. They appeared even to like one
another's society.

F 2

68 NATURAL HISTORY OF AQUATIC INSECTS CH.

' They never remain long a£ the bottom of the
water. Air is necessary for them, and this they take
in by the tail, which they raise from time to time to
the surface of the water. If they are by any means
prevented from passing their tails into the air they
become greatly agitated." [As in the larva of
Dytiscus, the only functional spiracles are the last
pair, opening at the tail.]

" The larva," continues Lyonnet, " moults three
times after escaping from the cocoon. It is well

known that when Insects
are about to moult they
remain for some time in-
active and without feeding.

o

At this time there is form-
ing within the old head,
the neck, and the hard en-
velopes of the legs, a new
armour, destined to replace
that which is about to be
cast off. The new armour,
before it has been exposed
to the air, is often quite
soft. As soon as the body
is freed from the old skin,
it becomes considerably

dilated. The Insect is still obliged to remain two
or three days without eating or exerting itself
actively, because many of its muscles are not as yet
firmly attached. Just after a change of skin the
head and thorax of Hydrophilus are white, and
one can then sec distinctly on each side of the

FIG 16. — Head of Hydrophilus larva
from under side.

I AQUATIC BEETLES 69

head six small black points, which are the eyes.
The head, as it becomes hard, turns black, and it is
then difficult to see without a microscope that there
are any eyes at all."

Lyonnet reared Hydrophilus larvae in order to
observe the process of pupation and the form of the
pupa, but found unexpected difficulties in his way.
u One larva," he says, "left the water on the 1st
of July, and travelled about my room. I put it back
into the aquarium. Some hours later another came
out, which I put into a trough with earth and grass.
It lived only two days, and such was also the fate of
two others. On the 2Oth of July I supplied earth to
two more, freshly emerged from the water, one of
which was larger by a quarter of an inch than the
other. They burrowed into the earth, and remained
there seven or eight days, when the smaller one came
out, and burrowing again at least three inches deep,
made a cell in which it stayed for two or three days,
lying on its back. Either my too frequent attempts
to see it, or some other circumstance, caused it to
destroy its cell, and this Insect also perished. About
the same time another larva which had begun to
make a cell destroyed it, apparently for the same
reason, and formed a fresh hiding-place in the earth.
There it made a new cell with a place of escape
on one side. On the 24th of August I removed
the earth which concealed it, and found the Insect
changed into a white pupa, though it was unable to
free itself from the larval skin. I endeavoured to
help it. It was easy to remove the bits of old skin
which clung to its body, but I did not venture to

70 NATURAL HISTORY OF AQUATIC INSECTS CH.

free the legs for fear of breaking them off. The
head was still entirely enclosed in the larval integu-
ment, and this was so hard that I had great diffi-
culty in removing it. I succeeded, however, without
wounding the pupa, but the head, being too tightly
packed in the corresponding part of the larva, had
not assumed its natural form, and instead of being
bent upon the thorax was stretched out. Moreover,
the legs, some of the tips of which I had broken off
in trying to free them from the larval skin, had taken
neither the form nor the arrangement suitable to the
pupa. Hence this Insect also was unable to complete
its transformation.

" The difficulties which I had encountered in pro-
curing healthy pupae made me suspect that I had not
given them earth sufficiently damp, and that the skin
of the larva must be slightly moistened in order that
the parts of the pupa may free themselves without
injury. I therefore took a larva which had been
wandering for fifteen days here and there on the
ground without offering to enter it. I placed it in a
large leaden box filled with earth much damper than
before. The larva entered the earth, and some days
after changed to a well-formed white pupa. On each
side of the head (on the fore part of the prothorax)
were three brown strong hooks. Two others of the
same kind were found at the hinder end of' the body.
These hooks being solid can contain no part of the
perfect Insect. I can imagine a would-be philosopher
saying, ' Tell me, you who suppose that everything
has its purpose, and that nothing comes by chance,
what is the use of these hooks ? It seems hardly

I AQUATIC BEETLES 71

possible that they can have any use at all, seeing that
they are possessed for a few days only by a pupa
buried in the earth, and that they are left behind when
the beetle emerges.' Such reasoning only carries

FIG. 17.— Pupa of Hydrophilus piceus, ventral view. X 2

weight when it is employed by one who has a minute
and intimate knowledge of the structure of the animal,
and of all its external circumstances. Since neither
the questioner nor myself possesses this knowledge,

72 NATURAL HISTORY OF AQUATIC INSECTS CH.

we should do better to say to ourselves : * Every
time that our weak perceptions have enabled us to
read the purpose of Nature, we have found the marks
of a wisdom superior to our own. It is therefore rash
.and arrogant to pronounce useless structures whose
purpose we are unable to assign. Although we cannot
imagine of what use these hooks can be to a pupa
buried in the earth, we may be well assured that they
have a definite purpose.' We shall presently see that
these parts are so necessary that the pupa would run
some risk of perishing if it were deprived of them.

" This Insect, although aquatic, breathes air. In its
larval condition, as we saw, it brings its tail from time
to time to the surface of the water. In the pupal
state it can no longer move its limbs, and it is this
apparently which compels it to leave the water in
order to undergo its transformation. It creeps out
upon the shore of the pond or ditch in which it has
hitherto lived, and there, in some moist place, it digs
out a cell, whose walls it secures by pressure, and prob-
ably also by gluing them together with a tenacious
substance which is seen to exude from the hinder
part of the body when the Insect is annoyed. Within
its cell the Insect rests for a time. Its body swells up,
and at the same time shortens. The new parts form
beneath the larval skin, which at length splits, and the
pupa emerges. This it does with ease when the skin
is moist, but with great difficulty, as we have seen,
when it has become dry. It was the lack of moist
earth which explained why several of my larvae re-
fused to enter the ground ; why others, having entered
it, came out again, and why yet others perished

I AQUATIC BEETLES 73

because they were unable to free themselves from the
larval integument.

" In the damp earth which the pupa requires, the
hooks described above fulfil a purpose unexpected by
us, but at the same time of great importance. The
skin of the pupa is very delicate. Lying on damp
earth it could hardly escape injury, and the weight of
the body might easily give it a distorted shape. But
the pupa protects itself from these dangers by assum-
ing an unusual attitude. It extends itself back
downwards in a horizontal position, and supports the
weight of its body by the three sets of hooks as upon
a tripod. In this attitude, though surrounded on all
sides by moist earth, it keeps its body from actual
contact with any object until it has assumed its final
shape.

" Thus we see how necessary are those hooks,
which at first sight appeared so useless. To decide
that this or that structure is superfluous because we
cannot guess its use is truly ridiculous in beings
whose information is so limited as ours."

Lyonnet's observations on the attitude of the pupa
and the use of the hooks are confirmed by the later
observations of Miger. The pupa of Hydrobius sup-
ports itself upon the floor of its cell in a similar way,
though here the spines cover the whole of the back.

" The pupa," continues Lyonnet, " is provided with
spiracles along the sides of the body, and these lead
us to suppose that it changes its mode of respiration,
and breathes in the pupal stage, not by the tail, but
by the sides of the body, as in many other Insects.

" When the time for the last transformation arrives,

74 NATURAL HISTORY OF AQUATIC INSECTS CH.

the eyes assume their black colour. Then the tips
of the mandibles, the claws of the foot, and lastly the
head and thorax take a brown colour. The legs
darken, the pupal skin is torn open, and a dark-
coloured Beetle emerges, destined to live henceforth
in the water."

Miger adds that the perfect Insect remains several
days underground before attempting to escape. At
length it forces a passage through the earth, being
aided by the flexibility of its wing-covers, and the
compressibility of its segments, which have not as yet
acquired any considerable degree of firmness.

The adult Hydrophilus may attain a length of over
if inches, and a breadth of about half as much
(45 x 20 mm.). With the exception of the Stag-
beetle it is the largest British beetle. The elytra or
wing- covers are, like the rest of the upper surface,
of a dark olive green. They are marked with faint
longitudinal sunk lines alternating with rows of dots.
The male Beetle is distinguished by the last joint of
the tarsus, which is dilated into a triangular plate,
and is believed to serve as a clasper. The species is
not uncommon in stagnant water near London, and
in the southern counties.

Hydrophilus often leaves the water by night, and
flies abroad. In winter it hibernates, burying itself
in the banks of the ponds which it inhabits. It is by
no means so powerful a swimmer as a Dytiscus. The
legs are less expanded, and are worked alternately
instead of simultaneously, so that the motion is
wavering and unsteady. The thorax and abdomen
are bevelled off to an angle on the middle line of the

I AQUATIC BEETLES 75

under surface, and here is found a stout and long
spine projecting backwards from the last thoracic ring,
which has been known to wound the hand when the
Beetle is carelessly held.

The mode of respiration of the adult Hydrophilus
is highly peculiar, and altogether different from that
of Dytiscus. It was first adequately described by
Nitzsch,1 whose account I find to agree in all essentials
with the facts which I have myself observed.

In adult Beetles there are usually two pairs of
thoracic spiracles (mesothoracic and metathoracic),
and from six to eight pairs more are borne upon the
anterior segments of the abdomen. Hydrophilus and
Dytiscus have each seven pairs of abdominal spiracles.
In Hydrophilus the foremost of these is large, and
those behind small, while in Dytiscus the last of the
series, near the hinder end of the body, is much larger
than those in front. When the Insect rises to breathe,
the hinder end of its body is inclined upwards to-
wards the surface of the water. In this position the
hindmost pair of spiracles give the readiest access
to the great air-tubes which run along the body, and
it is these which are enlarged. The large spiracles
of Hydrophilus however are not found at the tail end
of the body, but at the fore end (thoracic and first
abdominal), an arrangement which is convenient for
the taking in of air by the fore part of the tracheal
system.

Both in Dytiscus and Hydrophilus a large part of
the surface of the body is adapted to receive and

1 Archiv fur die Physiologie, von Reil und Autenrieth. Bd.
X. pp. 440-458, pi. IX. (1811).

76 NATURAL HISTORY OF AQUATIC INSECTS CH.

retain a pellicle or flattish bubble of air. Close-set
hairs are the simple means employed to prevent the
wetting of these particular tracts. Wrap a strip of
velvet round a stick, and dip it into water, or sprinkle
a few drops of water on a scrap of velvet. You will
see with what difficulty the water penetrates the
narrow spaces between the threads which form the
pile of the velvet. Close, outstanding hairs play the
same part in many an aquatic Insect.

The air-space beneath the wing-covers of Dytiscus
is described in connection with that Beetle. Hydro-
philus has a similar reservoir of air upon its back, but
it has in addition a peculiar and extensive reservoir
on its ventral surface.

On the under side of a Hydrophilus we can readily
see the hairy tracts which retain the film of air.
They cover the thorax and fore part of the abdomen
on each side. A prominent ridge along the middle
line, \vhich ends in the spine mentioned above, is free
of hairs ; so arc the greater part of the abdomen and
the thighs. The rest of the ventral surface is over-
spread in the living and submerged Beetle with a film
of air, which glistens like silver. The superficial air-
tracts extend also along the sides of the thorax and
abdomen, being bounded above by the overhanging
edges of the prothorax and wing-covers, and join the
reservoir of air on the back. The air-tracts appear
as transverse silvery streaks behind the head and
across the base of the wing-covers.

The spiracles open into these air-spaces, and the
submerged Hydrophilus replenishes from this source
the air within its body by a gentle rhythmical move-

AQUATIC BEETLES

77

ment of its abdominal segments, drawing in pure air
and breathing out the air vitiated by the oxidation of
the tissues. After a few minutes, more or less accord -

FIG. 18. — Hydrophilus piceus, male. Antenna (enlarged), and side view of head

with antenna in place.

ing to the activity of the Insect, the imprisoned air
within and without the body has become sensibly
impure, and then a fresh supply must be obtained

78 NATURAL HISTORY OF AQUATIC INSECTS CH.

from the atmosphere. The Hydrophilus creeps or
swims to the surface, and without difficulty assumes
a nearly horizontal position. Then the body is in-
clined a little to one side, so as to bring the angle
between the head and the prothorax on one side of
the body to the surface. A funnel-shaped depression
instantly appears at this angle. If the Beetle is
alarmed or suspicious, it sinks after a momentary
gulp of air, but if perfectly tranquil it may remain
for several minutes at the surface, drinking in air
energetically. The funnel-shaped depression remains
unchanged throughout the whole operation ; the
abdominal segments appear to move in and out, and
are alternately flattened and arched. There is also a
corresponding depression or elevation of the wing-
covers. During these various movements, which all
co-operate to enlarge or contract the air-spaces en-
closed within the body or lying beneath the wing
covers, the air-film on the ventral side may be seen to
dilate and contract alternately, and the body of the
Insect rises and falls a little. The channel of com-
munication between the spiracles and the funnel of
admission is bounded only by water on the under
side, yet the free passage of air is never checked.

We have next to consider how the funnel-shaped
depression by which the air passes in and out is
formed and maintained. The mere rising of the
Insect to the surface is not enough, nor is it enough
to bring one part of its air-bubble, where it is enclosed
by a mere film of water, to the surface. A film of
small extent might preserve its continuity for some
time, and so exclude the air, especially if the water is

3 AQUATIC BEETLES 79

in any degree impure, as must often be the case with
the water of a ditch or pond. The convenience, and
at times the safety, of the Hydrophilus require that
the moment it rises to the surface the film should
burst, and place the enclosed air in free communica-
tion with the atmosphere. Hydrophilus possesses in
its antenna an instrument for breaking the film cer-
tainly and without delay.

This antenna consists of nine joints. The basal
joint is large and curved, the next four joints smaller
and forming a flexible stalk. All these are readily
wetted by water, and serve, so far as is known,
merely as a handle to direct the movements of the
terminal joints. The terminal joints are four in
number. Each is dilated, except just where it is
attached to the joint beneath, and curved into a
semicircular plate. The enlarged joints form what
looks like the clubbed end of the antenna, which is
not really by any means so solid as it appears at first
sight. They are fringed on their concave sides with
long bristles, and are uniformly covered with fine
short hairs, which cannot be wetted by water.

The ordinary position of this antenna in the sub-
merged Beetle is horizontal, and extended in a gentle
curve. Springing just in front of the compound
eye, it passes backwards to the overhanging edge of
the prothorax. The four terminal joints lie in the
air-space ; the smooth basal joints are immersed and
wetted by the surrounding water.

When the Hydrophilus comes up to breathe and
directs upwards the very adjustable cleft between its
head and prothorax, the antenna of the same side

8o NATURAL HISTORY OF AQUATIC INSECTS CH.

is bent outwards. At the same time the enlarged
terminal joints hang downwards and form a well--
marked angle with the rest of the antenna. The
film which bounds the imprisoned air is thus drawn
upwards and outwards in the cleft, following the
movement of the antenna to which it clings. As
soon as any part of the unwetted terminal joints
comes to the surface, the film breaks, and a passage
is instantly opened from the air above to the air-space
beneath. The air-passage is enclosed between the
hairy side of the body and the vertically arranged
antennal joints, neither of which can be touched by
water. The size of the opening is determined by the
relative position of the antenna and the side of the
body ; the shape of the opening mainly by the tension
of the surface film and the varying pressure of the
water. It may give a better idea of the action of the
antenna to compare it to an arm, bent at the elbow
and with the hand hanging down, held at a little
distance from the side of the body, to ward off, not
a blow, but the water which would otherwise enter
the wider part at least of the cleft between the head
and prothorax.

A Hydrophilus may take in air so energetically
that the film enclosing the ventral air-space becomes
over-distended. Then bubbles escape at the base
of the third pair of legs every time the wing-covers
descend. Sometimes the Beetle becomes much
inconvenienced by the large quantity of air which it
has attached to its body, and has to drag itself down-
wards by clutching the stems of aquatic plants. I
once observed a Beetle thus rendered over-buoyant,

I AQUATIC BEETLES 81

which descended slowly and with great effort, until
apparently by the action of the basal joints of its legs
it dislodged a good-sized air-bubble. After this it
swam about freely.

The antennae are used only for breathing by the
submerged Insect. Their place as feelers is sup-
plied by the long and forward-directed maxillary
palps.

The food of the adult Hydrophilus is largely
vegetable, but it will prey upon small aquatic animals.
As some recent authors have said that the adult
Beetle feeds exclusively, or almost exclusively, on
animal substances, I may say that four Hydrophili,
kept in my aquarium, cleared it completely of small
weeds, devouring Nitella especially in large quantities,
and that in the absence of all animals bigger than
Daphnia they remained healthy and active.

\Ye now resume Lyonnct's narrative : " I was
particularly anxious to ascertain how the females
constructed the floating cocoon which encloses their
eggs. For this purpose I placed a few of these
Insects with some Duckweed in a large wooden
trough on the last day of May. On the 1st of June
I saw that one of the females, contrary to its usual
custom, was incessantly swimming about, and search-
ing on all sides. I expected that this was because
she did not find materials suitable for her work, and
as I had often seen a filamentous alga attached to the
cocoon, it occurred to me to give her some of this.
I floated it on the water by means of wooden
shavings, and next day, the 3rd of June, I found the
beginning of a cocoon, but the Insect soon abandoned

G

82 NATURAL HISTORY OF AQUATIC INSECTS CH.

her work, apparently because she had been disturbed
by several other kinds of aquatic Insects, which
lurked in the weed. I took them out of the trough,
and before long had the pleasure of seeing the female
Hydrophilus betaking herself to work under my eyes.
I found to my surprise that, like a Spider, she had her
spinneret at the hinder end of the body. By extend-
ing the hinder rings ever so little, and opening the
last of all, a nearly circular cavity appeared, in which
I could discover a whitish disc. From this two small
brown prominences were given off side by side.
Each enclosed a delicate conical tube, about a line in
length, and of dark-brown colour, stiff towards its
base, flexible and elastic towards the tip. These two
tubes formed the spinneret, and each contributed a
separate thread. They moved both together, and
exactly parallel to one another.

" The construction of the floating cocoon was
effected in the following way. At first, lying upside
clown near the surface of the water, the Beetle
buried the hinder part of her body and the two
hindermost pairs of legs in the alga, leaving the first
pair free, and making use of them to fit and mould
the alga to the end of her body. She then began
to weave the under side of her cocoon. While the
weaving was going on, she was careful from time to
time to press and flatten the growing cocoon, mould-
ing it with her fore feet against her body, and so
giving it the form of a flattened arch. After the first
piece, which was intended to form the upper side of
the cocoon, was finished, the Beetle turned over and
wove another piece, exactly the reverse of the first, to

I AQUATIC BEETLES 83

form the under side of the cocoon. The two curved
surfaces then were woven together, and the body of
the cocoon was finished, the work having occupied
about an hour and a quarter. For about two hours
after this the Beetle remained still, her back being
uppermost. At first her body was buried in the
cocoon up to the thorax, but one could now see that
she was gradually withdrawing it. During these two
hours of apparent rest she laid her eggs, not at hazard,
but in regular order, side by side, the pointed ends
uppermost. When the work had advanced so far
that her body was completely withdrawn from the
cocoon, she began to spin about the open mouth of
the cocoon, so as to gradually narrow the opening.
Then by spinning up and down she closed the end
of the cocoon, giving it a truncated appearance. The
next thing was to spin a little mast, which gradually
rose above the surface of the water, the fore part of
the Beetle during this part of the work being
constantly immersed.

" I do not know the use of this little mast. Perhaps
it enables the Insect to get rid of an excess of silky
matter. However this may be, the work was com-
pleted in about five hours, after which the cocoon was
left floating.

"On the 1 5th of July I observed an opening at the
foot of the mast, and saw, floating around, whitish
skins, either egg-shells, or the envelopes of the larvae.
Next day I saw a small larva emerge, and the day
after fifty more. From this point the history of the
Beetle has already been traced."

More than fifty years after Lyonnet's observations

G 2

84 NATURAL HISTORY OF AQUATIC INSECTS CH.

on Hydrophilus, Miger l described the life-history of
the large Hydrophilus.

Miger remarks that he was curious to find out how
an Insect living under water could construct a cocoon
floating on the surface. Early in May 1807 he found
several specimens, and placed them in a vessel of
water, together with the aquatic plants on which they
chiefly feed. At length a female was seen to under-
take the formation of a cocoon. Miger saw her
attach herself to the under side of a leaf which was
floating on the water. She placed her body across
the leaf, clasping it with her fore legs. The abdomen
was applied to the under surface of the leaf, and the
two tubes which form the spinneret could be seen to
be pushed in and out rapidly, while a white gummy
liquid was passed from them. This liquid was drawn
out into threads, which were attached to the leaf, and
gradually surrounded the abdomen, forming a semi-
circular pouch, in which the tip of the abdomen was
contained. After about ten minutes the Hydrophilus
turned sharply round, letting go the leaf and bringing
her head downwards, but without withdrawing the
abdomen from the cocoon. The leaf was now held
by the hind legs only, one being placed on each side
of the cocoon. The work went on steadily for nearly
an hour and a half, and through the transparent wall
of the cocoonthe movements of the spinneret could be
distinctly seen, until the gradual 'addition of threads
made it at last so thick and compact that the move-
ments of the spinneret could no longer be distinguished.

1 Miger3 s memoir is to be found in Ann. du Museum d?Hist.
Nat., Tom. XIV. (1809).

I AQUATIC BEETLES 85

Small bubbles of air were then seen to issue from
the cocoon. This was due to the displacement of the
contained air by eggs, which were laid side by side,
and glued together with a white liquid. In three
quarters of an hour the egg-laying was over. The
Beetle closed the cocoon slightly, and began to form
the pointed end. The wing-covers were now a little
opened, and their tips brought to the surface of the
water. The spinneret was uncovered, and its con-
tinuous and rapid movements could easily be followed.
The formation of the projection took more than half
an hour, at the end of which time the slender tapering
point rose about an inch above the water. The whole
operation took about three hours.

Miger observes that at the beginning of the work
the Hydrophilus is easily disturbed, but when egg-
laying has once begun she may be removed from the
water, and her cocoon cut open with scissors without
creating a stoppage.

He placed three females in a vessel of water with-
out any floating object. They were unable to form
a cocoon at all, though there passed from their
bodies a yellowish firm substance which contained
no eggs.

The air which fills the cocoon is derived from the
supply carried down by the Beetle. The cocoon is
always fixed to a floating object. " It is a mistake,"
says Miger, " to suppose that the turned-up point of
the cocoon serves as a mast. Such a notion could
only be drawn from an empty cocoon. It is not
unlikely that the drawn-out point serves for the
supply of air to the cocoon."

86 NATURAL HISTORY OF AQUATIC INSECTS CH.

Miger observes that three kinds of secretions are
made use of by the Hydrophilus. There is first a
sticky liquid, which when drawn out into threads
makes them cling to one another, and forms a com-
pact though flexible envelope. Another fluid serves
to glue the eggs together, while the third forms a
silky shining tissue, like that of the cocoons of Lepi-
doptera. This tissue is well suited for the admission
of air, but when submerged it allows the passage of
water also. The eggs arc regularly cemented together
side by side in a vertical position. The whole mass
of eggs is attached to the upper surface of the cocoon,
and has beneath it an air-space, which leads to the
opening by which the larvae emerge. This air-space
is only loosely closed by threads which are, however,
sufficient to prevent the entrance of the water. Since
the relatively heavy eggs are placed high in the
cocoon, and an air-space lies beneath, the cocoon
is liable to upset unless it is attached to a buoyant
body.

As soon as the larvae are hatched, they swell out to
twice their previous size, free themselves from their
integuments, and escape into the lower part of the
cocoon, where for about twelve hours they move about,
one over another.

Some additional particulars respecting the cocoon
of Hydrophilus arc furnished by Mr. A. G. Laker.1
He tells us that the average size is iij lines x lof
lines, and the height to the tip of the spike about 17
lines. The cocoons are buoyant, and when floating
freely about one third of their depth is out of water.
1 Entomologist, Vol. XIV. p. 82 (i 88 1).

I AQUATIC BEETLES 87

They are usually but not always attached to long
grass or floating leaves. " The spike consists of a
substance somewhat thicker and stronger than the
rest of the cocoon, and is hollow throughout the
greater part of its length, except that it is crossed
and recrossed inside with a dark thread-like substance,
thus somewhat resembling a horn stuffed with tow.
The apex of the spike does not, however, appear to
terminate in an orifice, but is closed. It does not
seem to me that this spike can serve as a balance to
the cocoon, because the nests are usually attached to
some kind of support. I may, however, mention that
I cut off the spike from two of the cocoons, and in
both these cases the eggs did not hatch ; it is, however,
possible that this may have arisen from some other
cause, although these particular cocoons appeared to
be similar in every respect to others of which the
eggs matured in due course. The cocoons from
which the spikes were removed subsequently sank.
These nests are so constructed that when floating
loose the spike retains its proper position, and even
if the cocoon be held so that the spike is parallel with
the water, and then suddenly released, it immediately
rights itself; if, however, the spike be partially sub-
merged and then released, the cocoon turns bottom
upwards. The number of eggs contained in each nest
is usually between fifty and sixty."

HYDROBIUS.

In shallow, grassy ponds there may commonly be
found a little blackish Beetle, about a quarter of an

88 NATURAL HISTORY OF AQUATIC INSECTS CH.

inch in length. The body has the oval convex form
and shiny surface of a Hydrophilus. The ventral
surface is overspread by a bubble of air. The antennae
are clubbed, and used in respiration as in Hydrophilus.
On close examination, however, it will be found that
the sharp spine, which projects backwards from the
under side of the last thoracic segment of the great
Hydrophilus, has disappeared in the smaller Beetle,
which is the Hydrobius fuscipes of entomologists.

In spring and early summer the full-grown Beetles
are abundant. They become scarce in July, and
about a month later the new generation, distinguished
for a time by the much lighter colour, makes its
appearance. In cold weather they bury themselves
in the mud. On watching a living Hydrobius, we
shall see it bring the cleft between the head and
thorax to the surface of the water when it requires to
breathe. Here a small funnel-shaped depression forms,
and leads air to the ventral surface. The antenna lies
in the cleft, and is apparently employed just as in
Hydrophilus, though the small size of the Beetle
renders it difficult to affirm this with absolute cer-
tainty. Living as it does in very shallow water,
Hydrobius finds it easy to expose the under side of
its abdomen to the air. When this is done, the
watery film bursts, and respiration goes on freely.
Whether by accident or of set purpose, the Beetle
often executes this movement. Even when floating
on deep water, it can turn on its back and throw
off the water from its abdomen.

The following note, together with a good deal of
other information respecting Hydrobius, I owe to

AQUATIC BEETLES

89

Mr. W. F. Baker : " Hydrobius is a vegetarian, and
may be found in any weed-grown pond, being gener-
ally accompanied by another Beetle of like habits,
Helophorus aquaticus. One or both of these is
pretty sure to be found at the right season in any
shallow pond covered with a carpet of Duckweed.
In such a pond there will be few or none of the Dytis-

FIG. 19.— Hydrobius fuscipes. X 8.

cidae (Acilius, Agabus, Dytiscus, &c.), which prey
upon small Beetles, for the mode of respiration of the
Dytiscidae requires open spaces, where they may
come to the surface and breathe without obstruction.
Any circumstance which discourages these carnivorous
Beetles favours the increase of Hydrobius and Helo-
phorus."

90 NATURAL HISTORY OF AQUATIC INSECTS CH.

In April the female Hydrobius lays her eggs in
white cocoons, which are to be found in great numbers
at or close to the surface of the water. They are often
slightly attached to blades of grass. The cocoon
is about as large as the Beetle which constructs it,
and is formed of silken threads woven together in
various degrees of closeness. A female may some-
times be found with the unfinished cocoon attached
to her body. One end of the cocoon (that first formed)
is firm and smooth ; the other end, which the Hydro-
philus shapes into a mast, is left loose and irregular
by Hydrobius.1 From each cocoon several larvae
issue. They are of minute size, but active and in-
tensely predatory. The large head, armed with
formidable jaws, which are incessantly opened and
closed, reveals at a glance the instincts of the creature.
Though constantly immersed in water, the larva, like
the pupa and the imago, requires a constant supply
of gaseous air. This it draws in by the tail. On the
dorsal surface of the tail a number of valves and flaps
(see fig. 20) form a shallow cup fringed with longer and
shorter setae, and into this no water can enter. When,
as is usually the case, the larva lies in very shallow
water, the tail-end is quite dry, and air is freely drawn
into the long tracheal tubes, which open under cover
of the fore edge of the cup. If accidentally pushed
into deep water, the larva hangs from the surface by
its tail, breathing all the time, or creeps along the

1 The eggs of Hydrobius hatch out readily when completely
submerged and wetted with water. The use of the cocoon in
this case appears to be rather to conceal and protect the egg
than to secure buoyancy.

AQUATIC BEETLES

bottom with a large bubble attached to its tail, which
clings to it with great tenacity and raises it above the
ground. While hanging from the surface the larva

FIG. 20. — Larva of Hydrobius fuscipes (magnified) ; head of do., under side; tail-
end, showing respiratory cup and tracheae.

*

executes a great variety of movements, bending its
body upwards, downwards, or sideways in search of
its prey. If kept in deep water for a long time it
drowns,

92 NATURAL HISTORY OF AQUATIC INSECTS CH.

Not a few aquatic Beetle larvae breathe after the
manner of Hydrobius, though the details of the
mechanism exhibit great variety. The larva of the
Dipterous fly, Stratiomys, employs the same principle
in a somewhat different way. But the closest parallel
is furnished by three other Dipterous larvae, those of

FIG. lii. — Pupa of Hydrobius fuscipes (magnified). The wings, enclosed by the
pupal wing-sheaths, are partly extended.

Dixa, Anopheles, and Pericoma, which are furnished
with a respiratory cup much like that of Hydrobius,
though more elaborate and more efficient The
older Hydrobius larvae are altogether buoyant, and
when out of their depth swim at the surface in a
serpentine manner.

£ AQUATIC BEETLES 93

By the end of June the larvae begin to pupate, and
during July the pupae may be found in great abund-
ance. The full-fed larva quits the water, and makes
a globular hole about half an inch deep in the mud or
clay of the bank, six inches or so above the water-
level. In this hole the pupa lies upon its back, kept
from contact with the moist earth by the long stiff
spines which stand out from its body. The abdomen
is flexible, and can be moved rapidly about when
the Insect is disturbed. Some larvae were found still
unchanged six days after leaving the water. Early
in August the fully developed Beetle begins to
show itself in numbers. At first whitish, its colour
darkens on exposure to light and air. It throngs the
shallow waters until cold weather drives it into the
mud, where it hibernates till spring.

Very soon after the emergence of the Beetle the
cocoon is formed and the eggs are laid. Hence all
the stages may be found together about the end of
August — eggs, young larvae, pupae, and adult Beetles.

DONACIA.

The " leaf-eating Beetles " (Chrysomelidae) include
a number of species which pass their early stages
upon submerged plants, and feed upon the roots.
The white water-lily (Nymphaea), Potamogeton, the
Arrowhead, the Sedges, the Marsh Marigold, the Bul-
rush, the Horse-tail, and other moisture-loving plants
yield shelter to the various species of Donacia and its
close ally Haemonia. D. crassipes is often found
abundantly upon Nymphaea or Sparganium. The

94 NATURAL HISTORY OF AQUATIC INSECTS CH.

female bites small round or oval holes in the leaves,
and through these apparently passes the eggs to the
under side, where she arranges them in a single or
double circle around each hole. The larvae, when
hatched, descend to the bottom, and begin to feed on
the roots. They exhibit no obvious adaptations to
aquatic life — no swimming organs, no gills, no peculiar
shape — but only the dirty-white colour, the semi-
cylindrical figure, the small, hard head, and the three
pairs of pointed legs, found in an ordinary larva which
buries itself in earth. When full grown, they are
about half an inch long. Their movements are very
sluggish.

Close examination with a magnifying glass reveals
just one peculiar structure. Towards the hinder end
of the body, on the 8th segment of the abdomen, there
project from the dorsal surface two slender curved
spines, and to the bases of these pass the longitudinal
air-tubes, which traverse the whole length of the body.
There are also two pairs of more flexible setae, which
fringe the last segments. At the roots of the spines
are a pair of small openings, which look like spira-
cles.

Roots of Nymphaea frequented by Donacia were
observed by Schmidt-Schwedt 1 to exhibit peculiar
scars. These were discovered with difficulty, owing
to the dark colour and uneven surface of the roots.
There was in each case a rough hole, made apparently
by the jaws of the larva when feeding, and, at a dis-
tance corresponding with the length of the body, a pair

1 Berl Entom. Zeits., Bd. XXXI. pp. 3^-334, Taf. V. figs,
i-ii (1887).

I AQUATIC BEETLES 95

of small slits. On microscopic examination these slits
were found to penetrate the epidermis of the roots, and
to lead to the small irregular air-spaces, which occupy
a considerable part of the interior of the roots. Some-
thing of this had been previously observed by Siebold,
who in 1859 described the larva as biting a hole in
the roots of Sparganium, passing the end of the
abdomen into it, pressing the spiracles by the help of
the curved spines close against the hole, and so draw-
ing the contained air into its body. Schmidt-Schwedt
believes that the pair of openings are made not by
the mouth but by the spines, and that the air is
drawn in by internal channels running along them.1

Aquatic Insects often find it a matter of some
difficulty to procure a sufficient supply of air, and
many ingenious contrivances appear as practical so-
lutions of the problem. None perhaps are quite so
remarkable as the present one. That the Donacia-
larva should have found out the air-reservoirs of sub-
merged roots, and possess special organs for tapping
them, is surely one of the curiosities of adaptation.

The pupal cocoon is a close-woven, oval capsule,
attached to the same roots as those upon which the
larva feed. On the attached side the wall of the cocoon
is deficient, and a good-sized hole, previously closed
by the root itself, appears when the cocoon is torn
away. A number of small holes, penetrating into the

1 Dewitz (Berl. Entom. Zeits., Bd. XXXII. p. 5, 1888) be-
lieves that in Haemonia, and presumably in Donacia also, the
spiracles serve for the admission of air to the body, as Siebold
maintained. Schmidt-Schwedt has (B. E. Z., 1889) reaffirmed
his original statement.

96 NATURAL HISTORY OF AQUATIC INSECTS CH. I

substance of the root, appear upon the plant when the
cocoon is detached, and it is'probably from this source
that the pupa derives its supply of air. Wounds in
the living tissues of the plant are as a rule quickly
repaired by a corky growth, but Schmidt-Schwedt
found that this was not the case with the holes bored
by the Donacia-larva in the roots of the water-lily.
They remain open so long as the access of water is
precluded by the cocoon, and only become closed by
cork when the cocoon is torn open by the emergence
of the Beetle. The under side of the adult Insect is
protected from \vetting by a close covering of silky
hairs.

The perfect Beetle appears in early summer ; the
larvae feed until autumn, and then pupate. The per-
fect Insect is believed to remain throughout the winter
in its pupal cocoon, for cocoons opened at this season
generally contain fully developed Insects and not
pupae.

CHAPTER II

FLIES WITH AQUATIC

THE GNAT (CULEX)

SMALL stagnant pools and ditches are the favourite
haunts of the larvae and pupae of the Gnat. A ditch
in a wood choked with fallen leaves is one of the best
hunting grounds, and in the summer months they
may be found by the thousand in such places.

The larva, when at rest, floats at the surface of the
water. Its head, which is provided with vibratile
organs suitable for sweeping minute particles into the
mouth, is directed downwards, and, when examined
by a lens in a good light, appears to be bordered
below by a gleaming band. There are no thoracic
limbs. The hind limbs, which are long and hooked
in the burrowing Chironomus larva, and reduced to
a hook-bearing sucker in Simulium, now disappear
altogether. A new and peculiar organ is developed
from the eighth segment of the abdomen. This is a
cylindrical respiratory siphon, traversed by two large
air-tubes, which are continued along the entire length
of the body, and supply every part with air. The

H

98 NATURAL HISTORY OF AQUATIC INSECTS CH.

larva ordinarily rests in
such a position that the tip
of the respiratory siphon
is flush with the surface of
the water, and, thus sus-
pended, it feeds inces-
santly, breathing uninter-
ruptedly at the same time.
It is of obvious advan-
tage to the larva that it
should in this way divide
the operations of breathing
and feeding between the
two ends of the body, for
the breathing must be done
at the surface, and the
feeding beneath the sur-
face. Observe that the
arrangement requires one
thing unusual in floating
objects — viz., that the body
shall be, in some parts at
least, if not altogether,
heavier than water. If it
were lighter than water

o

it could not maintain a
vertical position, but would
lie along the surface like

O

a stick, and would be quite

FIG. 22.- Larva of Gnat (Culex), in Unable tO SWCCp about in

side view.

search of the minute or-
ganisms which inhabit the water beneath. When

II FLIES WITH AQUATIC LARVAE 99

startled, the larva leaves the surface and sinks slowly
to the bottom by gravity alone, which shows that
the body is denser than the water. It does not
willingly remain below for any length of time, but
rises by a jerking movement, striking rapid blows
with its tail, and advancing tail foremost. When it
reaches the top, it hangs as before, head downwards,
and resumes its feeding operations.

How is it possible for a larva heavier than water to
remain floating at the surface without effort, as the
larva of the Gnat certainly appears to do ? The
possibility of such a thing turns upon the existence at
the surface of water or any other liquid of a contract-
tile surface-film. (See Introduction.) It is the con-
tractile force, or tension, of the film which rounds the
rain-drop and the soap-bubble. The film can be
stretched over unwetted threads, such as the threads
of a piece of dry muslin, or the fine wires of a piece
of varnished gauze, and thus supported it will not
only refuse to pass through fine meshes, but may
be made to hold up a rather considerable weight

of water.

If we take a solid body, capable of being wetted

by water, and place it in water, the surface-film will
adhere to the solid. If the solid is less dense than
the water, it will float with part of its surface out

•

of the water. Under such circumstances the surface-
film will be drawn upwards around the solid, and will
therefore pull the solid downwards. But if the solid
is denser than the water, its upper surface will be lower
than the water-level, and we can arrange matters so
that the surface-film around it will tend to pull the

H 2

TOO NATURAL HISTORY OF AQUATIC INSECTS CH.

solid upwards. Suppose that a solid of the same
density as water floats with part of its surface in
contact with air, and that weights are gradually added
to it. The result will be that the surface of the water
around the upper edge of the solid will become more
and more depressed. The sides of the depression
will take a more vertical position, until at last the
upward pull of the film becomes unable to with-
stand further increase of weight. If this point
is passed, the solid will sink. Before this point
is attained, we shall have the solid, though denser
than water, kept at the surface by the pull of the
surface-film.

This explanation will be followed more readily
with the help of a simple experiment

Take a float made out of a large cork, weighted by
a screw driven into its lower end. To the top of the
cork a small glass tube may be attached in a vertical
position. The lower end of the tube should be closed
and the upper end open. The weight must be
adjusted so that the top of the tube is only just out
of the water. Then add small weights one by one.
Small pieces of sheet metal can be placed by a
forceps on the cork. As the weight increases, the
tube slowly sinks until it becomes flush with the
water. Still no water flows in. We can go on
adding weights until the float depresses the surface,
as the reflections show, and yet, if the weight is not
excessive, the float does not sink. But if we take a
loop of thread wetted with water, and draw it across
the open mouth of the tube from side to side, the
surface-film will be drawn over the hole : then there

II FLIES WITH AQUATIC LARVAE 101

is nothing to keep the float at the surface, and it
settles steadily to the bottom.1

It is by virtue of this contractile force of the
surface-film that the Gnat-larva is enabled to main-
tain itself, against gravity, at the surface of the water.
The tip of the respiratory siphon is provided with
five flaps, which can be opened or closed by attached
muscles. When open they form a minute basin,
which, though its walls are cleft, does not allow the
surface-film of water to enter. \Yhen closed, the air
within the siphon is unable to escape. At the time
when the larva rises to the surface, the pointed tips of
the flaps meet the surface-film and adhere to it. The
attached muscles then separate the flaps, and in a
moment the basin is expanded and filled with air.
The surface-film is now pulling at the edges of the
basin, and the pull is more than sufficient to counter-
balance the greater density of the body of the larva,
which accordingly hangs from the surface without
effort. When the larva is alarmed and wishes to
descend, the valves close, their tips are brought to
a point, and the resisting pull of the surface-film is
reduced to an unimportant amount.

Swammerdam found it necessary, in explaining the
flotation of the larva of the Gnat, to suppose that the
extremity of its siphon was supplied with an oily
secretion which repelled the water. No oil-gland can
be discovered here or elsewhere in the body of the
larva, and indeed no oil-gland is necessary. The
peculiar properties of the surface-film explain all the

1 I owe this pretty little experiment to the ingenuity of Dr.
Stroud,

102 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. 23. — Larva of Gnat (Culex). 2, head and prothorax, ventral side. The com-
pound eyes, the folded legs, and other parts of the future fly are seen through
the transparent larval skin ; 3, antenna of do. ; 4, respiratory tube of do- ;
5, extremity of respiratory tube, end view

II

FLIES WITH AQUATIC

103

phenomena. The surface-film is unable to penetrate
the fine spaces between the flaps for precisely the
same reason that it is unable to pass through the
meshes in a piece of gauze.

After three or four moults the larva is ready for
pupation. By this time the organs of the future fly

FIG. 24. — Pupa of Gnat (Culex). Respiratory trumpet of do.

are almost completely formed, and the pupa assumes
a strange shape very unlike that of the larva.

At the head-end is a great rounded mass which
incloses the wings and legs of the fly, beside the
compound eyes, the mouth-parts, and other organs
of the head. Each appendage has its own sheath
(part of the proper pupal skin), and the appendages
are cemented together by some substance which is
dissolved or softened by alcohol. At the tail-end is

104 NATURAL HISTORY OF AQUATIC INSECTS CH.

a pair of flaps which form an efficient swimming-fan.
The body of the pupa, like that of the larva, is
abundantly supplied with air-tubes, and a communi-
cation with the outer air is still maintained, though
in an entirely different way. The air-tubes no
longer open towards the tail as in the larva, but to-
wards the head. Just behind the head of the future
fly is a pair of trumpets so placed that in a position
of rest the margins of the trumpets come flush with
the surface of the water. Floating in this position,
the pupa remains still so long as it is undisturbed,
but if attacked by any of the predatory animals which
abound in the fresh waters, it is able to descend by
the powerful swimming movements of its tail-fin.

Water is prevented from entering the trumpets
by the peculiar hairs which project in great numbers
from their inner surface. When the pupa first comes
up from below, the trumpets are overspread by a film
of water, which, becoming thinner and thinner by
evaporation, at length suddenly bursts.1

Why should the position of the respiratory organs
be changed from the tail-end in the larva to the head-
end in the pupa ? Not only the Gnat but Chi-
ronomus, Corethra and many other aquatic Insects
exhibit the same phenomenon. Evidently there must
be some reason why it is more convenient for the
larva to take in air by the tail, and for the pupa to
take in air by the head. We have already discussed
the considerations which seem to have fixed the res-

1 Useful information respecting the respiratory trumpets and
other organs of the Gnat pupa will be found in Dr. C, H. Hurst's
papers, (See later,, footnote tQ p, J44-)

II

FLIES WITH AQUATIC

105

piratory organs at the tail of the larva. Why, then,
need this arrangement be reversed when the Insect
enters the pupal stage ? There is now no feeding to
be done, and it surely does not signify how the head

FIG. 25. — Pupa of Gnat (Culex
nemorosus), showing the parts
of the fly enclosed in the
transparent pupa-skin. The
head and thorax are just freed
from the pupal skin.

FIG. 26. — Fly of Gnat (Culex nemorosus)
escaping from pupa-skin. This and the pre-
ceding figure are from specimens mounted
in Canada balsam.

is carried. Why should not the pupa continue to
breathe like the larva, by its tail, instead of developing
a new apparatus at the opposite end of its body, as
if for change's sake ? Well, it does not appear that,
so far as the pupa itself is concerned, any good

106 NATURAL HISTORY OF AQUATIC INSECTS CH.

reason can be given why the larval arrangement
should not continue. But a time comes when the
fly has to escape from the pupa-case. The skin splits
along the back of the thorax, and here the fly emerges,
extricating its legs, wings, head, and abdomen from
their close-fitting envelopes. The mouth-parts must
be drawn backwards out of their larval sheaths,
the legs upwards, and the abdomen forwards, so
that there is only one possible place of escape — viz.,
by the back of the thorax, where all these lines of
movement converge. If, then, the fly must escape by
the back of the thorax, the back of the thorax must
float uppermost. Otherwise the fly would emerge
into the water instead of into the air. Grantino- that

o

the back of the thorax must float uppermost in the
pupal stage, it is clear that here the respiratory tubes
must be set.

The mouth of the female Gnat is provided with a
case of instruments for piercing the skin and drawing
blood. The foremost of these is a tube split along its
hinder side, which lies in front of the rest, and is used
in suction. This, though long and slender, is stouter
than the delicate parts behind it, and it serves to
stiffen and protect them. Then come five long and
slender blades of great delicacy. Two pairs correspond
to the mandibles and maxillae of other Insects, though
here they are so simplified and attenuated that it
is not easy to make out the correspondence. The
maxillae are furnished near their tips with a row of
extremely minute saw-teeth. There is also a fifth
unpaired implement, which is an extraordinary
development of a part of the Insect's mouth which is

II

FLIES WITH AQUATIC

107

FIG. 27. — Fly of Gnat (Culex). male.

loS NATURAL HISTORY OF AQUATIC INSECTS CH.

usually quite inconspicuous. Besides these piercing
implements, the Gnat is provided with a soft flexible
sheath, which represents the labium. This takes the
shape of a tube, split along its fore side, which
surrounds and protects the delicate parts within.

>n

tr.6

ft

FIG. 28. — Head and mouth-parts of female Gnat, a, antenna (first three joints) ;
mxp, maxillary palp ; Ire, labrum-epipharynx ; /, labium ; m, mandible ; mx,
maxilla ; ft, hypopharynx. The separated mouth-parts are copied from Dimmock.

The extremity is divided into two lobes. AYhen the
female Gnat makes her attack upon the human skin,
she .plants the two divisions of the flexible labium
upon the spot selected. Then the piercing imple-
ments make their puncture, being guided and
supported by the labium, in the same way that the

II FLIES WITH AQUATIC LARVAE 109

fingers would be used to support a needle or other
slender instrument which we wished to pass into a firm
object, such as a piece of leather. As the piercing
instruments penetrate further and further, the labium
is bent backwards near the middle of its length, so as

o

to shorten it without withdrawing its support from
the piercing parts. When these preparations are
completed, the Gnat is in a position to draw the
blood into its mouth. One part of the oesophagus is
dilated, and forms a sucking-bulb ; in front of this is
a minute valve which apparently prevents the escape
of blood from the mouth. The momentary enlarge-
ment of the sucking-bulb reduces the pressure within,
and the blood flows up along the front split tube.
In the male Gnat these parts are a good deal simpli-
fied. The saw-teeth of the maxillae, the unpaired
lancet, and the sucking-bulb are all undeveloped,
and it is probable that the male Gnat, if it feeds
at all, obtains its food in a quite different manner
from the female. Some writers, however, maintain
that the male Gnat can bite. Why the female Gnat
should draw blood at all is a difficult question. It has
been surmised that a supply of highly nutritive fluid
is necessary for the formation of the numerous eggs,
but the force of this, as of every explanation which
has hitherto been suggested, is a good deal weakened
by the circumstance that Gnats abound in certain
regions where no quadrupeds are found, and very
few animals worth biting. They have been found in
immense numbers in uninhabited parts of Labrador,
on the tundras of Siberia and the desolate Kerguelen
Island. " Of the millions of mosquitoes which in the

iro NATURAL HISTORY OF AQUATIC INSECTS CH.

short Norwegian summer often thicken the air of the
Tromsdal — a wild valley within the Arctic Circle,
practically uninhabited either by man or probably by
beast — how few are ever likely to taste blood ! "l Sir
James Ross and his men found it necessary in their

A B

FIG. 29. — Antennae of Gnat. A, male; B, female.

polar expedition to wear gauze over their faces in the
summer months as a defence against Gnats.

o

The irritation caused bv the bite of the Gnat has

j

naturally led to the belief that poison is injected
into the wound. Whether this is true or not cannot
be positively stated. No poison gland has hitherto
been demonstrated, and there is some reason to
believe that the irritation of the wound is slight in
1 A. D. Michael, in Natural Science, Vol. I. p. 203.

II FLIES WITH AQUATIC LARVJE in

cold weather, and only becomes intense during great
summer heat.1

Nearly two hundred species of Gnats have been
described. Some of these inhabit hot countries, and
are popularly known as Mosquitoes, but to the
zoologist there is no species of insect which can be
defined as the Mosquito. Even in our own temperate
climate a burst of hot summer weather has led
persons who might be supposed to be pretty good
judges to give the name of Mosquito to our common
Gnat. "A few years ago a London Hotel, popular
with American visitors, was said to harbour Mosqui-
toes, which some of the visitors had brought with
them from the Southern States, but an examination
revealed the fact that the cistern was uncovered and
exposed, and was the breeding place for hosts of
Gnats." 2

The piping sound which the Gnat utters, and which
is often of itself enough to excite the greatest irritation
in its victims, is said to be due to the vibration of
the edges of its spiracles, occasioned by the passage

1 Some Gnats can subsist upon the juices of ripe fruits,
where they are unable to feed on the blood of animals.
Many two-winged flies of the section Nemocera have the
rostrum and mouth-parts remarkably prolonged, apparently
for sucking sap or nectar. Apetz has observed a Campylomyza
(Cecidomyidas) sucking a caterpillar, and both Culex and
Simulium are known to suck caterpillars. Culex and Anopheles
(Culicidas), some species of Ceratopogon (Chironomidae) and
Phlebotomus (Psychodidse), are various examples of blood-suck-
ing Nemocera. The substance of this note is taken from Baron
Osten Sacken's " Characters of the Three Divisions of Diptera,"
&c., Berl. Ent. Zeits. Bd. XXXVII. (1892).

