^iiiiiiiHi

!imii%ti

ftiinfiiftfiiiHitituiiuuiiiti

iim f \\\y

ilii :

MONOGRAPHS OF

THE ROCKEFELLER INSTITUTE

FOR MEDICAL RESEARCH

No. 13 November 15, 1920

THE EARLY STAGES OF TABANID^ (HORSE-FLIES)

By
WERNER MARCHAND, Ph.D.

NEW YORK

The Rockefeller Institute for Medical Research

1920

Copyright, 1920, by
The Rockefeller Institute for Medical Research

WAVERLY PRESS

The Williams & Wilkins Company

Baliimore, U. S. a.

^

at

MONOGRAPH No. 13

THE EARLY STAGES OF
TABANID^ (HORSE-FLIES)

BY
WERNER MARCHAND, Ph.D.

NEW YORK

The Rockefeller Institute for Medical Research

1920

A. L. MELANOER

From the Library of
J. M. ALDRiCH

MONOGRAPH OF THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH, NO. 13,
NOX'EMBER 15, 1920.

THE EARLY STAGES OF TABANID^ (HORSE-FLIES).

By WERNER MARCHAND, Ph.D.

{From !he Department of Animal Pathology of The Rockefeller Institute for Medical

Research, Princeton, N. J.)

Plates 1 to 15.

(Received for publication, April 23, 1917.)

CONTENTS.

Introduction 4

History of investigations 7

Chronology 15

Tabanidas, description of early stages 17

Bionomics 27

Special anatomy of tabanid larvae; Graber's organ 29

Chrysops (Meigen) , early stages in general 44

Chrysops himacidosa Neave 46

" callidiis Osten Sacken 46

" celer Osten Sacken 47

" dispar Fabricius 47

" Indus Osten Sacken 48

" longicornis Macquart 48

" magnifica, var. inornata, Austen 49

" moechus Osten Sacken 49

" mcerens Walker 50

" relictus Meigen 52

" vittatus Wiedemann 52

" wellmani Austen 54

Dorcaloemiis f adieus Austen 55

Gaslroxides ater Saunders 55

Goniops (Aldrich), early stages in general 56

Goniops chrysocoma Osten Sacken 57

HcBmatopota, early stages in general 66

Hccmatopota crudelis Austen 66

" decora Walker 67

" insaiiabilis Austen 67

" pluvialis Linne 67

sp 73

Hexatoma pellucens Linne 74

1

2 THE EARLY STAGES OF TABANID^

Tabanus, early stages in general 77

Tahanus albimedius Walker 78

" atratus Fabricius 79

" atrimanus Loew 91

" aitkimnalis Linne 91

" bicallosus Ricardo 93

" biguttatus Wiedemann 94

" bovinus Linne 97

" bromius Linne 99

" carolinensis Macquart 101

" corax Loew 101

" cordiger Meigen 103

" coslalis Wiedemann 105

" desertus Walker ' 106

" ditcBniatus Macquart 106

" epistates Osten Sacken 109

" fraternus Macquart 110

" Jronto Osten Sacken 110

" (Atylotus) fulvus Meigen Ill

" fuscipes Ricardo Ill

" glaucopis Meigen Ill

" grains Loew 112

" hilaris Walker 112

" ignotus Rossi 112

" insignis Loew 114

" king' Austen 115

" lasiophthalmiis Macquart 119

" laverani Surcouf ' 123

" lineola Fabricius 123

" maculatissimus Macquart 125

" medionotatus Austen 126

" melanoceros Wiedemann 126

" nagamiensis Carter 126

" nigerrimus Zetterstedt 127

" nigrescens Palisot de Beauvois 127

" obscui'ipes Ricardo 128

" (Neotabamts) ochrophilus Lutz 128

" orientis Walker 129

" par Walker 129

" pertinens Austen 132

" quatuornotatus {quadrinotatus) Meigen 132

" semisordidus Walker 144

" solstitialis Schiner 144

" speciosus Ricardo 145

WERNER MARCHAND 6

Tabanus spodopterus Meigen 145

" striatus Fabricius 145

" stygius Say 163

" sulcijrons Macquart 167

" tcenioJa Palisot de Beauvois 168

" (Neotabanus) triangulum Wiedemann 171

" trimaculatus Palisot de Beauvois 171

" tropicus Linne 171

ustus Walker 172

" variabilis Loew 172

" virgo Wiedemann 1 73

" vivax Osten Sacken 1 73

Unidentified species of Tabanus 176

Tabanus sp. No. .1 176

" " " 2 176

" " " 3 176

" " " 4 176

" " " 5 177

" " " 6 177

" " 7 178

" " 8 179

" " " 9 179

" " 10 179

" " " 11 179

" " <' 12 179

" " 13 180

" « 14 181

Parasites of the early stages of Tabanidae 182

Hymenopterous egg parasites; early discoveries 182

Phanurus tabanivorus, Ashmead, n. sp 183

" tabani Mayr 184

" eniersoni Girault 185

Telenomus benefactor Crawford 186

" kingi Crawford 186

Unidentified parasites recorded 186

Chalcid sp. 1 186

" " 2 186

" " 3 186

Ichneumonid sp 187

Parasites of the larvae 187

Discussion of Table II 188

Table II. Statistical table of results on early stages 190

Notes on methods of rearing and studying tabanids in early stages 196

Bibliography 200

Explanation of plates 204

INTRODUCTION.

The Tabanidge are a group of blood-sucking insects which, though
of less importance to medicine than mosquitoes and certain groups of
flies, are nevertheless of sufficient interest to the student of animal
pathology to warrant a closer study and a completion of our meager
knowledge of the life history of these insects. As the group consists
of many species spread over all parts of the world, it is natural that
the work done hitherto has scarcely advanced beyond the systematic
and descriptive stage ; and the lack of information on development and
early stages is especially felt by experimental science. In fact, al-
though Tabanidae play a considerable part in transmitting diseases
of domesticated animals, and in spite of the trouble which their
presence in large numbers causes to horses and cattle even under
normal conditions, our knowledge of their life history is very incom-
plete compared with that which we possess of other groups of blood-
sucking insects; for instance, CulicidcT. Only recently the study
of tabanid life histories has been undertaken from the point of view
of their economic importance, by King, British Government Ento-
mologist in the Anglo-Egyptian Sudan, Mitzmain in the Philippine
Islands, and others. However, these authors practically ignore
previous or contemporary literature on the subject of their investiga-
tion. Thus Mitzmain states that the literature on the subject is
meager and quotes only three publications. When rediscovering
Graber's organ in the larvse studied by him, he assumes that the organ
is peculiar to this species or has been overlooked by the previous
authors. The Zurich Index (Concilium Bibliographicum), however,
contains reference to several publications of European authors on
this organ.

I have undertaken a review of the literature on the subject because
the entomological literature is especially difficult of access, being
scattered over a number of small periodicals appearing in different
languages. Field and experimental work on entomological subjects
is often done at a great distance from large libraries, and it would be
an advantage to the individual worker, especially in the tropics, to

4

WERNER MARCHAND 5

have all the facts on a definite subject concentrated in one publication.
I have therefore decided to give not only a summary of the literature
with the bibliography, as is usually done in text-books, but also a
complete report of all the facts found in these publications, with very
little change in the wording, literal quotation of all authentic descrip-
tions, and with the reproduction of practically all the existing illus-
trations on the subject, sufficiently accurate to make the consultation
of the originals unnecessary. By this means I hope to relieve the
worker on the subject in question of the necessity of consulting
libraries, except the future monographer of the whole group, who will
prefer to go back to the original pubhcations. On the other hand,
as a full bibliography is given, in case of uncertainty about any detail,
the originals can always be consulted and compared with this report.

There is, however, another consideration of some importance which
has guided me in preparing this review; that is, the desirability of
having, generally, all the facts on one subject of research concentrated
in one language, preferably English. The diversity of languages is a
great hindrance for the progress of science, and from the point of
view of methodical unification, specialists in any branch of science
should be encouraged to concentrate all the literature on their own
subject in the language which they prefer to have used. Having
given considerable time to the study of this question, I cannot see
any serious objection to the more general adoption of English in
science. It is to be hoped that the temporary disadvantages to the
non-English-speaking scientists, resulting from its adoption would be
made up by the advantages of a great progress in science which would
undoubtedly follow. It is for the scientists of English-speaking
countries to pave the way, by conquering the field systematically,
subject after subject, and to facilitate the production of new work
by a complete assimilation of work previously done in foreign
languages.

I wish to express my indebtedness to the authorities of the American
Museum of Natural History and of the Philadelphia Academy of
Natural Science for giving me liberal access to their libraries and
collections, and to Dr. Theobald Smith, Director of the Department
of Animal Pathology under whose guidance this work was rendered
possible.

6 THE EARLY STAGES OF TABANID^

In grouping the material I have as far as possible followed a system-
atic arrangement, by bringing all the facts which could be referred
with certainty to a definite species under the heading of that species,
even where the facts appear to apply to a whole genus or to tabanids
in general. In each species the egg stage is first treated, followed by
the larval and pupal stages. Results obtained by one author in one
paper have, consequently, often been separated and rearranged. On
the other hand, no attempt has been made towards extensive gen-
eralization; the general statements on the family Tabanidae and the
more important genera are only such as are found already in the
literature, and are not to be considered as final or as expressions of the
author's opinion.

In the arrangement of the plates a slightly different plan has been
followed, as it seemed desirable to facilitate the comparison of corre-
sponding stages. I have therefore figured on Plates 1 and 2 all the
existing illustrations of the egg stage of different species; on the fol-
lowing plates the larvae and larval structures; following these the
pupge; and finally pupal structures, these being of especial sys-
tematic value. With this arrangement the different stages of the
same species are necessarily separated and figured on different plates.
In the details, the arrangement has been such as seemed most con-
venient. Occasionally an exception to the general order has been made
and a group of figures illustrating one species or genus appears
on the same plate.

HISTORY OF INVESTIGATIONS.

In 1760 the Swedish naturalist Degeer published a paper, in the
transactions of the Swedish Academy of Science, on the larval and
pupal stages of the European horse-fly, Tabamis bovinus, which he
found to be terrestrial in habit, and which he described in a paper
bearing the title Bromsarnas ursprung (The origin of horse-flies).
Degeer was a contemporary and compatriot of Fabricius, the first
great entomologist, and both continued the work of Linne. It is
stated by various authors that Fabricius' works also contain data on
the early stages of Tabanida3, though very few, but I have not been
able to find such data in the Systema Entoinologice (1775) or in the
Entomologia Systematica. The statement that the larvas of Chrysops
are found in the ground cannot have been made by Fabricius, in this
form, as the genus Chrysops was first established by Meigen (1803).
If Zetterstedt makes this statement in Diptera Scandinavice, it is pos-
sible that it is from verbal information or from some later writings of
Fabricius. Zetterstedt seems to doubt it, stating that he has seen
large numbers of newly emerged Chrysops near the margin of a lake,
which would indicate an aquatic habitat. The Swedish school for a
considerable time dominated entomological work in all countries
Degeer's observations are reprinted in his Memoires pour servir a
Vhistoire des insectes, which appeared in French in 1776 and also in
German translation (Goeze, 1882). These observations are given
without change by Macquart in Histoire naturelle des insectes. Dipteres
(1834), and by Westwood in England, the illustrator of the Arcana
Entomologica, in his paper Introduction to the modern classification of
insects (1840). To the same school belongs Wahlberg, who in 1838
published, in Swedish, accounts of the larval stages and mode of life
of many Diptera, which also contain, according to Brauer, notes on
tabanid larvae living inside lepidopterous larvee, probably in a pseudo-
parasitic way. In one of his papers on this subject I could not find
any reference to Tabanida^.

The progress of entomology consisted chiefly in a greater specializa-
tion as to definite orders which were systematically studied and

7

8 THE EARLY STAGES OF TABANID^

described, and in the greater attention which was given to the larval
stages. A new school of investigation on this subject arose in the
German-speaking countries with Vienna as its center, where Schiner
had become an authority on the Diptera of Austria-Hungary, and
where Brauer devoted himself to a comparative study of insect meta-
morphosis, particularly of dipterous larvae. It is to be noted that
Scholtz, in 1848, in Breslau, on an excursion in which von Siebold also
took part, discovered some tabanid larv£E, indicating an influence of
the growing science of general zoology on this subject. In 1854
Mann discovered oviposition in Tahaniis aulumnalis, and Kollar, at
the Vienna Museum, observed the eggs of various other species of
Tabanus, stating that since Degeer nothing had been known about the
development of these insects. In the following decades great progress
was made in the knowledge of dipterous and other insect larvae,
chiefly through the Vienna school. In 1868 one of Brauer's students,
Marno, found the larvae of Hexatoma pellucens, a tabanid not rare in
Austria, this being one of the first instances known of an aquatic
larval stage in this family. This larva was described more fully in
1883 by Brauer himself. In 1869 Brauer was able to describe the
larva and pupa of Hcematopota pluvialis {Regenhremse) , which is
terrestrial, giving at the same time a careful drawing of the mouth-
parts with correct interpretation of all the details. At about the same
time Brauer was v/orking on the classification of dipterous larvae.
The larva of HcBmatopola pluvialis was, however, described also by
Pejris in France in 1870, who found it in rotten pine wood, and gives
a figure showing the fine striations which are not shown in Brauer's
figures; and the same larva was described in 1875 by Beling, inspector
of forests in the Harz mountains, both apparently independent of
Brauer's studies. Beling was also the first to describe the pupa of
Chrysops, in 1882. On the other hand, von Friedenfels, who found the
larva of Tabanus auiumnalis in the salt lakes of Siebenbiirgen, belonged
to the Vienna naturalist group, and the larva which he had supposed
to be an annelid was identified by Brauer. The results of all these
investigations are briefly summarized by Brauer in 1883 in the third
volume of Die Ziveifiiigler des Kaiserlichen Museums zu Wien, in which
the larvae of Tabanus solstitialis , spodopterus, and bromius, and the
larva of Hexatoma pellucens are described.

WERNER MARCHAND 9

During the period just mentioned the microscopic study of lower
organisms had made continuous progress and the anatomical investi-
gators soon began to discover favorable objects of research in the
histology of insects. The study of sense organs of lower animals had
since Johannes Miiller become a favorable subject of research. On
the other hand, the native fresh water fauna furnished the most varied
and suitable objects for microscopic study, and this fact led to re-
peated observations on tabanid larvae from the point of view of general
zoology, the determination of the species at hand being omitted with
an almost disconcerting regularity. We have, however, noted that
tabanid early stages were observed, by Scholtz in Breslau in 1848, and
even the species was identified, as adults had hatched from the pupae
collected. In 1878 Graber, in Czernowitz, Austria, in his studies on
the sense organs of insects, discovered a peculiar otocyst-like organ
in a Tabanus larva, called Graber's organ by later authors, a dis-
covery which he connected with that of similar organs in the larva
of Ptychoptera by the zoologist Grobben. His paper called forth an
article by Krauss in 1879, a pupil of Brauer in Vienna, according to
whom this organ had already been demonstrated by Brauer in the
tabanid larva in his zoological course. Krauss also declared the larva
defined by Graber as a fly-maggot to be a Tabanus larva, and to belong
to Tabanus autumnalis. Whether the species was really this or a
similar one, no one can say at present. However, to Graber belongs
the credit of the independent discovery of this organ and its minute
description. In a later work (1882) on the chordotonal organs of
insects, Graber figured another tabanid larva with many anatomical
details, demonstrating in it the chordotonal organs which had been
discovered by Leydig in Corethra plumicornis. These discoveries,
especially the otocyst-like organ, gave rise to scientific controversy,
Lecaillon in France (1905 and 1906) assuming that the organ must
in reality be a gland, while Paoli (1907) in an extensive work under-
takes to prove that it is not an auditory, but a sound-producing organ.
At the same time Paoli advanced an interesting theory about the
manner in which the organ is developed, though without reference to
phylogeny.

With Lecaillon and Paoli an influence of economic entomology,
especially of agriculture, is notable in stimulating research, unques-

10 THE EARLY STAGES OF TABANID^

tionably a reaction on Europe from America. In America research on
the early stages of Tabanidae begins with Walsh (1863), who belongs
entirely to the old type of naturalist like Westwood and others, de-
scribing facts of nature half as pioneers of discovery in strange coun-
tries, half as pioneers of rehgion, and filled with a deep admiration of
the wisdom manifesting itself in creation. He is contemporary with
Osten Sacken, the first systematic specialist of American dipterology,
who published his Prodromus to a monograph of North American
Tabanidag. In Walsh's publication of previous work only Degeer
is referred to, and Walsh believes that his larva, which was proved
later to be that of Tabanus atratus, was the first instance of an aquatic
tabanid larva. In fact, though the observations of Wahlberg, Zetter-
stedt, Scholtz, and Kollar were not known to Walsh, these authors
give only indications but no full evidence of aquatic habits of life of
the larvag observed by them. While a member of the Boston Nat-
ural History Society, Walsh lived for some time in Illinois, where
tabanids were numerous and annoying, so that the economic side of
the subject inevitably received attention. We may say that with him
begins the American or economic school of investigations on this
subject. How curiously the economic considerations were mixed
with a profound confidence in the wisdom of nature may be seen from
the following quotation from his paper, with which few modern in-
vestigators will be found to agree, except perhaps the last sentence:

"The scheme of the creation is perfect, and nature is never at fault. It is
only when nature's system is but half understood that we heedlessly complain
of its imperfections. We blame the house-flies for annoying us, and fail to see
that in the larva state they have cleared away impurities around our dwellings,
which might otherwise have bred cholera or typhus fever. We execrate the
blood-thirsty mosquito, and forget that in the larva state she has purified the
water, which would otherwise, by its malarial efHuvia, have generated agues and
fevers. In all probability, when we rail at the Tabani, which torment our horses
in the summer, we are railing at insects which, in the larva state, have added
millions of dollars to the national wealth, by preying upon those most insidious
and unmanageable of all the insect foes of the farmer — subterraneous root-feed-
ing larvae."

The larva of Tabanus atratus is redescribed by C. V. Riley in 1870
in the second of his Missouri Reports, this being the first publication

WERNER MARCHAND 11

on a tabanid larva issued from an agricultural institution. Riley
apparently was interested in all insects and their development, and
in fact included in his Reports also some insects of no economic
importance.

A great step in advance was made when in 1895, Hart, in Illinois,
probably interested by Walsh, who had lived there, but probably also
not without some knowledge of the studies of the Vienna school, under-
took to investigate the entomology of the Illinois River on a large
scale. He described large numbers of new dipterous larvae, and inci-
dentally a number of tabanid larvae, notably those of Chrysops
vittatus, Tahanus stygius, nigrescens, lineola, and costalis, which were
not known before, and giving a preliminary classification by which
these larv£e could be separated. Hymenopterous egg parasites of
Tabamis, previously seen by Kollar, were then for the first time ex-
actly described by Ashmead, who cooperated with Hart. I do not
think that Hart could have achieved his results without a knowledge
of Brauer's monograph, which had, in fact, appeared ten years before,
and which, as stated, contains references to most of the previous
w^ork. But as he, apparently following Agassiz's principle, "Study
nature, not books," makes no mention of the previous literature on
this subject, I assume this to be the reason why later authors, Hine,
and even Lecaillon, have lost the thread of tradition pertaining to
their subject, and why important results such as the descriptions
of numerous larvae and the discovery of Graber's organ have been
overlooked by American and European authors. Shortly afterwards,
in 1899, the insect volume of the Cambridge Natural History, by
Sharp, appeared in England, which also does not mention the litera-
ture, but contains the description of an unknown tabanid larva,
referred to by Sharp as Tahanus {? Atylotus fulvus) .

In America new discoveries of tabanid early stages were made by
Hine, who, continuing Osten Sacken's work in this field, has become
the leading authority on the Tabanidae of North America, especially
from the systematic point of view. In 1903 he described the Hfe
history of Tahanus vivax; this paper was followed by a publication on
the Tabanidae of Ohio, which also contains notes on the early stages.
It is worthy of note that Hart's as well as Hine's first publications were
not issued by economic but by scientific institutions (Illinois Biolog-

12 THE EARLY STAGES OF TABANID^

ical Survey, Ohio Naturalist, Ohio Academy of Science) ; later the
connection with agriculture was established, first in Mine's Report on
the Tabanidae of the Gulf Coast (1903), then in his Habits and life
histories of some flies of the family Tabanidas (1906), in which the
early stages of Tabanus lasio phthalmus , and some others, are given
for the first time. The Tabanid£e, long known as a stock pest, were
at about that time suspected of carrying infectious cattle diseases
(anthrax and surra), as was evident from Salmon and Stiles' Emer-
gency report on surra published in 1902 in connection with a small
outbreak of surra in the United States following the importation of
Indian zebu cattle. The interest taken by entomologists in Hine's
work called forth some smaller publications in America, as in 1908
that of Walton, containing descriptions of the early stages of Goniops
chrysocoma, a peculiar species, and in 1909 that of Brimley, with notes
on the early stages of several other Tabanidse, one of them Tabanus
fronto, being terrestrial in habit. On Goniops chrysocoma, under the
auspices of the United States Department of Agriculture, a more
detailed study was made by McAtee in 1911.^ On the other hand,
in methods of control of Tabanidae, some progress was made by the
work of Portschinsky in Russia, in 1908; Lecaillon in 1905 in France;
and Paoli in 1907 in Florence, also undertook studies on Tabanidae,
which apparently were stimulated by work already done on this sub-
ject in America, Lecaillon referring to Hart, and Paoli, while resuming
the problem of Graber's organ, working at an agricultural school where
he undoubtedly received Hine's publications. Lecaillon's work (1905),
which acquaints us with the egg-laying habits of Tabanus quatuorno-
tatus, cites the observations on this species made by Kollar in 1854.
The progress made in the meantime by the Vienna school is, however,
ignored by him, which we understand if we realize that the wars of
1866 and 1870 had destroyed cooperation to some extent. The only
new author cited by Lecaillon is Hart, who, as we have stated, did
not quote the literature. At the same time, Lecaillon (1905) and
Hine (1906) are perhaps the first writers in whom the influence of
medicine on entomology is perceptible. The part that insects play
in the transmission of disease, the theory of v/hich is intimately con-

^ Specimens of the larva of this species had been found before by Pergande.

WERNER MARCHAND 13

nected with the discoveries of Theobald Smith (1893), Bruce (1895),
Ross (1897), Grassi (1900), and Reed (1901), had attracted attention
to all blood-sucking insects as possible carriers, and the influence on
the study of Tabanidae showed itself in the progress of knowledge of
their early stages. In the following period extending up to the
present, studies on these have been made largely under the auspices
of medical entomology. The progress resulting is considerable, the
number of publications on the subject within the last fifteen years
being about equal to that of papers published previous to 1900.

To the interest aroused in the study of Tabanidae from the medical
point of view we owe in the first place a knowledge of life histories of
tabanids of tropical countries, chiefly Africa and India, where the
importance of these flies as carriers of disease is paramount. In 1908
King was sent by the British -Government to the Sudan in charge of
economic entomology in the Wellcome Research Laboratory in Khar-
toum. He goes into the subject more extensively than any of the
previous authors, working out the life history of the black African
horse-fly, Tabanus bignUatus (1908) and of Tahanus ditceniatus,
tceniola, kingi, and par (1910), succeeding for the first time in causing
tabanids to oviposit in captivity, and giving minute descriptions of
larval structures, which appear cumbersome but may well serve to
separate the species described from others not yet known. At
about the same time we learn of tabanid larvae in India from Maxweil-
Leffroy and Howlett, Indian insect life (1909), a work, however,
written more from the point of view of agricultural interests, while
Baldrey and Mitzmain studied tabanids in captivity with the imme-
diate object of determining their part as carriers of disease. Baldrey
(1911-12) had Tabanus orientis oviposit in captivity, but made no
observations on the larvas. Mitzmain has made a very complete in-
vestigation on the life history of a single tabanid species, Tabanus
striatus (1913). Mitzmain worked in the Philippine Islands and dem-
onstrated experimentally the transmission of surra (1913) and anthrax
(1914) by this species. His omission of reference to previous litera-
ture has been commented upon, but as his work was done far away
from large libraries he was naturally handicapped in this respect.
While Mitzmain raised his flies entirely in captivity, Bainb ridge
and Fletcher (1914) reported the oviposition of the same species in the

14 THE EARLY STAGES OF TABANID^

free-living state. The text-book of medical entomology by Patton
and Cragg (1913), who worked in Madras, again gives descriptions
of egg and larval stages of various Indian tabanids, with information
on habits, methods of rearing, etc., giving also for the first time a
more complete idea of work previously done. The organ discovered
by Graber (1878) is illustrated and discussed by these authors;
Paoli's work on the same subject, however, seems to have remained
unknown to them. At about the same time (1914) Lutz in Rio de
Janeiro published the first meager observations on early stages of
tabanids in Brazil, having found larvae of Tahanus {Neotahanns)
ochrophilus and triangulum, and of one undetermined species. The
studies of Neave in the following year, 1915, furnish abundant infor-
mation on early stages of African tabanids, and new studies on South
American tabanid larvae were published by Bodkin and Cleare in
1916. Some new species of egg parasites have also been described in
recent years by Crawford from King's material, and by Girault (1916)
from Tahanus eggs from Dallas, Texas.

The various text-books on medical entomology and related subjects
which have appeared recently, besides those quoted by Bainb ridge
and Fletcher and by Patton and Cragg, those published by Goldi
(1913), Herms (1915), W. A. Riley and Johannsen (1915), and
Griinberg (1907), also Brumpt's Precis de parasitologic (1910), con-
tain a good deal of information on tabanid early stages, including also
quotations from previous literature.

While most of the recent work has been done by British and Amer-
ican authors, new and accurate studies on the oviposition of Tahanus
quatuornotatus have been made by Lecaillon (1911), and another
paper by Picard and le Blanc (1913) reports the early stages of another
species. In the latter year there also appeared a paper by del
Guercio on the tabanids of the rice fields of Bologna, which, however,
contains little of scientific value.

The complaint made by some writers of the meagerness of our
knowledge of the life history of this group, notably by Goldi, will
probably soon be unfounded, if our knowledge progresses as rapidly
as it has in the last few years.

South America, which is extremely rich in interesting tabanids,
and Australia, where until now apparently no work has been done,

WERNER MARCHAND 15

should yield many new discoveries. On the other hand, European
investigators have neglected the Tabanids somewhat, and it would
seem desirable to reinvestigate the larvae of the common Tabanus
bovinus, which since their discovery by Degeer in 1760 apparently
have not been found again.

CHRONOLOGY.

1760. Degeer bred Tabanus bovinus from terrestrial larva.

1775. Fabricius' Systema entomologiae, said to contain notes on Chrysops

larva.

1776. French translation of Degeer's Memoires.
1798. Fabricius' Entomologia systematica.

1834. Macquart cited the observations of Degeer.

1838. Wahlberg, in Sweden, published notes on the larvae as semiparasites.

1840. Westwood cited the observations of Degeer.

1842. Zetterstedt's observations on Chrysops.

1848. Scholtz observed pupae of three tabanid species near Breslau.

1854. KoUar published Mann's observations on oviposition of Tabanus.

1863. Walsh, in Boston, described the larva of Tabanus atratus (aquatic).

1868. Marno, in Vienna, found the aquatic larva of Hexatoma pellucens.

1869. Brauer described the terrestrial larva of Hcematopota pluvialis.

1870. Ferris, in France, independently described the same larva.
1870. Riley, in America, redescribed the larva of Tabanus atratus.
1875. Beling, in Germany, described (third) the larva of Hcematopota.

1878. Graber discovered an otocyst-like organ in the tabanid larva.

1879. Krauss claimed the discovery for Brauer.

1880. von Friedenfels found aquatic larva of Tabanus autumnalis in salt lakes in

Siebenbiirgen.
1882. Beling described the pupa of Chrysops relictus.
1882. German translation of Degeer's Memoires.

1882. Graber described the chordotonal organs of Tabanus {Chrysops?) larva.

1883. Brauer gave notes and illustrations on various tabanid larvae and briefly

summarized the literature.
1895. Hart, in Illinois, described several American species of tabanid larvae,

notably Tabanus stygius, lineola, costalis, and Chrysops vittatus.
1899. Sharp, in The Cambridge Natural History, illustrated larvae of Tabanus

{?Atylotus fulvus).
1903. Hine, in Ohio, described the larva and pupa of Tabanus vivax.

1903. Hine's Monograph on the Tabanidae of Ohio, with notes on life histories,

containing also the first notes, and on oviposition in Chrysops.

1904. Hine's studies at the Gulf Biological Station.

16 THE EARLY STAGES OF TABANID^

1905. Lecaillon, in France, gave detailed studies on the oviposition of Tabanus

quatuornotatus .
1903-06. Hjne's Gulf Biological Station Reports, containing new data on
oviposition.

1906. Hine's Report to the United States Department of Agriculture, describ-

ing early stages of Tabanus lasiophthalmus , sulcifrons (in part), etc.

1907. Paoli, in Florence, continued the work of others on Graber's organ.

1908. Portschinsky, in Russia, summarized the previous knowledge of tabanid

life histories, as an aid in the study and control of these insects.
1908. King, in Khartoum, described the early stages of Tabanus biguttatus.

1908. Walton, in Pennsylvania, described the egg and larva of Goniops

cJirysocoma.

1909. Brimley, in North Carolina, published notes on the larvae of Tabanus

fronto and other species.

1909. Maxwell-LefiFroy and Howlett, in Indian insect life, gave notes on the

larval habits of Indian tabanids.

1910. McAfee, in Washington, described further stages of Goniops.

1910. King, in Khartoum, published the life histories of Tabanus par, icBniola,

ditceniatus, and kingi, securing oviposition in captured specimens.

1911. Lecaillon gave additional studies on Tabanus quatuornotatus.
1911-12. Baldrey observed oviposition of Tabanus orientis.

1913. Mitzmain pubHshed the life history of Tabanus striatus, the carrier of

surra in the Philippines.
1913. del Guercio, in Italy, reported on the larva of Tabanus ignotus.
1913. Picard and le Blanc, in France, observed the larva of Tabanus cordiger.

1913. Patton and Cragg published a text-book of medical entomology, contain-

ing information on the larvae of several Indian tabanids.
19^4. Bainbridge and Fletcher, in India, gave additional notes on Tabanus
striatus.

1914. Lutz, in Rio de Janeiro, described the larvae of three Brazilian tabanids.

1915. Neave presented abundant data on the larval and pupal stages of African

tabanids.

1915. Riley and Johannsen, published a text-book, containing a few new

illustrations.

1916. Girault described a new egg parasite {Phanurus enter soni).

1916. Bodkin and Cleare published notes on the early stages of tabanids in
British Guiana.

TABANIDiE, DESCRIPTION OF THE EARLY STAGES.

The Tabanidae belong ecologically to the so called hydrophytic area,
as has been pointed out by Osburn. On the whole, they may be called
aquatic or semiaquatic, though not all of them pass their larval stage
actually in the water; at the same time they show traces of an adapta-
tion to plants which may serve to understand their phylogeny. The
larvae may be aquatic, semiaquatic, or even terrestrial in habitat,
but always live more or less hidden and are seldom seen. They are,
according to Malloch, *' found rarely among decaying leaves or in low
and somewhat marshy spots in fields." Neave and others, however,
have obtained them in quantity from the muddy banks of rivers and
brooks.

The eggs are laid in clusters of one or several layers, forming a com-
pact fiat or conical mass; they are rarely laid scattered, and never
singly. The clusters may consist of several hundred eggs, the eggs
being held together by a sticky substance. The whole mass is often
covered with a dark shining varnish or with a chalky substance.
These masses are deposited usually on plants or sticks at the edge of
ponds, streams, and lakes, seldom on dry ground, and sometimes also
on stones above water, but this is unusual .^ The single egg is spindle-
shaped or cylindric, and narrowed at the ends, white, yellow, or pale
brownish when first laid, but later often pigmented, brown or shining
black, the color being due to a pigmentation of the chorion. The
period of incubation is short, from 3 to 9 days, but in some cases the
hatching process is delayed for weeks by atmospheric conditions, while
the embryo is fully developed and may be caused to hatch by artificial
stimuli. All the eggs usually hatch at the same time, the young
larvae at first sticking together but soon losing their hold and tumbling
down. In many cases only part of the eggs of one mass give birth to
larvae, about half of them being infected by small hymenopterous
parasites. ^

^ Macquart's statement that the female deposits its eggs in the ground is
erroneous.

17

18 THE EARLY STAGES OF TABANID.E

The newly hatched larvae drop into the water or onto the ground,
according to the egg-laying habits of the species, and little is known
about them. The young larvae resemble the full grown in appearance,
but are generally more transparent, showing the internal organs, a
pair of black eye-spots, and Graber's organ in its primary condition,
which is later described. The chordotonal organs have been described
in young tabanid larvae. The nervous system is clearly visible. The
young larvae may be fed on tiny crustaceans, crushed insect larvae, etc.,
but are not always successfully reared.

Of the young larvae some have strongly developed tracheal trunks
which enable them to float, while others sink to the bottom.

All tabanid larvae are highly predacious, feeding on other insect
larvae, earthworms, and probably all animals they can get hold of,
not sparing their own kind. Wahlberg found them, according to
Brauer, semiparasitic in lepidopterous (noctuid) larvae. Perris
reports that he has found them under stones feeding on larvae of
Rhizotrogus and on Melolontha (imago?). According to Brauer,
young larvae burrow into other larvae (of beetles) and leave these
only when they completely fill the skin of the host. Brauer once
obtained the larva of Hamatopota from the larva of Helops lanipes,
from which, as it seemed to be molting, the tabanid larva made its
way. Walsh states that they feed on snails; Hine and others fed them
on earthworms. Young larvs can be fed on small crustaceans
(King). It is not known what food they take in nature, and also
whether they can subsist on vegetable food. Del Guercio's statements
on this point are probably erroneous. Mitzmain and others have
observed their cannabalistic tendency, at least in captivity. As Pat-
ton and Cragg have stated, the larvae of the large species and of
the nearly allied full grown stages of the smaller species feed almost
exclusively on earthworms whose body juices they suck out; this
explains why gregarine cysts {Monocystis) are not uncommonly found
in the alimentary tract of the imago. Malloch says that the food of
the species occurring in rivers is mostly tipulid and other larvae which
burrow in the soft banks of the rivers or occur in the river bottom or
in drift.

Structure of the Mature Larva. — All the known larvae of tabanids,
with the exception of the somewhat aberrant Gonoips, are elongate,

WERNER MARCHAND 19

cylindric, slender, tapering at both ends, capable of contraction and
extension, their body consisting of a head and twelve segments, the
last segment being very short. The larvae are eucephalous, with head
well developed but small, bearing the three-jointed antennae, at-
tached to the anterior angles of the head just above the palpi, the
basal joint being short, the others of varying length. A bunch of
stiff spines, either short or moderately long, is situated above each
antenna and on each side of the labrum. All the essential mouth-
parts are present and have been studied in many species. They are
well adapted for seizing the prey, and constitute a very formidable
apparatus (Patton and Cragg). The essential organs are the man-
dibles and first maxillae, and of these the former are the most powerful
weapons. Each mandible is a stout rod of chitin, slightly expanded
at the base, to which the muscles are attached, and narrowed distally
to a blunt point; the rod is curved downwards and forward, and on
its concave border has many coarse serrations. The maxillae are
similar in shape and general disposition, but are smaller and less
hea\dly pigmented, and are more pointed. Both pairs of appendages
can be thrust out from the head in a downwards and forward direc-
tion when the larva attacks its prey, by means of the protractor and
retractor muscles attached to the base of the rods. One sometimes
becomes aware of the existence of these organs when handling the
Iarv£e, as they are used in defense as well as in attack, and are cap-
able of inflicting a sharp nip, though they do not draw blood. The
maxillary palpi are simple and two-jointed, the distal joint being
much smaller than the proximal one. The dorsal and distal ex-
tremity of the head is projected forward as a short and fleshy labrum.
A labium is likewise present. The head is attached firmly to the
pharynx, and is retracted and exserted together with it. The pharynx
itself is an elongate chamber of the usual type and is easily seen in
the living larva. The eye-spots are placed on the dorsal side of
the pharynx, not on the outer cuticle, and follow the movements
of the pharynx. The pharynx leads posteriorly to the esophagus,
which is narrow, but wider and more muscular than that of the adults
It is usual to find that the esophagus is wider and more muscular in
insects whose food is solid or semisolid than in those which take only
fluid food.

20 THE EARLY STAGES OF TABANID^

The head itself, as far as it can be exserted, is strongly chitinized
and usually brownish in color, while the rest of the body is generally
pale, whitish or grayish, but in many species marked with a regular
dark pattern, which in others may be absent. Following the head
are the three thoracic segments, of which the first is the shortest,
and these are followed by nine abdominal segments, the last one
forming the respiratory tube, and the next to the last segment bear-
ing ventrally the anus on a fleshy prominence. The body surface is
generally smooth, well chitinized, and half transparent. The fourth
to tenth segments are provided at their anterior margin, or not far
from it, with a strong circular fleshy ridge, containing circular muscles,
and bearing the parapodia, eight in number when all are present, or
fewer if, as may be the case, the dorsal ones are lacking. The ventral
parapodia are usually better developed and placed somewhat closer
together. The parapodia may be retracted and exserted, and, pro-
vided with many curved spines outwardly, play a part in locomotion.
The ridges in some species are not very prominent, are usually cov-
ered with a fine pubescence rendering part of the body surface opaque,
and are often pigmented dark, forming the patterns spoken of. The
spinous processes of the prolegs pass through imperceptible gradations
into the fine pubescence of the dull or pigmented areas.

The larva of Goniops chrysocoma differs in appearance from the
other known tabanid larvas by being apparently much shorter and
thicker in the posterior half of the body than in the anterior where
the segments are much tapered and considerably longer, but as the
figures are based on specimens preserved in alcohol, we cannot pass a
definite judgment on their systematic position.^

Integument. — The integument in general can be divided into more
or less distinct dorsal, ventral, and lateral areas, which differ in their
finer structure, all being either longitudinally striated or more or less
smooth and shining. The lateral areas may be subdivided into an
upper, middle, and lower portion, marked off from one another and
slightly differing in structure. All these structural features of the
surface of the integument are probably vestigial traces of formerly
more marked characteristics, and are important in any effort to sepa-
rate different species of larvae.