- Enc. Brit.,\rv\. II. p. 866.

H2 NATURAL HISTORY OF AQUATIC INSECTS CH.

of a current of air, but this explanation has its
difficulties.

The female Gnat lays her eggs on the surface of
stagnant water, and the operation of egg-laying has
been carefully described by Reaumur. The Gnat
supports her body upon the four front legs, which
rest upon the margin of the pool or upon any floating
-object. The long hind legs are crossed, and in the
angle between them the eggs are received as they
pass one by one from the end of the abdomen. At
this time they are sticky and adhere to one another.
As the mass of eggs increases in size, the hind legs
alter their position and become more and more
parallel to one another. The operation of egg-laying
takes place in the early hours of summer days.

The egg-raft of the Gnat, though one of the
commonest objects in Nature, is apt to escape our
notice on account of its minute size, for it is less than

FIG. 30. — i, egg-raft of Gnat, from Reaumur; 2, a single egg; 3, egg removed
from ovary, with bladder-like appendage ; 4, end view of the appendage, showing
radiating lines.

a quarter of an inch in length. It was beautifully
described 150 years ago by Reaumur. The eggs of
the Gnat are cigar-shaped, and 250 or 300 of them

II FLIES WITH AQUATIC LARV^ 113

are glued together, so as to make a little concave
float, shaped like a shallow boat. The upper end of
each egg is pointed ; the lower end is provided with
a lid, through which the larva will ultimately issue
into the water. The Gnat in all stages, even while
still in the egg, requires an ample supply of air. It
is therefore necessary that the egg-raft should float at
the surface ; it is also necessary that it should always
float in the same position, so as to facilitate the escape
of the larva. This is effectually secured by a pro-
vision of almost amusing simplicity. Let us first
notice how efficient it is. If we take two or three of
these tiny egg-rafts, and place them in a jug of water,
we may pour the water into a basin again and again ;
the egg-rafts float instantly to the surface, and the
moment they come to the top they are seen to be as
dry as at first. The fact is that the surface-film
cannot penetrate the fine spaces between the pointed
ends of the eggs. The cavity of the egg-raft is thus
overspread by an air-bubble when accidentally sub-
merged. The eggs are kept from contact with water ;
the proper upper surface is made so buoyant that
the raft has great power of self-righting ; while, the
instant that it comes to the top, the excess of
water drains off, and the film bursts, leaving the eggs
perfectly dry on their upper surface.

CORETHRA

In still pools, especially such as are overhung by
trees, there may often be found in considerable abun-
dance the predatory larva of Corethra, also called the

I

ii4 NATURAL HISTORY OF AQUATIC INSECTS CH.

Phantom Larva from its remarkable and almost com-
plete transparency. This larva is about two-thirds of
an inch long. It often remains for a long time
together extended horizontally in the water at a
little distance beneath the surface, on the look-out
for its prey. Suddenly it disappears from view, and
when rediscovered is found to have moved an inch or
two, and to point in a different direction. The move-
ment is effected by a sudden lashing action of the
body, and in its promptitude resembles the jerk of an
electric telegraph needle.

By suitable tackle these larvae can be taken in any
required numbers, and it may be well shortly to
describe the instruments which are employed to
collect this and other aquatic forms of life. There
is nothing more generally useful than a hoop, five or
six inches in diameter, across which wire-gauze is
stretched. The hoop is attached to a socket which
receives the end of a long stick. It should be further
provided with a rim, upon which a muslin bag of no
great depth may be stretched. It is convenient to
sew the edge of the muslin bag round an india-rubber
ring, which can then be slipped over the metal rim.
The object of the gauze is to prevent sticks, leaves,
and other floating objects from entering the net.
Provided with such an apparatus as this, the collector
can sift the water in likely places until he supposes
that sufficient material is collected. Then the muslin
bag is slipped off from the hoop, and inverted into a
vessel of water. On account of its portability, the
vessel which I have found it most convenient to carry
is half of a large hollow india-rubber ball. For

II

FLIES WITH AQUATIC LARVAE

'/\;,

W

fetching up mud from the bottom of a deep ditch or

pool, the hoop and gauze

may be used without the

muslin. Another implement

which is worth carrying is

a large iron spoon with a

long handle.

The Corethra larva feeds
upon small aquatic animals,
such as Ephemera-larvae,
Daphnia, or Cypris. The
transparency of its body and
its immobility, except when
it instantaneously changes
its position, enable the larva
to wait for its victims with-
out risk of being observed.
Limbs would be here super-
fluous, and nothing can be
seen of them except a pro-
minence provided with a
double crown of minute
teeth, which can be distin-
guished by the microscope,
at the extreme end of the
tail. This represents, in a
greatly reduced form, the
pair of hooked feet found at

FIG. 31. — Larva of Lorethra. A,

the hinder end of the body dorsal view. B side view, x s.

* 1 he two pairs of air-sacs are seen

in Chironomus and Tanypus. LV-h!ifi£sti,an<J eighth sesments

J " behind the head.

A little farther in front, on

the lower surface of the body, is a vertical fin,

I 2

I/1

ii6 NATURAL HISTORY OF AQUATIC INSECTS CH.

composed of a row of feathered bristles. This no
doubt increases the sculling effect of the rapid
strokes of the tail, by which the larva executes its
sudden movements. Each feathered bristle is in-
serted by a forked base, and the whole row is con-
nected by the cuticle, which can be thrown into
folds like the valves of a concertina, or separated
at pleasure. The folds are independent in their
action, and some of them may be closed while others
remain open. This arrangement was first described
to me by Mr. Hammond. There is a similar vertical

FIG. 32. — Details of fin of larva of Corethra.

fin on the tail of the Gnat-larva : both have evidently
a common origin, and are used in the same way.

The head of the Corethra larva narrows forward.
When the larva first emerges from the egg it is
provided with a single pair of minute eyes, which are
little more than pigment-spots. Subsequently a
larger and many-facetted eye forms by the side of
this. Weismann believes that this eye is directly con-
verted into that of the fly. The antennae project in
front, and the mouth-parts lie beneath. The antennae
are not, as in most Insects, merely sensory organs,

II

FLIES WITH AQUATIC LARVAE

117

FIG. 33. — One element
of fin of larva of
Corethra.

but are prehensile, and play a great part in the
capture of the prey. Each consists of a simple basal
joint, which ends in four or five
long curved bristles. When at
rest, these bristles are directed
downwards and backwards, and are
flexed upon the basal joint. A
special muscle serves for the ex-
tension of the antenna, which is
brought back to its ordinary rest-
ing position by the elasticity of the
parts. The labrum is long and
narrow, and assists in the capture
of the prey. The mandibles re-
semble hands in shape, and bear
four or five long curved finger-like
teeth. The maxillae are mere stumps, and the labium
is reduced to a small triangular plate. Any small
animals captured by the antennae are crushed by the
mandibles and pressed into the mouth, but they are
not swallowed, as we might naturally expect. The
back of the mouth is closed by a circular fringe of
stout bristles arranged like a lobster-pot, or what
is in some parts of the country called a weel for
catching fish. The body of the victim is thus im-
prisoned and acted upon by the salivary secretion,
so that only the fluid products of digestion enter
the stomach. If the larva is gently pressed between
two glass slips, the pharyngeal tube can be everted,
and by a more natural process of eversion the in-
digestible parts of the food are voluntarily expelled
from the mouth of the living Insect. A somewhat

uS NATURAL HISTORY OF AQUATIC INSECTS CH.

similar weel may be found in the Gnat larva, and
also in the fish-like Amphioxus.

Respiration must be carried on chiefly through the
integument, for there are no spiracles. There is
however an imperfect tracheal system, consisting
mainly of two tubes which run the whole length of
the body. In the greater part of their length these
tubes contain no air, but in the thorax and again
towards the end of the abdomen, each of them dilates
and forms a good-sized air-sac (fig. 31). In the living
larva, as seen by the microscope, these air-sacs appear
black, owing to spots of pigment as well as to the
difference in refractive power between water and air.
The imprisoned air is more important from its
hydrostatic function than as a means of respiration.
The four air-sacs serve in fact as floats, and maintain
the larva without effort in a horizontal position near
the surface of the water.

In a fresh-hatched larva the tracheae and air-sacs
are altogether empty of air. How are the air-sacs
subsequently filled ? We have no definite informa-
tion on this subject. Gaseous air can be discharged
from solution in a watery fluid either by rise of
temperature or by diminution of pressure. Bring a
glass of water fresh from the pump into a warm room.
Bubbles of air shortly appear on its sides. Open a
soda-water bottle which was corked under .consider-
able pressure. The diminished pressure due to open-
ing occasions a copious discharge of gas. It is hardly
conceivable that the evolution of gas in the body of a
small aquatic Insect can be due to rise of temperature.
But there is no serious difficulty in supposing that it

ii FLIES WITH AQUATIC LARVAE 119

is caused by diminution of pressure. The sucker
implies such local diminution, and if a Simulium
larva, for instance, can employ negative pressure for
the purposes of attachment, it is conceivable that a
Corethra larva may be able to employ negative pres-
sure to extract a gas from its own blood for hydro-
static equilibrium. But until some mechanism able
to produce diminution of pressure has been found in
the air-passages of an Insect, the discussion is very
vague and unpractical. It will be remembered that
many other aquatic animals besides Insects can fill
some internal chamber of their bodies with a gas rich
in oxygen (see p. 37).

The pigment upon the air-sacs of the Corethra-larva
may conceivably help in the respiratory process.
Pigment is not unusual in the respiratory organs of
Insects. The large tracheal trunks of some Dragon-
fly larvae are tinged with purple, and the tracheal gills
of some Caddis-worms show a pale violet colour. Gas
may be secreted by a chemical process in which the
pigment plays its part.

The pupa floats in an upright position at the sur-
face of the water. At the head end is the great mass
composed of the head, wings and legs of the future
fly, all shrouded in the temporary larval skin. The
tail is broad, and like the tail-fin of a Lobster in shape.
It is the chief organ of locomotion possessed by the
pupa. When alarmed, the pupa descends a few inches
below the surface, by means of the powerful strokes
of its abdomen and tail.

When pupation takes place the tracheal system of
the larva becomes disorganised. Most of the air

120 NATURAL HISTORY OF AQUATIC INSECTS CH.

escapes to the exterior of the body, and lodges be-
neath the thorax, where it collects into a good-sized
bubble ; this is probably useful in maintaining the

B

FIG. 34.— Pupa of Corethra. A, ventral view ; B, side view.

upright position. The fore air-sac breaks up, but the
hinder one remains, apparently not in a functional
condition. A pair of respiratory trumpets, somewhat
similar to those of the Gnat pupa, are now the chief
means of communication with the outer air. They
are narrowed at the extremity, and open by a very
contracted slit.

The fly which issues from the pupa-skin is a small
Gnat-like Insect. The male has beautiful plumed
antennae. The female deposits on the surface of water

II

FLIES WITH AQUATIC LARWE

121

a flat, gelatinous sheet, in which the eggs are arranged
in spiral lines.

I will add a few words about another Gnat-like
Insect, with aquatic larva and pupa, which in its larval
state was described and figured long ago by De Geer,
and has since received the name of Mochlonyx. In
habitat, mode of feeding, &c., the earlier stages much
resemble those of Corethra. Both larva and pupa
are structurally as exactly intermediate between
Corethra and Culex as can well be imagined. Thus

\

FIG. 35. — Fly of Corethra plumicornis, male.

the tracheal trunks (air-tubes) are large in the Culex
larva, reduced in Mochlonyx, and rudimentary in
Corethra. The air-vesicles of the Culex larva are

122 NATURAL HISTORY OF AQUATIC INSECTS CH.

slight dilatations of the air-tubes. In Mochlonyx
they are greatly enlarged, and the connecting air-
tubes reduced in diameter, though filled with air. In
Corethra the vesicles are the only reservoirs of air,
the connecting tubes, though visible by the micro-
scope, being filled with a watery fluid. The Moch-
lonyx larva retains in a reduced form the breathing-
tube, which is so conspicuous in Culex, and which

FIG. 36. — Head of Fly of Corethra. The antennae (except their enlarged basal joints)

are cut away.

disappears in Corethra. These and other facts of the
same kind seem to indicate that the three larvae have
diverged from a common form, but what that common
form was I do not venture to conjecture.

CHIRONOMUS.

One of the commonest, and certainly one of the
most instructive, of aquatic Insects is the larva of
Chironomus. It abounds in ditches, water-butts, and
streams, especially dirty streams, and feeds upon
decaying vegetable matter. It may be readily recog-
nised by its crimson colour, which has suggested the

II

FLIES WITH AQUATIC

123

n •

4

$

/

popular name of Blood-worm.1 When of full size it
is nearly one inch long. These larvae protect them-
selves by making burrows out of
particles of earth or leaves, which
they weave together with the very
abundant secretion of the salivary
glands. When undisturbed, they
may often be seen to push the
head-end well out of the burrow
tor purposes of feeding ; at other
times the tail-end is pushed out
and waved to and fro in the water,
as a help to respiration. Now and
then they leave their burrows, and
swim through the water with a
lashing movement, twisting them-
selves into figures of eight. Occa-
sionally they rise to the surface, as
if to renew their supply of oxygen.
They are careless about finding
their way back to their burrows, for
in a short time they can glue to-
gether enough fresh fragments to
conceal themselves. The body of
the larva consists of a head and
twelve segments. The prothorax,
or segment next behind the head,
carries a pair of feet armed with
numerous hooks. These are used for grappling.

1 Under this name are grouped the larvae of several species—
C. plumosus, dorsalis, &c. The specific distinctions, whether
of the larva or of the fly, are in most cases very minute.

FIG. 37. — Larva of Chiro-
nomus. Near the head
are seen the rudiments
of the wings and legs of
the fly, enclosed within
the larval skin.

124 NATURAL HISTORY OF AQUATIC INSECTS CH.

Another pair of hook-bearing appendages is carried on
the last segment of the body, and is probably used for
holding on to the burrow. Caddis-worms too possess
a pair of hooked feet at the hinder end of the body,
and no doubt use them for the same purpose. The
last segment but one bears two pairs of long and
flexible tubes. These are filled with blood, which is
continually renewed by the pulsations of the heart,
and the exchange of gases necessary for respiration
is probably effected through the thin wall of the tube.
Four other smaller prominences on the last seg-
ment have a similar structure, and very likely are
used in the same way. The last segment also car-
ries on its upper surface two bunches of fine hairs
which are supplied by a pair of nerve-cells, and are
believed to be sensory in function.

When the larva is concealed within its burrow and
apparently at rest, an undulatory up-and-down move-
ment of the body is kept up, which continually renews
the water. At such times the four tubes are probably
actively engaged in the exchange of gases. The
same movement is practised by other case-dwelling
aquatic larvae, such as Caddis-worms and the cater-
pillar of Paraponyx.

The head, which is small in proportion to the body,
bears a short pair of jointed antennae, and two pairs
of eye-spots, which are mere pigmented patches,
without lenses. As usual, a sort of flap, the labrum,
hangs down in front of the mouth. This is furnished
with a very elaborate set of teeth, hooks and spines,
some of which are probably used to guide the fila-
ments of silk as they issue from the salivary glands.

II

FLIES WITH AQUATIC

125

There is a pair of strong, toothed mandibles, which
are the chief organs of mastication. The maxillae
are rudimentary, and the labium is represented by a
crescentic plate with strongly toothed fore-edge.

The transparency of the Chironomus larva is suffi-
cient to permit the study of its internal organs in a
living and uninjured specimen. Small larvae, say one-
quarter of an inch long, are best for the purpose. It
is a good plan to make a little cell of cotton wool or

c

FIG. 38. — Larva of Chironomus. A, head, dorsal side ; B, head, front view ; C, edge
of labium, with peculiar teeth and papillae.

silk upon a microscopic slide, to fill this with water,
place the larva in it by means of a fine brush, and
then gently lower the cover-glass. The imprisoned
larva can thus be studied microscopically. All the
external organs, the nervous system, the alimentary
canal, and the heart can easily be made out, and the
action of the jaws can be studied.

The alimentary canal begins with a long, narrow
oesophagus. This is telescoped behind into the crop
or proventriculus, which is of much greater diameter

126 NATURAL HISTORY OF AQUATIC INSECTS CH.

than the oesophagus, and shows three rows of promin-
ences on its outer surface. Behind the proventriculus
comes the stomach, which extends through about half
the length of the body. At its beginning the stomach
is nearly of the same diameter as the proventriculus,
but it narrows somewhat irregularly farther back.
The intestine succeeds to the stomach : close to its
beginning four long and slender tubes open into it.
These are the Malpighian tubes or excretory organs.
The vegetable food of the larva and the earthy
particles swallowed with it are usually to be seen as a
black mass, which does not fill the stomach, but is
restricted to a central channel of definite shape, a
good deal narrower than the stomach itself. By
dissection it can be made out that the stomach is
lined by a transparent chitinous tube, which protects
the delicate cellular wall from abrasion by rough
vegetable fragments or gritty particles. In the
Chironomus larva, as in other Insects, the oeso-
phagus and intestine are both furnished with such
a chitinous lining. The peculiarity of the Chironomus
larva consists in the prolongation of the cesophagal
lining as a free tube throughout the whole length of
the stomach.1 Digestion is probably carried on in
some such way as this. The vegetable food, already
masticated, passes slowly along the chitinous tube.
The fluid extracted from the food, containing most of
the nutritious matter, escapes at the free end of the
tube, and flows back into the outer compartment
of the stomach, external to the chitinous tube. Here

1 This peculiarity is found in some other Dipterous larvae, as
well as in certain Crustacea.

ii FLIES WITH AQUATIC LARV^ 127

it is absorbed, while the insoluble refuse is passed
directly into the intestine. A pair of salivary glands
is found in the fore part of the body. These are
hollow sacs lined by a single layer of cells. Each is
furnished with a duct. The two ducts unite in front,
and pour the salivary secretion into the mouth.
Whether the saliva takes any direct part in the diges-
tion of the food I do not know for certain, but it
undoubtedly furnishes the material for the silken
threads, with which the larva binds together the walls
of its burrow.

The student who is provided with a microscope
will find the cells which line the salivary glands an
excellent subject for study. The nuclei of these cells
are unusually large, and of complicated structure.
Each nucleus is a transparent spherical mass, enclos-
ing one or two smaller spheres, the nucleoli. Besides
the nucleoli, the nucleus contains a long cord, which
is wound into an irregular coil. This cord appears to
be made up of a number of discs placed end to end
like a pile of coins. If there are two nucleoli, each is
joined to one end of the cord : if there is only one
nucleolus, it receives both ends. The physiological
meaning of this intricate structure has not yet been
elucidated. It is an interesting but unexplained fact,
that nuclei of similar structure, though of very much
smaller size, occur in some plants, e.g. in the young
seeds of Fritillaria. It may be that the large size and
complex structure of the salivary nuclei of Chironomus
is connected with the very abundant secretion poured
forth by a comparatively small number of cells, but
this is mere conjecture.

128 NATURAL HISTORY OF AQUATIC INSECTS CH.

An inexperienced worker with the microscope, look-
ing for an interesting subject, which will neither tax
his knowledge nor his practical skill too heavily, will
find the salivary gland of Chironomus exactly suited
to his wants. For it is an object of considerable bio-
logical interest, and yet it can be studied without
mechanical appliances or special manipulative skill.
A few words of explanation will enable a mere novice
to make for himself most instructive preparations.

Lay a live larva on a clean glass slip, and cut off its
head with a scalpel. The transparent salivary glands
at once begin to flow out, together with the abundant
red fluid, which is the blood of the larva. By simply
laying a cover-glass on the glands, and examining
them with a quarter-inch objective, the nuclei can be
seen and studied. But the details of their structure
cannot be made out unless the tissue is stained. A
staining fluid applied to the yet living cells will
not act. (See p. 38.) It is therefore desirable to
kill the cells, and to do it rapidly without allowing
disorganisation of the nuclear structure to take place.
The reagents which accomplish this for us are called
fixing agents. Heat, such as that of boiling water,
will suffice in many cases. Momentary immersion in
strong alcohol is another means. The preparation
may be irrigated on the slide by a few drops of
absolute alcohol, which are afterwards to be drained off
and replaced by the staining fluid. Most microscopic
workers prefer osmic acid. A weak solution is em-
ployed, say \ per cent. Precautions, especially with
respect to the action of light, are requisite in the case
of osmic acid. It becomes reduced, and turns black,

II FLIES WITH AQUATIC LARV^ 129

and loses its action upon tissues if exposed to the light.
The bottle in use, which should be of small size, must
be coated with blackened paper, and kept in a wooden
box when not immediately required. A drop of
the solution fixes the cells and their nuclei very rapidly.
After a minute's exposure to the osmic acid, wash the
tissues in a few drops of dilute alcohol, and finally
with absolute alcohol. Then the stain may be applied.
Try methyl green, picro-carmine and haematoxylin
upon as many salivary glands, and learn the best
length of exposure. A few minutes will suffice.
Until practice has been got, it is well not to attempt
permanent preparations. Mount in dilute glycerine,
and do not omit to draw everything noteworthy. The
preparations will last for several days without further
precautions.1

In the living larva the heart may be seen busily at
work. It is a transparent tube, situated at the hinder
end of the body on the dorsal side. Two pairs of
valvular inlets admit the blood to the heart ; then the
contraction of its muscular walls drives the blood along
the narrow dorsal vessel, which passes above the
alimentary canal to the head. Here the regular blood-
channel ceases ; the blood escapes, and bathes all the
organs of the body, ultimately making its way back
again to the heart.

It is evidently a matter of some difficulty for the
Chironomus larva to procure a sufficient supply of air.
It lives in water, often in deep water, which abounds
in decaying organic matter, and can contain but little

1 The salivary cells and nuclei of the Chironomus larva have
been well described by Balbiani (Zool. Anzeigcr, iSSi).

K

130 NATURAL HISTORY OF AQUATIC INSECTS CH.

free oxygen. The burrows in which it lives furnish
an important defence against Fishes and other enemies,
but they still further increase the difficulty of securing
a sufficient supply of air. The larva can only freely
aerate its blood by leaving its burrow, and swimming
about near the surface. The thin-walled and trans-
parent appendages near the hinder end of the body
are probably of special service in taking up dissolved
oxygen. The tracheal system is rudimentary and
completely closed, and hence gaseous air cannot be
taken into the body. The dissolved oxygen, pro-
cured with much exertion and some risk, must be
stored up within the body of the larva, and used with
the greatest economy. It is apparently for this reason
that the larva of Chironomus contains a blood-red
pigment, which is identical with the haemoglobin of
vertebrate animals. The haemoglobin acts in the
Chironomus larva as it does in our own bodies, as an
oxygen carrier, readily taking up dissolved oxygen,
and parting with it gradually to the tissues of the
body.

It is instructive to notice that only such Chirono-
mus larvae as live at the bottom and burrow in the
mud possess the red haemoglobin. Those which live
at or near the surface have colourless blood, and a more
complete, though still closed, tracheal system. The
larva of the Gnat again, which has a large and open
tracheal system, and in all stages of growth inhales
gaseous air, has no haemoglobin at all. A list of the
many animals of all kinds which contain haemoglobin,
shows that for some reason or other each of them
requires to use oxygen economically. Either the'

II FLIES WITH AQUATIC LARV^ 131

skin is thick, and the respiratory surface limited, or
they are inclosed in a shell, or they burrow in earth
or mud.

We might expect to find that haemoglobin would
always be developed in the blood of animals whose
respiration is rendered difficult in any of these ways,
but any such expectation would prove to be un-
founded, and there are many animals whose mode of
life renders it necessary that oxygen should be stored
and economically used, which contain no haemoglobin
in their blood. Hence, while we have a tolerably
satisfactory reason for the occurrence of haemoglobin
in a number of animals whose respiratory surface is
limited, and whose surroundings make it a matter of
difficulty to procure a sufficient supply of oxygen, we
have to admit that many similar animals under the
same conditions manage perfectly well without hsemo-
globin. Such admission is not a logical refutation of
the explanation. I might fairly put forward the
baldness of mankind as at least the principal reason
for wearing wigs, and this explanation would not be
impaired by any number of cases of bald men who do
not wear wigs. The fact is that the respiratory needs,
even of closely allied animals, vary greatly, and
further, there are more ways than one of acquiring
and storing up oxygen in their bodies.

Either the storage-capacity for oxygen of the
Chironomus larva is considerable, or it must be
used very carefully, for the animal can subsist long
without a fresh supply. I took a flask of distilled
water, boiled it for three-quarters of an hour, closed
it tight with an india-rubber bung, and left it to cool

K 2

132 NATURAL HISTORY OF AQUATIC INSECTS CH.

Then six larvae were introduced, the small space
above the water being at the same time filled up with
carbonic acid. The bung was replaced, and the
larvae were watched from clay to day. Four of the
larvae survived for forty-eight hours, and one till the
fifth day. Two of them changed to pupae. Never-
theless, the water was from the first exhausted of
oxygen or nearly so.

The special provisions which enable Chironomus
larvae, or at least many of them, to inhabit waters
poorly supplied with oxygen, probably explain their
occurrence at great depths. They have been fished
up, together with larvae of Tanypus, from the bottom
of the lake of Geneva, and have been found alive at
considerable depths in the sea in Denmark, Maine,
and elsewhere.

I have described the adaptation of Chironomus
larvae to life at the bottom of slow streams and pools,
but we must notice that some of the species have
different habits, keeping near the surface of the water,
and supporting themselves on confervae, floating pieces
of wood and similar objects. It is interesting to see
the effect of this change of situation upon the respira-
tory organs. The surface-larvae have no tubules at
the hinder end of the body, and the blood rarely
exhibits a red tinge. I believe that the pupae which
succeed to these surface-larvae are similarly modified
to their peculiar conditions.

As the time of pupation approaches, the thorax of
the Chironomus larva becomes swollen, and its seg-
ments lose their distinctness. At this time the legs
and wings of the future fly may be seen beneath the

II

FLIES WITH AQUATIC LARVAE

133

skin. The pupa is distinguished from that of most
other aquatic Diptera by the tufts of respiratory fila-
ments which project from the prothorax. It lies half
buried in the mud at the bottom of the water with the
thorax and respiratory filaments projecting. These
are swayed to and fro by the continual
bending of the body. At the tail-end
two lateral flaps, fringed with long
bristles, form a broad fin, which is
used to a limited extent for locomo-
tion. The pupa virtually consists of
the body of the fly inclosed within a
transparent skin. The organs are al-
ready complete externally, and even
in microscopic detail they very closely
resemble those of the perfect Insect.
The parts are, however, as yet very
imperfectly displayed. The wings and
legs are folded up along the sides of
the body, and are incapable of inde-
pendent movement. The red colour
of the blood gradually assumes a
darker shade. After two or three days FlG 39>_Pupa of
the tracheal system, which was rudi- chironomus, side-
mentary in the larvse, but is now
greatly enlarged and extended, becomes filled with
air extracted from the water by means of the re-
spiratory tufts, and the pupa floats at the surface.
On the last day or so of the pupal stage, air passes
through the spiracles, and inflates the pupal skin.
At length the skin of the back splits, the fly extri-
cates its limbs and appendages, pauses for a moment

134 NATURAL HISTORY OF AQUATIC INSECTS CH.

upon the floating pupa-case, as if to dry its wings,
and then flies away. The whole business of extrica-
tion occupies less than a minute.

During the later larval stages, a number of new
organs are developed, which only become useful in
the winged Insect. Three pairs of long and slender
legs ultimately replace the very different larval feet ;
a pair of gauzy wings are added, and the head
becomes completely reconstructed. The powerful
mandibles and spinning appliances of the larva are
useless to the fly ; on the other hand, the fly will
want far more elaborate sense-organs than those of
the larva. The larva is an animal of very simple
mode of life, feeding upon dead vegetable matter at
the bottom of dark and slow streams ; the fly is a
nimble, aerial Insect, requiring keen senses to escape
danger and find a mate. The head of the larva is
accordingly both simple and small ; that of the fly
not only quite different in shape and far more complex
in structure, but absolutely larger. As a mere matter
of dimensions the head of the fly could not be con-
tained in the larval head.

Provision is made, not only in Chironomus, but in
very many Insects, for the gradual development of
new parts, differing materially from the old ones. An
Insect elaborates by secretion from its epidermis a
continuous integument or cuticle, which at length
becomes hard and firm. Whenever the cuticle proves
too small for the growing body within, the epidermis
is retracted, and develops a new cuticle within the
old one. The old cuticle ultimately bursts open, and
the Insect emerges, clothed in its new and very flexible

II

FLIES WITH AQUATIC LARVAE

135

covering.

Soon the new cuticle expands, and in the
course of a few hours, or it may be in a few days,
becomes as hard as that which has been cast. This
periodical operation of moulting gives opportunity for
gradual changes of shape. Some Insects, without
ever ceasing to run about and procure their food,
develop new organs, such as wings, by modification of
the shape of the cuticle at certain places. In the
later stages of a Cockroach, for
example, the edges of two of the
thoracic segments are found to
project backwards at the sides of
the body. Within these small pro-
jections wings are formed. They
are necessarily crumpled up to
allow of their being stowed away
into small spaces, and it is only at
the next moult that they become
expanded. By the same process
the long legs of a winged Insect
may be developed beneath the un-
broken surface of a footless larva.
There is a deep inward telescoping of the epidermis,
and from the bottom of the fold the limb begins
to grow outwards. (See Fig. 42.)

In a Chironomus larva the head, legs, and wings of
the fly are formed by such a process of infolding.
The operation occupies a considerable time, and
begins soon after the last larval moult, some weeks
before pupation sets in. The epidermis is retracted
from the larval cuticle, and first secretes a thin, new
cuticle, which will form the proper investment of

FIG. 40. — Larva of Chiro-
nomus. One of the
hind legs, showing its
crown of hooks, the
retractor muscles, and
the formation of a new
crown of hooks, prepar-
atory to change of skin.

136 NATURAL HISTORY OF AQUATIC INSECTS CH.

the pupa. This appears to remain plastic for a
long time, for it ultimately takes very exactly the
shape of the imaginal parts which form within it.
There is afterwards developed within it a second and
inner cuticle, which will ultimately form the integu-
ment of the fly. The two structures are always
distinct, but they are in most places in close con-
tact, and can only be distinguished by the micro-
scope. The pupal and imaginal cuticles do not at
all closely follow the larval skin, but become at
particular places folded far into the interior. The
folds which give rise to the head of the fly are two
in number and paired. They begin at the larval
antenna on each side of the head, and gradually ex-
tend further and further backwards. The object of
the folds is to provide an extended surface which
can be moulded, without pressure from surrounding
objects, into the form of the future head. On one
part of each fold the facets of the large compound
eyes are developed, another part gives rise to the
future antenna, a large and elaborate organ, which
springs from the bottom of the fold, and whose tip
just enters the very short antenna of the larva. The
folds for the head ultimately become so large that the
larval head cannot contain them, and they extend far
into the prothorax. Here a difficulty occurs. If the
generating cuticle of the prothorax were also to be
folded inwards, the future prothorax would take acorre-
sponding shape. But the prothorax of the fly has a
form dictated to it by the limbs which it bears, and
by the muscles to which it gives attachment. These
call for a great reduction in its length, and a peculiar

II

FLIES WITH AQUATIC LARVAE

137

shape, which it is not here necessary to describe. It
will be enough to realise that the epidermis of the
future prothorax cannot be sacrificed to the folds

FIG. 41. — Process of formation of the parts of the head of the fly in the larva of
Chironomus (male). A, the new epidermis thrown into complicated folds,
which have been cut away in places to show the parts within. B, the same parts
in horizontal section. Ic, larval cuticle ; //", transverse fold ; (/"', upper wall of
do. ; ;«, cut edge of new epidermis ; ant, larval antenna ; an, nerve to do. ; ant',
antenna of fly ; If, longitudinal fold ; o, eye of fly ; on, optic nerve ; an ', root of
antennary nerve ; br, brain ; as, oesophagus ; b, bulb of antenna of fly ; s, s, s',
blood-spaces.

which are to give rise to the head of the fly. All
interference between the two developing structures is
obviated by the provision of a transverse fold, which
pushes into the prothorax from the neck, and forms a

138 NATURAL HISTORY OF AQUATIC INSECTS CH.

sort of internal pocket. The floor of the pocket forms
two longitudinal folds, which prolong the folds origin-
ating in the larval head. The roof of the pocket
shrinks up and forms the connection between the head
and thorax of the fly. Ultimately the head-part is drawn
out, leaving the prothoracic structures unaffected.

The details of the formation of the head-parts of
the fly are too complex for description here.1 The
last stage of all, in which the paired folds, together
with the intervening tract, become drawn forwards
and moulded into a new head of simple convex
shape, can, however, be observed without serious
difficulty. Larvae about to undergo pupation are
easily distinguished by the thickened thorax. A
number of such larvae must be placed in saucers of
water, and observed continuously. The epidermis
and soft parts will be seen to b£ withdrawn from the
fore-feet, and about a minute later from the append-
ages near the tail. Then a bulge appears just behind
the head on the dorsal surface. The larval head,
suddenly emptied of its contents, slips round to the
lower surface. The deep infoldings described above
are everted, as when a sleeve is turned inside out.
The surface bearing the facets of the compound eyes,
which was internal, and the antennse of the fly, which
were deeply buried, now project from the exterior of
the newly protruded head. In the course of two or
three minutes, the head of the fly, lately sunk into
the body and forming a series of elaborate folds, is
everted, and takes the form usual in Dipterous flies.

1 See Miall and Hammond on the Development of the imago
of Chironomus. Linn, Trans., Vol. V. (1892).

II FLIES WITH AQUATIC LARVAE 139

The head is apparently protruded by blood-pres-
sure, set up by the contraction of other parts of the
body. When the head, still covered by the pupa-
skin, is once protruded, matters go on more slowly.
The old cuticle of the rest of the body is worked off
bit by bit, and it is not until some hours have passed
that the pupa gets completely rid of it.

It is interesting to the biologist to find various de-
grees of complication in the method of development
of new parts by infolding. In some Dipterous and
other larvae the folds which give rise to the imaginal
head are shallow, and the new parts are in close con-
tact with the corresponding organs of the larva, where
such exist. In the Muscidae, i.e. the family to which
the House-fly and Blow-fly belong, the formation of
the imaginal parts is incredibly complicated, though
the process is in principle the same as in Chironomus.
The degree of correspondence between the larval and
imaginal organs is one circumstance which affects the
extent of the infoldings. In Muscidae the fly differs
in almost every particular of external and internal
structure from the footless and almost headless larva,
and here the folds attain their maximum both of
number and complexity.

Nearly allied species of Insects, though closely
similar in the winged state, may differ considerably
as larvae or pupae. We may take an example of this
from Chironomus. In this genus there are two types
of larvae, each of which is succeeded by a special kind
of pupa. One group, which includes most of the
larger species, such as Chironomus plumosus, has two
pairs of respiratory tubules on the eleventh segment

140 NATURAL HISTORY OF AQUATIC INSECTS CH.

of the larva. The larva has red blood, and burrows
in mud at the bottom of streams and pools. The
pupae of such species have two bunches of respiratory
filaments. The second group, which is called by
Meinert the Motitator Group, is found at the surface
of the water on confervae, floating logs, etc. The
larva has no respiratory tubules ; the blood is usually
colourless ; the pupa has respiratory trumpets instead
of filaments.1

The origin of these two groups can only be
conjecturally traced. Both probably originated in a
Tipulid, which as a larva burrowed in the earth and
breathed air, and as a pupa was provided with two
respiratory trumpets opening by a very narrow cleft,
like those of Tipula. We may suppose that such a
larva betook itself to damp earth in the neighbour-
hood of streams, and gradually become more and
more aquatic. Subsequently a divergence of habit
arose in the aquatic forms. Some of the larvae
frequented the bottom of the water, while others
preferred to live at the surface. The bottom-feeders
developed the hollow tubes on the eleventh segment
as organs of aquatic respiration, and also haemoglobin
in the blood, as a means of storing up oxygen. Those
which took to life at the surface of the water, finding
oxygen easy to [obtain, were able to dispense with
both tubules and haemoglobin. Each group has its
characteristic pupa. One is adapted for living at the

1 Our knowledge of the early stages of Chironomus at
present extends to a very few species, and we do not know
whether the \vhole genus can be divided into these groups, nor
if so, how many species belong to each.

II FLIES WITH AQUATIC LARV^ 141

bottom, and provided with bunches of fine branching
filaments, which expose a considerable surface, and
are well adapted to the abstraction of oxygen from
the water, especially when waved to and fro by the
contractions of the bod}'. The pupae of the surface-
forms on the other hand have respiratory trumpets,
which are well suited for drawing-in gaseous air.
Such tubes would obviously be quite useless at a
depth of even one inch from the surface of the water.
It seems most likely that the surface-form is the more
primitive, and that the bottom-form is a special
adaptation of this. The primitive Tipulid, we may
be pretty sure, had neither respiratory tubules nor
haemoglobin in the larval stage. It is equally unlikely
that it possessed respiratory filaments in the pupal
stage. The majority of aquatic Diptera (Culex,
Tarry pus, Corethra, &c.) still retain with slight modi-
fications what we suppose to be the primitive form of
pupal respiratory trumpet. The pupa of Simulium,
on the other hand, which is immersed and fixed
beneath the surface of the water, has bunches of
filaments like Chironomus plumosus.

It is interesting to see in these Insects, how natural
selection can act upon the earlier stages of the life of an
Insect without materially affecting the structure of the
adult. The naturalist who should attempt to classify
aquatic Diptera by the larva and pupa, would place
Chironomus plumosus and Chironomus motitator in
different families, yet the flies seem to differ in nothing
of greater importance than the length of the first
joint of the fore tarsus. Other instances of the special
modification of provisional organs brought about in

142 NATURAL HISTORY OF AQUATIC INSECTS CH.

one species, while other closely allied species are not
so modified, are described by Fritz Miiller1 as due to
a process of falsification 2 set up by the struggle for
existence among free living immature forms. It is to
be regretted that our ignorance of the early stages of
nearly all the species of Chironomus prevents this
interesting question from being fully worked out at
present.

Certain facts in the development of the pupal re-
spiratory appendages point to a particular explanation
of the two forms, so widely different, which they
assume in nearly allied genera. When we watch the
developing thoracic appendages of the fly in such
Dipterous larvae as Chironomus, Culex or Simulium,
we find that they form a dorsal and a ventral series
each consisting of three pairs (see diagram).

i

u

0

u

u

•3

u

OS

2

0

0

0

i-G

rfi

1

o

rt

J

1

D1

D2

D3

V1

V2

V3

D2 and D3 will ultimately form the wings and
haltcres (poisers) ; V1, V2 and V3 the legs of the fly.

1 Facts for Darwin, Chapter xl.

2 The term falsification is not self-explanatory, and it is open
to objection as introducing quite irrelevant associations. The
obliteration of ancestral by adaptive characters is the thing
meant.

II

FLIES WITH AQUATIC LARVAE

143

D2"3 very early assume the appearance of undeveloped
wings, being broad, thin sheets, inclosed in folds of
similar shape. V1"3 soon take the form of long,
cylindrical processes, which in like manner, are in-
closed in tight-fitting cases. The case or sheath, in
both the dorsal and the ventral series, is merely the
outer part of the same fold which gives rise to the
appendage itself. Suppose the sleeve of a coat to be

FIG. 42.— i, Diagram to illustrate the formation of a new appendage : 2, one of the
developing legs of the Crane Fly (Tipula) in section.

partly telescoped or withdrawn into the upper part.
We shall then get an arrangement like that repre-
sented in the figure. A will be the shoulder of the
sleeve, C the wrist end, and B the part which comes
immediately in contact with C, separating it from A.
This is not unlike the folded-in and developing
thoracic appendage. A is a part of the new cuticle
which will help to form the wall of the thorax, C is

144 NATURAL HISTORY OF AQUATIC INSECTS CH.

the appendage itself, B the sheath of the appendage.
Now it becomes necessary to remark that Di will
give rise to the pupal respiratory organ, whether
tubular or tufted, further that Di may be in an early
stage somewhat wing-like, forming a thin, flat sheath.
What was its original purpose we cannot even con-
jecture. No Insect is known with a functional pro-
thoracic wing, and it is hard even to imagine an Insect
with three pairs of wings. The sheet-like prothoracic
appendage of the dorsal series may now curve round
until its edges approach and at last become united.
In this way the tubes or funnels which constitute the
pupal respiratory siphons of the Gnat,1 Corethra, &c.
are formed. Or the sheet may become supplied with
a network of large, branching tracheal tubes, which
repeatedly fork, as tracheal tubes in general do. If
we suppose that the intermediate substance, proving
superfluous, shrinks or is absorbed, the tubes will then
stand out as a tuft of forked filaments, exactly like
the pupal respiratory tufts of Chironomus plumosus,
Simulium, &c. I believe that this is the way in which
either a trumpet-shaped siphon or a tuft of branching
tubes is derived from a wing-like fold of skin.

The fly of Chironomus is a common object upon

1 Dr. C. H. Hurst, The Pupal Stage of Culex. Studies from
the Biological Laboratory of Owens College. This and the
same author's "Life-history of a Gnat" (Trans. Manchester
Micro. Soc., 1890) contain much interesting information.

I have adopted Dr. Hurst's view that " the first pair [of dorsal
appendages] become rolled up to form tubes, the respiratory
siphons, while the other two remain flat plates/' as agreeing
with the facts hitherto observed. I do not, however, consider
that it has been definitely proved.

II

FLIES WITH AQUATIC LARVAE

145

our window-panes, and would be called a Gnat by
most people. It can be easily distinguished from a
true Gnat, by its habit of raising the fore-legs from

FlG. 43.— Fly of Chironomus. A, male fly ; B, female fly ; C, antenna of male ;

D, antenna of female.

the ground when at rest. Gnats under like circum-
stances often keep the hind-legs raised. The Chir-
onomus fly is entirely harmless, and the mouth-parts
can neither pierce nor suck. Like many other Diptera,

146 NATURAL HISTORY OF AQUATIC INSECTS CH.

the flics of Chironomus associate in swarms, which
are believed to consist entirely of males. The male
fly has large plumose antennae with their dilated
bases almost in contact. In the female fly, the
antennae are smaller and simpler, and their bases arc
separated by an appreciable interval. The stomach
of the fly is usually empty or nearly so, and it is
probable that like many other winged Insects, it
never feeds.

The laying of the eggs of Chironomus is attended
with some special difficulties. It is convenient that
the eggs should be laid in water, and that they should
float on the surface, where they can get a fair supply
of air, and run no risk of being smothered in silt or
organic refuse, but they must not float free, for the
water of a running stream would carry them to great
distances, and perhaps lodge them in some very un-
suitable place, or even sweep them out to sea. These
requirements are met in the case of Chironomus and
some other Insects by laying the eggs in chains, and
moorincr them at the surface of the water. The eo-o-s

>-> oo

are invested by a gelatinous envelope, which swells
out the moment it reaches the water into an abundant
transparent mucilage. This mucilage answers more
than one purpose. In the first place it makes the
eggs so slippery that Birds or Insects cannot grasp
them : it also spaces the eggs, and enables each to
get its fair share of air and sunlight. The gelatinous
substance appears to possess some antiseptic property
which prevents water-moulds from attacking the eggs.
Long after the eggs have hatched out, the transparent
envelope remains unchanged. During the summer

II

FLIES WITH AQUATIC LARVAE

147

months, the egg-ropes, nearly an inch long, of some
very common species of Chironomus may readily be
found on the edges of a stone fountain in a garden,
or in a water-trough by the side of the road. The

A

FIG. 44. — Egg-masses of Chironomus. A. egg-rope of C. dorsalis, divided into sec-
tions, to show both sides ; B, twisted fibres, which traverse the egg-rope ; C,
egg-mass of another species of Chironomus ; D, egg-mass of a third species ; E,
part of do. (more highly magnified) ; F, developing eggs, two stages.

eggs are arranged upon the outside of the rope in
loops, which bend to right and left alternately, form-
ing sinuous lines upon the surface. Each egg- rope
is moored to the bank by a thread, which passes
through the middle of the rope in a series of loops,

L 2

148 NATURAL HISTORY OF AQUATIC INSECTS CH

and then returns in as many reversed and overlapping
loops, so as to give the appearance of a lock-stitch.
The thread is so tough that it can be drawn out
straight with a needle without breaking. If the egg-
rope is dipped into boiling wrater, the threads become
apparent, but in the natural state they are invisible,
owing to their transparency. The mucilage is held
together by the threads intenvoven with it. The
loops can be straightened without injury until the
length of the rope is almost doubled. If stretched
beyond this point the threads become strained, and
do not recover their original shape when released.
By means of these threads the whole mass of many
hundreds of eggs is firmly moored, yet so moored that
it floats without strain, and rises or falls with the
stream. The eggs get all the sun and air which they
require, and neither predatory Insects, nor Birds, nor
water-moulds, nor rushing currents of water can
injure them. All the species of Chironomus are not
alike in the formation of their egg-chains. See for
example Fig. 44.

It is always interesting to the biologist to find the
same contrivance made use of by animals of different
kinds, and I have therefore thought it worth while to
mention a number of other animals which coat their
eggs with jelly.

Insects yield many examples. One of the most
familiar is furnished by the Caddis-flies. Sometimes
(Mystacides) they lay their eggs in a roundish
gelatinous disc, a quarter of an inch across, which is
attached to an aquatic leaf, and contains hundreds of
eggs arranged within it in a regular spiral line. Some-

ii FLIES WITH AQUATIC LARVAE 149

times the egg-mass of a Caddis-fly takes the form of
a rope. One such egg-rope (species unknown) I have
examined. It was about two inches long, and con-
tained very many grass-green eggs, which hatched
out and yielded Caddis-worms.

The marine worm, Sagitta, also lays eggs imbedded
in jelly.

Some aquatic Snails make use in a variety of ways
of the same contrivance. The Pond-snails (Limnaea,
Physa) lay their eggs in strips of jelly, which are
attached to water-plants or submerged stones.
Bythinia lays its eggs in three long rows on stones or
water-plants (Jeffreys). In Planorbis the egg-mass
takes a globular shape (Jeffreys). In Valvata the
globular gelatinous mass is invested by a capsule, and
attached by a short stalk (Jeffreys). Natica lays its
eggs in broad strips of jelly upon shells or stones.
The Sea-snail, known to zoologists as Philine or
Bullsea, lays its e^crs on mud- banks in the form of a

J O o

small oval mass about as big as a hazel-nut. Upon
the surface of this the eggs are disposed in intricate
lines. There is a suspensory cord too, which super-
ficially resembles that of the Chironomus egg-rope,
and points to the necessity of mooring the eggs.
Many Nudibranch Mollusca employ a transparent,
gelatinous envelope for the protection of their eggs.
The eggs of Pteropods are found floating at the sur-
face of the sea, imbedded in a gelatinous substance,
and forming long, cylindrical strings. Many Cepha-
lopods lay their eggs in branched, gelatinous masses
attached to fixed objects.

Fishes furnish several examples of the same thing.

150 NATURAL HISTORY OF AQUATIC INSECTS CH.

The spawn of the Angler Fish (Lophius) spreads out
on the surface of the sea as a sheet, eight to ten feet
square. The Perch discharges her eggs enveloped in
a transparent substance, which she rubs off against
stones. They form a mass of loosely interwoven
strings, several feet long. The female Ceratodus also
lays her eggs in long, gelatinous strings.

Lastly, the Frog furnishes the most familiar ex-
ample of a transparent egg-mass. Laid in stagnant
water, the Frog-spawn does not require to be moored.
The eggs are so slippery that Insects or Birds, ex-
cepting only the broad-billed Duck, cannot, grasp
them. They are spaced so that each gets its fair
share of sunlight and air. Then the slimy covering
detains bubbles of air which expand in the wrarmth of
the sun's rays, and buoy up the eggs. Frogs' eggs,
when fresh-laid and free from bubbles, are slightly
heavier than water, and sink. After no long time
they float at the surface, except when the weather is
unusually colu. The presence of antiseptic proper-
ties in Frog-spawn is attested by the persistence of
the empty egg-envelopes for weeks after the Tadpoles
have hatched out. In the end they often become
overgrown by green cells, and in that condition are
eaten by the Tadpoles. The spawn of the Toad
forms long ropes. This is probably a provision for
hatching in slow streams. The eggs of the Frog
might possibly be carried away in such places, but
the egg-ropes of the Toad, which are many yards
long, get entangled in weeds, and never float far.