' See also Goniops in general, p. 56.

WERNER MARCHAND 21

The structure of the chitinous membrane has been studied in de-
tail by Lecaillon in Tahanus quatuornotatus Meig. The membrane
consists of three layers or zones, the innermost soft and of greatest
thickness, the two outer ones hard and elastic; of the latter the inner
(median) is much thicker than the external one. By staining methods
the three layers can be differentiated. The fine striation of the
cuticle is formed by the median layer, which is thickened in the form
of narrow longitudinal ridges. The hairs are formed exclusively of
the external layer. The muscles to the integument are attached to
the median zone.

The muscular system is strongly developed, consisting of the re-
tractor and extensor muscles of the head, and probably similar muscles
governing the retraction and extension of the syphon; also of muscle
fibers arranged more or less vertically to the integument, and circular
muscles which function in the contraction of the circular segmental
ridges and the retraction and exsertion of the parapodia. The mus-
cular system is partly illustrated but not described by Graber (1882)
(Plate 8, Fig. 100, a). Lecaillon (1906) notes the muscular inser-
tions in the integument, stating that the muscles pass through the
inner chitinous layer, losing here their transverse striation, and being
inserted in the median layer. Various muscles go to the organ of
Graber, described later, by means of which this organ is displaced in
various directions. (Plate 9, Figs. 104 to 106, and Plate 10, Figs.
112, 115, and 116.)

Alimentary Canal (Plate 7, Fig. 98). — This has been studied by
Patton and Cragg in Tahanus alhimedius (also in other species?). It
is much shorter and less complicated than that of the larva of
Stomoxys calcitrans, described by the same authors. The esophagus
opens into a short, cylindric proventriculus, which is also a highly
muscular structure, and is sharply distinguished from the succeeding
part of the gut by its clear translucent appearance in the fresh con-
dition. The mid gut extends from the proventriculus to the hind
end of the body, and is thrown into one or two simple coils, not con-
stant in their position. It is separated from the proventriculus by a
short constriction, and is again constricted at the posterior end, just
anterior to the opening of the Malpighian tubes. Between these
points the lumen is wide and is thrown into numerous sacculations

22 THE EARLY STAGES OF rABANIDE

by the contractions of the muscle fibers in the wall. The mid gut
is, at least in the species studied by Patton and Cragg, of a strik-
ing orange-red color in the fresh condition, and is filled with a semi-
solid mass of a light chocolate color, which oozes out if the wall is
punctured in dissecting. The hind gut is short and simple, and is
coiled up in the posterior end of the abdomen. The total length
of the gut is about twice the length of the body of the larva. The
salivary glands are simple and tubular, and bear a remarkable resem-
blance to those of the adult insect.

Respiratory System. — The respiratory system resembles that of the
mosquito larva. There are two large lateral tracheae which run the
whole length of the body on each side of, and slightly dorsal to the
alimentary canal. According to Patton and Cragg (1913), these
communicate with the external air through an opening, which can be
closed, on a small prominence on the dorsal surface of the penultimate
segment. This statement appears to be erroneous. Brauer (1883)
states that the tabanid larvae are usually metapneustic, the last
segment with a vertical respiratory fissure or the last two segments
forming a respiratory tube. According to Malloch (1917), they are
always metapneustic. The respiratory tube is of varying length,
sometimes very short, sometimes more slender and tube-like, or form-
ing a short acute spine. The two main air-filled trunks are much in-
flated in some species, enabling the larvae to float at the surface of
the water, and narrow, almost filiform in other species. As they pass
forward, they give off slender branches of tracheae to the body cavity
in each segment. In the anterior part of the body the tracheal
stems, in the species where they are generally much inflated, become
slender, describing one or several semicircular loops and tapering into
fine tracheae going to various organs. The tracheal branches going
to the body wall of each segment may be homologous with those
leading to the abdominal spiracles in amphipneustic larvae.

Dorsal Blood Vessel. — The dorsal blood vessel is easily seen in trans-
parent larvae and ends blind, in the cases observed, before the elev-
enth segment.

Malpighian Tubes. — The Malpighian tubes are, to judge from
Patton and Cragg's figures, four in number, in young larvae irregu-
larly crowded in the posterior part of the body between the two
tracheal trunks, and of brownish, greenish, or yellow color.

WERNER MARCHAND 23

Fat Body. — No observations of the fat body have been made, ex-
cept that Graber (1878) describes it, in Tahanus autumnalis (?), as a
network of pale lobes and trabecute with dendritic tracheal expansions,
lying dorsally and above the new organ described by him.

Gonads. — These organs, which must be present in the larva, have
not been noticed or described by any author.

Nervous System. — ^The nervous system of the larva consists, ac-
cording to Graber, who studied the young larva of Tabanus {Chry-
sops ?), of a large upper and lower cephalic ganghon, and a chain of
large ganglia arranged like beads and extending about half the body
length. According to Graber, this would be the only instance of the
nervous system having this structure (Plate 8, Fig. 99).

Sense Organs. — The eyes have been mentioned; their function is not
clearly understood. Antennas and palpi are probably olfactory.
Tactile bristles occur on the body surface and have been studied by
Graber who finds them connected with several ganglion cells (Plate
8, Fig. 100, h). Chordotonal organs, similar to those in Corethra,
are found laterally in each segment near the surface of the integu-
ment, and are of the mono-, di-, and triscolop type (Graber), and are
connected with the trachea?. According to Graber, in small larvae,
the chitinous body walls themselves act as a tympanum, thus ex-
plaining the absence of any special tympanal apparatus (Plate 8,
Fig. 100, a).

At the posterior end of the larva, situated dorsally and adjacent
to the blind posterior end of the heart, on the eleventh segment, is
an organ of unknown function, called Graber's organ, after its dis-
coverer (1878). It consists of a pear-shaped sac, the broader end of
which is anterior; the posterior end narrows down to a fine tubule
which opens on the integument of the body between the last and the
next to the last segments. Within this sac there is a series of capsules
set one behind the other in the long axis, and within each of these
capsules is a pair of small, black, pyriform bodies, each attached to the
anterior side of the capsule by a delicate pedicel (Plate 9, Figs. 103 to
108, and Plate 10, Figs. 109 to 120). These bodies diminish in size
from the anterior end, the first being considerably larger than the
rest. They are easily seen through the integument of the living
larva. The outer sac is an invagination of the integument, and as

24 THE EARLY STAGES OF TABANID^

such has a chitinous lining. Graber believed that the structure is
an auditory organ, but Lecaillon supposed it to be a gland. Berlese,
according to Patton and Cragg, does not consider that it has been
proved to be a sense organ, though it is well supplied with nerves.
Paoli has shown that the young larva has only two pedunculate
bodies, and that with each successive molt a new capsule with two
more such bodies is formed. Having observed that Tahanus larvae
are able to produce a slight crackling noise, Paoli believes the struc-
ture to be a sound-producing organ, but it is dilfhcult to understand
what ecological meaning such an organ could have.*

The pupae of tabanids are classified as orthorhaphous (Brauer),
opening in the act of hatching with a dorsal longitudinal slit. They
strongly resemble lepidopterous pupae {Mumienpuppe, pupa obtecta),
having wing and leg cases firmly attached to the body and covered
with it by a chitinous membrane. The pupal body is subcylindric,
abruptly pointed or rounded anteriorly, and tapering somewhat poste-
riorly; it is generally yellowish brown to ferruginous brown, finely
wrinkled, and has a lateral tuft of hairs on each abdominal segment.
On either side of the head are the antennal sheaths, pointing outwards,
and on each side of the median line two large tubercles, each with a
central hair; below these are two raised areas with sharp edges, sepa-
rated by a deep ridge. Lower still there is a pair of elevations, also
with raised edges, and on the ventral surface of the head one or more
tubercles. The segments of the thorax are indistinct; the mesothorax
bears the large raised ear-shaped spiracles. The abdominal seg-
ments are free and about equal in length, and have one or more fringes
of hairs near their hind margins; the second to the seventh segments
inclusive have well marked lateral areas, covered with long hairs
which are continued into the dorsal and ventral surfaces and corre-
spond to the lateral areas in the larval integument. These hairs in-
crease in length from before backwards, and are best developed on the
seventh segment. The eighth segment is short, and is provided with
six projecting spurs or teeth, and with a large anal tubercle. In the
male the tubercle is ribbed and bounded anteriorly by a continuous
fringe of strong spines; in the female the tubercle is smaller and the

* For details concerning the structure of this organ, refer to pp. 29-43.

WERNER MARCHAND 25

fringe of spines is widely separated. The pupas of many of the
larger species of Tahanus have in addition a lateral tuft of spines situ-
ated on a ridge. Patton and Cragg were the first to notice structural
differences in the male and female pupas.

Brauer (1883) says that he classifies the tabanid larvae with the
Leptidae and Acanthomeridae (to which they are undoubtedly nearly
related, as also to the Asilidae) under Homoeodactyla tany stoma, of
which the larval characteristics are:

Larva meta- or amphipneustic or with tracheal gills, usually with eyes at the
sides of the head capsule, the latter usually hidden in the following rings, generally
more or less retractile, and behind it eleven or twelve body segments. Mandi-
bles hook-hke; extending out of or under them in their concavity are the maxillae,
which are soft-skinned, and the laterally prominent maxillary palpi; antennae
short. Labrum hooked or horn-like, exserted. Pupa a free "mummy-pupa."

Malloch (1917) proposes to unite the Tabanidae with the Leptidae
to a superfamily Tabanoidea, having the following characteristics:

^^ Larva. — Head small, wholly or partly retracted, permanently retracted
portion with an arcuate dorsal plate over the longitudinal rods; mandibles strong,
hook-like, curved downward; maxillas well developed, wholly or largely mem-
branous, the palpi well developed; antennae distinct, pedunculate. Body cy-
lindrical, with or without pseudopods; lateral abdominal spiracles absent in
Tabanidas, small, but no lateral spiracles distinguishable in Leptidae; apical
spiracles in a vertical fissure in Tabanidae, exposed and separated in Leptidae."

"Pupa. — ^Head with strong cutting armature; antennae with or without
distinct annuli. Thoracic respiratory organs sessile. Wings and legs closely
fused to each other and to thorax; fbre tarsi overlying midpair, the latter over-
lying hind pair, the pairs successively longer, hind pair not extending beyond
apices of wings. Abdomen with seven pairs of lateral spiracles; segments armed
with transverse series of slender bristles which become progressively stronger
from base to apex of abdomen."

Brauer's characterization of the Tabanidae is the following:

Family TabanidcB. — Larva usually metapneustic. Eyes distant from the mouth-
parts, situated laterally and back of the head capsule; the latter divided by a
fissure at the inserted end, at the hind end open and prolonged backwards into
long rods, hidden in the following segments. Mandibles hooked, often serrated
at the border; labrum forming a hook-like projected septum between them.
Body eleven-segmented, often girdled with retractile fleshy tubercles which are
often developed as prolegs at the ventral side only. Last segment with a ver-

26 THE EARLY STAGES OF TABANID^

tical respiratory fissure, or the last two segments forming a respiratory tube.
Pupa free, without crovvn of hooks at the fore end, antennal sheaths laterally
exserted, parallel to the head. Betv/een them four swellings or ridges, arranged
in a curve and formed by chitinized folds. Above them three tubercles, forming
a triangle (anlage of ocelli) and behind them outwards two larger tubercles.
Beneath the v/ing sheaths on the inner margin of the sheath of the compound
eyes on each side arc two small tubercles lying one above the other. Anal seg-
ment with six divergent cone-shaped pointed hooks. Spiracles behind the head
and on the seven abdominal segments distinct and large, the former with kid-
ney-shaped or ear-shaped margin (Therioplectes) and often very large. The
pupa rests in the ground or, in some species (Therioplectes), remains in the water.

Malloch describes the body of the larvae as "circular in transverse
section, elongate, tapering at both ends, and with encircling loco-
motor swelhngs at the segmental sutures in all genera except Goniops."

Malloch's description of the tabanid pupa, based on the study
of a number of species, is the following:

"Pupa. — Head without projecting thorns. Thoracic respiratory organs ses-
sile, connected subcutaneously with a large cavity on each side of median line
close to anterior margin of prothorax. [Plate 13, Figs. 153 and 154.] Wings
and legs rather short. Abdominal armature consisting of 1, or 2 closely con-
tiguous, series of bristles on each dorsal segment except first, and a weaker trans-
verse series on ventral segments; apical segment ending in six stout processes
which are more or less radiate and pointed. [Plate 13, Figs. 162 and 163.]"

On the habits of the larvae before pupation, facts have been pub-
lished by Neave (1915).

Before the larvae have reached their full growth, which in many
cases signifies the beginning of a resting period, they usually lie
buried in the mud, head downwards, with their syphons projecting
immediately above the surface of the mud or of a shallow layer of
water above it if it is present. In the resting stage the syphons do
not seem to be used and the larvae remain several inches below the
surface for weeks or even months. This is presumably an adapta-
tion connected with a climate in which there is a very marked dry
season, and, consequently, a risk of the mud in which they are lying
more or less drying up. Pupation appears to take place in normal
circumstances an inch or more below the surface, though occasion-
ally in captivity individuals pupated lying horizontally upon it.
The pupa is normally upright in the mud, and after pupation, as soon

WERNER MARCH AND 27

as the case has hardened, it works its way up by means of its rings
of spines and the aster (a name proposed for the terminal whorl of
spines), until the pupal head lies just below the surface, being often
visible from above. The pupa at first is usually of a pale yellowish
or greenish color, but darkens as the imago develops within, the
process beginning with the eyes.

Neave observed that in normally fine weather the imagos of all
species almost invariably emerged between noon and 3 p.m. They
were more irregular during spells of dull and rainy weather. The
process of emergence seems to be similar in individuals of a genera,
whether Chrysops, HcBmatopota, or Tabanus. The head of the pupa
splits in the median dorsal line and the imago rapidly emerges until
only the end of the abdomen, which is at first enormously elongated,
remains in the pupa case. The wings at this stage are milky white
and the darker markings, if any, are barely visible. The unago
usually remains in this position for two or three minutes before
completely leaving the pupa case. It is capable of flight very soon
after this, but if undisturbed sits for about half an hour on any
suitable object near by while the wings dry and assume their normal
coloration, and the abdomen its normal shape. During this period
several drops of a milky white fluid (the meconium) are passed
through the anus.

Bionomics.

On the bionomics of the tabanids in the early stages we possess
many interesting notes by Neave (1915). From the results of a
year's collection of adults in one locality, and from other evidence,
it is probable that the majority of tabanids, at least of Nyasaland,
have only one brood a year. This is certainly true of nearly all the
species of Tabanus and of Dorcalosmus fodiens Aust. It is possible,
however, that certain species of Chrysops and Hcematopota may be
double-brooded; much doubtless depends on the larval food supply,
climate, etc.

According to Neave, in the case of many species the larva grows
very slowly after hatching and often takes six months or more to
become full grown. It then, especially in species of Tabanus, goes
through a resting period, during which it remains buried in mud or
sand, sometimes at a considerable depth. In contrast to the lengthy

28 THE EARLY STAGES OF TABANID^

larval stage, the pupal period is short, varying, in Neave's experience,
from 10 to 16 or 18 days, according to the species and the climatic
conditions, the longer period being usual for the larger species of
Tabanus.

One difficulty connected with the question of these flies having
more than one brood a year arises from the fact that even in larvae
from the same batch of eggs the rate of growth is extremely variable,
and consequently the processes of pupation and emergence do not take
place simultaneously in a certain proportion of the individuals. Some
of the remainder take longer to reach maturity, others seem to pass
through an extended dormant period. The adults arising from these
emerge at irregular intervals, often months later. This probably
explains the capture of odd specimens of any species long after the
usual season. It would also seem not improbable that individuals
which miss their normal season for pupation in some circumstances
continue in the larval stage until the following year. Thus Neave
possessed in his laboratory in January and February examples of
larvae of Tabanus corax, some still in a dormant state and others not
yet mature, the season for the adult flies being over by the begin-
ning of January. It would appear that these would not have pro-
duced imagos until the following December, though Neave was un-
able to decide this point on account of his return from Nyasaland to
England. Hine states that the larvae of Tabanus stygius — probably —
hibernate twice before giving the adult.

Surcouf and Ricardo, however, assert that in spite of the opinions
to the contrary the tabanids seem to have, at least in Algiers, often
two generations in a year. These authors have taken from a horse
several fresh specimens in May (1908), while the same species is taken
in France and Algiers in the fall. Maxwell-Leffroy and Howlett state
that at Pusa (Bengal) there are apparently three broods of Tabanus
yearly, flies emerging at the beginning and end of the hot weather
(about February and June) and at the end of the rains (October).
These statements should be compared with those made by Neave.
Hibernation takes place in the larval state in India in all species
which were observed by Maxwell-Leffroy and Howlett; and the same
holds good for all known American species (Hine) , larvae being found
late in the fall and again in early spring. No pupae have been found
in the winter.

WERNER MARCHAND 29

Special Anatomy of Tabanid Larva. Graher's Organ.

Almost the only detail of the structure of tabanid larvae which has
been studied more extensively is a peculiar organ situated on the
dorsal side of the segment before the last, and discovered first by
Graber (1878), who, while conceding that it has very little resem-
blance to an auditory organ, nevertheless may have to be placed in
this category. As the situation of the organ is rather unusual if it is
auditory, Graber recalls the discovery by Grobben of similarly situ-
ated organs in the larva of another fly, Ptychoptera contaminata
(Tipulidae), and of which the auditory function seems fairly certain.
Whether similar organs are found in other dipterous larvae is not
known; Malloch (1917) does not mention them. The organs seem to
occur in all tabanid larvae, but to be more easily seen in the young
stages; they have been overlooked by some later investigators, nota-
bly Hart (1895) and Hine (1901-06), but have been seen and figured
by Walton (1908) for Goniops chrysocoma (only as dots), and by
Mitzmain (1913) for Tahanus striatus, all of whom have apparently
overlooked the European literature on the subject, Mitzmain assum-
ing that the organ must either be peculiar to the larvae observed
by him, or have been overlooked by the previous authors.

Graber found the organs in the abdomen of the larva, at the
sides of the two last segments. They consist of a cornea-like infla-
tion of the integument which is covered inside with its epithelium.
The open inner side of this vesicular invagination seems to be closed
by a membrane supported by radial elastic fibers, the fundus of which
can be extended by means of a strong muscle. According to Graber,
the presence of a special nerve, which, however, approaches the
vesicle without any specialized termination, is proof that we are
dealing with a sense organ.

However, what characterizes them as auditory organs are two or
three otolith-like bodies found in the liquid contents of the capsule.
These abdominal otocysts consequently differ from antennal structures
described elsewhere by Graber, first, by belonging to the integument
itself, second, by their inner wall not being formed by chitin but by
the epithelium. As a consequence of this, the development of cuticu-
lar auditory hairs is naturally also suppressed, and a new type of

30 THE EARLY STAGES OF TABANID^

auditory capsules is developed, which Graber distinguishes as oto-
cysts without cilia (aciliate) from other known otocysts (ciliate
otocysts).

The organs were found by Graber (Spring, 1878) on a dipterous
larva collected with other material, on the bottom of a mud pond
in a brick factory. The larva (Plate 9, Fig. 103) was when extended
about 40 mm. long, cylindric, strongly spindle-shaped, and pointed
at the end, and except the dark intestinal system, of an almost glassy
transparence. Eleven segments were counted, which with the ex-
ception of the terminal ones, were surrounded by a girdle of papilli-
form retractile processes. Unfortunately, as literature was not acces-
sible, the species could not be determined.

In order to see the organ in the fresh condition, the larva is laid
on its ventral side and fastened with an elastic holder. The organ
(Plate 9, Figs. 104 to 108) lies in the median line of the dorsal side,
behind the termination of the dorsal blood vessel, and immediately
behind the border incision between the ninth and tenth segments.
By focusing from above, at first the body cuticle becomes visible,
which derives from the presence of longitudinal ridges the extra-
ordinary elasticity characteristic of these larvae. When focusing
farther down, the epithelium is seen, consisting of flat polyhedric
cells, and following this a network of pale lobes and trabeculae with
dendritic tracheal expansion; namely, the fat body. Immediately
underneath these tissues lies the organ in question, which is conse-
quently completely separated from the integument. It consists of
a pale capsule 0.3 mm. in length and pear-shaped, its free and broader
end directed upwards, while its pointed hind end is prolonged into a
narrow tube. The capsule being of considerable size is visible even
at slight magnification. Moreover, it is not likely to be overlooked
because of the intensely black bodies included in it, which strongly
contrast with the light background.

In detail the structure of the organ appears to be peculiar. From
the capsule proper with its tube the nerves and muscles attached to its
anterior end may be distinguished. The whole capsule together with
the terminal tube appears, as Graber says, to be a cecum-like in-
vagination of the ectoderm. Unfortunately the origin of the ter-
minal tube could not be found by Graber; it apparently lies in the

WERNER MARCHAND 31

last body segment and is in no way connected with the intestine,
which terminates at the end of the segment before the last, or with
the sexual organs. In Graber's account it remains questionable
whether the terminal tube, as was supposed, really originates in the
integument. The tube and the capsule forming its enlarged terminal
part behave histologically like a glandular formation. The main layer
is an epithelium consisting of large cells. This is seen best, especially
after treatment with 35 per cent potassium hydroxide, at the distal
(head) end of the capsule. The cells appear as elongated, sac-like,
pale compartments, separated from one another by bridge-like septa,
always showing a dark nucleus accompanied by a small nucleolus.
In the remaining part no distinct cell borders can be distinguished,
but only large granulated nuclei. In the terminal tube the latter are
arranged alternating behind one another, similarly as in the narrow
secretory ducts of true glands. The epithelial tube is covered at the
outside by a thin homogeneous covering which may be considered as
a tunica propria.

The structures belonging to the third stratum, that is, to the
chitinous membrane, are peculiar. Leaving aside certain complica-
tions, we find first a chitinous capsule corresponding to the epithelial
one, which is prolonged into a narrow canal corresponding to the
terminal tube. This canal pursues, inside the terminal tube, an un-
dulating course, curved alternately to the right and left, suggesting
the muscle in the stem of a Vorticella.

The chitinous capsule enclosed within the epithelial vesicle is
often very thin and delicate compared with the thickness of its
matrix, and is perfectly transparent. Its free (inner) surface, how-
ever, shows at high magnification small tegula-like overlapping
scales. In this chitinous capsule the black bodies already mentioned
are enclosed.

At low magnification these bodies appear as simple homogeneous
globules. At the highest magnification, however (Plate 9, Fig. 107),
they are found to be hollow chitinous structures with a somewhat
wrinkled surface, which after the fashion of a volumetric flask are
prolonged into a narrow hollow stem or peduncle. The bodies re-
main perfectly black and opaque even when boiled in caustic potash ;
consequently they seem to have very thick and strong chitinous

32 THE EARLY STAGES OF TABANID^

walls. Sometimes it seems as if they are filled with a dark tough
substance, projecting in the shape of a papilla or of a granulated
string into the lumen of the peduncle. Eight of these pedunculate
bodies, as Graber calls them, were present in the larva examined,
arranged in four pairs, lying one behind the other, giving to the
vesicle the appearance of an internally segmented organ. The
bodies of the first two pairs are of about the same size, 30 ji in di-
ameter, the length of the peduncle 26 ix, its width at the tip 1.8 ix.
The bodies of the two following pairs immediately touching one
another are considerably smaller than the rest, their diameter being
only 20 /x.

The compound nature of the organ appears more distinctly than in
the serial arrangement of the pedunculate bodies, in the repetition of
the covers in which they are enclosed. There are not four such covers
or secondary sacs, however, but only three, as the two posterior pairs
of pedunculate bodies are contained in a common envelope. The
first sac is formed by the capsule itself; that is, by its rounded (head)
end. The pedunculate bodies are observed hanging by their pedicel
from the slightly invaginated upper wall, and, like the following ones,
turned from the inside towards the outside in a somewhat oblique
direction. The space in which the bodies are found, however, is
separated also from the remaining lumen of the capsule, and, as far
as could be seen, by a transverse septum issuing from the side walls.

The second pair of pedunculate bodies is, however, surrounded by
an independent sac entirely separated from the main capsule, the
fixation of the pedunculate bodies themselves at the inner upper wall
of the secondary vesicle being exactly the same as that in the first
capsule. This second sac, however, is not completely closed behind
the pedunculate bodies, but is forming here only a neck-like con-
traction, its walls otherwise being continued into the following sac.

The third inner sac resembles the second one in all essentials, but
encloses, as already mentioned, two pairs of pedunculate bodies,
which again, in a manner analogous to that in the former cases, are
inserted into its walls.

Probably the last two sacs are evaginations of the fundus of the
common capsule, fitted into one another like the sheaths of an onion.
In this way the first pair is surrounded by one envelope, the second

WERNER MARCHAND 33

by two, and the third and fourth by four envelopes. The accessory
parts of the organ have also been studied by Graber.

At a moderate magnification it is possible to discern, in addition
to the tube extending backwards from the capsule, two other liga-
ments originating at the sides of the pear-shaped part, so that the
whole organ appears to be held in place by three ligaments, one of
which is the terminal tube. The two lateral ligaments extend ob-
liquely, crossing the two large tracheal trunks, in a forward and out-
wards direction, inserted in a place not determined by Graber but
probably lying in the seventh segment. At higher magnification the
anterior ligaments are found to be muscles, and immediately behind
the place of origin of these muscles two pairs of nerves are seen at-
tached to the capsule. The first nerve being rather thin forms imme-
diately at the head of the capsule a thick ganglion-like swelling, in
which several pale nuclei are discernible. The character of the end-
ing of this nerve has not been determined; Graber assumes that it
enters into connection with the capsular epithelium. Nothing more
is known about the termination of the second nerve which is much
thicker than the first one.

All the parts described, the capsule, the tenninal tube, the muscles,
and nerves are connected with one another and with the surrounding
organs, as also with the tracheae and with the band-like extensions
of the heart muscles, by a peculiar connective tissue. The latter
shows the greatest affinity with certain elastic reticulate tissues of
higher animals. These tissues play a part, according to Graber, in
the extraordinary changes in the relative situation of the parts
during the movements of the larvae. Graber discusses the possible
function of the organ he described, assuming that it must be either a
gland or sense organ. The presence of a duct leading to the exte-
rior seems to indicate the former, but the arrangement of the con-
tents seems to contradict this view.

If the organ is considered a gland, this would imply that a secretion
formed by the epithelial cells is discharged into the chitinous capsule.
In this case, however, no explanation is found for the presence of the
pedunculate bodies, not to mention the various interior secondary
capsules. The pedunculate bodies have no free opening, and while
they are hollow and possibly have been formed by invagination of

34 THE EARLY STAGES OF TABANID^

the walls of the capsule, the place of invagination seems to have
disappeared.

Graber therefore decided that the organ must be a sense organ,
which, however, according to its peculiar structure and position can-
not be either an organ of touch, smell, taste, or sight, and is conse-
quently placed among the auditory organs. Graber also believed
that he found the two main features of a cystoid auditory organ,
namely a cyst-like cell system, connected with a nerve, and an in-
ternal fluid medium which may carry the sound. There were lacking
only the characteristic elastic appendages of the auditory cells.

Graber assumes that under certain conditions the auditory cells
may be stimulated by waves of sound even in the absence of special
end organs, in the same way that the light-perceiving elements of the
eye in part are also affected by light-waves in the absence of special
terminal organs as found in highly developed organs of sight.

The presence of the peculiar pedunculate bodies seems, more than
anything else, to have determined Graber' s view, in as far as they
form a good analogy to the otoliths commonly found in auditory or-
gans of the cyst-like type. The pedunculate bodies are comparatively
heavy, thick- walled bodies and are attached to the cysts by means of
thin and probably elastic filaments or strings ; they may be compared
with the clappers of a bell, which facilitate their function as otoliths.

Graber classifies all auditory organs occurring among insects, as
(1) elementary organs, consisting of isolated hearing cells or auditory
hairs, (2) cystoid auditory organs, gymnotocysts, and chitinotocysts,
found in crustaceans and many insects, and (3) tympanal organs, pro-
vided with auditory rods as found in Orthoptera. The chordotonal
organs are related to the latter and differ by the absence of a tympanal
membrane. The organ in question is classified among the cystoid
organs.

Henneguy found the same organs in small larvae resembling larvae
of Stratiomys (Lecaillon), (Plate 10, Fig. 109), which undoubtedly
belong to Tabanus. The organ in this instance (Plate 10, Fig. 110)
is different. It comprises a cellular mass situated in front of the
organ and evidently connected with the dorsal vessel, a capsule fol-
lowing it, containing two pigmented bodies arranged as described by
Graber, and, finally, a cyst prolonged backwards into a filament but

WERNER MARCH AND 35

not containing pigmented bodies. Henneguy is inclined to place the
organ among the chordotonal organs.

Lecaillon, having in 1904 sent some larvae of Tabanus quatuorno-
tatus to Henneguy, studied the organ at the latter's suggestion. He
records a number of facts which may facilitate further research on
the subject. The principal results of his observations are as follows:

1. Graber's organ exists in all larvae of Tabanus quatuornotatiis. It should
consequently be sought for among the tabanids, and possibly also in related
families.

2. It already exists in the larvae hatching from eggs, but it continues to de-
velop as the larvae grow.

3. In very young larvas its structure corresponds to that described by Henne-
guy; in fact he made his observations on very young larvae.

4. As the larvae grow, the primitive structure becomes more complicated and
more closely resembles the structure described by Graber. This author ob-
served full grown larvas.

5. Consequently it can be stated that Henneguy 's description applies to the
young condition of the organ, and that of Graber to the fully developed condition.

6. In Lecaillon's observations, as shown by the figures, the pigmented bodies
do not always remain regularly arranged in pairs, contrary to Graber's descrip-
tion. They are seen frequently placed one behind the other in the posterior duct
of the organs. Sometimes there is an odd number of them in this duct, which
means undoubtedly that these granules can sometimes be discharged or expelled.

7. While nothing can be said with certainty about the function of Graber's
organ, Lecaillon considers it glandular in nature rather than a sense organ.

The figure of the larva given by Lecaillon (Plate 10, Fig. 117) was
obtained from a living larva of Tabanus quatuornotatus one month
old, seen in dorsal aspect with light falling through, a larva at this
stage being still very small. Two other figures (Plate 10, Figs. 118
and 119) give the aboral extremity of two other larvae slightly older
than the first one.

As seen in these figures, the main part of the organ of Graber oc-
cupies the median and dorsal region of the next to the last segment, in
almost its entire length, and is situated not far from the two large
tracheal trunks, which open, one near the other, at the end of the last
abdominal segment. It consists mainly of an oval sac or cyst pro-
longed posteriorly into a tube. The oval and dilated part is subdi-
vided by a septum in two cysts placed one behind the other. Careful

36 THE EARLY STAGES OF TABANID^

study of this region shows that the two cysts are not completely sepa-
rated and, on the contrary, communicate with one another in their
axial region; in other words, the septum which separates them is
incomplete. The tubular region which prolongs the posterior cyst
begins towards the middle of the segment next to the last and opens
in the median line not far from the openings of the two tracheal tubes.
This tubular region is, consequently, about as long as the dilated por-
tion of the organ. The fundus of the sac, that is its anterior region,
has thicker walls than any other part. There is in this region a
cellular mass similar to that described by Henneguy in the organ of
the undetermined larva which he examined.

The ink-black bodies can be found, according to Lecaillon, in both
capsules and in the tubular portion. The first capsule always con-
tains two which are situated anteriorly, one to the right, the other to
the left of the median line. By carefully examining their arrangement
it is found that they are attached to the cellular mass at the fundus
of the cyst, each by a small pedicel, as has been described by Graber
and by Henneguy, whose observations are confirmed by Lecaillon.
These pedicels do not show on Lecaillon's figures, which, as he says,
were drawn from living larvae under conditions when the details were
not clearly visible.^ The second capsule contains sometimes also two
of the black bodies, arranged as in the first capsule, but often it
does not contain any of them.

The tubular region often contains no black bodies. But frequently
and especially in older larvae, it almost always contains a varying
number. The maximum number which Lecaillon found was sLx, but
there may be only five, four, three, or two, even in larvas of the same
age and living under the same conditions. As the caliber of the tubu-
lar region is narrow, the black globules are placed here in a length-
wise row, and not in pairs placed transversely, as in the capsules.

As Lecaillon wished to ascertain whether the black bodies could
eventually be expelled to the exterior, a larva which had six globules
in the tubular portion of the organ was placed under observation.
Eight days afterwards Lecaillon found that the tube contained only

^ I found that living tabanid larvae up to 15 mm. in size can be examined by
means of a compound microscope, without suffering injury from the pressure
necessary to hold them in place.

WERNER MARCHAND 37

two globules, four evidently having been expelled. At about the same
time another larva of the same age was found to contain only the
two bodies of anterior cyst; all the others had been thrown out.®
In summarizing Lecaillon states:

(1) The black bodies are formed by the cellular mass which constitutes the
fundus of the capsule. (2) The globules are detached periodically to be expelled
to the exterior after a longer or shorter time. (3) They are regenerated at the
bottom of the cyst when the others have been expelled.

These conclusions of Lecaillon appear only partly warranted.
From observations made hitherto there is no indication of a relation
of the discharge of these bodies and their new formation.

Concerning the nature of the black bodies, Lecaillon states that they
are not hard as if they were mineral granulations ; they are not affected
by reagents, as acid fixation fluids; they may be cut when the organ is
sectioned. Consequently they seem to be a pigmented material,
excreted by the cellular mass which forms the bottom of the cyst.
Lecaillon alludes, in suppport of this view, to the fact that the larvas
of Tabanus quatiwrnotatus have a completely white body though the
substances found in the digestive tract (organic relics) are usually
blackish. The pigment derived from the alimentary substances, or
formed normally in the body, might be excreted by means of this
particular gland.

Paoli in Florence in 1907 studied the same organ in various taba-
nid larvce which, on the authority of Bezzi, are said to belong prob-
ably to Tabanus cordiger Meig. or Tabanus autumnalis L. Paoli,
however, asserts that in all probability many species of this family
possess Graber's organ more or less developed and also with slight
modifications in details of structure, according to age and species.
For the description of the larvae studied by Paoli see pages 91-93,
Tabanus autumnalis.

According to Paoli's description, Graber's organ is found in the vis-
ceral cavity, situated dorsally under the subcutaneous muscular strata,
in the anterior third of the eighth abdominal segment between the
two large tracheas (Plate 10, Figs. 115 and 116). It is a little pear-

® I have been able to confirm Lecaillon's observations in the larvae of Tabanus
atratus (1916).

38 THE EARLY STAGES OF TABANIDiE

shaped cyst with a slight constriction in its wider part; its pointed
end is turned hindwards. Inside this cyst are found, in the cases
examined, several pairs of black rounded chitinous pedunculate
bodies; each pair of these bodies is in turn enclosed in a special chit-
inous capsule filled with liquid. The attachment of the muscles and
nerves to this cyst are described in detail. The cyst is continued,
at its pointed end, into a contorted canal which opens in the deepest
place of the infolding between the eighth and ninth abdominal seg-
ments (Plate 10, Fig. 115). The canal and the cyst are covered in-
side with chitin and completely surrounded by the hypoderma, so as
to warrant the conclusion that the whole organ must be an invagina-
tion of the integument, as Graber had supposed, though he was
unable to see the opening of the terminal tube.

According to Graber, the first cyst contains in its interior four pairs
of pedunculate chitinized bodies; the first pair being in a closed cap-
sule, the second inside another capsule, the latter not completely
closed posteriorly, and finally the remaining two pairs contained in a
common sac, open posteriorly.

Paoli states that the number of pedunculate bodies is not limited
to four pairs, but varies according to the age of the larva. Graber also
said (in his later work), that in the young larva there is only one
pair of such pedunculate bodies, and that the second pair is formed
later. Henneguy, who examined a newly hatched larva, found only
one pair.

Paoli found four pairs in larvae about 1 cm. in length, but in speci-
mens of larvae in later stages of development he always found a greater
number, up to seven pairs in larvae of almost 3 cm. in length. Each
pair was found enclosed in a capsule which remains more or less open
posteriorly, but which is always present; the largest of them, that is,
that which is located anteriorly, has the thickest walls and is formed
last, and it is only this one which is really living and in function,
while all the others are dead and have no function whatever, differing
little from one another, especially the capsules enclosing the smallest;
namely, the oldest pairs of pedunculate bodies.

Each of the pedunculate bodies has the shape of a slightly elliptic
ball, their surface being smooth or sometimes rugose. They are
strongly chitinized, black, frail, and breakable, as under pressure

WERNER MARCHAND 39

they may be reduced to small fragments; the pedicel is inserted ob-
liquely, equal to the latter in length. The pedicel itself is chitinous
but not black; at most it is brown in the neighborhood of its inser-
tion in the globule; at the other end it has a widening by means of
which it fastens itself in a little cavity of the wall of the chitinous
capsule. The pairs of pedunculate bodies are attached each one
to the central portion of the anterior wall of the capsule itself, and the
pedicels are a little divergent so that the two spherical bodies are
located a certain distance from one another, suspended in the liquid
which fills the cavity of the capsule. The pedunculate bodies are of
dimensions increasing in the direction of the head; in a larva with
four pairs these bodies measured for each pair 11, 15, 16.5, and 18 jx
in diameter.

Paoli explains the probable manner in which these cysts and the
tube are formed. It has been stated that the latter opens into the
innermost part of the sulcus which limits the eighth abdominal seg-
ment from the ninth; Graber did not see this opening but supposed
that it was to be found in the last — ninth — segment and certainly
independent of the alimentary canal and of the genital organs.

The origin of this organ is thought to be a sac-like invagination of
the integument (Plate 10, Fig. 113). The most external part forming
the terminal tube, the more internal one a cystiform enlargement,
from the bottom of which the two pedunculate bodies originate, the
latter consequently being nothing but modified hairs derived from
the hypoderm surrounding the capsule.