The egg-ropes of Chironomus are well worth the
attention of the naturalist for another reason, viz : —

ii FLIES WITH AQUATIC LARVAE 151

the ease with which the development of the larva
within the egg can be studied. In the first place, they
can always be procured during the summer months.
They are so transparent as to admit of examination
under high powers of the microscope as living objects,
and as they require no sort of preparation, they may
be replaced in the water after each examination to
continue their development. This saves all trouble in
determining the succession of the different stages — a
point which usually presents difficulties to the embry-
ologist. The whole development of the egg of
Chironomus is completed in a few days (three to six,
according to temperature), and it is therefore an easy
matter to follow the process throughout with the help
of three or four chains of eggs.

The development of any complex animal is a diffi-
cult study, and involves as a first requisite a good
anatomical knowledge of the adult organs, but even
a beginner can learn something from the study of
living eggs of Chironomus. In an early stage he will
see a transparent outer layer of cells, completely
inclosing the granular yolk.1 A little later the outer
layer thins away in one place, and in the end exposes
the yolk at that point, cutting off what will ultimately
be the head of the larva from the tail. Then the
transparent body becomes segmented, and the rudi-
ments of the antennae and jaws begin to appear. The
integument is pushed in at the head-end, so as to
form a long narrow tube which reaches up to the yolk.
This tube gives rise to the mouth and oesophagus.
A similar pushing-in of the integument at the tail-
1 See Fig. 44, F., left-hand embryo.

152 NATURAL HISTORY OF AQUATIC INSECTS CH.

end gives rise to the intestine. The stomach is
formed out of the cells which immediately inclose
the yolk. The hooked feet of the prothorax and of
the last segment form comparatively late. The
elongate larva is eventually coiled up within the
narrow egg ; it escapes by the bursting of what we
must call the egg-shell, and begins life on its own
account. Until the first moult takes place, the larva
differs in some respects from older larvae. The four
respiratory tubes on the ventral side of the eleventh
segment have not yet been developed, and the blood
has not yet assumed its red tinge. Moreover the
head is larger in proportion to the rest of the body
than at a later time, and contains, as in most Insects,
the brain, or fore pair of ganglia. Before long the
brain is slowly retracted into the prothorax, where it
is found during the rest of the larval period. The
relatively small head, which grows very slowly, could
not apparently furnish the requisite space for a brain
in addition to the powerful muscles of the jaws. It is
only after pupation that the head, that is, the new-
formed imaginal head, once more incloses the brain.

TANYPUS.

Another larva, closely related to that of Chironomus,
is found in similar situations. This is the larva of
Tanypus. It is nearly colourless, but is often tinged
with a reddish colour, which, on closer examination, is
found to be due to the contents of the alimentary
canal. The heads of Chironomus larvae, remaining

c">

undigested in the stomach, show the source of

II

FLIES WITH AQUATIC LARVAE

153

the reel colour. In one case no fewer than seven
heads of Chironomus larvae were found in a single

FIG. 45. — Larva and pupa of Tanypus maculatus. The four lower figures represent
the circular gelatinous egg-mass ; a developing egg, in side-view ; the tail plates
of the pupa, front view ; and the partly united, hook-bearing, prothoracic feet
of the larva.

Tanypus. Crustacea (sometimes still alive) arc
often to be seen in the crop. The head is narrow,

154 NATURAL HISTORY OF AQUATIC INSECTS en.

the antennae rather long and completely retractile ; they
can be withdrawn into the head by means of a special
muscle and protruded again by blood-pressure. Re-
traction of the sensitive organs and even of the entire
head is not unusual in burrowing Insect-larvae. The
first joint of the thorax is unusually long, and is
furnished with a pair of hooked feet, joined together
at the base. These are used like a crutch to propel
the body of the larva. At the hinder end of the
body is another pair of hooked feet, also long,
narrow and stiff. De Geer says very aptly that the
larva moves about as if it had wooden legs. The
extremities of the legs, which bear clusters of hooks,
are however retractile. The tracheal system is better
developed than in the Chironomus larva, but does
not, so far as I can discover, open to the surface.
The Tanypus larva makes tubes like those of Chiro-
nomus, but does not keep so close to them, and may
often be found free, swimming through the water
with a serpent-like movement or clinging to sub-
merged objects. It occasionally thrusts the fore part
of its body out of its tube or burrow, and strikes the
water with it. The movement is rhythmical, and
may be kept up for a long time. The object is no
doubt to promote aeration of the blood. In captivity
the larva seldom constructs tubes. The pupa gener-
ally keeps below the surface, but can come up to
breathe by means of a pair of respiratory trumpets.
When alarmed, it sinks, and often holds on to objects
at the bottom of the water by means of its tail. The
pupa is further provided with suckers on the abdomen,
which enable it to hold on to solid objects with great

ii FLIES WITH AQUATIC LARV^ 155

ease. Meinert, who has given the best account of
the larva and pupa of Tanypus, says that the suckers
are circular depressions outside the dorsal shields of
the abdomen. The pupa of Tanypus varius shows
them most distinctly : here they are borne in pairs by
four segments (3-6). When the pupa has attached
itself by a single sucker, it can turn about without
losing its hold.1

The eggs of Tanypus are described by Hammond2
as circular gelatinous masses, adhering to floating
objects. The eggs are arranged in double rows, along
about eight straight and parallel lines which extend
across the disc.

CERATOPOGON.

One of the commonest Dipterous larvae found at
the surface of ponds is the long slender worm-like
larva of Ceratopogon bicolor. It is commonly met
with entangled in the confervae which float at the
surface, and its extremely slender body seems well
adapted for travelling in a serpentine way through
a mass of vegetable threads. The head is very long
and slender ; there are no limbs. At the tail-end
is a crown of long bristles, which can either be ex-
tended backwards or curved forwards at pleasure.
These bristles are probably a means of attachment
and locomotion. The surface of the body is marked
with fine longitudinal lines, which look as if they
had been cut upon the cuticle by a microscopic

1 De eucephale Myggelaruer, p. 83.

2 Postal Microscopical Journal.

156 NATURAL HISTORY OF AQUATIC INSECTS CH.

graver. The digestive
tube is straight and
simple, and appar-
ently adapted to the
wants of a carnivor-
ous animal. I have not
been able to identify
the small particles of
food which are occa-
sionally seen within it.
There is a pair of very
delicate air-tubes run-
ning along the body.
The pupa swims at the
surface, and can attach
itself to floating objects
by means of two spines,
which project from the
extremity of the ab-
domen. It does not
completely free itself
from the larval skin,
but carries the loose
and shrivelled envelope
about, attached to its
tail. Some other species,
of quite different mode
of life, do the same.
Most species of Cera-
topogon are not aquatic
in their early stages,

FIG. 46.—!, Larva of Ceratopogorr, 2, head fou^ }}ve under the bark
of do. ; 3, mandibles ; 4, double outline,

showing flexion of the body in opposite ^f J^eeS. £. DufoUH IS
directions while swimming.

II FLIES WITH AQUATIC LARV^! 157

described,1 as living during its early stages in the
sap which flows from wounded elms. It creeps and
even swims about in this juice with great activity.

The flies hatched from aquatic larvae have naked
wings, while those produced from tree-haunting larvae
have hairy wings.

Some species of Ceratopogon sting, and inflict
painful wounds upon man. They are occasionally to
be found among other Tipulids upon the window-
pane, but their natural resort is the branches of trees
and shrubs.

Like most aquatic Diptera, Ceratopogon passes the
winter in the larval form; the pupa and imago appear
in May or June, and young larvae abound in the
latter part of summer. The eggs are laid 100 or
more together in star-shaped clusters among the
confervse.2

DlXA.

The larva of Dixa was first described by Reaumur
in a memoir published in 17 H-3 It is, he says, only
seven or eight lines long. There are eleven segments
behind the head. The larva is almost always bent
double into the shape of a siphon, so that the head
and tail come close together. The bend is at the

1 Laboulbene, Ann. Soc. Enlom. de France, 4e Ser., Tom. IX.
(1869).

2 Gercke, Verh. d. Vcr. f. natur-w. Unterhalt. Hamburg,
Bd. IV., p. 222-8 (1877). The larva and pupa have been well
figured and described by Meinert, De eucephale Myggelarver
(1886).

3 Mem. de I' A cad. Roy ale de Paris.

158 NATURAL HISTORY OF AQUATIC INSECTS CH.

sixth segment, and there are five segments on either
side of it, but as the five segments towards the tail
are longer than the five next to the head, the body
is not bent exactly in the middle. The larva keeps
close to the edge of still water. Only the head, the
tail, and the segment "next to the tail, he says, are
constantly in the water. He observed that when the
larva was attached to the side of a vessel, only the
head and tail being in the water, if he inclined
the vessel so as to cause the water to rise towards
the larva, it immediately crept a little higher. If the
vessel was inclined the opposite way, so that the larva
was left dry, it made haste to regain the water.
When creeping, the middle of the body moved fore-
most. The larva was provided with legs, though
they were too small to be seen properly with a
magnifying glass. The legs, he says, are attached
to the back, and the Insect always lies on its back,
with the mouth turned upwards. The legs were
short, and resembled the prolegs of a Silkworm,
being terminated bv hooks. The first four le^s were

O J O

directed towards the head, the others towards the
tail. Since the animal is bent double, all its legs
point the same way, and combine to propel the body
with the sixth segment foremost. The larva could
work its legs, not only from front to back, but also
from side to side. When submerged, it extended its
body, and swam like a worm. It appeared never to
venture into deep water voluntarily, but to regain
the shore as soon as possible. The mouth was
surrounded by prominences and tufts, which were
incessantly drawing water into the mouth, and thus

n FLIES WITH AQUATIC LARVAE 159

bringing a supply of the minute organisms which
serve the larva for food.

The description of Reaumur, given above in a
condensed form, is in the main a true and satisfactory
account. He was, however, mistaken in supposing
that the larva turns its back upwards, and that the
legs are borne upon the dorsal surface. Moreover
Reaumur was quite ignorant of the transformations
of Dixa. His successor, De Geer, in this as in many
other cases, added fresh information on points left

A.

incomplete by Reaumur. De Geer1 reared the fly,
which he identified as a kind of Tipula. He gives
descriptions and figures of all the stages. De Geer
shared Reaumur's mistake as to the dorsal and
ventral surfaces. Meinert has given the best modern
account of Dixa, and has supplied excellent figures
of the details of structure of the larva and pupa.2

The structure of the larva can be gathered from
the figures here given (Fig. 47). The body is bent,
as Reaumur says, into the form of a siphon, the fifth
and sixth segments behind the head forming the
bend. The surface which comes uppermost is, as
usual, the dorsal surface, and the locomotive organs
are ventral. They consist of two pairs of pseudopods
or prolegs, armed with hooks, and borne upon the
fourth and fifth segments. On the eighth, ninth, and
tenth segments are bunches of setae which answer to
some extent the same purpose ; these are called legs
by Reaumur. At the tail-end and on the dorsal
surface is a respiratory cup, of the same nature as

1 Hist, dcs In sect es, Tom. VI. p. 380.

2 De encephalc Myggelarver (1886).

160 NATURAL HISTORY OF AQUATIC INSECTS CH.

that found in the larvae of Hydrobius and Pericoma.

Valves and flaps and
outstanding processes,
all fringed with long
hairs, exclude the water
from a shallow sunk
space, upon which the
longitudinal tracheal
trunks open.

The body is bent
from side to side, not
from above downwards.
Hence the whole of the
proper dorsal surface re-
mains dorsal. Six seg-
ments (5--io) are pro-
tected on the dorsal side
by as many shields, all
fringed by setae.1

Reaumur's statement
that the two ends of the
body are in the water,
while the middle is out
of the water, is liable to
mislead. It would be
more correct to say that
the two ends are turned

FIG. 47. — Larva of Dixa, in side-view and
from above. The species here figured
is not described in any of the systematic
books which I have consulted. The
fly is 6 mm. long, of yellow colour, and
with uniformly clear wings.

1 These shields are not
found in all species of Dixa.
They are absent in the larva

j

figured by Meinert, which
occurs also in England.

ii FLIES WITH AQUATIC . LARV.E i6r

towards the water, while the bend is turned away from
it. The larva frequents pools which are overgrown
with vegetation. It creeps upon leaves which rise a
little way out of the water, and lies upon them with
its head and tail close to the edge of the water. The
bend is commonly above the water-line, but is com-
pletely wetted ; it does not break the surface-film.
Even when the larva has crept above the water-line it
is, strictly speaking, submerged, for it takes with it a
watery film ; the body is bent into a V, the apex of

FIG. 48. — Head of larva of Dixa, side-view.

which travels foremost, and each half of the V alter-
nately shoves the body onwards. Larvae placed in a
cup of water often wander to a distance and perish
by drying up. The respiratory cup is the only part
of the body which is dry, and this is in free com-
munication with the air. The head is commonly
sunk into the water, but close beneath the surface.
The ciliary organs which surround the mouth are
constantly employed, like those of a Gnat larva, in
sweeping microscopic particles into the gullet, and
this action is greatly promoted by the extraordinary

M

162 NATURAL HISTORY OF AQUATIC INSECTS CH.

mobility of the head, which can be bent completely
back till it touches the prothorax, or screwed round
till the line of the mouth lies in the axis of the
body. If the larva should slip into deep water, the
respiratory cup remains free from water, and buoys
up the tail. If the whole body is sunk beneath
the surface, a bubble is carried down enclosed in
the fringes of the respiratory cup. The larva when
thus submerged, swims energetically after the manner
of a Chironomus larva, twitching its body this way
and that, and making pretty rapid, though un-
steady progress. It can readily regain the surface
of the water.

The larva of Dixa makes use of the surface-film
not only to buoy up its tail, and to keep water out of
its air-filled basin, but also as an aid in climbing.
When it leaves the water it takes with it a close-
fitting, contractile film, which binds it to any wettable
surface with which it comes in contact. This use of
the film will become more evident if we suppose that
we have to climb a slippery pole. We grasp it as
hard as we can with arms and legs, and should find
the task greatly facilitated by a strong band of india-
rubber, which embraced both body and pole, and
spared us the necessity of grasping. The Dixa larva
is bound to the surface of a leaf by its contractile
water-film, and can therefore employ all its strength
in shoving itself along.

The same principle is turned to account by a
Gyrinus when it creeps up a smooth surface, though
here only the ventral surface of the Insect is
wetted. Mr. Scourfield has shown that certain

II

FLIES WITH AQUATIC

163

Copepod Crustacea creep out of water by a similar
use of the surface-film.1

The attitude, the mode of breathing and the mode
of feeding observed in the larva of Dixa, are curiously

fffi

id!

/(

FIG. 49. — Pupa of Dixa in side-view, and from ventral aspect.

like those of a certain Gnat larva, Anopheles. So
close is the resemblance that an experienced entomo-
logist has, in a published paper, mistaken one for the

1 Journ. Linn. Soc. Zool, Vol. XXV. p. 18 (1894).

M 2

164 NATURAL HISTORY OF AQUATIC INSECTS CH.

other. There are few better examples of adaptive
resemblance.

The pupa of Dixa is coiled into a nearly circular
disc, the abdomen being bent forwards beneath the
thorax. It has a pair of respiratory trumpets behind
the head, and either floats in a vertical position at
the surface of still water, or lies surrounded by the
vegetation which grows on the margin. It moves
but little, so far as I have seen.

The larva is of a full black colour, and very incon-
spicuous upon decaying vegetation. It may be found
from spring to late autumn, though in the height of
summer, when the transformation occurs, the larvae
become scarce. Though very few English naturalists
have seen them, they are probably common enough.

DlCRANOTA.

At the bottom of muddy pools and slow streams
there may often be found small red worms, half
buried in the mud. They attain a length of perhaps
a couple of inches. The head-end is concealed in
mud. The tail-end, which is more slender, projects,
and waves to and fro in the water, as if to effect the
aeration of the blood. If withdrawn from the mud,
the worms throw themselves into tight coils, and
sometimes form a tangled mass at the top of the mud.
The red colour is due to haemoglobin in the blood,
which courses along fine vessels, as in the Earthworm,
to which this animal, which is called Tubifex, is
rather closely related. Tubifex moves or withdraws
itself into its burrow by means of hooks projecting

II

FLIES WITH AQUATIC

165

from its body. There are four
rows of these, and each ends in
two points. This little aquatic
worm constitutes the favourite
food of more than one Insect;
which pursue it, as the larva
of the Ti°;er-beetle does the

o

Earthworm. I have lately in-
vestigated the structure and
habits of a Dipterous larva
which lives upon Tubifex. This
is the larva of Dicranota. The
full-grown larva is about three-
quarters of an inch long, and of
a dirty white colour. Com-
monly it buries itself in gravel
and mud, and may be fished
out of the bed of a stream con-
taining Tubifex. It creeps with
some rapidity through the mud,
and rarely exposes itself to view.
It can also swim about in the
water with a lashing or serpen-
tine movement not unlike that
of the Chironomus larva. Now
and then it leaves the water
altogether. The food of the
Dicranota larva is readily ascer-
tained by observing that the
stomach contains a red sub-
stance after Tubifex has been
supplied to it. The peculiar

'•J&iLLi

/ It'll I \M, ,'
' -••••'.•.-> . ,

*'. ',

•/. v/'

• 'A\')

N^/M

FIG. 50. — Larva of Dicranota
bimaculata. This and fol-
lowing figs, are copied from
a paper by L. C. Miall in
Trans. Entont: Soc., 1893.

1 66 NATURAL HISTORY OF AQUATIC INSECTS CH.

double crotchets of the worm can also be seen by the
microscope in the stomach or intestine. Like some
other carnivorous animals, the larvae can endure a
long fast. Before I knew of what their food con-
sisted, I kept them in small vessels filled with earth
and water, and found that they lived and moved
about actively for several weeks without food. It

seems to be well established that
certain Ticks, e.g. Argas persicus
and Argas reflexns (the Canterbury
Tick) can live for at least four
years without drawing blood.1 In
order that the Dicranota larva may
be able to pursue the worms, and
follow them into the depths of their
burrows, it is necessary that it
should be able to travel with
tolerable speed through mud and
gravel. Its body is accordingly
furnished with feet suited to such
a mode of locomotion. Five seg-
ments near the hinder end of the
body are provided with paired feet,
which resemble the false feet of
caterpillars. Each of these is furnished at the tip
with three circles of hooks, the terminal ones being
the longest.2 Almost the whole of the surface of the

^D

body is covered by a dense growth of minute pointed
hairs, which are directed backwards. In order that

A D. Michael in Natural Science, Vol. I. p. 202.
2 These circles are not complete, but interrupted variously,
as wab pointed out to me by Mr. J. J. Wilkinson of Skipton.

II

FLIES WITH AQUATIC LARV^l

167

the head may not take any harm from rubbing

against hard objects, it can be

completely retracted into the

body.

The most notable features of

the mouth-parts are a pair of

mandibles with long curved

teeth. The top of the head is

defended by a strong shield,

from the under side of which,

along the middle line, hangs

downwards a vertical plate. This

plate, as well as the rest of the

shield, gives origin to muscles,

and especially to the powerful

muscles which move the jaws.

There is a pair of small antennae,

and two rudimentary eye-spots on each side of the

head. Towards the tail are three pairs of tapering

prominences ; the hindmost pair is very long, and

forms the extremity of the body. These appendages

are supplied with rela-
tively large air-tubes,
and are probably of use
in aerating the blood.
Other means of respira-
tion are provided in the
form of spiracles. As
in a good many other
Insect-larvae, which in-
habit water or burrow in the earth, there is only one

pair of spiracles, which are situated at the hinder end

FIG. 52. — Head of larva of
Dicranota, ventral surface.

FIG. 53. — Head of larva of Dicranota,
retracted into thorax.

i68 NATURAL HISTORY OF AQUATIC INSECTS CH.

of the body. A large tracheal tube is connected with
each spiracle, and runs along the body to the head,

giving off many branches to the
various organs. The spiracle itself
consists of a solid cone received
within a wider tube, or vestibule.
The cone is long and tapering, and
apparently impervious. The be-
haviour of the living larva would
seem nevertheless to indicate that the
spiracle is actually employed in re-
spiration. When it comes to the
surface of the mud, it commonly pro-

FIG. 54.-Spiracle of trU<deS the hinder Cnd °f

This is furnished with the three pairs
object. transparent of prominences already mentioned,

which are probably tracheal gills, and
also with spiracles. We can therefore understand
how it is that the larva seems to be indifferent whether
its tail projects into air or water. It will lie for a
loner time in either condition without sioms of un-

o o

easiness. When it re-enters the water, after its tail
has been exposed to the air, a bubble can often be
seen attached to each spiracle.

The pupa, as in most other aquatic Dipterous
Insects, is provided with a pair of respiratory
trumpets placed just behind the head. These are
flattened from before backwards, and have a rather
sharp edge. \Vithin each trumpet is an expanded
tracheal tube, which blends with the outer integument
along the edge. The line of junction is perforated
by a regular row of small oval apertures, which no

II

FLIES WITH AQUATIC LARVAE

169

o

doubt serve for the admission of air to the trachea ;
their minute size must constitute an effective pro-
vision against the entrance of water or dirt. Along
the back of the pupa are a number of transverse rows
of minute spines. Almost the entire surface of the
thorax is roughened in the same way. Five of the
abdominal segments bear paired prominences on the
ventral side. By means of these
spines and fleshy processes, the pupa
can travel through the mud, and come
to the surface, when the fly is ready
to emerge. A pupa, when laid upon
damp mud, readily moves about, and
in no long time establishes itself in
a convenient position just below the
surface. Many other pupae of the
most diverse sorts, which bury them-
selves in earth or wood, have some
provision of the same kind. The
pupa of the Daddy-long-legs is armed
like that of Dicranota. Every col-
lector of Moths will recollect the spiny
rings of the pupae of the Clear-wings,
the Goat-moth, the Wood Leopard,
and the Swifts. The roughening of
the abdomen in such cases gives it a
sufficient hold on surrounding objects
to enable the Insect to creep to the
surface before the final change takes
place, thus avoiding the damage to gauzy or plumed
wings, which would inevitably result if the moth or
fly emerged in a narrow gallery, or below the surface

o o

o

FIG. 55. — Pupa of
Dicranota, ventral
surface.

i;o NATURAL HISTORY OF AQUATIC INSECTS CH.

of the ground. As in many other aquatic Diptera,
the pupa is notably smaller than the larva. In
Dicranota it is only three eighths of an inch long,
less than half the length of the larva.

I have never seen the eggs of Dicranota, but the
structure of the glands of the female fly leads me to
suppose that they are enveloped in a slimy mass or
egg-rope, and laid in water.

I have described the anatomy of the larva and
pupa pretty fully in the Transactions of the Entomo-
logical Society (1893). Since the publication of my
paper I have ascertained that Dicranota is not at
all uncommon in Yorkshire, and probably in other
parts of the country also.

PTYCHOPTERA.

In shallow muddy pools the larvae and pupae of
species of Ptychoptera are often plentiful. The
whitish bodies of the larvse can sometimes be seen
floating near the surface or lying above the mud, but
more frequently the body is buried, and only the tip
of the long slender tail protrudes through a small
hole. Sometimes the mud is pitted with scores of
such holes, like the places in which we find the little
red worms called Tubifex. If we dig with a trowel
in the mud so pitted, a plentiful supply of the larvae
can often be obtained. They are to be found
throughout the year, but the pupae occur only in the
spring and summer.

The body of the larva might easily be mistaken
for that of an Eristal'.s, on account of its long and

II

FLIES WITH AQUATIC LARVAE

i/i

retractile tail. It can, how-
ever, be distinguished from
Eristalis by the much more
complete head, and by the
presence of two slender ap-
pendages, one at each side
of the root of the tail. A
minute examination reveals
many other points of dif-
ference ; indeed, the two
larvae only resemble one
another in certain super-
ficial characters, which im-
mediately depend upon
their similarity of habitat.

The head of the larva
is hard and small, not un-
like that of the Chironomus
larva, but with many differ-
ences of detail. There is
a pair of eye-spots, and two
minute antennae, which can
only be made out by close
examination. The mouth-
parts are a good deal like
those of Chironomus.

Behind the head come
twelve segments, and it is
necessary to notice how
these are adapted to the life
of an animal which has to
creep in and through mud.

B

FIG. 56. — A, Larva of Ptychopterapalu-
dosa ; B, tail of larva of Ptychoptera,
much magnified, showing convoluted
air-tubes and lateral appendages.

172 NATURAL HISTORY OF AQUATIC INSECTS CH.

Most of the segments are thickened behind, and on
each of these annular thickenings a circle of stiff,
backward-directed setae (bristles) is borne. Like the
differently arranged setae of the Earth-worm, these
give a sufficient hold to enable the animal to advance
by alternate expansion and contraction of its seg-
ments. The action is aided by three pairs of hook-
bearing prominences, or pseudopods, of the same kind
as the prolegs of caterpillars.

The body is provided with numerous sensory hairs,
and also with two pairs of small vesicles situated on

the sides of the body in the
tenth and eleventh seg-
ments. These are described
by Grobben l as filled with
a fluid, in which float re-
fractive and transparent

FIG. 57.— Trachea (air-tube) of larva rrlobllles three as a rule to
of Ptychoptera, showing constric- JUlCh, III

tion. The upper figure shows a «orU vPQlVlf* A «?npria1
cross section of the same, and eacn VCSlClC. /\ bpCUdl

illustrates the provision for readily Ciir>r»1ip« the VPQtVlp
enlarging or contract mg the tube. SUppllCS IllC VCSlClC,

dilates to form a ganglion-
cell, and then tapers to its termination in the wall
of the vesicle. These vesicles are considered by
Grobben to be auditory, but no adequate reason is
assigned.

At the root of the long tail there is on each side a
long and slender appendage, which contains a rela-
tively large trachea. This is no doubt, as Grobben
says, a tracheal gill. A more important respiratory
organ is the long and retractile tail itself, which can,

1 Sitz. d. k. Akad. d. wiss. Wien. 1875.

ii FLIES WITH AQUATIC LARVAE 173

as in the Eristalis larva, be adapted to a varying
depth of water.

At the extremity of the tail are two openings ! by
means of which air is passed into the two tracheal
tubes, which traverse the length of the tail, and are
continued through the whole length of the body. In the
centre of most of the segments the tracheae become
greatly enlarged, the intervening portions, from which
many branches are given off, being comparatively
narrow. Each tube, therefore, resembles a row of
bladders connected by small necks. A cross-section
shows that the tubes are not cylindrical, but flattened,
and that while the lower surface is stiffened by the
usual parallel thickenings, the upper surface is thrown
into two deep, longitudinal furrows, so that it is
readily inflated into a cylindrical shape, and readily
collapses again when the air is expelled. It seems
likely that the buoyancy of the larva can thus be
regulated, and a larger or smaller quantity of air
taken in as desired.2

The larva, thus fortified against any interruption of
its supply of air, lies at its ease on or in the water,
stretching out its long tail to the surface, and break-
ing the surface-film, like the Eristalis larva, by the
extremity of the appendage.

The pupa has a pair of respiratory tubes, which
are carried on the thorax, close behind the head.
One of these tubes is very long, the other very short
and altogether functionless. The long tube is twice

1 Grobben, loc. cit.

2 Grobben observes that in very young larvae the main
tracheae are cylindrical, and of uniform diameter.

174 NATURAL HISTORY OF AQUATIC INSECTS CH.

as long as the body, and tapers very gradually to its

free tip. Here we find a
curious radiate structure,
rather like the peristome
of a moss-capsule, which
seems adapted for open-
ing and closing. There
is, however, no orifice
which the most careful
scrutiny has succeeded
in discovering. A deli-
cate membrane extends
between the teeth, and
prevents any passage in-
wards or outwards of air
in mass. The tube en-
closes a large trachea, the
continuation of one of
the main tracheal trunks.
This is stiffened by a
spiral coil, but at inter-
vals we find the coil de-
ficient, while the wall of
the tube swells out into
a thin bladder. How-
ever the tube is turned,
a number of these
bladders come to the
surface. As the pupa
lies on the surface of
the mud, the filament

floats on the top of the water, and the air

X '00

FIG. 58. — Pupa of Ptychoptera paludosa,
with long respiratory tube. A second
(rudimentary) tube is also seen. '1 he
figures to the left show the extremity of
the tube, and a portion of the middle,
both highly magnified. The trachea
lies in the tube. Both thin out at the
bladder-like swellings, and come close
together, as if to allow a free inter-
change of gases.

II FLIES WITH AQUATIC LARVAE 175

is renewed without effort through the thin-walled
bladders.1

The head and thorax of the pupa are defended and
roughened by many minute projections. The abdomen
is softer, but armed with five circles of fine spines.
Thus, as in Dicranota and many other Insects, the
pupa is enabled to creep to the surface of the ground
before the extrication of the fly. Like many other
Dipterous pupae, it has a fair degree of mobility, and
can travel about in the ground, so as to choose a
situation as convenient as possible in respect of
dampness.

SlMULlUM.

In brisk and lively streams the little, blackish
larvae of Simulium can sometimes be found in count-
less numbers. They attach themselves by choice to
water-weeds, such as float-grass, water-cress, water-
crowfoot, and the like. They may also be found on
stones, but I believe that the larvae found in stony
streams belong to a different species. In a stream
where all the necessary conditions are fully satisfied,
where there is a never-failing supply of well-aerated
water, plenty of submerged foliage, and plenty of
microscopic organisms, the larvae sometimes abound,
looking like small black worms, five-eighths of an
inch or less in length. When the leaves are examined
by the eye alone, without handling, perhaps hardly a

1 Grobben describes minute orifices in the outer wall of
these bladders which I have not myself seen. It is often a
matter of the greatest difficulty to decide whether a minute and
transparent membrane is completely closed or perforated.

176 NATURAL HISTORY OF AQUATIC INSECTS CH.

single larva will be seen. They are clustered for the
most part on the under side, and only become visible
when the leaf is plucked or turned over. They are
most numerous where the current is brisk. In York-
shire I find them plentiful in weedy streams which
come down from the moors, and especially in the
rapids above waterfalls Hagen describes the larvae
and pupae of Simulium pictipes as abounding in the

FIG. 59. — Group of larvae of Simulium attached to a stone.

rapids of the Ausable River, in the Adonirack Moun-
tains. Here the pupa-cases are fixed to the rocks in
clusters, which resemble small wasps' nests.1 Though
the larvae require well-aerated water, it need not be
pure. Streams contaminated with sewage often con-
tain them in great numbers.2

1 Proc. Boston Soc. Nat. Hist. Vol. XX. pp. 305-7 (1880).

2 The species which I have chiefly studied appear to be S.
latipes (stony streams) and S. reptans (on water- weeds in rapid
streams).

II

FLIES WITH AQUATIC LARY.£

177

The food of the larva is altogether microscopic.
I have found the stomach filled with the flinty valves
of Diatoms and Desmids, with here and there bits of
a small Crustacean. A pair of fringed appendages,
one on each side of the head, are employed, like
the similar apparatus of the Gnat larva, in sweeping
particles into the gullet.

The body of the larva is cylindrical, and dilated in
the hinder part of the abdomen. The mouth-parts
are represented in Fig. 61. In addition to these are
seen the fan-like appendages, each provided with

FIG. 60.— Head of larva of Simulium, dorsal view, showing eye-spots, antennae, and

fringed appendages.

about fifty long filaments, which are feathered along
one side, and sweep the food into the gullet. Since
the subsistence of the larva depends upon these
delicate organs it is not surprising to find that great
pains are taken to keep them from being clogged.

N

178 NATURAL HISTORY OF AQUATIC INSECTS CH.

By the help of a lens the larva can often be seen
combing them out with its mandibles. There are two

pairs of eyes, which are
reduced to mere pigment
spots, and small three-
jointed antennae.

There are two pairs of
le^s, as in the larva of

o '

Chironomus, and these end
in coronets of hooks. But
in the Simulium larva
there is no burrow to cling
to, and the leers become

FIG. 61. — Head of larva of Simulium,

ventral view, showing mouth-organs, greatly modified. Each pair

The eye-spots of the dorsal surface

are seen through. COalcSCCS tO form a single

organ, which is mainly

employed as a sucker. The fore pair, borne upon
the first thoracic segment, is however occasionally
used in grasping, being opposable to the head. The
action of these suckers is easily demonstrated by
placing a larva fresh from the stream in a saucer of
water. It creeps about like a Leech, applying the
two ends of its body alternately to the smooth sur-
face of the saucer. Even in a rushing stream it
appears never to be detached against its will. The
hooks, not unlike in form to the hooks found on the
legs of the Chironomus larva, form concentric rows
on the extremity of the fused legs of the Simulium
larva. They are probably useful to prevent slipping
upon the smooth leaf or slimy stone. In the suckers
of a Cuttle-fish we find the same contrivance resorted
to. Each sucker is a cup whose cavity can be en-

II FLIES WITH AQUATIC LARV^ 179

larged by muscular pull, so that the pressure within
the cup becomes less than the external pressure.
The edge of the cup is roughened, as in Simulium.
by numerous minute teeth, which prevent slipping.

The larva, when feeding, holds on by its tail-sucker,
and sticks its body straight out into the current. But
if the current is unusually strong, the larva, when not
feeding, often doubles up its body, and holds on to
the surface of the leaf by both suckers.

The life of a submerged Insect in a rapid current
has of course its own special difficulties. The Simu-
lium larva has to creep from leaf to leaf, to change
its position as the stream rises and falls, and to avoid
dangerous enemies. Of these enemies aquatic larvae,
and especially Caddis-worms, are the commonest and
most formidable. In moving about there is always
risk of being accidentally dislodged, and the conse-
quences of this might be serious. For if a larva
should let go or miss its hold in a rapid stream, what
is likely to happen ? It seems inevitable that it will
be swept away, and who knows where it will come to
rest ? The little rivulet which I am accustomed to
visit for the purpose of observing this larva is a
bright clear stream, flowing over water-cress and
brook-lime and forget-me-not. A few feet lower
down it ends in the wide and stony Wharfe, a stream
of quite different character, in which I have never
been able to discover a single specimen of the same
species. Other brooks in which the larvae are plen-
tiful, empty themselves into rivers unsuited to an
Insect of habits so peculiar — muddy, sluggish, or
brackish. But this difficulty has been provided

N 2

iSo NATURAL HISTORY OF AQUATIC INSECTS CH.

against, and I find that the larva is seldom or never
swept away, even when its haunts are invaded by a
groping naturalist. Its swimming power, I should
explain, is inconsiderable, altogether insufficient for
propulsion against a rapid stream. If seriously
alarmed, the larva lets go, and immediately dis-
appears from sight. But by watching the place
attentively, we shall before long see the larva work-
ing its way back, and in a
minute or two it will be found
attached to the very same leaf
from which it started, or to
some other leaf, equally con-
venient, which it happens to
fall in with. I found the diffi-
culties of observation in fast-
flowing water crowded with
leaves very great, until at last

XfO

it occurred to me to push a

FIG. 62.— Head of larva of Simu-

Hum, showing on left side the white plate in amonsf the

position and form of the new fc>

larval antenna and fringed ap- Ipovpc Then trie rlarl-

pendage, which will become ICdVCb. ddl K-

external at the next moult ; on coloured larvp^ herame r>er
right side the new imaginal larVSE

antenna, which will become fart]\7 evident on the \vhite
external after pupation. ICC11J

ground, and I was able to see

exactly how they managed. When disturbed by the
plate, some of them let go, and drift a few inches away.
They are not very easily frightened, and most of them
remain holding on by their sucker. Those which quit
the leaf remain stationary in the torrent or nearly
so, and on close observation a thread, or perhaps a
number of threads, become visible on the white
ground. These threads are in general stuck all over

II FLIES WITH AQUATIC LARVAE 181

with small vegetable particles, like fine dust, which
make them much more apparent. The threads ex-
tend in all directions from leaf to leaf, and the larva
has access to a perfect labyrinth, along wrhich it can
travel to a fresh place by help of the current and
with the speed of lightning. I suppose that it grasps
the thread with its prothoracic clawrs, for when it
comes to rest it is always found holding on by them.
To recover its first position is not difficult if the net-
work of threads is intact, or if the larva has even a
single thread to grasp. Sometimes it hauls itself up
hand over hand, like a Leech or a Looping caterpillar,
applying its two suckers alternately to the thread.
It can also creep along its thread by means of the
prothoracic hooks only. I have often seen the body
swept into a position at right angles or nearly so
with the thread ; then the larva seems to be holding
fast by its head. When thus attached, it can travel
against the stream at a slow but quite appreciable
rate.

Although the larva commonly slides along a thread
previously made, and easily seen to be an old one by
the small particles which cling to it, it can upon a
sudden emergency spin a new thread, like a Spider
or a Geometer larva. The new threads are perfectly
clean, and consequently invisible even on a white
ground so long as they are submerged. I got proof
of the formation of fresh-spun threads by dislodging
larvae which had attached themselves to leaves wiped
clean a minute before. But the clearest proof is got
by suddenly lifting out of the water a leaf with many
larvse upon it. One or two are pretty sure to let

i82 NATURAL HISTORY OF AQUATIC INSECTS CH.

themselves drop a foot or more in the air, and the
thread can be seen to glisten in the sun, and to
lengthen itself at the pleasure of the Insect. The
Simulium larva, possessing this power of instantan-

FIG. 63. — Late larva of Simulium, showing changes within the larval skin, fl, fi, fi,
fore, middle and hind legs of fly ; ra, respiratory appendages of pupa ; ?u, wing
of fly; /i, haltere of fly.

eously manufacturing a rope, can hardly be taken at
a serious disadvantage.

The salivary glands, which secrete the silk, are
unusually large in this larva. They extend the whole
length of the body, and then bend forwards for a
third of its length. The tracheal tubes are large,
and give off a network of fine branches to the skin.
These appear to absorb oxygen from the water, for
there is no opening into the tracheal system.

When the time for pupation comes, special pro-
vision has to be made for the peculiar circumstances

II

FLIES WITH AQUATIC

183

in which the whole of the aquatic life of the Simulium
is passed. An inactive and exposed pupa, Ir-ke that
of Chironomus, may fare well enough on the soft

FIG. 64.— Simuiium. A, larva in outline, with salivary glands. B, coronet of hooks
at tail-end of larva. C, part of silken cocoon of pupa, showing its interwoven
fibres, the product of the salivary glands.

muddy bottom of a slow stream, but such a pupa
would be swept away in a moment by the currents
in which Simulium is most at home. Before pupa-
tion the Insect constructs for itself a kind of nest,

1 84 NATURAL HISTORY OF AQUATIC INSECTS CH.

not unlike in shape to the nests of some Swallows.
This nest is glued fast to the surface of a water-weed.
The salivary glands, which furnished the mooring-
threads, supply the material of which the nest is
composed. Sheltered within this smooth and taper-
ing cocoon, whose pointed tip is directed up-stream,

A B

FIG. 65. — A, four pupae of Simulium in their cocoons, attached to aquatic stem ; B,
pupa of Simulium, removed from its cocoon.

while the open mouth is turned down-stream, the
pupa rests securely during the time of its trans-
formation.

When the cocoon is first formed, it is completely
closed, but, when the Insect has cast the larval skin,
one end of the cocoon is knocked ofif, and the pupa
now thrusts the fore-part of its body into the current

II FLIES WITH AQUATIC LARV^ 185

of water. The respiratory filaments, which project,
immediately behind the future head, just as in
Chironomus, draw a sufficient supply of air from
the well-aerated \vater around. The rings of the
abdomen are furnished with a number of projecting
hooks, and as the interior of the cocoon is felted by
silken threads, the pupa gets a firm grip of its
cocoon. If it is forcibly dislodged a number of the
silken threads are drawn out from the felted lining.

A serious difficulty now appears. The fly is a
delicate and minute Insect, with gauzy wings. How
does it escape from the rushing water into the air
above, where the remainder of its life has to be
passed ? This was a question upon which I spent
much thought, but in vain. It appeared to me for
many months completely insoluble. In the stream
at my feet I could see countless pupae. There were
also empty cocoons, containing, as close examination
showed, the cast pupal skin of Simulium, still attached
to the felted lining, and the last larval skin as well.
On the bushes which overhung the stream innumer-
able Simulium flies were resting. How had they
made their way in safety after the last transformation,
although their structure was to all appearance ex-
tremely ill adapted to such a manoeuvre, through the
strong current into the upper air ? I was at last
informed by Baron Osten Sacken of a paper written
by Verdat, seventy years ago, in which the emergence
of the fly of Simulium is described.1 During the

1 Gesch. d. Simulien, Natunv. anz. d. allg. Schiv. Gesellsch.
(1822). Translated by Osten $acken, in dwer. Ent, Vol. II,
p. 229.

1 86 NATURAL HISTORY OF AQUATIC INSECTS CH.

latter part of the pupal stage, which lasts about a
fortnight in all, the pupa-skin becomes inflated with
air, which is extracted from the water, and passed
apparently through the spiracles of the fly into the
space immediately within the pupal skin. The pupal
skin thus becomes distended with air, and assumes

FIG. 66. — Respiratory appendage of pupa of Simulium, consisting of four branched
tubes. Removed from beneath larval skin, and therefore unexpanded.

a more rounded shape in consequence. At length it
splits along the back, in the way usual among Insects,
and there emerges a small bubble of air, which rises
quickly to the surface of the water and then bursts.
When the bubble bursts, out comes the fly. It
spreads its hairy legs, and runs upon the surface of
the water to find some solid support up which it can

II FLIES WITH AQUATIC LARV^ 187

climb. As soon as its wings are dry, it flies to the
trees or bushes overhanging the stream.1

The ascent of the Simulium fly through the waters
of a rushing stream is rendered safe by the hairy
surface of the body. The surface-film, by which the
air is imprisoned, clings to the hairs, and refuses to
enter the fine spaces between them. In the same
way a covering of velvety hairs prevents the Diving
Spider, as well as many diving Insects, from wetting
that part of their bodies which bears the spiracles.

The flies of Simulium congregate, sometimes in
extraordinary numbers, on bushes and low trees near
the streams which harboured the larvae and pupae.
The males are distinguished from the females by
their large heads, a great part of which is covered
by the enormous compound eyes. In the male, but
not in the female, the upper part of each eye is built
up of larger facets than the lower half. No reason
is known for this curious feature, which is not un-
paralleled among Insects. The thorax is of large
size, the abdomen small. The legs are little used for
walking, but are continually exercised in what look
like exploring movements, the flies patting the
surface of a leaf with their feet. They are said to

1 Verdat's original description is graphic and worth quota-
tion. When the fly reaches the surface of the water, he says,
" tous ses membres se deploient a la fois comme par explosion.
II se tient facilement sur 1'eau, ou il marche par le secours des
pelottes de 1'extremite des tarses, qui sont alors bien developpes,
et il gagne une tige voisine sans se mouiller. En tres peu de
temps ses ailes sont deployees et affermies. II quitte son ber-
ceau en prenant son essor, se repand partout dans la campagne
pour y chercher des alimens et sa femelle."

1 88 NATURAL HISTORY OF AQUATIC INSECTS CH.

suck the juices of leaves and the honey-dew secreted
by Aphides. Verdat says that they distend their
abdomen with blood. De Geer says that they attack
large, smooth caterpillars, sucking their blood. The
caterpillars, he adds, do not take much notice of
them.1 The males love sunshine, and may be seen
in swarms high in the air. The females generally
remain at lower levels.2

In some countries the flies of various species of
Simulium have proved a serious plague by the irri-
tation which they cause to men, horses, and cattle.
In Hungary and on the lower Danube great losses
have resulted. The bite raises a small blister, and
rouses the cattle to fury. They rush about wildly,
rubbing their hides against rocks or trees. The
corner of the eye is the part of man said to be aimed
at. Swarms of these flies have greatly annoyed
certain Australian exploring expeditions. In Britain
the flies of Simulium are perfectly harmless.

The pupae and flies may be found in great numbers
in April and May, and again in August. The Insect
winters as a larva, gaining maturity in spring, and
laying eggs, which produce the autumn brood.

The co'Q's are laid on stems and leaves of water-

oo

plants, hundreds being attached together in one
gelatinous mass. They have a thick shell, and are
of a yellow colour. They are like Chironomus eggs
in shape, but are much less transparent.

1 Memoires, Vol. I. p. 328.

2 Osten Sacken, Bcrl. Entom. Zeits. Bd. XXXVII. (1892).

II FLIES WITH AQUATIC LARVAE 189

STRATIOMYS.

The curious, history of this Insect has been de-
scribed for us by two master-hands. Swammerdam l
bestowed upon it all his skill and industry ; and
Reaumur2 independently worked over the same
ground. Reaumur's fourth volume bears the same elate
(1738) as the second volume of the Biblia Natures,
and Swammerdam's investigations were probably
not accessible to him at the time of writing.

Several species of Stratiomys are common in
Europe. The commonest and best known is S.
chamseleon, which is that figured by Swammerdam
and probably one of those figured by Reaumur.3

The larva of this species is, when full grown, from
two to three inches long, being capable of consider-
able elongation according to circumstances. The
body is flattened and elliptical in cross-section, nar-
rowed at each end, especially at the tail-end, which is
very slender. Behind the small, horny head come
eleven segments, which increase greatly in length and
diminish in breadth towards the tail. Each of the
first four segments is overlapped by the succeeding one,
while the remainder are each overlapped by the seg-
ment in front, so as to facilitate a telescopic retraction
at either end of the body. The skin is tough and
flexible, and defended by a peculiar structure, to

1 Biblia Natures, pp. 649-694 ; PI. 39-42.
1 Memoir es, Tom. IV. Mem. VII. VIII.

PI. 22 and part of PI. 23. A number of Reaumur's figures
belong to the closely related Odontomyia.

NATURAL HISTORY OF AQUATIC INSECTS CH.

mm

iiJiJSiS

I'iuHsss

FIG. 67.^-Stratiomys chamaeleon. i, larva; 2, hrva floating at surface of water;
3, larva descending ; 4, pupa within larval skin (2, 3, and 4 are from Swammer-
dam, Biblia Natures) ; 5, head of larva, dorsal view, a a. marks the attachment
of the thoracic integument ; 6, head of larva, ventral view. The ventral wall is
incomplete behind, and the pharynx and gullet are exposed ; 7, piece of integu-
ment ; 8, do. in section, with conical, calcareous nails ; 9, a single calcareous
nail, surface view ; 10, spiracle, lying in centre of tail-coronet.

II FLIES WITH AQUATIC LARV^ 191

be described later on, against the attacks of predatory
animals. The colour is in some degree protective,
consisting of shades of green, brown, and yellow, and
exhibiting an irregular pattern of spots and streaks.
Stagnant waters of small extent, and especially
ditches, are its favourite haunts. The larva of
Odontomyia, which is very similar to that of Stra-
tiomys, prefers the shelter of Duckweed and other
aquatic plants. Stratiomys occurs in various parts of
England, but is reputed rather rare, perhaps because
so little attention has been paid by English naturalists
to aquatic Diptera.

The head of the larva is greatly reduced, and con-
sists of a middle part, black, horny and pointed, which
is separated by deep clefts from the lateral parts. In
these clefts lie the palps, which bear strong hairs. A
number of hooks arm the mouth. The food consists
of microscopic organisms swept into the mouth by
the palps.

The extremity of the tail forms a beautiful coronet
of branched filaments, about thirty in number, which
are accessory to respiration. When this coronet is
expanded it forms a basin open to the air and im-
pervious to water, by reason of the fineness of the
meshes between the component filaments. Were the
larva provided with a basin of the same proportions
formed out of continuous membrane it might float
and breathe perfectly well, but would find it hard
to free itself neatly and quickly from the surface-film
when some sudden emergency rendered it necessary to
descend. As it is, the plumed filaments collapse and
their points approach ; the side branches are folded

192 NATURAL HISTORY OF AQUATIC INSECTS CH.

in, and the basin is in a moment reduced to a pear-
shaped body, filled with a globule of air, and reaching
the surface of the water only by its pointed extremity.
Down goes the Stratiomys larva at the first hint
of danger, swimming through the water with swaying
and looping movements, somewhat like those of

FIG. 68. — Tail-coronet of larva of Stratiomys chamaeleon. From Swammerdam,

Biblia Naturce.

Chironomus. When the danger is past, it ceases to
strueele, and floats again to the surface or swims

^5 & *J

upwards by a lashing movement. The pointed tip
of its tail-fringe pierces the surface-film, the fila-
ments separate once more, and the floating basin
is restored.

In shallow water, the last one or two segments are
usually bent upwards, so as to reach the surface.

n FLIES WITH AQUATIC LA.RV& 193

There are no limbs. The larva moves in the water
by vigorous flexion of its body, first to one side and
then to the other. It can also creep on earth or
mud. Swammerdam observes that, like other Dip-
terous larvae, it drags itself along by its mouth when
out of the water. Alternate contraction and exten-
sion of the segments, aided by the stiff hairs which
stand out from the body, are also used as a means of
propulsion.

The wider part of the body lodges the alimentary
canal, which is very long and thrown into complicated
folds. The salivary or silk glands are tolerably large.
A pair of very large air-tubes run along the body.
Beginning as narrow tubes in the head, they dilate
until they attain half the width of the body, gradu-
ally narrowing after this until they terminate in
oval spiracles which occupy the centre of the
coronet. There are nine pairs of spiracles on the
sides of the body, which are not open, but take the
form of black, circular plates. Branches from the
longitudinal air-tubes pass to them on the inside.
The second segment, which will ultimately carry the
wings, has no spiracles. In the same way the second
and third segments of Insect-larvae which are destined
to acquire two pairs of wings bear no spiracles, though
closed tracheal tubes running to the surface may
often be discovered by dissection in the places where
spiracles would be expected.

Stratiomys changes to a pfipa in early summer.
The larvae prefer to leave the water at this time, and
bury themselves in the earth or in the floating vege-
tation of a ditch. Swammerdam says that even in a

o

194 NATURAL HISTORY OF AQUATIC INSECTS CH.

vessel containing nothing but water their transforma-
tion was not hindered. The Insect becomes rigid ;
its skin hardens and contracts to some extent ; the
last segments become distorted. On opening a larva
which exhibits these marks of change, the animal will
be found to have shrunk and changed its shape. It

now presents the form of the
future fly, but is shrouded in a
silky cocoon, the secretion of
the salivary glands, and occupies
only the fore part of the space
within the larval skin (Fig. 67, 4).
Some entomologists have even
taken the pupa to be a parasite
which had eaten up the larva of
the Stratiomys, and undergone
transformation in the empty
skin.1 Remnants of the air-tubes
and the cast lining of the in-
testine moor the pupa, as it were,
to the tail of the larval skin.
The great reduction in size of
Stratiomys during transforma-
tion is an extreme case of what
may often be remarked in In-
sects. The larva gets some unknown advantage from
a length and bulk which would apparently be incon-
venient to the flying Insect.

The peculiar structtire of the larval skin now
acquires a special value, especially when the pupa is

1 Westwood, Classification of Insects, 1 1.. p. 532, quotes such
cases.

FIG. 69. — Pupa of Stratiomys,
removed from larval skin.
From Swammerdam,.Z>/$/;Vz
Natures.

II FLIES WITH AQUATIC LARVAE 195

not buried in earth, but floating on water. The larva
can descend into the water when attacked, but the
pupa is too buoyant, and too much encumbered by
its outer case, to execute any such manoeuvre. Pro-
vision has accordingly to be made for the protection
of the helpless pupa against its many enemies. It is
probable that hungry Insects and Birds mistake the
shapeless larval skin, floating passively at the surface,
for a dead object. The considerable space between
the outer envelope, or larval skin, and the body of the
pupa may keep off others, for the first bite of a Dytis-
cus or Dragon-fly larva would be disappointing. Still
further security is gained by the texture of the larval
skin itself. The cuticle consists of two layers. The
inner is comparatively soft and laminated, while the
outer layer is impregnated with calcareous salts, and
extremely hard (Fig. 67, 7, 8). The needful flexibility is
obtained by the subdivision of the hard outer layer.
Seen from the surface, it is broken up into a multitude
of hexagonal fields, each of which forms the base of
a conical projection, reaching far into the softer layer
beneath. The conical shape of these calcareous nails
allows a certain amount of bending of the cuticle,
while the whole exposed surface is protected by an
armour, in which even the pointed mandibles of a
Dytiscus larva can find no effective chink.

Swammerdam once found a pupa of Stratiomys
dead within its larval skin. In the abdomen were
larvae and pupae of a parasitic Insect (perhaps an
Ichneumon or a Tachina) which had sprung from eggs
laid within the Stratiomys. Neither water nor hard
skin can exclude these cruel and insidious enemies.

O 2

196 NATURAL HISTORY OF AQUATIC INSECTS CH.

The pupa-stage lasts for a very few days and then
the fly emerges. The larval skin splits across the
third and fifth segments, and also along the middle
line between these segments.1 The fly then pushes its
way out, and Reaumur tells us that it suffers no in-
convenience if it emerges from a floating larval skin.
The hairs which cover all parts of its body prevent
it from being wetted, and it can stand upon the water
as upon solid ground. The deserted larval skin, which
served as a cocoon, contains nothing but torn tracheal
tubes and the cast lining of the gullet and intestine.

When the fly first appears its wings are crumpled,
but they soon unfold by injection of blood into their
veins and of air into their air-tubes. At this time
they bleed if cut, which is not the case afterwards.

The fly has at first sight a Bee-like appearance.
Its abdomen is broad and flat, of a full black with
yellow patches and spots. The wings when at rest
lie one over the other, do not cover the sides of the
body, and project beyond it. The antennae are
simpler in structure than those of the Blow-fly ; they
are commonly described as 3-jointed, but the so-
called third joint is long and 5 -ringed, as if formed
by the consolidation of as many originally distinct
joints. From the back of the thorax two spines
stand out towards the abdomen, and hence the name
Stratiomys, which means the " armed Fly." The
winged Insect haunts flowers, especially heads of
Umbellifers, and is most often seen in the neighbour-
hood of pools and ditches.

1 The exact position of the splits seems to vary according to
the species.

II FLIES WITH AQUATIC LARV^ 197

The structure of the antennae is greatly relied upon
by systcmatists as a mark of the leading groups of
Diptera, and the antennas of Stratiomys are con-
sidered as intermediate between the many-jointed
antennae of the Nemocera (Culex, Chironomus, Tipula,
&c.) and the highly specialised and peculiar antennae
of a Blow-fly or House-fly. This intermediate
position of Stratiomys gives a special interest to the
manner of its transformation. It is the simplest in
structure and life-history, and the most like a Nemo-
ceran, of all the Diptera which retain the larval skin
as the outermost covering of the pupa.1 Here we
find the artifice in its simplest form ; it is protective
and useful, but does not as yet involve serious altera-
tion in the enclosed pupa. When the larval skin of
Stratiomys bursts to allow of the escape of the fly, it
bursts after the Nemoceran manner, that is, by split-
ting longitudinally, though there is also transverse
fission. In the Muscidae the anterior segments of the
larval skin are detached by a transverse fissure only.
The protection afforded by the hardened larval skin
seems to have led in the Muscidae to a very peculiar
and interesting process of destruction and regenera-
tion of all the tissues of the body. The old organs
break up, and the new organs are developed apparently
anew, but really from rudiments developed during the
larval stage. (See Chironomus, p. 135.) This pro-
Some species of Ceratopogon do not free themselves from
the larval skin at the time of pupation, but allow it to encumber
the tail of the pupa. The pupa of a certain species of Subula
pushes out at the head end of the larval skin and remains
sticking there.

198 NATURAL HISTORY OF AQUATIC INSECTS CH.

cess, once thought to be peculiar and unique, exhibits
gradations too, and passes by many intermediate
steps into a mere moult without change of shape.
To trace and study these gradations is one of the
most interesting of the biological problems cf the
future. Those who attempt to connect the Nemocera
with the Muscidae will find that Stratiomys is an
important link, illustrating as it does the origin of a
change which has profoundly modified the life-
history.

The eggs are said to be attached in overlapping
rows to the underside of aquatic plants.

THE RAT-TAILED MAGGOT (ERISTALIS).

The naturalist who comes across the Rat-tailed
Maggot, a common inhabitant of stagnant pools,
especially such as are foul with decaying organic

i

matter, should turn at once to that memoir by
Reaumur in which he speaks of " two-winged Flies
which resemble Bees, Wasps and Hornets." For the
sake of those who have not Reaumur at hand I will
give the substance of his remarks, but I cannot hope
to keep any of the charm of the original. Reaumur
discourses in a leisurely way, talking freely of his
personal experiences, and handling his subject with
the light and amusing manner of an accomplished
writer, who must needs be clear, but above all things
hates pedantry. It is hard upon so pleasant an
instructor that he should be compressed until all' his
individuality is lost, and those who can should by all

II

FLIES WITH AQUATIC LARVAE

199

means read Reaumur's own words
rather than my abstract.

He remarks in the first place
that there are Dipterous flies so
curiously like Bees and Wasps that
even an experienced entomologist
will hesitate to handle them for
fear of being stung. They have
the colour, the size, and very nearly
the shape of the formidable Insects
which they imitate, and like them
they haunt flowers. Some of them
pass their feeding stage in water.
The larva of Eristalis is one of
these. It bears a very long tail,
on which account Reaumur pro-
posed for it a name which is still
current, that of the Rat-tailed
Maggot (ver a queue de rat}.

The larva has seven pair of short
feet set at nearly equal distances,
the first being near the head, and
the last not far from the root of
the tail. The extremity of each
foot is circular, and armed like the
prolegs of a caterpillar with a large
number of fine hooks ; the feet
can be retracted at pleasure. The
hooks are set in a double circle,
the circle next to the tip being
made up of larger and less numer-
ous hooks than the other. Thq

K?V.«rr"V*'**

r<- it .. •/ • "v .« ..

'..v-'i?

FIG. 70. — Larva of Eris-
lis, ventral view.

200 NATURAL HISTORY OF AQUATIC INSECTS CH.

extremity of the foremost pair of feet is not cylin-
drical as in the rest, but flattened, and the hooks
project from it like the fingers from a hand.1

The larvae of Eristalis were first brought to Reau-
mur singly, and had been found on the ground. He
did not then know that the larva is really aquatic.
One day however he emptied a bell-glass which con-
tained water as well as the remains of many Insects
and leaves. The mud at the bottom was black and
putrid. In it were seen more than two hundred long-
tailed larvae. He washed them in clean water, when
they were seen to be of whitish colour. They were
divided into parcels, and put into separate vessels,
which were filled with water to a depth of two inches.
Before long the use of the long tail appeared. At
first the larvae moved about in various ways, some
swimming through the water, others creeping on the
side of the vessel or at the bottom ; but in less than a
quarter of an hour all were at rest, their long slender
tails just reaching the surface. Some had the hinder
end of the body nearly vertical and in line with
the tail ; others were creeping on the bottom, the tail
being nearly at right angles with the body. Some
had the tail straight, while in others it was thrown
into a single or double curve.

Since the water in the vessel was about two inches
deep, the tails were of that length also — very long for
larvae whose bodies were at most two thirds of an
inch in length. To ascertain whether they could
lengthen them yet more, Reaumur added water to

1 The larva can sometimes be seen to creep upon the under-
side of th$ surface-film by means gf its feet.

II FLIES WITH AQUATIC LARVAE 201

the depth of another half inch, when the tails were
lengthened to the same amount. Again and again
water was added, and the tails became lengthened in
proportion. When however the depth grew to five
and a half or six inches, the larvae could no longer
reach the surface from the bottom. Some crept up
the side of the vessel, while others floated up to such
a height that they could maintain their communi-
cation with the air.

The organ, by means of which the larva is enabled
to breathe when its body is several inches under
water, deserves close examination. So transparent is
the tail, and indeed the rest of the body, especially in
young larvse, that the internal parts can be studied
almost as well as if their integument were a tube
of glass. It is easy to see that the tail consists of
two tubes, one sliding within the other. The outer
and wider tube looks like a prolongation of the body-
wall ; it is thrown when at rest into innumerable
transverse wrinkles. Within it is a second tube, a
considerable part of which is brown or nearly black.
The smaller tube is the special respiratory organ.
The tail can be lengthened by protruding this inner
tube from the outer one. But besides this mode of
adjustment, the outer tube is capable of elongation
and contraction, being furnished with innumerable
annular fibres and increasing in length as it diminishes

o o

in diameter. When most extended it is no thicker
than a stout thread, and the inner tube resembles
a horse-hair.

Mr. J. J. Wilkinson furnishes me with the following
additional information respecting the tail of the

202 NATURAL HISTORY OF AQUATIC INSECTS CH.

Eristalis larva : — " The tail is divisible into three
sections. There is an outer sheath, immediately

FIG. 71. — Diagrams of tail of larva of Eristalis larva; i. fully extended ; 5, com-
pletely retracted ; rt, respiratory tube ; chs, channelled sheath ; 7t'i, wrinkled
sheath ; ^, body of larva. In 3 and 4 the retractor muscles are shown ; 6 shows
the tracheae within the respiratory tube and channelled sheath (J. J. Wilkinson).

continuous with the integument of the body ; an
intermediate sheath, marked by about forty longi-
tudinal furrows, and hence called 'the channelled

II

FLIES WITH AQUATIC LARVAE

203

sheath' by Batelli ; and an inner ' respiratory ' tube,
which encloses the tracheae. The outer sheath is
much wrinkled transversely, especially when con-
tracted. It bears great numbers of minute hair-like
setae. Batelli says that it is also provided with setae
shaped like Roman swords, and he figures these
(Tav. I, Fig. 11). I have found similar objects on both
the outer and the channelled sheath, but by examina-
tion with a high power could see
that they were diatoms, and not
setae. Four pairs of tubercles bearing
bunches of setae project from the
outer sheath. There is one pair at
the base, where it joins the body, one
pair at the tip, and two intermediate
pairs. These are set at equal dis-
tances, and mark out the surface of
the sheath into three equal tracts.
The intermediate or channelled sheath
is exposed when the tail is com-
pletely protruded, but during retrac-
tion it disappears more or less within
the outer sheath. Between the furrows are regular

o

longitudinal rows of small hooklets. The inner or re-
spiratory tube is marked with innumerable transverse
folds or wrinkles. Its extremity is firmer and more
solid than the rest ; here are the external openings of
the two tracheal trunks, surrounded by five plumose
setae, which spread out and cling to the surface-film,
thus keeping the tip of the tail above water ; when
forcibly submerged, they carry down a bubble of air.
Two tracheal trunks, prolongations of the great longi-

FIG. 72.— Extremity
of respiratory tube
of larva of Eristalis.

204 NATURAL HISTORY OF AQUATIC INSECTS CH.

tudinal air-tubes of the body, run through the tail side
by side. In cross-sections made through the respira-
tory tube they appear
kidney-shaped, the con-
cavities facing each other,
but in front of the base of
the respiratory tube they
become cylindrical. When
the tail is completely ex-
tended the tracheae are
straightened out, but dur-
ing contraction of the tail
they are thrown into nu-
merous and regular coils,
which lie in the hinder
part of the body.

" Retraction of the tail
is largely effected by six
muscles which are attach-
ed to the junction of the
respiratory tube and the
channelled sheaths. When
these muscles contract the
respiratory tube is with-
drawn bit by bit into the
channelled sheath, and
ultimately into the outer
sheath (see Fig. 71), the
external surface of the

FIG. 73.-D:agram of internal structure channelled sheath beCOITl-

of larva of Eristalis, showing the inrr a r fViP Qamp> rim^ fnlrlprl
retracted head, the tracheal system,
and the telescopic structure of the tail
(J. J. Wilkinson).

.nvir

imv

TVn'c maw crr\ nn

i ru§ may go on

II FLIES WITH AQUATIC LARV^ 205

until only the tip of the respiratory tube projects
from the outer sheath, the channelled sheath disap-
pearing altogether. If further retraction is necessary
the muscles of the outer sheath itself come into
action, and this sheath shortens to any requisite ex-
tent. Its hinder end is furnished with a strong

d>

circular or sphincter muscle, which grips the respira-
tory tube at such times and prevents it from slipping
out. The base of the respiratory tube then be-
comes curved or even doubled up within the body,
reaching in some cases almost as far as the head
(Reaumur), and the coils of the tracheae become very
numerous and close. Reaumur and Batelli are in-
clined to suppose that the coiled tracheae play the
part of springs and help to effect the protrusion of
the tail, but this I cannot think to be correct. The
tracheae, even if it be admitted that they can act as
springs, have no base from which to act. The pro-
trusion of the tail is, I think, entirely due to pres-
sure set up within the body by contraction of its
muscular wall. If you take a dead larva with con-
tracted tail between the finger and the thumb, and
gently press the contents of the body backwards, the
tail will suddenly shoot out to its full length.

" The tracheae are continued forwards into the
body, becoming dilated into large air-sacs. They are
in communication with a pair of anterior spiracles
which appear on the upper surface of the first seg-
ment. These spiracles are not functional in the
larva."

Reaumur says that he often saw a bubble appear
at the tip of the tail in a submerged larva. At first

2o6 NATURAL HISTORY OF AQUATIC INSECTS CH.

no bigger than a pin's head this bubble dilated con-
siderably, then contracted, and then dilated again, as
if air had been forced out of the tracheal tubes and
sucked up again.

I have also to thank Mr. J. J. Wilkinson for the
following account of the very peculiar pharynx of
the Eristalis larva, based upon his long-continued
observations and inquiries. A slighter description of
the organ has been published by Batelli.1

" The head of the Eristalis larva is soft and capable
of changing its shape. It is rounded in front, and
when extended, shows two small sensory papillae,
which are carried on the under side. Behind these
is the opening of the mouth. The pharynx is lodged
in a brown chitinous capsule, below which can some-

v\\.\

^(/(M%f& -_

P^mvW'ftJ v j

<^^m*ttL.J •* ^. -^

^\\V\tft>^
.- VflOMMUl^ •

^>^i^;

FIG. 74. —Head of larva of Eristalis, ventral view, showing mouth and sensory

papillae.

times be seen a fleshy process which resembles the

tongue of the larva. The larva often puckers up its

mouth, so as to conceal its jointed appendages, and

1 Bull. Soc. Entom. ItaL, 1879, pp. 77-120, pi. 1-5.

II

FLIES WITH AQUATIC

207

even the first pair of legs. The fore part of the body
is covered with short reddish spines, which vary in
shape according to the species
examined. In some larvae they
are single-pointed, in others
three- or four-pointed, while in
a third kind they divide into
from sixteen to twenty points.

" The larva possesses a re-
markable sucking and filtering
apparatus in its muscular
pharynx. This is situated in theFlG.7S._Sen?orypapilteoflarva
second and third segments, and aLEngaangiiIith The "erves

inner

forms an oval body, held in its P*pi"a is two-jointed,
place by a number of muscles, and stiffened by a
complicated chitinous skeleton. Two thin and broad
chitinous bands extend along the hinder part of the
pharynx ; these become narrower and more solid in
front, where they give off processes for muscular
attachment and for the support of the mouth-organs.
Here the chitinous bars unite, and form a strong
arch, completed below by a stirrup-shaped piece.

" When the pharynx is opened, it is found to enclose
t\vo chambers, an upper and a lower, which are separ-
ated by a horizontal floor of unusual construction.
The upper chamber is lined by a loose membrane,
which dissolves in potash solution. This can be
dilated by special muscles. From the hinder end of
this chamber proceeds the cesophagus, the entrance
to which is controlled by a muscular valve. The
lower chamber also has a loose lining, which is chitin-
ous, and supported by the chitinous skeleton already

208 NATURAL HISTORY OF AQUATIC INSECTS CH.

mentioned. Both chambers are subdivisions of one
cavity, and are only distinguished here for con-
venience of description.

The partition or floor between the two chambers
is formed by nine ribs, which are long, narrow and
deep, resembling planks set at equal distances, upon
their edges. They are attached beneath to the wail

FIG. 76. — Cross-section of pharynx of Eristalis larva, showing below the fringes of
barbules, some open and some closed ; above these, the upper chamber ; and
towards the top of the figure the muscles in paired masses, with tracheal tubes.

of the lower chamber, but their upper edges are free,
and carry fringes of numerous and close-set barbules.
The outermost rib on each side bears only one fringe,
which springs from the inner and upper border of the
rib ; each of the remaining seven ribs has a pair of
fringes, and in cross-section resembles the letter Y,
the two arms being the fringes, and the stem the
supporting rib.

ii FLIES WITH AQUATIC LARV^ 209

" The lower chamber communicates with the
mouth-opening by a short passage, which is imper-
fectly divided along its length by a fleshy, tongue-like
prominence, the tip of which is beset on its upper
surface with innumerable spines. Near the openings
of the two branches of the passage, the inner wall of
the mouth is armed with a pair of shell-like plates,
concave towards the mouth, and bearing upon their
concave surfaces numerous prominent ridges.

" The apparatus is employed in the following way.
The food of the larva consists of organic particles,
scraped from submerged objects by the hooklets
around the mouth. When the larva is feeding, its
movements resemble those of a pig, working over a
heap of refuse with its snout. As soon as a quantity
of particles have been detached, the larva dilates the
upper chamber of its pharynx by means of the
attached muscles. Then the water, bearing the par-
ticles with it, rushes into the mouth, and passing
through the lower chamber, is sucked into the upper
chamber. The current, as it sweeps past, lifts the
fringes of barbules. Then the upper chamber con-
tracts, and the superfluous water escapes by the
mouth. In doing so, it closes the valve-like fringes.
The particles arc left clinging to the barbules, and
only the filtered water passes out. \Yhen a sufficient
quantity of food has been collected, the passage into
the oesophagus is opened, and the whole mass is
swallowed.

" The roughened surfaces of the shell- like plates
lie on cither side of the tongue-like prominence
previously mentioned. A great multitude of cross-

P

210 NATURAL HISTORY OF AQUATIC INSECTS CH.

A. — Pharynx expanding. Hanging valve and fringes open. Water and food

passing into upper chamber.

B. — Pharynx contracting. Hanging valve and fringes closed Water driven out,

food retained in upper chamber.

C. — Pharynx contracting. Transverse fold closed down upon inferior pharyngeal
arch (" stirrup "). Food and water forced into oesophagus.

FIG. 77. — Pharynx of Eristalis larva. Three diagrammatic sections, illustrating the
action of the parts (J. J. Wilkinson). In C the " shell-like plates" are seen to
the right of the figure. Between and behind them is the "tongue-like
prominence (triturator)." The muscles which raise the roof of the upper
chamber are indicated in all the sections by oblique lines.

II FLIES WITH AQUATIC LARV^ 211

bars, which in some places pass into teeth, stand out
from the ridges. How the plates are employed has
not been discovered. They cannot be protruded from
the mouth.

"Just in front of the pharyngeal fringes, and in the
floor of the passage, is a transverse row of backward-
directed setae.1

" Batelli aptly compares the action of the fringes
to that of the strainers of the Baleen Whales."

The larvae feed upon organic matter, decaying
leaves, &c., found in stagnant water. Reaumur found
that when kept in clean water and supplied with
bread, they lived until they had completed the larval
stage.

When the time for pupation comes the Rat-tailed
Maggot quits the water, its body becomes encrusted
with earthy particles which adhere in consequence of
the sticky exudation from the skin. When placed
on damp loose earth, they enter it and undergo their
transformation beneath the surface. For convenience
of observation Reaumur kept some larvae which were
ready to change in empty boxes. They glued them-
selves to the wood, and if only slightly attached by
the hinder part of the body, the transformation was
successfully effected ; but when the whole of the
under side stuck to the wood, the larva perished.

These Insects, like the Blow-fly and Stratiomys
form a cocoon out of the larval skin. The colour
changes from white to yellow ; the tail becomes

In the larva of the Blow-fly the sensory papillae, the shell-
like plates and the pharyngeal strainer or its equivalent can be
observed. See Lowne on the Blow-fly, 2nd ed., Figs. 5 and 8.

P 2

212 NATURAL HISTORY OF AQUATIC INSECTS CH.

wrinkled, but still retains a considerable length ; the
body shortens and becomes rather thicker than
before. After ten or twelve days the skin is hard
and opaque, and to a great extent detached from the
pupa within.

After twenty-four or thirty-six hours from the time
of pupation two pairs of horns appear at the head
end. The fore pair, is about half the length of the
other, and the two pairs often curve towards one
another. The fore pair, when diligently observed, are
found to be present in the larva, but they become
much more conspicuous at the time of pupation by
the smoothing out of the adjacent parts. The large
posterior pair appear for the first time shortly after
pupation. These horns are really the respiratory
organs of the pupa, as is clearly shown by the large
tracheal trunks which enter them.

The process of pupation is similar in Eristalis and
the Blow-fly, though more rapid in Eristalis. Twenty-
four hours after the posterior horns have appeared,
the proboscis, the wings and the legs may be seen
by opening the cocoon. The long tail has now no
connection with the body of the pupa. Near its base
are found a number of coiled tracheal tubes, collected
into a bundle. The main tracheal trunks, however,
persist.

In a favourable season the fly is ready to emerge
eight or ten clays after pupation begins. The part
of the cocoon contiguous to the four horns is pushed
off in two unequal pieces. When this happens two air-
vesicles appear, one attached to each of the posterior
horns ; the smaller horns each receive a large trachea.

II FLIES WITH AQUATIC LARVAE 213

Some other two-winged Flies, such as the Blow-fly,
which form a pupal cocoon out of the larval skin,
emerge by pushing off the end of the cocoon. For
this purpose the head of the fly is provided with a
vesicle capable of dilation or contraction. Reaumur
says that he took it for granted that the fly of
Eristalis would emerge in the same way. Finding a
pupa in which the part of the cocoon adjacent to the
respiratory horns was burst open, he looked into
the cavity and saw part of the body of the fly
alternately elongating and contracting itself. This
he naturally supposed was the head, but on closer
examination it proved to be the tail-end of the body.
Shortly before the fly emerges it must turn itself
completely round in its cocoon, a manoeuvre of some
difficulty, seeing how nearly the body fills the cavity
in which it lies.1 The flies appear in early spring,
and again in autumn.

Several species of Eristalis or allied Flics came
under Reaumur's notice. Though agreeing in essen-
tial points of structure and mode of life, they differed
in size, coloration (especially in the winged state) and
to some extent in the situations selected for the
deposit of the eggs.

More than one species every year laid eggs in
buckets partly filled with water, which were kept
in Reaumur's garden, and he was able to observe
the operation. The winged female flew round the
inside of the bucket, as if to explore its surface, and

1 Mr. J. J. Wilkinson finds that the supposed reversal of the
body in the larval skin is a fable. The fly comes out head
first (1903).

214 NATURAL HISTORY OF AQUATIC INSECTS CH.

now and then touched the water with her legs. When
satisfied as to the suitability of the place, she alighted
a few inches above the water-level, and extended her
abdomen considerably. The pointed and curved ex-
tremity was several times protruded and again re-
tracted. Then the female flew or walked away, and
repeated the operation in a fresh place. On search-
ing the spots where the fly had settled, which were
usually in the crevices between the staves of the
bucket, eggs, sometimes to the number of twenty,
were found attached to the wood. They were white,
of oblong shape, and not unlike those of the Blow-fly ;
the surface was finely shagreened. The eggs were
viscid, and adhered to the wood.

Living trees often contain cavities formed by decay.
When these become filled by rain water, they form
favourite sites for the egg-laying females, especially
in autumn. One species which is found on trees has
the larval tail much shorter than common. Before
pupation the larva attaches itself to some object by a
viscid exudation from the skin. In another species
which also passes its larval stage in the cavities of
trees the tail is very short, and flanked by a pair
of conical fleshy prominences. At the extremity of
the tail the two posterior spiracles are plainly to be
seen. At the head end are two small yellowish
projections which also communicate with the trachcal
trunks. These arc the anterior spiracles. A little
behind these and more external comes a pair of
double hooks. There is also at the extreme fore end
of the body, and just above the mouth, a pair of
fleshy prominences, as in other rat-tailed maggots.

II FLIES WITH AQUATIC LARVAE 215

The seven pairs of ventral feet are set close to
the middle line, and each is fused with its fellow.

Baron Osten Sacken 1 has shown that the resem-
blance of the fly of Eristalis to a Bee has caused the
curious belief of ancient times that Hive-bees could
be produced by the putrefaction of dead animals, and
especially from the carcasses of oxen. " The original
cause of this delusion lies in the fact that a very
common fly, scientifically called Eristalis tenax (popu-
larly the Drone-fly), lays its eggs upon carcasses of
animals, that its larvae develop within the putrescent
mass, and finally change into a swarm of flies, which
in their shape, hairy clothing and colour, look exactly
like Bees, although they belong to a totally different
order of Insects. Bees belong to the order Hymenop-
tera, and have four wings ; the female is provided
with a sting at the end of the body : the fly Eristalis
belongs to the order Diptera, has only two wings
and no sting."

Ovid, Virgil, and a crowd of less famous authors
vouch for the long-established belief in the miracle.
But the prescription for raising a swarm of Bees
given by Florentinus, a Byzantine author of about
the tenth century A.D., is more explicit and solemn
than any other extant. Ostcn Sacken translates his
instructions thus :-

"Build a house, ten cubits high, with all the
sides of equal dimensions, with one door and four

1 On the O.i -en-born Bees of the Ancients (Bugonict) and their
relation to Eristalis tenax , a two-winged Insect. Heidelberg,
1894. London : R. H. Porter. Published in shorter form in
Bull. Soc. Entom. Ital.^ 1893.

216 NATURAL HISTORY OF AQUATIC INSECTS CH.

windows, one on each side ; put an ox into it,
thirty months old, very fat and fleshy ; let a number
of young men kill him by beating him violently
with clubs, so as to mangle both flesh and bones,
but taking care not to shed any blood ; let all the
orifices, mouth, eyes, nose, &c., be stopped up with
clean and fine linen, impregnated with pitch ; let
a quantity of thyme be strewed under the re-
clining animal, and then let windows and doors be
closed and covered with a thick coating of clay, to
prevent the access of air or wind. Three weeks later
let the house be opened, and let light and fresh air get
access to it, except from the side from which the
wind blowrs strongest. After eleven days you will
find the house full of Bees, hanging together in
clusters, and nothing left of the ox but horns, bones
and hair. When the shed is opened, small white
animalcules are seen, resembling each other, motion-
less ; then they are observed to grow little by little,
to develop wings, and to take the colour of Bees ; the
wings are at first short and tremulous, unfit for flight ;
the limbs are weak ; finally, desirous of light the
Insects start and knock against the windows, pushing
each other."

Virgil's description of the process is very similar.
The chamber to be built must have four windows ; a
young ox is to be slain and his flesh pounded with-
out shedding of blood ; thyme and cassia are to be
strewed. The words " trunca pedum primo" (with
stumps of feet) apply exactly to the larva of Eristalis.

Wasps and hornets were similarly supposed to
proceed from the carcasses of horses. Ostcn Sacken

ii FLIES WITH AQUATIC LARVAE 217

attributes this belief to the swarming of the bee-
like Bot-fly of the horse (Gastrophilus equi) about
horses. The belief would be strengthened by the
occasional productions of the wasp-like Helophilus
(a close relative of Eristalis) from carcasses of the
horse, though any other dead body would do as
well.

It may be asked whether the ancient husbandmen
were not acute enough to find out that their Bees,
formed according to prescription, laid no honey ; and
further, why no Insects, likely to be mistaken for
Bees, are seen to issue from carcasses in our own
time and country. The experimenters of old were,
we must suppose, satisfied to let their swarms of arti-
ficially produced Bees fly abroad. It was enough in
their eyes to have saved the race from extermination.
Bee-keepers, even in the Augustan age, no doubt
knew better ways than this of getting swarms into
their hives, and the supposed generation of Bees from
carcasses was probably a belief rather than a method
likely to be tried by practical men. The answer to
the second question may be given in Ostcn Sacken's
o\vn words : — " The- larva of Eristalis, being aquatic,
requires a pool of stagnant water containing putrescent
organic matter. A carcass, left in the open air, is at
once attacked by the ordinary carcass-loving flies,
Liicilia, Callipliora., &c. During the second stage of
putrescence a pool of corrupt liquid is formed about
the carcass, and then is the time for Eristalis to
appear. In the present age of sanitary police a
carcass would never reach this stage." But in Lap-
land Zetterstedt, as quoted by Ostcn Sacken, found

218 NATURAL HISTORY OF AQUATIC INSECTS

Eristalis hovering about the carcass of a sheep which
was immersed in putrid liquid.

The habits of Eristalis tend to bring it about the
habitations of man. Hence it is not altogether sur-
prising to learn that it employs the facilities afforded
by steamers and the like to increase its range. Like
the pig, and still more like the rat and the cockroach,
it bids fair to follow traders over the whole world.
Baron Osten Sacken, whose unrivalled knowledge of
Diptcra makes his testimony conclusive, gives the
following particulars as to its immigration into fresh
countries.

Before 1875 Eristalis tenax was only known to occur
in the temperate regions of the old world. During
twrenty years of residence in North America Osten
Sacken never met with a specimen until on Nov. 5,
1875, he found one on a window at Cambridge, Mass.
Next year he observed several specimens at Newport,
R.I. The species has now completely overrun the
United States, and has occurred in Canada. The in-
vasion proceeded not from E. to \V., as might have
been expected, but from \V. to E. Eristalis tenax
has also been found in New Zealand, but not as yet
in Australia.1

Note. — An account of the aquatic larva and pupa
of Pericoma (see pp. 92, 160), by L. C. Miall and N.
Walker, will appear in Trans. Entom. Soc., 1895.

1 Details are given in Baron Osten Sacken's Oxen-born Bees.

CHAPTER III

AQUATIC HYMENOPTERA

DOWN to the year 1862 the large orders of Hy-
menoptera and Orthoptera had not been ascertained
to contain a single aquatic species.1 Since that date
the Orthoptera have been found to include some
semi-aquatic Grasshoppers native to Ceylon and Java
(genus Scelymena). The Hymenoptera have proved
to contain quite a number of forms parasitic upon
aquatic Insects. Amongst these are two minute
Insects, described by Sir John Lubbock in the paper
cited above. Polynema natans2 was found swimming
in pond-water by means of its wings. It lived several
hours under water without sign of inconvenience,
though provided with air-tubes, and breathing
apparently through spiracles in the usual manner.3

1 Lubbock, Linn. Trans., Vol. XXIV. p. 135 (1863).

2 Westwood (Linn. Trans., 2d Ser. Vol. I. p. 584) thinks that
Sir John Lubbock's Polynema should be placed in the genus
Anaphes.

3 This is Lubbock's account. Ganin, in the paper cited
below, says that there is no trace of tracheal tubes in any
stage. The wings, he adds, are hollow sacs filled with blood,
and act as gills.

220 NATURAL HISTORY OF AQUATIC INSECTS CH.

An immersion of upwards of fourteen hours, how-
ever, caused insensibility and apparent death. Both
males and females enter the water. In the same pond
was found a second species of similar habits, referred
to a new genus, Prestwichia.

" Whereas Polynema natans swims with its wings
and uses its legs apparently only for walking, the

FIG. 78. — Polynema (Anaphes) natans, side-view of female, X 30. From Lubbock.

present species (Prestwichia) when under water, holds
its wings motionless, and uses its legs as oars.
Though they are neither flattened nor provided with
any well-developed fringe of setae, still they seem to
serve their purpose pretty well, and the motion of
this species is more rapid than that of the former.

in AQUATIC HYMENOPTERA 221

Both species are fond of creeping along the sides of
the vessel in which they are kept, or on the leaves
and stems of aquatic plants ; but very frequently they
quit their support and swim boldly out into the open
water. As the motion in Polynema natans is caused
by the wings, it might almost be called a flight ; owing
however to the density of the medium and partly to
the direction in which the wings act, the movement,
though not inelegant, is slow, and is rather a succes-
sion of jerks than a continuous progression. ... I
was unfortunately unable to ascertain whether they
could fly ; taking my opportunity when they were
out of the water, I teased several specimens of Poly-
nema natans with the point of a needle, but never
succeeded in making one take to its wings, at least
not in air. When walking on the water, however,
they sometimes started off suddenly, but always kept
close to the surface, so that it rather seemed as if
they were carried by some tiny gust of air. We
might almost wonder how an animal like this, with
no apparent weapons of defence and no great powers
of speed, could maintain itself in this world of com-
petition. Protected, however, in its early stages by
the victim which it is destroying, it is exposed to its
enemies but for a short period of its life; and if, like
many of its allies, the eggs of other Insects are the
prey which it seeks, speed may be of comparatively
little importance. However this may be, we find in
the two Insects now under consideration, no pecu-
liarities indicative of an aquatic life. Many water-
insects are more or less boat-shaped, and both in
form and position the legs are admirably adapted to

222 NATURAL HISTORY OF AQUATIC INSECTS CH.

serve as oars ; others again use the hinder part of
the body as a fish does its tail. Here we find no such
arrangements. In both cases the head is broad ; in
Prestwichia the hind legs have, indeed, some scattered
hairs, but not more so than its terrestrial allies ; nor in
Polynema do the wings appear in any way modified
to adapt them to their new function. In conclusion,
if Polynema natans and Prestwichia aquatica had
been extinct species no palaeontologist would have
suspected that they were aquatic ; in the present
state of our knowledge there is nothing in their
structure which would have suggested such an idea." l

Both Polynema and Prestwichia are minute, about
a millimetre long.

Ganin l has traced the development of Polynema
(Anaphes) natans or some closely allied species.
The eggs are laid and the early stages passed inside
the eggs of a common Dragon-fly, Calopteryx virgo.
The female Calopteryx lays her eggs in the cellular
tissue of the leaves of the white water-lily (Nymphaea),
and it is to discover and penetrate them that the
Polynema enters the water. As a rule one egg is
laid by the parasite in one egg of the Dragon-fly. If
more than one is laid in the same egg, only one
attains complete development. The egg of Polynema
is flask-shaped, having a short and slender neck at
that end where the tail of the larva will ultimately
appear. The young larva is motionless and of very
simple form. It is enclosed within a transparent
cuticle, which again is enclosed by the egg of the

1 Lubbock, loc. cit.

2 Zeits. /. wiss. ZooL, Bd. XIX. p. 417 (1869).

in AQUATIC HYMENOPTERA 223

Dragon-fly. The larva before long casts its envelope
and appears as a segmented worm-like animal with
peculiar provisional organs and especially with a pair
of stout hooks at the head end. Rudiments of the
permanent appendages then begin to appear. When
all the contents of the Dragon-fly's egg have been
devoured, which happens in a few days from the time
of penetration, the Polynema pupates. The larval
appendages dwindle or are cast off by moulting. The
last segment enlarges greatly and is converted into
the abdomen of the imago. The fly emerges after a
pupal stage of ten or twelve days.

We have scanty information respecting a number
of aquatic Insects, which, like the terrestrial Ichneu-
mons, lay their eggs in the bodies of living Insects.
The larvae hatch out and devour the bodies of their
hosts little by little, delaying fatal injury until the
parasite is full grown. One such form, Agriotypus
preys upon Caddis-worms. The female was long
ago observed to enter the water, and later on Siebold
reared it from cases of Caddis-worms. The history
of the parasite has recently been more fully explored
by Klapalek,1 so well known by his researches into
the life-history of the Trichoptera. Klapalek finds
that in Bohemia Agriotypus commonly attacks the
cases of a Caddis-worm known by the name of Silo
pallipes. On warm days in April the Agriotypi may
be seen swarming like Ants about the banks of the
brooks, and also flying above the water. The
females descend stems and grasses into the water,
and creep under stones on the bed of the stream
1 Ent. Month. Mag. 1889, p. 339.

224 NATURAL HISTORY OF AQUATIC INSECTS CH.

-s

FIG. 79. — Agriotypus armntus. A, imago ; B, pupa ; D, larva ; E, Agriotypised case
fjf Silo ; G, section of ditto, showing r/1, fore operculum ; wl, hind ditto of case ;
5, remains of Silo larva ; fig, pupa of Agriotypus ; e, remains of cast larval skin
of ditto. From Klapalek'.

in AQUATIC HYMENOPTERA 225

in search of victims. The larva of the parasite
spends its whole life under water and inside the case
of a Caddis-worm. Its host is not mortally injured until
it has prepared for pupation. Like a healthy Caddis-
worm, it makes its case fast and closes it up. Then
the Agriotypus larva devours it, and crams the
remains into the hinder part of the case. It then
proceeds to moor the case by a long band, which it
secretes from its own salivary glands, and which is
the external indication of an agriotypised Caddis
Within the case it spins its cocoon, changes to a
pupa about September, and winters before emerging
as a winged fly. Klapalek believes that this aquatic
Ichneumon is very common.

CHAPTER IV

AQUATIC CATERPILLARS

AMONG the commonest of Moths are some small,
delicately marked species which abound in the
neighbourhood of ponds, lakes and marshes. The
patterns on their wings have suggested the popular
name of China Marks. One of the species so des-
ignated has in its first stage the habits of an ordinary
caterpillar. It feeds upon the Elder, and is common
in gardens. Other species feed upon the leaves of
water plants, and are in various degrees adapted to
an aquatic life. We owe to Reaumur the first
mention of an aquatic caterpillar, that of Hydro-
campa nymphseata, from which is bred the Brown
China Marks Moth. He visited, about the middle
of June, a pond covered with the floating leaves of
Potamogeton natans, and found that the plants
harboured numerous larvae, which fed upon the leaves,
and also used them as a means of concealment. In
some cases a small bit of the leaf was bitten out to
an oval shape, and fastened by silk threads to another
part of the leaf, forming a flattish lens-shaped prom-
inence. In most cases, however, both leaves were

CH. IV

AQUATIC CATERPILLARS

227

bitten through, and the two pieces, quite similar in
shape and size, fastened together with silk, formed a
movable sheath which protected the body of the larva,
as it crept about in search of food.

The larva itself is about an inch long when full
grown. The body is thickish towards the middle,
but narrowed at the two ends. It appears smooth to
the eye. Its colour is nearly white, but with a tinge
of yellow ; the upper surface of the thorax is darker.
The head, which is small and half-retractile, is dark
brown or nearly black. Like most land caterpillars,

FIG. 80. — Larva of Hydrocampa nymphseata, side-view.

it is provided with sixteen legs, the last pair being
very short. The spiracles or breathing pores have
the usual number and arrangement. Although the
larva is constantly submerged, it breathes air like
other caterpillars ; Reaumur found that when it was
oiled, so as to cut off the supply of air through
the spiracles, it died in less than a quarter of an
hour. If the larval case is opened under water, it is
found to be full of air, and the larva perfectly dry
within it. The larva is indeed incapable of being
wetted. A multitude of very fine and close-set
prominences cover its whole body, and between

Q 2

228 NATURAL HISTORY OF AQUATIC INSECTS CH.

these the surface-film is unable to penetrate. The
case is not completely closed. An opening is ne-
cessary in order that the larva may push out its head
and the fore-part of its body for the purpose of
feeding or locomotion. We should be inclined to sup-
pose that, on these occasions, the sheath would fill with
water, but this does not happen. The edges of the
opening are elastic, so that they open only under
pressure, and then to just such an extent as allows
the body of the Insect to protrude. The unoccupied
cleft is so minute that the surface-film of water is
unable to penetrate. Reaumur describes the opera-
tion of making a sheath. The larva lays hold of the
under side of the leaf, and gnaws a piece of it into
the desired shape. Then it carries the piece in its
mouth, and brings it against another leaf in such
a position that the two under-surfaces are in contact.
The leaves are a little concave beneath, and thus,
when brought together, they do not lie flat, but enclose
a lens-shaped cavity. The bits of leaf are then fixed
by threads of silk, and the second leaf is cut to the
same size and shape as the first. The largest sheath
which I have seen was if in. by f in. The larva
is commonly found on Potamogeton early in the
summer, but later on, when the leaves of the
Water-lily appear, it seems to prefer these both
for food and for the manufacture of its sheaths. It
is often found upon another aquatic plant, viz. the
Water Starwort (Callitriche verna). Before pupa-
tion, the Insect creeps up the stem of a tall water-
plant, and attaches its sheath to this well out of the
water. The sheath is now lined with a cocoon of

iv AQUATIC CATERPILLARS 229

close-woven silk.1 The pupa resembles that of a small
terrestrial Moth. It breathes by three pairs of well-
developed spiracles, carried on prominent warts which
stand out from the sides of the thorax. There
is a short ventral projection of unknown use. The
cmereence of the Moth has not been described.

o

Reaumur discovered the eggs of the Moth, which
are laid on the under side of the leaves, near the
edge, and enclosed, as is so often the case with
aquatic Insects, in a transparent jelly. From forty
to one hundred eggs are arranged with great regu-
larity in one patch, and according to Reaumur, the
female Moth protects the mass by glueing a piece
of leaf over the whole.2 The larvae hatch out in July
or August, and at once begin to make their own
sheaths.

Miiller has added some interesting facts to
Reaumur's account.3 He tells us that the young
larva, which is about rS mm. long, is completely
wetted with water. Its spiracles are rudimentary,
and do not communicate with special air-tubes.
The surface of the body has at first no pointed
prominences, such as those which at a later time
prevent the skin from being wetted. It must depend
for its supply of air upon the oxygen dissolved in
the water, though it has no gills or branchial organs
of any kind. The sheath which it makes for its own

1 Reaumur says that the cocoon is attached to a submerged
leaf, but this does not agree with what I have observed.

2 I have found the eggs completely exposed.

3 " Beob. an im Wasser lebenden Schmetterlingsraupen,"
von G, W, Muller, Zool.Jahrb. VI. p. 617.

230 NATURAL HISTORY OF AQUATIC INSECTS CH.

protection is wetted, full of water, and fixed. The
larva at this time excavates the leaves of Potamogeton,
scooping out the soft green substance for its nourish-
ment. It moults several times, without any import-
ant change in its form or habits. The cold season
comes round, and there is still no change, except
that the larva feeds less. The following April it
begins to feed actively again, and makes a fresh
sheath, but still its mode of respiration remains
unchanged. In May and June, it for the first time
begins to make sheaths filled with air, and its cuticle
is now provided with long conical prominences, in-
terspersed with smaller ones, which act like velvet,
and prevent the surface of the body from being wetted.
The spiracles now open for the first time, and the
larva breathes the enclosed air. After this it moults
once or twice more before undergoing pupation.

Buckler1 has related the history of Hydrocampa
stagnata. The larvae feed on the Bur-reed (Spar-
ganium), riddling the pith with holes, and biting away
the sides of the leaves. They hibernate in the cold
months, begin to feed again in April, and emerge as
Moths in the height of summer. The full-grown
larva is -J in. long, and tapers both ways from the
third segment. The pupal sheath is made of bits of
leaf woven together, and moored to a floating leaf.
It lies horizontally, either partly or entirely submerged.
Sheaths are also attached to submerged parts of the
plant. Inside the sheath is a cocoon of white silk,
quite dry and filled with air, which closely invests
the pupa.

1 Entom. Monthly Mag., Vol. XIV. (1877-8), pp. 47-103.

iv AQUATIC CATERPILLARS 231

The Duckweed, so common on stagnant water, har-
bours another species of aquatic caterpillar, Cataclysta
lemnae, which produces the Moth called the Small
China Marks. The larva of this species is smaller
than that of nympJiceata, and of variegated brown
colour. The top of the thorax is darker, the head
small and yellowish, and usually concealed in the pro-
thorax, which is horny and shining. The spiracles
are very minute. This larva makes its sheath of
leaves of Duckweed, attached together by silk, or of
the hollow stem of some aquatic plant, closing in such
cases one of the open ends with leaves of Duckweed
or bits of larger leaves. It exhibits two distinct
larval stages, like those of nymphceata, an earlier
branchial stage, which lasts till the second moult, and
a later tracheate stage. The life-history is completed
in one season, and there appear to be two generations
in the year. The pupa may JDC distinguished from
that of nymphceata by its long antennary sheaths,
which reach to the end of the body. The female
Moth runs about on the surface of the water, and lays
eggs, three or four together, on the under side of the
leaves of Duckweed.

The species described above, though living in water
throughout their earlier stages, are nevertheless during
most of their lives air-breathers. An allied caterpil-
lar, Paraponyx stratiotata, possesses a more complete
adaptation to aquatic life, namely, branchial respira-
tion. Paraponyx was first described by De Geer, to
whom we are indebted for an excellent description.
The larva of this species frequents the spiny leaves
of Stratiotes, the Water-soldier, a plant which is a

232 NATURAL HISTORY OF AQUATIC INSECTS CH.

favourite resort of a number of Snails, Leeches,
Polyps, and other animals. Paraponyx also occurs
on Anacharis (Buckler).

De Geer found a Paraponyx larva walking on the
leaf of Stratiotes, under water. He placed it in a
vessel, and next day found that it had cut out a piece
of the leaf, and made a sheath for itself, bringing
the two concave sides together, as nymph&ata does
with Potamogeton. The larvae feel greedily on the
leaves of Stratiotes. The sheath allowed the water
to enter freely, and the larva was completely wetted.
In some cases, the bases of the leaves were merely
tied together with silk, and not cut. The larva
grows to upwards of an inch in length. It is of a
pale transparent or yellowish green. The head is
brown, deeply cut out behind, and capable of being
retracted into the prothorax. The antennas are
longer than in most caterpillars, and several-jointed.
Along the sides of the body are found transparent
branchial filaments supplied by special air-tubes.
Some of the filaments are single, others spring three
or four together from a common point. There are
six tufts on some segments, eight on others, not
counting the single filaments ; the prothorax has none.
The branchial filaments are passive, and unlike those
of the Ephemera, cannot be separately moved.

Though the branchial filaments are fixed to the
body, the body performs regular respiratory move-
ments, like those of a Caddis-worm or a Chironomus
larva. Buckler tells us that night and day, at in-
tervals of from one to three minutes, the larva rapidly
undulates its body up and down, the. hinder end, by

iv AQUATIC CATERPILLARS 233

which it holds on, being motionless. The movement
lasts for twenty seconds. At these times the filaments
are depressed and pointed forwards, instead of radiat-
ing as usual. All the segments, except the second,
third and last, carry black spiracles, those of the
second, third and fourth abdominal segments being
much larger than the rest. The spiracles are not
however open, and are functionless. De Geer imitated
Reaumur's experiment of oiling the stigmata. The
result was very different in the case of Paraponyx ;
one larva when immersed in olive oil lived eight days,
and spun against the side of the vessel.