In this way it is evident that in the young larva which has not
yet undergone any molt there will be a primitive organ with only one
pair of pedunculate bodies. At each successive molt the chitinous
stratum of the organ remains in place, involved by a new stratum
formed around it by secretion from the hypoderm, and the old cap-
sule is pushed back distally. The new layer consequently forms a
new cyst with a transverse division separating two halves in the
anterior of which two new pedunculate bodies are formed, this being
the new organ, while the posterior half consists of. all that existed
before. In this manner, with the successive molts, an increase in the
number of the cysts and of the pedunculate bodies is brought about;
the capsule wall which is thinner when it belongs to a younger stage;

40 THE EARLY STAGES OF TABANED^

is no more easily visible in the final form. This h3T>othesis is, accord-
ing to Paoli, the only one which explains satisfactorily the manner
in which this strange organ is formed, and it is sufiSciently supported
by observation. In fact, the whole organ is covered by the hypo-
derm, the latter not being very different from that which covers the
body integument internally.

In a preparation of a whole larva which had molted recently
Graber's organ was made up of five pairs of black pedunculate bodies,
and one pair, anteriorly located, had the pedunculate bodies already
formed but entirely colorless; this capsule was evidently recently
formed and its pedunculate bodies were not yet completely chitinized
and fully colored.

The further fact, that the number of pairs of pedunculate bodies
increases with the age of the larva, clearly demonstrates that they
are formed successively, and inasmuch as we have to deal with an
organ dependent on the integument, these successive formations must
have something to do with the molts as modifications of the integu-
ment. All the cells and nuclei described by Graber as surrounding
the cyst and the tube consequently are merely cells and nuclei of the
hypoderm which surrounds the whole organ.

Paoli discusses further the muscles and ligaments observed by
Graber. It is found that the muscles as well as the nerves are at-
tached to the anterior capsule, which alone forms the living and
functioning part of the organ, while the remaining part is to be con-
sidered dead, having functioned in previous stages of development.

While Graber described only one pair of muscles, Paoli observed
four pairs and stated that probably Graber's second pair of nerves
corresponds to the second pair of muscles. ,Two large anterior
muscles are attached to the anterior part of the capsule with a large
surface of insertion extending from the dorsal posterior side of the
capsule upwards almost to contact with one another. These muscles
are those seen by Graber who assumed that they were attached
with the other end in the fifth abdominal segment, the seventh as
he erroneously assumed. According to PaoH, they are shorter and
inserted in the sulcus between the sixth and seventh abdominal
segments; they diverge anteriorly passing across the two large tracheal
trunks. By the contraction of these muscles the organ is displaced
towards the head.

WERNER MARCH AND 41

The second pair of muscles corresponds probably to the elements
which Graber assumed to be a second pair of nerves; their diameter
is about the same as in those of the first pair; they are attached at
the sides of the posterior parts of the cyst, extending backwards
almost parallel to one another, and finally are found to be inserted
at the sides of the opening of the terminal tube, in the furrow be-
tween the eighth and ninth abdominal segments. The action of these
muscles consequently is directly opposed to that of the first pair,
and displaces the organ posteriorly.

The two pairs of muscles described are the most important and are
attached at the lateral walls of the cyst; the two following pairs are
attached to the posterior border of the ventral side of the cyst. Of
these, the third pair, more exteriorly located, is formed by two di-
vergent muscles which are inserted at the sides of the ventral body
wall, somewhat behind the sulcate prominence bearing the anus.
The two muscles of the fourth pair are also attached to the ventral
body wall, near the median line, in front of the lip-like hook-bearing
expansion. Hence it appears that these two pairs of muscles pull
the organ in the ventral direction, between the two large principal
tracheae.

Membranous ligaments are also described by Paoli. In addition to
the four pairs of muscles and the terminal tube, the organ is held in
position by membranes, the largest of which has the shape of a coni-
cal cap involving the capsule anteriorly and attached to the poste-
rior extremity of the dorsal blood vessel, which terminates not far
from the limit between the seventh and eighth abdominal segment.
Other membranes, one on each side, envelop the basis of the muscles
of the second pair, but Paoli was not able to find the insertion of
their free end.

Concerning the nerves belonging to the organ, Graber claimed
that it was richly innervated, describing two large nerves which
formed ganglion-like masses near the capsule. These nerves, how-
ever, are muscles, according to Paoli, which are slightly inflated at
their base. It is to be noted, however, that the ganglion-like swell-
ings were reported by Graber not for the large nerves, but for the
first thin pair. Paoli found two nerves which anastomose and fuse
to a short tract, further on redividing into two branches, one more

42 THE EARLY STAGES OF TABANTD^

slender which goes to the base of the muscles of the second pair,
one larger which, extending below the muscles themselves to the an-
terior part of the capsule, corresponding to the pedunculi, while it is
not possible to affirm with certainty that it has a relation with these.
Paoli seems not to attach any importance to the behavior of this
nerve.

All the muscles, in addition, receive in their course the termina-
tions of other nerves. Those of the first pair, not far from their
place of insertion in the cyst, receive two branches derived from a
nerve which bifurcates in the neighborhood of the muscle itself, in
fact so near that it gives the appearance that the nerves go to the
cyst.

All the nerves and muscles described are related to the head cap-
sule which encloses the first pair of pedunculate bodies. The re-
maining section, which contains the dead parts of the organ, receives
towards its middle the endings of a slender nerve which, while it is a
single one in the region of the hypoderm, branches into a large num-
ber of thin ramifications.

Concerning the function of Graber's organ, Paoli likewise dis-
misses the hypothesis of a glandular function, there being no indi-
cation whatever of the presence of glandular cells.

That the organ should have an auditory function is also thought to
be of little probabihty, especially as it would differ from other cystoid
auditory organs, by having its inner surface covered with chitin,
bearing no hairs, but instead of them, pedunculate bodies. Otoliths
are, moreover, practically never found among insects, and the organ
in question would be unique also in this respect. Furthermore, in
this organ, there seems to be no special preponderance of nerves as one
would suppose to be the case in a sense organ, but rather a prepond-
erance of muscular apparatus, in as far as numerous long muscles are
connected with it in various directions, these muscles being richly
innervated. The organ is consequently, according to Paoli, destined
chiefly to be set in motion or to be extended in various directions;
in fact, displacements were observed in anterior, posterior, and ver-
tical directions, and also lateral displacement seems to be possible
if the muscles acted on one side alone.

WERNER MARCHAND 43

Paoli, in seeking for the function of this organ, recalls his own
observation of the larvae producing a crackling noise, under water
as well as in the air, similar to the sound produced by small electric
discharges. He thinks that these sounds might be produced by-
means of this organ. By the action of the muscles the cyst would
be subjected to deformations by means of which the pedunculate
bodies would be caused to hit against one another. The elastic
membranes also would aid in this process as well as the chitinous walls
of the capsule. The two large tracheae would serve as resonators.

The objection that such an organ of sound should lie at the surface
of the body rather than in its interior, is met with the consideration
that these larvas are aquatic, living below the surface of the water
or in the mud, where the possession of an organ producing sound in
the air would be of little value. The song of the Cicada becomes
much feebler when the insect is submerged under water, but the
larva in question produces its sound as well under the water as in
the air. As the pedunculate bodies are enclosed in a liquid, the con-
duction of the sound would also occur easily in a liquid element. If
the peculiar sound produced by the larva was not caused by the action
of this organ, there would be, according to Paoli, no other organ by
which it could possibly be produced. Paoli has consequently de-
cided to assign to the organ a sound-producing function.^

^ I have in the meantime found occasion to observe tabanid larvas, and was
eager to test Paoli's observations. In fact the crackling noise was freely produced
by full grown Tabanus atratus larvai, and also, in harmony with Paoli's state-
ment, it was chiefly heard when the larvae were disturbed and defending them-
selves with their sharp mandibles. The coincidence of the two phenomena was
so close that I am bound to assume that the sound is produced by means of the
mandibles. As is well known, the mandibles of tabanid larvae are strongly chit-
inized and provided with a serrate inner edge. Very near this and slightly be-
low, the maxillae, also chitinous, are situated. In the act of biting, the mandibles
are suddenly exserted with considerable force and it is conceivable that in this
act their serrate edge strikes the maxillae, producing a sharp sound. The action
itself may have some physiological importance, in as far as it serves in the lacera-
tion of the skin or body wall of animals attacked, while the sound produced ap-
pears to me more accidental, as, for instance, the gnashing of teeth in carnivora.
Hence it is not likely that Graber's organ has the function ascribed to it by
Paoli. Also Paoli's opposition to Graber's view is not well founded in as far as
he himself assumes that the pedunculate bodies are modified hairs. They could

CHRYSOPS (MEIGEN) , EARLY STAGES IN GENERAL.

In Fabricius' Systema Entomologies and Eniomologia Systematica
I have found no statement of the early stages of any species of this
genus.

Zetterstedt says that, according to Fabricius, the larvae live in the
ground. However, Zetterstedt himself saw a large number of appar-
ently new emerged adults at the borders of a lake, which would
indicate an aquatic habitat.

According to Hart, the larvae and pupae of Chrysops, as well as the
imago, are distinguishable from those of Tahanus by the antenna!
structure. Otherwise the Chrysops larvae closely resemble in
structure small or young Tahanus larvae. The dull pubescent annuli
are partly present in Chrysops, but the longitudinal lateral lines, ex-
cept on the prothorax, are shining and almost entirely without pu-
bescence. There is very little pubescence here, however, in some
young Tahanus larvae. The species described by Hart {Chrysops
vittatus Wied.) is easily recognized by the dark patch on the last
segment.

The distinguishing characteristics given for larvae of Chrysops
compared with those of Tahanus are the following: "Last antennal
joint much longer than the one preceding, dorsal areas of thorax
striated like those of abdomen." Neave compared the syphon of the

consequently take the place of the auditory hairs which are missing, and as the
pedunculate bodies, which in reality are modified hairs, would assume the func-
tion of otoHths, there would be no exception to the general rule that true oto-
liths are not found among insects. Functionally the organ could well belong to
the group where it was placed by Graber. Also the fact remains that, even
according to Paoli, nerves go to the fundus of the cj^st, where the pedunculate
bodies are attached. The problem cannot be regarded as solved by Paoli's
hypothesis. On the other hand, to assume a true auditory function seems equally
hazardous; it seems more probable that the organ might play a part in the senses
of equiUbrium and orientation, which, however, can only be determined by
further investigation.

44

AVERNER MARCHAND 45

Chrysops larva with that of Hcsmatopola (Plate 5, Figs. 73 and 74),
and found the syphon much more elongated in Chrysops.

Beling (1882) is the first to- describe the pupa of a Chrysops, which
had been found at the edge of a brook. A Chrysops larva which
pupated became known first through Hart (1895). This larva re-
mains, until recently, the only one described. After Hart, our
knowledge is increased chiefly by Hine, especially with reference to
oviposition and egg stage, on which Patton and Cragg have also
made observations. The larvae and pupae of several African species
became known through Neave's work (1915).

The little Chrysops pup^e have longer antennae, and the thoracic
spiracular prominence is more nearly in a vertical plane than in
Tabanus, its inner edge being more strongly elevated. The lower
free edge is crossed by sharp folds, making it serrated. In Chrysops
the abdominal spiracles are subcylindric near the apex; the spinose
fringes consist of long teeth only; and the terminal teeth are long
and rather narrow at the base.

The distinguishing characteristics of the pupae are the following:

"Antennae surpassing adjacent margin of head; fringes of abdomen of long
spines only; inner margin of thoracic spiracular prominences sharply elevated,
lower margin serrate-edged; abdominal spiracles slender, subcylindrical near
apex; size small."

The eggs of Chrysops are deposited, as we learn from Hart, "in one
flat tier, forming an oval or diamond-shaped area, pointed at one or
both ends." We know, however, from Hine's observation on Chry-
sops celer, that there are exceptions to this rule and there are
species of Chrysops which oviposit in several layers after the fashion
of Tabanus. Most of the eggs of Chrysops are black, according to
Hine, and are placed in a single layer, but there are exceptions to
this, for the eggs of Chrysops celer are never darker in color than
brown, and are placed in at least three layers one upon the other.

With regard to the habits of the pupal stage, Hine's observations
are of interest, as he saw around fresh water ponds myriads of pupa
skins of Chrysops with just the anterior end projecting above the
surface of the ground.

The following are the notes which we possess on various species
of Chrysops, arranged alphabetically.

46 THE EARLY STAGES OF TABANID^

Chrysops himaculosa Neave. — An African species, allied to Chry-
sops centurionis Aust., discovered by Neave (1915) on Mt. Mlanje,
in southern Nyasaland. A typical male and female and one other of
each sex were bred in October and November, 1913; one female was
collected in November, 1912.

Of the four individuals above mentioned three were bred from col-
lected pupas and the fourth from a larva which resembled that of
Chrysops longicornis, though considerably larger, with a somewhat
less strongly pigmented anal segment and with well marked hairs on
the syphon.

The hooks of the pupal aster, especially the upper and middle
pairs, are decidedly elongate (Plate 13, Fig. 167, a, h, c).

Imago, larvEe, and pupae had been taken on the banks of a wooded
stream.

Chrysops callidus Osten Sacken. — A common and widely distrib-
uted species, recorded from most of the eastern states, reaching
Florida in the South, Indiana in the West, but apparently not occur-
ring in northern New England. Hine has described the oviposition
of this species, indeed has watched the entire process of oviposition,
which usually occupies from 20 minutes to half an hour, during
which time something like one to three hundred eggs are laid.

The female alights on the leaf head downwards and begins to push
the tip of the abdomen forward towards the sternum of the thorax,
placing the protruding end of an egg against the leaf. This end sticks
fast and she then moves the tip of the abdomen backwards until
normal position is reached and the egg is free. By the same move-
ment one or two eggs are then placed to one side of this one and two
or three on the other side of it. The unfinished end is soon observed
to be V-shaped, the female moving gradually forward and placing the
end of the abdomen to one side of the V and depositing eggs along
down until the apex is reached, then changing the tip of the abdomen
to the outer part of the other side of the V and placing eggs along it
down to the apex on this side.

This process is kept up, the female changing regularly to the outer
part of the opposite side of the V each time the apex is reached.
Between 9 o'clock and noon seems to be the favorite time of day for

WERNER MARCHAND 47

oviposition, as with other species of Chrysops and Tahanus; Hine has
seldom observed females ovipositing at other hours of the day.

The eggs when first laid are clear white but gradually get darker
until they become permanently dark brown or black. It took eggs
of Chrysops callidus 5 or 6 days to hatch, and it required about a day
longer in the case where eggs were kept in the shade the whole time
than in cases where the eggs were in the sun during the day.^

The writer has made some additional observations on Chrysops
callidus, which are published elsewhere.

Chrysops celer Osten Sacken. — A species recorded from Maine,
Massachusetts, Pennsylvania, New Jersey, Ohio, and North Caro-
lina. This species is common in Ohio during the latter half of May.
The eggs have been observed commonly by Hine along the margin
of ponds and artificial lakes, clinging to various kinds of foliage over-
hanging the water. The female has been observed ovipositing on
different occasions, and is the only species of the genus placing its
eggs in masses composed of layers one above the other as in Tahanus.

Chrysops dispar Fabricius. — This species is widely distributed in
India, Ceylon, Malay, and adjacent parts. We possess some infor-
mation on its early stages through the work of Patton and Cragg in
Madras (1913).

^In the summer of 1916, I collected the eggs of Chrysops callidus 0. S., in
Princeton, N. J., from which the young larvae were obtained. These are half
transparent, whitish, and their structure is analogous to that of most tabanid
larvae, differing, however, from the young larvas of Tahanus, as far as known to
me, by having the main tracheal trunks not inflated, but of about equal diame-
ter all along their course. This character ("not provided with air sacs") is,
however, given by Patton and Cragg for all "small tabanids" observed in Mad-
ras, including species of Chrysops, Tahanus, and Hcematopota. It causes the larva
of Chrysops callidus to sink to the bottom instead of floating to the surface. The
larvae of this species molt soon after hatching, and can live under water for a
considerable time, but as all the larvae died in young stages, I am unable to
say what their further habits are, and whether they spend most of their life in
the water or later on invade the mud of the shore. At a pond where large
numbers of Chrysops had oviposited the previous summer, in spring no larvae
could be detected in the mud of the margin, while other tabanid larva3 were
numerous.

For further details see Marchand, /. N. Y. Entomol. Soc, 1917, xxv, 149.

48 THE EARLY STAGES OF TABANID^

Chrysops dispar lays its eggs, as do all the small species of tabanids
observed by these authors, invariably on blades of grass just at the
edge of a shallow stream, or on the leaves of the lotus plant at the
edges of small ponds, but never over deep water.

In Madras the smaller species of tabanids always lay their eggs in
the afternoon, commencing about 4 p.m. Chrysops dispar has been
seen ovipositing as late as 7 p.m. An egg mass of this species, on a
blade of grass, and also a single egg, is figured by these authors (Plate
1, Figs. 14 and 15), but no description of the larva is given.

The larvae of the smaller species of tabanids (including Chrysops
dispar) contain no air sacs, according to these authors, and if they
fall into deep water they die. It is important to recognize this in
breeding experiments and to place them in trays with only a little
water.

Pat ton and Cragg's text-book contains notes on smaller species of
tabanids some of which may apply to this species, but are not clearly
referred to it.

Chrysops indus Osten Sacken.— This species is recorded from New
Jersey, New York, Canada, and Ohio. According to Hine, it appears
in Ohio usually by the middle of May. It is the first species to
appear in the spring and females have been observed ovipositing on
plants growing along the margin of a small lake on the University
grounds in Columbus, Ohio. The eggs are placed in single layers on
grass blades that hang over the edge of the water.

Chrysops longicornis Macquart. — This is an African species, the
early stages of which have become known through the work of Neave
(1915). The species is the most abundant of the genus in the neigh-
borhood of Mt. Mlanje, southern Nyasaland, where Neave's investi-
gations were carried on. The flies prefer well wooded localities and
Neave thinks it probable that all stages exist throughout the year.

The larvae were first discovered at the end of August, 1913, and to
Neave's surprise, in view of the habits of the adults, were found in
the mud of a small marsh and stream bed in an open spot with only
comparatively thin woodland near it. Many other examples were
subsequently taken, both in similar places and in less unexpected
spots on the banks of wooded streams, etc. Except for an occasional

WERNER MARCHAND 49

freshly emerged individual, the adult flies were not taken in these
open places and appear therefore to migrate from them after emer-
gence and to return to them for the purpose of oviposition. If this
is the case, it is another example of the possibilities of error in search-
ing for the breeding place of a species in the spot most frequented by
the adults.

The larvae, figured by Neave but not described, were obtained in
considerable numbers from September on, a few being still obtain-
able even in January and February. In the figure (Plate 5, Fig. 72)
the eleventh segment shows a broad dark pigmented band, dorsally
reaching somewhat anteriorly to the middle of the segment, ventrally
including the anal prominence, while the tube-shaped posterior third
of the segment appears to be free from pigment, as also the slender
tube-like twelfth segment (syphon) . The circular ridges show only
traces of pigmentation. A good illustration of the pupa is given
(Plate 11, Fig. 130), which, however, shows no peculiar characteristics
by which it could be differentiated from a Tahanus pupa. The pupal
asters of male and female pupae are shown and differ considerably
(Plate 13, Fig. 171, a, h).

Chrysops magnifica, var. inornata, Austen. — An African species,
according to Neave, not rare in the Mlanje district of southern
Nyasaland, during the rains from October to April.

This species was bred from the larva by Neave, but the larva so
closely resembles that of Chrysops longicornis Macq. that Neave
never succeeded in separating it satisfactorily, since in both species
the usually distinctive characters of the pigmented anal segment and
syphon were variable.

The pupal aster (Plate 13, Fig. 166, a, h) resembles that of Chry-
sops longicornis, but the middle pair of hooks is stouter and somewhat
more curved, and the shape dift"ers somewhat, especially in the male.
The pupal asters of the male and female are illustrated.

Chrysops mcechus Osten Sacken. — A species recorded from New
Jersey, District of Columbia, Illinois, Ohio, and Kentucky. Hine
has observed the females ovipositing on foliage overhanging a mill
race at Georgesville, Ohio, June 4, 1899, but does not describe the
eggs.

50 THE EARLY STAGES OF TABANID^

Chrysops moerens Walker (synonym cestuans van der Wulp). —
Recorded from Illinois, North Dakota, Wisconsin, and Ohio. This is
the species in which oviposition was first observed by Hart (1895),
who saw these flies from August 3 to 10 flying among the marginal
rushes of Fourth Lake, Sand Lake, and Slough Lake in Lake County,
Illinois, and ovipositing on the stems of the rushes.

The egg is described by Hart as follows:

"Egg. — [Plate 1, Fig. 2, egg masses.l Length 1.6 mm., diameter 0.25 mm.,
cylindrical with rounded ends, straight or slightly curved, smooth, slightly opaque,
cream color when laid, becoming dark fuscous brown, placed in a single flat
layer, obliquely stacked as in Tahanus, about one fourth of the length of each
egg being visible at the surface, the remaining three fourths being covered by
those stacked against it. The mass is about 10 mm. long and 3 or 4 mm. wide,
its outline variable, usually diamond-shaped, both ends pointed, or at one end
short or truncate, making it more or less triangular."

Eggs of this species were observed by Hine (1906) at Sandusky,
Ohio, during the first days of July and were present in varying num-
bers during the following two months.

While the female is ovipositing she is not easily disturbed; conse-
quently one has an excellent opportunity to watch the procedure.
An illustration given by Hine (Plate 1, Fig. 1) was made from a
photograph of a living specimen which was found in the act of egg
laying and was carried, with the leaf, to the laboratory where the
picture was taken. "During the whole time," says Hine, *'she con-
tinued ovipositing without showing any signs that she was aware of
what was going on or that she had any concern for the welfare of
her eggs."

The method of placing the eggs is similar to that recorded for
Chrysops caUidus.

The female alights on the leaf with her head downwards and begins the process
by pushing the tip of the abdomen forward towards the under part of the thorax
and placing the protruding end of an egg against the leaf. The end sticks fast
in consequence of the glue-like substance which accompanies it, and she then
moves the tip of the abdomen back to its normal position, thus freeing the egg.
By similar movements one or two eggs are placed on one side of the first, and two
or three on the other side of it. The unfinished end soon becomes V-shaped;
she moves slowly forward and Hfts the tip of the abdomen to one side of the V

WERNER MARCHAND 51

and places eggs along it downwards until the apex is reached; then changes to
the other side of the V and places eggs along it, downwards to the apex on this
side. It was noted in specimens of this species that sometimes a female would
place as many as three rows of eggs on one side, one after the other, before chang-
ing to the opposite side. It is only necessary to study a mass of these eggs in
order to see the precision with which the different specimens are arranged.

The eggs are placed on various aquatic plants, often standing in
rather deep water and at times as much as 20 rods from shore. Hine
always found them on scattering plants around the edges of grassy
areas and not back among the dense growth; consequently they are
easily seen, not only on account of their conspicuous location, but also
because of their shining black color, which contrasts strongly with
the green leaves to which they are attached.

Their coloration renders the egg clusters conspicuous, and Hine
suggests that hand picking might be of consequence in the control of
Chrysops. In order to demonstrate what could be done in the way of
gathering eggs of this species, Hine went out on the morning of July
17, in Ohio, in a small rowboat, and collected for an hour. At the
end of this time a count showed 433 masses, and an average of 250
specimens to each mass, a result obtained by counting several and
striking the average, giving a total of 108,250 single eggs taken as a
result of the hour's work.

Eggs of Chrysops mcerens, laid from 8.45 to 9.30 a.m. on July 13,
hatched before noon of July 19, thus making the incubation period
6 days. This is the shortest incubation period Hine has observed
for any species of Tabanidae.

Of the larvae, Hine says that after hatching they drop into the
water, and he states that, in the natural breeding grounds of the
flies, it is almost impossible to find the very small larvae after they
have dropped from the eggs and have become more or less scattered
among the debris which is usually plentiful in these places.

In order to ascertain whether the young larvae can be killed by
spreading a film of kerosene on the surface of stagnant water over
which eggs are placed, Hine (1906) performed some experiments. A
tank of water was used, on the surface of which half a pint of kerosene
was placed to each square yard of surface. Sparganium leaves to
which the eggs were attached were brought in from the marsh and

52 THE EARLY STAGES OF TABANID^

put into a bottle as one would arrange a bouquet, and this was placed
on the bottom of the tank so that the parts of the leaves to which the
eggs were attached were a foot or more above the surface of the
water which contained the layer of kerosene. Even under these
conditions an exact count could not be obtained, because the kero-
sene appeared to affect different specimens differently. Some were
killed very quickly, some died after an hour or more, while others did
not appear to suffer particular inconvenience from the treatment.
Further observation is necessary to be able to give conclusive state-
ments regarding the matter.

Chrysops relictus Meigen.— A European species, and the first
species of the genus in which reliable data were obtained, by Beling
(1876), on the pupal stage. From three pupae found on July 16,
1876, in the sand of the border of a small meadow brook, two imagos
were produced on July 24 and 25. The third did not develop.

Beling's description of the pupa is as follows:

Pupa. — 12 mm. in length, 3 mm. in diameter, of dirty brownish yellow color.
Head shining, strongly brownish anteriorly; lower frontal margin with four
broad, rounded teeth in a transverse row ; above these teeth two small tubercles,
each with two stiff brown moderately long hairs; further down posteriorly two
similar tubercles separated by a larger distance and each bearing only one
such hair. Dorsally, at the border between the head and thorax, are two brown-
ish ear-shaped longitudinal ridges diverging posteriorly. Antenna! sheaths
laterally appressed to the head, short, terminating in a point, not much marked.
Abdomen nine-segmented, brown, with blackish segmental incisions, less shining
than the head and the leg and wing cases. First abdominal segment very short,
deeply emarginated in the middle of the anterior margin; third to eighth abdominal
segments inclusive dorsally near the posterior margin with a transverse row of
densely placed backwardly appressed, pale bristle-like teeth of unequal length,
gradually becoming longer on the posterior segments and extending also over the
ventral surface of the segments. Anal segment ending in six claw-like spines
arranged divergently, of which the two upper ones are slightly smaller than the
remaining four.

Chrysops vittatus Wiedemann. — A common species distributed all
over the eastern United States as far west as Kansas and Iowa.
This is the first species of Chrysops of which the larva has been de-
scribed (Hart, 1895).^ The larvae were found, according to Hart, in

^ Recently, in the spring of 1917, 1 have found larva; of this species from which
a male imago was reared.

WERNER MARCHAND 53

connection with those of Bittacomorpha, Limnophila, and Sialis, in
the weedy, swampy little stream at Station I of his entomological
survey of the Illinois river valley. They were quite common here,
occurring in the mud and the mats of dead stems, rarely floating at
the surface. The first were seen March 28, but they continued to
occur up to April 15, increasing slightly in size. In the breeding cage
they burrowed in the mud and through the vegetation.

In the latter part of May the water was allowed to dry up and on
the 28th all that remained was poured off. From June 1 to 3 three
pupae were formed in the damp mass of dead vegetable matter resting
on the mud in the cage. Two imagos emerged June 9, both males,
the third failing to transform.

The coloration of the larva readily distinguishes it from other
known tabanid larvae. Hart's description of larva and pupa is quoted
below.

"Larva. — [Plate 3, Fig. 39. P" Length 10-15 mm., diameter 1.6 mm. Head
light colored, mouth parts pale, tips of maxillary palpi in line with end of labrum;
body whitish, a mottled appearance within, at the middle of the body."

"Dorsal and ventral areas striate, striae entire, distinct, and not very fine,
the lateral striation a little finer, that of the prothorax very fine, with a small
smooth spot adjoining the smoother surface of its ventral area; the latter shorter
than the dorsal, not including the anterior pair of setae; median sulcus scarcely
dull-pubescent. Meso- and metathorax with lateral impressed lines, and
dull-pubescent pale annuli, but the lateral lines almost without pubescence.
Fleshy false feet of abdominals rather prominent, dorsal pair united into one,
there being no narrowing near the median line; annuli very pale except on the
last two or three segments; last segment white basally, the remainder covered
with dull blackish microscopic pubescence reaching forwards to the anal promi-
nence, a triangular extension each side of the middle above, often a small spot
accompanying each; respiratory tube whitish, spine sometimes projecting."

"Tracheal trunks sinuate posteriorly, crossing and recrossing in front of the
middle."

"Pupa. — [Plate 13, Figs. 154 and 157.] Length 9-10 mm., diameter 2 mm.
Light brownish, ferruginous, obsoletely transversely wrinkled, head and thorax
shining, abdomen duller."

^^ In the quotations, as well as in the rest of the paper, the figure numbers
refer to the illustrations in the present monograph.

54 THE EARLY STAGES OF TABANID^

"Antenna! sheaths not very thick at base, surpassing the marginal angulation
above them; carinated tubercles not prominent, lateral notches broad and shallow,
palpal sheaths indefinite, rather distant; setiferous tubercles scarcely darker;
ocellar tubercles replaced by pale dots. Rima of thoracic spiracles (Plate 13.
Fig. 157) strongly elevated from inner side, so that the flat top of the promi-
nence is nearly vertical, the upper edge of the rimal border forming a sharp carina
and its anterior extremity ending at the tip of the marginal extension in an acute
angle; the free lower edge is crossed by sharp ridges, giving it a serrate profile;
rima less curved at middle, more strongly at each end, scarcely hooked;
inner notch with radiating striations."

"The abdominal fringes consist of a single row of pale spines on each segment,
rather long except dorsally on the second, where they are shorter and thicker.
The abdominal spiracular tubercles arise from a slight elevation, tapering from
a comparatively small base as far as middle, thence nearly cylindrical to
apex, which bears a subcircular rima; on the anterior slope a small transverse
groove, not longer than the rima; tubercle about as high as its basal diameter.
Last segment with six nearly equal terminal teeth, their points marking the
angles of a hexagon; slender, even constricted at base, twice as long as their
diameter near base. Lateral spines almost wanting; ventral fringe in front of
anal tubercle in male; a tuft of about five spines on each side in place of this fringe
in the female."

Chrysops wellmani Austen. — An African species, which, according
to Neave, occurs in fair numbers near Mt. Mlanje, southern Nyasa-
land, from September to January.

The larvae of this species (Plate 3, Fig. 50), differ strikingly from
any of the other Chrysops larvae seen by Neave, in their strong pig-
mentation. In the figure it is seen that while segments 1 to 3 are
free from pigmentation, segments 4 to 11 show the regular tabanid
pigmentation on the transverse circular ridges, leaving an area around
the prolegs free from pigment. There are no longitudinal stripes, but
on segment 11 there are two subdorsal dark spots (in addition to the
broad posterior band) , traces of which are also visible in the figure of
the syphon of Chrysops longicornis.

The larvae were obtained in the beds of forested streams with those
of Chrysops longicornis and Chrysops magnifica, var. inornata, but
were much less common. They were found only between the middle
of October and the end of November. There is considerable differ-
ence in the hooks of the aster in the two sexes of this species, the
upper and lower hooks, especially the former, being much reduced

WERNER MARCHAND 55

in the female. The pupal asters of both sexes are shown (Plate 13,
Fig, 168, a, h)}^

Dorcalcemus fodiens Austen. — -An African species, on which some
observations have been made by Neave (1915) in southern Nyasaland.

Some examples of what may be the larvae of this species were found,
though the point could not be decided, as Neave was obliged to leave
before they reached maturity. They were captured during Decem-
ber, January, and February in some swampy ground on which a patch
of maize was growing. These larvae were of fair size, some 30 mm. in
length ; the syphon was very short and had a distinct pigmented ring
on the anal segment resembling that in HcBmatopota larvae. In the
more mature specimens traces of intersegmental pigment were present.

At the time these larvae were captured no other larvae of so large a
species were obtainable, and as Dorcalcemus fodiens is the only large
tabanid which is on the wing in March and April, there are grounds
for thinking that the larvae belonged to that species.

Gastroxides ater Saunders. — A black species of a genus allied to
Chrysops, with elongate slender legs and long antennae, occurring in
Pusa, Bengal, India (Maxwell-Leffroy and Kowlett, 1909). The
larva is said to breed in hollow trees. No more details are available.
A picture of the adult is given by Maxwell-Leffroy and Howlett.

^^ In the summer of 1916 I found eggs of another species of Chrysops which
dififered considerably in appearance from the two types described hitherto.
These eggs were found on the under surface of leaves of the yellow pond lily
{Nuphar), four or five clusters in all, and one on a Poniederia leaf. The
eggs were white, and placed vertically on the surface of the leaf, one beside the
other, the cluster being roundish when seen from above, presenting an even
dull white surface formed by the tips of the eggs, and with clean-cut vertical
outer walls. After a few hours these egg clusters turned somewhat brownish,
but never became very darl^. In one case the female fly was observed oviposit-
ing, but escaped before being captured. It appeared to belong to Chrysops tini-
vittahis or possibly Chrysops lugens. The larvas hatched within 5 days, molted
soon after hatching, and did not differ appreciably from those of Chrysops callidus.

For further details see Marchand, J. N. Y. Entomol. Soc, 1917, xxv, 149.

GONIOPS (ALDRICH), EARLY STAGES IN GENERAL.

The following brief characteristics of the early stages of Goniops
are given by Malloch (1917).

"Larva. — Mandibles stout, slightly curved, apically truncated; antennae
elongate, 3-jointed, basal joint stout, tapering apically, about twice as long as
apical 2 combined; apical joint much shorter than preapical; maxillary palpi
2-jointed, the apical joint slender and distinctly shorter than the basal. Thoracic
segments very distinctly tapered anteriorly, abdomen stout, roughly oval in
outline, the whole body appearing pyriform or slightly club-shaped; abdominal
segments with rather irregularly arranged transverse series of locomotor tuber-
cles; spiracular chamber in form of a vertical slit."

"Pupa. — Head without projecting thorns; antennal sheath short, curved
downward. Prothorax about one-third as long as mesothorax; wings short,
extending to apex of first ventral abdominal segment; apices of hind tarsi slightly
surpassing apices of wings. Armature of dorsal abdominal segments consist-
ing of stout thorns in a transverse series, 2 of which, near middle of segments
2 to 7, are much stronger than the others; later the series are discontinued some
distance from margins; apical segment with three strong thorns on each side,
between which are several weaker protuberances."

The eggs are, as stated by Malloch, (after Walton (?), McAtee
(?)), usually deposited on the under side of leaves of various plants,
and when the larvae hatch they drop to the ground, living after-
wards among the decaying leaves and other vegetable debris. They
are probably predacious like other Tabanida?.

The egg masses are parasitized by a proctotrypid parasite.

McAtee briefly compares the larva and pupa of Goniops, apparently
the only native pangoniid genus of which these stages have been
described, with the comparatively well studied immature forms of
the true tabanids, such as Tahanus and Chrysops.

Hart says (quoted by McAtee) :

"All the larvae of Tabanidae studied agree in the following general characters:
Body tapering at both ends, which are somewhat pointed; skin shining and glassy,
with opaque markings of a microscopic felted pubescence. The palpi have
short thick joints; the basal joint of the antennae is quite short, and there is a
bunch of stiff diverging recurved hairs between each antenna and the median

line above."

56

WERNER MARCHAND 57

The full grown Goniops larva, on the other hand, is pyriform and
not at all pointed at the ends. Its skin, except on the head and pro-
thorax, is not shining, but everywhere opaque and wrinkled or tuber-
culate. The palpi have long slender joints; the basal joint of the
antenna is very long, much exceeding the two terminal joints, and the
hairs on the antennal flap are flexible and attached to the surface
of the head. The double second joint of the antenna of first stage
larvae is noteworthy.

Of the tabanid pupa Hart says:

"The mesothorax is one-half longer than the prothorax and the second to
seventh abdominal segments are encircled by continuous fringes of slender spines."

In Goniops the pupal mesothorax is three times longer than the
prothorax, and the fringes of spines on the abdominal segments are
not continuous but interrupted and definitely grouped.

Goniops chrysocoma Osten Sacken. — A North American species,
known from not very frequent captures, from New York, New
Jersey, Pennsylvania, Ohio, and Maryland. Taken in June and
July, usually on foliage.

Goniops chrysocoma (Plate 2, Fig. 28) is the only representative of
the Pangoniinae of which we know anything of the early stages, but
in this species our knowledge is rather complete owing to the obser-
vations of Walton and McAtee.

Walton was the first to observe oviposition accurately and to
describe the egg. However, Walton relates that the first fly seen by
him was presented to him by Mr. Warren S. Fisher, of Highspire, Pa.,
who took a female in the act of ovipositing on a leaf of what proved
to be Angelica on July 4, 1907, near Highspire. The plant overhung
a small more or less permanent ditch of water, and it was naturally
inferred that the larva might be aquatic in habit, in common with
others of the family.

However, on June 14, 1908, while collecting on a dry hillside, in a
brush patch, about five miles to the eastward of the former locality,
Walton found another female of this species in the act of oviposition
on the under side of one of the leaves of a small oak sapling, to which
his attention had been attracted by a peculiar buzzing sound. The

58 THE EARLY STAGES OP TABANID^

female made no effort to escape, indeed it required considerable force
to remove the insect from her position near the eggs. The immedi-
ate locality was a hillside pasture lot, half covered with scrub oak and
berry bushes, dotted with clumps of false indigo. The nearest water
was a small overgrown ditch some 60 feet distant.

On June 18 Walton visited the same spot near Highspire, with the
hope of securing additional data, and was rewarded by finding another
fly in a similar position on a leaf of the wild cherry, thirty feet dis-
tant from the water.

The eggs (Plate 2, Figs. 28 to 32), "which are deposited upon the
under surfaces of the leaves of various herbaceous plants and trees"
(Walton), in a three-sided pyramidal heap, are described by him as
"yellowish white in color, about 1.5 mm. in length, slender, slightly
curved, and resembling those of many other flies in general appear-
ance." One of the heaps contained 534 eggs by actual count (Plate
2, Fig. 29).

The eggs of Goniops have been further observed, according to
McAtee, during each of the years 1908, 1909, and 1910, on Plummer's
Island, Maryland. On June 26, 1908, Mr. H. S. Barber collected a
female and a large greenish white egg mass, which was laid on the
under side of an oak leaf about 8 feet above the ground. The larvae
hatched June 28.