The Insect passes the winter as a larva, and in
captivity at least, feeds the whole time. In June De
Geer observed that cocoons were formed between the
young leaves of Stratiotes, well below the surface of
the water. The cocoon is grey outside, but lined with
close white silk. A large opening is left in the outer
envelope of grey silk, whose function is uncertain.
De Geer thinks that it may serve for the escape of
the Moth, or else for allowing free access of water.
The pupa has three pairs of prominent spiracles. It
is submerged, but the cocoon which it inhabits is
filled with air. Pupae removed from the cocoon, and
placed in saucers of water, lived for several days5
but ultimately perished without producing Moths.
Others when taken out of the cocoon and placed in
air, perished quickly. Others again, when placed
with one side of their bodies in water and the other
side in the air, survived, and produced Moths. These
facts appear to indicate that the stigmata are actually
employed by the pupa for respiration. As the cocoon

234 NATURAL HISTORY OF AQUATIC INSECTS CH.

is completely submerged, the emerging Moth must
float up through the water. The Moths, which are
by no means uncommon, appear in June, July, or
August, and may be seen disporting themselves on
the surface of the water. They lay greenish eggs on
floating fragments of leaves, and the eggs hatch out
in eight days.

There is, as we have seen, some discrepancy in the
accounts of the pupation of Hydrocampa nymphseata.
Some observers have recorded that the pupa-case
is attached to submerged leaves, while others have
found it fixed to stems well out of the water. But all
agree that the pupa-case of Paraponyx is often to be
found attached below the surface, and it becomes a
question of some interest how the air-breathing pupa
can under these circumstances obtain the moderate
supply of gaseous air which it requires. A fact
contributed by Dr. Schmidt-Schwedt1 may furnish
the answer to the question. Schmidt-Schwedt found
that beneath the pupal cocoon of Paraponyx the
leaf of Stratiotes2 is pierced by a number of small
holes, which, as thin sections prove, pass into the
air-filled spaces which are so common in aquatic
plants. It may be that Paraponyx, like Donacia,
draws its supply of air, during the pupal stage,

1 Berl. Entom. Zeits., Bd. XXXI. p. 332 (1887). The
author does not believe that these openings in the leaf are
actually employed as a source of respirable air, but his reasons
are of a general kind, relating to the excess of oxygen in the
gas evolved from a living leaf, and not very convincing to
myself.

2 He calls it by mistake Pistia stratiotes, a very different,
plant.

IV

AQUATIC CATERPILLARS

235

from wounds made in the tissues of the leaf, but
the essential facts still require confirmation by other
observers.

Another aquatic caterpillar of which very little is
known, is Acentropus niveus. The larva has not
been described, but it is said to live on Potamogeton
pectinatus. The pupa makes a silken cocoon, and
the Moth emerges in July.

CHAPTER V

CADDIS-WORMS (TRICHOPTERA)

A CONSIDERABLE part of this chapter is translated
from one of the most interesting and characteristic of
Reaumur's memoirs. The original is to be found in
the third volume of his History of Insects^

This is one of many excellent memoirs by
Reaumur which we have occasion to cite. A short
account of the naturalist who did so much for the
study of Insects seems due to his memory.

Reaumur was born at Rochelle in 1683. He was
enjoined to study law, the profession of his father,
but science led him away, and as his means rendered
him independent of a lucrative profession, he became
a student of Mathematics and Physics. His first
publications treated of Geometry, and were so highly
valued that when only twenty-four years of age, he
was elected a member of the Academy of Sciences.
In 1 710 he undertook the direction of a vast and
important work, the first great treatise on technical
education. This was the Description dcs Arts et

1 The passages extracted from Reaumur are put in inverted
commas.

v CADDIS-WORMS 237

Metiers, published by the Academy with the help of
the French Government. It led Reaumur to make
a long series of discoveries of much practical and
scientific interest. We find him investio-atinsf the

o o

tension of cords, and proving that, contrary to the
belief of practical men, they are weakened by torsion
(twisting). He experimented on the ductility of
metals, the lustre of pearls and fishes, the growth of
shells, the colour of the turquoise, the crystalline form
of certain metals and the manufacture of steel, which he
is said to have introduced into France. Tinplate too,
which had always been imported from Germany, was
first made in France according to the method taught
by Reaumur. He endeavoured to reproduce the
porcelain of the Chinese, but failed to get the desired
product, though he discovered the kind of white
glass still known as Reaumur's porcelain. He brought
into practical use the ancient method of hatching
eggs by artificial heat, showed how to preserve eggs
from putrefaction by greasing the shell, and sug-
gested improvements in the hanging of carriages and
the fitting of axles. Among his countless researches
are treatises on the production of spiders' silk for
commercial purposes, on the purple of the ancients,
on the sealing of tubes by mercury, on gold in the
rivers of France, on the Torpedo and its electrical
organ, on the fossil shells of Touraine, and on the
phosphorescence of shells.

The thermometer by which Reaumur is best known
to the public was constructed on an original plan.
He divided the liquid contained in the bulb and the
tube up to the freezing point of water into a thousand

238 NATURAL HISTORY OF AQUATIC INSECTS CH.

equal parts. The degrees were marked so that each
corresponded to an increase of volume equal to one
of the thousand parts. It happened that the alcohol
(not pure alcohol) which he used to fill the thermo-
meter dilated ilnnr °f its volume during the rise of
temperature from the freezing to the boiling point of
water, and Reaumur was thus led to divide the cor-
responding part of the tube into eighty degrees.
Hence the scale which was long used, almost to the
exclusion of others, in continental countries.

Among the many physical inquiries pursued by
Reaumur are those which relate to the evaporation
of snow and ice during frost, and to the action of
freezing mixtures.

Natural History was a favourite study with
Reaumur throughout his life, and gradually took
almost entire possession of his time and energies.
Among his early researches are a description of the
organs of locomotion in Star-fishes, and experiments
to prove that the Cray-fish and Lobster have the .
power of reproducing lost limbs. In 1734 began the
publication of his magnificent History of Insects •,
which though incomplete, fills six quarto volumes
abundantly furnished with plates. The life-histories
of many Lepidoptera, Caddis-flies and Aphides, of
the larvae which prey on Aphides, of the Cicada, of
gall-making Insects, of various Diptera, of the Saw-
flies, of Bees and their economy, of Wasps, Ichneu-
mons, Ant-lions, Ephemerae, Dragon-flies, and lastly of
the Horse-tick, Hippobosca, occupy these elaborate and
beautiful memoirs. The patient and sagacious observa-
tions here recorded in languageof transparent clearness

v CADDIS-WORMS 239

have been the delight of generations of naturalists,
and will be so again when the rage for names and
lists has abated. The History of Insects includes no
Beetles or Orthoptera. These would no doubt have
been added in time, but for the accident which in 1757
caused Reaumur's death. He left a multitude of
portfolios filled with notes and observations.

In one great zoological controversy of that age
Reaumur took up a hesitating and upon the whole
an unfortunate position. The prevalent opinion
respecting corals was, that the polyps were flowers,
the soft cortex a plant, and the hard part mere
unorganised stone. The researches of Count
Marsigli (1711) were believed to give strong support
to this interpretation. Peyssonnel in 1727 communi-
cated to Reaumur, and through him to the Academy
of Sciences in Paris, observations which were based
upon long study of the Mediterranean corals. He
pointed out that the so-called flowers could extend
and contract their parts, that they persisted through-
out the year, and that they had the composition of
animal bodies. Reaumur produced Peyssonnel's
communication, but kept back the author's name,
subsequently giving as his reason that Peyssonnel
would have run risk of contempt as the propounder
of views so adventurous. Reaumur proceeded to
give his own views as to the nature of corals, which
were completely orthodox, and completely mistaken.
Many years later the discovery of the Hydra and
Plumatella by Trembley reopened the question, and
Reaumur now perceived his mistake. He encouraged
fresh inquiries, candidly accepted their results, which

240 NATURAL HISTORY OF AQUATIC INSECTS CH.

were altogether favourable to Peyssonnel, and pub-
lished in the preface to the sixth and last volume
of his memoirs a clear and eloquent exposition of
the animal nature of Zoophytes. This was in 1/42.
Peyssonnel never received due recognition of his
services to Natural History, and his views were
treated coldly until they were at length enforced with
greater knowledge by John Ellis, to whom fell the
rewards of the discover}-.

Reaumur spent much of his time at his country
residences, one in Saintonge, the other at Bercy near
Paris. He avoided public offices, and his fortune
rendered him careless about increase of wealth.
When the Regent Orleans granted him a pension of
12,000 livres, he asked that it should be allotted to
the Academy of Sciences for technical experiments.

Except for a quarrel with Buffon, which no longer
interests us, Reaumur's scientific career was happy
and tranquil. He was of kindly disposition, and a
favourite in all circles

Reaumur writes in a fluent and animated style.
He does not disdain personal incidents, illustrations
from every-day life, nor anything which may kindle
the reader's interest. Some have called him diffuse,
but his supposed diffuseness is only the leisurely man-
ner of a writer who addresses a wide circle of readers.
He is never in a hurry, never enters into dry and
technical discussions, never wastes his reader's time.
His speculations, even when antiquated, are worth
the attention of those who would enter into the
thoughts of a sagacious and most productive in-

vestigator.

v CADDIS-WORMS 241

I now begin the Natural History of the Caddis-
worms with some extracts from Reaumur.

" One of the objects," says Reaumur, " which makes
most impression upon the common run of travellers,
is the variety of dresses put on by the inhabitants of
different countries. We may well wonder that men
who have only one end in view, viz. to protect them-
selves against the cold, should make use of so many
different means of attaining their object. What a
difference there is between the dress of men in civil-
ised and in barbarous countries ! How great a
variety in the dress of the different civilised countries,
and of the different savage countries ! The changes
of fashion show clearly that the form of our dress is
not always founded upon use or convenience. Insects,
like men, form coverings for themselves out of very
various materials, but the covering of each species is
invariable. Nature has taught each what is best
suited for its own use. Insect larvae made use long
before us of skins, wool, cotton and silk ; but besides
these, they employ other materials which never enter
into our own clothing. Some of these larvae make
themselves coarse clothes, which might be compared
to those of savages ; others display a higher degree
of skill. We shall give some examples of both
kinds.

" Belon tells us that the old French name for these
aquatic larvae is charrces [carts, or anything dragged
along]. They are found in small streams and brooks,
in ponds and lakes, in a word in any piece of water
which has plants living in or around it. Vallisnieri
observed them carefully, and was the first to give an

R

242 NATURAL HISTORY OF AQUATIC INSECTS CH.

account of them in his " Gallery of Minerva." ] He
says that they feed upon leaves of celery, ranunculus,
and water-dock as well as upon other aquatic plants.
[Caddis-worms are usually vegetable feeders, but not
exclusively so. They devour larvae of Chironomus,
Ephemerae and other aquatic animals. Some are
altogether carnivorous, e.g. the Hydropsychidae.
When hungry, Caddis-worms devour one another.
They do not attack the end of the case, which can be
defended, but tear it open in the middle.]

" Among these aquatic larvae there are many
species which I have not learnt to distinguish. The
commonest of all is much larger than the case-forming
larvae which I have just mentioned as living upon
land.2 None of them are true caterpillars, but all
chancre to four-winged flies. We will consider the

o o

stages through which they pass, after we have ex-
amined their mode of life.

" The body of these larvae is lodged in a silken
tube, to the outside of which are fastened fragments
of different substances, selected for the purpose of
strengthening and defending it. The sheaths may
be quite irregular, rough and prickly, or smooth and
symmetrical. When the old sheath becomes too
narrow or too short, the larva makes a fresh one.3
Sometimes the new sheath differs more from the cast-
off one than our dress of to-day differs from that of our

1 Works, folio edition, Vol. I. p. 37.

2 Case-forming Lepidoptera.

3 It is much more common for the larva to enlarge its case
by adding to the wider end, and cutting off the old and useless
part.

v CADDIS-WORMS 243

grandfathers. But with these aquatic larvae it is not
a question of caprice or fashion ; their dress is care-
fully adapted to the actual needs of life. They
employ very different materials, and the kind of
material largely affects the dress which they put on.
They make use of whole or nearly whole leaves,
bits of leaves, and these of a great number of species,
or little sticks and straws. Others use seeds,
roots, grains of sand and gravel, or the shells of
water-snails and bivalves ; in short, all the materials
which can be found in water are employed by par-
ticular Caddis-worms. In some sheaths one only of
these materials is employed, and these are the most
neatly constructed. In other sheaths a number of
different materials are made use of, so that the larva
is dressed, so to speak, in rags and tatters, and its
covering is altogether shapeless.

" Every sheath is a hollow cylinder with an open-
ing at each end. The fore end, out of which the
head and the six limbs can be passed, is wide.
The hinder end is narrow, and closed by a circular
silken plate with a hole in it. [The hinder end
is often quite open. Some species make cases of
uniform width.]

" If the sheath is made of leaves, it may take a
flattish form, wide in proportion to its depth, but
there are few of this kind. [Limnophilus pellucidus.]
In general the shape is cylindrical. Some sheaths
are covered externally by bits of rushes or small
twigs glued together and arranged lengthwise. Some-
times the rushes are so artfully joined that they look
like a grooved cylinder without joints, but it is rare to

244 NATURAL HISTORY OF AQUATIC INSECTS CH.

find one in which some shred does not spoil the look
of the sheath by projecting beyond the rest. This
irregularity, as we shall presently see, is really
necessary to the complete suitability of the sheath. In
the cases of the larvae of Phryganeidae the bits of
leaves are arranged with great regularity in a spiral
manner. (See Fig. 82, I, p. 256.)

" A larva sometimes picks up a reed, split into two.
If its sheath is entirely made up of small bits, and
therefore wants firmness, it will fasten these bits of
reed to the outside, bringing them as close together
as possible. Some sheaths are made up of such
fragments of reed ; in others small straws, such as
those of Equisetum, are arranged transversely to the
axis of the body [Limnophilus rhombicus], so that
the cross-section shows a circle within a polygon, the
angles of the polygon being produced. Such cases
are naturally very rough, but they often exhibit a kind
of symmetry.

" Lastly there are sheaths which are built up, partly
longitudinally and partly transversely, the fragments
bcino; ill-arranged and wanting in symmetry. Here a

O O O J *

large irregular piece of wood has been fastened on,
there a bit of stone, or it may be a shell. Some are
entirely made of bits of shell, all the shells belonging
to one species. [This is often the case with Limno-
philus flavicornis.] I have seen some which are made
up of little bits of the shells of water-snails, and others
made of bivalve shells, the t\vo valves being still
united. Singularly enough, they are sometimes built
up of living animals. What should we think of a
savage who should cover his body, not with furs, but

v CADDIS-WORMS 245

with living musk-rats and moles ? This is what these
Caddis-worms do. The snails and small mussels are
still alive within their shells, though they are fastened
down to the sheath of the Caddis-worm so completely
that they cannot move on their own account.

" It is common enough for the Caddis-worms which
make their sheaths of fragments of shell to fasten
along each side a stick which passes beyond the
sheath at each end. [Limnophilus lunatus, Anabolia,
Halesus radiatus, &c.] These sticks may be as long
again as the rest of the sheath, and almost equal to
it in diameter. Such a sheath is like a house fixed
between two beams higher than itself. In other
instances there is only one such stick, and sometimes
the sticks thus added are short and stout.

" Most of the sheaths of Caddis-worms are so
heavy that they would make a trying load for the
Insect if it walked upon dry land. Living in the
water, however, the Caddis-worms can creep along
the bottom or go up and down the weeds with an
ease which shows they are not over-burdened. [There
are species which make heavy cases of gravel, and
never leave the bottom.] It is plain, when we come
to consider it, that a sheath which would be too
much for the strength of the Insect in air may be
easily carried about in water, if the mean specific
gravity of its different components is pretty nearly
equal to that of water. This is the reason for the
irregular and awkward pieces of wood which so often
spoil the symmetry of the sheath. The Caddis-worm,
though it cares little about the shape of the vegetable
fragments which it fastens to its sheath, is generally

246 NATURAL HISTORY OF AQUATIC INSECTS CH.

careful to select such as have a suitable specific
gravity.

" The larvae cannot swim, or can only swim badly,
and they usually creep about. In the act of creeping,
the head and fore-part of the body are thrust out of
the sheath, and the larva drags itself along by its six
legs. This operation is often greatly facilitated by
the buoyancy of the vegetable fragments attached to
the sheath. [Some young larvae can swim very
fairly. In spring and early summer a small Caddis-
worm, which is common in ponds, may be seen
to swim rather rapidly through the water with the
help of its long, fringed hind-legs. The movement
is a little jerky, but the larva, carrying its case, can
move through the water in any direction in a straight
line.]

" Caddis-worms have six hard and jointed legs.
They have no false abdominal feet, like those of
caterpillars. The head is protected by armour of a
brown colour. The first ring of the thorax is of the
same texture and colour, and bears a short pair of
legs. The second ring is also brown and hard, and
carries the second pair of legs ; the third ring is
yellowish with brown spots, and carries the third
pair. The rest of the body is made up of nine rings,
which are white, soft and transparent. On the first
of these nine abdominal rings are three fleshy
prominences more or less conspicuous in the different
species. One of these projects from the upper surface
of the ring, while the other two are lateral ; all can be
retracted at pleasure. They are hollow at the tip,
and when the Insect is taken out of the water, they

v CADDIS-WORMS 247

are kept moist by a liquid which flows from them.
Is it possible that these prominences help in respira-
tion ? "

Mr. T. H. Taylor has, at my request, examined
these retractile processes with care. His inquiry is
still incomplete, but has already yielded results of
interest. The processes are believed to be best
developed in the Phryganeidae, but as yet little has
been done to study the various families comparatively.
The processes in question are absent in those Caddis-
worms which construct no movable sheaths. The
median tubercle may be the most completely retractile
of the three ; l the lateral ones are armed at their
extremities with peculiar scales, notched at the
extremities.

The fact that these processes are absent in the
Caddis-worms without sheaths suggested that they
might possibly be used to regulate a flow of water
through the sheath. On supplying powdered carmine
to a living Caddis-worm, the presence of a current of
water, passing in at the head-end, is readily demon-
strated. A Caddis-worm, removed from its case,
performs undulating movements with its abdomen.
Other tube-dwelling larvae, such as those of Chiro-
nomus and Paraponyx, maintain a current of water
in contact with their bodies in a very similar way. In
order to study the movements of a Caddis-worm
under something like normal conditions, I proposed
to Mr. Taylor to supply a naked Caddis-worm with

1 This is the case with Phryganea grandis ; in some other
Caddis-worms the median tubercle is not more retractile than
the others.

248 NATURAL HISTORY OF AQUATIC INSECTS CH.

small plates of mica, in the hope that a transparent
case might be obtained. This expectation was
perfectly realised. The Caddis-worm made a shapely
and sufficiently transparent case of mica, and per-
mitted the following observations to be made, which
I quote from Mr. Taylor's notes :-

" When Phryganea grandis is at rest it is retracted
within its sheath, the tips of the legs alone projecting
from the opening. The front part of the animal as
far back as the third abdominal segment is arched,
and comes close up to the top of the sheath, while
the rest of the body reclines on the floor to which the
hooks at the tail end are firmly attached. The
animal is not however still, but at intervals waves its
body up and down, causing a stream of water to flow
through the case from the head end. The undulations
begin at the second abdominal segment, and pass
backwards to the tail end just as in a piece of string
which is fixed at one end and waved up and down at
the other. The rest of the body is kept steady so
that the stream of water has a free admission. This
seems to be effected by the lateral processes of the
first abdominal segment, which can be pressed against
the sides of the case and kept from slipping by the
spines which cover their points. The median tubercle
is different in form and structure from the lateral
ones, and can be protruded or retracted with great

ease.'

1 The trial succeeded most readily with Limnophilus rhom-
bicus and an undetermined Caddis-worm. Phryganea grandis
was very reluctant to cover its body with mica, and only did so
after long exposure.

v CADDIS-WORMS 249

Supposing that the lateral processes are mainly
used for securing the body to the inside of the
sheath, the use of the adjustable median process can
be guessed at. It is the third screw, so to speak,
which tightens the whole arrangement, and regulates
the distance of the body from its enclosing tube.
The thoracic legs, it is obvious, are otherwise employed
when the animal is creeping or feeding.

I have lately seen a passage of Dr. Schmidt-
Schwedt's,1 which shows that he had arrived at a
similar explanation of the processes in question.

It is singular that Reaumur should offer the im-
probable suggestion that these retractile processes are
respiratory organs, and then go on immediately to
describe the true respiratory organs of the larva.

" The rest of the abdominal rings are provided with
organs whose use it is not easy to describe. They
consist of a great many white and soft filaments, of
which there are two bunches in the upper half of each
ring. At times the Insect moves them about, or
spreads them like a plume. At other times they are
laid flat, and meet across the back. I am much
inclined to suppose," says Reaumur, " that these offer
some analogy with the gills of Fishes.'' [The tracheal
gills of Caddis-worms vary in their arrangement in

1 He says that the larva moves its abdomen up and down,
most probably to renew the water wrhich bathes its body, and
that the fringed sides, by increasing the breadth of the abdomen,
add to the effect. " The fleshy processes of the first abdominal
segment find their use here ; the body is kept by them more in
the middle of the sheath, and the flow of water on all sides of
it is thus greatly promoted." Zacharias' Tier-und Pflajizenivelt
des Susswassers. Bd. II. p. 97 (1891).

250 NATURAL HISTORY OF AQUATIC INSECTS CH.

the different species, but are usually wanting or im-
perfect in the thorax and the first and last abdominal
segments ; they are in several rows and either simple
or branched. Fine air-tubes lead from them to the
main air-trunks of the body.]

;<At the hinder end of the body is a pair of horny
hooks, which take firm hold of the under side -of the

B

FIG. Si. — Phryganea varia. A, Larva ; B, its case ; C, a tracheal gill.

sheath. Indeed, if we attempt to remove the larva
by force, it can almost be pulled in pieces before
it will let go ; to remove it without violence from
its sheath, it is best to split open the sheath.
[Anglers know of a readier method than this;
A pin, gently inserted into the narrow end of the

V CADDIS-WORMS 251

sheath, soon dislodges the larva.] On looking at
the under side of the head with a lens, we see that
the mouth-parts are tolerably like those of caterpillars
and other vegetable-feeding larvae. On each side of
the mouth is a pair of strong toothed mandibles, well
adapted for cutting and shaping the materials of
which the sheath is composed, or for dividing the
food. Further back is a pair of maxillae with three-
jointed palps, while the labium is represented by
three pointed triangles, which taper backwards and
are armed with teeth along the fore-edge. [The
labium is often of simple structure and ends in two
minute palps.]

" The silken lining which comes next to the body
of the larvse shows that they can spin, and it is not
difficult to see the thread, either with a lens or with
the naked eye, for it is much coarser than that of
most caterpillars. I believe that the thread issues
from the same place as in caterpillars. I have traced
it to the middle of the labium, where the spinneret of
a caterpillar is situated. However, I have since sus-
pected that the spinneret of the Caddis-worm is not
situated here, but further back, in front of the base of
the first pair of legs. Here is a soft curved promin-
ence directed towards the head, which looks like a
spinneret. I have not, however, seen any thread
attached to it. [The curved spine between the fore-
legs is not the spinneret, but an organ whose function
is as yet unknown. The true spinneret is close behind
the mouth. Reaumur does not say anything about
the silk-gland. As in Chironomus, Simulium, etc.,
the secretion of the salivary glands forms the silk.

252 NATURAL HISTORY OF AQUATIC INSECTS CH.

The glands arc enormously developed in Caddis-
worms, and may be three times as long as the
body.]

"If we remove a Caddis-worm from its sheath
without injuring either, and then lay the two side by
side, the Caddis-worm at once enters its case, head
foremost. It is not so foolish as some other kinds of
larvae which fail to recognise their own sheath when
they have once left it, and would rather make a new
one than put on a second time the covering which
has been stripped off them. The front opening is
the only one by which the Caddis -worm can enter.
As it enters head first, its position is now reversed,
but the sheath is sufficiently wide for the Caddis-
worm to turn round inside. I have seen a Caddis-
worm with its head and limbs projecting as usual
from the front opening, by which it had re-entered
head first the day before, after being forcibly ex-
tracted.

u Though Caddis-worms will return to their sheaths,
it is not because they are too lazy to make new ones,
\Yishing to see them at work, I took a Caddis-worm
in the beginning of April, stripped it of its case, and
put it into a glass dish with various bits of leaves soaked
in water. In less than an hour it had made a nc\v sheath
of fragments of leaves. It is true that the sheath

o

was rather shapeless, and seemed to be made of odd
pieces slightly fastened together. However, it served
to conceal the body completely, and the Insect could
drag it about wherever it went.

" I turned this Cadclis-worm out of its case a second
time, as I had not sufficiently observed its mode of

v CADDIS-WORMS 253

proceeding. This time I put it into a saucer of white
earthenware, half full of water, taking care to throw
in bits of straw, hay and wood. For nearly three-
quarters of an hour the larva crept about feeling the
sticks and straws without deciding how to make use
of them. They did not appear suitable to the work
in hand. Perhaps they were too light, for they were
not yet perfectly soaked in water. Then I broke up
the two sheaths from which the larva had been
successively expelled, and threw them into the water.
Some of the fragments floated while others sank to
the bottom. I threw in also bits of leaves, and it was
not long before the Caddis-worm had found what it
wanted. It fixed upon a fragment of leaf nearly as
long as itself, and much broader. The hinder part of
the body moved up and clown so that the bunches
of filaments carried upon it were seen to wave to and
fro. While this was going on, the head was busily
employed. The larva cut off bits of the leaf, and
passed its head to and fro between them, as if in the
act of spinning, and very shortly the pieces were
fastened together. Thus the work went on until in
a short time the sheath concealed the fore part of the
body. It gradually grew until the whole body was
covered, but so far the work was very rough, and
spaces could be seen between the fragments. The
sheath was moreover too large for the body, and
accordingly several pieces were slipped inside and
fastened down one after another. There happened
to be in the saucer a branch of a water plant with
nearly cylindrical leaves of about the dimensions of
a common pin. The Caddis-worm cut several bits

254 NATURAL HISTORY OF AQUATIC INSECTS CH.

out of these leaves, and fastened them to its sheath^
placing some of them transversely in the manner
already described. When the outside of the sheath
had taken the desired form, the larva worked at the
inside, spinning a silken lining of a substantial kind.
I have since seen several larvae of the same species
at work, making new sheaths, lengthening or
strengthening the old ones, adding new fragments
cither to increase or diminish its weight, and all that
I have been able to see was merely a repetition or an
adaptation of the operations just described. [The
cases of small stones or sand are the hardest to make.
A Caddis-worm will complete one of these in five or
six hours.]

"In some species (such as Phryganca grandis) the
sheath appears to be rolled spirally round, like a
ribbon wound upon a stick. I have seen bits of oak
leaf arranged in this way along some very large sheaths
which I found in a pond in the Bois de Boulogne.
Some of these were only covered for part of their
length by bits of broad leaves, the rest being occupied
by narrow strips arranged side by side so as to form
a spiral band extending to the higher end of the
sheath. The larvae which inhabit these last sheaths
have two parallel curved bands on the front of the
head.

" The larvae of a very small species [Triaenodes
bicolor] are also covered by a spiral band, and look
as if a green ribbon had been bound round them from
head to foot. The dress of the larvae is like our own
in one respect ; it is much handsomer when it is new.
The colour changes in time from a beautiful green to

v CADDIS-WORMS 255

a dirty brown. The spiral in this last species consists
of a vast number of pieces, which increase in size by
imperceptible gradation. We have often to use a lens
to see that the band is not perfectly continuous. The
leaves of which it is is composed are usually leaves of
duckweed cut square. I believe that the fly of this
species carries its wings crossed one upon another
and laid horizontally.

" All Caddis-worms find it difficult to live in water
which is contaminated, or in small vessels. They
will last longer out of the water than in a very scanty
supply or in foul water. M. Baron sent me several
larvae of a common species from Lucon to Paris by
post. When they arrived they had been five or six
days out of water ; however, they were still alive and
I saw them walk."

The Caddises exhibit something of that versatility
which characterise the whole class of Insects. Thus
we have species with moveable cases (the great
majority) ; 'species with fixed cases or retreats which
the larva quits at pleasure ; forms adapted to stagnant
water, others adapted to life in torrents. Some marine
species will be noticed in Chapter XI I. Lastly, there
are terrestrial forms. The iarvse of Enoicyla -live in
moss at the foot of trees, often far from water. They
have no tracheal gills. The female flies are practically
wingless, an indication in other orders of Insects of
indiscriminate feeding and a plentiful supply of food.
E. pusilla occurs in England, and other species are
known.1

The variety of the cases manufactured by Caddis-
1 McLachlan, Triclioptcra of the European Fauna.

256 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. 82. — Cases of various Caddis-worms, i. Phryganea gr.mdis : 2, Limnophilus
rhombicus ; 3, 4, 5, L. flavicornis ; 6, L. pellucidus ; 7, Anabolia nervosa ; 8,
Limnophilus lunatus ; 9, Stenophylax nigricornis (larval case) ; 10, do. (pupal
case); n, Hydroptila MacLachlani ; 12, 13, Molanna angustata ; 14, Serico-
stoma ; 15, S. pedemontanum, anterior closing membrane; 16, posterior ditto;
17, Micropterna nycterobia, posterior ditto. 1-8, and 14 are from MacLachlan,
fi. G0ssi/>, 1868, the rest from Klapalek.

v CADDIS-WORMS 257

worms is very great. A number of forms, characteristic
of particular species, genera, and families are here
noted. The information given is for the most part
taken from McLachlan's Trichoptera of tJie European
Fauna.

Bits of leaves or stems placed side by side so as to form a
spiral band, passing many times round the case, which is open
at both ends, tapering when small, parallel-sided later. Still
water. Fam. P /try garni dee. Tricenodes forms a similar but
more tapering case.

Case of entire or nearly entire leaves of trees, flattish and
irregular. Still water. Glyphotcelius.

Case of fibres or sticks placed transversely, sometimes
with shells added. Still water. Limnophilus rhombicus.

Various materials, but especially shells of water mollusks.
Still water. Limnophilus flavicornis.

Case of sand, bits of grass, etc., to which long pieces of wood
are added. Still water. Limnophilus lunatus.

Curved cylindrical tube of fine sand. Still water. Limno-
philus vittatus. When the tube, as in this case, is curved, the
convex side is next to the back of the larva.

Case cylindrical, of bits of dead leaves, joined by their edges ;
closed with stones before pupation. Shallow running water.
Limnophilus auricula.

At first of vegetable fragments arranged obliquely to form a
cylindrical tube, afterwards of stones only. Pupal case shorter,
and more conical, closed with stones, and attached a number
together to stones. Still water. Limnophilus griseus.

Case of sand, cylindrical, scarcely curved, mouth-end oblique.
Pupal case shorter, attached a number together by the mouth-
end. Running water. Limnophilus extricatus.

Case of small stones and vegetable matter. Long twigs
attached to one end. Mouth-end oblique. Running water.
Anabolia.

Case of small stones, fixed or lodged in the bed of a stream.
Stenophylax.

Case of small stones. In swift, clear streams. Micropterna
(see Fig. 82. 17.)

S

258 NATURAL HISTORY OF AQUATIC INSECTS CH.

Case of vegetable fragments arranged longitudinally or
obliquely. Often with a long piece of wood at one end. Pupal
case closed with stones. Running water. Halesus radiatus.

Case curved and tapering behind, of stones and vegetable
matter, the stones at the tail-end. Swift streams ; the larva
attaches its case in the strongest currents. Halesus auricollis.

Case of small stones or sand, cylindrical. Streams. Seri-
costoma.

Case flattened, often with larger stones attached at the sides.
Rivers and streams. Goera.

Case quadrangular, of vegetable matter. Rivers and canals.
Brachycentrus.

Cases slender, straight, tapering tubes of fine sand, attached
a number together to the stones of streams. Oligoplectrum.

Case of sand, spirally wound like a snail-shell. S. Europe,
America, £c., &c. In still water. Helicopsyche.

Case of fine sand, curved, cylindrical. Still or running
water. Fam. Leptoceridc?.

Case of sand overspread by a sandy shield which projects
over the head of the larva. De Geer calls this a " Tortoise-
shell case." Still water. Molanna.

Case of sand, slightly curved, cylindrical. Tail-end closed by
a blackish membrane with a central slit. Before pupation the
mouth is closed by a single stone. Clear, sub-alpine streams.
Odontocerum albicorne.

Straight or slightly curved, cylindrical cases of fine sand ;
tail-end cut off before pupation. Usually in rivers and streams.
Leptocerus.

Case of sand, with long twigs attached. Slow streams, canals
and tanks. The larva has the legs, especially those of the third
pair, unusually long. Mystacides.

Case much curved, of fine sand or mud lined with silk. In
the spray of waterfalls. Adicella.

Case of hardened silk. Standing and running water. Setodes.

Cases fixed, of stones, occasionally inhabited by several larvae.
Still or running water. Some Hydropsy chides.

Larva loosely covered by silken threads, forming a snare.
Pupa in solid stony case. Clear rapid streams. Plectrocnemia.
Larva covered by vegetable debris, &c., until pupation. Rapid
streams. Polycentropus.

v CADDIS-WORMS 259

Larva makes silken channels on the surface of stones, which
become covered with slime, but prepares firmer cases before
pupation. Streams and springs. Tinodes.

Cases of small stones. Pupa enclosed in a brown cocoon,
instead of lying free within the case. Torrents. Rhyaco-
philidcE.

Cases of silk, movable, with foreign bodies attached, resem-
bling small seeds, slit at each end, at either of which the larva
can protrude its head. Sometimes moored by threads. Fixed by
strong threads before pupation. Standing or running water.
Fam. Hydroptilidce.

I now resume Reaumur's account. " It is not only
in the manufacture of their sheaths that these larvae
exhibit industry. They have subsequently to be
transformed into pupae in order to emerge as winged
Insects. The pupa is no better fitted to defend itself
against the attacks of its enemies than the chrysalis
of a Moth. Carnivorous Insects abound in the water
as on land or in the air. Accordingly the Caddis-
worm, before undergoing its transformation, provides
for its safety during the time when it can no longer
defend itself. It does not, however, quit its sheath,
but closes up the two ends with plates composed
of silk or of some other material. In its pupal con-
dition it continues to breathe air dissolved in water ;
to secure a constant supply of fresh water, the plate
which closes each end of the sheath is pierced with
holes like a strainer or grating, made of coarse threads
of silk. Thus the pupa is furnished with an ample
supply of fresh water, flowing through its sheath,
while it is sufficiently protected against carnivorous
Insects.

The pupae may be found in their cases in May
and June, but there are probably some which pass the

S 2

260 NATURAL HISTORY OF AQUATIC INSECTS CH.

winter in this state. In the month of March I have
found in the water cases with perforated ends, and on
opening them the pupa was seen. I placed some of
these pupae in water, where they lived several days
during which time they waved the hinder part of
their body to and fro [no doubt for purposes of
respiration].

" These pupae possess some peculiar features which
deserve a moment's attention. The body is white
with a citron tinge. On each side there is a narrow
black band extending over the four last rings. The
tail ends in a small fleshy fork. On the back are
bunches of white respiratory filaments, and also four
or five brown spots, which under the microscope are
seen to be furnished with spines directed backwards.
The four wings are plainly to be seen folded and
enclosed in the pupal skin, as in Lepidoptera. The
head is small in comparison with the body. On each
side is a great black eye, and a little below this a beak
like that of a parrot. Above this beak may be seen a
tuft of hairs which gives the whole head some re-
semblance to that of certain Birds, but on close
examination it can be made out that what at first
sight looked like a beak consists of two hooks [the
mandibles], which unite in front, and are crossed
at the points. These hooks lie beneath a fleshy
projection which carries the tuft of hairs. They are
different in shape from any possessed by the larvae.
The Caddis-fly has no such organs at all. These are
therefore parts peculiar to the pupa ; of what use can
they be, seeing that the pupa takes no food ? It
is probable, as Vallisnieri supposes, that they are

CADDIS-WORMS

261

useful only at the time when the fly is about to
emerge, and that they serve chiefly or solely to break

=dm B

c

D

FIG. 83. — Pupa of Hydropsychid Caddis-worm. A, case of small stones, seen from
attached side, with enclosed pupa ; B, larva within case, preparing to pupate ;
C, pupa ; D, pupal head, showing large hooks (mandibles).

down the gritty partition which closes the sheath.
Vallisnieri has seen in the end of June pupae which

262 NATURAL HISTORY OF AQUATIC INSECTS CH.

had broken down the perforated ends of their sheaths,
and had partly emerged. He has also seen such
pupae transformed into flies. I have had flies which
emerged in the beginning of April, having probably
passed the winter as pupae in their sheaths.

" Sometimes the sheath of the larva is longer than
necessary for it in the pupal stage. In such cases
perforated partitions are formed across the tube at
two points, but if the larval case is short, the per-
forated plates are fixed at the two extremities. I
have seen at times the perforated ends of the pupal
sheaths move in and out according as the Insect
sucked the water in or expelled it. These pupal
sheaths are generally attached to some fixed object,
and the sheath is fixed before the perforated plates
are added."

The pupal case is often a good deal shorter than
that in which the larva dwelt. Before pupation the
larva makes provision for its security during the rest-
ing stage. Various plans are employed by different
species. Some close the ends of the tube with the
silken sieves described by Reaumur ; others (Limno-
philus auricula, L. griscus, Halesus) with stones ;
Odontocerum with a single stone. Micropterna
sequax is said by Pictet l to make its case at first of
leaves ; as the larva grows it adds stones, until at
last the case is entirely stony. When pupation ap-
proaches, the case is lengthened with larger stones,
and closed at the wide end. Then the larva proceeds
to bury the narrow end vertically in the mud, until only
the large stones at the wide end are visible. To accom-
1 Recherches sur les Phryganides, p. 133.

V CADDIS-WORMS 263

plish this it turns round in its case, and works with its
head and legs projecting from the narrow end till
a hole is made of suitable size to receive the case,
after which it resumes its ordinary position. The
cases are commonly secured by silk threads to stones
or other objects.

De Geer describes the process of pupation and
the escape of the fly.1 Having found cases just
closed with silken sieves, he opened one, and found
within it a larva not yet transformed. It had, how-
ever, lost the power of moving its limbs, which
seemed paralysed, though the abdomen was capable
of vigorous movement.2 Next day the larval skin
was cast. The pupa now exhibited all the parts of
the imago, enveloped in a pupal skin, but not glued
down, as in a Lepidopterous chrysalis. The antennae
and legs were free, a point of importance in con-
nection with the emergence of the imago.

The fly does not emerge beneath the water, adds
De Geer. The pupa first extricates itself from its case
by the help of the hooks described above. Then it
climbs up the stones or aquatic plants until it gains
the air, attaches itself by the claws of its legs, and
throws off the pupal skin. The pupa, when liberated
from its case, swims and runs about with surprising
agility. Swimming is mainly effected by the pair of

1 Mem. sur les Insectes, u, 546, 516.

The mobility of the abdomen of the pupa is a point of
practical importance. The respiratory movements kept up by
the abdomen of the larva are equally necessary in the next
stage, and the pupa has the lateral fringe of hairs, the tracheal
gills, and every other requisite for maintaining a flow of water
through its case.

264 NATURAL HISTORY OF AQUATIC INSECTS CH

intermediate legs, which are broad and fringed with
hairs. Sometimes the fore legs are fringed also.
Fritz Miiller remarks that certain pupae of Caddis-
flies, which are not submerged, want these hairy
fringes, an interesting verification of the interpretation
commonly accepted.1 The pupa commonly swims
back downwards. Some of the smaller species do not
creep out of the water, but float to the surface, and
the fly in such cases escapes from the floating pupa
skin after the manner of a Gnat.

The dorsal surface of the pupal abdomen commonly
bears a number of hooks or groups of hooks, which
are no doubt used, like the very similar hooks of
many Lepidopterous and other pupae, for movement
within the cocoon.

Some mention should be made of the Caddis-
worms which do not carry about cases. These belong
to the families Hydropsychidae and Rhyacophilidae,
and are found usually in rapid streams. They have
retreats, often common to several larvae, which are
fixed and not portable, often consisting of stones
fastened by silk threads to one another and to a
larger stone. The larvae seek their food abroad,
lurking in crevices. They are sometimes, if not
always, carnivorous, preying upon Ephemerae, Simu-
lium larvae, and the like. In these larvae the lateral
fringe of hairs upon the abdomen and the retractile
processes of the first abdominal segments are wanting,
a confirmation of the explanation given of these struc-
tures, viz., that they aid in promoting a stream of
water through the confined tube of the ordinary

?, March ?95 1879.

v CADDIS-WORMS 265

Caddis-worms. The tracheal gills may be numerous
and unusually large, or altogether absent. The anal
feet are unusually long and powerful, two-jointed,
and armed with strong hooks for grappling. Just
before pupation, the larva constructs a pupal case
of stones firmly fastened together. Enclosed in this,
its respiratory requirements are much the same as
those of ordinary Caddis pupae. Accordingly tracheal
gills, if not present earlier, now appear, and the sides
of the abdomen are extended laterally by processes
which resemble, in their effect at least, the ordinary
lateral fringes.

I am indebted to Mr. T. H. Taylor for the follow-
ing interesting particulars respecting the mode of
life of the larva of Plectrocnemia :-

" Plectrocnemia finds its home in streams where
the water flows swiftly over a stony bed. If a stone
be lifted out, the under side is often found to be
covered with patches of mud from which brown larvae
emerge and begin to crawl over the surface. The
muddy particles are evidently held together by some
binding substance, and the whole forms the retreat
of the Caddis-worms, corresponding to the cases of
Phryganea. When a larva is placed in a vessel
of clear water, it at once begins to explore its
new quarters, and eventually selects a site for its
dwelling. This is made of silken threads secreted
by the large silk glands, and when completed the
structure consists of a tube considerably longer
and broader than its occupant, and open at both
ends. It is supported and strengthened by a mesh-
work of silken threads, which spread out for a

266 NATURAL HISTORY OF AQUATIC INSECTS CH.

considerable distance, and are attached to the sur-
rounding objects.

" Through this trans-
parent case it is easy to
observe the larva cling-
ing by its feet and anal
hooks to the threads,
and to watch its move-
ments. Like other Cad-

I

dis-worms, it sets up a
current of water through
the tube by swaying its
body up and down. The
respiratory stream enters
at the head end, and
flows out behind.

" Plectrocnemia pos-
sesses no gill-filaments,
and apparently the only
structures analogous to
gills are five finger-like
processes on the last
segment. These are cap-
able of being completely
withdrawn. The absence
of processes on the first
abdominal segment is
interesting. Assuming
that they serve to keep
the larva steady by pi cs-

sing against the side of the case it is essential that
they should act upon a rigid support. Here, however^

FIG. 84. — Larva of Plectrocnemia (Fam.
Hydropsychida;).

V CADDIS-WORMS 267

there is no firm resisting surface, and the processes
are not developed. Had they been true respiratory
organs, we may conjecture that they would have
been retained to supply the place of tracheal fila-
ments, which are altogether wanting.

" From time to time the larva turns round in its
case, and even leaves it for a short space. Generally,
however, it remains quiet inside, apparently on the
alert for prey. If a Chironomus or other small
aquatic larva approaches, it is almost certain to get
entangled in the network of silken threads. At once

o

the Caddis in its retreat perceives the presence of a
possible victim. The long hairs which cover the body
are possibly tactile, and reveal slight disturbances of
the silken network. The Plectrocnemia then pro-
ceeds warily to determine the cause of the disturbance.
Should the Chironomus be entangled near the middle
of the tube, the Caddis-worm does not hesitate to bite
its way through the side, and its jaws very soon quiet
the struggles of the prey.

" There is some resemblance between the snare of
the Plectrocnemia and the web of a Spider, but the
Plectrocnemia is effectually concealed by the mud
which clings to its retreat. In captivity it forms a
web which is free from foreign particles, and allows
all its manoeuvres to be observed."

"The Caddis-fly," says Reaumur, "exhibits some
marked peculiarities of its own. In a classification
of Flies, we might assign to them the name of papilio-
naceous, that is, Flies which at first sight resemble
Butterflies. These papilionaceous Flies have four
wings, of which only the two upper are visible when

268 NATURAL HISTORY OF AQUATIC INSECTS CH.

the Insect is at rest. In this position the upper wings
are nearly flat above, then they bend at an angle and
slope downwards. These two upper wings are mode-
rately transparent, but appear opaque when they
overlie the others. It is their opacity which causes
them to resemble Butterflies' wings, but on close
examination we see that they have none of the scales
which are characteristic of the wings of Lepidop-
tera [but are clothed with hairs]. The lower wings
are very transparent [and usually folded fan-wise] ;
they consist of a colourless or slightly bluish gauzy
membrane. The fly immediately after emergence,
and for some days subsequently, has a greenish tint ;
then it gradually turns dark. The six legs are long,
but the fly does not stand high above the surface on
which it rests. The antennae are very long, longer
than the body, gradually tapering, and of many
joints. The eyes are compound, like those of other
Flies and Butterflies. The mouth bears implements
very unlike those of the larva or pupa. There are
[no mandibles, but] four palps, two above and two
below, besides a very small proboscis [labium ? The
tarsi are five-jointed]."

The esters of Caddis-flies are laid in water or on

o o

water-plants, or on trees overhanging a stream, or
sometimes far from water.1 They are often of green
colour, and are laid many together in a mucilage
which swells out as soon as it comes in contact with
water, forming a cylindrical egg-rope, or in some
cases a flat disc. As soon as the larvae arc hatched,

} McLachlan, Entom, Month. Mag., Vol. XVI., p. 135 (1878).

v CADDIS-WORMS 269

they begin to make cases for themselves. We have
very little information as to the actual process of
egg-laying. Female Caddis-flies have been seen to
descend into the water, and are not uncommonly
found in a dirty condition, as if they had been
immersed in muddy water. The last joints of the
legs are often much dilated in the female, but not in
the male fly. This points to resting on the surface
of the water during egg-laying.

Entomologists are by no means agreed as to the
proper position of the Caddis-flies (Trichoptera) in
the system of Insects. Reaumur treated them as a
particular sort of Lepidoptera, calling them teignes
aquatiques. In favour of a close alliance with that
order might be alleged the four wings, clothed with
hairs, essentially agreeing with the scales of Lepi-
doptera, and sometimes not unlike them in form ;
the attitude of the wings and antennae of the resting
Insects, which is very like that of some small Moths ;
and the reduction of most of the mouth-parts of the
imago to functionless rudiments, as in many Moths.
The maxillary and labial palps are, however, much
more complete than in most Lepidoptera.1 Even the
mode of life of the larva, the aquatic habitat, and the
sheath or case can be matched among Lepidoptera.
The pupa of Micropteryx, a genus of Tineina (small
Moths), resembles a Caddis-pupa in the circumstance
that its appendages are not glued down, but distinct
and mobile. Moreover, the large pupal mandibles
are used to extricate the Moth from its subterranean

1 In Micropteryx and some other Lepidoptera the palps are as
well developed as in most Trichoptera.

270 NATURAL HISTORY OF AQUATIC INSECTS CH.

cocoon in a very Trichopterous manner.1 These
considerations are of various degrees of importance,
and some of them would not carry much weight with
the systematist. Those naturalists who, like the
writer, are compelled by their ignorance of systems
to take all their views on such matters from other
people, will acquiesce in Mr. McLachlan's opinion
that in any linear arrangement the Trichoptera and
Lepidoptera must not be widely separated.

It might be supposed that Caddis-worms, being
immersed in water as well as protected by a case,
would be free from the destructive attacks of Insect-
parasites. Mr. McLachlan has, however, mentioned a
Dipterous parasite which infests Limnophilus, one of
the Caddis-worms, and a short account of an
Ichneumon (Agriotypus), which preys upon another
Caddis-worm, is given in the section on Aquatic
Hymenoptera (p. 223).

Many of the contrivances of Insects are not con-
fined to a single order. The singular habit of laying
eggs in living larvae is shared by Hymenoptera
and Diptera ; Lepidopterous, Dipterous, Hymen-
opterous, and Coleopterous larvae mine in the thick-
ness of a leaf ; Hymenoptera, Diptera, Coleoptera,
and Rhynchota practise the art of making galls on
plants ; societies with a common dwelling exist
among Hymenoptera (Bees, Wasps, and Ants), and
also among Pseudo-neuroptera (White Ants) ; the
Dipterous Leptis vermileo imitates the pit-falls of the
Ant-lion (Neuroptera). Hence it is in accordance
with the great adaptability of the Insect-organisation
1 Dr. T. A. Chapman, Trans. Ent. Soc., 1893, p. 255.

v CADDIS-WORMS 271

that more than one order should furnish examples of
larvae which protect themselves by cases constructed
out of foreign particles. Not only the Caddis-worms,
but certain Beetles, Lepidoptera, Diptera, and
Neuroptera have learnt this art. The larvae of
Clythra construct leathery cases which they drag
about with them ; the head is protruded from one
end, and when full-fed, the Insect changes to a pupa
within its case. The larvae of Cryptocephalus form
cases too.1 Among Lepidopterous larvae the Coleo-
phoridae, the Tineidae, and the Psychidae are case-
builders, commonly carrying their cases about and
undergoing their transformations within them. Some
of these land caddises are described by Reaumur in
his most graphic manner in the memoir cited.
Chironomus and Tanypus (Diptera) construct fixed
cases out of vegetable fragments, and Lyonnet2
figures and describes what is probably a Chironomus,
of which he says that it weaves together the confervae
of the ditches, and forms out of them a sheath open
at both ends, and enlarged towards the middle. The
sheath is so flexible as to yield to all the undulating
movements of the inhabitant. The larva drags
itself and his sheath along by its mandibles and fore
legs. If the sheath becomes entangled, it quits it
and soon makes another. The Insect pupates within

1 Fowler, British Coleoptera, Vol. IV., pp. 286-7.

2 Rech. sur FAnat.,&>c.,dedijf. Especes tflnsectes,^. 179-183,
pi. xvii., Fig. 2. De Haan, the editor of this posthumous work
of Lyonnet, names (with some hesitation) the Insect a Tanypus,
but its habits are surely those of. Chironomus. See Additional
Notes, p. viii.