In 1910 McAtee found four egg masses on July 3 and two on
July 10. One of the first four egg masses was collected. The larvae
hatched July 7. Another had been deserted by the female by July
10. The outer layers of eggs were black, and from these issued, on
July 11, numerous proctotrupids, which Mr. J. C. Crawford says are
Telenomus, probably an undescribed species. The two remaining
egg masses of the lot found July 3 were covered by the females
until July 10 (Plate 2, Fig. 28), a period of a week, during which
time many eggs were added. These eggs, and the two masses discov-
ered on July 10 as well (Plate 2, Figs. 30 and 31), were hatched by
July 17. They were deposited on the under sides of Eupatorium,
Benzoin, and Hamamelis leaves. Some of the empty egg cases (Plate
2, Fig. 32, McAtee), usually clung to the leaf after hatching, but in
one instance not the slightest trace remained of an egg mass on a

WERNER aiARCHAND 59

witch hazel leaf. A fly heard by Mr. E. A. Schwarz giving its pecu-
liar buzz on July 13, and which undoubtedly was ovipositing then,
was located by McAtee on July 17. On July 24 the female was
absent and the eggs were hatching. The larvas, dropping to the
ground, immediately burrowed in.

These observations show that the female Goniops guards the egg
mass sometimes for a week at least, a fact which has escaped Wal-
ton's attention; that this precaution does not always prevent para-
sitism; that the period of incubation varies; and that the larvae are
fitted for a subterranean life upon which they enter as soon as
hatched.

Walton placed the two batches of eggs found by him in breeding
jars and on the evening of July 25 the first larvae made their appear-
ance. The second lot appeared two days later. The eggs are, ac-
cording, to him, yellowish white when deposited and change but
little if any in color before hatching. The process of hatching was
observed. The larvag were quite lively when hatched and "it was
a curious sight to see them come tumbling out of the eggs by dozens
when the cluster was brought under the bright light." According
to him, the period of incubation is from 7 to 10 days.

According to McAtee, eggs have hatched in from 2 to 11 days
from the date of collection. But from the fact that eggs are added
to the mass for several days, and that all hatch at the same time,
it must be inferred that the eggs within the body of the female keep
pace in development with those laid. To determine the true period
of incubation observations must cover the process from the laying
of the first egg to hatching (McAtee).

All the egg masses found on Plummer's Island in 1910 (McAtee)
were on the steep north slope of the principal elevation of the island,
which is a well shaded, cool, and damp locality. The finding of
seven egg masses in this area of less than an acre in one season shows
that Goniops chrysocoma is not uncommon locally, even though little
is known of it and recorded captures are not numerous (McAtee).

As to the larval stage, I have already related Walton's observa-
tions on the hatching process. The larvae obtained by him were
divided into three lots, the first placed in earth entirely submerged
in water, the second in dampened sand without food of any kind, the

60 THE EARLY STAGES OF TABANIDiE

third in a jar of damp earth together with some small angleworms.
Twelve hours later the first lot were all dead; both of the other
lots were lively and apparently in good condition. The second group
continued to live, without food for about ten days and then died.
The remaining group lived for some weeks but finally died also, the
angle worms being alive and uninjured.

Walton concludes that the larva is terrestrial (as confirmed by
McAtee) and the period of incubation is from seven to ten days
(see above).

The young larva (Plate 7, Fig. 93, a, b, c) is described as follows:

"The freshly hatched larva is slightly more than 2 mm. in length, slender, but
capable of contracting the body into an almost spherical mass; in color it is pale
yellowish white, semitranslucent. The head, which is capable of being entirely
withdrawn into the first thoracic segment, bears several pairs of antenna-Uke
appendages and an obtusely pointed chitinous hook."

"On each side of the median line of the body, within the second thoracic seg-
ment, there is a distinct pinkish spot; also on the last segment there is a pair of
round black spots resembling stigmata [Graber's organ (?)]; elsewhere the body
seems to be absolutely devoid of hairs or tubercles."

McAtee's description of the larva goes much more into detail.
He also described the adult larva from a specimen collected by Mr.
Theodore Pergande, and the pupa collected by him. In fact the
greater part of McAtee's article is devoted to the description of the
newly hatched and full grown larva and pupa of this species. In the
description of the larvae the segments are numbered from the head
backwards. McAtee does not accept the customary use of the terms
pro-, meso-, and metathorax for the three anterior segments of the
larvae, but speaks of them as the first, second, and third body segments,
which they really are. In the newly hatched larvs they are scarcely
differentiated from the following segments. In the full grown larvae,
while distinguishable by the surface markings, their exterior features
are homologous with those of more posterior body rings. In com-
paring and describing them, therefore, it is more natural to use
numerical designations. McAtee's description of the young larva
follows :

''First Stage Larva;. — The average length of first stage larvae of Goniops chry-
socoma which have been preserved in alcohol is about 1 mm. In life they are

WERNER MARCH AND 61

about twice as long. The larvae are not tuberculate, but the margins of each
segment from the third to the tenth, especially the front margins, are more or less
raised into low rounded rings. On a larva with arched body definite transverse
impressions behind the anterior fleshy annulus of each segment are apparent
under magnification. They render the annuli conspicuous enough, in fact, to
give an impression as of false feet to the naked eye observing the larvae crawling."

"The mouth parts are exceedingly minute and hard to observe. In arrange-
ment they suggest those of the full grown larva, described below, and the homol-
ogies have been made out accordingly, and, it is hoped, successfully. The
drawing [Plate 7, Fig. 95, a, h] is strictly diagrammatic and is made up from a
number of studies of larval heads, none of which showed all the parts in the posi-
tion used in the drawing."

"Labrum {Ibr.) short, pointed, black-tipped, and slightly curved downward.
Labium triangular, not bifid as in full-grown larvae. Maxillae {mx.) fleshy,
truncate-conical, with a short downwardly projecting lobe on the inner side of the
distal end; palpus (mxp.) arising from the end of the maxilla, first joint long,
somewhat enlarged distally, tipped by a number of short rods or spines, one of
which is larger and blunt. It may be considered a second palpal joint surrounded
at the base by a group of spines. Mandibles {md.) fleshy, blunt-tipped, crenu-
late on lower edge, lying just inside of maxillas. Antennas {ant.) straight, tapering,
directed forward; basal joint as long as first palpal joint, somewhat expanded dis-
tally, second joint double, one of its divisions longer and apparently tipped with
a seta."

"First segment of body slightly inflated, first and second segments convex
above, flattened beneath, lower lateral edge rather prominent. Second and
third segments with two or three longitudinal furrows on each side. Second seg-
ment with two conspicuous well separated elongate brownish spots visible (ap-
parently somewhat under the surface) on the dorsal aspect. Hind margins of
segments becoming more undulate posteriorly, markedly so on ninth and tenth
segments. Last segment with two round black spots, (spiracles) Graber's organ
(?), set close together on the median dorsal surface; this segment with two more
prominent ventral tubercles, two similar lateral ones, and other minute tubercles."

Both Walton and McAtee have seen the organ of Graber in the
young larva of Goniops; the former speaking of a resemblance to
spiracles, while the latter seems frankly to believe that he is dealing
with spiracles. For a possible derivation of this organ from spiracles
we possess, however, no evidence, except that in tabanid larvae all
spiracles except the terminal ones have disappeared.

Full Grown Larva (Plate 7, Figs. 92, a, b, c, and 94). — The full
grown larva described by McAtee is one of two collected by Theodore
Pergande, near Cabin John Bridge, Maryland, April 13, 1899, They

62 THE EARLY STAGES OF TABANID^

were found under stones covering the opening of mouse burrows.
The color in life was gray. The general color of the preserved
specimens was dark brown, the head black. The total length, when
the head was retracted, is given as 17 mm., the greatest diameter
7.5 mm. With the head fully drawn out the larva measures 21 mm.
For the details McAtee's description is quoted:

"Head convex above, flattened beneath, its lower lateral edge a well marked
ridge made by the stretching of the skin over the prolonged basal supports of
mouth parts. Anterior part of head marked oflf from posterior part by a band of
very thin wrinkled skin; anterior fold of this band beginning dorsally just in front
of two large lateral smooth areas containing the indefinite bluish white eye spots;
fold descending obHquely over side of the head, ending ventrally between bases
of maxillae. Anterior part of head and areas surrounding eye spots, with glassy
surface, remainder covered with thin wrinkled skin.'

McAtee describes the mouth-parts in detail (Plate 7, Fig. 95, a, h)

"Epistoma and labrum {Ibr) forming a thin lance-like projection between
upper parts of oral apparatus. Upper edge of labrum grooved from opposite
middle of antennal flap {anf.) to base of the mandibles {md.) ; provided with an
unequally two-lobed caruncle just back of the upturned tip {Ibr.). Higher lobe
of caruncle and tip of labrum each bearing a solitary anteriorly directed seta
[see Fig. 95, a]. Lower edge of labrum applied to labium {lb.)\ latter flat; thin
lateral strips diverging from a distinct ramus behind being chitinized, the
remainder flexible. Labium ending in a pair of rounded conical fleshy lips;
flexible portions closely set with short yellow hairs."

"Mandibles black, claw-like, blunt, sheathed at base by lobes of maxillae {mx.).
The latter thin, flexible, following the curve of the mandibles; their slightly
forward curved tips surpassing mandibles; their lower edge and inner side pro-
vided with numerous yellow hairs."

"Palpi (mxp.) arising from external basal flaps of maxillae. First palpal joint
inwardly and downwardly curved, second setiform, slightly curved downward
and overlapping mandible."

"Arising in the arch between epistoma and side of the head is a flap (anf.)
which seems to be part of the antennal apparatus. It follows edge of epistoma
nearly to base of mandibles, and curving down is attached to the posterior third
of first antennal joint. From this point clear around the curve to where it paral-
lels the epistoma the flap is fringed with long yellow hairs. First joint of antenna
(ani.) slightly incurved (thus being directed outward), nearly three times as long
as the two terminal joints together. Second joint conical, tapering gradually,
directed forward and downward; third setiform and directed downward."

WERNER MARCHAND 63

"Anterior part of first segment very finely tesselated, the granules being
arranged in irregular longitudinal rows. The head retracts into the posterior
part of this segment, whose exterior is a longitudinally striate thin membrane
inflated to gibbous barrel shape [Plate 7, Fig. 92]. The line of separation between
the parts of this segment is marked by a ring of fine fleshy crenulations. Second
and third segments surrounded about the middle by undulate crinkly thin-edged
folds [Plate 7, Fig. 92, a, b, c] with five symmetrically placed backward angula-
tions in each fold on each side. Parts of these two segments posterior to the
folds with as many longitudinal sulci as there are angles in the fold, and longitud-
inally striate with fine irregular wavy ridges. Anterior third of second segment
finely striate lengthwise. Part of second segment just in front of fold and ante-
rior portion of third segment granular."

"On the fourth and following segments these folds are broken up into series of
fleshy tubercles, on both sides of which the surface of the segments is raised into
low ridges, which on the anterior segments have a few, and on the posterior seg-
ments several, low protuberances. The fourth and following segments more
prominently ridged transversely on the dorsal area and longitudinally on the
ventral area. A trace of longitudinal striation remains on posterior of three
ridges (on each segment just described) or protuberances representing it. The
fleshy teeth derived from the medium segmental folds largest and most numerous
on the middle of lateral area of each segment, where they are heaped up into
irregular elevations, with two more prominent points forming a series alongside;
these elevations marked off by deep impressions both above and below and be-
coming more prominent posteriorly. There are three tubercles above (supra-
lateral series) lateral prominences and about five below (infralateral series) on
each segment behind the fourth."

"The fleshy fold is continuous across the dorsal area of fourth segment [Plate
7, Fig. 92]; back of this, dorsum of each segment marked by a depressed compara-
tively smooth elliptic area. These areas bounded in front by a varying number
of thin fleshy teeth and posteriorly by a series of low broad longitudinally striate
protuberances. Two of the latter fall into a series down the median line of the
segments, on the ninth and tenth of which they become closely approximate,
much more prominent, and round pointed. The first of the series (supralateral)
of these tubercles above the prominent lateral elevations bounds the dorsa' de-
pressions at the sides and stand in a series along the back."

"Ventral series [Plate 7, Fig. 92] of protuberances marked off from infra-
lateral series by a hiatus, by lower and thinner teeth, and by the forward arch-
ing of the series. It consists essentially of two stronger lateral teeth with a
varying number of less prominent ones between. On the tenth segment the
series is shortened and the middle elements are almost lacking. Between these
arched series of teeth and posterior ridges on the ventral surface of the segments
(which are represented by about four low round protuberances) are depressed
areas similar to those of the dorsal surface. These are bounded at sides by

64 THE EARLY STAGES OF TABANID^

first protuberances of infralateral series. Each segment from fourth to tenth
has a pair of impressed dots on inner pair of elevations of the posterior longitu-
dinally striate ridge. Segments six to nine have four of these impressions, one
outside of each of the median pair."

"On the last or eleventh segment the anus is a semicircle with the convex side
downward, overhung by four prominent tubercles in bilaterally symmetrical
pairs. Mouth of the air tube a smooth oval surface just above the anal tubercles.
It has a vertical slit and is surrounded by a projecting crenulate frill."

Pupa (Plate 11, Fig. 121, a, b, c). — One of the full grown larvae
collected by Mr. Pergande pupated and a female imago issued May
29, 1899. McAtee's description follows.

"Length of the pupal shell 19 mm. ; greatest diameter 6.5 mm. Head and thorax
of the pupa a hghter, abdomen a darker ferruginous. Head and thorax finely
and irregularly wrinkled; anterior half of each abdominal segment finely wrinkled
transversely, posterior half with wrinkles less distinct or absent (especially on
ventral surface), but very finely and closely punctuate. This makes the color
appear more intense, in places almost orange. Middle of the segments, except
the first, surrounded by an interrupted fringe of definitely grouped sharp pointed
spines, the larger of which tend to be serrate."

"Vertex of head marked by a narrow rounded longitudinally wrinkled trans-
verse ridge. In the depression between this ridge and the antennal prominences
and in front of the extremities of the ridge are two outwardly directed setae.
Antennal sheaths short, appressed, downwardly curved, conical, arising from
two low wrinkly protuberances (the antennal prominences above mentioned).
These prominences separated by a deep fold, and from them curved and diverg-
ing impressed lines run down the face. Below each antennal sheath is a widely
separated vertical pair of setae."

"Prothorax longitudinally wrinkled, except for a smooth area behind and
below each antennal sheath. A setiferous tubercle stands above each of these
smooth areas. Prothorax angulate in the median line behind. Mesothorax
three times as long as the prothorax, bearing two spiracular tubercles near an-
terior lateral angles (about opposite the middle of each lateral half of the pro-
thorax). These tubercles similar to those described below, but compUcated by
flexures of the walls. They bear at the summit upwardly arched crescent-
shaped rimae or air slits. The only setae I can find on the mesothorax are one on
each side directly back of these spiracular tubercles. Metathorax very short in
the median line, but somewhat longer at the sides, which have two rounded angles
anteriorly. Wing pads and leg sheaths, the latter slightly the longer, almost
covering ventral surface of the first abdominal segment."

"Middle of each side of first to seventh abdominal segments with elevated
round polished knobs bearing on the posterior portion of their summits the pos-

WERNER MARCHAND 65

teriorly arched crescentic spiracles. Second to seventh segments with a sharp
pointed backwardly curved spine directly posterior to each spiracular tubercle.
A short distance above this spine is a similar one, and between these two from
one to four shorter ones. Some shorter teeth occur also both ventrally and
dorsally from the stout spines. These lateral spines become stronger poste-
riorly."

"There is a definite break between the groups of lateral spines and the weak
spines forming the lateral elements of the dorsal series. This break is marked by
sharply impressed lines on segments 2 to 7 [Plate 11, Fig. 121]. Dorsal seriesof
spines on second to seventh segments consisting of a pair of stout spines on each
side of the median line [Plate 11, Fig. 121], the pair on the seventh segment being
most widely separated. On each side of the mid-dorsal pair are about three
other symmetrically placed strong spines. Between the larger spines are vary-
ing numbers of shorter ones, and gradually diminishing small ones terminate the
series on each side. All spines sharp pointed and curved backward. Ventral
surface of the second to seventh segments with smaller spines, having median
pairs of stronger teeth, most widely separated on the third segment and nearest
together on the seventh. There is a tendency for one of the minor spines on
each side of the median pair to be larger than its fellows. These smaller spines
of varying number but maintaining their series across venter of the segments,
interrupted only by the stronger ones and diminishing gradually on each side.
A wide hiatus exists between the last of the ventral series on each segment and
the group of spines near the spiracular tubercle. Eighth or terminal segments
with three strong spines on each side, connected by series of weaker points
[Plate 11, Fig. 121]. The pair made by the uppermost of these strong lateral
teeth is more widely separated than the corresponding ventral pair. In each
case the interspace (that is, the dorsal and the ventral area of the segment) is
devoid of points, except for small ones immediately adjacent to the large spines.
The location of the larval anus is marked by a rounded transversely wrinkled
knob and the spiracular eminence consists of two conspicuous tuberculate pro-
jections surmounted by sharp pointed downwardly curved spurs."

This is, so far, tlie longest description of a tabanid larva and pupa
which we possess.

Of early stages of the many exotic Pangoniinae nothing appears to
be known.

HiEMATOPOTA, EARLY STAGES IN GENERAL.

We owe to Neave notes on the early stages of this genus. He
collected at Mt. Mlanje in southern Nyasaland, Africa. The larvas
were by no means easy to obtain as compared with those of Chrysops,
a few individuals of three species only being found in September and
October. Later in the season, in January, considerable numbers of
the larvae of an unidentified species were obtained, perhaps those of
Hcematopota insatiahilis or an allied species. The larvae all seem to
resemble each other closely and it is very difficult to distinguish
specific differences in them, though they differ in a marked manner
from those of any other genus of tabanids seen by Neave. The dif-
ferences between these larvae and those of Chrysops are discussed
under that genus (Plate 5, Figs. 73 and 74). The limitation of the
pigmented areas to the anal segment and the abruptly truncated
syphon seem to be characteristic of and peculiar to this genus, as also
are the very short, sometimes almost invisible pseudopodia.

The genus Hcematopota is represented by a large number of species
in the Old World, especially in tropical Africa; the species are less
numerous in the New World and the genus is entirely absent in
Australia.

Hcematopota crudelis Austen. — An African species, originally de-
scribed from German East Africa, and by Neave found by no means
uncommon near Mt. Mlanje, in southern Nyasaland, in the months of
October and November.

The first larvae of Hcematopota found in Africa, by Neave, belonged
to this species. Several of the larvae were found in September and
October, and, though many were lost, two males and three females
eventually emerged in October and November.

The larva, which is figured (Plate 6, Fig. 85) differs from that of
Hcematopota insatiahilis only in that the pigmented areas are more
pronounced round the anus and at the base of the syphon.

The pupal aster (Plate 13, Fig. 170) has hooks of nearly equal length

and forms a very regular star. There is no true dorsolateral comb,

though one female has a minute knob-like process in place of it.

The pupal aster is figured.

66

WERNER MARCH AND 67

HcEmatopota decora Walker. — An African species, which prefers
low and dry country, therefore not common in the Mlanje district of
southern Nyasaland, where Neave made his studies. However,
Neave has acquainted us with the larval and pupal stages.

The larva was of the usual Hcematopota type, but the pigmented
areas on the anal segment were an orange color. This might, how-
ever, have been due to the mature condition of the larva, which was
about to pupate. The pupal aster is regular in shape. A dorso-
lateral comb consisting of three spines is present. Pupal aster and
dorsolateral comb of a female are figured (Plate 13, Fig. 169, a, b).

Hcematopota insatiahilis Austen. — An African species, abundant
near Mt. Mlanje, southern Nyasaland, about November and early
December, according to Neave (1915).

A few larvae of Hatnatopota insatiahilis were obtained, and from
some of these, three females were bred between November 7 and 10,
1913. The three posterior segments of "what is believed to be the
larva of this species" are figured (Plate 6, Fig. 86). There seems to be
no pigmentation except a narrow double ring at the posterior end
of the eleventh segment. The circular ridge on the tenth segment
shows no pigmentation. The twelfth segment is very short and
rounded.

The pupa is figured (Plate 11, Fig. 131) but not described. The
pupal aster of this species (Plate 13, Fig. 172, a, b, c) is very remark-
able, the upper hooks being reduced to mere knobs, while the middle
pair are enormously enlarged. A well marked dorsolateral comb
is present. Pupal aster of the female in side view and from behind is
figured, also the dorsolateral comb.

HcEmatopota pluvialis Linne. — One of the commonest and most
widely spread tabanids in Europe, known also to torment the rein-
deer in Lapland, where it occurs in incredible numbers. It is the
representative of a genus spread all over the world but absent in
Australia, while in Africa it is represented by a large number of
species.

In the more recent literature nothing is found concerning the life
history of this species, the earlier work having escaped attention of
recent investigators.

68 THE EARLY STAGES OF TABANID^

Zetterstedt believed that the larvae of Hcematopota live in the dung
(Scholtz).

Scholtz reported, as early as 1850, having found the pupae of this
species; we shall give his statement with the other facts available on
the pupal stage.

Larva and pupa of this species have subsequently been described
by three authors, independent of each other. All agree that they are
terrestrial in habit.

Brauer (1883) once obtained the larva of Hcsmatopota from the larva
of Helops lanipes (Coleoptera, Tenebrionidae) , from which when it
seemed to be in process of molting, the tabanid larva came to light.

Brauer (1869) is the first to give accurate descriptions of the
early stages (Plate 6, Fig. 83, a-d), with the exception, however,
of the egg state, which remains unknown in Hcsmatopota pluvialis,
while it has been observed in other species of Hcematopota.

Brauer found, in an excursion made in June, 1869, to Lang En-
zersdorf, Austria, near the railroad embankment, in a region of scat-
tered Cottonwood {Popuhis alba), in entirely dry soil, a white grub,
about 20 mm. in length and 3 to 4 mm. in diameter, which was un-
fortunately only briefly examined with a lens. When in the evening
Brauer began to examine it more closely the grub had pupated, and
it was only from the remainder of the larval skin that the character
of the larva could be ascertained. However, the description obtained
from this meager material is very accurate.

The larva (Plate 6, Figs. 83, a, b, 84, and 91) is cylindric, twelve-segmented
'(head included). The head (Plate 6, Figs. 84 and 91) is formed after the type of
ttabanid larvae, almost completely differentiated, the chitinous plates diverging
.behind and exceeded in length by two long rods in the middle, which are con-
tinuous with the middle chitinous plate of the head and with the labrum. The
latter is narrow, somewhat curved downwards, widened at the tip, and ciliated
laterally. Close to the sides are the hooked, downwardly curved parallel man-
dibles, their margin being convex and serrate. Below the mandibles and partly
within the concavity formed by them are the maxillae, which are soft in appear-
.ance. Their basal part is globular and slightly spiny, the distal part is formed
by a finger-like piece; at the side of the latter, externally, a two-jointed palpus of
twice its length, the two joints being of equal length. The terminal joint, more-
over, is widened and excavated to spoon shape. Above and at the sides of the
.mandibles are the antennae, of which the distal joints are cylindric and simple,

WERNER MARCHAND 69

the first one thick and long, the second short and narrow. There are no bristles
at the base of the antennae. The eye-spots are small, perhaps more distinct in
life, and situated behind the middle of the head. The whole head is deeply
retractile. The body is pure white in color with many longitudinal striae, and
bearing, on the fourth to tenth segments, small fleshy warts on the lateral and
ventral sides, each segment having four. These warts can be retracted, giving
the larva an almost smooth appearance. The last segment ventrally with a
thick half globular anal prominence, and posteriorly with a short conical structure
with a vertical two-lipped fissure. The lips of the latter are strongly chitinized,
transversely sulcated, each of them leading to one of the main tracheal trunks;
they represent the posterior spiracles of the larva, while the small anterior
spiracles are situated on the posterior half of the second segment.

The whole structure of these respiratory organs indicates that the
larva had not lived in water or left it only for pupation, as, according
to Brauer, in all tabanid larvae of terrestrial habitat the posterior
spiracles are analogous in structure, while the aquatic larvae of Ta-
hanus autumnalis and Hexatoma pellucens are enabled to extend the
last segment like a telescope, showing a similar fissure only at the end
of this tube.

A much more detailed description of this larva is given byPerris
(1870), who found it twice, and each time in a single individual,
in the decayed wood of old pine stumps. Of the two specimens
found one was sacrificed to be studied; the second one metamorphosed
and produced, to the surprise of the observer, the species Hcematopota
pluvialis. This fly, as he says, is very common in regions of France
where there are no pines, and so abundant in the Landes that he
could not have missed the larvae if the stumps or the bark of dead
trees were its normal place of development, as he had especially
studied all insects inhabiting the pine. Consequently, Perris as-
sumes that the insect has deviated from its ordinary habits. How-
ever, the larvfe were found in the pine wood detritus itself where they
apparently had lived, and in which the larva mentioned changed to
pupa after three months.

A translation of Perris' description of the larva (Plate 6, Fig. 82, a-e)
is given below :

Larva (figured) 12 mm. in length. Smooth, hard, leathery (coriaceous), cylin-
dric, attenuated at both ends, anteriorly more than posteriorly, very finely and
densely striated in a longitudinal direction.

70 THE EARLY STAGES OF TABANID^

Under ordinary conditions the head, which is small in comparison with the
body, reddish and semicoriaceous, appears conical; but when protruded for-
ward, it forms a sort of neck posteriorly. It is retractile to the extent that it
can disappear completely in the first segment. The front is protruded into a
fleshy mouthpiece which Perris assumes to be the labrum. To the right and
to the left of this a black horny linear piece is seen, which is curved anteriorly
and prolonged posteriorly into a rod, becoming more and more thinned out, which
is observed through the transparent tissues and which is inserted deep in the
first segment. These two rods are at first parallel, then somewhat diverging.
They serve, according to Perris, as places of attachment to the muscles which
move the head and the linear piece mentioned, which are the mandibles. Below,
two fleshy subconical pieces are seen, fused together at their base, of which the
inner is separated from the outer one only at the tip of the latter. The external
piece, which is a maxilla, is surmounted by a rather long palpus of two joints,
the first of these joints one and a half times as long as the second; the inner
piece is the lobe of the maxilla. Above, on each side of the head, opposite to the
base of the maxillae, a rather long antenna is situated, which appears to consist
of three joints, of which the first is very short, the second four or five times
longer, the third slender and about one-half as long as the preceding.

This larva has, consequently, a protruded labrum, two horny serrate man-
dibles, two maxillae with their lobes, two maxillary palpi consisting of two joints,,
two antennae consisting of three joints, all extending beyond the head.

The body is formed of eleven segments; the first, or prothorax, truncated
conical, is one and a half times as long as all the other thoracic segments, which
are slightly shorter than the abdominal segments. The first six of the latter
have eight fleshy and retractile protuberances, two dorsal ones distant from one
another, two on each side placed close together, and two ventral ones placed at
a distance from one another like the dorsal ones. These protuberances are of
great help to the larva when it starts to walk. In this act, the larva presses
the head against the surface on which it is to move and contracts the body, at the
same time retracting the protuberances; then, freeing the head and exserting the
protuberances which, all around, exercise a strong pressure, it pushes forward.
The last segment of the abdomen, which is somewhat rounded at the sides, is
rather suddenly and considerably narrowed down posteriorly, where it terminates
in a disk in the center of which a reddish vertical slit is observed with the aid
of a lens. At high magnification this slit is recognized to be a spiracle the lips of
which are transversely striated. In seeking for the remaining respiratory orifices^
two of these oval and reddish spiracles are found at the posterior border of the
prothoracic segment, one on each side. Under the last segment, in its anterior
third, there exists a fleshy prominence, of the shape of an ellipsoid placed trans-
versely, and retractile; it is traversed by a longitudinal fissure or sulcus which
indicates the anus. This prominence also serves to aid locomotion.

WERNER MARCHAND 71

The body of the larva is completely covered, except in the anal region, with
very fine and densely placed longitudinal striae easily seen with the lens. They
certainly play a part in the contractions and dilatations of the body.

The larvae of this species liave been found and described again by
Behng (1875). One larva was obtained by him on March 24, 1873,
from hog dung which had been collected from a meadow and piled
up at the edge. This larva" was kept in a room which was not heated,
and pupated on June 24, producing a male imago on July 6, after a
resting period of twelve days. From December, 1873, Beling fre-
quently found single larvae in the soil of cultivated fields, several
times also in the earth of hedges bordering on meadows. The larvae
which were collected were kept in a glass jar partly filled with humus
and continued to live for months in this medium, without injuring
one another. However, at the time when metamorphosis was to
take place, they killed one another, apparently for the purpose of
feeding on the killed individuals by sucking out their contents, and
of the ten larvae finally only one remained, which pupated on June
25, 1874, and produced a female imago on July 5, after 11 days of
pupal condition.

According to Beling, the larvae resemble closely those of Tabanus
hromius in color, appearance, and habits, but are at once distin-
guished by being shorter and thinner, and by the much finer punc-
tated^^ longitudinal striation. Concerning localities of capture,
Beling states that the Hcematopola larvae occur chiefly in humus of
cultivated fields, while the larvae of the terrestrial Tabanus species
mentioned are found usually in grass-covered or meadow soil.
Beling 's description of the larva is translated below.

Larva. — 10 to 12 mm. in length, 2.5 mm. in width, somewhat contractile; when
fully extended, up to 15 mm. and even more in length, twelve-segmented, pale
yellow, strongly and glassy shining, with very fine irregular (punctated) longitu-
dinal striation, somewhat narrowed in front and behind, and especially during
locomotion narrowed down to spindle shape in the anterior part, with small
narrow brown head completely retractile into the first segments, appearing, when
seen from the dorsal side of the first two segments, as a brown longitudinal
stripe, broadened behind and ending in two points. The last body segment is
shorter and narrower than the preceding one, rounded at the posterior border,

^^ The German word used is " Nadelrissig."

72 THE EARLY STAGES OF TABANID^

with a cushion-like prominence on the middle of the rounded part. The promi-
nence is often in the shape of a short thick cylinder, truncated at right angles,
bearing in the middle the vertically placed stigmatal fissure. Anus prominent
on the under side of the last segment. On the under side of the fifth to the elev-
enth segments inclusive are transverse locomotor swellings near the anterior
margin, forming six longitudinal rows, of which four on the ventral side are
arranged in two pairs, not far distant from one another, while in the outer rows
ridges are placed at a greater distance from one another and are also placed farther
back from the anterior segmental border.

Scholtz (1850) reports having found the pupae of this species, to-
gether with those of Tahanus autumnalis and Tahanus tropicus, on
an excursion with Prof, von Siebold and Dr. von Frantzius in the
neighborhood of Breslau, in 1850, at the edge of a pond which was
entirely covered with Lemna, and the water of which was polluted
from manure piles surrounding it. Here the pupae were found
quite near the edge under a thick moist mass of decaying Lemna,
together with the pupae of Stratiomys , Syrphus, etc.

A translation of Brauer's description (1869) is as follows:

The pupa (Plate 6, Fig. 83, c, d) 15 mm. in length, slender, with no spines at
the anterior end, bearing here only two small tubercles. The wing and leg
cases do not extend beyond the first abdominal segment; the second to tenth
abdominal segments bear a circular armam.ent of bristles, the last segment end-
ing with a thick and only slightly diverging fork which is more or less embedded
in the cast larval skin.

From Brauer's pupa there appeared, after two weeks, a male of
the species in question.

More precise in detail is the description given by Ferris (1870),
obtained from the cast pupal skin:

The pupa (Plate 6, Fig. 82, /, g) presents all the parts of the perfect insect.
The organs of the head can be guessed rather than seen. Two frontal tubercles
(trouques), surmounted by a bristle, seem to be the cases for the antennal bases;
two other oblique tubercles overlapping with their borders, in the region of the
mouth, seem to hide the palpi. Between these two tubercles and those pre-
ceding them two patterns are formed by small elevated lines, the upper ones
forming a parenthesis, and the lower ones a sort of brace. Exterior to these
patterns there exists, on both sides, a transversely triangular prominence, with
its summit overlapping a little the border of the head. I (Perris) consider these
prominences, to judge from their location and form, as the cases for the distant

WERNER MARCHAND 73

division of the antennas. Below there are two small tubercles placed obliquely
one below the other and surmounted by a little hair, and on the vertex four
tubercles arranged transversely, the outer ones larger and bearing a hair. The
dorsal part and the sides of the thorax bear some small hairs. The abdomen,
consisting of seven segments, presents on the posterior thirds of each of the
first six a circular series of stiff bristles, unequal in length, thick at their base
and directed backwards. The last segment is small and ends in a small crown of
six conical diverging teeth, the two upper ones a Httle smaller than the others.
Rather near the anterior border of the thorax two obtuse protuberances arc seen,
somewhat more excavated and striated at the summit, crossed by an oblique fur-
row and inside showing a depression, the thoracic spiracles of the nymph. There
are also visible the very conspicuous marks of six abdominal spiracles; they are
situated laterally on the first six segments.

It is conceivable that the tubercles, the hairs, and the bristles with which this
pupa is armed, are very useful when it wants to move or to reach the open air in
order that the perfect insect should not be hindered in the act of eclosion. In
the eclosion the shell of the pupa splits all along the median and dorsal side of
the thorax.

Beling also describes the pupa of Ilcematopota pluvialis.

Pupa. — 10 to 13 mm. in length, about 2.2 mm. in width, cylindrically rounded,
of a dirty brownish yellow color, slightly shining, somewhat narrowed behind.
Below the two small front teeth placed beside one another, and above the frontal
margin two small tubercles, not much farther distant from one another, and
each bearing a long bristle. Above this tubercle, som.ewhat distant from it, is
another separate hair, and, in addition, on each side of the thorax are two hairs,
one directly located behind the other. Abdomen nine-segmented, first segment
very short, hardly reaching to the fourth part of the length of the second seg-
ment; third to eighth segments inclusively surrounded, near the posterior mar-
gin, with a circle of rather densely placed brownish bristle-shaped teeth, of un-
equal length, appressed backwards, and shorter on the ventral than on the dorsal
side. Last abdominal segment small, bearing in the middle of its under surface
a row of similar bristle-like teeth as the other segments, and at the end with six
outwardly divergent comparatively strong teeth with blackish tips, arranged in
a hexagon of which the upper two are shorter and weaker than the remaining
ones.

Hcematopota sp. — Two undetermined species of Hcematopota were
observed by Patton and Cragg (1913) in Madras, to oviposit, as all
the small tabanids observed by these authors, "on blades of grass
just at the edge of a shallow stream, or on the leaves of the lotus
plant at the edges of small ponds, but never over deep water."

74 THE EARLY STAGES OF TABANID^

These species of Hmnatopota would then fall under the heading
of those larvae which, according to Patton and Cragg, have no air
sacs and, consequently, perish when falling into deep water. Like
the other small species, these Hcematopota '^ spread out their eggs in
one or more layers on blades of grass." The eggs of one species of
Hcemalopota are dark gray when deposited, according to the same
authors. No descriptions or figures are given. The eggs are said
to be torpedo-shaped (not cylindric) in all the smaller tabanids in
Madras, which should hold good for Hcematopota.

Hexatoma pellucens Linne. — A European species, according to
Schiner common in Austria, attacking man as well as animals.

In the middle of June, 1868, in a garden in Breitensee, Austria,
Ernst Marno found in a reservoir 8 feet deep, filled with rain water
and dung fluid, together with larvae of Eristalis and Culex, some dip-
terous larvae of about 25 mm. in length, which were on the dorsal
side beautifully mottled with black. Some of them kept in this
water had died the following morning and were preserved in alcohol.
Brauer, who was consulted, pronounced them to be tabanid larvae
of an unknown species.

Early in July, after a heavy rain, larvae were again found, which
in color and appearance seemed to be the same species, and of which
in the middle of July when they apparently were full grown, a consid-
erable number were collected from the mud of their habitation.

Of these, however, many perished which were continually replaced
by new captures. On this occasion also empty pupal cases and some
full ones were found; the latter had already made some progress
in the transformation to the perfect insect, but had fallen into the
water and had soon died and decayed.

Brauer determined the contents of these pupte, as well as a dead
fly which was found in the water, as belonging to Hexatoma. However,
a search for the pupse in the soil surrounding the reservoir was in
vain, and there remained some doubt as to the identity of the larvae
and pupae.

The larvae captured by Marno were kept by him in a shallow
flower pot, which was filled half with moist earth and half with
moist decaying foliage, as it had been observed that the larvae ap-

WERNER MARCHAND 75

parently showed a preference for this environment and had been
seen to work their way into a decaying leaf. They did well under
these conditions; only a few of them died, and on July 22, in the
afternoon, the first soft pupa was found, the larval skin still adher-
ing to it, and identical with those found dead in the water.

A few days later, when a search was again made for the larvae,
none of them were found, but instead it was discovered that they
pupated in the interspaces of the somewhat defective brick walls of
the reservoir, and here some pupae as well as larvae were collected,
of which the latter were more or less contracted in preparation for
their metamorphosis. From these pupae the first fly was obtained
on July 27, in the morning.

The duration of the larval stage is about three weeks; the duration
of the pupal stage Marno could not ascertain, leaving his material
to Brauer to study.

The larva (Plate 6, Figs. 89, a, b, and 90, a, b, c) has been de-
scribed by Brauer (1883), who reports it to be aquatic, being found in
pools and puddles. A translation of Brauer' s description is given
below.