272 NATURAL HISTORY OF AQUATIC INSECTS

its sheath. Lastly, the larvae of Hemerobius and
Chrysopa (Neuroptera) cover their bodies with the
skins of Aphides.

It must not be supposed that little or no progress
has been made in the investigation of the Natural
History of the Trichoptera since the last century.
Much excellent work has been done of late years,
especially by Pictet,1 McLachlan,2 Fritz Muller,3
Klapalek,4 and Morton.5 We have now fair accounts
of the structure and mode of life of many species ;
the indispensable work of classification, which was
hardly even attempted by Reaumur, has made satis-
factory progress, and the special student is thereby
spared much confusion and loss of time. Unfortu-
nately it is hardly possible to give to a beginner any
intelligible account of this valuable mass of detailed
work. Those young naturalists who seek for work more
promising than records of parish distribution would
do well to master and extend the results achieved by
the writers named. Such work will bring them into
actual contact with the problems of nature.

1 Reck, pour servir a FHistoire et rAnatomie des Phrygan-
ides (1834).

2 Monograph of British Caddis-flies (1865) ; Revision and
Synopsis of the Trichoptera of the Eitropeaji Fauna (1874-80).

3 Zeits. / luiss. ZooL, XXXV. (iSSi) ; Zool. Anz., 1879;
Ent. Soc. Trans. , 1870.

4 Arch, der Naturw. Landesd. von Bohmen, VI. 5 ; VIII. 6.
(1888-93).

5 Ent. Month, Mag. 1882.

CHAPTER VI
SIALIDAE

THE ALDER FLY (SIALIS)

ONE of the very commonest of aquatic Insects is
the larva of Sialis lutarius. It is found in ponds and
slow streams. A muddy bottom is one requisite ;
hence Linnaeus' specific name, lutarius, which means
muddy. The larva creeps about actively on or in the
mud, seeking its food, which consists of other aquatic
animals. I find that in captivity it will eat Caddis-
worms and Ephemera-larvae, and authors mention
having found fragments of Insects in the alimentary
canal. The length of a full-grown larva is about an
inch, and the greatest width one sixth of an inch.
The body tapers a little towards the head, and very
gradually to the slender and long-drawn-out tail.
The six legs are strong and well suited to walking.
Each ends in a tarsus of two joints, bearing a pair
of pointed claws. The head is provided with a pair
of antennae, a pair of long, curved and pointed man-
dibles, which sufficiently indicate the carnivorous
habits of the larva, and the other mouth-parts usual
in biting Insects. From the sides of the first seven

T

274 NATURAL HISTORY OF AQUATIC INSECTS CH.

abdominal segments, or rather from the membranous
inter-segmental spaces just in front of them, stand
out as many pairs of long, tapering, five-jointed

B

FIG. 85. — Mouth-parts of larva of Sialis. A shows, in order from without inwards,
the antenna, mandible, maxilla, and labium. B, the mentum (base of labium).

appendages.1 These are fringed on opposite sides
with hairs in regular series, and each is traversed by a
sinuous, branching tracheal tube. A stream of blood,

1 Pictet (Ann. Sci. Nat. ZooL, 2* Serie, V. pi. 3) shows eight
pairs of four-jointed gills, and his figure has been copied by
other authors. Roesel has the right number, viz. seven, but his
figure is too small to show the number of joints.

VI

THE ALDER FLY

275

carrying along colourless corpuscles, can be seen by
the microscope to flow
at each pulsation of the
dorsal vessel into the
space which surrounds
the trachea. These ap-
pendages are tracheal
gills, very like those
of many other aquatic
larvae. There are no
tracheal gills to the
eighth and ninth seg-
ments. To the ninth
abdominal segment is
attached the long tail,
which appears to repre-
sent a tenth segment. It
is fringed, and contains
two tracheal tubes, much
resembling indeed two
ordinary tracheal gills
completely fused to-
gether. In the usual
position the tracheal
gills curve upwards and
backwards, and it is
only in the dead larva

•

that they lie confusedly
as in Fig. 86. It is
said in some books that
they can be employed
as fins for

swimming,

FIG. 86.— Larva of Sialis. The jointed re-
spiratory appendages of the abdomen
are in the living larva curved upwards
and backwards.

T 2

276 NATURAL HISTORY OF AQUATIC INSECTS CH.

but this is not the case. The larva creeps by means
of its six legs, sometimes aiding its progress by an
undulating, side-to-side movement of the abdomen.
In a confined space it may often be seen to move
its abdomen in vertical undulations, which cause a
swirl in the water, and no doubt promote respiration.

Several aquatic larvae, not at all closely related to
Sialis, have similar tracheal gills. Gyrinus and
Orectocheilus are examples taken from the Coleoptera.
One peculiarity of the gills of Sialis, viz. their division
into joints, is shared by a Dipterous larva, Dicranota.
It is worth notice that aquatic Insects, living side by
side under identical conditions, should show so great
a variety of respiratory organs. Some have tracheal
gills, not all of one kind ; others breathe by spiracles ;
others have no special respiratory organs at all.

Though the Sialis larva lives in water, and can
spend months without coming to the surface, as I
have found by observation of captives, it can survive
a rather prolonged exposure to air. Full-grown
larvae, taken out of the water before the time of
transformation has fully come, and placed on damp
earth, walk about till they find a convenient crevice,
and then bury themselves. They live and retain the
power of active movement for days when thus se-
cluded. This faculty of enduring exposure to dry
conditions is useful, as will shortly be seen, on two
occasions, which inevitably occur during the lifetime
of the Insect — viz., immediately after hatching, and
before pupation.

In May or June, when the larva is full-fed, it quits
the water and seeks a place where it can bury itself

vi THE ALDER FLY 277

in the earth. For this purpose it will often travel far
from its native pool, and I have lately found one
creeping on the surface of the ground six yards from
water, though the season was dry, and the soil com-
mon garden mould, not particularly retentive of
moisture. This larva had climbed up a concrete wall,
made its way through a thicket of Cotoneaster, and
reached an open flower-bed. When the Insect has
found a place to its mind, it enters the earth, ex-
cavates a little cell, casts the larval skin, and is trans-
formed into a pupa, which has the legs and wings free
from the body, though enclosed in special sheaths.
There are spines on the abdominal rings, as usual
in pupae which have to make their way to the surface
of the earth ready for the emergence of the imago.
After two or three weeks the fly emerges — a heavy
awkward Insect with black body, and four large and
coarse wings, which are clear but with conspicuous
black veins. The female fly is one third larger than
the male. At first sight the winged Insect might
be taken for a Caddis-fly, for the wings are rather
opaque, longer than the body, and form a sort of roof
sloping steeply away on either side. The body is
stouter and heavier than that of a Caddis-fly, the
antennae shorter in proportion, though still of good
length, and the mouth-parts much more complete.
The wings too differ considerably from those of the
Caddis-fly. A winged Sialis flies heavily, and is
easily caught. When threatened it tries to escape by
running rather than by taking wing. The female
lays patches of dark brown eggs on leaves, stones or
palings not far from the water. There may be several

278 NATURAL HISTORY OF AQUATIC INSECTS
hundred eggs in one cluster, cylindrical with rounded

f

ends, and closely packed together. From the free
end of each egg a small pointed and whitish pro-
jection is given off. When the larvae hatch out, they
travel to the water, where they have to spend about a
year before attaining the winged state. I have often
seen the fresh-hatched larvae wriggling out on leaves
many yards from the nearest stream or pond. How
they find out the way, and whether or not many
perish from taking a wrong direction, I do not know.
Anglers have called this the Alder Fly, no doubt
because it is often found on the alders which over-
hang streams. In some old books it is called May
Fly, a name which is better suited to this Insect than
to Ephemeridit, which emerge in various months of
the summer season

CHAPTER VII

PERLID.E

STONE-FLIES (PERLA)

IN streams, and especially in clear streams which
flow down from the hills along stony beds, the larvae
of Perla are plentiful during early summer. Lift any
good-sized stone, as large as the hand or larger, and
it is likely that two or three dusky objects, resembling
small Shrimps, will run along it to seek the side which
is turned away from the light. In the stream which
I have most diligently frequented two aquatic larvae
occur side by side on such stones. One is that
of Baetis (Ecdyurus) fluminum, an Ephemerid ; the
other that of a Perla. The Baetis is smaller than the
Perla, though a good inch in length when of full size,
and has three tails. The Perla is somewhat larger,
and has only two tails.

In June the larvae of Perla leave the water, and
shortly after the winged flies emerge. The stones
which border the brook are now strewn with larva-
skins, curiously like living larvae in shape, but dry
and empty. Along the back of the thorax is the

280 NATURAL HISTORY OF AQUATIC INSECTS CH.

gaping slit by which the fly escaped. The eyes,

legs, jaws, and wing-
sheaths, as well as the
rings of the body, are
so perfect that the
outer form of the
larva can be described
from such a cast skin.
Not only these exter-
nal organs, but also
the lining membrane
of the great air-tubes,
and that of the fore
part of the alimentary
canal, including the
armature of the giz-
zard, can be made out
in the cast skin, a
proof that these struc-
tures are formed by

FIG. 87. — Larva of Perla bipunctata
(bicaudata).

FIG. 88. — Under side of head
of larva of Perla bipunctata
(bicaudata).

folding of the outer integument into the respiratory
openings and the mouth.

vii STONE FLIES 281

It is easy to observe the escape of the fly. A full-
grown larva creeps out from the stream, sometimes to
a distance of several feet, and seeks a stone on which
it can securely fix its hooked feet. Then the old
skin begins to swell. Air is passed into the narrow
space between it and the new skin beneath, which
though still soft and flexible, is now complete in all
its parts. Before long the back of the thorax splits
lengthwise. First the new thoracic skin shows itself,
then the head and antennae appear, and the wings are
drawn out of their sheaths. The legs of the fly are
next freed, and when these have got a sufficient hold
of the ground, the abdomen and tail-filaments are
withdrawn. The newly emerged fly is still soft and
pale-coloured, but in a few hours or even, in some
species, in a few minutes, the organs of locomotion
become firm, and the Insect takes to flight.

In spite of its large wings the imago is a poor flier,
turning with difficulty, and showing little power of
avoiding an obstacle. It is easy to capture it with a
net or even with the hand. Like a Grouse, it whizzes
along with great effort and nearly in a straight line.
The habits of the winged Perla do not call for great
agility. The males and females mate on the
ground, near the place of emergence. The fertilised
eggs, which are oval and black, project from the
end of the abdomen of the female, being loosely
fastened together by a transparent skin. These are
dropped into the water, and the work of the winged
generation is at an end.

From these eggs small aquatic larvae issue, which
may be found plentifully in the same streams during

282 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. 89 — Perla bipunctata (bicaudata). Mouth-parts of do. ; egg of do.

vii STONE FLIES 283

late summer and autumn. They hide themselves
during winter, but reappear early in spring. From
this time they feed actively, capturing and devouring
all the weaker inhabitants of the stream, until the time
comes to quit the water. I have found in the stomach
as many as five empty heads of Chironomus larvse.
The jaws of the larva resemble those of a Beetle or
Cockroach. The mandibles are stout, and armed
with small pointed teeth ; both pairs of maxillae are
well developed, and provided with good-sized palps.

It would naturally be expected that a larva so
entirely aquatic as that of Perla would possess re-
spiratory organs of the nature of gills, and this is
actually the case. At the base of each leg, and just
above it, is a tuft of filaments, supplied by fine,
branching air-tubes. A little behind each of these
bunches of tracheal gills, and in the flexible skin
which unites the thoracic segments, or unites the last
thoracic segment to the first abdominal, is another
bunch. There are thus twelve thoracic gills (six
pairs) in the larva of Perla. Finally, just inside the
base of each tail-filament is another gill.1 It will be
noticed that, as in Crustacea, the gills are placed where
the movements of the appendages can either wave
them to and fro, or set up a swirl in the water. The
living larva rises and sinks continually (when not
travelling about) by a slight movement of the legs,
and this movement is probably of use in respiration.

In an Ephemera larva the tracheal gills are com-

1 Palmen, Morph. d. Tracheen-sy stems (1877). The descrip-
tion in the text applies to certain large species, such as Perla
bipunctata.

284 NATURAL HISTORY OF AQUATIC INSECTS

pletely cast at the last moult, when the spiracles open
for the first time and allow air to pass directly into the
tracheal tubes. In Perla, however, the tracheal gills
persist, though in a greatly reduced form, after the
spiracles have opened. In the winged fly the gills
can be discovered with a lens in exactly the same
places where they occur in the larva. This fact,
which was established by Palmen, was of great use
to him in refuting a view which had been extensively
adopted — viz., that spiracles are derived from pre-
viously existing tracheal gills. It was supposed that
the casting of the gill in all cases opened a way into
the tracheal system, which was previously closed. In
the winged Perla, however, minute tracheal gills, pro-
ject from every thoracic spiracle. They are probably
quite useless. The chief theoretical interest of Pal-
men's arguments is that they dispose of a theory
which implied that Insects were primitively aquatic,
and became adapted to aerial respiration by the con-
version of their gills into air-slits. On the contrary,
it is on all grounds probable that Insects were primi-
tively terrestrial, and that the various adaptations to
aquatic conditions are secondary. Where Insects
possess gills, these are not primitive organs pos-
sessed by the more typical Insects, but relatively late
acquisitions, developed as it were casually, and with a
marked relation to the peculiar needs of the family
or small group in question. No organs of an Insect are
more variable in position, number and structure than
the larval gills, and this is one proof that they were
not inherited from a common ancestor, but acquired
by many small groups independently of one another.

CHAPTER VIII

MAY-FLIES (EPHEMERIDyE) l

THE first useful account of the life-history of an
Ephemera that we meet in zoological literature is to
be found in Swammerdam's Biblia Natures. The
species investigated by him is now named Palingenia
longicauda. The larva is common in the large, slow
and muddy rivers of Holland. Swammerdam's his-
tory, a good deal abridged and annotated, is here
reproduced in English as an introduction to the May-
flies or Ephemerae. Some particulars respecting the
author, taken from Miall and Denny on the Cock-
roach, are prefixed.

Swammerdam's great posthumous work, the Biblia
Natures, contains about a dozen life-histories of In-
sects worked out in more or less detail. Of these the
May-fly (published during the author's lifetime, in
1675) is the most famous ; that on the Honey Bee the
most elaborate. Swammerdam was ten years younger
than Malpighi, and knew Malpighi's treatise on

1 The names used by modern naturalists are employed wher-
ever the species described by Swammerdam and Reaumur can
be clearly identified.

286 NATURAL HISTORY OF AQUATIC INSECTS CH.

the Silkworm — a not inconsiderable advantage. His
working-life as a naturalist comes within the ten years
between 1663 and 1673 ; and this short space of time
was darkened by anxiety about money, as well as by
the religious fanaticism, which in the end completely
extinguished his activity. The vast amount of highly-
finished work which he accomplished in these ten
years justifies Boerhaave's rather rhetorical account
of his industry. Unfortunately, Boerhaave, whom we
have to thank not only for a useful sketch of Swam-
merdam's life, but also for the preservation of most
of his writings, was only twelve years old when the
great naturalist died, and his account cannot be taken
as personal testimony. Swammerdam, he tells us
worked with a simple microscope and several powers.
His great skill lay in his dexterous use of scissors.
Sometimes he employed tools so fine as to require
whetting under the microscope. He was famous for his
inflated and injected preparations. As to his patience
it is enough to say that he would spend whole days
in clearing a single caterpillar. Boerhaave gives us a
picture of Swammerdam at work which the reader
does not soon forget. " His labours were superhuman.
Through the day he observed incessantly, and at
night he described and drew what he had seen. By
six o'clock in the morning in summer he began to
find enough light to enable him to trace the minutiae
of natural objects. He was hard at work till noon,
in full sunlight, and bareheaded, so as not to obstruct
the light ; and his head streamed with profuse sweat.
His eyes, by reason of the blaze of light and micro-
scopic toil, became so \veakened that he could not

vin MAY-FLIES 287

observe minute objects in the afternoon, though the
light was not less bright than in the morning, for his
eyes were weary, and could no longer perceive
readily."

Swammerdam's Life of an Ephemera follows :-
"The winged Palingenia (Fig. 91) is an Insect with
four wings, two small antennae, six feet, and two long,
hairy filaments, which stand straight out from the
hinder end of the body. It lives at most five hours in
the winged state. Every year it appears on the banks
of the Rhine, the Maas, the Waal, the Leek, and the
Yssel, appearing on the surface of the water about the
feast of St. Olof and St. John. The flies may be seen
in the air for three days together. Those which appear
on the first day die the same evening, and the same
thing happens on the second and third days. Then
a whole year has to elapse before they are seen again.
" When the female has emerged from the water
and cast off her skin, she passes the contents of
the double ovary into the water, but first she moves
to and fro on the surface of the water as if in sport,
and flits about with rapid exploring movements.
Immediately after the eggs are passed into the
water, they are fertilized by the male,1 which has
previously emerged and cast off a delicate mem-
braneous skin. The eggs, thus passed into the water
and fertilized, sink slowly, and are scattered over
the mud at the bottom of the stream. The form
of the eggs contributes to this mode of dispersal ;

1 This is a mistake. The eggs are fertilized in other
Ephemerae, and no doubt in Palingenia too, while still in the
body of the female.

288 NATURAL HISTORY OF AQUATIC INSECTS CH.

they are rounded discs, which sink at different rates ;
hence if a number of the eggs are gently lowered
into the water on the point of a knife, they are
seen to spread as they descend.1

" How long the eggs lie on the bottom of the
river, and how many days elapse before the larvae
emerge, is known only to God. Something might be
made out upon these points if any one would search
the bottom of the stream at frequent intervals, or
keep the eggs in a basin with water and mud.

" Some time after the descent of the eggs, a crowd
of minute worms, each with six legs, makes its ap-
pearance. These do not differ in shape from the
older larvae. Their growth is so slow that after a
year, viz., in the following June, they are only a third
of the length of the larvae then ready to enter the
winged state. At the end of another year the larvae
are twice as long, but three years are required .before
they attain their full size.2 Not only do these larvae
of different ages differ in size, but also in the degree
of development of the wings. The small larvae, a
year old, exhibit no trace of wings ; after two years
the wings are visible, and enclosed in special sheaths ;
at the end of the third year they are quite plain and
ready to burst forth.

1 If a handful of ivory counters are thrown into deep water,
some descend edgewise, and reach the bottom long before
others, which happen to take a horizontal position. Some such
difference of position and not of specific gravity, seems to be
referred to by our author.

2 Confirmation of this is desirable. Some Ephemeras spend
one, and some two years as larvae, but no other case of an
Ephemera remaining a larva for three years is on record.

viii MAY-FLIES 289

" These larvae are rarely, if ever, seen swimming in
the water. It is true that they can swim by a kind of
serpentine movement, bending the head now down,
now up, but they always keep to the banks out of
reach of the current. They can very rarely be seen
out of the mud, in which they make for themselves
long cylindrical and horizontal burrows.

" As the Bees in their own wonderful way make
homes out of wax, so the larvae of the Ephemera
excavate out of mud the tubes in which they dwell. If
they are taken out of their tubes, they can only creep
readily when the bottom is so flat as to support the
whole length of the body. Although they are ordin-
arily immersed in water and can swim, I have found
on taking a number of them out of their tubes, that
they at once fall on their backs as if paralyzed, and
are not able to right themselves. But within their
burrows they can move quickly backwards and
forwards. The same is true of various other larvae
which burrow in trees, fruits, leaves or galls. The
Cossus when taken from its hole in a tree covers its
whole body with a web, and supported by this, it is
able to make a new hole in the wood, but if left with-
out support, or some fixed object against which the
body can be pressed, it is quite unable to bore. The
larva of the Ephemera is so helpless outside its tube,
that if while swimming in the water it ceases to exert
itself, it falls at once to the bottom, and there lies upon
its back.

" As soon as the larvae escape from the egg, they set
about making their burrows, and these are gradually
increased in size as the larva grows. The All-wise

U

290 NATURAL HISTORY OF AQUATIC INSECTS CH.

Creator has provided them with limbs suitable to their
work. The fore limbs are fitted for digging, as in
the Mole or Mole-cricket. Besides their feet, the
larvae are also provided with jaws, each with two teeth,
like the pincers of Crabs, and these are well suited for
working in the mud. The larvae may be seen at work
if they are placed in water mixed with a little mud.
If the mud is not sufficient to cover the whole body,
they conceal first the head, then the body, and then
the tail, endeavouring all the time to complete their
dwelling. Anglers assure us that when the water of
the rivers sinks, the larvae work deeper and deeper
into the mud, rising again as the water rises. This, I
suppose, is indispensable for their breathing. They are
provided with many air-tubes for distributing the air
throughout the body. Access of air would be stopped
if they remained below after the water had risen. I
have often found that when taken out of their tubes and
placed on wet sand, they much prefer creeping out of
the water to burying themselves in the sand. The
reason of this may be that the warmth of the water is
injurious to them, as well as the want of mud.

" As to the food of these larvae, it would be hard
to say what it was without the help of dissection,
but by this means I have discovered that they feed
upon mud only. Whenever the body is opened, mud
is found in the stomach and intestine. In the same
way certain caterpillars feed upon the same substance
of which they make their homes.1

1 Like Earthworms and many other burrowing animals, the
larva of Palingenia probably feeds not upon the mud itself, but
upon vegetable fragments scattered through it. The food of

vni MAY-FLIES 291

" The full-fed larvae pass from their burrows into
the water, and thence into the air, but as no animal is
without its enemies, so these larvae, as soon as they
enter the water are pursued by Fishes, and when
they leave the water for the air, they are immediately
liable to be devoured by Birds. All anglers know
that the larvae make excellent bait for Fishes. When
they emerge from the water, they are often thick as
falling flakes of snow. Hence the Dutch proverb, Het
isser soo digt> als Haft (as thick as May-flies). At all
times of the year when the weather is favourable for
fishing, these larvae form an excellent bait. When the
waters are high, it is not easy to fetch them out of
the mud, and it becomes necessary for some one to
strip himself naked and go down after them. I have
sometimes sent a man into the water to procure a
supply of larvae for dissection. The larvae are tena-
cious of life, and live long on the hook, but when
taken out of their tubes and placed in water mixed
with earth, they do not live more than two days.
To keep them alive, they should be placed in wet
sand or mud. In this way I have seen the older
larvae live four days, the younger ones eight days ;
but when completely submerged, they cannot long
be kept alive.1 If it is desired to send live larvae to

the larvae of different species varies considerably. Some feed
upon living Crustacea, and have a " weel " in their throat to
prevent the escape of the victims. I have seen such a structure,
not unlike that of Corethra, very plainly in a cast skin of an
Ephemera larva, which I was unfortunately unable to deter-
mine.

1 It seems strange that a larva which is habitually submerged
should perish when submerged in captivity. The explanation

U 2

292 NATURAL HISTORY OF AQUATIC INSECTS CH.

a distance, there is no better way than tying to-
gether stems of the great reed, and causing the larvae
to creep into them.

" The body of the larva is divided into fourteen
distinct segments. The first forms the head, the

o

three next the thorax, and the
remaining ten the abdomen.
Upon the head are seen the
eyes, and just beneath them the
delicate antennae, divided each
into five joints ; below these are
the mandibles, and again below
these the hairy, membranous
maxillae, which are like those of
Crabs or Shrimps. The first
pair of thoracic limbs are adapt-
ed by their shape for digging.
Their most powerful action is
outward, and in this way they
are able to throw out the mud
as a Mole throws out earth.
Each of the fore-feet is formed
of four joints and a single claw.
The second thoracic segment is
protected above and below by

a shield-like plate ; it bears a pair of limbs, each
consisting of five joints and a claw, and on the

probably is this. Under natural conditions the larvae of Palin-
genia inhabit slowly running water, which not only supplies
dissolved air for respiration, but brings down the food upon
which they subsist. The captives were no doubt kept in vessels
of standing water. Few aquatic animals can long survive the
change from running to standing water.

FIG. 90. — Larva of Palingenia
longicauda, male. From
Swammerdam, Biblia Na-

vin MAY-FLIES 293

sides of the segment [in advanced larvae] are the
sheaths of the wings. Air-tubes ramify upon the
sheaths, and shortly before a moult, the wings may
be seen within them, folded up in a wonderful
and beautiful manner. The third thoracic segment
bears the second pair of wings, which are much
smaller than the first, and also a pair of five-jointed
legs. The first segment of the abdomen is smooth
and provided with no appendages, the next six
segments are furnished with gills, which are in
incessant movement. In the Lobster, Crab, and
Sepia,1 which in many respects approach the structure
of Insects, we find the gills formed and arranged
in nearly the same manner, though with this difference
that in Crabs and Lobsters the gills are enclosed by
a hard carapace, while both in them and in the Sepia
they are more concealed than in the Palingenia-larva.
The eighth and ninth segments of the abdomen are
simple and smooth. The tenth or last abdominal
segment is furnished with three hairy filaments ;
besides these, there are two small curved appendages
(claspers) which are not so conspicuous in the female.2
To these are added in the male yet another pair of
small appendages beneath the others.

" The colour of the larvae, when very small, is pale
blue, verging upon green. This colour belongs rather
to the viscera than to the skin. The eyes are dark

1 The correspondence between the gills or any other organs
of Sepia, and those of an Insect, is very far from close.

2 On the under side of the last abdominal segment of female
Ephemeras, a pair of minute appendages can be made out,
which are the parts intended by Swammerdam.

294 NATURAL HISTORY OF AQUATIC INSECTS CH.

brown, and the back speckled with dark spots, which
become larger with age.

" The male larvae are distinguished in the first
place by the eyes, which are twice as large as in the
female. The body of the male, however, is much
smaller than that of the female, as in all other Insects
which I have observed.1 The effect of this difference
in size, is that space is allowed for the vast number
of eggs formed in the female. The male has longer
tail filaments, and possesses moreover three or four
appendages, placed partly on the sides and partly
beneath, which in the female are inconspicuous or
wanting altogether.

" The larva is harmless and inoffensive. If roughly
handled, it bends its head towards the breast, and
stiffens its body. There is nothing more wonderful
in these creatures than the play of the branchiae
which stand out from both sides of the body. They
move so regularly, distinctly, and incessantly, as to
excite admiration in the beholder.

" I come next to consider the transformation of
this Insect. The change is effected so rapidly that it
seems to consist merely in slipping off two integu-
ments and unfolding certain appendages. In order
to make it quite plain what is the difference between
the swimming larva and the flying Insect, I will first
describe the internal organs as they occur in both
stages. Here I follow a path trodden by no one

1 Since Swammerdam's time a good many exceptions to this
rule have been noted, among the rest, the Stag-beetle, Dynastes,
Megasoma, many Dragon-flies, the Honey-bee and other Bees,
Methoca, £c. See Darwin's Descent of Man, chap. X.

vin MAY-FLIES 295

before. I will not however lament, with Clutius, the
rarity of books dealing with this subject. Nature is
the best revealer of her own wonders. Though books
may be useful, if they truly represent the phenomena
of nature, I pity those who trust to the observations
of others, and impose fictions upon themselves and
their readers.

" I will describe the methods which in the year
1670 I employed to investigate the anatomy of the
larva.

" The male, which is easily distinguished by its
large eyes, is fastened with the finest needles, back
downwards, upon black paper or linen in a wooden
dish. Then having cut through the skin, we see a
watery fluid flow out, which is the true blood of the
animal, though it is not red in colour as in the Earth-
worm or in Quadrupeds. I have found nothing bettei
for opening the skin than fine scissors, for lancets,
however sharp, tear the parts. Then with the finest
scalpel, or the point of a needle, the skin is gently
separated from the parts beneath. Under the skin
is found a delicate membrane, and beneath this the
muscles of the body-wall, some of which pass directly
from one segment to another, others obliquely or
transversely. Others again serve for moving the
limbs. Within the muscles is a very delicate mem-
brane, upon and within \vhich lies the fat, consisting
of minute white vesicles. Next we find the cesophagus,
stomach and intestine. The cesophagus, like a thin
thread, passes from the mouth into the thorax, and
expands to form the crop. The crop, when distended
with food, or with air injected from a fine glass tube,

296 NATURAL HISTORY OF AQUATIC INSECTS CH.

is smooth externally, but internally it is thrown into
a reticulation of folds. It appears to be provided
with multitudes of fine vessels, but if these are care-
fully examined with a magnifying glass, it will be
found that they are really air-tubes, which supply all
parts of the body, whether external or internal.
Beyond the crop comes the stomach, then the small
intestine, the large intestine or colon, and lastly the
rectum. The small intestine is provided internally
with a number of folds resembling the valvular folds
of the small intestine of man. In the ' colon are
longitudinal muscular valves, rather like those which
form the manyplies of the ruminant stomach. The
delicate rectum leads to the exterior of the body. A
pair of muscles is attached, one on each side, to the
rectum, and serves for pressing out its contents. Since
the larva feeds upon mud, we commonly find the crop
and intestine filled with mud. When the time of trans-
formation is approaching, the animal ceases to feed,
as also do the Cossus, the larva of the Bee, and the
Silkworm. Hence the intestine becomes transparent
at the time of transformation.

" A pair of tracheal trunks wind in a serpentine
manner along the sides, and send branches to all the
organs. The tracheae consist of numerous rings held
together by delicate membranes. At the time of
moult the lining of the air tubes is cast, though I have
not seen this in Palingenia. It is, however, very
conspicuous in the Silkworm, where, at the time of
moult, hundreds of the delicate air-tubes cast their
lininor membrane. I find it difficult to discover the

^5

external openings of the air-tubes, since they do not

vni MAY-FLIES 297

open into the mouth or throat as in other animals.
After long examination, I believe that I have dis-
covered the openings on the under side of the thorax,
nearly in the same place as I afterwards found them
in Grasshoppers, where, however, they are easily seen.
Since the Palingenia-larva lives in water and mud, it
is natural that the openings should be narrow and hard
to discover.1 From these observations it is clear why
the Palingenia larvae, when the water of the river rises,
creep upwards and betake themselves to new tubes,
in order that they may get the air which they require
to breathe. For the same reason they follow the
water as it sinks, lest they should be surrounded with
air, and dried up. When the air-tubes are examined
in an Insect which has been dead some days, so that
the viscera have turned black, they appear like pearls,
or bright silver on a dark ground. The firmness of
their texture prevents them from rapid decay, and
they preserve their shape for a considerable time.
To make sure whether air is really contained in these
vessels or not, it is only necessary to compress them
with the point of a needle under water, when bubbles

1 Swammerdam is mistaken in supposing that there are such
openings in the larva of Palingenia or other Ephemeridae. At
times of moult indeed the future spiracles are momentarily
opened to permit the removal of the old lining of the air-tubes,
which comes off with the cast skin. Lubbock (Linn. Trans. .
Vol. XXV., p. 480) says that this does not happen with the larva
of Chloeon dimidiatum, except at its last moult. Only in the
imago (winged fly) and the sub-imago (a stage peculiar to
Ephemeridas, in which the imago is shrouded in a thin skin,
formed within the proper pupal skin) do the thoracic nnd
abdominal spiracles serve for the admission of air.

298 NATURAL HISTORY OF AQUATIC INSECTS CH.

of air will be seen to issue. In a dried larva torn
across, the air-tubes are very easily seen, for they
retain their shape, and remain open when all the
other parts have dried up. The six large gills which
stand out from each side of the body are all provided
with large air-tubes, three to each gill, as also are the
five pairs of golden-yellow fins beneath, by means of
which the larvae swim.1 Some other observations
which I had made upon the gills and their vessels
have been lost, and I cannot recollect what they
were. I do not know, for instance, what is the use
of the plume attached to the first pair of gills, and
I can now give no more information about them than
can be gathered from the figure.2

" The heart lies on the dorsal surface as in other
Insects. It swells out in the middle of each segment,

•

Just as Malpighi represents it in the Silkworm. I do
not agree with this author in saying that the larva is
furnished with more than one heart, and I have only
very rarely seen any contraction of the heart in the
larva of Palingenia.

" The nerve-cord consists of eleven ganglia. From
the first of these, or brain, the optic nerve can be seen

1 In Palingenia, as in many other Ephemerida?, each tracheal
gill consists of a plate or lamina and a bunch of respiratory
filaments. Swammerdam calls the laminae " gills," and the
filaments " fins," though both form parts of one set of organs.

2 In addition to the six gills described by Swammerdam
there is another in front, probably the plume referred to in
the text. The lamina of the fore pair in Swammerdam's figure
is too large in comparison with those which succeed (Vayssiere?
Org. des Larves des Ephemerines, Ann. Sci. Nat. Zool. 1882,

P- 5)-

vni MAY-FLIES 299

to spring. In the same way the nerves of the body
are given off from the ten other ganglia. Paired
connectives issue from each ganglion, and unite it to
the next, so that the cord appears to be divided along
a considerable part of its length. The nerve-cord may
be demonstrated in an uninjured larva, by inflating
the body with air from behind. By this means the
cord is pressed against the integument, and can be
seen with a lens, or even with the naked eye.

" There are two ovaries in the female, resembling
the ovaries of Fishes. Each is supplied with in-
numerable air-tubes. The eggs are so small that they
cannot well be observed without the aid of a lens.
They are of a flattened oval form, and of white colour.
Their small size in comparison with that of the adult
Insect is explained by the fact that the larvae grow for
three years before they arrive at maturity.

" The emergence of the fly takes place in warm and
still weather. Shortly before emergence, the wings
are observed to become prominent, though still
enclosed within the larval skin. The intestine is
emptied, and the colour of the animal changes in
consequence.

" When all is ready, the larvae quit their burrows,
and swim freely in the water. The time of emergence
is usually towards evening, and always in the summer
months. In the year 1671, I saw the Palingenia
flying about from the I3th of June.

" When the larvae have left their burrows, they
make their way with all speed to the surface, and the
transformation is effected with such rapidity that even
the most attentive observer can make out little, except

300 NATURAL HISTORY OF AQUATIC INSECTS CH

that the winged fly suddenly darts out from the midst
of the water. It is possible to observe what is taking
place by going on the stream in a boat, catching with
the hand larvae which are about to emerge, and
slightly pressing the thorax so that they cannot
disengage themselves from the larval skin.

" How, we may ask, are the wings expanded with
such speed ? We might suppose that they are pro-
vided, like the wings of Earwigs, with muscles, which
can suddenly expand the wings, or at pleasure fold
them up again within their narrowr sheaths. But the
wings of the Palingenia are not provided with
muscles, and it appears that they are expanded by
sudden injection of blood from the heart. If the
wings at the time of emergence are wounded, they
bleed freely. Air injected into the air-tubes may also
help to render them stiff and dry. If the wings of a
larva ready for emergence are cut off and thrown
upon the surface of water, they can be seen to expand,
though detached from the body.

" When the Palingenia has quitted the water it
makes its way at once to some place where it can
settle, and then it immediately casts a second very
delicate and transparent skin, which covers the whole
body. This second moult, following so rapidly upon
the former one, does not materially change the shape
or appearance of the Insect.

" At the time that the Palingenia leaves the water,
the external form of the body undergoes conspicuous
change. The gills are completely cast, leaving only
small rudiments on the sides of the abdomen. At
the same time the jaws, the larval feet, and the

VIII

MAY-FLIES

301

sheaths of the wings and tail-filaments are shed ; the
antennae, which were previously conspicuous, become
far more delicate, and also much shorter in the
winged Insect than in the larva. The eye, which
was flat and smooth in the larva, is changed into a
many-faceted, compound eye. The legs and tail-

FIG. 91. — Fresh-emerged fly of Palingenia longicauda, male. From S\vammerdam;

Biblia Natures.

filaments become as long again as they were in the
larva, while the third or middle tail disappears

altogether.1

1 In those Ephemeridas which, like Palingenia, lose the
middle filament at the last moult, it is observed that there are
only two filaments in the fresh-hatched larva. To these the
third is added. When the fly emerges, the third filament is no

longer to be seen.

302 NATURAL HISTORY OF AQUATIC INSECTS CH.

" When the winged fly has become completely free,
it again seeks the water, and floats upon the surface
from time to time, as if in sport. Now and then it
rests upon the water with the help of its long tail-
filaments, the wings being folded together. The
hairs, with which the tail-filaments are so abundantly
provided, prevent the Insect from sinking. The same
may be observed in other Insects, like the larva of
the Gnat, which employ their hairs and the air
entangled in them, to float upon the surface. I
believe, but am not certain, that while the male
Palingenia during transformation moults twice, the
female moults once only.1 The male may be dis-
tinguished by the larger size of the eyes, which are
twice as great as in the female. The tail-filaments
of the female are shorter by a third than those of the
male. The colour of the male is more ruddy than
that of the female. Lastly, the male, in addition to
the two long tail-filaments, has four minute appen-
dages, which can hardly be made out in the female.

" The female deposits her eggs upon the surface of
the water, where they are fertilized by the male, as in
the case of Fishes.2

" In the adult condition these Insects do not eat, as is
the case with various other Insects also. The mouth-
parts and alimentary canal of the winged fly are
completely useless and empty. Hence the lightness

1 That is, the female remains in the subimaginal skin, or
transparent covering described above, while this is cast by the
male. This statement is not confirmed by observation of other
Ephemerae.

2 This is a mistake, as mentioned above

vni MAY-FLIES 303

of the body, and the short duration of the winged
Insect. After emergence they do not live more than
five hours at the most, that is from six or half past
six in the evening to eleven. I have observed the
length of their duration in the winged state by keep-
ing the winged Insects in a box.

" Vast numbers perish at the moment of their
emergence, being devoured by Fishes. Those which
escape from the water are devoured in large numbers
by Swallows and other Birds. If they escape this
peril, and return to the surface of the water, many
fall a prey to Fishes. At the time when the winged
Palingenia abounds, it has been observed that the
Trout are fatter and better flavoured.

"It is well known to all persons who frequent
rivers, that the Ephemerae appear in continual suc-
cession for three days. A few may be seen on the
fourth or even on the fifth day.

" What I have related concerning Palingenia
differs in many particulars from the account given
by Mouffet, Aldrovandus, Johnston, Clutius and
others, but I am far from judging or blaming them.
It may easily be that these naturalists have described
a different species from mine, but I would recommend
every one who desires to know the truth about these
things to learn himself of Nature, for Nature can
teach in a short space of time more than any one can
learn in a long course of years from many books.

" While studying the life-history of Palingenia, I
became acquainted with other species. None of these
live so short a time in the winged state as Palingenia.
In the month of June 1670, at Sloten near Amsterdam,

30J NATURAL HISTORY OF AQUATIC INSECTS CH.

I saw in the evening a great crowd of small Flies,
a little larger than Gnats. So many settled on my
clothes that I was completely covered with them,
and great numbers left their thin pellicles behind on
my clothes. I observed that they betook themselves
to the water, and sported there like Palingenia. The
larvae of this second species do not live in mud, or
make tubes, but live for the most part on stony or
sandy bottoms. They are tougher and hardier than
Palingenia, and their skin is more like that of a Crab
or a Shrimp. They have gills along the sides of the
body. If any one in the month of June takes stones
out of the Rhine, the Leek, or other of our native
streams, he will see many of the larvae clinging to
them. I have seen the same thing in the Loire, the
Seine, and other rivers of France."

Reaumur gives us a very lively account of one
species of Ephemera (Polymitarcys virgo), besides
notices of some others. The following is a condensed
translation of the twelfth memoir of his sixth
volume.

" Many kinds of Flies are doomed to perish on the
very day on which they emerge, and are hence called
Ephemerae or Dayflies. Some do not even see the
light of the sun ; they emerge after he has set, and
die before he rises again ; others live an hour or even
half an hour only.

" The wings of the adult Ephemera are trans-
parent, and shorter as well as broader than those
of common Flies. There are two pairs,1 the fore

1 Sometimes reduced to one, the hind pair being undeveloped
in certain species.

vin MAY-FLIES

305

pair being considerably the larger. When the Insect
is at rest, the wings are carried vertically, as in most
Butterflies.

" The body ends behind in two or three long fila-
ments, which are very fragile. The winged Ephemerae
often have these appendages more or less imperfect.

" Though the duration of the winged Ephemeras is
so short, they live a long time as larvae in the water.
Swammerdam supposes that the species described by
him remains three years in the larval stage. Some
species known to me live two years as larvae, and
many others about one year.

" During its aquatic life the Ephemera undergoes
no conspicuous change, except that the pupa, or
nymph, exhibits on its thorax the sheaths of wings,
which are altogether wanting in the larva.1 There
are six jointed legs, a head furnished with biting
jaws, and slender, many-jointed antennae. The ab-
domen is ten-jointed, and ends in two or three long
filaments.

1 The aquatic period of an Ephemera does not admit of
definite subdivision into stages. The fresh-hatched larva has
commonly neither tracheal gills nor any trace of wings. Rudi-
ments of gills shortly appear, and become larger and more complex
at each succeeding nioult. Before they have attained their com-
plete development the rudiments of the wings are often apparent.
The gradual acquisition of the imaginal structures is facilitated
by the unusual number of moults. In Chloeon dimidiatum as
many as twenty-one larval moults have been observed by
Lubbock. "In moulting the insect does not split the skin of
the thorax, as is so generally the case, but merely that on the
upper part of the head. It is wonderful, indeed, how it can
escape through so small an orifice" (Lubbock, Linn. Trans.,
Vol. XXIV., p. 68).

X

306 NATURAL HISTORY OF AQUATIC INSECTS CH.

" Some Ephemera larvae make burrows ; others
lead a free life. The sides of the abdomen bear a
number of paired tufts, which are moved up and
down with great rapidity. They have been taken by
some authors for fins, but it is enough to refute this
notion to observe that they are moved most ener-
getically when the larva remains stationary. They
are gills, as their microscopic structure shows. In
some Ephemerae they stand out from the sides of the
body like the oars of a galley,1 in others they are
upright or curved over the back.2 There may be six
or seven pairs of gills, beginning at the first or second
abdominal segments ; the last three segments never
bear gills. In the larva of Ephemera vulgata each
gill consists of two branches of nearly equal size, which
are each fringed with filaments like a feather. Two air-
tubes traverse each branch and filament. In another
Ephemera larva [that of Chloeon dipterum] the gills
fo-rm plates or leaflets, each of which is doubled in
two, and the tracheae, instead of projecting, run in the
thickness of the leaflet. These gills work backwards
and forwards, and often all are moved simultaneously
though sometimes the last pair remain motionless
when all the rest are moved.3 In the larva of Poly-
mitarcys virgo each gill consists of two long and

1 This is the case, for example, with the larva of Baetis
(Ecdyurus) fluminum, a large species found abundantly be-
tween and beneath the stones of clear streams.

2 Ephemera vulgata, Polymitarcys virgo, &c.

3 The last pair of gills in the Chloeon larva have the leaflet
flat, and not folded in two. In the six other pairs the two
leaflets are more distinct from one another than Reaumur's
description implies.

VIII

MAY-FLIES

307

narrow leaflets, traversed
by central tracheae, and
fringed along the fore edge
by close-set filaments.

" Polymitarcys is the
commonest species in the
neighbourhood of Paris,
and the one best worth
study. The winged fly ap-
pears for three or four days
in succession in most years,
and affords a curious spec-
tacle. On coming out of
doors in a morning, \ve see
the pavement strewn all
over with pretty flies rather
like Butterflies. Sometimes
they lie so thick as to hide
the pavement completely,
and this sight is repeated
three or four days running.
Our surprise is increased
when \ve learn that these
Insects issue from the
rivers, in which they pass
their early stages. It is
hard to imagine that so
many can come forth from
the water at the same time
-all the more because on
searching the river one
might fail to find a single

FIG. 92. -Larva of Polymitarcys
virgo, X 6. From Vayssiere. The
tracheal gills of the left side are
displayed.

X 2

3o8 NATURAL HISTORY OF AQUATIC INSECTS CH.

•

Ephemera there, though we should expect to find
them heaped together by the thousand.

" The larva of Polymitarcys rarely swims free in the
water ; it has a burrow excavated in the earthy bank
of the river. From the level of the water for two or
three feet downwards the bank is everywhere drilled
with holes a quarter of an inch or less in diameter.
The holes take for the most part a horizontal direc-
tion. The passage returns upon itself to a second
outlet close to the first, as may be seen by slicing a
clod of the earth horizontally, when a U-shaped
burrow is exposed, the two limbs being separated
only by a slight partition of earth, which is often
broken down. The larva can therefore enter or leave
its burrow without being obliged to creep backwards.
Earth of the consistence of clay or marl is best suited
to the purpose ; gravel cannot be made to answer.
The lining of the burrows is much finer than the
surrounding earth, and appears to be laid down in an
even layer by the larva. The burrow is proportioned
to the size of its inhabitant ; a full-sized one is more
than a quarter of an inch in diameter, and over two
inches long, measured from one end to the other
along the curve. In such a burrow the larva is freely
exposed to the water, while protected from the
attacks of Fishes. Its food consists of earth, from
which it removes by digestion all nutritive matter.

" I suppose that the larval stage occupies two years,
and that in the last few months of the second year
signs of wings appear ; at this time half-grown larvae,
destined to manure in a subsequent season, are
abundant.

viii MAY-FLIES 309

" The head of the larva is furnished with slender,
many-jointed antennae, a pair of long, curved mandi-
bles, well suited to the excavating of earth, and
behind these with two pairs of maxillae, the hind pair
being united to form the labium. The fore legs are
stout, as if for digging ; they are directed forwards,
as also are the legs of the second pair, while those of
the third pair turn backwards. I have seen larvae,
when taken from their holes and placed in a bucket
of earth and water, use their fore legs to make a fresh
hiding-place.

" The larvae of Polymitarcys are very hard to please
in the matter of fresh water ; they perish in four or
five days if placed in large buckets full of water ;
those of some other species, such as Chloeon dipterum,
on the contrary, will live for months in shallow
vessels without renewal of the water, and will there
undergo their transformation.

" The Ephemera described by Swammerdam
(Palingenia longicauda) emerges about St. John's
Day ; the fly of Polymitarcys much later, viz., in mid-
August. Palingenia appears about six o'clock in the
evening, Polymitarcys at sun-down or later. Differ-
ences of temperature and other less obvious causes
render the appearance of the flies earlier in some
seasons than in others.

" In 1738 I resolved to attend to the emergence of
the fly, and engaged an angler of Charenton to warn
me when the first signs appeared, which were expected
between St. Laurence's Day and " Notre Dame
d'Aout," that is, between the loth and I5th of August.
This year the flies appeared on the i8th. On the

310 NATURAL HISTORY OF AQUATIC INSECTS CH.

19th I received warning from my angler, and the
same day, three hours before sunset, I took his boat
to examine the banks of the Marne and Seine. Where
the shore was level and sheltered from the wind,
heaps of dead Ephemerae could be seen. During this
excursion by water I removed some clods of earth
which were riddled with holes, and placed them in a
large bucket of water as near as possible in their
original position. The sides of the clods exhibited
larvae partly or completely exposed. At length the
sun set. At that time Ephemerae were to be seen
flying here and there over the Seine, but at half past
seven the number showed no important increase. I
crossed over into the Marne, where there seemed to
be still fewer. At about eight o'clock the coming on
of evening and the flashes of an approaching thunder-
storm caused me to return into an arm of the Marne,
which washes the stairs leading from my own garden.
I had the bucket containing the Ephemera-larvae
carried into the garden, but as soon as the men had
reached the top step they cried out that a prodigious
number of Ephemerae were emerging. I seized one
of the lanterns with which they had come to meet
me, and ran to see what was going on. All those
clods which stood out of the water were covered with
the flies, some beginning to cast their skin, others
nearly free, others ready to take flight, while in the
water beneath were larvae whose transformation was
less advanced. While I was examining the contents
of the bucket the storm broke, and forced me to take
refuge in the house. Before leaving the bucket I
threw over it a cloth, to prevent the Ephemerae from

vni MAY-FLIES 311

flying away. In half an hour I was able to return to
the garden and take off the cloth, when the number
of winged Ephemerae was found to be greatly in-
creased, and still increasing. Many took to flight ;
still more were drowned. The element lately so
essential to their life had now become deadly to
them.

" Ephemerae, attracted by the light, came from a
distance and were drowned in the bucket. In order
to spare them, and to examine more carefully those
which were uninjured, I covered the bucket once more
with the cloth, and held the light beneath. The cloth
was speedily covered with Insects, which were taken
off in handfuls.

" So far I had taken no notice of what was going
on by the river, but now the exclamations of rny
gardener, who had gone to the foot of the stairs,
attracted my attention. I then saw a sight beyond
all expectation. The Ephemerae filled the air like
the snow-flakes in a dense snow-storm. The steps
were covered to a depth of two, three, or even four
inches. A tract of water five or six feet across was
completely hidden, and as the floating Insects slowly
drifted away, others took their place. Several times
I was obliged to retreat to the top of the stairs from
the annoyance caused by the Ephemerae, which
dashed in my face, and got into my eyes, mouth, and
nose.

" It is singular that nocturnal Moths, which shun
the light of day, should be attracted by the lights in
our rooms, and still more singular that these Ephe-
merae, which emerge only after sun-down, and perish

312 NATURAL HISTORY OF AQUATIC INSECTS CH.

before sunrise, should be drawn so powerfully towards
a lantern. The person who held the light had a bad
time of it ; in a few moments he was covered with the
flies, which came in all directions as if to overwhelm
him. The luminous sphere about the light was
crossed at all angles by the orbits of the circling
Insects, which after performing one or two revolutions,
fell to the earth.

" After half an hour or less the swarms were less
dense, and by ten o'clock only a few scattered Ephe-
merae could be seen on the river, while no more came
round the light.

" Next day the Ephemerae appeared in undiminished
numbers, but on August 2ist they hardly amounted
to one-third of the previous flights. Each day they
appeared between a quarter and half past eight ;
from nine to half past nine they filled the air, while
by ten o'clock they had almost ceased to fly.

"In the afternoon of the 2ist the air was cold for
the time of year, the thermometer standing at I7°R.
(7O°F.).1 One might suppose that heat would hasten
the transformation of these larvae, as it hastens the
development of the chrysalis or the egg of other
Insects. However, in spite of a relatively low
temperature, the flies emerged at the same hour as
before.

" On the 22nd the thermometer fell to I5°R. (66°F.);

1 I have given in brackets the accepted equivalents on Fahren-
heit's scale of the readings quoted by Reaumur from his own
thermometer. They are obviously too high. Reaumur's ther-
mometer was somewhat differently graduated from any modern
ones.

viii MAY-FLIES 313

it rained several times during the morning and poured
all the afternoon. In the evening the Ephemerae
appeared at the usual time, though in smaller num-
bers, most of them having already emerged. Rain
was falling, though in less quantity. It appears,
therefore, that whatever the weather on the day of
emergence — warm or cold, sunshine or rain — the
Ephemerae quit the water at a fixed hour.

" What becomes of the prodigious swarms of Insects
when they no longer fly through the air ? They are
for the most part already dead or dying. A large
part fall into the river from which they issued. The
Fishes enjoy a feast, and the Erench anglers speak of
the Ephemerae as manna — e.g., they say the manna
has begun to appear ; there was a good fall of manna
last night.

" Whether devoured by Fishes or not, those which
fall into the water soon perish. More lingering but
not less certain is the fate of those which descend
upon the banks or the neighbouring fields. Heaped
one upon another, and unable to move, they die by
inches, the last survivors perhaps seeing the rising of
the sun.

" I returned to Paris on the 22nd at ten o'clock at
night, but left instructions to have the Ephemerae
watched on the following days. They appeared in
constantly diminishing numbers for four or five days
longer.1

1 Polymitarcys does not occur in the British Isles, though it
is plentiful in the large rivers of the Continent. I have seen
swarms of winged insects, rivalling those described by Reaumur,
at Dinant on the Meuse.

3i4 NATURAL HISTORY OF AQUATIC INSECTS CH.

" The Ephemerae issue from the water in order to
perpetuate the species. As soon as they emerge, the
females are ready to lay their eggs, and actually do
so. The rapidity with which they cast the larval
skin is truly wonderful. We cannot take our arm
from the sleeve of a coat more readily than the
Ephemerae extricates its abdomen, wings, legs, and its
long tail-filaments from their sheaths. During the
operation they rest upon objects standing out of the
water or upon the water itself. The thorax splits
lengthwise, and the rest of the business of extrication
is over in a moment. I have often tried to stop it
half-way, in order to see how the wings are folded in
their sheaths. But though I crushed the head as soon
as it appeared, the act of emergence went on all the
same. So also when I plunged emerging Ephemerae
into spirits of wine, they completed their moult before
perishing. The long tail-filaments are often broken
during extrication. The cast skin is sometimes
carried up into the air, clinging to the tail-filaments,
and an Ephemera in this state seems twice as long as
usual.

" The fly of Polymitarcys virgo is about three-
quarters of an inch long, not counting the long tail-
filaments. The hind wings are very small in com-
parison with the fore wings, which are in outline like
those of many Butterflies, though they are not scaly,
but transparent. The fore legs are very long and
directed forwards,1 the others short, perhaps too short,
for the fly cannot readily rise from a flat surface and

1 Probably, as in other Ephemerae, these long fore legs are
used by the male to seize the female.

VIII MAY-FLIES 315

take wing. I found that when Ephemerae fell upon a
napkin spread over my knees, they could only rise
into the air with the help of their long tail-filaments,
which gave a momentary support to the body.

" The male fly of Polymitarcys has the middle
tail-filament much shorter than the two others, and
only from one-sixth to one-eighth of their length.
Two pairs of appendages project from the under side
of the extremity of the abdomen of the male, which
seem to correspond to the clasping organs of other
male Insects.

" The eggs produced by the female fly are intended
to be laid in the water of the river, but as if in
ignorance she often deposits them on any object on
which she may happen to alight. The eggs are
grouped in two oval masses, from one-third to a
quarter of an inch in length. I counted 350 eggs in
one such mass. The short life of the winged female
compels her to deposit her 700 or 800 eggs at once,
without much discrimination of likely and unlikely
places. When the bucket containing a lantern and
covered with a cloth attracted Ephemerae from a
distance, many of these laid their eggs on the cloth.
The female turns up the extremity of the abdomen
nearly at a right angle, and the orifices of the two
oviducts, which open behind the sixth [the figure
seems to indicate the seventh] abdominal segment,
are exposed ; from these the t\ 'o clusters of eggs are
passed out at the same time. When the eggs have
been discharged, two vesicles, apparently filled with
air, project from the orifices of the oviducts ; these
are probably either principal or subsidiary aids in the

316 NATURAL HISTORY OF AQUATIC INSECTS CH.

act of expulsion. The air, which dilates the vesicles
and expels the eggs, may be taken in by the two
pairs of large thoracic stigmata.

" Ephemerae which are not dazzled by light or
otherwise led astray skim the surface of the water,
and support their bodies upon it by means of the
tail-filaments while engaged in egg-laying. The
specific gravity of the eggs is higher than that of
water, and they fall at once to the bottom. There
the eggs soon become scattered, for the jelly in which
they are imbedded is soluble in water. I have placed
several egg-masses over night in vessels of water ;
next morning the eggs were spread over the bottom,
quite loose, though in regular groups. If placed in
alcohol they cohere, for alcohol does not dissolve the
jelly.

" How are the eggs fertilised ? How can time be
found for their fertilisation in the brief space during
which the winged Ephemeras are active ? I have no
precise information to give on these points ; the time
of emergence renders it hard to observe directly what
happens.1

" Swammerdam has observed the habits of an
Ephemera (Palingenia longicauda) which emerges
earlier in the day than Polymitarcys, and takes to
flight more than two hours before sunset. He thinks
that in this species the eggs are fertilised by the
males in the water, after the manner of Fishes. But
it is hard to understand how eggs which, like those
of Polymitarcys, fall to the bottom as soon as they

1 De Geer has cleared up this difficulty, as will be seen
further on.

vni MAY-FLIES 317

are liberated, can be thus fertilised.1 Again, how is
it that I never saw the males fertilise the egg-clusters,
which had been laid on a cloth or napkin ? If the
female could be so mistaken as to lay her eggs where
they could not possibly be hatched, why should not
the males fertilise them in such situations ?

" I prefer to suppose that the males unite with the
females, but that the act of union is very brief.
Indeed, I have seen something of the sort, though
only by the help of lights held near the surface of the
water. I with others caught pairs of the Insects
which appeared to be attached. When holding a
napkin over my knees, I saw the males mount upon
the females and apparently unite with them. Lastly,
the male is provided with appendages near the
extremity of the abdomen which seem to be suitable
for grasping the female.

" I do not know how long it takes for the eggs to
hatch, for those which I saved were put into a vessel
of water, which was not changed. Running water is
probably necessary.

" The great fecundity of the females and the
sheltered situation in which the larvae dwell, explain
the extraordinary numbers of the winged flies.
Every year is not, however, equally prolific in swarms
of Ephemerae. The short duration of the winged
state requires the assemblage of great numbers of the
same species. If these Insects emerged a few at a
time and on many successive days, the males and

1 It is pretty certain from subsequent observations that the-
eggs of Ephemeras are fertilised in the body of the female
before they are laid.

318 NATURAL HISTORY OF AQUATIC INSECTS CH.

females could hardly meet before their brief life was
ended. To some allied species, which emerge in
scanty numbers, a longer span has been conceded.
One such Ephemera I kept alive for six or seven
days, and it might have lived longer had it not
been restrained from flying abroad."1 [Here ends
Reaumur's account.]

De Geer supplements the observations of Reaumur
by a number of interesting particulars. In Sweden,
where he lived, the prodigious swarms of Ephemerae
do not occur, though many species are found in
smaller numbers.

The first kind which he describes is Ephemera
vulgata, whose larva is very common in ditches,
ponds and slow streams, where it burrows in the mud
or hides beneath stones. It can walk on the bottom,
and also swims at times, drawing up the legs, and
propelling itself by an up and down movement of the
abdomen. The larval antennae are rather long, and
the slender mandibles (which are crossed) extend
considerably in front of the head. There are six
pairs of tracheal gills, which extend over the back
and point towards the tail ; each consists of a double
leaflet fringed with long, hollow filaments, in which
fine air-tubes run. The legs are flattened and fringed
with hairs. The three tail -filaments are longer than
the body, and each is fringed with hairs on both
sides. [In some species — e.g. Chloeon dipterum and

1 Observations and experiments made since Reaumur's time
render it probable that the life of an Ephemera may be pro-
longed by captivity. Mating and egg-laying seem to hasten
the end.

VIII

MAY-FLIES

319

Polymitarcys — the lateral filaments of the larva are
fringed on the inner
side only.]

" The recently emerg-
ed fly," says De Geer,
" settles on trees, plants,
walls, &c., near the water
which harboured the
larva. Here it fixes itself
by the hooks of the feet,
usually with the head
downwards, and rests
until the last or sub-
imaginal moult is at
hand.1

" The fore legs of the
male fly are very long,
as in other Ephemerae,
and raised from the
ground, so that they
look like antennae. They
can be brought down

1 In some Ephemendae I
have seen the sub-imago
(imago enclosed in a trans-
parent skin) emerge from
the larval skin under water,
and float up. In other cases
(Baetis fluminum, for in-
stance) the larva creeps out
of the water, like Perla, and
the cast larval skin is left on
the shore.

FIG. 93. — Larva of Ephemera vulgata, X 5.
From Vayssiere. The fore and mid legs
are cut short, and all the tracheal gills of
the left side, except the first, are cut
across. A small tuft of hairs appears
close to each gill, which probably serves
to keep off foreign particles.

320 NATURAL HISTORY OF AQUATIC INSECTS CH.

for walking, but owing probably to their unusual
length, the gait of the male Insect is slow and
awkward. At the extremity of the fore leg are two
flattened appendages, diverging from one another,
and armed with minute spines, as if to improve the
holding-power of the legs.1 In the female the fore
legs, though not so long, are raised, as in the male,
The male has also claspers at the end of the abdomen,
much larger compound eyes than those of the
female (as in many other Insects) and longer tail-
filaments.

" The eggs are discharged all at once, in the form
of a yellowish-white, oblong disc, which is dropped
into the water, probably as the female flits over the
surface. It sinks at once, and the eggs separate from
one another.2

" The flies emerge at the end of May or the begin-
ning of June, always towards sunset. Some hundreds
may be seen together, dancing in the air, usually
around some tall tree. The males greatly exceed
the females in number. They rise by the action of
the wings, and when they have ascended five or six
feet above the tree, allow themselves to sink slowly
the wings being outspread and motionless, and the
long tail-filaments raised and widely separated. Late

1 In some species the fore leg of the winged male has a sort
of kink or angle, which may be used to hold the female (Baetis,
Chloeon).

2 The process of egg-laying seems to exhibit great variety in
the different species of Ephemeridae. Females of Baetis flu-
minum, Chloeon dipterum, and Chloeon Rhodani have been
found among stones under water, as if they had descended to
lay their eggs there.

viii MAY-FLIES 321

in the evening they all disappear, and rest upon
plants throughout the night and the following day.
The duration of the winged state was not determined
with certainty, but it is short. The flies continue to
emerge for fifteen days or more in succession.

" I have often seen the union of the male and
female. Hio;h in the air I have seen a male out of

o

a swarm of Ephemerae seize a female, and remain
attached to her ; they flew away to the top of a wall
where they settled together. The wall was too high
for me to see distinctly, but I could make out that
one of the pair, no doubt the male, curved the
abdomen round as if to seek the female. Swammer-
dam was no doubt much in error in supposing that
the eggs of Ephemerae are fertilised in the water.

"After publishing these imperfect observations, I
was several years later amusing myself by the con-
templation of the aerial dances of swarms of Ephe-
merae, composed of males only, as they usually arc,
when I perceived that if a female mixed in the swarm,
as often happened, two or three males pursued her,
until one of them succeeded in flying away with her.
The pair ordinarily sought the top of a wall or the
summit of a tree, but two or three couples placed
themselves conveniently for observation on the leaves
of a neighbouring bush. I could see that the male
placed himself beneath the female, and curved his
abdomen upwards to gain the orifice of the oviduct.
The affair was ended in a moment, and the male flew
away. I should have been glad to observe the use of
the long fore legs and of the claspers of the male, but
my attention was fixed upon the movements of the

Y

y.2 NATURAL HISTORY OF AQUATIC INSECTS CH.

abdomen of the male, and it is impossible to take
notice of a number of things at once.

" Captive Ephemerae of this species survived two or
three days, while those of Chloeon diptcrum survived
several days after acquiring wings." l

We owe to F. J. Pictet, a celebrated naturalist of
Geneva, some interesting particulars of the life-history
of Ephemerae. His classification of the larvae accord-
ing to their mode of life is instructive, though it is to
be regarded rather as a description of some well-
marked larval forms than as a zoological system of
any kind. He arranges the larvae as belonging to
one or other of the following groups : —

I. Burrowing Larva.- -Tv this section belong the
Palingenia described by Swammerdam, the Polymit-
arcys described by Reaumur, and with less pronounced
habits the Ephemera vulgata, which forms the chief
subject of De Geer's observations. The body is long,
the head small, and the fore legs adapted for digging.
We may add that the respiratory organs bear many
filaments, and are therefore suited to passive aeration
of the blood by means of a current of water, as well
as capable of active respiratory movements. These
larvae inhabit slow streams with shores of clay or
earth, and make their burrows in the banks. Their
food is finely divided organic matter.

Leptophlebia connects groups I and II. It has
the tracheal gills filamentous, but united by a mem-
brane at the base. The legs show no adaptation to
^ occurs on the banks of the Rhone in

fc>-

1 Winged individuals of this species have since been kept
alive more than three weeks.

vin MAY-FLIES 323

company with Heptagenia, sheltering itself in the
mud.

II. Larvce with flattened /^//^.--Ecdyurus flumi-
num, a large species found in clear, stony streams in
hilly districts, or Heptagenia, will serve to exemplify
this section. All parts of the body are flattened out ;
the head is semi-circular and the legs very broad.
This flat shape enables the larva to cling to stones,
and escape the full force of the stream. The leaflet
of the gill is well developed, and the filaments small.
They make no burrows, but when disturbed they
swim with tolerable ease. They are carnivorous, and
prey upon small aquatic larvae.

III. Swimming Larvce.- -The common Chloeon,
which is found in pools, ditches, and small rivulets,
is adapted for swimming in preference to other modes
of locomotion. The legs are weak, but the tail-
filaments, all broadly fringed on both sides, form an
efficient tail-fin. The tracheal gills consist of leaflets
without filaments. The larvae feed on small animals
which they find among aquatic plants.

IV. Creeping Larva --Here Pictet places some
larvae of which Ephemerella is the best example.
Not possessing limbs adapted for digging, nor means
of rapid locomotion of any kind, it inhabits streams
of moderate swiftness. The tracheal gills, which are
reduced to four pairs, are protected by large and firm
leaflets. To escape its enemies and pounce unseen
upon its prey, it covers its body with a thin layer of
mud, which clings tenaciously to the fine hairs with
which the integument is furnished. Some Perlidae,
as Pictet remarks, have the same instinct. So has an

Y 2

D

G

FIG. 94-.— Larva of Raetis (Ecclyurus) fluminum. A, the larva; 1), mquth-pam of
do. ; C, one of the tracheal gills; I), the t\vo wing-covers of the right side ; E,
mandible ; F, maxilla and its palp ; G, labial palp.

CH. vin MAY-FLIES 3-5

aquatic Beetle (Georyssus pygmaeus) which is found
adult under stones in running water. Helophorus
and Elmis also disguise themselves with mud,1 and I
suspect that Nepa is yet another example.

Such groups as these are merely short descriptions
of the adaptive modifications of particular species or
small groups of species. It would be easy to add to
the number by adopting a different principle of forma-
tion. Vayssiere, for example, defines five groups of
Ephemerae according to the modifications which the
larval respiratory organs may assume.2

His group I. includes those larvae in which the
tracheal gills consist of both leaflets and filaments

o

(This nearly agrees with Pictet's group of Burrowing
Larvae).

II. receives those in which the tracheal gills consist
of leaflets only.

III. has the filaments of the tracheal gills protected
by an overlying leaflet.

IV. has the hinder tracheal gills overlaid and pro-
tected by the enlarged leaflet of the second pair, the
first pair of gills becoming practically suppressed.

V. contains some very peculiar and interesting
larvae in which the mid-thoracic segment extends
backwards over all the tracheal gills, forming a special
respiratory chamber, which opens behind or by three
special apertures of small size. In this last group
the arrangement becomes similar in principle to that
adopted in Decapod Crustaceans (Lobster, Crayfish,

1 Kirby and Spence, Vol. II., p. 258.

2 Org. des Larves des Ephemerines, Ann. Sci. Nat. Zoo!., 1882,

P- 33-

326 NATURAL HISTORY OF AQUATIC INSECTS CH.

&c.) except that the external wall of the respiratory
chamber is formed in the Ephemerae by a backward
extension of the midthorax, instead of by a downward
extension of the carapace, as in the Decapods. To
this peculiar group belong a singular American form,
Baetisca, and an Old World genus, Prosopistoma,
found in the Rhone, Garonne, &c., and described by
Dr. Em. Joly and Vayssiere.

The eggs of Ephemeridae are often furnished with
singular appendages which in some cases prevent their
being swept away by currents of water. In Ephemera
vulgata l the eggs are described by Dr. H. Grenadier2
as possessing striated caps of reddish-brown colour
which invest both poles of the egg.3 A mushroom-
shaped stalk serves as a base and springs directly
from the end of the egg. This is of firmer consistence
than the striated part, whose numerous and close-
set fibres radiate regularly from it. The fibres had
previously been described by Leuckart 4 as bundles
of spermatozoa, but Grenadier points out that
they are to be found in all stages of development
within the larval oviduct, and not merely after fer-
tilisation, as Leuckart's interpretation requires. No
definite information as to the function of these striated
egg-caps has been obtained. According to Leuckart,
the eggs of some species of Ephemeridae bear them
at one end only, others at both ends.

Grenacher has also described what appear to be

1 Grenadier's words are : larvae belonging to Ephemera s. str.

2 Zeits.f. wiss. ZooL, Bd. XVIII. (1868), pp. 95-98.

* A similar striated cap is seen in the ovarian egg of the Gnat.
4 M tillers Arch. 1855, p. 200.

vin MAY-FLIES 327

long anchoring threads attached to the eggs of
Ephemera vulgata. From 8 to 12 of these are fixed
to a zone which encircles the egg transversely towards
one end. The whole arrangement is repeated at the
other end of the egg. Each thread bears a small
knob at its free end, and these knobs apparently
become entangled at the bottom of the stream and

o

moor the eggs.

CHAPTER IX

DRAGON-FLIES (ODONATA)

SOME Insects have at length impressed the popular
imagination with great distinctness, and among these
is the Dragon-fly. It is a little remarkable that
until modern times no Insect, except such as are
either useful to Man, or noxious in a striking degree,
should have commanded serious attention. During
the last hundred and fifty, or, at most, two hundred
years, mankind in Western Europe has begun for the
first time to take an interest in the natural history of
Insects without reference to their profitableness or to
the annoyance which they cause. The ancients cared
little for any Insect except the Bee, and no other
Insect plays a considerable part in classical literature,
though the Wasp, the Gnat, and the Gad-fly figure
occasionally in poetry or fable. It is a curious proof
of the general indifference to Insects that, until com-
paratively modern times, no literary use was made of
the extraordinary and easily observed transformations
of Butterflies. Indeed, there is, I believe, no familiar
Greek word of the classical period for a Butterfly. A
real popular interest in Natural History founded upon
observation is one of the latest fruits of the Revival

CH. ix DRAGON-FLIES 329

of Learning. During the course of the eighteenth
century, the discoveries of Swammerdam and
Reaumur slowly made their way into the thoughts
of the people, and some acquaintance with the life-
history of Insects is now to be counted upon in every
reader. The death-like repose of the chrysalis and
the emergence of the Butterfly, the short life of the
winged Ephemera, and the transformation of the
Dragon-fly from a sluggish larva lurking in pools to
a glorious winged creature flying swiftly through the
air, are now among the every-day illustrations of the
preacher and moralist, and form a highly character-
istic feature of modern literature.

It is easy to see how the Dragon-fly in particular
came to strike the imagination. Its size and beauty*
its singular form, its swiftness and strength, are con-
spicuous, to every observing eye. In its habits and
life-history we find the most emphatic contrasts.
There is the contrast between a sordid life in a
muddy pool and the animated life of a creature
strong in flight. The graceful form and lovely
colours of the winged Insect are combined with
fiercely carnivorous tastes. The slender body,
adorned with rare and pure colours, the delicate
gauzy wings, and the great eyes with the play of
light in them, should be associated, we fancy, with
the sweet juices of flowers or some other delicate
food. But when we capture a Dragon-fly on the
wing, and open its mouth, we find it filled with a
black mass of small Insects, held over for mastica-
tion at a time of repose.

In the days when as school-boys we found the

330 NATURAL HISTORY OF AQUATIC INSECTS CH.

summer half-holiday all too short for the exploration
of pond and stream, the Dragon-fly larva, with its
dingy colours, its forbidding shape, and its predatory
habits, excited the keenest curiosity. We saw it
stretch out its great paw and secure an unsuspecting
victim. Some of us watched with eager eyes the
emergence of the winged fly. Most of us were
satisfied, however, to read a prosaic abridgment of

FIG. 95. — Libellula depressa. From Charpentier.

that vivid description of the process of extrication,
which, with so much else, we owe to Reaumur.
Either by observation or by reading we all know
enough of that marvellous change to feel the beauty
of Tennyson's verses :-

To-dav I saw the dragon-flv

*• O J

Come from the wells where he did lie,
An inner impulse rent the veil
Of his old husk ; from head to tail
Came out clear plates of sapphire mail.
He dried his wings : like gauze they 'grew,
Thro' crofts and pastures wet with dew
A living flash of light he flew.

ix DRAGON-FLIES 331

The larva: inhabit water, for the most part standing
water, where they subsist by capturing living animals
of all kinds, Insects, Snails, Tadpoles, and even
Fishes. The skin is moulted several times during
this stage, which is believed to occupy as a rule a
year or more. For some time before the emergence
of the winged Insect, the rudiments of the wings
become externally visible, and in this state the Insect
is sometimes called a pupa. There is, however, no
stage corresponding to the resting-stage of the
Lepidoptera.

The larvae of the numerous species of Dragon-fly
exhibit external features by which it is easy to
distinguish and group them. Some of the smaller
species have three pointed, semi-transparent plates
(tracheal gills) projecting from the end of the body.
These are recognised at a glance as Agrionidae. In
the remaining Dragon-fly larvae, the tracheal gills
are not conspicuous. If the abdomen is compara-
tively broad and flat, and shorter than the hind legs,
we know that the larva belongs to the Libellulidae.
In ^Eschnidae the abdomen is slender and longer
than the hind legs.

The larva is not at all nimble. It often remains
for hours nearly motionless, clinging with its legs to
some water-weed. When it walks, the movement is
steady but not rapid. It has another mode of loco-
motion which may be called swimming. In larvae
of Agrionidae the abdomen is bent first to one side
and then to the other. This sculling movement
drives the body forwards with moderate speed.
Libellulid larvae can propel themselves more rapidly,

332 NATURAL HISTORY OF AQUATIC INSECTS CH.

but not continuously, by the sudden ejection of water
from the tail. In air the water can be spirted to a
distance of several inches. It is, however, only a very

o: « §

c 15
u t* £

Ji

rt ""

x £ £
2 -.o

D D

•C "o

'g-S

->

^

4) C

C U O

.SJ3'*3

T3 •" O,

gs-s
'

vo £- c

CN « S

^ u
6

slow or inattentive animal which could be captured
by the direct pursuit of a Dragon-fly larva. It trusts
chiefly to its sombre coloration and its motionless

ix DRAGON-FLIES 333

attitude. The larva clinging to a stem in the shady
recesses of water-weeds is not easily distinguished,
and the absence of movement removes the chief risk
of discovery. In the same way, the Hare in long
grass, the young Plover on a heath, the Trout in a
pool, are all concealed from view by their not very
close imitative coloration, but when they move, even
an ill-trained eye observes them. Until they move,
the eye is not focused upon them at all, but a mov-
ing object is noticed even when out of focus. The
policy of the slow Dragon-fly larva is to lie still and

FIG. 97. — Head of larva of ^Eschna, in side-view, with mask projected to seize prey.

wait till its victim comes within easy reach. Then of
course it must be quick as lightning. It does not
usually make a spring, but stretches out an arm-like
appendage of the head, and seizes its prey. The
appendage so employed is a peculiar modification of
a pair of limbs attached to the head, and called the
second pair of maxillse.1 In Insects these appendages
form the third pair of jaws, and are fused more or less
completely into a labium or under-lip, which closes in
the back of the mouth. The labium of the Dragon-
fly larva is carried on a jointed arm, and usually

1 Other explanations of the origin of these parts have been
given, and the question is not easily settled in a decisive way.

334 NATURAL HISTORY OF AQUATIC INSECTS CH.

much expanded at the end. Side-pieces, correspond-
ing to the labial palps, are attached, and there is
commonly a pair of spines or claws, which secure the
struggling victim. The details of the structure vary
according to the species. When the larva is at rest
the apparatus is folded up, the broad joint being
spread over the front of the mouth, while the arm is
bent backwards between the fore-legs. In Libellulid
larvae the side-pieces can be brought together in the
middle line like the jaws of a rat-trap, which they
further resemble in their toothed edges. No more
murderous weapon exists among animals than this.

We must next consider how the submerged larva
obtains its supply of air. In the broad-bodied
Libellulid larvae, a pair of large spiracles can easily
be seen on the dorsal surface behind the head (be-
tween the narrow prothorax and the mesothorax, i.e.
between the fore and middle segments of the thorax).
In the larvae of Agrionidae the same spiracles are
hidden, but can be made out by careful search.
Another pair of thoracic spiracles can be discovered
by dissection. These openings have been usually
supposed to be completely closed during the aquatic
stage. Reaumur found that oiling the spiracles did
not destroy life, as it does in ordinary air-breathing
Insects, and this fact has often been quoted as
decisive proof that the spiracles are not really open
in the Dragon-fly larvae. Reaumur himself draws no
such conclusion, but allows that the oil possibly did
not adhere to the wet surface, or that the spiracles
closed and shut out the oil. He might have given a
third explanation as an alternative, viz., that the larvae

ix DRAGON-FLIES 335

do not depend upon their spiracles even when the}'
are open, but are able to breathe in other ways
also.

The late Dr. Hermann Dewitz left behind him
some valuable experimental notes, which show,
among other things, that the respiration of Dragon-
fly larvae is effected in more than one way.1

He endeavoured, in the first place, to ascertain
whether the spiracles, and especially the large thoracic
spiracles, were as completely closed as had been
assumed. For this purpose he employed various
methods. Sometimes the larva was placed in more
or less dilute alcohol, when a train of bubbles issuing
from one (seldom from both) of the large thoracic
spiracles would often establish the fact that they \verc
pervious. Gentle warming, and even cooling, pro-
duced like results in many cases. When a thin glass
beaker, containing a live larva, was held in the hand,
the rise of temperature was often enough to cause
bubbles to issue from the spiracles. Useful informa-
tion was got also by placing living larvae in water
that had been boiled to get rid of dissolved air, and
afterwards cooled. The larvae were enabled to reach
the surface by climbing up a stick, and note was
taken of the way in which they came up. If they
came up head foremost and pushed part of the thorax
out of the water, this was a distinct indication that
they relied mainly upon their thoracic spiracles.
Some larvae, under these circumstances, come up
backwards, with the head down, and push the tip of
the abdomen out of the water — an obvious proof
ZooL Am. Bd. XIII., pp. 500-504; 525-531 (1891;.

336 NATURAL HISTORY OF AQUATIC INSECTS CH.

that the process of rectal respiration (to be explained
later on) is more efficient than tracheal respiration.
In other experiments, either the spiracles or the
rectum was closed by collodion.

By such means as these Dewitz ascertained that,
in young larvae of yEschnidae, the thoracic spiracles
are not pervious, though those of adult larvae, almost
ready for transformation, are. In boiled water young

/Eschnid larvae always
bring the tip of the abdo-
men to the surface, older
ones either the tip of the
abdomen or the thora-
cic spiracles. Young

FIG. 98.— Extremity of abdomen of larva ^Eschnid larvae die wheil
of ./Eschna. The plates are separated ,1 -111

to allow the admission of water. the TCCtUm IS Closed by

collodion ; older larvae

endure the stoppage of either rectum or spiracles,
but die if both are stopped.

In Libellulidae the fore pair of thoracic spiracles
are developed and become pervious much earlier than
in ^Eschnidae. In old larvae the spiracles are more
important as a means of respiration than the rectum.

Young larvae of Agrionidae have the thoracic
spiracles pervious, but not fully functional. Dewitz
thinks that the muscular apparatus for inspiration by
the spiracles is not yet developed. When placed in
boiled water they crept up the stick, but never
pushed any part of the body out of the water. The
surface-layer was apparently sufficiently aerated for
their purpose. Sometimes they waved the abdomen
to and fro, as if to increase the efficiency of their

IX

DRAGON-FLIES

337

tracheal gills. Advanced larvae are able to use their
spiracles, and when placed in boiled water, creep up
the stick head foremost, throwing the body into a
horizontal position when the surface is reached, so as
to expose the spiracles to the air.

FIG. 99. — Part of respiratory organ in rectum of /Eschna larva. One of the six
muscular bands occupies the middle of the figura. The ventral tracheal trunk
is seen giving off branches to the wall of the rectum. Another trunk, more
dorsally placed, is shaded. A third, which runs above the rectum, is omitted.

From all these observations it is clear that the
larvae of Dragon-flies, especially in their later stages,
fill their tracheal tubes directly from the atmosphere
to a much greater extent than had been previously
supposed.

338 NATURAL HISTORY OF AQUATIC INSECTS CH.

Though the spiracles arc often open during at
least the later part of the larval existence, they
cannot, of course, be employed when the Insect is
actually submerged, and other means of respiration
are required in addition.

An Agrionid larva is provided with three leaflets
at the tail-end, which in Calopteryx are nearly as
long as the abdomen. These are useful in swimming

X 6~0

FIG. ioo. — Part of three rows of respiratory folds from cuticular living rectum of
^T.schna larva. The shaded parts are abundantly supplied with tracheal tubes.
The leaflets appear to be connected with a central trachea, but this is not really
the case.

but microscopic examination shows that they serve
for respiration also. Each leaflet is covered with a
network of air-tubes, and, we can have little doubt,
absorbs dissolved air from the water, passing it into
the main tracheae, which supply the chief organs of
the body. In the other families the abdomen ends
in five valves, three of which are commonly larger
than the rest, and can be brought to a point or widely
separated at pleasure. When they open, they disclose
the outlet of the intestine, which is guarded by three

IX

DRAGON-FLIES

339

fleshy folds, corresponding in position to the three
valves. Within these is a somewhat capacious
chamber (the rectum) whose wall exhibits a compli-
cated and interesting structure. Six thick longitu-
dinal bands, separated by thin and flexible membranes,
seem to be intended to allow great distention without
risk of stretching the delicate epithelium or cellular
layer. Each longitudinal band bears a double row
of transverse folds, which
enormously increase the
epithelial surface, and at
the same time lodge
tracheal branches. Oustalet
estimates that there are
no fewer than 24,000 of
these folds. The tracheal
branches enter larger and
regularly arranged air-
tubes, which in turn open
into the main trunks run-
ning the length of the body.
A large volume of water

can be sucked into the rectum at pleasure, and from
this the tracheae draw a fresh supply of oxygen.
The vitiated water can either be expelled gently, or,
when the larva requires to propel its body forwards,
with considerable force. An Agrionid larva, Calop-
teryx, has similar rectal gills, besides external
tracheal gills.

Under strictly natural conditions, Dragon-fly larvae
may increase the effectiveness of their respiratory
organs by means already mentioned in connection

Z 2

FIG. 101. — A small part of one leaflet,
highly magnified, showing many
fine tracheal branches. The por-
tion shown is marked by a small
circle in Fig. 100 (lower left-hand
corner).

340 NATURAL HISTORY OF AQUATIC INSECTS CH.

with Dewitz's experiments. Thus, Agrionid larvae
sway their abdomen from side to side, and ^Eschnid
larvae often come to the surface and take air directly
into the intestine. These expedients are most fre-
quently resorted to when the water is foul.

The escape of the Dragon-fly from its larval skin
is described by Reaumur in his usual lively style,
and I think that the reader will like to have his
account as nearly unabridged as possible.1

" Most, perhaps all, of the larvae of Dragon-flies,"
says "Reaumur, " must live ten or eleven months
beneath the surface of the water before they can
undergo transformation ; there may possibly be
Dragon-flies produced in autumn from eggs laid in
spring ; if so, the larvae, enjoying the season most
favourable to growth, develop faster than the rest.
[Such cases are not certainly known to occur, though
they are probable enough ; certain large Dragon-flies
are believed to live two and even three years as
larvae.] However this may be, from April to the end
of September, and even to the middle of October,
Dragon-flies emerge every clay.

" The change which is to convert an aquatic Insect
into an inhabitant of the air takes place out of the
water. All the larvae, however, which one sees either
wholly or partially out of the water are not absolutely
ready to become winged. Often when they have

1 The original is to be found in the eleventh memoir of
Reaumur's sixth volume. Here as elsewhere Reaumur is not
very careful to distinguish the various species described, but
many of his observations seem to have been made upon yEschna
cyanea, which is very common both in France and in England.

ix DRAGON-FLIES 34 1

crept an inch or two out of the water they return
thither, after taking a breath of air ; but those which
have made a longer excursion, it may be several feet
upon the land, and especially those which are found
clinging fast to stems or branches, are making ready
to cast their skin and to become flies.

" I have found larvae belonging to the same species
which underwent their transformation an hour or two
after issuing from the water, while others passed a
whole day before liberation. The operation takes
some time, but the observer will be so much occupied
by the interest of the details, that he will not willingly
leave it half finished. One can read, so to speak, in
the eyes of the larva, whether it is ready for trans-
formation or not. Within a quarter or half an hour
of the time of liberation the eyes, which were pre-
viously dull and opaque, become bright and trans-
parent. This change does not really take place in
the larval eyes themselves, but is due to the pressing
close against them of the eyes of the fly, which have
already acquired all their brilliancy.

" In order to enjoy again and again the spectacle
of the transformation, it is desirable to collect a con-
siderable number of larvae of the same species and
to keep them in water. When the cast skins show
that some of the flies have emerged, the margin of
the water should be searched at different hours of the
day for larvae which have crept out. They commonly
rest some time in order to dry themselves thorough ly
before proceeding further. At length the larva begins
to seek a place convenient for its operations. It often
grasps a plant and holds on to it with its head

342 NATURAL HISTORY OF AQUATIC INSECTS CH.

upwards. The two hooks with which each leg is
provided secure the body in its place.

" When the larva has thus fixed itself, internal
changes take place. The first visible effect is that
the skin cracks along the back of the thorax, and the
thorax of the fly becomes visible through the cleft.
The part which first appears swells, and, like a wedge,
helps to enlarge the cleft, which extends to the fore-
edge of the thorax, then along the neck, and finally
to the level of the eyes, whence lateral clefts proceed
on either side. At this time the fly is able to dilate
its head like a Blue-bottle fly under like circum-
stances. In proportion as the cleft becomes pro-
longed, more and more of the body of the fly is
exposed. At length the head is completely with-
drawn, and appears so large, that one has difficulty
in believing that it could immediately before have
been enclosed within the larval head. "When the
head and thorax are free, the legs are drawn out of
their sheaths, and to disengage these more readily,
the fly arches itself backwards. At this time there
can be seen on each side two white strands, which
remain attached to the thoracic part of the larval
skin. These are [the linings of] the four great
tracheal trunks, and are drawn out from the four
thoracic spiracles. After this the fly throws her head
further and further back, until at length the body is
supported only by the hindermost segments, which
still remain enclosed in the larval skin.

" At this time the legs of the fly are completely
free and distant from their sheaths. For two or
*;hree minutes they are moved about, as if to try their

ix DRAGON-FLIES 343

flexibility. Then they come to rest again, and are
kept perfectly still. \Yhcn I first saw an emerging
fly thus motionless, I thought it dead or dying, ex-
hausted perhaps by its efforts. It remained more
than a quarter of an hour in this position, and I have
seen others motionless for nearly half an hour. I
was about to cease observing in despair, when I dis-
covered that the fly had been merely waiting till its
limbs had gained the necessary firmness, and till its
strength was restored. The next operation required
much exertion. The body, which was previously
bent backwards, was, by a sudden spring, arched in
the opposite direction. The head was brought back
to the top of the sheath, the legs clasped the sides
of the cleft, and attached themselves firmly. It is
plainly essential that the hooks should become com-
pletely rigid before this manoeuvre could be executed.
After this the Dragon-fly had to extricate the hinder
part of her body. She bent the abdomen double,
withdrew it completely, and then extended it in a
straight line.

" Though the Dragon-fly was now free, it had a
very different appearance from those which range the
fields. It seemed deformed. The abdomen, though
longer than the sheath from which it had issued, had
not yet acquired its full length. The wings seemed
little larger than when they were enclosed in their
sheaths. They were turned edgewise, laid side by
side, and folded up like a fan, or like a leaf in the
bud. Not only were they folded along their length,
but also transversely. The wings expanded so rapidly
that it was difficult to get a faithful drawing made of

344 NATURAL HISTORY OF AQUATIC INSECTS CH.

them. At this time the fly carefully avoids spreading
her wings, for though they afterwards become firm as
sheets of talc, they are at present soft as wet paper.
It is important to avoid the slightest derangement,
or even contact with one another. The abdomen is
carefully bent into such a position as to avoid touch-
ing them. As the wings expand, we can see the
veins spreading further and further apart and the
folds becoming effaced. The expansion is apparently
due to the injection of liquid into the veins. When
I took a fly which had perished during extrication, I
found that the wings were soft, and could be expanded
by the fingers ; but as soon as they were released,
they returned to their original folded condition.
The expansion of the wings is usually completed in
less than a quarter of an hour, but when first expanded
they are not yet firm or dry enough for flight. More
than two hours are often necessary before the wings
can be spread out horizontally, and two or three
hours more are required before they are able to
support the weight of the body. All the time that
the wings are expanding, the abdomen is being
gradually prolonged. When the Dragon-fly first
emerges, its colours are faint, but they gradually
increase in distinctness and brilliancy."

The process of egg-laying has its difficulties, for,
while the female Dragon-fly is apparently ill-fitted to
enter water, the larvae which escape from the eggs
cannot endure more than a short exposure to air.
The plan pursued seems to vary in the different
species. It is said that some Dragon-flies merely
drop the eggs, singly or in groups, into the water.

IX

DRAGON-FLIES

345

In other species the female, accompanied by the male,
enters the water, holding on to submerged plants,
and deposits the eggs in convenient places. Several
species make incisions into succulent aquatic plants,
and lay a single egg in each incision. Diplax is said
to lay her eggs in a gelatinous rope.

CHAPTER X

POND-SKATERS, WATER-SCORPIONS AND WATER-
BOATMEN (RHYNCHOTA)

TllE order Rhynchota or Hemiptera is distinguished
by the absence of conspicuous transformation and by
the suctorial proboscis. The peculiar texture of the
wings indicated by the word Hemiptera — half of the
fore-wing being membranous, while the basal half is
firmer— is not found throughout the order. These
Insects are active, and seek their food in all stages.
The larvae have no wings until shortly before the
emergence of the imago, when, as -in Dragon-flies and
many other Insects, stunted wing-cases appear, which
enclose the perfect wings of the adult. In some
Rhynchota, hence called Homoptera, both pairs of
wings are alike in texture and usually membranous.
In the Heteroptera the fore pair are firmer than the
hind pair, especially towards the base, and form
elytra or wing- covers, somewhat like those of Beetles.
To this sub-order belong a number of very common
aquatic Insects. They are all predatory, feeding
upon small Insects or Crustaceans, and, like most
other Heteroptera, emit a secretion of unpleasant

CH. X

POND-SKATERS

347

smell when crushed or
seized. Many Heteropoda
are dimorphic or even poly-
morphic, that is, within
one species are found two
or even three kinds, irre-
spective of sex. The adult
fern ale, especially, may have
, short win^s, or

' '*^*

no wings at all. Dimor-
phic species are common
among the aquatic Heter-
optera. It is noticeable
that none occur among the
tree-haunting species of the
sub-order. Wingless forms
may have their advantage
in species which live and
breed in water, or lead a
comparatively stationary
life on land. They are not
for one thing tempted to
incur the greater risks of
the air. But at least an
occasional winged form is
necessary for migration to
fresh sites.

Just as in the carnivorous
Coleoptera we find some
(Gyrinus) which skim over
the surface of the water,
and others (Dytiscus, etc.)
which seek their prey in

FIG. 102.— Hydrometra stagnorum,
and rostrum of do.

348 NATURAL HISTORY OF AQUATIC INSECTS CH.

FIG. 103. — i, Gerris thoracica (the centre of the thorax should be paler) ; 2, wing-
less imago of Gerris najas ? ; 3, hind wing of G. thoracica ; 4, rostrum of
do. ; 5, transverse section of abdomen of G. najas.

x POND-SKATERS 349

the water below, so among the aquatic Rhynchota
some run on the surface without being wetted, while
others arc almost constantly submerged. The surface-
life is that practised by Hydrometra and the Pond-
skaters.

Hydrometra, sometimes called the Water-gnat, is
distinguished by its peculiar shape and habitat. The
whole Insect is very slender, and the head is drawn
out to a length of one-third that of the body.
Mounted on long slender legs, the Insect walks
sluggishly on the top of the water, or creeps about
the grassy banks.

Closely related to this are the much more lively
Pond-skaters (Gerris). Like Hydrometra they have
a narrow, elongate body, but the head is not so pro-
longed. The fore-legs are prehensile, and said to be
used for capturing and holding living prey.1 As often
happens in such cases, they are separated from the
next pair by the prolongation of the first segment of
the thorax on which they a\re borne. The second
and third pairs of legs are long and slender, and
support the weight of the body. When the Pond-
skater darts about the surface of the water, the inter-
mediate legs are most active, while the hind pair
steer the course. Not only the legs but the whole
body is covered with a velvety pile, which renders it
incapable of being wetted. Though the legs do not
break the surface-film, they depress it, making little
dimples in the water. When the Insect stands on the
surface of a shallow sandy pool in the sunshine the

1 This is De Geer's account. I believe, however, that Gerris
is timid, and prefers a victim already dead or nearly so.

NATURAL HISTORY OF AQUATIC INSECTS en.

FIG 104. — 6, advanced larva of Gerris sp. with wing-cases and incomplete abdomen ;
7, larva, probably of wingless form ; 8, abdomen of half-winged imago ; 9, fore
tarsus, Uvo-jointed ; 10, n, extremities of lancets ; 12, inner pair of do., inter-
locking and forming a tube.

X

POND-SKATERS

351

FIG. 105. — Very young larva
of Gerris, natural size, and
magnified.

dimples cast shadows on the bottom, each surrounded
by a bright ring, due to refraction of the rays which
pass through the curved surface.
But the central shadow, cast by
the body of the Insect, has no
bright ring, and when the animal
moves, no bright ripples appear
about the central shadow7, such
as accompany the movements
of the legs. The body does
not actually touch the water.1
Meinert observes that Gerris has
three modes of moving about on
water, (i) drifting by wind or
current, (2) skating, and (3)
leaping. A small species, G. lacustris, can leap a few
inches on the surface of water, but a greater distance
on solid ground. Gerris dives occasionally but not
often, and never when avoiding pursuit. An allied
species, Velia currens, though less nimble than Gerris,
swims under water more readily. At times it can be
seen to run, back downwards, on the surface-film,
completely immersed in the water ; its abdomen at
such times glistens with the air-bubble which over-
spreads it.

All these surface-Rhynchota live by sucking the
juices of dead or dying Insects blown or in some
other accidental way brought to the surface of the
water. The beak is driven into the bod)7, and kept
there for a considerable time until the available juices
have been extracted.

1 Meinert, Saertryk af Entomol. Mcddelelser, 1887.

352 NATURAL HISTORY OF AQUATIC INSECTS CH.

Since so much depends upon the efficiency of the
legs, it is not surprising that the Pond-skaters spend
a good deal of time in cleansing and brushing them,
often standing meanwhile upon three legs only. They
winter as adults, and may be seen darting about on
water in cold weather. The larvae are plentiful in

FIG. 106. — i, Velia currens, wingless form ; 2, fore tarsus of do. ; 3, rostrum ;
4, 5, lancets ; 6, transverse section of abdomen.

summer ; they arc to be seen more often than the
adults diving beneath the surface. The hinder seg-
ments of the larval abdomen are small and retracted,
so that at first sight the abdomen seems to be wanting
altogether. The Pond-skaters are commonly, if not
always dimorphic, the adults being either long-winged

WATER-SCORPIONS 353

or short-winged. The eggs are enveloped in mucilage,
and attached to submerged plants.

We pass next to those aquatic Rhynchota which
seek their food beneath the surface of the water..
Though primarily adapted to a submerged life, they
leave the water at times to seek a fresh haunt, and
fly very well. These excursions are usually made at
night.

The family Nepidae (Water Scorpions) includes
two very common forms, Nepa and Ranatra, which
agree in the possession of fore-legs completely modi-
fied as instruments for seizing live prey. The antennae
are short and concealed. The respiratory tube con-
sists of two long spines. Each spine is hollowed, and
forms a demi-canal. When brought together, they
make a tube, whose continuity is maintained by a
multitude of hook-like bristles which project from the
opposed edges. The tube conducts air to a pair of
spiracles situated near the hinder end of the body
and at the base of the tube ; these are the only
apertures for the admission of air. A Nepa or
Ranatra may sometimes be seen to creep backwards
along a submerged weed until the tip of its breathing-
tube breaks the surface of the water. The mechanism
for interlocking the two halves of the tube closely
resembles that found in the proboscis of a moth.
The fore-legs of Water-scorpions are solely prehensile.
They are strong and pointed, but without the usual
small terminal claws. They can be folded in two
like a clasp-knife. The fore-legs of Mantis, a preda-
tory Orthopterous Insect, furnish a close parallel in
many ways to those of the Nepidae. The Water-

A A

354 NATURAL HISTORY OF AQUATIC INSECTS CH.

scorpions are rather sluggish in their movements, and
creep about in a leisurely way on the bottom of
ponds. Nepa feeds mostly on small Insects, Ranatra
upon the Water-flea (Daphnia) and other aquatic
animals. The form of the body differs remarkably
in the two genera. Nepa is broad and flat, Ranatra

n

FIG. 107. — A, Water-scorpion, Nepa cinerea. B, fore leg, enlarged, showing
groove in femur which receives the rest of the limb. C, egg.

extremely elongate. Ranatra is nearly black ; Nepa
of a dark ash-colour, and very inconspicuous except
when the wings are spread, when the top of the
abdomen is seen to be of a brick-red. The eggs of
both are laid one by one in holes pierced in the leaves
of aquatic plants, or in notches cut in submerged

stalks. From one end of the egg stand out filaments

X

WATER-BOATMEN

355

(seven in Nepa, two in Ranatra), whose minute struc-
ture shows that they supply the egg and embryo with
air. The fresh-hatched larva of Ranatra has no
respiratory tube, and perhaps does not take in
gaseous air at all. In the larvae of all Nepidae the
tube is short.