Larva, when extended, slender and spindle-shaped; when contracted, obtusely
cylindric, the anterior and posterior segments being retracted in the latter case.
The body consists of a head capsule and eleven segments. The cuticle has a
striped appearance because of the many closely arranged longitudinal furrows.
The first three segments and the under surface are bone-white (yellowish white),
the dorsal surface of the remaining segments shows beautiful grayish brown pat-
terns on a white background, brought about by microscopic pubescence. The
head capsule (Plate 6, Fig. 90, a, b, c) is narrow, formed like that of all tabanid
larvae, compressed, and can be retracted into the second segment. It consists
of four plates which posteriorly are not connected, and of which the two middle
ones are drawn out into long rods. The lateral plates show on half their length
on the outside an irregularly rounded and convex eye-spot on each side. At the
anterior end is a comb-like upper lip reaching above and between the mandibles.
On each side of its base there is a cushion-like projection thickly beset with mostly
two-pointed bristles, on which the antennae are inserted, which are short and two-
jointed. The first joint is cylindric, the second divided lengthwise into two
parts. The exterior point is shorter than the interior. Below and to the inside
of the antennae are the hooked and more strongly chitinized mandibles, which
may be moved up and down parallel to one another. Their surface is trans-
versely sulcate; consequently the anterior convex margin is serrately incised.

76 THE EARLY STAGES OF TABANED^

Within the concavity of the mandibles the softer maxillae are placed. The ex-
tremity of the latter, curved hook-like in the same way, runs parallel with the
mandibles, so that to the observer from the side, two hooks appear superposed
one on the other. The maxillary palpus projects obliquely towards the outside
and above, and has a thick cylindric basal joint and a thinner simple terminal
joint of double length. Beginning with the third body segment, there appears a
short intermediate segment, on which rather long bristly hairs are inserted. On
the ventral side several circular protuberances with bristles are noticed on the
intermediate segment. ^^ The fourth to tenth segments show on the dorsal side
two, and on each side one spinous protuberance. The last segment has, on the
ventral side, the paired anal prominence, covered with finer and stronger hairs
and surrounded by a three-sided furrow.

At the basal angles of the anal prominence small wart-like swellings appear.
From the last segment a respiratory tube, may be extended which is rather sharply
pointed and narrowly cone-shaped, showing at the tip a vertical fissure between
two thickenings, the endings of the main tracheae of the metapneustic larva.

The pupa is a free pupa, and is found in the ground.

The head of the pupa bears on each side above the base of the antenna a
small wart with a bristle, and above and below the antennal sheath, on the eyes,
similar but slightly smaller warts. B ehind the prothorax are the anterior spiracles,
placed on small cylindric prominences which do not project far. The wing and
leg cases reach the posterior margin of the first abdominal segment.

Length of the larva: 27 to 30 mm.; width in the middle 4 mm.

^'The German word used here is " Zwischenwulst" which means a swelling
lying between the segments.

TABANUS, EARLY STAGES IN GENERAL.

Much that can be said about Tahanus in general has been said
already under the family Tabanidae.

The eggs of Tahanus are not appreciably different from those of
Chrysops, except that the latter resemble, according to Patton and
Cragg, more those of the smaller species of Tabanus, the eggs of the
larger species being of shghtly different form. According to Hart,
the eggs of Tabanus are always laid in clusters composed of several
layers, but recent observations have shown that from this rule ex-
ceptions are frequently met with, when the eggs are spread out in
one layer or even loosely scattered. As a rule, however, the eggs are
laid in a cone-shaped or bilaterally symmetrical mass containing
several layers of eggs and attached to various objects as specified
under the heading of Tabanidas.

The larvae and pupae, studied by Hart, form two groups, which may
be called the lineola and the atratus groups. The larvae show the
most distinctive characteristics in the structure of the surface es-
pecially of the prothorax, which has lateral, dorsal, and ventral
shining areas, Hmited in front by an opaque pubescent annulus.
Comparing the anterior extension of these shining areas, Hart found
them all of about the same length in Chrysops and the lineola group
of Tabanus, although this length varies in Chrysops. The lateral
prothoracic areas are extensively invaded by the anterior annulus in
the atratus group of larvae, the striated shining space becoming basal,
being not more than half as long as the dorsal area. The upper
lateral space of the mesothorax is not closely striate, and is quite
shining. In all except some of the atratus group the remaining lat-
eral striation, including that of the prothorax, is not much finer, and
is also shining, but in the others, as in atratus, the prothoracic and
abdominal lateral striation becomes microscopically fine and even
subopaque. A smooth spot near the lower hind angle of the pro-
thoracic lateral area furnishes another good differential character.
The dorsal and ventral striation varies in extent according to age,
but the thorax is striated above in Chrysops, and smooth or nearly

77

78 THE EARLY STAGES OF TABANID^E

SO in Tahanus. The three types of coloration observed are well
shown by the three Tahanus larvae figured (Plate 3, Figs. 40, 41,
and 42).

PupcB. — The great difference between the abdominal spiracles and
terminal teeth of the lineola group and those of the atralus group is
evident from Hart's figures (Plate 13, Figs. 162 and 163). In both,
the thoracic spiracles are in a plane nearly parallel to the adjacent
surface and the spinous fringes contain long and short spines. The
preanal fringe in the lineola group shows more or less of a chitin-
ous web, uniting the bases of the spines.

A pair of short appressed palpal sheaths on the lower surface of the
head, resembling the antennal sheaths, differs slightly in position in
related species of the atratus group.

The species of Tahanus have been classified by various authors,
but for convenience, I have listed them here "n alphabetical order.

We possess data on early stages of fifty-five determined, and four-
teen undetermined species of Tahanus. Those of some of the worst
stock pests still await description.

Tahanus alhimedius Walker, — This is one of the commonest Indian
species of Tahanus, according to Patton and Cragg, and occurs almost
all the year round in Madras.

Observations on the life history and early stages have been made
by Patton and Cragg (1913). The species oviposits in a variety of
situations, but most frequently on the leaves of some plant over-
hanging deep water. The egg masses have also been found on small
rocks in the bed of a stream, and on pieces of string hanging over
house drains; on one occasion a mass was found on a papaya tree, at
the foot of which water was occasionally allowed to flow. The num-
ber of eggs laid is between 500 and 600. The eggs, as in all the
larger Indian species mentioned by Patton and Cragg {Tahanus
striatus, alhimedius, and speciosus) are subcylindric with tapering
ends. The eggs of Tahanus alhimedius measure 1.9 to 2 mm. in
length and -4 mm. in breadth. The eggs are brownish white when
deposited A chalcid parasite, which was not identified by Patton
and Cragg, regularly destroys large numbers of egg masses of this
species, and of striatus. This species, as the larger species of India in

WERNER MARCHAND 79

general, oviposits any time during the day, but as a rule in the
morning.

When "one of the larger species" (including Tabanus albimedius) is about to
oviposit "it alights on the leaf or blade of grass with its head downwards; it then
thrusts the tip of the abdomen forwards under its thorax, and deposits an egg,
which adheres to the leaf owing to the sticky substance which accompanies it.
The abdomen is then returned to its original position, and as soon as the next
egg is ready to be laid, is again flexed, and the second egg is placed at one or
other side of the first. In this way three or four eggs are laid on one side of
the first and three or four on the other. The mass at this stage has the shape
of a V. The fly now moves forwards, and, raising the end of the abdomen to
one arm of the V, places a number of eggs down the side until the apex is reached;
she then changes over to the other side and deposits eggs all down that arm up
to the apex. In the end a raised compact mass of eggs is built up, which, if
examined with a lens, demonstrates the precision with which the eggs are placed
in reference to one another."

The larva of Tabanus albimedius has not been described and also
not figured, by Patton and Cragg, though they evidently had it
under observation. A good characterization of tabanid larvae in
general is given, but the differences of the various species are not
entered upon.

Of the larva of this species, by the same authors, dissections have
also been made, and detailed illustrations are given of the mouth-
parts (Plate 7, Fig. 97), and of the alimentary tract (Plate 7, Fig. 98).

The figure given of the head shows the mouth appendages: labrum,
maxillary palpus, maxilla, antenna, and mandible; also the lateral
area of spines behind the antennae. On the figure the antennae appear
to be two-jointed, while the authors state that they are three-jointed
in all tabanids.

The figure given of the alimentary tract shows the esophagus,
proventriculus, the narrow salivary glands, apparently inserted in
the pharynx, the mid gut, hind gut, and Malpighian vessels, four in
number.

Patton and Cragg were the first to publish anything on the ali-
mentary tract of a tabanid larva based on dissection.

Tabanus atratus Fabricius. — A species inhabiting eastern North
America, as far west as Colorado, conspicuous by its large size and

80 THE EARLY STAGES OF TABANID^

uniformly black color. About its early stages we are informed by
Walsh, C. V. Riley, Hart, and Hine. As Hine summarizes, the eggs
are deposited around marshy places on grass and sedges, and the
larvae are found by digging in the mud.

Hart (1895) is the first to give notes about its oviposition, which
was noticed in Illinois, in July and August, and egg masses then be-
came frequent. The last date for the imago was August 15. On
August 1 1 , a female was observed ovipositing on the side of a wooden
frame standing over the water. The egg mass was placed in a breed-
ing cage, and one week later, on the 18th, many larvae hatched
from it. Another egg mass of the same form and appearance, placed
on the dark bark of a stick projecting from the water, was brought
in July 27, from which hatched on August 4 larvae apparently of
this species.

Hart's description of the egg mass and egg is given below:

"Egg Mass. — Blackish-brown, subconic, with oval base, 10-15 mm. long
and 8-10 mm. wide, height 5-7 mm.; sides convex or concave, apex correspond-
ingly rounded or pointed; eggs pointing obliquely upward and towards one
end, both sides meeting upon that end in a more or less prominent longitu-
dinal crest. The eggs are stacked in four or five tiers, one above another, and
gummed together into a firm mass."

"Egg. — Length 2.5-2.7 mm., diameter 0.4 mm. Dark brown, subcylindrical,
ends more or less tapering and curved, surface minutely rugose and subopaque."

Hine completed these observations in 1906. According to him, the
eggs of Tabanus atratus are generally placed in masses of various sizes
on the leaves and stems of grasses and sedges and other plants growing
in marshy or wet ground, but not necessarily in the water. A single
mass may contain as many as 500 eggs, but often they are smaller
and they may be larger; they are white when first placed, but soon
turn brownish. The mass is very convex and composed of several
layers, one above the other, the bottom layer being attached to the
surface of the leaf or stem and the other layers each to the one that
was placed before it. Each egg is elongate, spindle-shaped, between
2 and 3 mm. in length, and narrowed at each end. A female was
observed, in Ohio, ovipositing on June 23 at 11 o'clock. The eggs
were taken and kept in a room out of the sun, where they hatched on
the morning of July 2 before 6 o'clock, thus requiring an incubation

WERNER MARCHAND 81

period of nearly nine full days. It has been proven that the eggs
of tabanids hatch more quickly when exposed to the sun during the
day (see Chrysops), where they are usually deposited; therefore the
time given is probably too long for eggs under natural conditions.

There is no definite way, as far as our knowledge goes, according
to Hine, of telling the eggs of the black horse-fly from those of other
species of this genus, but being a large species the masses are much
larger than in some others, and are more convex than usual. The
particular place of oviposition is in a measure characteristic.

It is of interest to note that the egg masses of Tabanus atratus are
often found to be parasitized.

When the two egg masses mentioned and described by Hart had
produced larvae they were placed in a dry vial, and a little later it
became evident that both masses had been parasitized by Hymenop-
tera, minute black imagos emerging freely in the vial. An exami-
nation of one of the masses showed that about one-half of the eggs
had been infested. Examples of the imago were sent to Mr. W. H.
Ashmead, who found the species to be a new one. It was described
by him as Phanuriis tahanivorus (Ashmead).

I give the description of this insect at the end of the chapter on
Tabanus, under the heading, Parisites of the early stages of Tabanidae.
The illustration of this insect, given by Hart, has been omitted.

Hart is also the first to tell us about the early larval stages of
Tabanus atratus. From the two egg masses he obtained larv«
hatched as already stated, on August 4 and on August 18. In the
latter case, the incubation period was determined to be one week.
The larvae were at this time commonly found in water among the
vegetation, less commonly in the sand of the shore, and young indi-
viduals became frequent.

Larva, Newly Hatched {According to Hart). — "In this stage the lateral areas
are sculptured similarly to those of the adult, but the dorsal and ventral areas,
though shining, are rather sparsely striated. Traces of the dark markings are
visible, especially on the posterior segments."

According to Hine, the larvae, when first hatched, are about 3 mm.
in length, white, and with a narrow darker shade at the union of
each tv/o segments. As soon as they drop to the ground they begin

82 THE EARLY STAGES OF TABANID^

to burrow and are soon beneath the surface, where they cannot be
seen. At first these larvse are very hard to see on account of their
small size; consequently not much has been learned of their habits
under natural conditions; but when they are nearly grown they are
to be found in a variety of places. We now come to a discussion of
the adult larval stage.

Walsh (1863) gives a description of an aquatic Tahanus larva which
he was unable to identify as the imago obtained from this larva was
in too bad a condition to be identified, having remained many weeks
without care, in the breeding jar. From his descriptions, however,
and also from his remarks on the imago, it is evident that the larva
belongs to Tabanus atratus. There is in Massachusetts no other
Tahanus known of this size and with wholly brownish black wings,
and the description of the larva agrees with that given by later
authors for undoubted Tahanus atratus larvae. However, some of
Walsh's smaller specimens may belong to Tahanus stygius or other
species.

Walsh found his aquatic larva, on many different occasions,
"amongst floating rejectamenta." On one occasion he found six
or seven specimens in the interior of a floating log, so soft and rotten
that it could be cut like cheese. Once he discovered a single speci-
men under a flat submerged stone, in a little running brook. Finally,
once he met with one alive, under a log, on a piece of dry land which
had been submerged two or three weeks before, whence it appears that
it can exist a long time out of the water. Walsh had, on several pre-
vious occasions, failed to breed this larva to maturity, and the only
imago he had was obtained in 1861 from larvae which, suspecting
them to be carnivorous from the very varied stations in which they
had occurred, he had supplied with a number of fresh water mollusks,
but the habits of which, in consequence of having been away from
home, he was unable to watch. On September 2, 1863, he found a
nearly full grown larva among floating rejectamenta, and between
that date and September 23 this larva devoured "the mollusks of
eleven univalves" (genus, Planorhis) from one-half to three-fourths of
an inch in diameter; and on three separate occasions observed it work
its v/ay into the mouth of the shell. In this operation the pseudopods
were energetically employed and Walsh found, on cracking the shells

WERNER MARCHAND 83

after the larva had withdrawn, that a small portion of the tail end of
the animal was left untouched — no doubt in consequence of his
being unable to penetrate to the small end of the whorl in the shell —
and also the skin of the remaining part and the horny tongue
membrane.

Walsh's description of the larva of Tahanus atratus, including that
of the imago which serves to establish its identity as to species, is
given here:

" Tabanus — ? Imago male. Blackish. Legs blackish, wings brownish fuscous.
Length 0.70 inch. Expanse L30 inch. One decayed specimen came out between
June 14 and July 14, 1861, from a larva found early in June."

"Larva, from the living specimens, obtained August 14, 1860, and September
2, 1863. Length 2.25 inches when extended, 1.75 inches when contracted; di-
ameter 0.25 to 0.30 inch. The specimens found in 1863, 0.25 inch shorter.
Body cylindrical, twelve-jointed, the three or four terminal joints much tapered
at each side of the body, but more so anteriorly than posteriori}^, and joints one
and eleven, each with a retractile membranous prolongation at tip. Joints one
to ten are subequal; eleven is about two- thirds as long as ten, and twelve about
one-fourth as long, and 0.5 inch in diameter. Color a transparent greenish white,
paler beneath, on the anterior and posterior margins of joints two to eleven, the
anterior annulus laterally connected with the posterior by two to four dark green
lines. On the dorsum of four to nine, and more obscurely on ten, a dark green
basal triangle, extending half way to the tip; joint one with paler markings, and
with no dark annulus behind; joint twelve entirely fuscous. Plead small, appar-
ently fleshy, pale, truncate-conical, 0.03 inch wide, and about 0.04 inch long in
repose, inserted in joint one without any shoulder. The trophi occupy two-
thirds of its length, but it has a long cylindrical internal prolongation, extending;
to the middle of joint two, which is som.etimes partially exscrted, so that the headi
becomes twice as long as before. All the trophi are pale and apparently fleshy
except the mandibles, which are dark colored and evidently horny, and they^
have no perceptible motion in the living insect. The labium is slender, a little
tapered, and three times as long as wide, on each side of and beneath which is a
slender thorn-like decurved brown-black mandible. The labium resembles.
the labrum, but is shorter, and on each side of it is a slender palpiform but exar-
ticulate maxilla, extending beyond the rest of the mouth in an oblique direction.
No palpi. On the vertex are a pair of short fleshy exarticulate filiform an-
tennjE, and there are no distinct eyes or ocelli. In the cast larva! integument
the entire head, 0.25 inch long, is exserted, and is dark colored and evidently
horny, all the parts retaining their shape except the antennze, labrum, and
labium. The whole head has here the appearance of the basal part of the
leaf of a grass plant, clasping the origin of the maxillae on its posterior half, and

84 THE EARLY STAGES OF TABANID^

bifurcating into the somewhat tapered cylindrical mandibles on its anterior half.
The maxillae are traceable to two-thirds of the distance from the tip to the base
of the head, scarcely tapering, bent obliquely downward at two-thirds of the
way to their tip, and obliquely truncate at tip. On the anterior margin of ven-
tral segments four to ten, in the living insect, is a row of six large fleshy roundish
tubercular retractile pseudopods, the outside ones projecting laterally, and each
at tip transversely striate and armed with stout bristly pubescence; on the ante-
rior half of ventral joint eleven is a very large, transversely oval, fleshy, whitish,
retractile proleg, with a deeply impressed, longitudinal stria. On the anterior
margin of dorsal joints four to ten is a pair of smaller, transversely elongate, re-
tractile, fleshy tubercles, covering nearly their entire width, armed like the
pseudopods, but not so much elevated as they are. No appearance of any
spiracles. Anus terminal, vertically slit, with a slender retractile thorn, 0.05
inch long, visible in 1860, but not in 1863. Head, and first segment or two,
retractile."

When handled, the larva is, according to Walsh, "very vigorous and
restless," and burrows with great strength between the fingers, and
even on a smooth table walks as fast as any ordinary caterpillar,
either backwards or forward; when placed in a vessel of water it
swims vigorously, twice the length of its body at every stroke, by
curving its tail around laterally, sometimes to the right, sometimes to
the left, so as to touch the side of the fourth or fifth joint, and then
suddenly lashing out with it. In such a vessel it always keeps close
to the surface, and at the end of every stroke, and also when in re-
pose, elevates the anal slit out of the water, on which occasion Walsh
once saw a bubble of air attached to it. In the breeding jar the
larva scarcely ever comes to the surface, but burrows among the
decayed wood, aquatic plants, etc.

The larva described by Walsh differs, according to him, remarkably
from the one described by Degeer {Tabanus bovinus L.), in having
ventral pseudopods as well as dorsal ones. Walsh says that it might
be supposed that the dorsal tubercles were branchial, "but for the fact
that they are found in the earth inhabiting species described by De-
geer," and that like the aquatic larvas of Pnonoxyphon discoideus
Say (Coleoptera) it has a branchial apparatus issuing from the anus,
and the short retractile anal thorn, observed in 1860, was the form
assumed by this structure when out of the water. This assumption
by Walsh is partly confirmed in so far as the terminal spine con-

WERNER MARCHAND 85

tains the openings of the tracheal trunks. On the other hand, this
spine has no connection with the anus, which is not terminal, as
Walsh assumes, but lies in the fleshy tubercle of the eleventh seg-
ment, where Walsh observed "a deeply impressed, longitudinal stria."

The larva described by Walsh and which undoubtedly was Tahanus
atratus, at least that specimen from which he obtained an imago, is
found according to him from the beginning of June to the beginning
of September, at which latter time he also met with a specimen only
half the length of the full grown specimen (possibly belonging to a
different species).

The pupa has also been described by Walsh, and the description is
given here without change:

''The pupa (from the pupal integument) is cylindrical, suddenly rounded at the
head, and tapering a little in the last two abdominal joints; the color is a very
pale, semitransparent, yellowish brov/n. The mouth is represented by six tu-
bercles, hexagonally arranged, above which, upon each side, is a trigonate three-
or four-jointed antenna, pointing outwards. The pronotum commences immedi-
ately behind the antennae and bears on its anterior dorsal submargin a pair of
reniform tubercular spiracles, the mesonotum, to which the wing cases are at-
tached, is twice as long as the pronotum, and bears on its anterior dorsal margin
a pair of obliquely placed reniform tubercular spiracles, three times as long as
the prothoracic ones. Then follows a very short metanotal piece, about one-
seventh as long as the pronotum, bearing no spiracle, which is succeeded by
eight subequal segments, all but the last bearing on their lateral dorsal surface a
subbasal round tubercular spiracle. The first of these eight segments is simple
and extends to the tip of the wing cases i^"* the others are all furnished two-thirds
of the way to their tips with an annulus of appressed bristles directed back-
wards. The anal thorn is very robust, having a diameter of one-half the last
abdominal segment, and is squarely truncate as soon as its length is half its width,
and terminates in six small robust thorns, arranged in a regular hexagon. Length
0.97 inch; greatest diameter 0.21 inch. One specimen."

The second author to describe the larva of Tahanus atratus was
C. V. Riley (1870). His description of the larva is quoted by Salmon
and Stiles. ^^

^■^ 'T beheve this first spiracle-bearing segment to be metanotal, as also the
corresponding piece in the pupa of Midas fulvipes, u. v.," (Walsh).
^^ Salmon and Stiles, Emergency report on surra, pp. 100 and 101.

86 THE EARLY STAGES OF TABANID^

"The larva [Plate 3, Fig. 46] is a large twelve-jointed cylindrical affair, taper-
ing at each end, of a transparent, highly polished, glassy, yellowish or greenish
appearance, shaded with bluish green and furnished above and below as in the
figure (figure given by Riley) with large roundish sponge-like tubercles which
are retracted or exserted at the will of the insect. Though the external integu-
ment is so transparent that the internal structure is readily visible, yet this
integument is firm and the larva most vigorous and active, burrowing with great
strength either backward or forward in the earth and between one's fingers when
it is being held. Placed in water it will swim vigorously by suddenly curhng
round and lashing out its tail, but it is apparently not as much at home in this
element as in the moist earth, for it is restless and remains near the surface with
the tip of its tail elevated into the air. When the water is foul it moves about
actively near the surface, but when it is fresh it remains more quietly at the
bottom."

This specimen, which Riley succeeded in breeding, was sent to
him by Mr. Adolph Engelmann, of Shiloh, St. Clair County, 111.
It was found by Mr. William Cooper, of the same county, about ten
feet from a small but permanent body of water. Mr. Cooper at first
took it to be a leech, but when he attempted to catch it it immediately
commenced burrowing into the ground.

The larva is declared by Riley to be semiaquatic, for it is at home
either in moist earth or water. This specimen was kept for over two
weeks in a large earthen jar of moist earth well supplied with earth-
worms. It manifested no desire to come to the surface. Riley
found several pale dead worms in this jar, though not able to say
positively whether they had been killed and sucked by this larva.

Hart (1895) reports that he has taken the larva of Tabanus atratus
in every month of the season except June, at which time they had
mostly reached the pupa or imago stage. They seemed to prefer
the sandy shores, and were taken abundantly May 17 at Hart's
Survey Station C, by running through a coarse sieve the surface
layers of sand of the shore near the wave-washed margin. Station
C is characterized by Hart as being located near the outlet of Quiver
Lake, the shores being here near together and sheltered. The east
bank was sandy, with a muddy coating over the part which is exposed
at low water, while the west shore was of black mud grown over with
willow trees and overflowed in moderately high water. The water
was clear and on both sides thickly filled in summer with algae and

WERNER MARCHAND 87

other aquatic plants, having little or no current in ordinary stages.
The same process of sifting was repeated June 25, and not a single
larva was found. Individuals placed in breeding cages failed to
transform. A pupa was collected June 30, from which the imago ap-
peared July 17. A cast pupal skin was also picked up July 18. Sev-
eral imagos were taken in the same locality between May 23 and
June 22.

During the winter good sized larvae sometimes occurred in dip-net
collections, and March 18 they were again found to be common at
Hart's Station C, in loose drifts, partly frozen, left by an early spring
rise. The previous year they were common in April far from the
margin, among sticks, logs, and other drift, marking the higher stage
reached by the water on March 19 of that year. These situations
remained moist for a long time, harboring a large variety of aquatic
forms, some of which completed their transformations successfully
while others apparently failed, the river remaining low and the
weather dry.

Hart's description of the mature larva (Plate 3, Fig. 41) is as
follows :

^' Larva, Mature. — Length 45-55 mm., diameter 6-7 mm. Transparent
whitish with a greenish tint, marked with conspicuous dark brownish or greenish
fuscous, paler in younger specimens."

"Lateral prothoracic striated areas less than half as long as the dorsal, stria-
tion microscopically fine and opaque or scarcely shining, a small smooth spot on
the anterior margin of the striated area, resting on the lower lateral line; remain-
ing upper lateral areas of thorax much more coarsely and sparsely striate and
shining; middle and lower thoracic areas — often much reduced, or even entirely
covered by the lateral stripes — with distinctly finer and closer striation, but
still shinmg; abdominal lateral areas with still finer striation, nearly as fine as
that of the prothorax and feebly shining; dorsal and ventral areas all smooth
and shining, rarely a few broken striae about their margins, at the base of the
prothorax or on the anal segment."

"Dark annuli distinct, broad, including false feet, transverse pale spot imme-
diately in front of dorsal tubercles narrow or closed up in the mature larva; on
the abdomen above, each annulus usually extends back on the median line in a
triangular prolongation, often nearly attaining the next annulus, less developed
in younger larvae. Prothoracic lateral space occupied in front of the striated area
by a dark opaque quadrate spot, extending from the anterior annulus. Lateral
stripes of meso- and metathorax broad, at least the upper ones widened poste-

88 THE EARLY STAGES OF TABANIDiE

riorly, the lateral edges of the dorsal areas therefore parallel behind the middle
of the segment, as seen from above; lateral stripes of abdomen, especially the
intermediate ones, more or less abbreviated and broken up posteriorly except on
the segment next the last. In these stripes the punctures of the upper and
lower rows are indicated by rounded pale dots, and those of the inner rows by
elongate dots. Last segment with broad dark annuli about base of respiratory
tube and around anal prominence, with lateral connections; also more or less
invaded above by the basal annulus, often leaving there only a pair of pale spots
posteriorly. Often a dark spot in the anterior angles of the ventral space on the
seventh abdominal, and one behind the anal dark ring."

"False feet moderately elevated, with coarse whitish pubescence more or less
tipped with fuscous or with brownish in younger larvae, dorsal pair narrowly con-
nected over median line. Main internal tracheje usually subparallel, sinuated,
not very conspicuous, although easily traceable. Stigmatal spine rarely visible."

A few larvag have been obtained by Hart, which are Kke atratus
except in one particular — the surface of the body, especially of the
anterior abdominal segments, shows a fine undulate wrinkHng re-
sembling the sculpture of the pupa, but smoother. As the specimens
showing this appearance were shrunken and in bad condition, it is
surmised that it is an effect of letting the alcohol get too weak and
then changing to strong alcohol.

Hine has taken the larvae while digging in the ground in the vicinity
of ponds, from under stones on ditch banks, from the water with
dip-nets, and occasionally in most unexpected places. He says,
however :

''If one is looking for them he is likely to meet with more or less disappoint-
ment, as the finding of one specimen does not indicate necessarily that others
may be taken under the same conditions. The fact that specimens have been
taken from floating logs and debris suggests that they may be transported for
longer or shorter distances in this way, and during high water be stranded upon
ground which, when the flood subsides, is high and dry and far removed from the
bed of the stream. Since the species in all its habits is closely associated with
water and wet ground, this seems to be the only way of explaining the appear-
ance of larvae in dry soil and in places remote from where the eggs are laid."

nine's description of the larva follows, being in many ways more
brief and accurate than that of Walsh given previously.

"Full grown larva nearly 2 inches in length. General color yellowish white,
with wide dark brown bands at the union of each two segments. Prothoracic
segment on each side with two lateral grooves, which do not quite reach the

WERNER MARCHAND 89

posterior border of the segment, and a dorsal (?) groove continued for the entire
length. These grooves and a number of irregular dots on the posterior part are
dark colored, while the remainder of the segment is light. Mesothoracic segment,
on each side, with four longitudinal grooves, which reach nearly the entire length.
The dark markings on this segment include a narrow anterior border, the lateral
grooves, and a number of irregular dots near the posterior margin. The meta-
thoracic segment is like the last except that the dark color on the anterior mar-
gin is wider and the posterior, instead of being dotted, is uniformly brown. The
abdominal segments are each similar to the metathoracic, but the dark mark-
ings in the region of the lateral grooves are more or less abbreviated. Last ab-
dominal segment with two pairs of dark markings; the ventral pair extend the
whole length of the segment and are connected just behind the anal prominence
by a cross-band; the dorsal pair are oblong, somewhat irregular in outline, and
extend from the anterior margin to beyond the middle of the length. At the an-
terior ventral border of each of the first seven abdominal segments is a trans-
verse series of prolegs, three on either side of the midventral line. These pro-
legs are located within the dark transverse bands, but are lighter in color than
these and prominent enough to be seen easily. Above the prolegs on either side
of the middorsal line is a small swelling which appears as a rudimentary proleg;
before the two is a distinct transverse light spot still within the dark area."

"The head of the larva is very small for so large an insect and the mouth
parts are minute. The mandibles consist of two strongly chitinized pieces, and
work by being pushed endwise backward and forward. When drawn in, the
anterior ends point directly forward, but when protruded, these same ends point
downward and backward, thus forming a pair of hooks by means of which the
prey is held. The larva is able to protrude its mandibles very quickly and to use
them very effectively on soft-bodied invertebrates on which it is known to feed."

W. A. Riley and Johannsen (1915) have also figured the full grown
larva of Tahanus atratus (Plate 3, Fig. 45), and Malloch (1917) fig-
ures the head and first segment of the larva (Plate 5, Fig. 77).

On the pupal stage we have information, in addition to Walsh's
description, by C. V. Riley, Hart, and Hine.

Riley reports that his larva transformed to pupa within the ground
during the early part of July ; it remained in this state but a few days
and the fly issued July 13, and soon made its presence known by its
loud buzzing inside the jar. It was a perfect specimen, and the
pupal integument was sufficiently firm and polished so that by care-
fully washing off the earth an excellent cabinet specimen was ob-
tained, which retained almost the exact form and appearance of a
living pupa. Before the escape of the fly, which was effected through

90 THE EARLY STAGES OF TABANIDvE

a longitudinal fissure on the back of the head and thorax, reminding
one of the mode of escape of the harvest flies {Cicada), this pupa by-
means of the horns with which it is furnished had pushed itself up
to the surface of the earth.

The pupa itself (Plate 12, Fig. 140) is, according to Riley, nearly
an inch and a quarter in length and a third of an inch in diameter.
It is cylindric, slightly curved, as in the figure, rounded at the head
and tapering at the extreme hind portion. The abdominal seg-
ments are, all but the first one, provided with a ring of fine yellow
bristles, pointed backwards. There is a stout thorn at the anal ex-
tremity, bearing six other thorns.

The pupa state lasts but a few days and before the emergence of
the fly the pupa is pushed to the surface of the ground by means of the
bristles and thorns of the abdomen, with bending movements of the
body.

It splits along the dorsal line and the fly emerges, leaving the pupa
case in perfect condition.

Hart's description of the pupa (Plate 13, Fig. 163) contains some
additional details.

"Pupa, Male. — Length 30-v35 mm., diameter 7.5 mm. Yellowish fuscous
with a brownish tint, thorax not paler. Palpal sheaths distinct, short, very
narrowly separated by a depressed space. Abdomen roughly wrinkled and sub-
opaque. Spiny fringes tipped and annulated with black. Otherwise as in the
pupa (female) of T. stygius'^ (described by Hart in the same paper).

Hine (1906) gives another description of the pupa:

"Pupa [Plate 11, Fig. 124; Plate 13, Fig. 164] about U inches in length.
Color brownish yellow. Antennal and other tubercles of the head darker than
the surrounding parts. Prothoracic spiracle slightly elevated, clear brown in
color, reniform and oblique, rima gradually curved to near the dorsal end, where
a distinct hook is formed by a sharp bend. Abdominal spiracles nearly round;
rima of the first short and gradually curved and with a slight hook at the dorsal
end. Terminal teeth [Plate 13, Fig. 164] arranged in pairs, a ventral pair and a
pair on each side formed by a dorsal and a lateral tooth. The distances between
these teeth are variable; the two dorsal are nearest together, then follows the
distance between a dorsal and a lateral, the distance between the two ventral,
while the distance between a ventral and a lateral on each side is greatest of
any."

WERNER MARCHAND 91

This appears to be all that is known about the early stages of
Tabanus atratus.

Tahanus atrimanus Loew. — An African species, common near Mt.
Mlanje in southern Nyasaland, having a preference for the neighbor-
hood of wooded streams, and being there most abundant in Novem-
ber and December.

Neave collected and bred the larvae. On October 25 a number of
larvae were taken in the Ruo River among the roots of grasses in
running water, but occasional individuals were also found in the
mud in other wooded streams. Imagos, bred from these, began to
emerge on November 25. At the same spot in the Ruo River some
other larvae were found which may be those of Tahanus pertinens,
but they were not bred to the adult stage.

The larvae, which are figured (Plate 3, Fig. 52) are strikingly dis-
tinct from those of the apparently closely allied Tabanus variabilis
Lw. They are of a somewhat opaque yellowish color, with rather
faint brown pigmented areas. The pseudopodia are well devel-
oped, as is the case with other species found in running water, and
there are well marked hairs on the syphon. Though invisible in
preserved specimens, two pseudopodia of considerable length are
present immediately anterior to the anus.

The pupa is a clear orange-yellow color and the aster is remark-
able for the erectness and large size of the dorsal pair of hooks,
especially in the female. The dorsolateral comb consists of a few
widely spread spines.

The larva (Plate 3, Fig. 52), the pupal aster of both sexes, and the
dorsolateral comb of the female are figured (Plate 14, Fig. 173, a, b, c).

Tabanus autumnalis Linne. — Scholtz (1850) reports finding the
pupae of this species, which is common in Europe, with those of
Tabanus tropicus and Hcematopota pluvialis (see also these species),
on an excursion in the neighborhood of Breslau in June, 1850, at the
edge of a pond covered with Lemna, the water of which was com-
pletely polluted from manure piles surrounding it. The pupae, of
which the flies emerged after a few days, were found quite near the
edge under a thick moist mass of decaying Lemna, together with
Stratiomys and Syrphus pupae.

92 THE EARLY STAGES OF TABANID^

The larvae were first found in 1851, by Brauer and Goszy, on the
shores and shallow places of the "Wienfluss" (Danube ?); also the
pupae were found. Kollar, 1854, gives only these few remarks on
the subject.

Raillet says (quoted from Brumpt) : The larva of this species lives
in the water; it breathes the air at the surface of the water, like the
larvae of mosquitoes; they may be destroyed likewise by spreading
kerosene oil on the swamps where they occur.

It is possible that the larvae in which Graber discovered the organ
named after him were of this species (Paoli).

Surcouf and Ricardo (1909) report on their observation of an egg
mass of Tahanus autmnnalis L. at Lamballe (C6tes-du-Nord) in
August, 1907. The eggs formed a dense mass attached to a reed,
and the female, turning the head towards the moist ground, re-
mained immobile, making no effort to escape when it was captured.
Unfortunately, the observers waited until the following morning to
take the plant with its roots, and in the meantime the grass had
been cut.

Graber (1882) gives the illustration of a newly hatched larva of
Tabanus, which is supposed to belong to Tahanus autumnalis (Plate
8, Fig. 99). In this larva Graber studied the chordotonal organs,
tactile bristles, and other microscopical details (Plate 8, Figs. 99 to
102) . These details are discussed in the chapter on Tabanidae, descrip-
tion of early stages (see also Tahanus sp. Nos. 10 and ^<?,page 179).

Tahanus autumnalis (Plate 8, Figs. 99 to 102; Plate 9, Figs. 103 to
108) is the species in which the organ of Graber was probably dis-
covered for the first time. Krauss (1879) informs us that Brauer, in
the spring of 1875, demonstrated in his entomological class the larvae
of Tahanus autmnnalis, which had been bred from eggs, and called
attention to a peculiar and undescribed organ in the abdomen.
Drawings of the organ were made by Krauss but not published;
according to the latter, they agree so completely with those pub-
Kshed by Graber that it is likely that Graber' s " Fliegenmade" was
in fact the larva of Tahanus autumnalis.

Krauss adds the statement that the larvae of this species, in contrast
to tabanids — the larvae of which live in damp localities, manure,
decaying plant material, or in the ground — spend all their develop-

WERNER MARCHAND 93

ment in the water, leaving it only for pupation, which takes place
at the edge of ponds, in the mud, among plant detritus, etc.

Krauss expresses the opinion that the organ will doubtless also be
found in the imago, and in that case the mystery surrounding it may
be cleared up.

Von Friedenfels (1880) reports on a larva which he found in the salt
lakes of Siebenbiirgen. This larva was aquatic and transparent and
was first thought to be an annelid. When sent to Brauer in Vienna,
it was identified by the latter as a larva of Tahanus autumnalis.

Tabanus bicallosus Ricardo. — A smaller species of Tahanus, recorded
from Pusa, Bengal, and Madras, by Patton and Cragg, to whom we
also owe some notes on its early stages.

As in all the small species observed by these authors, the eggs are
laid on blades of grass just at the edge of a shallow stream, or on the
leaves of the lotus plant at the edges of small ponds, but never over
deep water.

The eggs, as in all the smaller species in Madras, are torpedo-shaped,
the eggs of this species measuring 1.1 mm. in length and 0.2 mm. in
breadth. Figures of the eggs are given (Plate 1, Figs. 12 and 13).

The egg mass figured by Patton and Cragg seems to consist of two
layers spread out one over the other, and the whole "moulded into
the hollow of the blade right up to the tip" (as in other small Indian
tabanids) .

The mature larva of Tabanus bicallosus is figured by the authors
(Plate 4, Fig. 62), but no description is given. On the drawing the
eleventh and twelfth segments appear to be completely fused (?),
dorsal pseudopods are well developed, which should indicate an
aquatic mode of life. There seems to be a color pattern similar to
that of the larva of Tahanus atratus.