The remaining aquatic Rhynchota which we have
to describe are the Notonectidae or Water-boatmen.
These Insects are very common in stagnant waters,

FIG. 108. — Water-boatman, Notonecta glauca.

one of them, Notonecta, occurring in profusion where
decaying organic matter provides subsistence for the
animals on which it preys. In the Water-boatmen
the hind-legs are developed as oars, and become
flattened, fringed with hairs, and useless for loco-
motion on land. Of the two genera most commonly
met with, one, Corixa, swims about like a Water-
beetle, with the back upwards, but the other, Noto-
necta, with the back downwards. Like most other

A A 2

356 NATURAL HISTORY OF AQUATIC INSECTS CH.

aquatic Rhynchota they fly very well. Corixa flies
at night, and has been seen to fly into an open
window in August, being attracted by a light. The
head in both genera is transversely elongate, the eyes
large, and the antennae concealed, as in Nepidae.
The mouth, as in other Rhynchota, is suctorial, and
armed with a stout proboscis, rather long in Noto-
necta, short in Corixa. The proboscis is piercing, and

FIG. 109. — Rostrum and setae (mandible and toothed maxilla) of Notonecta glauca.

capable of inflicting a slight wound on the hand of
anyone who seizes the Insect. The body is velvety
and overspread with a glistening air-film, which is
conspicuous on the under-side, but elsewhere hidden
by the wings. Notonecta is inflated by air to such
an extent that it is very buoyant ; it has to cling to
weeds or other fixed objects when it wishes to remain
below, and as soon as it relaxes its hold it rises to
the surface. This uncommon buoyancy is probably

X WATER-BOATMEN 357

•

connected with the peculiar attitude of the Insect.
The air is so lodged in and around its body as to
make the ventral side the lighter, and this floats
upwards. When it reaches the surface, the tips of
the expanded oars and the tail press against the
surface-film and keep the head submerged. In this
position the Notonecta breathes at ease, but is ready
to make a sudden spring without suffering any delay
from the difficulty of wetting its body. Now and
then it swims back upwards for a few strokes, but
soon assumes its ordinary position. Corixa is much
less buoyant, and can remain as long as it pleases at
the bottom of a glass bowl, where there is nothing to
cling to. The hindermost abdominal spiracles, we
might expect, would be large in Notonecta, consider-
ing that when it reposes, only the extremity of the
body is in contact with the air. In reality they are
very minute. Larger spiracles, well defended by
hairs to prevent the accidental entrance of water, are
found on the sides of the thorax. To these the air
is led by a singular passage. The side of the body
which floats uppermost is keeled along the middle
line. On each side of the keel and between it and
the lateral edge, runs a long row of elastic hairs,
while a second and parallel row runs along the edge
itself. These rows of hairs enclose a watertight
covered way, leading to the thorax, and along it the
air is guided to the large thoracic spiracles. Now
and then one of the long oars is drawn over the
abdomen, as if to urge the air more rapidly along
the passage.1 Corixa breathes in a different fashion.
1 Schmidt- Schwedt, in Tier- und Pflanzenwelt des Suss wassers.

358 NATURAL HISTORY OF AQUATIC INSECTS CH.

The tail is not brought to the surface, but the air
passes directly into the thoracic spiracles. In cold
weather Corixa, like other aquatic Rhynchota, buries
itself in the mud at the bottom of a pool, and is often
smeared with mud when it reappears in early spring.1

FIG. no. — Corixa Geoffroyi.

Schmidt-Schwedt says that Corixa uses its fore-legs
to play a tune upon the snout. Captive specimens
were observed to utter a tolerably loud and sustained
note of an evening. The Insects were at such times
submerged. He observed that the note was produced
simultaneously with the movement of the fore-leg.

1 Douglas, Ent. Mon. Mag., Vol. III. p. 25.

x WATER-BOATMEN 359

Notonecta is said to make a noise like the word chew
repeated three times running. The Insect at such
times rubs its fore-legs together.1

Bruyant has lately described a similar music exe-
cuted by the very minute Sigara minutissima.2 This
little Water-boatman occurs among Myriophyllum
and Ceratophyllum in the lake of Chauvet (Auvergne)'
and has also been found in the freshwater sponge.
Though only a millimetre (^ in.) long it makes a
very distinct stridulation, and the characteristic sound
led to the discovery of the Insect in a new site. The
middle and hind legs are adapted for swimming.
The fore-legs are very short, and end in a single-
jointed tarsus, which is armed with stiff hairs. These
play upon the rostrum and make a noise like that of
a comb rubbed along a sharp edge.

Sigara is not the only aquatic Insect which inhabits
the freshwater sponge. The larva of Sisyra, a
Neuropterous Insect, is very frequently found in the
larger pores. A Caddis-worm (Leptocerus) burrows
in the substance of the sponge, and in summer an
undetermined Dipterous larva is occasionally met
with in the same situation.3

The process of egg-laying by Notonecta has been
studied by Regimbart.4 The Insect attaches itself
firmly by its fore and middle legs to a submerged
stalk, buries its rostrum in the plant, as if to gain
additional support, and makes an incision uith the

1 P. Rcdfern, in Brit. Assoc. Report, 1859. Sections, p. 173.

2 Comptes rendus, 1894.

3 Weltner, in Tier- tend Pflansenwelt des Siisivassers,

4 Ann, Soc. Entom. France, 1874, p. 204.

360 NATURAL HISTORY OF AQUATIC INSECTS CH.

armed ovipositor. The work is executed by a back-
ward and forward sawing motion, and occupies about
a minute. Then the egg is passed into the incision,
about one third of its length remaining outside ; the
exposed extremity corresponds to the head of the
future embryo. It will be seen that this is nearly
the same procedure as that adopted by the female
Dytiscus (p. 41, Fig. 3).

The eggs of Corixa are in general glued to sub-
merged objects. In certain species they are very
numerous and densely massed. This is particularly
the case with the eggs of two Mexican species (C.
mercenaria and C. femorata) which have long served
as an article of food to the Mexicans. They are
gathered in the lakes of Chalco and Texcuco, which
adjoin the city of Mexico. Reeds are set in bundles
in the shallows of the lakes, and upon these the
Water-boatmen lay their eggs, which are gathered
and detached by beating. They are made into cakes
with meal, and are said to have an agreeable acid
taste. Great deposits of former layings of eggs are
said to form at the bottom of the lakes, and to grow
steadily in extent.1

A Sialid Insect of North America (Corydalus
cornutus) forms large flat, rounded masses, composed
of two or three thousand eggs, on the leaves of trees
overhanging water. An extinct Corydalus of Tertiary
age has left countless fossil eggs in the freshwater
beds of the Laramie group, Colorado. The eggs are
massed into elongate, cigar-shaped bundles, which
are 5 cm. (2 inches) long. Each is estimated to

ie^ Tom. III. p. 846*

x WATER-BOATMEN 361

contain about two thousand eggs. The fossil eggs
were probably laid in water, like those of the Mexican
Corixae, and it is not unlikely that the lakes of
Mexico may hereafter yield masses of fossil eggs
like those of the Laramie beds.1

1 Scudder, " Tertiary Insects of North America," U.S. Geol.
Surv. of the Territories, 1890.

CHAPTER XI

THE WATER SPRING-TAIL (PODURA)

I WILL begin the account of the Water Spring-tail
by a short notice of De Gecr, the Swedish Reaumur,
to whom we owe the first account of this animal,
and also a host of accurate and interesting observa-
tions on almost all kinds of Insects.

Baron Charles De Geer was born in 1720. He
came of a Dutch family long settled in Sweden.
One of his ancestors, Louis De Geer, came over from
Holland in the time of Gustavus Adolphus, and in-
troduced better methods of smelting iron than had
previously been known. He set up foundries for
cannon, brass-works and manufactories of small
arms. He brought Belgian workmen to the cele-

o o

brated mines of Dannemora, and formed them into
a colony which lonq; retained its foreign character.

J o o

Louis De Geer acquired great wealth, which he spent
freely upon public and benevolent objects. He was
ennobled, and founded a family distinguished for
power of more than one kind. The naturalist, Charles
De Geer, was brought up in Holland. It is said
that his interest in Insects was roused at an early

CH. xi THE WATER SPRING-TAIL 363

age by some silkworms given him to rear. He
studied at Utrecht, and afterwards at Upsal, where
he was one of the pupils of Linnaeus. He inherited
large property from an uncle, and, among the rest,
succeeded to an interest in the mines of Dannemora.
These had been inundated by water, and De Geer
contrived machinery to free them and render them
workable again. He was skilful in the management
of his large estates and liberal in his gifts, especially
to charities and schools. His Mcmoires pour servir
a r Histoire des Insectcs occupy seven large quarto
volumes with many plates, and include descriptions
of upwards of 1,500 species. De Geer is a less
pleasing writer than Reaumur, whom he endeavoured
to imitate, but more concise and methodical. He
had the advantage of early training in the system of
Linnaeus, and instead of the picturesque but at times
rather indefinite names employed by Reaumur, he
employs a precise nomenclature. De Geer is always
painstaking and trustworthy, and many of his de-
scriptions are clear, spirited, and full. The first
volume of his JMcmoires appeared in 1752, and we
are told that its sale was so discouraging that he
burnt many copies to relieve his indignation. When
he decided, in spite of this want of public interest,
to carry on the work, he sent free copies of the later
volumes to those who had been so enterprising as to
subscribe for the first, The first volume is now very
rare, and the difficulty of obtaining the entire work
has appreciably hindered the knowledge of De Geer's
researches. The German translation is much used in
place of the original French. De Geer died in 1778,

364 NATURAL HISTORY OF AQUATIC INSECTS CH.

and his seventh and last volume appeared shortly
afterwards in the same year. This last volume gives
his views upon the classification of Insects. He
regarded the characters drawn from the wings as of
primary importance, but attended also to the structure
of the mouth-parts and to the mode of transformation.
His system was on the whole the best of its century.

The first paper ever published by De Geer l was
written at the age of twenty. In it he describes four
small Insects belonging to what is now called the
order of Thysanura.2 The Thysanura are all of small
size and wingless ; they undergo no transformation ;
many leap about by means of a forked appendage
attached near the end of the abdomen to its under-
side, and are hence called " Spring-tails." Leeuwen-
hoeck had previously found and described Thysanura.
but De Geer's account is fuller and more exact. Of
his four species two, now called Podura aquatica and
Isotoma palustris, were found on the surface of water,
De Geer's account of Podura aquatica is here quoted
in brief.

In the month of February he observed good- sized
black patches floating on small pools of water.
Closer examination showed that the patches were
made up of small Insects clustering together and
moving incessantly. Though resting on the water,
they were never wetted. When touched with a stick
the Insects dispersed, but gathered together again as

1 Ada Soc. Reg. Scient. Upsal, Vol. I. (1743), p. 279. The
paper was read in 1740.

2 Seven species are described by De Geer in the last volume
of his Memoires (1778).

XI

THE WATER SPRING-TAIL

365

soon as the stick was taken away. They were only
one twelfth of an inch long, and entirely black. The
body was cylindrical, divided into segments, and
tapering sharply at the hinder end. The head
carried a pair of short, mobile, four-jointed antennae
and a number of simple eyes. The legs were short,
and ended in single claws. The tail, as De Geer
very naturally calls the forked appendage, could be
bent forwards beneath the body,
and then, being forcibly ex-
tended, could throw the Insect
like a skip-jack into the air.
De Geer remarks that the spring
remained extended for a little
while after the Insect alighted,
and was then gently drawn for-
wards. The act of leaping
seemed to him clumsily per-
formed, for the Podura generally
fell on its back, and wriggled
about to regain its natural
position. When undisturbed, it
moved on the surface of the
water by a slow creeping action.

White specks were seen floating on the water
among the black masses, and these De Geer found
to be cast skins of the Podura.

De Geer gives a clear figure of the forked spring,
and notices also a little rounded prominence between
its branches. This is now considered to be an organ
of adhesion, and De Geer himself so describes it at
a later time in another Thysanuran. Many of the

FIG. in. — Podura aquatica.
X 30. From Lubbock.

366 NATURAL HISTORY OF AQUATIC INSECTS CH.

terrestrial species are able to support themselves on
smooth vertical surfaces by help of this adhesive
organ, which opens by a slit, and gives passage to a
fleshy sucker.

The aquatic Spring-tail, says De Geer, perished in
two or three hours when removed from the water,
and its body shrunk as if by evaporation. He was
inclined in 1740 to suppose that the tubercle mentioned

FIG. 112.— Isotomapalustris (Aquatica cinerea of De Geer). a, the forked appendage
or spring ; b, the adhesive organ, or ventral sucker, seen also between the hind
legs in the upper figure ; c, claws at extremity of leg. x 60.

above sucked up water for the supply of the body
when the Spring-tail floated on its native pools.
Having placed some of these Insects in a vessel
completely full of water, he observed that they crept
to the bottom, and lived for several days. They did
not apparently possess any power of swimming.

The smaller and less common Isotoma palustris is
also described, but De Geer gives little information
as to its habits.

It is easy to find aquatic Podurae in ditches and

xi THE WATER SPRING-TAIL 367

ponds, for they are plentiful and widely distributed.
During the summer months they spring, like Sand-
hoppers, from the surface of the water, and then
repose tranquilly upon it. They are black, with a
mixture of white individuals, which are fresh-moulted.
How is it that they are never wetted by the water ?

The microscope shows that the body is covered
with short and close-set hairs. Into the minute
spaces between these the surface-film cannot pene-
trate, and the trifling weight of the body is insufficient
to press it as a whole beneath the surface. Were the
body once wetted, it would be hard for the Insect to
force its way through the surface-film into the air.
It is equally hard for it to make its way through the
surface-film into the water.

What has the size of the creature to do with its
power of swimming or sinking ? you may ask ; the
size surely does not affect the specific gravity of its
body. No, but it affects the proportion which the
contour of its body bears to its weight. In similar
figures the contour increases or diminishes directly
as any linear dimension, but the weight as the
cube. A minute Insect has accordingly a far
greater contour in proportion to its weight than a
large Insect of precisely similar form, and therefore
encounters greater resistance from the surface-film.
The resistance in the case of the aquatic Spring-tail
is, as we have seen, so great that it can leap from the
surface of the water and alight upon it again without
ever breaking through. I see that in various books
on Natural History this power is attributed to the
extraordinary elasticity of the spring. But no elas-

368 NATURAL HISTORY OF AQUATIC INSECTS CH.

ticity would suffice if the Insect were of large size.
Take a stone of half a pound weight, and throw it
upon water ; it sinks at once. Crush the same stone
to fine powder, and scatter a small quantity upon
the water ; the lighter particles will float for a long
time.

Can the aquatic Spring-tail descend beneath the
surface ? I was long persuaded that it could not.
In my aquarium this species abounds, and propagates
itself from year to year, feeding, as I suppose, upon
Duckweed and other floating vegetation. I have
never once found it beneath the surface. But out of
doors and in wintry weather I believe that it descends,
as so many other aquatic Insects do, for shelter from
the cold. It may also, I have some reason to sup-
pose, feed on submerged vegetation when the surface
is free of weeds. I once brought home a collecting
bottle full of decaying vegetable matter and water.
When I came to look at it closely, two or three
Podurse were floating on the surface. After an hour
or two there were eight or ten. This led me to think
that they must have come up from the depths where
they had previously been lurking. Once at the
surfao , they were unable to descend again. I placed
a number in a beaker half full of water and saw them
run and leap upon the surface. I chased them with
a small rod till they became excited and alarmed, but
they were wholly unable to escape by sinking. Even
when alcohol was added to the water, and the surface
tension lowered thereby, the dead bodies of the
Podurae floated as before. It was only when they were
placed upon strong alcohol that they became wetted,

XI THE WATER SPRING-TAIL 369

and after a considerable time were seen to sink. But
where water-plants with floating leaves grow, the
Podura^ can descend at pleasure. They will not
attempt to do so if they are well off at the surface ;
but when food is scarce, or the temperature low, they
may occasionally be seen to drag themselves through
the surface-film by grasping the stem with their
strong and hooked legs, and so descending. When
the Podura climbs downwards into the water it is
still unwetted, and enclosed by a film of air which
clings to its hairy body. It can remain for many
hours submerged without inconvenience.

Like some aquatic Beetles the Podura probably
buries itself for weeks together in the mud at the
bottom of ponds, and thus escapes the severity of
winter cold.

The small, metallic-coloured, carnivorous Flies of
the family Dolichopodidae commonly seek their prey
in the neighbourhood of streams and pools. Hence
the naturalist In search of aquatic Insects cannot fail
to find them almost daily and hourly, sometimes in
swarms, sometimes singly. They come to rest on the
grasses, herbs, or bushes near to water, on stones in
the bed of a stream, or even on the surface of the
water itself. Some rival the Pond-skaters in the
agility with which they dart to and fro upon the
surface of rapid streams ; others hover incessantly in
the spray of waterfalls. At least one species of
Dolichopodidae preys upon Podura aquatica, and it
is probable that many flies fresh-hatched from a
variety of aquatic pupae fall victims to these swift
and destructive enemies.

B B

CHAPTER XII

INSECTS OF THE SEA- SHORE

IT has been a surprise even to professed naturalists,
even to men who have made Insects the study of
years, to discover that Insects have contrived to
occupy not only the surface of the earth, the soil
beneath the surface, the air, the freshwater streams,
and the freshwater lakes, but even the sea beach and
the sea itself. Seventy years ago, that is, previous to
Audouin's study of the habits of Ae'pus, very few
entomologists had thought it worth while to search
for Insects among the refuse cast up by the tide.
That air-breathing Insects should betake themselves
to places where they might be covered by the salt
tide, seemed incredible, even to those who were
familiar with the many Insects which haunt our
rivers and ponds. And yet, even before Audouin's
time, Lyonnet, Treviranus, and Straus-Diirckheim
had observed the singular power of certain Insects
to endure complete immersion. Their observations,
however, failed to impress the minds of naturalists.
The supposed impossibility of Insect-life on the sea-
shore between tide-marks was based upon no experi-

CH. xii INSECTS OF THE SEA-SHORE 371

mental evidence, but merely upon vague impressions.
Had it come into the minds of the naturalists of 1820
to make the trial which Plateau made in 1872, they
would have better appreciated the hardiness of ter-
restrial Arthropods, and the ease with which they
submit to hours and even to days of submersion in
water. Plateau took a number of land Beetles of
various kinds, devoid of special organs for carrying
down a supply of air, and found that they could
endure complete immersion in fresh water for two,
three, or even four days. The Beetles became torpid,
but recovered on being restored to the air.1 Many
aquatic Insects, which dive with the greatest ease, do
not survive more than a very fe\v hours under like
conditions. Gyrinus struggles violently when detained
beneath the surface, and perishes in about three
hours, while Aphodius becomes perfectly still and
senseless, but recovers after an immersion of fifty
hours.2

Insects submerged in water behave as a rule like
Insects placed in an atmosphere devoid of free oxy-
gen. They cease to struggle in a few moments, and
pass into a state of torpor, which may be greatly
prolonged without fatal results. Certain larvae and
pupae live for weeks or months buried deep in earth
or in vegetable refuse. It is almost impossible to
kill certain Insects in a moderate time by carbonic

1 Rech. physico-chimiques sur les Articules aquatiques (Bull.
Ac. Roy. Belg. 1872).

2 Gyrinus is an Insect of much more active habits than
Aphodius, and the aquatic Insects experimented upon by
Plateau are in general more active than the terrestrial species.

B B 2

372 NATURAL HISTORY OF AQUATIC INSECTS CH.

acid or nitrogen, though they quickly become un-
conscious and motionless in such gases.

The saltness of sea-water might be expected to
prove disagreeable if not injurious to Insects, but
there is little proof that such is actually the case.
Insects, when forcibly submerged, survive about as
long in salt water as in fresh. Many Insects are not
easily wetted by water. The hairs with which some
are covered and the dense, glossy chitin of others
prevent effectual wetting. The surface-film of water
will not pass into small openings, such as the mouth,
or the spiracles, or the spaces between close-set Hairs.
Larvae however, which live in water, and are tho-
roughly wetted by it, cannot endure the change from
fresh water to salt. I placed thirteen living Chiro-
nomus larvae in salt water, and found that in four
hours five had died, while the rest were extremely
languid. In eight hours all were dead. By gradual
acclimatisation even such larvae as these can make
themselves at home in salt water. Packard dredged
up live Chironomus larvae in Salem harbour, and not
a few Dipterous larvae of various species have been
found established in brine-vats. Plateau has drawn
up a list1 of near eighty species of Insects and
Arachnida, which, though they cannot swim and
though they breathe only gaseous air, inhabit the
sea-shore, and undergo daily or at least frequent
submersion. The list will no doubt be largely
increased by the future labours of naturalists.

The Insects of the sea-shore are mostly very small.

1 " Lcs Myriopodes Marins," Jour, de VAnat. et Phys.

(1890).

xii INSECTS OF THE SEA-SHORE 373

Very few species attain such a size that they can be
seen at a distance of four or five feet ; a length of
four millimetres is somewhat uncommon.

On the verge of the beach, especially where it is
flat and sandy, there will commonly be found a line
of black sea-weed, cast up by the spring-tides. In
the damp and fermenting mass beneath the surface
of the sea-weed, innumerable maggots thrive, and
produce throughout the year black Flies, which are
rather like the common House-fly, but have longer
wings and smaller bodies. The Flies sometimes
emerge in dense swarms, but more commonly settle
here and there on the heap of weed, especially in
cold weather. They occasionally fly a few miles
inland and visit flowers. The black pupae are buried
in the weed. This is the Dipterous Insect known
to entomologists as Coelopa frigida. The maggots
and pupae easily survive a short immersion in salt
water, to which they are occasionally, but not fre-
quently, subject.

Another Dipterous fly is very common on the
shore. This belongs to the same large family,
Muscidse, as the House-fly, but to the Acalypterate
division, and is named Actora aestuum. It is of small
size and abounds in summer on the wet sand and foam
at the very verge of the tide. The larvae are be-
lieved to feed upon thrown-up seaweed (Fucus vesicu-
losus), and are covered at every tide. The body is
ten to fifteen millimetres long, cylindrical, and covered
with short hairs. It is a good deal like the Leather-
jacket (larva of the Daddy-long-legs), and like it
bears a number of conical fleshy processes at the

374 NATURAL HISTORY OF AQUATIC INSECTS CH.
tail end. Ei^ht of these form a semicircle on the

o

dorsal side, and in their midst are two spiracular
plates, each bearing three spiracular openings. Four
others stretch horizontally along the chord of the
semicircle, and another pair lie one on each side of
the anus on the ventral side.1

A small Chironomid, probably ThalassomyiaFrauen-
feldi, is common on our shores. The larva has been
described by Johnston as an Annelid, under the name
of Campontia cruciformis. It closely resembles the
Blood-worm of our ditches, except that it has a green
instead of a red tinge. A specimen of the larva has
been dredged off the Isle of Man from a depth of
over ten fathoms, and Mr. Swainson has often found
them on Zoophytes upon the sea-shore at St. Anne's-
on-the-Sea. The flies are often plentiful on the edge
of the retreating tide.'2

Several species of the brilliant, raptorial Dolicho-
podidae (see p. 369) haunt the surf and breakers of
the sea-shore, where no doubt they find their prey.

The Beetles of the sea-shore include a host of small
Carabidae, predatory Insects, which lurk under stones
or burrow in wet sand. The terrestrial species have
been known to enter water in pursuit of their prey.
Bertrand, for instance, observed Carabus clathratus
chasing aquatic Insects in water five centimetres
deep.3 The Carabidae of the sea-shore usually under-

1 H. Gadeau de Kerville, Soc. Entom. de France, Vol. LXIIL,
p. $2 (1894). The fly has not been reared from the larvae de-
scribed, but it is probably that of Actora aestuum.

2 Swainson, Brit. Naturalist, June, 1894.
,3 Quoted by Plateau, loc.

XII

INSECTS OF THE SEA-SHORE

375

go daily immersion, and become motionless when
covered by water.

Two European species of Aepus have been more
carefully studied than the rest of the marine Cara-
bidae. They run about on stones, sea-weed, sponges,

FIG. 113. — Aepus Robinii.

etc., at low water. Being like most other Carabidae,
incapable of flight, they cannot avoid the rising tide.
As soon as it reaches them, they creep under stones
and remain motionless. The body is flattened, and
covered in every part with hairs which entangle air
(Audouin). There is a large pair of air-sacs in the

376 NATURAL HISTORY OF AQUATIC INSECTS cil.

A

abdomen, as I learn from Mr. Hammond, which arc
no doubt useful during prolonged submersion. The
eyes are very curious. A chitinous plate protects and
almost entirely covers them, leaving only a small
round central hole. The form of this plate suggests
that it may be employed as a kind of pin-hole
camera. Mr. Hammond, who called my attention
to this peculiarity, has drawn it in fig. 114. The
tibia of the fore leg is provided with a peculiar comb-
like organ (sec fig. 116), which
looks as if it mistfit be useful for

o

cleaning the antennae or limbs.
Such an apparatus is not uncom-
mon in Beetles. I am indebted
to Mr. W. F. Baker for the fol-
lowing remarks upon the tibial
comb :-

" If the fore leg of the common
Beetle, Pterostichus niger, be ex-
amined closely, a deep notch pro-
tected by a large spine will be
found on the lower side of the
tibia ; this notch is armed with
stiff, fine hairs. I was for a long
time unable to imagine what pur-
pose it could serve. After inspect-
ing all the Beetles in my collec-
tion, I found that this peculiar
feature was restricted to the Geo-
dephaga, though in some of the
Staphylinidae a somewhat similar structure, consist-
ing of a spine and fine; hairs without any notch, could

FIG. 114. — Eye of Aepus.
A. External surface
showing facets. B,
Perforated plate, front
view. C, section.

XII

INSECTS OF THE SEA-SHORE

377

be observed. Among the Geodephaga species of
Broscus, Callistus, Harpalus, Amara, Calathus,
Anchomenus, Clivina, Dyschirius, Chleenius, Ptero-
stichus, Stomia, Aepus, Lori-
cera, Demetria, and Bembi-
dium have the tibial comb
in its complete form. All the
species which possess it fre-
quent moist, dark and dirty
places, hiding under stones or
logs of wood, burrowing in
the earth, or creeping under
the decaying bark of trees.
The comb on the fore leg of
the Honey-bee suggested that
the similar structure in Beetles
might be employed for the
purpose of cleansing the an-
tennas from dirt contracted

in confined situations. To test this, I smeared with
gum the antennae of Pterostichus niger, and set
the Insect free in a box whose bottom was strewn
with fine sawdust. After running wildly about and
seeking to escape the Beetle made a careful examina-
tion of its prison, waving its antennae, and feeling the
ground with them continually. The sawdust soon
began to clog the antennae, and then the Beetle would
stop, raise its fore leg, bring the antennae into the
comb by a movement of the head, and draw it through,
thus effectively clearing it of loose particles. One
piece of sawdust clung so tight that three distinct
pulls were required to remove it. The Beetle then

FIG. 115. — Abdomen of Aepus in
outline showing air-sacs in situ.

378 NATURAL HISTORY OF AQUATIC INSECTS CH.

endeavoured to cleanse its comb with the tarsus of
the middle leg, but went about it in a very clumsy
fashion.

" I was then satisfied that the comb is really used
to cleanse the antennae. I next endeavoured to find
out how Beetles which have similar habits but no

comb cleanse their antennae. It would
take too long to relate all the experi-
ments which I have made, but to be
brief, they do what is necessary by
the help of the spines on the tarsus
(Carabus), or those on the tibia or
tarsus (Notiophilus, Leistus, and
Nebria), or by rubbing the antennae
against plants and other objects
(Elaphrus). The operation often takes
up a considerable time where no
comb can be employed."

Aepus is often found among small
molluscs (Rissoa, Lasaea) and is be-
lieved to prey upon them. Another
little Carabid, Cillenum or Bembi-
dium laterale, can often be seen in
thousands upon refuse washed up
by the tide. It is regularly submerged.

The family of Staphylinidae (Rove-beetles) in-
cludes several marine species. Bledius is common on
sandy shores near high-water mark. There are fif-
teen British species. They prefer a smooth expanse
of sand thinly coated with mud, and here throw up
little hillocks, half an inch or so in height. Hundreds
of these hillocks form one colony. On sandy shores

FIG. 116. — Part of
fore leg of Aepus.
with comb upon
tibia.

xii INSECTS OF THE SEA-SHORE 379

the patches inhabited by Bledius may be so numerous
as to form a nearly continuous line miles in length.
The Beetle shovels out the sand backwards with
its hind legs. When covered by the tide, which
is an unusual event, it remains still in its burrow.
On warm evenings just after sundown they fly about
in great numbers, never rising high, but keeping near
the surface of the sand. The same species which
haunt the shore occur also in marshy ground in inland
places near the sea. Bledius arenarius is about three
and a half millimetres long ; its body is narrow and
rather cylindrical.

A Carabid Beetle, Dyschirius, preys upon the larvae
and pupae of Bledius. It is a restless creature, con-
stantly running in and out of the galleries, and has an
insatiable appetite. It is of about the size, or smaller
than its victims. Several species, both of Bledius and
of Dyschirius, occur on our coasts. A number of
other marine Staphylinidae have been described.
Micralymma is common on rocky shores, and re-
sembles Ae'pus in its habits.

A few marine Caddis-worms are kno\vn. One is
found between tide-marks in Lyttleton Harbour, New
Zealand, and forms its sheath of the coralline sea-
weeds.1 The Rev. A. E. Eaton has found plenty of
small Hydroptilidae in salt or brackish streams on
the borders of the Sahara. Mr. Swainson has fished
up a living Oxyethira from the incoming tide of
the Ribble, but it is not yet proved that this Caddis-
worm is a true inhabitant of the sea.2 Philanisus

1 MacLaehlan,/<?tf77Z. Linn Soc., " Zoology," vol. XVI. p. 417
(1883), ? British Naturalist* Dec,, 1894,

380 NATURAL HISTORY OF AQUATIC INSECTS CH.

plebius and a species of Molanna are additional
instances.1

Not even the waters of the sea can exclude Hy-
menopterous parasites. M. Moniez2 has discovered
a minute Proctotrupid under stones on the sea-shore
at Aigues-Mortes in company with marine Crustacea.
The habits of this Insect have not been investigated.

o

Aepophilus, so called, because it is often found in
company with Aepus, is a small Bug, only three
millimetres in length, which inhabits fissures of rock
not far from low water. The adult Insect is found
only in the month of October.

A number of marine Thysanura have been dis-
covered. There are also False-scorpions, Acari, and
at least two Centipedes which inhabit the shore. It
is much to be desired that the mode of life of this
varied and interesting assemblage of marine animals
should be more closely inquired into.3

The Insects of the beach hardly ever swim. On a
wave-beaten shore the mere attempt to swim would
be extremely risky. But far out at sea Insects may
safely trust themselves to the surface of the waves.
Their action is not however to be called swimming-
They run on the surface of the water, like a Velia or
a Hydrometra.

The Insects in question are Halobates and Halo-

1 " Challenger" Reports, " Zool.," Vol. VII.

2 Rev. Biol. du Nord, 1894.

Professor G. S. Brady and I are endeavouring to collect
information concerning the Insects of the sea-shore, and would
value the assistance of naturalists who have opportunities of
observation.

XII INSECTS OF THE SEA-SHORE 381

batodes. Several species are known. They inhabit
all the great tropical and sub-tropical oceans, often
hundreds of miles from land, appearing on the surface
in calms, or when there is a swell without wind-waves.
They appear of a brilliant white, either wholly or as to
their legs, but this is due entirely to a bubble of air
which overspreads the close-set hairs of the integu-
ment. When removed from the water or examined
closely, they are seen to be dark-coloured. They run
rapidly on the sea, and are said to dive occasionally.
What becomes of them in stormy weather, and where
the eees are laid and hatched, has not been found out.

oo

They feed upon the floating bodies of dead marine
animals, and may be seen to run out from such
objects when alarmed by the approach of a boat.
These Insects belong to the Rhynchota (Hemiptera),
and in some respects come pretty near to such forms
as Hydrometra or Velia. They are peculiar in never
exhibiting a trace of either pair of wings. The fore
legs are of moderate size, the middle legs of enormous
length, and in Halobates fringed with hairs. The
hind legs are not so long as the middle legs. The
whole animal has a superficial resemblance to a long-
legged spider. The mesothorax, on which the middle
legs are carried, is extremely large, and the abdomen
of insignificant size.

CHAPTER XIII

THE CONTRIVANCES OF AQUATIC INSECTS

THE aquatic Insects described in the foregoing
pages have been selected on no philosophical prin-
ciples. They are merely such as I have happened to
come across in my rambles by pond and stream.
But even this chance collection of aquatic species
may yield interesting results if studied from a suitable
point of view. Every independent worker has his
own notions as to the point of view which is par-
ticularly important and interesting, and I fear that
we are apt to be unjust to those who do not share
our tastes. I will not again occupy time and space by
justifying my own preference, which inclines me
chiefly to study the wealth of contrivance which is
exhibited by these obscure creatures.

Let us first of all run over the modes of locomotion
practised by aquatic Insects. The Pond-skaters stand
or run upon the surface of the water, which they
dimple but do not break. The Water Spring-tail or
a small Gerris can leap from the surface-film, and
alight upon it unwetted.1 The Whirligig Beetle

1 A small Copepod Crustacean, Pontcllina mcditerranea, is
said to do the same. (Ostroumoff in Zool. Ans., October, 1894.)

CH.XIII CONTRIVANCES OF AQUATIC INSECTS 383

(Gyrinus) darts to and fro upon the surface, changing
its course every moment by slight adjustments of its
peculiar paddles, but when alarmed it dives into the
depths beneath. The larvae of the Gnat, Dixa, Peri-
coma, Anopheles, Stratiomys, and Hyd robins suspend
themselves from the surface-film by an unwettable
basin at the tail-end of their bodies, which admits
air to the spiracles, and at the same time allows the
head and jaws to sweep through the water in search
of food. Some aquatic larvae (Eristalis, for in-
stance) may be seen at times to creep upon the
underside of the surface-film. The same manoeuvre
is practised by other aquatic animals, such as the
leech-like Triclads, the Pond-snails, and Cyclas.
These have been erroneously said to creep upon the
air. A very simple experiment will show that the
body is at such times completely immersed. If a
little lycopodium powder is sprinkled over the water,
the light grains are not parted when the animal creeps
beneath them. In some cases, for instance in the
Triclads, the body is heavier than water, and sinks in
a moment if it is detached from the surface-film. A
drop of water let fall upon the creeping animal will
send it to the bottom. The Water-boatman (Noto-
necta) is buoyant, and comes to rest in an inverted
position, with its tail and the extremities of its oar-like
legs resting upon the surface. It is then well placed
for intercepting drowned or drowning Insects. When-
ever it chooses to leave the surface, its oars propel it
swiftly through the water, and the air is so distributed
throughout its body that it swims most easily in its
resting position, that is with its back downwards.

384 NATURAL HISTORY OF AQUATIC INSECTS CH.

The resting position, moreover, brings the spiracles
within reach of a fresh supply of air, though not in
this Insect actually to the surface. Dytiscus when at
rest has its spiracles fully exposed to the air. If
such Insects are disabled, they have only to cease to
struggle in order to gain the surface in that position
which is most convenient for breathing. Swimming
beneath the surface is accomplished with various
degrees of efficiency. The winged Dytiscus is per-
fectly adapted for submerged navigation. The even
contour, the glossy surface, the powerful oars, fringed
with bristles, working simultaneously, and feathering
during the return stroke, the steering action of the inter-
mediate legs, the arched back, and the angulated keel
will delight those who can appreciate a finished piece of
mechanism. The soldering of the broad bases of the
oar-like hind legs to one another, and to the back part
of the thorax as well, stiffens the body, and it may be
prevents the loss of power which would result from
the incessant squeezing out of water enclosed between
the basal joint and the body. I suspect that there is
a further mechanical advantage in this structure which
I am at present unable to explain. Polynema travels
about in the water by the help of its wings, like a
Penguin. Many long-bodied Dipterous larvae move
by a lashing action, striking the water sideways and
instantly reversingthe stroke. In Corethra and the Gnat
the lashing action of the larva is aided by a fin com-
posed of close-set bristles. The Simulium larva runs
about its network of silken threads in the water of a
rapid stream, holding on by one or both of its clusters
of hooks, and throwing out a new thread whenever

xin CONTRIVANCES OF AQUATIC INSECTS 385

necessary. Some Dragon-fly larvae swim by striking
the water with the long abdomen and its terminal
plates ; others make sudden rushes by the violent
expulsion of water from the intestine. Slow larvae
creep by legs, hooked tubercules, backward-pointing
hairs, and the like. Not a few pupae, both aquatic
and terrestrial, bear spiny rings which aid in creeping,
and enable them to make their way to the air when
the fly is about to emerge.

Or we may review the contrivances which enable
aquatic Insects to capture their food. The perforated
suctorial mandibles and the mouth-lock of the
Dytiscus larva, the mask of the Dragon-fly larva, the
" weel" which entangles the struggling victims of the
Phantom larva (Corethra), the clasp-knife mechanism
of the prehensile fore legs of the Water Scorpion, the
snare of the Plectrocnemia larva, the ciliary organs of
many aquatic Dipterous larvae, and the pharyngeal
strainer of the Eristalis larva are varied in design,
perfect in execution. But our admiration for these
wonderful implements is mingled with another feeling.
The weapons of carnivorous Insects are a little too
like a collection of instruments of torture.

The respiration of aquatic Insects may be alto-
gether like that of the terrestrial species, except that
the gaseous air to be taken into the body is not so
readily procured. We have seen how the wing-covers,
the felted back, and the dorsally set spiracles of the
winged Dytiscus combine to secure an ample supply
of air every time that the Insect floats or swims to
the. surface. Hydrophilus and Hydrobius employ
their peculiar antennae in breathing ; Notonecta its

c C

386 NATURAL HISTORY OF AQUATIC INSECTS CH.

oar-like legs. Donacia and perhaps the pupa of
Paraponyx tap the air-chambers of submerged plants.
The larvae of Ptychoptera and Eristalis have respira-
tory tails, which can be protracted or withdrawn, but
each has a mechanism peculiar to itself. The long
respiratory filament of the Ptychoptera pupa is alto-
gether different from both. Nepa and Ranatra
breathe by a pair of valves, which cohere to form a
tube, and conduct air to the spiracles. Other aquatic
Insects breathe only dissolved air, and this in many
different ways ; by the outer skin, or by bunches of
thin-walled tubes, occupied in some cases with
tracheae, in others with blood-spaces, or by flat plates
which can be waved to and fro, or by intestinal folds.
Tube-dwelling larvae have their own difficulties and

o

remedies. Chironomus larvae, Caddis-worms, and
some aquatic caterpillars which inhabit sheaths made
out of leaves, maintain a flow of water through their
tubes by incessant undulation of the body. This
operation is aided by peculiar structures, such as
fringes of bristles, retractile processes which regulate
the distance of the body from the wall of the tube,
and so forth. The Chironomus larva employs
haemoglobin as a means of storing up oxygen. The
spiracles of almost any Insect, whether aquatic or
terrestrial, abound in singular features, nearly all un-
studied at the present time. The spiracles of Dytiscus
are among the wonders of nature ; it would take a
lifetime to master their details.

Attack and Defence call forth yet more contri-
vances. Protective resemblance, concealment within
burrows, by webs, by portable cases, by fixed cases,

xin CONTRIVANCES OF AQUATIC INSECTS 387

by the transparency of the body, are a few of the arts
practised by aquatic Insects, either in self-defence or
as a means of pouncing unseen upon their prey.

The egg-laying of aquatic Insects is attended with
special difficulties, some of which spring from the fact
that the female fly is in general ill-fitted to enter the
element in which the earlier stages have to be passed.
These fresh difficulties are met by fresh contrivances.
The egg-ropes of Chironomus, the egg-raft of the
Gnat, the anchoring threads of the eggs of Ephemera,
the floating cocoon of Hydrophilus, are adaptations of
peculiar interest. Dytiscus, Notonecta, Ranatra, and
certain Dragon-flies lay their eggs in incisions made
in submerged plants. But even these carefully
hidden eggs are searched out by such egg-destroyers
as Polynema, which lay in them their own eggs, from
which proceed the parasites which will in the end
devour their undeveloped host.

I might go on to enumerate fresh contrivances
under such heads as the constructions of aquatic
Insects, the emergence of the winged fly, the defences
of resting pupae, and so on, but the subject is in-
exhaustible, and the heaping-up of examples proves
wearisome. In past ages all these would have been
cited as fresh proofs of a wisdom and beneficence
which we are unable even to comprehend, far less
to imitate. And whatever the cool scientific spirit
may have done to weaken that interpretation, the
beauty and completeness of natural adaptations come
upon us at times with overwhelming force. It is
with peculiar interest that we observe how physical
principles, the discovery of which is among the latest

388 NATURAL HISTORY OF AQUATIC INSECTS CH.

triumphs of human science, were long ago turned
to practical account by some of the most insignifi-
cant of Insects. Many of these natural contrivances
secure the safety, and increase the happiness of
living creatures. But there is a dark side to the
picture.

In the Theodicees and Natural Theologies, which
now fill whole libraries, we shall find little mention
of the murderous instincts of animals, or of the
cruel weapons which are placed at their service/ Yet
the ravages of carnivorous Quadrupeds, Birds, and
Insects are as obvious as any of the operations of
nature. The misery caused by internal parasites
is less conspicuous, but even more perplexing. The
Dragon-fly and the Dytiscus are not altogether
pleasing objects of contemplation, but the Ichneumons
seem to go beyond all that the cruelty of despots
has imagined. That innumerable Insects should be
eaten up alive by parasites, which spare the vitals of
their victims lest they should perish too soon, that
this miserable existence should be protracted until
pupation, so that the devourer should find a safe
shelter during its own resting-stage — these are facts
which show us that Nature is not everywhere bene-
ficent. Of the same kind are the facts related con-
cerning some of the fossorial Hymenoptera. I can
never read Fabre's Souvenirs Entomologiques without
a shudder. The operations which he has brought to
light with such diligence, and described with such
animation, impress me much as do the tales told of
the Well of Cawnpore or the Ice-tower of Avignon,
What may be the solution of the mystery, and how

XIII CONTRIVANCES OF AQUATIC INSECTS 389

so much benevolent foresight can be reconciled
with so much cruelty, it is not for the naturalist to
explain, though the mere naturalist finds it hard to
shake off these thoughts when they have once come
up in his mind.

The beginnings of mercy and unselfishness appear
in the behaviour of wild animals to their young.
Social animals are called upon for a more habitual
exercise of the same attributes, and it is perhaps the
social state which has chiefly moderated the selfish-
ness of Man ; it is the social instinct which leads
him to pity even the humblest victims of the struggle
for existence.

When we have to tell what we have seen and found,
it is our business to give a true account, disguising
nothing, and keeping nothing back. But let us be
careful not to speak as if our little plummets had
sounded the depths of the universe. Those who have
surpassed their fellows in the improvement of natural
knowledge, have always been the first to admit that
what they have come to know is lost in the infinitude
of the unknown.

INDEX

INDEX

ABRIDGED LIFE-HISTORIES, 20

Acentropus, 235

Actora, 373

Aepophilus, 380

Aepus, 375

Aeschnidse, 331

Agrionidse, 331

Agriotypus, 223

Alder Fly, 273

Anglers' Names for Aquatic

Insects, 28
Anopheles, 92, 163
Aquatic Beetles, 30
Caterpillars, 226
Flies, 97

Hymenoptera, 219
Insects, Anglers' Names for,

28

Contrivances of, 382
Equilibrium of, 15
Groups of, 28
Wintering of, 18

BAKER ON HYDROBIUS, 89

on Tibial Comb of Beetles,
376

Batelli on Eristalis larva, 203

Beetles, shapes of, 6
food of, 6

Bledius, 378

Buckler on Hydrocampa, 230
on Paraponyx, 232

Burgess on mouth of Dytiscus
larva, 45

CADDIS WORMS, 124, 223, 236
Marine, 379

Carabidae, Marine, 374
Cataclysta, 231
Caterpillars, Aquatic, 226
Ceratopogon, 155, 197
Chironomus, 122, 197, 251

Development of, 151
Ccelopa, 373
Collection of Aquatic Insects,

, IJ4
Contrivances of Aquatic Insects,

,382
Corethra, 113, 121, 141, 384,

385
Conxa, 355

Corydalus, Eggs of, 360
Culex, 97, 121, 141

DE GEER, LIFE OF, 362

on Caddis-worms, 263

on Ephemera, 318

on Mochlonyx, 121

on Paraponyx, 231

on Thysanura, 364
Development of Chironomus,

151

of Polynema, Ganin on, 222

Dewitz on Respiration of Dragon-
fly larvae, 335

Dicranota, 164

Diffusion through living mem-
branes, 38

Diptera, Aquatic, 97

Divided Eyes, 34

Dixa, 92, 157

Dolichopodidae, 369, 374

Dominance, Marks of, S'

Donacia, 93

D D

394

INDEX

Dragon-flies, 328
Dragon-fly larvae, 385

Escape from larval skin,

340

Dyschirius, 379
Dytiscus, 39, 89, 384

EGGS OF AQUATIC INSECTS,

146, 326, 360

Ephemera, De Geer on, 318
Ephemeridas, 285

Pictet on, 322

Vayssiere on, 325
Equilibrium of Aquatic Insects,

15

Eristalis, 198, 383

FLIES, AQUATIC, 97

GANIN ON DEVELOPMENT OF

POLYNEMA, 222
Gerris, 349, 382
Gnat, 97, 145
Great Water-beetle, 61
Gyrinus, 30

H/EMONIA, 93
Halobates, 380
Halobatodes, 380
Helophorus, 89
Hemerobius, 44, 45
Hurst on Gnat pupa, 104, 144
Hydrobius, 87
Hydrocampa, 227
Hydrometra, 349
Hydrophilus, 61
Hymenoptera, Aquatic, 219

IMAGINAL FOLDS, 135

Insects, Dominance of, 8
in Brine, 2, 372
of the sea-shore, 370
originally terrestrial, 4
variety of habitat of, 7

Invasion of the Waters by In-
sects, 3

Isotoma, 366

KLAPALEK ON AGRIOTYPUS,
22;

LAKER ON HYDROPHILUS, 86
Lepidoptera, Aquatic, 226
Libellulidce, 331
Live Natural History, 24
Lowne on Suckers of Dytiscus,

55
Lubbock on Polynema, 219

Lyonnet on Hydrophilus, 64
Life of, 62

MAY-FLIES, 205

McLachlan on Caddis-worms,

.257

Meinert on perforated mandi-
bles, 45

on Dixa, 159
on Tanypus, 155

Micralymma, 379

Micropteryx, 269

Miger on Hydrophilus, 73, 84

Mochlonyx, 121

Mosquito, 109, in

Moult of Insects, 134

Miiller, G. \V., on Hydrocampa,
229

Muscidre, 139, 197

Myrmeleon, 44, 45

NEFA, 353

Nepidse, 353

Nitzsch on Hydrophilus, 75

Notonecta, 383

Notonectidte, 355

ODONATA, 328

Osten Sacken on Oxen-born

Bees, 215
on Eristalis in America, &c.,

218
Oxyethira, 379

P^EDOGENESIS, 23

Palmen on Perla, 283, 284
Palingenia, 285
Paraponyx, 124, 231
Parthenogenesis, 23
Peri coma, 92, 218
Perla, 279

Phantom Larva, 114
Pictet on Ephemeridse, 322

INDEX

395

Plateau on Suckers of Dytiscus,

55, 59
Plectrocnemia, 265

Podura, 362
Polymitarcys, 304
Polynema, 219, 304
Pond-skaters, 349, 382
Prestwichia, 220
Ptychoptera, 170

RANATRA, 353

Rat-tailed Maggot, 198

Rtaumur, Life of, 236

on Egg-raft of Gnat, 112

on Eristalis, 198, 213

on Escape of Dragon-fly

from larval skin, 340
on Hydrocampa, 227
on Polymitarcys, 304

Respiration of Aquatic Insects,
118

Rhynchota, Aquatic, 346

SCHIODTE ON PERFORATED

MANDIBLES, 45
Schmidt- Schwedt on Corixa,

358

on Donacia, 94
on Notonecta, 357
on Paraponyx, 234
on retractile processes or

Trichoptera, 249
Scourfield on Copepods and

Surface-film, 162
Sea-shore, Insects of the, 370
Sialis, 273

Siebold on Donacia. 95
Sigara, 359
Simmermacher on suckers of

Dytiscus, 55
Simulium, 141, 175, 251

Sisyra, 359

Sponge, Insects living upon the

Freshwater, 359
Spring-tail, Water, 362, 382
Staphylinidae, Marine, 378
Stone-fly, 279
Stratiomys, 92, 189
Swammerdam, Life of, 285

on Gnat larva, 101

on Palingenia, 285

on Stratiomys, 189
Surface-film of Water, 12, 99,

367

TANYPUS, 141, 152

Taylor on Plectrocnemia, 265

on Retractile Processes of

Trichoptera, 247
Thalassomyia, 374
Thysanura, De Geer on, 364

Marine, 380
Tipula, 140
Trichoptera, 236

VAYSSIERE ON EPHEMERID^E,

325
Verdat on Simulium, 187

Viviparous reproduction, 23

WATER AS A SPHERE OF LIFE,

i

Boatman, 355
Scorpions, 353
Spring-tails, 302, 362
Surface-film of, 12, 99, 367
Whirligig Beetle, 30, 382
Wilkinson on Pharynx of Eristalis
larva, 206

on Tail of Eristalis larva,
20 1

PRINTED IN GREAT BRITAIN BY

RICHAKD CLAY & SONS, LIMITED.

13UNGAY, SUFFOLK.

\