The pupa is likewise figured (Plate 11, Fig. 135), but not described.
In addition, figures are given of the abdominal tip of male and fe-
male pupae (Plate 12, Figs. 144 and 145), showing the arrangement of
the six terminal teeth. In the male pupa attention is called to the
ribbed anal tubercle and the continuous fringe of spines in front of
it, as differing from the condition in the female, which has a simple
anal tubercle and an interrupted fringe of spines in front of it.

94 THE EARLY STAGES OF TABANID^

This is the first and only attempt to separate the sexes of tabanids
in the pupal stage.

Tahanus higiittatus Wiedemann. — The early stp^^es of this species,
which is widely spread throughout Africa, occurring from the Cape
to the Egyptian Sudan, have been described by King (1908).

The eggs, which were obtained in the vicinity of Taufikia, Sudan,
in the marshes, are deposited in a rounded mass (Plate 1, Fig. 18)
on grass and reeds overhanging a pool. One egg mass that was
counted contained about 450 eggs. The entire act of oviposition was
not timed, but it lasted well over half an hour. When a female is
ovipositing, although usually exceedingly shy, the stem on which she
is resting may be plucked and carried away or put into a bottle with-
out disturbing her. Having deposited the eggs, she covers the mass
with a creamy white secretion which turns black after a short time.

The egg is spindle-shaped, slightly more pointed at one end than
at the other, and white in color. Length 2.5 mm. The eggs under
observation hatched in about eight days, but possibly under natural
conditions, exposed to the sun, the incubation period would be
shorter. On hatching, the larvae fell into the water, swam to the
sides, and buried themselves in the mud.

The larvas can swim only on the surface of the water, and progress
either by a telescopic movement or by lashing vigorously from side
to side.

Several methods of rearing them were tried. The majority were
placed in a large glass vessel containing mud, living grass, and water.
Some were put into jars containing only water, others in dishes con-
taining moist sand, others again in vessels containing sand and
water so arranged that there was a pool at one end of the vessel and
moist sand at the other.

The larvas in the vessel containing mud, grass, and water did well,
but many were devoured by predacious insects — e.g., dragon-fly
larvae — introduced by accident in the mud and water, and others
perished owing to the grass dying and fouling the water during
transit. Eventually sand and water in Petri dishes were found to be
best, as it could be kept clean and the larvae easily located when
wanted.

WERNER JIARCHAND 95

At first they were fed on tiny crustaceans dredged from rain pools,
but during transit, when these could not be obtained, scraps of
freshly killed raw meat and congealed blood from the bodies of
gorged mosquitoes were substituted. After arriving in Khartoum
their diet consisted of earthworms, ■ as a plentiful supply of these
could always be procured.

They grew very slowly and at greatly varying rates. Two larvse
hatched from one egg batch on June 11 measured respectively, five
weeks later, 4 mm. and 15 mm. Owing to their telescopic nature
it was exceedingly difficult to measure them accurately, so the figure
must be taken as merely approximate. They did not appear to be
cannibalistic in their habits, as several of various sizes were reared
in the same dish and sometimes kept short of food, but were never
seen to attack each other. When one died, however, its comrades
usually devoured it. When not feeding they spent most of their
time buried in the sand, with just the tips of their respiratory syphons
showing. If the sand was allowed to dry they became very restless,
and would make continual efforts to escape from their jars until
water was given them again.

Early in August, when they were about eight weeks old, they ceased
feeding and were then transferred to jars containing sand to a depth
of 6 cm. They descended to the bottom of these jars and were still
there when, about six Vv^eeks later, the author went to England on
leave.

On January 28 King returned to Khartoum, and it was found that
the jars then contained several dead adults — all males — a few dead
pupae and larvae, and a single live larva. This last died early in
February without having reached the pupal stage. The empty pupal
cases were all sticking up out of the sand, the pupae having evidently
worked their way up from the bottom of the jars by means of their
abdominal spines. In several instances the old larval skin had
remained attached to the caudal teeth of the pupal case.

Neave (1915) also found the larvae of this species in the valley of the
lower Shire and its tributary, the Mwanza, in southern Nyasaland.
Here, in the sand and mud on the banks of the rivers, large num-
bers of larvae were obtained, of which the majority belonged to this
species, and these were chiefly found in mud among the Phragmites

96 THE EARLY STAGES OF TABANID^

reeds on the banks and in the back-waters, etc., of the Shire River.
They often occurred, especially if the mud was inclined to be dry,
at a depth of as much as 6 or 8 inches. The locality figured by
Neave represents a large fiat sand bank partly covered with Phrag-
mites, at the border of a river about 50 feet in width.

The lower part of the Mwanza River was dried up at the time of
Neave's visit, except for a few isolated pools, on the banks of v/hich
many larvae of Tabanus biguttatus were found in the mud. Many
others were subsequently found in the Ruo valley in November.

The larva is, according to Neave, of a type which has many repre-
sentatives, being of a clear white color, except in individuals about to
pupate, with well defined pigmented bands and spots between the
segments and on the anal segment (Plate 5, Fig. 69). Several species
have larvae of this type, only varying in the amount of pigment and
in the distribution of it on the anal segment.

King's description of larva and pupa follows:

The larva (Plate 5, Fig. 75, o, b), when nearly hatched, is white in color, but later
assumes a grayish to yellowish tinge.

Mandibles (Plate 5, Fig. 78) black, slightly serrated and with two tufts of
curved hairs at the base.

First thoracic segment anteriorly brown. Laterally placed on the second and
third thoracic segments are brown comb-like marks with the four teeth pointing
backwards.

On the anterior portion of each abdominal segment — with the exception of the
eighth — are two brown annuli, or rings, encircling the body. The former of these
two rings is usually covered by the posterior margin of the preceding segment.
The hinder ring bears a double line of fine black hairs and also a number of small
fleshy projections or pseudopods. The eighth abdominal segment serves the
purpose of a respiratory syphon. Its posterior margin is brown and from its ex-
tremity can be extruded a small process terminating in stigmata. A brown
curved longitudinal mark is situated on either side of the eighth segment. The
anus is placed at the base of this segment. Length 35 mm.

The pupal case is chestnut-brown in color, with the thoracic tubercles darker.
Each of these tubercles bears a spine. The abdominal segments are apically ringed
with backward projecting spines. The anal segment terminates in a cluster of
six teeth (Plate 12, Fig. 148), the dorsal pair larger than the lateral and ventral
pairs. Length 20 mm.

The pupal aster has been figured by Neave (Plate 14, Fig. 175,
a-d); it is of the regular type and the dorsolateral combs have

WERNER MARCHAND 97

long spines, which are, however, shorter and stouter (especially in
the female) than those of Tabanus cor ax.

Tahaniis hovinus Linne. — This is one of the commonest species of
the genus in Europe, the largest species in Sweden. It has a wide
distribution and occurs even in South Africa.

The larvae were found by Degeer (1760) in the soil of a meadow.
Placed in earth which was renewed from time to time some of them
gave pupae. I give Degeer's original report in translation from the
German.

The larvae of these flies, which have not been recorded by any one before me,
live in the ground, and I have described them already in the Swedish Proceedings.^®
When in the month of May, I turned over the earth of a meadow, I found many of
them, and having placed seven or eight in a jar with fresh earth which was renewed
from time to time, I was aware, on June 12, 1760, that one of them had turned
to a pupa, and had in part crawled out from the earth, being still fixed in it by the
abdomen. I dug for the remaining larvae, but found only three of them, which
pupated later, and in the same way protruded with half of their body from the
earth. At last I found one more small dead larva. I presume that those which
pupated have eaten up the others.

The largest of these larvae (Plate 4, Fig. 59, a, c) was about W inches long,
and in the middle 3\ lines in diameter (as seen in the figure), very similar
to the larvae of the large crane-flies. The body was cylindric, of almost equal
diameter all along its course, narrowed and pointed at the head, twelve-seg-
mented, the last segment small and wart-like.

Color whitish gray and yellowish; a number of black transverse bands formed
by the transverse ridges and some transverse stripes in the incisions between the
segments. The small head is shining brown. Under the lens the skin appears
also shining, and covered with very fine longitudinal ridges.

The head is elongate, horny, bearing several parts difficult to distinguish as they
are constantly in motion, also with two small short antennae, some pointed parts
placed below, and two large horny black mandibles placed above the latter, as
long as the head and curved downwards. If a larva was held between the fingers,
it exserted the hooks, attached itself to the skin with them, and pulled the con-
tracted body up. Probably it burrows its way in the earth by means of these
hooks. While resting, the head is withdrawn into the first segment, and this in
turn into the second segment, so that the anterior part becomes of the same
diameter as the rest of the body. The two last segments can also be retracted
into the third last one, in a similar way.

^® Bromsarne's Ursprung.

98 THE EARLY STAGES OF TABANID^

The last segment (Plate 4, Fig. 59, c) has the aspect of a small soft textured
conical wart; at its end there is a small, elongate protruded horn-Hke brownish
yellow piece placed vertically, which I assume to be a spiracle, as a longitudinal
fissure is discernible. Under the segment before the last, near the segment
preceding it, there is a fleshy prominence, also with a longitudinal fissure in the
middle, the anus.

On the body seven prominent blackish ridges are found situated anteriorly
on the fourth, fifth, and the following segments, to the tenth, including the
latter. At the sides and below fleshy warts take the place of feet. In con-
tracting the rings the interior parts are drawn in.

The nymph (Plate 4, Fig. 59, b, d), 1 inch in length, of the same thickness as
the larva, cylindric excepting the last segment, which is much smaller. It trans-
formed into a large horse-fly while I had expected that a large crane-fly would
appear. The color was brownish gray; darker at the abdomen and the other
parts. The abdomen was eight-segmented; at the posterior margin of each seg-
ment was a fringe of long gray hairs. On the last segment at the tip there are six
hard horn-like points (Plate 4, Fig. 59, d), by means of which the nymph works
its way out of the earth. Head, thorax, and wing cases, or the. whole thoracic
region together shorter than the rest. Anteriorly, on the head, there are two
small brown tubercles, each of them bearing a hair, probably spiracles. At the
side of each tubercle an elongate and equally brownish point, appressed flatly
to the head, directed backwards and jointed in the middle, probably the anten-
nal sheaths. If the nymph is touched, it carries out worm-like movements with
the abdomen.

Early in July the fly hatched out. The pupal skin splits open dorsally along-
side the thoracic shield, and at the sides of the head. Internally, there are in
the cuticle of the head two long horn-like spines, the function of which is unknown
to me.

These observations of Degeer are remarkable for their precision. A
number of details which have been incorrectly represented in tabanid
larvae by several later authors, as the number of segments, position
of the anus, etc., have been correctly described. Nothing essential
is omitted. Degeer, in 1760, calls attention, like del Guercio in
1913, to the similarity between certain tabanid larvas and those of
crane-flies, but while the latter is satisfied to state that there is no
appreciable difference between them, Degeer's description of the
Tabanus larva remains fully characteristic of the group and prob-
ably also of the species, though before the larvas hatched, he had ex-
pected them to become tipulids.

WERNER MARCHAND 99

Westwood (1840) has briefly translated Degeer's description, and
I give his summary to complement the above rather imperfect trans-
lation :

The larva of Tabaiius bovinus Degeer (L) is found in the earth, and is of an
elongated subcylindric form, attentuated at each end, especially in front; it is
destitute of feet, twelve-jointed, having the head distinct, narrow, elongate,
horny, armed with two strong curved hooks, antennae, and palpi; the fourth to the
tenth segments having an elevated dorsal papillose ridge used in progression;
the terminal segment is minute and tuberculiforni; the pupa is naked, incomplete,
elongated, subcylindric, with six spines at the end of the body; the margins of
the abdominal segments cihated, and the forehead bi-tubercled."

Westwood also reproduced Degeer's figures, though very imper-
fectly; also Macquart, 1834.

No further observations have been published on the early stages of
this species, since Degeer.

Tabanus bromius Linne. — ^The larvas of Tahanus hromius live, ac-
cording to Beling, by preference in the grass-covered soil of meadows,
fields, and similar places. They are often brought to the surface by
moles, and in the spring and summer, 1874, Beling found them, as
also later on the pupae, especially frequently on meadows in fresh
mole hills. The pupal stage lasts usually between two and three
weeks, and the imagos begin to appear in the second half of the
month of June. The larvte feed by sucking the contents of earth-
worms, larvas and pupse of other insects, and, if there is lack of other
food, of their own kind, but they seem to be able to subsist in case
of necessity on earth alone, in which, as Beling says, he kept them
for months until pupation,

I assume that the larvce, if they subsist on earth, will in fact sub-
sist on its organic contents in a similar way, as is the case with earth-
worms. Moreover, small earthworms and insect larvae are easily
overlooked by the observer if the earth used is not specially sterilized.

Larva. — Length up to 16 mm., width 4 mm., color pale yellow, strong, silky,
shining, with distinct dense and fine longitudinal striation, with twelve body
segments, excepting the small narrow brown head, which can be retracted into the
first segment. The latter in the middle of dorsal surface with a wall like longi-
tudinal prominence, which is covered in its anterior part with short hairs, and on

100 THE EARLY STAGES OF TABANIDiE

each side is provided with a rounded densely brush-like denticulate tubercle.
Antennte not discernible, palpi very short, one-jointed, thick. First body seg-
ment short, obtusely cone-shaped, retractile, and merging into the second seg-
ment without distinct border. Last segment short and much narrower than all
the preceding, cupola-shaped, bearing on the anterior part of the under side the
more or less prominent half circular anus crossed by a deep longitudinal furrow or
fissure. The tip of the last segment a roundish, wart-shapcd, retractile promi-
nence, bearing the vertical stigiratal fissure. Scgm.ents five to ten inclusive with
marginal swelHngs anteriorly. In this ridge or ring eight protuberances are dis-
tinctly observable, of which two smaller ones are located on the dorsal side,
one on each side, and two pairs, placed near to one another, forming the loco-
motory swellings of the under side. The surface of the retracted head capsule
visible through the cuticle of the first segments appears as an extensive brownish
yellowish area limited by two broad blackish stripes which, diverging from the
middle, are pointed and drawn out behind.

The pupa is likewise described by Beling, a translation of which
follows.

Pupa. — Pupa 16 to 18 mm. in length, 3.6 to 4 mm. in width, obtuse at the
anterior end, slightly narrowed behind, of dirty greenish yellow color, smooth,
somewhat shining, and if older, with blackish borders of the segments. Above
the front five to six small flat brownish tubercles or warts, of which the two
lowest are enlarged and tooth-hke, the two upper ones flatly truncated, bear-
ing a short blackish brown hair in the middle of the truncation. On each
side of these frontal tubercles is a short tooth-like outwardly directed spine.
Above the frontal tubercles are two small wart-like tubercles each bearing a
short black bristle. Lower ventral side of thorax on each side with two
small brovv'n warts bearing a backwardly directed closely appressed hair,
and with an elongate black spot, these together forming two rows diverging
posteriorly. Abdomen nine-segmented, first segment very short, third to eighth
segments inclusive on each side with two parallel longitudinal depressions rather
distant from one another, together forming two depressed Hnes running laterally
alongside the abdomen. In the middle of these depressions on each of the six seg-
ments mentioned and in its anterior part a short spine-shaped prominence and a
similar prominence also on the second segment immediately behind the wing
cases. Third to eighth abdominal scgm.ents near its posterior margin with a
circle of densely placed yellowish to brownish or even blackish hairs of uneven
length and more or less completely appressed backwards to the surface. Last
abdominal segment with six short thick outwardly diverging spines with dark
brown tips, of which four are placed in a horizontal row; namely, a strong one
on each side and two weaker ones in the middle. Wing and leg cases not very
strongly marked, of the same color as the rest of the body, reaching to the anterior
border of the third abdominal segm^ent.

WERNER MARCH AND 101

The pupa of Tahanus hromius L. has been figured also by Surcouf
and Ricardo (1909); the previous observations by Beling have not
been observed by these authors.

This pupa (Plate 12, Fig. 151) was collected in the earth of a rail-
road embankment at Longny, Orne, France, by M. E. Cordier, from
whom the authors received it. The adult died in the act of hatching;
it proved to be a male of Tahanus hromius L. ; having left the pupal
shell half way, it had been unable to free itself in spite of numerous
efforts as proved by the pupal shell being unusually drawn out in
length. The manner of the hatching of the adult is evident from
the examination of this rare piece; the head causes the top of the
shell to burst and, consequently, to break open longitudinally, under
the pressure of the insect, down to the base of the first abdominal
segment, which remains intact like the following segments. When
the adult has reached this stage of its development, it takes flight,
and goes in search of food as soon as it is dry,

Tahanus carolinensis Macquart. — ^Of this North American species,
Malloch has given us, in his synoptic table, the following data on the
pupal stage:

'Pupa. — Dorsal abdominal segments, except first, armed with an irregular
transverse series, or two such series, of very stout thorns, their bases very much
dilated, slightly caudad of which series there are sometimes a few widely separated,
much longer spines. Seventh dorsal abdominal segment with 10 or 12 moderately
long thorns in a continuous transverse series, slightly cephalad of which is a
transverse series of very stout thorns longitudinally in line with the spaces be-
tween the thorns of the posterior scries."

No illustrations are given.

Tahanus corax Loew. — A large species with wings uniformly dusky
to the edge of the apex. Habitat, Africa. According to Neave,
common on the southern side of Mt. Mlanje, southern Nyasaland, in
the more wooded areas within the belt of heavy rainfall. On the
wing from the end of November to the beginning of January,

We are indebted to Neave (1915) for a knowledge of its life history.

The flies were kept by Neave in a comparatively large cage made of
mosquito netting and wood, each partition measuring about 5 by 4
by 3 feet, in which were boxes containing grasses and growing plants.

102 THE EARLY STAGES OF TABANID.E

The mosquito netting was loosely attached to the cage, so that the
shock to a fly in striking against it was minimized.

One or two captured females of Tahanus corax oviposited in these
cages, the process in one instance taking nearly an hour, between
3 and 4 p.m. Many egg masses of this species were also obtained in
the bush, always on reeds or grasses overhanging mud. While the
female is ovipositing she is not easily disturbed, as in the case of Ta-
hanus higultatus (King), and on one occasion one of Neave's collec-
tors brought him the reed, fly and all, from more than a mile distant
without disturbing the female. The egg mass with its cement cov-
ering is pure white when first laid, becoming dark gray as it hardens.
The cement which covers these egg masses must be, according to
Neave, water-proof and insoluble, as some individuals from them
succeeded in hatching even though the egg mass had been kept in
70 per cent alcohol for two days.

The egg masses of Tahanus corax are of the usual tabanid type, all
the individual spindle-shaped eggs being laid with their long axes
in the same direction. In the cases observed hatching took place
about the fifth day, but this would be likely to be lengthened or
shortened in some cases, according to the temperature. The process
of hatching takes place very suddenly. The egg mass splits in the
midline, following the long axis, and the small larvae emerge almost
simultaneously, forming a large quasi-viscous drop which falls bodily
from the reeds, etc., into the water or mud below.

A number of newly hatched larvae were obtained from collected
egg masses, generally found on reeds overhanging swampy ground.
The young larvse grow very slowly at first. Neave figures the
syphon and anal segment of one of these young larvae. The figure
(Plate 10, Fig. 120) shows the pecuHar Graber's organ, somewhat
tongue-shaped or triangular, and, according to Neave, attached by
fine strands of muscle from each of the three corners, apparently to
the body wall. It lies above the gut immediately below the dorsal
integument and seems to be capable of motion independent of the
general body movements. The organ contains a number of pairs of
small black pyriform bodies.

Neave obtained the larvae of Tahanus corax in considerable num-
bers. Adult specmiens are large, from 40 to 45 mm. in length, and

WERNER MARCHAND 103

distinct, having a thick rough integument of a dull reddish color,
with a tendency to two more definite patches of darker red on the
dorsal portions of the last two segments. The general coloration
appears to be largely due to the presence of foreign bodies in the
rough skin. The s3q)hon is very short, as seen in the figure (Plate

5, Fig. 67).

These larvae were most ferocious cannibals in all stages and very
troublesome in the laboratory, as they seemed to have unlimited
power of wandering about, even over dry surfaces. They fre-
quently succeeded in reaching the receptacles in which other species
were kept, and in destroying the larvae in them.

In the pupal aster (Plate 15, Fig. 181, a, b, c) the dorsal hooks of the
male are somewhat larger than those of the female. The dorso-
lateral comb is large and composed of very long and fine spines. The
pupal asters of male and female, and the dorsolateral comb of the
female are figured.

Tabanus cordiger Meigen. — A species described by Meigen (1820),
and widely spread in Europe, chiefly in the South.

Brauer (1883) is the first to figure the larva of this species (Plate

6, Fig. 88, a-d), but gives no description.

Picard and le Blanc report that they found, on March 4, 1913, in the
trunk of a poplar at the edge of the Mousson River, near Montpellier,
France, an elongate, whitish larva, pointed at both ends and having
a ridge, or prominent ring, on each segment, a larva which appeared
to them to be that of a tabanid. The larva was placed in a jar con-
taining pieces of decaying wood from the poplar tree in which it
had been found, and was left without any other food.

The exact time of pupation was not observed, but in the meantime
it seemed justifiable to assume that the larva was satisfied with the
vegetable food which it obtained from the decaying wood, since one
month after its capture it had not yet transformed. On June 10, a
male Tabanus was observed in the room, which had just hatched, and
was determined by Dr. Villeneuve as Tabanus cordiger Meig.

The larva lived, at the time of its capture, in the stump of a poplar.
The wood of the stump was not yet completely decayed, but rather
soft and very moist. A careful search did not reveal any other

104 THE EARLY STAGES OF TABANID.^

larvae of tabanids or other insects, while the felled trunk enclosed
numerous larvae of Tipulidae, Asilidae (Eryx), and some larvas of
beetles (Lamellicornia).

Picard and le Blanc's observations are insufficient for an exact mor-
phological description of the larva, but a brief description of the
pupa is given, based on the pupal shell left after the hatching. A
translation of this description follows:

Pupa. — (Plate 12, Fig. 136.) Slightly curved inwards ventrally in its posterior
part; measures 20 mm. in length, 4 mm. in width, and 3 mm. in height. The
anterior part (Plate 13, Fig. 155), extending to the scutellar segment, and com-
prising the head, thorax, legs, and wings, is smooth and unarmed, except for some
dorsal hairs. The posterior part of the pupa (Plate 13, Fig. 158) is composed of
seven segments, surrounded at about the second third from their base with a
crown of stiff and unequal bristles. The tubercles described by Surcouf in an
unidentified pupa were not observed. The last segment is terminated by two
three-toothed tubercles, at the base of which is found ventrally a cup-shaped
formation which possibly corresponds to the anus of the larva. The preabdom-
inal segment and the first six abdominal segments bear on each side a spiracle
placed towards the base of the segment, at the height of the anterior third. The
crown of bristles of the seventh segment is incomplete and interrupted dorsally.

The emergence of the adult causes at the anterior part of the pupal shell a fis-
sure which separates it into three wings, two dorsolateral ones and a ventral
one of irregular shape. The latter, in the shape of an elongate hexagon of 4 mm.
length by 3 mm. width, shows a characteristic ornamentation which can be
traced on the figure given.

At the anterior part two tubercles are found, each bearing a bristle; somewhat
below two slight ridges which meet in the median line, further down two doubled
tubercles, placed between two triangular thickenings, finally, more posteriorly, two
pairs of small tubercles between which two depressions with inner convexities
are found. A little in front of the point of attachment of the piece, two lateral
triangular expansions are observed. It is probable that the first two tubercles
correspond to the antennae of the larva and the others to the mouth-parts.

These observations show, according to Picard, that certain species
of tabanids may have wood-inhabiting larv«, and that their habits
are more variable than one would suppose. Picard' s paper is accom-
panied by three figures.

The observations made by Paoli on Graber's organ (page 37) have
possibly been made on this species.

WERNER MARCHAND 105

Tabanus costalis Wiedemann. — Habitat, eastern North America.
The species is very common, and a pest to stock. Of its early stages
v/e know Kttle. There are no data available on its oviposition, and
although Hine (1903), while speaking of the control of tabanids in
Louisiana, mentions that oviposition in tabanids often takes place
not over water but over damp ground, and that "one finds the eggs
of costalis and a number of other species in such places quite fre-
quently," he gives no description of the eggs.

The larva seems, according to Hart, to be normally a terrestrial
larva. He has found it two or three times in the earth of corn fields
in Champaign County, Illinois. The dates given are May 31 and
June 4. Specimens were placed in a breeding cage, and an imago of
costalis was secured from them.^'' Hart believes that the species is
single-brooded. His description of larva and pupa follows:

^^ Larva. — Length 20 mm., diameter 2.7 mm. Prothorax with lateral shining
areas about as long as the dorsal, coarsely striate, a smooth spot near center of
disk; dorsal and ventral areas of thorax smooth, a few striae on those of meta-
thorax, especially posteriorly; remaining areas moderately striate, lateral areas
of abdomen a little more finely striate than the others; all more or less shining."

"Dark annuli pale, narrow, longitudinal stripes scarcely present; false feet
with dull pubescent crests, their sides rather finely striate; a narrow dark annu-
lus at base of respiratory tube, another around base of last segment, enclosing
anal prominence and giving off a pair of lateral stripes, the lower one longer; no
projecting spine seen."

'^ Pupa (from defective cast skin of male). — Length 20 mm., diameter 3
mm. Light fuscous brown, shining; abdomen smoothly wrinkled, slightly opaque;
prothoracic spiracular tubercles slightly but nearly equally elevated, free margin
rounded at tip, rima not vertical, evenly arcuate, slightly hooked in front."

"Abdominal spiracular tubercles small, subtriangular, narrower behind, ob-
liquely subconical, much shorter than basal diameter, bearing a very small sub-
circular rima; fringes formed of unequal pale spines, the longer ones sparse on
seventh segment above; outer terminal teeth twice as large as lower pair, directed
laterally and slightly backwards; upper pair smallest, directed upwards; ventral
fringe of last segment not noticeably webbed; lateral tufts rather high, not near
ends of ventral fringe."

Hart's material was, as he says, "not in the best condition for
accurate comparisons."

^^ I have since found the larva of this species in the muddy banks of Lake
Carnegie in Princeton, N. J., which shows that it is not always terrestrial.

106 THE EARLY STAGES OF TABANID.E

Tabanus desertus Walker. — South America, British Guiana. The
larval and pupal stages have been observed by Bodkin and Cleare in
the coast region of British Guiana. Numbers of larvce were found in
a damp accumulation of sweepings situated at the end of a drain
leading from a large cattle pen. Several of these larvae were secured
and kept in the laboratory under frequent observation. Although
supplied with a quantity of suitable food the largest larva eventually
consumed its companions and pupated. After sixteen days a female
Tabanus desertus emerged. The pupal aster (terminal end of the
pupa, showing the arrangement of the spines) is figured (Plate 15,
Fig. 180, a, h).

Tabanus ditaniatiis Macquart. — The distribution of this tabanid as
given by Austen is wide. In Africa it occurs from the Transvaal in
the South to Egypt in the North, while outside the bounds of Africa
it is found in Baluchistan, India, Ceylon, China, and Japan. In the
Anglo-Egyptian Sudan it occurs fairly commonly in the South, but
until 1910 it had not been recorded from the northern provinces, when
King began his studies on this species.

The larvae were taken early in March, 1910, in a small water chan-
nel, locally known as a "Gadwal," on the estate belonging to the
Sudan Plantation Syndicate Ltd., at Zeidab, Berber Province. The
water was for the most part overgrov^^n with a covering of green
slime, and when this was cleared away a few larvae could generally
be seen on the surface. On stirring up the mud at the bottom and
edges of the water more appear, while if one waited for an hour or so,
specimens would continue to rise. They were apparently living in
the mud at the bottom of the pools and coming periodically to the
surface to breathe. They could be seen rising to the surface by a
lashing motion, and if left undisturbed would after a few seconds
sink out of sight again.

Some forty odd larvae of various sizes were taken on March 9 and
placed in a jar containing water, slime, and hollow grass stems; most
of these had disappeared by the next morning, the larger ones having
devoured the smaller ones. On March 10 more than a hundred were
obtained, and, together with the survivors from the previous day,
were divided among three jars (only three being available), two con-

WERNER MARCHAND 107

taining wet mud, and the third water with hollow grass stems and
other debris. Earthworms were provided as food, but were not
taken very readily; the larvag seemed to prefer to eat each other.
They were brought to Khartoum on March 1 1 , and the following
morning each of the thirty-three which were still living was placed
in a separate jar containing clean river sand and water. They fed
freely on tiny earthworms, but their numbers steadily decreased
until about April 16, when thirteen survivors, having attained
maturity, ceased to feed. Up to this stage, if the sand in which they
were living was allowed partially to dry out, they became very rest-
less until water was given them again, but hereafter they preferred
sand which was only slightly damp. In appearance as well as habit
they altered considerably at this stage of their existence. While
young and growing, they possessed well developed prolegs and con-
spicuous dark dorsal markings; now, however, their prolegs became
small, and in color they appeared uniform yellowish white.

These thirteen larvae were left undisturbed until May 26, when one
specimen was washed and was found to have pupated, probably
within the previous two days, as the eyes had not begun to show the
color which they acquired later. On the following day, two more
pupae were discovered in the sand. Prior to pupating the larvae
had made a number of tunnels in the sand, and the pupas were lying
in a more or less upright position in the tunnels and near the surface.

On April 28 King left Khartoum, and traveled in the provinces
until May 30, by which date one larva had died and twelve com-
pleted their life cycles, producing eight females and four males.
The first had emerged on April 29 or 30, so the period passed in the
pupal stage was probably about six days.

King's detailed description is given in about the following words.

Immature Larva. — (Plate 3, Fig. 47; Plate 4, Fig. 57, a-d.) Length 18
mm. Color yellowish white, with dark markings composed of pubescence.
Mandibles dark brown to black, slightly serrated. Anterior margins of the
meso- and metathoracic segments dark, except on the venter. A ring of pseudo-
pods, eight in each ring, two dorsal, two lateral, four ventral, on the anterior third
of each abdominal segment except the eighth, well developed, except the dorsal
pairs on the first and second segments, and bearing spines or hooks. Spines are
also present between the pseudopods on each ring. The rings on the first and

108 THE EARLY STAGES OF TABANID^

second segments edged before and behind with dark pubescence, especially on
the dorsum, the pubescence extending between the dorsal and lateral pseudo-
pods, thus enclosing the dorsal pseudopods in a dark ring. On each of the third
to the seventh segments inclusive is a patch of dark pubescence between the
lateral and the dorsal, and between the dorsal pseudopods, three patches on
each ring, the median patch being conspicuous. To the naked eye these me-
dian patches constitute a median dorsal line of black dots.

On each of the third to the sixth segments, inclusive, are two patches of dark
pubescence, immediately anterior to the dorsal pseudopods. The posterior mar-
gin of the eighth segment bears dark pubescence. The surface of the larva
other than that bearing pubescence is shiny and longitudinally striated.

Mature Larva. — (Plate 4, Fig. 57, b, Fig. 64, a, b, c.) Length 25 mm. Color
yellowish white. Mandibles dark brown to black, slightly serrated. Thoracic
segments shining and longitudinally striated, except the anterior margins, which
are opaque and pubescent. On the prothoracic segment are five longitudinal
grooves, one ventral, two sublateral, two subdorsal, not extending to the posterior
border. On the meso- and metathoracic segments are eight such grooves, four
on either side. The first abdominal segment bears one pair of ventral pseudo-
pods;'^ the second segment one pair of ventral and one pair of lateral; the third
to the seventh, two pairs of ventral and one pair of lateral. Traces of most of
the other pseudopods are present, especially of the dorsal pseudopods on the
fourth to the seventh segments. The pseudopods bear small colorless spines or
hooks, and similar though smaller spines are situated between the pseudopods
and on the dorsum of the first, second, and third segments where the pseudopods
are wanting. On the dorsum of the first and second segments these spines con-
stitute a double band. The posterior third of each abdominal segment is shiny
and longitudinally striated. The anus is edged with pubescence. The syphon
when exserted appears rather shorter than the eighth segment.

Pupal Ca^e.— (Plate 11, Fig. 129; Plate 12, Fig. 141.) Length 17 mm.
Color yellowish brown, thoracic tubercles and abdominal spiracles darker, the
former bearing hairs. On the posterior third of the second to the seventh ab-
dominal segments is a ring of backwardly pointed spines, shortest on the second
segment and longest on the seventh. The eighth segment (Plate 12, Fig. 141,
a, h) terminates in a coronet of six teeth, chestnut brown in color, darker at the
tips, the lateral pair by far the largest, the dorsal and ventral pairs being about
equal in size, the former sometimes slightly the larger. The dorsal pair arises
from between the lateral teeth, the four teeth constituting a row. Ventrally
placed to this coronet are two rows of similar teeth, each row consisting of from
two to five teeth, the two rows together constituting an interrupted trans-
verse row. These teeth are unequal, and vary in size and number in different
specimens.

^^ King uses the term "pseudopod" instead of "proleg" which should be used,
pseudopod being used in protozoology.

WERNER MARCHAND 109

The pupa, when first formed, is yellow with a greenish tinge, especially on the
thorax. Later, as the imago develops, the eyes show as deep maroon and the
thorax becomes generally darker.

On Tabanus ditcBniatus we possess also some notes by Patton and
Cragg, who observed the species in Madras (1913).

Tabanus ditceniatus, being a small species, oviposits, according to
these authors, "on blades of grass just at the edge of a shallow stream,
or on the leaves of the lotus plant at the edges of small ponds, but
never over deep water." It should then fall under the heading of
those tabanids in which the larvae are said to have no air sacs and to
die when falling into deep water.

The eggs of Tabanus ditcBniatus measure, according to Patton and
Cragg, about 1.2 mm. in length, and 2 mm. in breadth, being slightly
more slender than those of Tabanus bicallosus Ric, an Indian
species, studied by Patton and Cragg (1913).

The egg is figured by Patton and Cragg (Plate 1, Fig. 17), evidently
showing a dark band placed subapically, the extreme tip being also
dark on the figure. The egg mass of the same species is figured as
spread out in a single layer on a blade of grass (Plate 2, Fig. 21).

The mature larva is also figured (Plate 4, Fig. 61); attention is
called to the short stout syphon tube.

The pupa is figured by the same authors (Plate 11, Fig. 133), and
the eighth abdominal segment of the pupa is figured (Plate 12, Fig.
147) to show the arrangement of the terminal teeth and the anus.

With these observations our knowledge of the species is quite
complete.

Tabanus epistates Osten Sacken. — -Of this North American species,
Malloch (1917) has given us, in his synoptic table, the following data
on the pupal stage:

Pupa. — Dorsal abdominal segments, except first, armed with an irregular
transverse series, or two such series, of very stout thorns, their bases very much
dilated, slightly caudad of which series there are sometimes a few widely sepa-
rated much longer spines. Seventh dorsal abdominal segment with the poste-
rior transverse series consisting of two long, widely separated spines on the
middle portion, and several, closely placed, on each lateral extremity which are
but little caudad of the much shorter thorns of the anterior series. The portion

110 THE EARLY STAGES OF TABANID.E

of head-capsule between bases of antennae slightly elevated, rounded, fairly
rugose, and not carinate or divided below; abdominal armature moderately strong,
distinctly biserial laterally."

No illustrations are given.

Tabanus fraternus Macqiiart. — An African species, but uncommon
in so damp and well vi^ooded a locality as Mt. Mlanje, where a single
female was bred by Neave on December 16, 1914, from a locally
collected larva, which was not recognized as distinct from that of
Tabanus tcBniola.

The pupal aster (Plate 15, Fig. 182, a, b) resembles that of Tabanus
maculatissimus in having a small papilla on each side of the midline.
The dorsolateral comb consists of only a small number of spines,
which though rather long are fairly stout. The pupal aster and
dorsolateral comb of the female are illustrated.

Tabanus fronto Osten Sacken. — -This American species has been
recorded from North and South Carolina, Texas, and Florida. We
know nothing of its egg-laying habits, but the species has been bred
repeatedly from larvae, by Brimley, in Raleigh, North Carolina. The
larvae of this species occurred freely in the soil in Brimley's garden,
in a comparatively dry locality situated on the crest between two
water sheds, the nearest permanent water being at least a quarter
of a mile away. These larvae are described as white with pale brown
transverse bands, and transform into pupae in June or July, and into
flies some two or three weeks later. The earliest date on which an
adult emerged was July 4, which is also the earliest date on which
Brimley has seen the species in the fields. Two larvae, which had
been preserved in alcohol, were yellower than the trimacidatus larva
described by Brimley, but showed no trace of the pale brown bands
which exist in life. The largest of the two measured 36 mm. long,
and was taken July 5, while the smaller one was ZZ mm., taken on
March 31. Both, as also the preserved trimaculatus (?) larva, were
well, but not abnormally, extended.

The only pupa reported of this species was found by Brimley under
a stone in his back yard.

Although horse-flies do not generally breed away from water,
Tabanus fronto seems to be an exception, as larvaa have been taken in

WERNER MARCHAND 111

Brimley's garden in several different years, while the adults occur
more commonly in the garden and house than any other species of the
family, the flies quite frequently entering the house, while newly
emerged specimens have been noted on a number of occasions.

Tahanus (AtylotusY^ fulvus Meigen. — -A European species. Sharp
describes a larva (Plate 4, Fig. 60, a~d) which he says may belong to
this species. This description follows:

"In a larva, probably of this family (Tabanida;) , found by the writer in the
shingle of a shallow stream in the New Forest (England), the annuli are replaced
by seven circles of prominent pseudopods,^^ on the abdominal segments, about
eight in each circle, and each of these feet is surmounted by a crown of small
hooks, so that there are fifty or sixty feet distributed equally over the middle
part of the body without reference to upper or lower surface. The figures of the
larva of T. cordiger, by Brauer, and of Hcemalopota phivialis, by Ferris, are some-
thing like this but have no setae on the pseudopods."

Tahanus fuscipes Ricardo. — An African species reported by Neave
from Lake Chilwa, southern Nyasaland, in January, 1914, in circum-
stances which render it extremely probable that they had bred in
mud some distance from water, which had been hard and dry for
some portion at any rate of the dry season. Neave is inclined to
think that in this and other mid season species, such as Tabanus
claritihialis Ric. and Tahanus sandersoni Aust., the larvae hibernate
fully fed at the beginning of the dry season and only pupate when the
next season's rains release them from the hard ground.

Tahanus glaucopis Meigen.— A European species, occurring in
Austria as well as in Scandinavia, but evidently rare. According to
Brauer, the larvae have been observed by Wahlberg (1838) in noctuid
caterpillars. I was not able to find the statement quoted by Brauer,
but it may be correct, as Wahlberg made numerous observations on
parasitic and semiparasitic dipterous larvae.

In as far as most of the Bombyliidae are truly parasitic, and tabanid
larvae often burrow deeply into their prey, it is not unlikely that
tabanids belonging to the Pangoniinae which are apparently related

^^ Sharp (Insects, p. 483) uses the generic name Atylotus, which is a subgenus
of Tabanus.

^^ Sharp, like King, uses pseudopod for proleg.

112 THE EARLY STAGES OF TABANID^

to the Bombyliidae, should be of more pronounced parasitic habits.
In Tahanus glaucopis we have probably to deal with an occasional
parasitism, not typical for the species.

Tahanus gratus Loew. — An African species, and, according to Neave,
a very common one and one of the earliest on the wing in southern
Nyasaland, occurring sometimes in August.

Larvse and pupae were collected 'n this month, the first individual
emerging in the laboratory on September 1. The larva (Plate 3,
Fig. 51) is moderately pigmented, though compared with that of
Tahanus insignis the pigmented areas are nearly confined to the edges
of the segments and are not nearly so dark in color. The syphon is
somewhat longer than in that species. A few larvae were obtained
in the stream beds near Mt. Mlanje and a small series in Portuguese
territory to the east of the mountain early in October.

The pupal aster (Plate 15, Fig. 186, a-d) is of the normal type,
being regular in outline. The spines of the dorsolateral comb are
much reduced, especially in the male.

Tahanus hilaris Walker, — A species of Bengal, Assam, and South
India, on the early stages of which we have some notes by Patton and
Cragg (1913). According to these authors its habits of oviposition
are the same as in Tahanus striatus, that is, it oviposits as a general
rule on blades of grass, pieces of stick, etc., at the edge of a river,
stream, or pond. The eggs when deposited are brownish white.
No figures or descriptions are given.

Tahanus ignotus Rossi. — A species not listed in Schiner's Fauna
Austriaca, and, according to del Guercio, synonymous with Tahanus
alhipes Fabr. Del Guercio has given a somewhat unsatisfactory
account of its life history, and its early stages. The species has
appeared, according to del Guercio, in extraordinary numbers, in
company with tipulids, in the rice fields of the region of Bologna
(Italy), and the larva is said by him to have caused considerable dam-
age to the rice fields.

The species, of which del Guercio furnishes a brief description,
not stating, however, on whose authority the insects were identified,
appeared in the region (Molinella) from the first ten to twenty days

WERNER MARCHAND 113

of June. The flies are found on plants like Arum, Sagitlaria, Typha,
etc., on which they move about frequently with a buzzing sound.

The eggs are deposited on the leaves of the plants mentioned and
others, forming circular crusts, which were seen abundantly after
oviposition had taken place, that is, during the whole month of
July, while some belated specimens may oviposit as late as August.
A more detailed description of egg or egg cluster is not given.

Larvae were never found in the fall, but half grown larvs were found
in winter, most of them being ready to transform in April or May of
the following spring.

Without insisting on the strangeness of the fact, de Guercio re-
ports that these larvae are so similar to those of the tipulids found
by him {Tipula oleracea L.) that without an accurate examination
it would not be possible to distinguish them. They are more robust,
more cylindric, but of the same color; they are in the same way pro-
vided with brushes of hairs on the segments of the body, but the
mouth-parts have mandibles of about twice the size and thickness of
those of the tipulids, from which they differ only by the kind of
velvety combs found on the clypeus and in the whole anterior part
of the head. The figure to which del Guercio refers, is given (Plate
7, Fig. 96, a, h) for comparison. The habits of these larvae are said
to be the same as those of the tipulids with which they are found,
and to behave in the same way when transforming into pupae.

I have given this description as it is the purpose of the present
report to give a full account of all that is known about the early
stages of Tabanidae. But it may be permissible to express a reason-
able doubt whether del Guercio did not eventually figure two tipulid
larvae, having taken one of them, erroneously, to be a Tahanus. As
no reference is made to any previous descriptions of tabanid larvse,
excepting (see below) Tabanus autumnalis, and in this case also the
literature is not quoted, and as there is no mention of the structures
common to all known tabanid larvae, as the prolegs, syphon, respira-
tory tube, serrated mandibles, the assertion that the larva of this
Tabanus can hardly be distinguished from that of Tipula oleracea
cannot be taken seriously.

The transformation takes place, according to del Guercio, from the
end of May to early June, when the larvae leave the rice fields in
large numbers for their metamorphosis in the dams and meadows.

114 THE EARLY STAGES OF TABANID^

While the Tipula appears in two generations, the life cycle of the
Tahanus comprises only one generation, which begins at the end of
spring of one year and is completed towards the middle or end of
spring of the following year.

The larva of Tipula and those of Tahanus are, as del Guercio in-
sists, of the same dark gray color as the soil in which they live and re-
semble one another not only in shape but also in many of their organs,
including the mouth-parts, in such a way that there remains no
point of resemblance, even in this regard, between the larva of the
Tahanus of the rice field and that, for instance, of Tahanus autumnalis,
the larvas of which differ from it in every detail.

The pupa, however, is figured by del Guercio (Plate 11, Fig. 123)
as the pupa of this Tahanus, for comparison with the tipulid pupa
(Plate 11, Fig. 122), and is in fact in every way a real Tahanus pupa,
except that on the figure no spiracles are visible, which were appar-
ently overlooked by the author in making the drawing.

Del Guercio' s statement that the damage done by tipulids and
tabanids in the rice fields resulted in the destruction of the whole
cultivation, has probably to be corrected in that as far as any larvae,
the tipulids may possibly have been responsible for the damage; for
the tabanids, however, this is very unlikely, and, as long as del
Guercio has, in fact, made no observations whatever on the food of
his larvae, we are bound to assume that the Tahanus larvae, which
were apparently numerous, but of which we are still awaiting de-
scription, fed on the larvae of Tipula oleracea, and were in this in-
stance certainly not injurious.

I wish here to call attention to the fact that Degeer has already
spoken of a great regemblance between tabanid and tipulid larvae;
in fact, before he knew that the larvae found by him were larvae of
Tahanus, he expected crane-flies to hatch. However, his descrip-
tion, the first one ever given of a Tahanus larva, establishes beyond
doubt the characteristics of these larvas by which they differ from
those of tipulids.

Tahanus insignis Loew. — {Tahanus sharpei Aust.). An African
species, common in southern Nyasaland, near Mt. Mlanje, from No-
vember to March.

WERNER MARCHAND 115

The very characteristic and strikingly pigmented larva (Plate 3,
Fig. 53) was, according to Neave, common in the mud of the forested
streams from the end of September. It may be distinguished at a
glance from other similarly pigmented species by the white trefoil-
shaped area on the dorsum of the anal segment. This is a voracious
and predacious larva and troublesome to keep in the laboratory for
that reason.

The pupal aster is of the normal type, the spines of the dorsolateral
comb being few in number but somewhat long. A small series of
adults was bred by Neave from the larvae above described. A cer-
tain number of these flies belonged to the type of Tabanus sharpei
Aust., and the two forms, the larvas of which are identical, are con-
nected by a great variety of intermediates.

The larva is j&gured (Plate 3, Fig. 53), also the pupal aster of the
male and the dorsolateral comb of both sexes (Plate 15, Fig. 184, a,
b, c). In the figure, the larva looks much like that of Tabanus atratus
but the lateral stripes are poorly developed except on the thoracic
and on the ninth and tenth segments, while on the eleventh segment
they are more completely fused with the transverse ones into a broad
pigmented band.

Tabanus kingi Austen. — Tabanus kingi, recorded from Khor Arbat,
Sudan, Africa, is a species superficially resembling Tabanus tceniola,
and allied to species of Tabanus from Abyssinia at present unde-
scribed. The life history has been worked out by King (1910), who
observed the species in Khor Arbat, in a locality consisting of a
stream of slightly brackish water running in a gorge on a rocky hill.
On emerging from the hills into the plain the stream disappears in
the sand. In the autumn, during the brief rainy season, it comes
down in sudden overflow, and is then of considerable size, but in April,
the month in which these observations were made, it is, except where
there are pools, not more than a few inches in depth. The bed of
the stream is stony and there is little or no vegetation on its banks.

The female fly deposits her eggs in a rounded mass on a rock
rising sheer from the water generally slightly overhanging, and
from six to fifteen inches above water level. Rocks chosen for
this purpose overhang comparatively deep pools, from eighteen

116 THE EARLY STAGES OF TABANID^

inches upwards, in which the water moves but slowly. Such rocks
occur only here and there, in the mile or so of stream searched, only
three rocks bearing traces of having been used by this tabanid for
purposes of oviposition being found. On one of them were the re-
mains of several hundred egg masses lining a small crack in the face
of the rock from 2 to 3| feet above the water level. As none of the
fresh egg masses found were situated more than 15 inches above
water level, these old masses had probably been deposited when that
level was higher. Altogether seven females were taken in the act of
ovipositing, and several more were seen. No particular time of the
day seems to be chosen for the act; one was found ovipositing at
11.40 a.m. and another at 4.40 p.m., and unlike Tabanus higuttatus
Wied., the only other horse-fly King has observed ovipositing in the
field, this tabanid does not lose her natural wariness while engaged
in depositing her eggs. In fact, she is often more difficult to capture
then than when merely sunning herself on a rock.

The egg masses, figured by King (Plate 1, Fig. 20), vary in size,
and no count of the number of eggs contained was made, but the
average mass is believed to consist of about 500. When freshly laid
the mass is glistening white and can be seen from a considerable
distance, but within a few hours it takes on a mottled gray hue
which so closely resembles the color of the rock that it is not easily
detected. While the fly is occupied in laying her eggs, numbers of a
tiny Hymenoptera assemble and proceed to add their eggs to the
mass, continuing to do so after the fly has gone. From some twenty
egg masses collected from the rocks about equal numbers of this
egg parasite and of the tabanid larva were obtained. Specimens of
these Hymenoptera were sent for identification to the Scientific Sec-
retary of the Entomological Research Committee. They proved to
be a new species of Chalcidas, and have been described by Mr. J. C.
Crawford, of Washington, under the name of Telenomus kingi (foot-
note to King's paper, by Mr. Guy A. K. Marshall) .

One horse-fly''^ taken in the act of ovipositing completed her egg
laying in a collecting box on the evening of April 13. These eggs had

^^ King frequently uses the Arabic word "seroot," for horse-fly, by which the
flies are known in the Sudan.

WERNER MARCHAND 117

hatched by the morning of April 19, the incubation period being there-
fore about five days. Under normal conditions, exposed to the sun,
it may possibly be less. The larvae from these eggs were allowed to
fall from the egg mass into a basin containing water and stones and
were provided with portions of earthworms, and tiny coleopterous
and dipterous larvae obtained from wet moss. They refused to feed,
however, and all died; probably at this stage of their existence they
require brackish running water.

In places the stream at Khor Arbat is very shallow and ripples
over and around stones; under these stones larvae of various sizes,
mostly nearly mature, were taken. Apparently stones which were
not quite or were barely covered with water were chosen by the
larvae in order that they might come up to breathe without losing
their hold and so be in danger of being carried away by the current.
Usually only a single larva was found under one stone, and, in every
instance where two or three were together, a mortal combat was
taking place. If a larva was placed on one's hand it would at once
endeavor to drive its mouth hooks through the skin, and where the
skin was thin, it would succeed in inflicting a sharp pricking pain.
Owing to these cannibalistic habits the number of larvae which
could be transported was restricted to the number of vessels available,
so, though nearly two hundred were taken from the stream, only
forty-two were brought alive to Khartoum. There they were placed
in jars containing coarse sand, brought from Khor Arbat, and water,
and fed on medium sized earthworms. They took these willingly
when hungry but appeared to need food only every two or three
days. King left Khartoum on April 28, Captain W. B. Fry taking
care of the larvae during his absence, and on King's return on May
30, the majority of these larvae were dead. One, however, had com-
pleted its life cycle and seven were still living. Six of these seven pu-
pated during the next few weeks but died as pupae. The pupal
period is probably about six days, for one which pupated on May 5
appeared to be mature on May 11, when it perished.

One empty pupal case was taken under a stone in the bed of the
Khor Arbat stream; the fly must have crept up the stone through
several inches of running water before gaining the air.

118 THE EARLY STAGES OF TABANID^

Although this tabanid, according to King, in its adult form closely
resembles Tabaniis tccniola Pal. de Beauv., in its larval stage it
differs markedly from that species. The larva is admirably adapted
for clinging to stones in rapidly running water, its usually long pseudo-
pods with strong hooks, being retractile and capable of being used
as suckers. None of the other tabanid larvae observed by King
possessed an anal proleg.

Besides the seven specimens mentioned above as having been taken
in the act of ovipositing, two more were caught sucking blood from
Camels. No males were seen.

King describes the early stages in about the following words.

Egg. — Length 2 mm. Color white, becoming darker as the embr}'0 within
develops. Spindle-shaped.

Mature Larva. — (Plate 4, Fig. 56 and Fig. 65, a, b, c.) Length 35 mm. Color
pale gray to dusky gray to deep chestnut-brown. Mandibles dark brown to
black, long and powerful, slightly serrated. Anterior margins of meso- and meta-
thoracic segments dark. A smooth shiny pale area on the dorsum of each thoracic
segment; on the prothorax this area is concave anteriorly, convex posteriorly,
and with parallel sides. On the meso- and metathoracic segments it appears to
the naked eye diamond-shaped, though sometimes it is actually hexagonal. On
the venter of the prothorax are two shiny pale longitudinal areas, each bearing
several long black hairs arising from a single pore; a similar but larger area is
striated on each of the meso- and metathoracic segments, bearing two similar
tufts of hair. On either side of the meso- and metathoracic segments are three
longitudinal areas not extending to the margins of the segments, longitudinally
and deeply striated. On the anterior margins of the meso- and metathoracic
segments on either side are four paler lines extending backwards, to form the
divisions and edges of the three striated areas. On the anterior third of each
abdominal segment except the eighth is a ring of pseudopods, eight in each ring,
two dorsal, two lateral, four ventral, except on the first segment, where the
dorsal pair is wanting. The dorsal pseudopods are never well developed, and,
with the exception of those on the fifth, sixth, and seventh segments, unprovided
with hooks. The lateral and ventral pseudopods are very long and bear at the
apcies long, strong hooks, chestnut-brown in color, sometimes darker at the
tips. On the median pair of ventral pseudopods on the fourth, fifth, and sixth
segments, these hooks form a complete circle, but on the remaining pseudopods
bearing hooks the circle is incomplete. Immediately below these hooks is a
row of tiny spines. Immediately behind the ventral pseudopods on the first to
the seventh segments inclusive is a shiny striated area. On the venter of the
eighth segment, anteriorly placed to the anus, is a pseudopod equal in size and
similar to the ventral pseudopods on the other segments, and bearing an incom-

WERNER MARCHAND 119

pJete circle of hooks. Scattered over the surface of the larva are occasional
black hairs. The syphon, when exserted, is shorter than the eighth segment, and
bears a number of black hairs. The dark appearance of the larva is due to tiny-
dots of pubescence arranged closely together, except on the shiny areas mentioned
above.

The skin of the larva frequently bears scars of old wounds.

Pupal Case. — (Plate 12, Fig. 137 a, h, c.) Length 20 mm. Color yellowish brown,
thoracic tubercles and abdominal spiracles darker, the former bearing hairs. On
the posterior third of the second to the seventh abdominal segm.ents inclusive
is a ring of backwardly pointing spines, shortest on the second segment and
longest on the seventh. The eighth segment terminates in a coronet of six
teeth, chestnut-brown in color, darker at the tips, the lateral pair by far the
largest, the dorsal and ventral pairs being equal in size. These teeth are ar-
ranged roughly in a circle. Ventrally placed to this coronet are two rows of
five comparatively thin spines, of varying length, together constituting an inter-
rupted transverse row. Dorsolaterally placed to the coronet are two rows of
spines similar to the ventral row.

The dorsum of the abdomen is sometimes clothed with black pubescence
arranged in four longitudinal stripes. On the sixth and seventh segments these
stripes merge and on the seventh segment the pubescence is confined to the
posterior third. The pubescence is wanting on the dorsum of the eighth segment
but is present on the venter of the seventh and a small patch is situated immedi-
ately below the coronet on the eighth segment.

The pupa when first formed is yellowish. Later, as the imago develops, the
eyes show as dark spots with a greenish tinge and the thorax becomes generally
darker.

Tahanus lasiophthalmiis Macquart. — This species, which is widely
spread and occurs, according to Hine, in eastern North America,
Columbia, and Chile, has been reared by Hine (1906) from the egg
to the adult. The fly is one of the earliest of the genus to appear in
the spring, adults having been taken at Columbus, Ohio, as early as
May 20, and it is common during the first half of June. The eggs are
placed in masses on various plants that grow in low, wet ground, but
Hine has not observed them over water. The masses are shiny
black when fully colored, rather small for members of the genus,
only slightly convex, and accompanied with an unusual amount of
cementing material, which nearly obscures the form and arrange-
ment of the individual eggs. The mass suggests somewhat a drop of
tar or other black substance fastened to the surface of a leaf of the
common cattail reed (Typha latifolia), a sedge, or some other plant.

120 THE EARLY STAGES OF TABANID^

The eggs are usually deposited after June 10, and the specimens
from which larvae for rearing hatched were taken in Medina County,
Ohio, on a common sedge found growing near the outlet of a small
spring. They were collected June 28 and hatched the next day and
the day after. As Hine had not been successful up to this time in
keeping very young larvae for any length of time, it was decided to
try different methods of treatment in order to find out, if possible,
that which is best suited to their requirements. Some were placed
in a jar containing water only (No. 1) ; others in a jar containing water
with a couple of inches of sand in the bottom (No. 2). A third jar
(No. 3), in which larvae were placed, contained wet muck, while the
fourth lot (No. 4) was placed in a jar containing moist sand to the
depth of about 3 inches, covered over the top with a quantity of fine
leaves of water plants. In all the breeding jars were placed plenty
of small crustaceans and other minute invertebrates procured from
water by means of a finely-meshed sieve.

It was soon observed that the larvae in breeding jar No. 4 fed on
the crustaceans, and at the end of a few days showed a distinct in-
crease in size. Those in the jars containing water soon died, and jar
No. 3 did not appear to be a success, so all but No. 4 were aban-
doned. The larvae in this last, however, were separated, and placed
in similar jars, one specimen in each, and reared to full size, the
adult fly being procured the following spring.

Since, as stated, three of the four jars started were soon abandoned,
what is said hereafter regarding the method used in rearing pertains
to the single one retained. A glass jar was selected so that the
actions of the larvae could be observed; a small jar seemed desirable
because the larvae are predacious and eat their own kind as readily
as anything else, for which reason it is necessary after a short time
to place only a single specimen in a jar; also, even a small receptacle
furnishes plenty of room and the long series which it is desirable to
have takes as much space in the insectary as one cares to give to a
single species. Only the quantity of sand and other material neces-
sary to success should be placed in the breeding jar, as it is desirable
once in a while to look this material over carefully in order to locate
the very small specimens and find out what they are doing.

WERNER MARCHAND 121

Half pint jelly glasses were found by Hine to be well suited for the
purpose and easily obtainable. Covers proved to be desirable in
order to prevent too rapid evaporation of moisture, but a small per-
foration or two in them was necessary to furnish ventilation. As the
muck which was tested as soil for the Jars grew much mold, clean lake
sand was chosen as decidedly preferable for the purpose. The cov-
ering of plant material mentioned furnished a resting place for the
small crustaceans offered for food, and the larvte themselves seemed to
choose to remain in it in preference to burrowing into the sand, al-
though they were apt to be found in any part of the jar. Algae made
good material for covering, but only a small amount could be used,
and too much water was detrimental as either in excess tended to
cause decay, and consequently a bad odor, which was observed to be
unfavorable to the insects. The principal point in favor of the algae
was that they contained no hollow stems or large pieces into which
the larvae could crawl, but still, being composed of small soft par-
ticles, furnished a mat in which they could hide. When it was de-
sired to locate these larvae it was easily done by picking the mass to
pieces. As odors, which are often fatal to the larvae, were likely to
develop from the material put in for food and also from other sources,
it was found necessary to watch the jars continually, giving them a
thorough cleansing once in a while, and perhaps putting in fresh sand
and plant material occasionally.

Larvae when first hatched were about 2 mm. in length; they grew
rather slowly, but in fifteen days after hatching had doubled their
length. They fed readily on the small crustaceans which were given
them. It was impossible to give these small crustaceans their proper
surroundings, so many of them died, and it was observed that the
young larvae fed on these as well as on the specimens v/hich they
killed themselves. The larvae could be seen crawling about in the
jars; they appeared to remain very near the upper surface of the
sand most of the time, and when food was scarce did much crawling,
but when food was plentiful satisfied their appetites and hid among
the plant material where they remained quiet.

A difference in size in the various larvae soon became apparent, and
the older they became the greater was this difference. On July 23,
twenty-five days after hatching, some specimens measured as much

122 THE EARLY STAGES OF TABANID^

as 7 mm., while others measured only 3 mm. At this date angle-
worms were given for food, and were accepted readily, and appeared
to be as satisfactory as the crustaceans, but it would seem that the
latter are preferable for the stage just after hatching.

On July 27 some of the larvas were 10 mm. in length, and on August
2 the same specimens measured 12 mm.; thus at this stage they grew
more rapidly than when they were younger. They fed actively till
about the middle of September, when they had become apparently
full grown, or 25 mm. long. Length in the larvae of tabanids is, ac-
cording to Hine, not a satisfactory means of indicating the size, for
the segments telescope on one another in such a way that it is difficult
to take two measurements exactly alike, but an endeavor was made
in this case to make the different measurements similar, so I believe
that those given are considered sufficient to indicate the comparative
sizes of the different ages. After September 15 the few specimens
remaining alive buried themselves in the sand of the breeding jars
and were quiet most of the time until March 10, when one pupated,
the adult emerging on the 25th of the same month. The others died
before the pupal stage was reached. Hine has noted that larvs of
various species of tabanids taken from their natural habitats during
the winter did not produce adults in the spring much before the
same species appeared naturally, but in this case, where the specimens
was kept under artificial conditions during its entire life, the adults
appeared almost two months earlier than is normal in nature.

Hine's description of the larva follows:

"rAe mature larva [Plate 3, Fig. 44] is not notably different from those of
other species of Tabanus so far as form and appearance are concerned. The color
is a dirty white with a pinkish shade over most of the body; the prolegs are not so
prominent as in many species, and on this account specimens appear somewhat
maggot-like. On either side of the body is a longitudinal row of very small black
spots or specks, one to each segment and located just above the ventral pro-
legs; these spots are lacking on some of the anterior and some of the posterior
segments; their presence appears to be characteristic of the species, at least so far
as my acquaintance with different larvae goes. Mature specimens are about 25
mm. in length."

Hine has not taken the larva of this species in its natural habitat,
therefore he cannot say where it is to be found, but he thinks that

WERNER MARCHAND 123

it lives in debris, or in the ground around low places near where the
eggs are laid.^^

" The pupa [Plate 1 1 , Fig. 127] is somewhat dusky in coloration, the thorax being
almost black. The terminal teeth of the abdomen [Plate 13, Fig. 159] are quite
different from those of any species studied so far, and these differences alone
make its determination easy. The dorsal and lateral teeth are much larger than
the ventral, the lateral being much larger than any of the others; the ventral
teeth point almost directly backward, while the direction of the others is largely
upward. The thoracic spiracle is rather small and nearly longitudinal, its rima
is curved, but no distinct hook is formed at the posterior end. Length 18 mm."

We owe our knowledge of tlie early stages of Tabanus lasiophthalmus
entirely to Hine.

Tabanus laverani Surcouf. — An African species, being rare in Neave's
locality, IVIt. IVIlanje in southern Nyasaland, where only occasional
specimens were taken.

A single female was bred on November 25, 1913, from a larva col-
lected near Neave's headquarters. The larva did not belong to the
pigmented type like that of Tabanus gratus and resembled the larva
of Tabanus variabilis in bearing lateral prolegs on the anal segment.
It was, however, of a yellower color and less transparent than that
species, and lacked any pigmentation on the syphon.

The pupal aster (Plate 15, Fig. 185, a) is remarkable for the great
size and elongation horizontally of the middle pair of hooks. The
dorsolateral comb is reduced to tv/o very short processes, the main
combs on the last segment being also of this character, as may be
seen from the view in profile.

The pupal aster and dorsolateral comb of the female are figured
(Plate 15, Fig. 185, a, b,c).

Tabanus lineola Fabricius. — A common species, inhabiting eastern
North America, common in states as wide apart as IVTassachusetts,
Ohio, New Jersey, and Louisiana.

We owe to Hart a description of larva and pupa. The larva closely
resembles the young of nigrescens, and was not separated from it at
first. Examples were taken at Hart's collecting station, C, near the

^- 1 have since found (May, 1917) the larva of this species in the muddy bank
of a rapid flowing brook in the neighborhood of Princeton, N. J.

124 THE EARLY STAGES OF TABANID^

foot of Quiver Lake, Illinois, over sand, mud, and algas vegetation;
at Station I, in the bed of the slough with grass, rushes, and willows,
and a very shallow stream of spring water when the river is low; and
at Station H, on the Illinois River below Havana, Illinois, where it is
narrow, (the east bank steep and sandy, a layer of mud over sand at
lower levels, the water quickly deepening, considerable current, a
little vegetation; on the west bank mud, steeply sloping, trees but
almost no vegetation, decided current), on April 14, 15, and 30;
and in Flag Lake, (shallow, muddy, bordered with rushes, thick with
floating vegetation), on April 27 — as shown by specimens preserved.
The larvae were also taken April 8 and June 15 and 24 in Sand Lake,
Lake County, Illinois, and in ponds in Kane and Champaign Counties.
I give Hart's description of the larva.

"Larva. — Length 20 mm., diameter 2.7 mm. Prothorax with lateral shining
areas about as long as the dorsal area,^^ striation about the same as that of the
upper mesothoracic area, no noticeable central smooth spot, a small one on the
lower margin posteriorly; remaining lateral areas a little more finely and closely
striate; dorsal and ventral areas of thorax nearly smooth on disk, with basal
striae; those of abdomen with moderately close striae, more or less interrupted on
disk; all areas more or less shining."

"Surface whitish, dull pubescent markings very light brown but distinct, annuli
narrow, crests of false feet also dull pubescent, their sides striate; lateral stripes
of thorax distinct, slender, not dilated posteriorly, lateral edges of dorsal areas
of thorax diverging. An opaque dark ring about base of respiratory tube, and
another encircling anal prominence, above it usually three light brown spots."

"Main internal tracheae rather thick and noticeable, subparallel, not strongly
sinuate, at least back of the middle. Terminal stigmatal spine often protruded."

Of Tahanus lineola Hart has obtained three pupas on May 18 of
different years. Imagos were obtained from these on May 27, 29,
and June 6. The tabanid pupae develop, according to Hart, much
more rapidly in hot weather than in cold, and to this fact is probably
due the difference in time of emergence. Another pupa was taken at
Matanzas Lake on August 24. Hart's description of the pupa follows:

"Pupa—iVlEite 12, Fig. 149; Plate 13, Fig. 162.] Length 19 mm., diameter 3
mm. Light ferruginous brown, shining, abdomen roughly wrinkled and sub-

^^ I cannot confirm this character from material collected in Princeton, Spring,
1917.

WERNER MARCHAND 125

opaque. Palpal sheaths indistinct, not distant; tubercles not dark; ocellar tuber-
cles indistinct or wanting; thoracic spiracular tubercles slightly but nearly equally
elevated, free margin rounded at tip, rima not vertical, evenly arcuate, slightly
hooked in front."

"Abdominal spiracular tubercles subtriangular, narrower behind, obliquely
subconical, much shorter than basal diameter, bearing a small subcircular or short
and strongly arcuate rima; on anterior slope a transverse groove, usually longer
than the rima; fringes formed of unequal pale spines, only one or two long spines
above on seventh segment; outer terminal teeth much longer than the others,
directed laterally and upwards, the tips of the four upper teeth about in line.
Fringe anterior to anal prominence showing a chitinous webbing between the
bases of the spines, so that the separated tufts of the female look like a pair of
broad low teeth with several spiny points; lateral tufts low down, near ends of
ventral fringe, formed of short spines."

The pupa is also figured by Malloch (from Hart's material), (Plate
12. Fig. 149).

Oviposition and eggs are not known. Also it remains to be deter-
mined whether there are one or two broods of this species in one
year. Adults were taken as early as May 17, and as late as September
27, with dates of capture in June, July, and August.

Tahanus maculatissimus Macquart. — ^An African species com-
mon in the neighborhood of Mt. Mlanje, southern Nyasaland, where
Neave obtained data on the early stages.

The larvae were found in mud in a partially dried up stream, and
from these a few individuals of the adult were bred during Novem-
ber. The larvae were obtained in Portuguese territory to the east of
Mt. Mlanje. These larvae were not, however, at the time distin-
guished from those of Tahanus biguttatus, of which they were thought
to be immature examples. The figure (Plate 5, Fig. 71) is from
other individuals, obtained subsequently, which are believed by
Neave to belong to this species. This figure shows the pigmentation
pattern of the eleventh segment considerably developed, while the
dark circular band at the posterior end of the segment is compara-
tively narrow.

The pupal aster (Plate 14, Fig. 178, c) is normal except for a pa-
pilla on each side of the middle line, about the middle. There is a
well marked dorsolateral comb, consisting of comparatively short
stout spines. The pupal aster and dorsolateral comb of the female

126 " THE EARLY STAGES OF TABANID^

and the dorsolateral comb of two different male individuals are fig-
ured (Plate 14, Fig. 178, a~d).

Tahanus medionotatus Austen. — An African species, to which Neave
attributes, with some doubt, a series of specimens collected in southern
Nyasaland in 1915. Of this species five males and three females
were bred between the end of September and the beginning of
November.

The pale colored larva has rather long prolegs, a ring of pigment of
varying width around the base of the syphon (Plate 5, Fig. 68), and
another ring around the anus, which is usually prominent in the
living larvae. There are also present two prolegs immediately ante-
rior to the anus, but these are not visible in the somewhat contracted
preserved specimen and therefore are not shown in the figure. The
pupal aster (Plate 14, Fig. 174, a, h) resembles that of Tahanus
ohscuripes in having a large and even longer, but less horizontal,
middle pair of hooks. Not only the dorsolateral but also most of
the lateral comb is absent.

Tahanus melanoceros Wiedemann. — A species recorded from the
Atlantic States from New Jersey south.

Late in March, 1909, Brimley in Raleigh, North Carolina, found,
v/hile looking under stones in a small clear woodland stream, a
Tahanus larva which was quite lively and seemed thoroughly at home
in clear water. He kept the specimen in a bottle with some wet leaves
and practically forgot it. However, on May 18 it had transformed
to a pupa, and thirteen days later, on May 31, a male of Tahanus
melanoceros emerged from the pupa. The larva was approximately
the same size as the trimaculatus larvae collected by Brimley, and was
like them white without darker bands.

Tahanus nagamiensis Carter. — ^An African species, only recently
described, of which a single female was captured by Neave on the
Malosa River, the Anglo-Portuguese boundary south of Mt. Mlanje,
southern Nyasaland, on October 8, 1913. A male was bred from a col-
lected pupa on September 27, 1913.

The pupal aster (Plate 15, Fig. 179, a, h) somewhat resembles that
of Tahanus laverani in the great development of the middle pair of

WERNER MARCHAND 127

hooks, the dorsolateral comb is absent, a characteristic which Neave
has not seen in the pupa of any other species of Tahanus, except in
Tahanus medionotatus . The other combs are, however, striking and
characteristic, as may be seen from the figure.

Tahanus nigerrimus Zetterstedt. — Of this European species we have
some indication that it is lignicole in its larval stage. Scholtz (1850)
expresses the opinion that perhaps some Tahani live in their early
stages in the detritus of old tree trunks. He found, in July, 1848, in a
deep wooded ravine near Charlottenbrunn, on the decaying trunk of
a tree of Fagus sylvatica, a newly hatched but already fully colored
specimen of Tabanus nigerrimus Zett., which had previously been
found in Sweden and Norway.

Tahanus nigrescens Palisot de Beauvois. — This species is nearly
related to Tahanus stygius, occurring in the Atlantic States (New
Jersey), and also in Illinois. Of its oviposition nothing is known,
but Hart has observed the larva and given us a description of the
pupa.

An undersized larva, supposed to belong with those of stygius,
pupated May 18, and on June 1 produced an imago of Tahanus
nigrescens. Most of the larvae, treated in Hart's paper as stygius
were very uniform in size and characteristics; Hart, though unable
to separate the two species at this stage, believes that the bulk of
his material at least was stygius. The imago of nigrescens had been
taken previously near the Mississippi, in southern Illinois, on
August 10.

^ Pupa, Male. — Length 25 mm., diameter 5.5 mm. Palpal sheaths narrowly
separated, about one fourth as far apart as the setae borne by the large frontal
tubercles, a smooth depressed space between them, without tubercles. Lobes
of carinate transverse ridges of head more rounded and separated by a deeper
notch than usual. Abdomen a little more shining and more smoothly wrinkled.
Otherwise not different from the pupa of stygius (female) next described."

From Malloch's analytical table the following data are obtained
on the pupa:

"Pupa. — Dorsal abdominal segments, four to six at least, with one or two
transverse series of short irregular spines, the bases of which are not much dilated,

128 THE EARLY STAGES OF TABANID.^

and slightly caudad of these is a transverse series of closely placed, very long,
slender bristles. Length at least 25 mm.; abdominal spiracles with very much
elongated vertical rima, the upper and lower extremities slightly curved forward.
The long spines on dorsal abdominal segments either black-tipped or all pale,
without a black preapical ring; short spines in anterior dorsal series slender and
very uneven. A small but distinct tubercle just in front of base of middle leg in
addition to and some distance above the one bearing the paired hairs."

Malloch's material, which v^as partly the same as that collected
by Hart in 1895, contained only one specimen of nigrescens (in very
poor condition).

The prothoracic spiracle of the pupa has been figured by Malloch
(Plate 13, Fig. 156).

Tahanus ohscuripes Ricardo. — An African species, occurring in
southern Nyasaland, chiefly on the plains in Portuguese territory to
the east and south of Mt. Mlanje, in October and November.

A single male was bred by Neave from a pupa collected on Octo-
ber 1. The pupal aster, which is figured (Plate 15, Fig. 183, a, h)
somewhat resembles that of Tahanus laverani and has the same large
middle pair of hooks. The dorsolateral comb is reduced to a single
knob-like process.

Tahanus {N eotahanusY^ ochrophilus Lutz. — A common Brazilian
species. On its early stages we possess some notes by Lutz in Rio de
Janeiro (1914).

Larvae of this species were found in muddy or sandy soil below and
at the sides of a small brook with slowly flowing water. Large
quantities of sand had to be sifted to obtain them. As food, tubi-
fex was given. The full grown larvae measured 30 mm. Color
creamy white, integument shining and transparent, intestine reddish
or blackish. Tracheag silvery. Mandibles dark, serrate beneath.
Digestion slow, four to five days. Before pupation, contracted; in
spontaneous death, expanded.

Culture in moist sand, but for observation purposes moss in glass
jars was preferred. The larvae have to be kept in the dark during the
intervals of observation. The pupal stage lasted about ten days.

^^ The few species of Neotahanus of which early stages are known are here class-
ified with Tahanus. There is no doubt that the genus Tahanus could be subdi-
vided into smaller genera or subgenera but opinions on this point are not definite.

WERNER MARCHAND 129

The eyes change color after two or three days. Eye coloration and
stripes of abdomen visible two to three days before hatching. Pupal
shell very transparent.

Tahanus orientis Walker. — A Tahanus occurring in North India,
Nepal, Bhutan, particularly in high elevations; according to Baldrey,
"by far the largest of the species caught at Muktesar (7,500 feet);
it is caught at the beginning of March and appears then to be quite
full grown (owing to its torn and worn out condition)."

According to Baldrey, who tried to breed the species for experi-
mental purposes, it begins laying its eggs during the last weeks in
April, and all flies dissected up to June 19 were found full of eggs.

The eggs when first laid are of a creamish white color, but after a
few hours this changes to a cigarette-ash color. They are laid in
batches resembling bunches of bananas.

Four lots of eggs were found in jars containing iiies; in two cases
the only fly in the jars was found dead.

Tahanus par Walker. — Of this African species we possess full de-
scriptions of the Qgg, larval, and pupal stages by King (1910), who
not only secured oviposition in captivity, but also bred the species
through from the egg to the adult stage.

Occasional specimens of this tabanid were met with on the White
Nile from Gebelein southwards. In the country behind Bor females
were abundant and seemed to spend their time resting on foliage,
waiting for approaching animals which they would attack at once.
No eggs could be found, though a careful search was made in all
the places that were considered likely to serve as breeding grounds;
hence a number of females, gorged with blood, were placed in a
breeding cage, in which was also a dish containing mud, water, and
growing grass and weeds. The flies fed on sugar and water, and
though the majority died within the first two days, the survivors
eventually produced three small batches of eggs (Plate 1, Fig. 10).
The eggs obtained in this manner were deposited on May 23 and 24,
on the under sides of the leaves of a water weed. Unlike the eggs of
most members of the genus Tabanus, they were not closely packed
in a rounded mass, but placed vertically and separately though in a

130 THE EARLY STAGES OF TABANID^

cluster. The single egg is "spindle-shaped, about 1.15 mm. in length
and white in color, becoming darker as the embryo within develops."

The eggs obtained hatched on May 30, consequently after an incu-
bation period of six to seven days, and the tiny larvae (Plate 1, Fig.
11) were divided into three lots, and placed in glass basins containing
mud, water, and growing grass. These basins for purposes of refer-
ence were labelled A, B, and C. At the time when the eggs hatched,
King was in the Sudan region, where it was impossible to land and
obtain any subterraneous insect larvae or tiny fresh water crustaceans
for them, so they were offered the expressed stomach contents of
gorged female ticks — Rhipicephalus simus — taken from a dog. A
few fed once or twice but the majority refused, and all buried them-
selves in the mud.

On June 11, the larvae from A were transferred from mud to clean
river sand and water, and given freshly killed mosquito larvae. They
fed on these readily and grew apace, though at greatly varying rates.

The larvae in B were also given mosquito larvae from June 11, but
they refused to feed, and the mud in which they were living was sev-
eral times allowed to dry up. On July 11 they were placed in clean
river sand and water, and at once began to feed and grow.

On July 19, when King returned to Khartoum, their diet was
changed, owing to the difficulty of obtaining mosquito larvae, to
freshly killed and bruised earthworms. They did not take readily
to this food, and some died, while others disappeared from the basins.
At the time it was thought that they had begome cannibals, but
eventually it was found that they were being taken by mice. The
stock of larvae from A and B had by this time become reduced to
one, which appeared to be full grown and so was killed and preserved.

On July 26 the larvae from C were transferred to clean river sand
and water. It was then fifty-seven days since they had emerged
from eggs, and they had spent a great part of that time in a dry
cake of mud. Occasionally this mud had been moistened, and food
offered them, but they had very rarely taken it. Most of them were
alive, but with the exception of a few which were slightly larger than
when just hatched, they had not grown at all. They now under
more favorable conditions fed readily on a mixed diet of earthworms
and mosquito larvae and grew, some rapidly, others more slowly.

WERNER MARCHAND 131

On September 3 and 4 one pupated, lying on the surface of the sand,
partly submerged in water, and six days later gave rise to an adult
female. By October 18 several more had completed their life cycles,
and on that date, as King was proceeding to England on leave, the
remaining ones were killed and preserved.

All those that pupated did so on the surface of the sand, some high
and dry, others half in and half out of the water. "Probably," says
King, "under more natural conditions, the pupal stage would be
passed buried in the soil — the structure of the pupal case seems to
indicate this."

The average pupal period was from six to eight days.

The following are practically King's descriptions of larva and pupa
(that of the egg having been quoted previously) of Tabanus par.

The mature larva (Plate 3, Fig. 48) when fully extended measures about 13.5
mm. Color white with a grayish tinge. Mandibles dark brown to black, ser-
rated. On the anterior third of each abdominal segment except the eighth is a
ring of pseudopods, eight in each ring — two dorsal, two lateral, and four ventral —
except on the first abdominal segment, where the two dorsal ones are wanting.
On the second abdominal segment the two dorsal pseudopods are very small.
The pseudopods are largest on the third, fourth, and fifth abdominal segments,
and are always more developed on the ventral than on the dorsal surface. Each
pseudopod bears a crown of colorless spines or hooks, and there are patches of
dark spines between the pseudopods. The spines on the dorsal sections of the
rings on the first and second abdominal segments are dark. The anus is situated
ventrally, at the base of the eighth segment, and is fringed with blackish hairs.
The syphon tube consists of two segments, and, when exserted is as long as the
eighth abdominal segment.

The pupa (Plate 12, Fig. 142, a, b) is from 12 mm. to 15 mm. in length and
at first yellowish white in color, becoming darker as it nears maturity. The eyes
show plainly through the pupal case as dark greenish purple. The empty pupal
case is yellowish brown, the thoracic tubercles and the spiracles being darker
than the surrounding parts. On the apical third of the second abdominal seg-
ment is a fine ring of backwardly pointing spines. Similar but broader rings,
bearing longer and stronger spines, are on the third, fourth, fifth, and sixth
abdominal segments, and one of intermediate breadth on the seventh. The
eighth abdominal segment terminates in a coronet of six teeth, in color shining
brown, becoming darker at the tip. The dorsal pair are smallest and close to-
gether, the ventral pair next in size and wider apart, and the lateral pair longest
and arising from almost the same level as the dorsal pair. Ventrally placed to
this coronet of teeth are two rows of small teeth, from two to four in each row,

132 THE EARLY STAGES OF TABANID^

together forming an interrupted transverse row. These teeth are of unequal size
and vary in their relative size in different specimens.

Tabanus pertinens Austen. — ^An African species, according to Neave
usually confined to comparatively low-lying country where the river
beds are of a sandy nature. In southern Nyasalalid it was found com-
mon on the Mwanza River, in the Shire Valley, as early as the end of
July, but did not occur on Mt. Mlanje, though a few specimens were
taken by Neave at some distance from the mountain.

A pigmented larva (Plate 3, Fig. 54), which it was thought might
belong to this species, was taken in some numbers in the Shire River
in August, and in the Ruo River in October. It was found in both
cases in water amongst the roots of grasses or water plants, and
seems to prefer rivers with a sandy bottom and banks. The striking
larva is remarkable for the development of the dorsal prolegs, which
perhaps are associated with its comparatively free-swimming exist-
ence. The prolegs are also present immediately anterior to the anus.

Tabanus quatuornotatus {quadrinotatus) Meigen. — ^Common in
Europe. This was the first species of tabanid in which the eggs and
the act of oviposition became known. The credit of the discovery is
due to Joseph Mann, Curator of the Imperial and Royal Zoological
Cabinet in Vienna, who, according to KoUar's report in 1854, during
a naturalist's trip to Carniola, Austria, in May and June, 1854, suc-
ceeded in observing in a damp meadow at Wippach the female of
Tabanus quatuornotatus in the act of oviposition. Mann's own words,
quoted by Kollar, are translated as follows:

On June 25 at 11 a.m., I found a Tabanus just beginning to deposit eggs on a
grass blade; as it did not attempt to escape, I cut off the blade and took it home.
Towards 2 p.m. the Tabanus flew to the window; I at once looked at the grass
blade and found on it a cluster of eggs completed. These were wax-yellow at
first, later on they took on a grayish color, and two days afterwards they
appeared almost black.

Mann later on found, in the field and also on other plants, several
egg masses similar to the first, bringing them all to Vienna.

By this discovery it was known for the first time that the Tabanus
does not deposit the eggs on the ground where, according to Degeer,
the larva lived, but on plants; further, that the Tabanus does not

WERNER MARCIL\ND 133

scatter its brood as many other species of flies do, but deposits the
eggs in a single mass.

The eggs were counted, and their number found to be about 350
to 400. Also the hatching of young larvae was observed, and the
duration of the egg stage found to be from ten to twelve days (Kollar) .

From the egg masses collected by Mann, besides young Tahanus
larvae, also ichneumonid flies (parasitic Hymenoptera) were observed
to hatch; to Kollar and Mann consequently credit is to be given for
the first discovery of hymenopterous egg parasites in tabanids,
though no description is given (see under Parasites of the early
stages of Tabanidae, page 182).

The publication of Kollar has been quoted in detail by Lecaillon, to
whom we owe further studies on the early stages of Tabanus quat-
uornotatus, chiefly on the oviposition of this species. Lecaillon
found the species, with a related form, Tahanus hromms, abundantly
at Gouy, Aisne, France, in May, 1904. Oviposition of Tahanus
quatuornotatus was observed on May 24, 1904. In the afternoon of
a sunny day, on a wooded and not especially damp hillside, a female
was found sitting immobile on a dry branch of a weed {Origanum),
the head turned downwards, at a distance of 35 to 40 cm. from the
ground. The eggs already laid formed a conspicuous white mass.
On coming nearer, Lecaillon ascertained that the Tahanus, which
under every other circumstance is likely to take flight, was not dis-
turbed and continued to oviposit. Lecaillon broke off the branch
with the insect on it, placing the whole in a Jar. The female con-
tinued to oviposit for about ten minutes, when it ceased and left
the branch. The egg has been figured by Lecaillon.

Lecaillon calls attention to the fact that, as already stated by
Kollar, the ovipositing female is somehow indifferent to what is
going on around her, a fact which should be taken into account in
understanding the habits of adult tabanid flies.

Concerning the early stages, we learn from Lecaillon's observations
on this species that oviposition in some tabanids may take place in
comparatively dry and not necessarily in a damp environment. In
both Mann's and Hart's observations it took place in a moist environ-
ment. As among insects frequently not only the necessities of em-
bryonal but also of larval life may be anticipated from the manner

134 THE EARLY STAGES OF TABANID^

and conditions of oviposition, Lecaillon thinks that as in tabanids
oviposition may take place, according to species, either in a dry or
in a damplocahty, "not only the embryo but also the larva" can live
under very variable conditions of humidity. Lecaillon considers that
observations on the larval life of Tabanus qualuornotatus verify this
hypothesis.

Later, in 1905, again large numbers of egg masses of this species
were found by Lecaillon, and his first observations could be generalized
to some extent. The eggs were always laid on warm sunny days
in early June about the middle of the day, and almost without ex-
ception on dried out twigs of various herbaceous plants. In fact, of
sixty egg masses collected in 1905, not a single one was found on a
green leaf or stem. During the subsequent years, few exceptions to
this rule were noticed. In 1907, however, Lecaillon found an egg
cluster fastened to a green stem of a grass, and in 1908 a cluster fas-
tened to the dry branch of a tree (bouleau) which had been rammed
into the ground. Lecaillon discusses (1906) this fact, putting the
question whether we might be dealing with mimetic resemblance.
This he thinks hardly probable. The females, avoiding the green
objects, would alight only on twigs of grayish or blackish color. More
likely it seems to Lecaillon that the females oviposit on dry twigs
because these are more rigid than the green stems. In fact, it is
observed that during the act of oviposition, the stems of weeds on
which the insects alight are curved, especially when agitated by the
wind. On the dried out twigs the Tabanus is in a much more stable
situation and can with greater facility give to the egg mass its rather
complicated form.^^

The places where eggs had been deposited were in all cases exam-
ined on woody slopes distant from any water courses. This fact
alone is certainly sufficient to prove that the larvae of Tabanus
qualuornotatus are not aquatic but terrestrial.

The egg masses are found in large numbers assembled in certain
places on hillsides which evidently present all the favorable conditions
to which the ovipositing females are adapted. This circumstance is

^^Lecaillon's explanation is more than doubtful. Many species of tabanids
complete their egg masses on quite slender reeds and grasses. Also for Tabanus
quatuornotatus it remains doubtful by what stimuli it is attracted to the twig.

WERNER MARCHAND 135

favorable to the destruction of the eggs in the case of very dangerous
species which should be brought under control. The egg masses are
in fact very conspicuous and may be collected easily and quickly.
All egg masses observed were placed about 30 to 50 cm. above the
ground.

The latest date on which oviposition was observed was June 14;
however, it may be that it takes place even later, as Lecaillon found
eggs containing larvs as late as July 22 (1907). The period of re-
production extends in this species, at least in certain years, from the
end of May to the end of June. The period given in Lecaillon's
note in 1906, of two to three weeks, was evidently too short.

A detailed description of the egg mass is given (Plate 2, Figs. 33
to 38). The egg mass presents the aspect of a roughly subconical
body which under natural conditions is placed with the base below and
the vertex above. The axis of the cone is arranged about parallel to
the branch or stem which serves as support, and is placed in such a
way as to cut lengthwise through the conical surface and to be partly
enveloped by it. However, it would not be accurate to speak of the
egg mass as subconical, as certain authors (Brauer and Hart) have
done. The body of the mass is distinctly bilaterally symmetrical,
as shown in the accompanying figures. The plane of bilateral sym-
metry is determined by the stem which serves as a support and by a
crested longitudinal line placed on the side opposite the egg mass
(Plate 2, Figs. 34 and 35). On the other hand, it is possible, with
regard to the mass, to distinguish a vertical direction or orien-
tation, a horizontal anteroposterior one, following the plane of sym-
metry, and a horizontal lateral one, placed at right angles to this
plane. In the anteroposterior direction the base of the mass has its
greatest dimension (5 mm. in a specimen examined). In the two
other directions the dimensions are about the same and somewhat
smaller than in the preceding one. Posteriorly, the surface adhering
to the support is much larger than the opposite surface anteriorly.
Finally, in the specimens examined, the base of the mass was not
flat but perceptibly and progressively excavated from the periphery
towards the center.

The egg mass is composed of eggs which are placed quite regularly
on top of each other, if observed vertically, or placed in successive

136 THE EARLY STAGES OF TABANID^

horizontal layers, if observed horizontally. The eggs are, moreover,
glued together by a substance which hardens after deposition, and
causes the eggs to adhere strongly to each other. It goes without
saying that the number of eggs contained in a horizontal layer dimin-
ishes if we proceed from the base of the mass towards its summit.

The color of the egg mass, which is white at the moment of deposi-
tion, is changed rapidly into brown and later into black. The change
of coloration begins some time after oviposition is completed; it ap-
pears at first at the summit of the mass, thence it spreads towards its
base, which after a few hours has turned almost completely dark
brown or black. This change of coloration in the egg masses of taba-
nids has already been observed, as Lecaillon emphasizes, by Mann
and also by Hart. However, Mann, who observed the same species
as that studied by Lecaillon, speaks of a wax-yellow color which later
passes into blackish brown, and Lecaillon assumes that Mann has not
noticed the primitive white color of the eggs when first laid, or that he
possibly dealt with a different species. Hart's statement that in the
eggs of Chrysops moerens {cestuans) the first color is cream, seems bet-
ter to agree with Lecaillon' s observation. This may possibly be
merely an inexactness of terms, as one author may call an object
white which another would call pale yellow or cream.

The change of coloration which takes place in the egg mass a
short time after oviposition must be considered as advantageous, as
we have to deal with the substitution of a protective color (black)
for a white color which as actually observed, renders the eggs very
conspicuous to the eye.

According to Lecaillon, the color of insect eggs may be due either
to a coloration of the yolk, or to a coloration of the egg envelopes.
Changes of coloration can consequently be due to changes taking place
inside the egg, or to changes taking place in the envelopes. In the
case of Tahanus quatuornotatus , the substitution of the white color
by a dark color is due to a brown pigment which develops in the
chorion after oviposition, probably by the slow effect of air and
light. This pigment is very strongly developed and hides com-
pletely the contents of the egg. Consequently even at the end of
embryonic development when the eggs contained wholly white
larvae, they still are in external appearance completely black.

WERNER MARCHAND 137

The pigment is, according to Lecaillon, not altered by alcohol,
while in eggs in which the color is due to the yolk the latter is usually
decolorized by the action of alcohol.

The female constructs its egg mass, depositing the eggs one after
the other, beginning at the anterior end. At last, when the mass
has reached a certain width the eggs are found arranged in horizontal
and more or less regular layers, while the lower border of the mass
forms a pronounced projection, the beginning of the depression of
the lower surface. While the female is sitting with head turned
downwards the tip of the abdomen is placed upwards, towards the
point where the egg is to be laid.

The eggs themselves are of curved shape, they measure about 2.5
mm. in length and only 0.5 mm. in width. Their convex side, which
is chiefly visible in those which lie at the periphery of the egg mass,
is always directed towards the outer side of the latter.

Lecaillon calls attention to the significant regularity of tabanid egg
masses. Among the greater part of insects which deposit their eggs
freely on leaves, twigs, walls, etc., the egg mass consists generally of
an agglomeration of any shape which it may assume under the con-
ditions presented by the object on which the eggs are laid. In
Tahanus quatuornotatus , however, the case is different; the egg mass
has a very complicated structure, which to a great extent is inde-
pendent of the shape of the supporting object, and of a strictly bi-
lateral symmetry which can certainly be regarded as a characteristic
of evolutionary perfection. From Hart's figures it is seen that the
same character is found also in the egg mass of Tahanus atratus and
Chrysops ccstuans (mcerens); and it is probably generally found
among tabanids.

In a later paper, in 1911, Lecaillon reports additional observations
made on this species in 1906, 1907, and 1908, on the conditions un-
der which the eggs are laid, and a more amplified description of the
egg masses themselves.

Since the first publication, a considerable number of egg masses of
Tahanus quatuornotatus were examined and a number of individual dif-
ferences were found in the arrangement of the egg masses. Some of
the more interesting forms met with were figured in the plate added
to the second publication. The egg mass previously described is

138 THE EARLY STAGES OF TABANIDiE

consequently incomplete; moreover, they are found frequently in
places where this Tabanus oviposits. But in many cases the egg
masses have been found to be much larger than the one described in
1905. In cases where the egg mass, fastened to a twig, reaches its
maximum of development, it is much more prolonged vertically
than the one figured in Lecaillon's first report, and it possesses a
second plane of symmetry which is horizontal and at right angles to
the principal plane of symmetry passing through the stem which
serves as support and through the anterior tip which presents the egg
mass. But only rarely will the egg mass be found entirely com-
pleted in this manner, and all the intermediate forms are found
between that described in the first publication and that described
later, with two planes of symmetry.

Frequently the egg masses are found grouped together on the same
stem, numbering two, three, or four (Plate 2, Figs. 35 to 38).

Finally, as stated also in Lecaillon's short note of 1906, two egg
masses are frequently not only contiguous but actually form the
continuation of each other (Plate 2, Fig. 3?>). Of these joint egg
masses Lecaillon is uncertain whether they are produced by one
female or by two different females. Probably the latter, inasmuch as
in the figures given their size is considerable ; also from our knowledge
of the egg-laying instinct of the female, it does not discern between
the eggs laid by itself and eggs laid by another; consequently it will
be likely to continue the process of oviposition started by another
female in case it, by chance, had alighted on a stem where already
other females were laying. As during the process of laying the
single egg is always placed beside another previously laid, one would
expect that in starting an egg mass, preference would be given to a
place where eggs are already present instead of places where only a
smooth surface is given.

Sometimes the herbaceous stem which supports the egg mass is
very fine, and in these instances the latter surrounds it almost com-
pletely, even from behind, following the median line of the stem.
In other cases the egg mass is attached to the top of the spike of a
grass plant; in this instance it envelops only a small part of the
object, so that from behind the egg mass appears as of considerable
width.

WERNER MARCHAND 139

Besides, the egg masses are never found of a strictly geometrical
form, and numerous irregularities in details of their structure are
always observed. In general, however, it can be said, that the eggs
are placed horizontally and in rows which when observed horizontally
or vertically always present a more or less regular arrangement.

On the larval stage of this species we possess the results of Lecail-
lon's first studies (1905), amphfied by later additions. At the time
of Lecaillon's studies knowledge about tabanid larvae was still very
meager; he believes that the larvae are predacious, and there are
certain species which are aquatic and others in which the larva lives
in the ground or in moist soil.

Lecaillon made an attempt to raise the larvae of Tahanus guat-
uornotatus from the egg, but did not succeed completely, as many of
the larvae died in spite of all precautions; but he was able to keep
them alive for several months and some of them were still living when
the paper was pubhshed (December 15, 1905), having survived for
more than six months. Lecaillon made some observations on the
hatching of the larvae.

According to Mann, the duration of embryonic development in this
species is ten to twelve days. Lecaillon placed eggs laid May 24 in a
moderately damp atmosphere (the stem with the eggs was simply
placed in a crystallizing dish), and observed that on June 6, thirteen
days after oviposition, no hatching had taken place. Taking two
eggs out of the mass, he saw, as a result of the slight pressure or
traction to which they had been submitted, the hatching of com-
pletely developed larvae.

On the following day the egg mass was vigorously rubbed, and
as a result all the larvae hatched at once, showing that embryonic
development had been completed in all the eggs. The difference in
time appearing in Lecaillon's observation as compared with that of
Mann is said to have no significance as the duration of embryonic
development is very variable, in a given species, according to the
temperature in which the eggs are kept.

The medium duration of embryonic development was, according to
Lecaillon's first publication, from twelve to thirteen days. But
later he found that the larvae generally do not hatch as soon as they
are completely developed; they may remain inside the egg shells for

140 THE EARLY STAGES OF TABANID^

a period which seems to be rather long. In 1907, Lecaillon found
that eggs which were laid on June 11 had not yet hatched on June 8.
Eggs collected on July 1 1 , which had certainly been laid a long time,
produced larvae on July 22. Embryonic development is, however,
completely terminated at the end of twelve or thirteen days, be-
cause, after that period, it is sufficient to rub the egg mass only
slightly in order to cause the larvae to hatch.

It can be assumed, consequently, that the larvae do not leave the egg
shells in which they have developed, until they need food; from the
time of their formation to that of hatching, they probably find an
efficient protection inside the egg shells. Perhaps also the egg shell
which has to be pierced by the larvae in the act of hatching is of a vari-
able resistance according to the conditions of the environment. In
a damp environment, for instance, the shell would be less resistant
than in a dry atmosphere, and the larvae could hatch much more
easily in the first than in the second instance. The larva appears in
this case to be adapted to remaining enclosed in the chorion of the
egg for a time long enough to wait for favorable atmospheric condi-
tions. Lecaillon has not made exact studies to determine this point.

When the larvae hatch, they remain sometimes for a few moments
on the surface of the egg mass. They carry out very varied move-
ments of contraction and finally fall to the ground into which they
burrow immediately. Often they are also attached to one another,
forming sort of batches which finally lose hold on the egg mass and
fall to the ground. They are, as Lecaillon says in 1906, extremely luci-
fugous (negative phototropic), and they could not be easily destroyed.

Immediately after hatching, the larvae are white in color and this
color is retained. The body is also somewhat transparent, and the
arrangement of different internal organs can be observed with facility.
It is seen that in the middle intestine there remains a rather large
quantity of nutritive yolk, a remainder of the primitive egg contents.
At the level of the yolk the intestinal tract is opaque. When placed
in damp earth, the young larvae appear not to take food for several
days.

In order to ascertain whether these larvae were carnivorous, Le-
caillon supplied them with ant nymphae freed from their cocoons,
with flies killed previously, and with larvae of Chironomus. He

WERNER MARCHAND 141

found that they did not attack these with great alacrity, but seemed
rather to avoid their contact when placed near them. However,
after having pierced Chironomus larvae by means of a pointed scalpel,
Lecaillon observed some of the larvae to suck up the blood, their in-
testinal tract assuming a characteristic red color; he even saw them
penetrate into the body of the animal in order to devour it. On the
other hand, it was observed that the larvae placed in moist earth ab-
sorbed the organic detritus, giving their intestine a distinct dark
color. These facts led Lecaillon to assume that the larvae of Tabanus
quatuornotatus can feed, according to circumstances, on animal or
vegetable matter in process of decomposition, and probably also on
certain soft larvae selected by the experimenter.

Lecaillon goes on to investigate the conditions of humidity favor-
able for the larvae. Generally he left them in very moist earth. But
once, having placed within the crystallizing dish, which contained the
larvae, another dish the under side of which was wet and rested on the
former, from which the earth had been partly removed, Lecaillon found
that fifteen larvae had assembled under the smaller dish, between the
two glass surfaces, and were completely submerged in water. He
subsequently placed a number of larvae in a cup filled with water,
and observed that they appeared to be at home in the water, remain-
ing in it and making no attempt to leave it. Later, Lecaillon placed
the larvae which he intended to rear in mud taken from an aquarium ;
they remained completely burrowed in this mud and could even find
nourishment in it.

On the other hand, Lecaillon repeatedly allowed the earth in which
larvae were kept to dry out completely and found that the larvae not
only did not suffer but remained quite active.

His conclusion was that the larvae of Tabanus quatuornotatus are
capable of adapting themselves to a variable degree of humidity, and
can live, at least for some time, in dry ground as well as in water.
Concerning the range of humidity, and also concerning food habits,
these larvae, as probably other tabanid larvae also, appear not to be
adapted to limited conditions, in contrast to many other insects.

In his later publication, Lecaillon emphasizes the great resistance
of the larvae of Tabanus quatuornotatus to humidity and draught, ex-
plaining the fact by the presence on the body surface of a chitinous,

142 THE EARLY STAGES OF TABANID^

thick, and impermeable integument. In acetic sublimate, at a tem-
perature of 24°, the larvae remain alive for more than half an hour.
It may be mentioned that many dipterous larvae behave in this rela-
tion exactly like the larvee of tabanids.

Lecaillon has again tried to rear larvae of this species which had
hatched on June 27, 1905. He found that they succeed best in moist
ground. In February, 1906, a rather large number of these larvae
were still alive; they had grown comparatively little, measuring only
7 or 8 mm. in length, against 2 or 3 mm. at the time of their birth.
Some lived for a year without having reached a much greater size
than this. Some of the larvae were still found alive in July, 1906.

Under the conditions mentioned, the intestinal tract of the larvae
almost always contains particles of soil, giving them a blackish color.

From these observations it is probable that in Tahanus quatuorno-
tatiis the duration of the larval stage is more than one year. However,
Lecaillon believes that this need not necessarily be, as it is apparent
that the larvae when kept in moist earth do not find exactly the con-
ditions under which they live in nature. Their growth is consequently
much diminished and cannot be compared with normal growth.
This subject, however, requires further investigation.

The chitinous cuticle of the larva was also studied by Lecaillon
(1906). It is thick and resistant, as genierally in dipterous larvae.
Its thickness in a larva of 630 ix in diameter being about 17 ^t. This
resistant and impermeable covering protects the larva efficiently.
It may be left in the water, for instance, for several hours without
being appreciably injured. It can be left, on the other hand, in dried
soil for a considerable time without being killed. If one wishes to
fix the larvae in toto, even in the case of very small larvae, great diffi-
culties are encountered because of the impermeability of the chitinous
membrane.

The free surface of the chitinous layer appears to the unaided eye
to be smooth and devoid of hairs. But with a lens it is found that it
presents numerous longitudinal parallel ridges and numerous hairs of
reduced length. This characteristic has been found to be a general
feature of tabanid larvae, and the examination of the striae can fur-
nish the means to distinguish the genera and species.

WERNER MARCHAND 143

The finer structure of the chitinous layer is easily studied on cross
sections of the larva which is fixed and stained before. On section
the chitinous layer is found to consist of three zones; first, an inner
zone, which is in immediate contact with the hypothelium; second, a
median zone, resting on the former; third, an external zone, which
limits the body of the larva exteriorly. These three zones present
different characteristics.

1. The inner zone is highly developed; it is much thicker than the
two other layers taken together. It is of a lamellar structure, as on
cross-section it presents numerous concentric striae. This inner zone
is slightly affected by acid stains.

2. The median zone is much less thick than the inner, but much
thicker than the external zone. This zone or layer forms the thick-
enings which project exteriorly in the form of longitudinal ridges on
the surface of the integument. It follows that it is much thinner
at the level of the intercostal furrows than at the level of the ridges.
It appears not to be lamellar in structure, but rather to present at
the level of the ridges vertical striae. Finally, its most remarkable
property is its great affinity for basic stains.

3. The external zone is extremely thin; unless high magnifications
are employed to observe it, it easily escapes detection. It covers the
median zone in the intercostal furrows as well as on the ridges, and
is not thicker at their level. It is not affected by either basic or acid
stains. This layer forms the hairs. Neither the median nor the
internal zone takes part in the formation of hairs.

It follows from the difference in affinity with regard to stains,
shown by the two first layers, and the inability to stain the outer
layers, that the three chitinous layers are easily distinguished by a
combined staining method. By means of magenta red and picric
indigo carmine, for instance, the inner zone becomes greenish and the
median zone dark red, while the external layer remains colorless.
Similarly, by a combination of hematoxylin and light greeti, the inner
layer stains green and the median layer deep black, the external
layer always remaining colorless.

The descriptions of the chitinous integuments of insects do not
completely harmonize with those given by Lecaillon of Tabanus
quatuornotatus. Lecaillon assumes two reasons for this fact: first,

144 THE EARLY STAGES OF TABANTD^

the staining methods employed have undoubtedly been insufficient;
second, the different types of insect may show important differences
in the integument. If the integument is not very thick, it may
often be of simpler structure. But only exact observations can de-
termine this point. In the meantime, Lecaillon points out that the
structure of the integument of this larva suggests strongly that de-
scribed by Dubosq in the myriapods.

Lecaillon describes also the attachment of the muscles on the chit-
inous membrane of the larva of this species. The muscular fibers in-
serted in the integument do not end at the epithelium but cross the
chitinous layer, becoming attached at the median zone. But the stri-
atlon of their fibers disappeared where they reach the integument,
and the whole section contained between the hypothelium and the
median chitinous zone is formed by non-striated fibrils forming a
sort of a tendon. All the muscles attached to the integument present
this type of insertion. Lecaillon points out that the middle and
external layers form the really hard and resistant part of the chitin-
ous membrane of the integument, as Dubosq has shown in the myria-
pods. This aids in the understanding of the mode of insertion of the
muscle fibers in the highly contractile larvae of tabanids, in this
region of greatest resistance. About Graber's organ, in this species,
see page 35 and Plate 10, Figs. Ill to 116.

According to Bezzi, the Tabanus larva in which Paoli studied
Graber's organ, may have belonged to this species (Paoli). See
p. 37 (Special anatomy of tabanid larvae).

Tabanus semisordidus Walker. — The egg masses of this South Ameri-
can species have been observed in British Guiana, by Bodkin and
Cleare, in 1916, to be deposited on the leaves of aquatic grasses and
in some instances on the leaves of young rice plants. They are laid
in a little bundle consisting of twenty or more cigar-shaped, shining
black eggs adhering to one another and to the leaf surface.

Tabanus solstitialis Schiner. — A common European species.
Brauer (1880) states that the pupa is found in the water, being green
in color and provided with large ear-shaped anterior spiracles. The
larvae are not mentioned.

WERNER MARCHAND 145

Tabanus speciosus Ricardo. — In Madras, India, but rather rare;
feeding on cattle with Tabanus striatus and albimedius, and similar in
size and appearance to the latter.

According to Patton and Cragg, this species lays its eggs always
on the leaves of water-lilies growing in deep water. The number of
eggs laid is between 500 and 600. The process of oviposition is said
to be similar to that of Tabanus albimedius, but Tabanus speciosus, in-
stead of forming a V-shaped mass as is usually the case with the
larger tabanids, lays its eggs in a round heap, which it then plasters
over with a chalk-like substance, almost completely covering the
eggs. The egg mass is figured (Plate 1, Fig. 19),

The larvce of the larger Indian tabanids, including this species, are
powerful swimmers and have air sacs connected with their tracheal
tubes, so that they can float or sink at will. A description of the
larva is not given.

Tabanus spodopterus Meigen. — A common European species. Ac-
cording to Schiner, very common in Austria, at the Neusiedler See.
Brauer (1883) gives an illustration of the larva (Plate 6, Fig. 87, a-f),
which he found, in the month of May, under dry leaves on the
ground.

Tabanus striatus Fabricius. — The early stages of this species have
become known through Mitzmain's work, which is the most thorough
investigation into the life history of a tabanid species which we pos-
sess. We cannot enter here upon his results in as far as the habits
of the adult fly and its part in the transmission of disease is con-
cerned, but have to limit ourselves to report on the results obtained
on the early stages of this species.

The eggs of Tabanus striatus (Plate 2, Figs. 22 to 27) were not found
in the field, but Mitzmain was able to obtain them from females
ovipositing in captivity. A large case was built for this purpose
provided with a tank of water, growing plants, and two animals
(carabaos) as a source of food for the flies. In a short time females
were observed feeding on the hosts and several were found ovipositing
in various places about the enclosure.

The time selected for egg laying under the conditions provided was
invariably during the early afternoon, never later than 2 o'clock.
This was observed in nearly fifty instances.

146 THE EARLY STAGES OF TABANID^

The eggs were laid in a compact mass either extended on a flat
surface or surrounding various attached objects, usually of small
diameter, such as projecting splinters of wood, suspended fibers of
jute sacking, fine brass wire, a single animal hair, and coarse iron
wire. Upon these materials the eggs were laid in an ellipsoidal
form sometimes surrounding the objects completely or nearly so.
On one occasion two egg masses were found upon a small splinter of
wood which they entirely enveloped. The surfaces of the egg masses
were continuous, so that the double mass resembled a single large
one. When eggs were found deposited on a flat surface, on two occa-
sions a leaf was the object selected. These were leaves of an orna-
mental plant which was used for shade purposes in the breeding cage.
The plant grew close to the cement water tank in the breedmg cage.
In all other instances the eggs were deposited upon woodwork on the
sides and ceihng of the cage, invariably upon the shaded portions, as
the under side of beams and partitions. In egg laying upon flat sur-
faces there was a strikingly constant geometric form. Usually the
form assumed was roughly a pentagon with a biconvex center.

At the beginning of oviposition (Plate 2, Fig. 24) usually two eggs
are deposited in the position of an inverted V. Three to four eggs are
then laid on either side of the apex of this V, and then one side and
then the other is built up, rather irregularly at first, until the sides
of the pentagon are completed. The eggs are laid cleanly and defi-
nitely, each line slightly overlapping the preceding one. When the
eggs are laid in the extended order, they are deposited three or four
layers in depth, but usually as many as six layers are required to
complete the mass when the eggs surround a convex object.

In the process of laying, "the body is held away from the egg mass,
the legs being planted firmly." When the eggs are attached to one
of the objects mentioned above, the insect stands with head down-
wards, the fore legs suspended alongside the head, the hind and middle
legs supporting the weight of the body. At the first movement, the
anal end of the body is bent towards the thorax under the abdomen,
and with a Siight jerk the egg is laid, while the brush-like appendage
of the ovipositor exudes a tiny drop of liquid coating the egg as it is
deposited. The movement of deposition is very much like squeezing
a bit of pasty material from a collapsible metal tube.

WERNER MARCHAND 147

In several counts that were made by Mitzmain, the fly was observed
to lay with clock-like precision at the rate of ten eggs per minute.
This did not vary whether the attached object was above or below
the fly. In three instances observed, the process occupied from forty
to forty-five minutes. Both the beginning and completion of the
egg-laying process were without deliberation, the insect walking
away from the mass of eggs and flying off as soon as the last egg was
deposited.

When disturbed during oviposition, the insect does not fly and
can readily be carried without attempting to escape. While in the
act of laying, if interrupted and dislodged from the position, it imme-
diately begins to deposit a new egg mass. This was twice repeated
with one female, and three distinct egg masses were deposited, all
identical in geometrical arrangement.

The eggs of this species of Tahanus are laid with very little cement-
ing material. The cement used is a transparent substance "and not
dark and opaque as found to exist in the species described by Hine"
{Tahanus lasio phthalmus) . The cement provided by this species
was tested and found to be water-proof, as well as insoluble in various
grades of alcohol and xylene.

The eggs (Plate 2, Fig. 25) when laid are a pale clay-yellow, but
within twenty-four hours become slightly darkened with an ashy
gray tinge. Microscopically fine black striations can be seen running
lengthwise for nearly 0.5 mm. from the end opposite the micropyle.

The shape of the individual egg is that of the muscid t)^e with
more sharply pointed ends; it is not quite spindle-shaped. Several
eggs were measured and found to average in size 1.6 by 0.4 mm.
The size of the mass varies from 9 to 12 mm. in length by 6 to 9-
mm. in breadth.

The number of eggs laid in a mass varied greatly. In four masses
counted there were respectively 270, 340, 417, and 425. Ten masses
dissected from the bodies of killed flies were found to average 405;
the greatest number found in any female was 495.

Mitzman gives a detailed description of the hatching process.
Two egg masses were observed microscopically during the entire
process of hatching, and fourteen egg masses were noted as to the
length of the incubation period. The minimum period observed

148 THE EARLY STAGES OF TABANID^

was three days and the maximum five days. Four days is probably
the average length of time required for incubation. It was ob-
served that the degree oi temperature and moisture influenced the
time of hatching. Slight changes in either of these factors can be
used to control the time of emergence from the egg. Mitzmain's
observations on the hatching of an egg mass are the following:

"Twenty-two hours previous to the hatching of the embryo, certain unmistak-
able activities were discernible in the egg. The first signs of these were seen in
the two eggs which formed the nucleus for the egg mass and which are the first
eggs laid. These movements, as indicated by either of the dark eye spots, could
be seen with a hand lens at intervals of a few seconds; their action was similar
so that of the bubble in a spirit level. In about an hour the movement was
teen to be rather general in the egg mass, accompanied in the eggs first laid by
an alternate collapsing and distending of the exochorion. This action is the
result of the torpedo-like movement of the head capsule of the embryo in the di-
rection of the micropyle of the egg. The movement is effected by the piston-like
action of the apophyses of the cephalopharynx, which appear to work alternately,
bringing the saw-toothed mandibles in contact with the micropyle canal. These
movements proceeded uninterruptedly during the hours of the night, the only
change observable being that the body segments of the embryonic larva became
better defined. At 4.25 the next morning the segments of the embryo could
easily be counted through the chorion. The dorsal surface of the exochorion
was seen to be sHghtly shrivelled."

"Fully one hour and thirty minutes intervened during which there was no
action worth noting. This quiescence was interrupted by a sudden remarkable
activity of all of the visible eggs of the mass. At 6.08 there was a general up-
heaval of the surface of the egg mass, an agitation within the eggs, and an alter-
nate collapsing and distending of the eggshells. At 6.10 the first layer of eggs
gave birth to a silvery horde of young larvae, which at 6.12 had crawled from
view [Plate 2, Fig. 27]. Then ensued another spasmodic agitation giving birth
to another lot of larvae, which crawled from the mass of empty eggshells [Plate 2,
Fig. 26]. The emergence which is effected by the head structures is aided by the
posterior protuberances, which functioning as prolegs push the body of the larva
clear of the eggshells."

Mitzmain's discussion of the egg stage and of the hatching process
is followed by a description of the morphology and habits of the
young larvae.

Immediately after emerging from the egg, the young larvae (Plate
5, Fig. 80, a, b) seek concealment. In nature, no doubt, resort would
be had to the convenient water course where aquatic plants, drift-

WERNER MARCHAND 149

wood, and stones would be the probable hiding places. The larvae
under observation became very active and crawled out of the slender
dish, a height of 9 cm., and fell into the water of the basin pro-
vided. When collected and placed in a deep glass vessel with some
water, the entire mass took refuge behind the filter paper in the glass.
There they crowded side by side with their syphons projecting from
the upper edge of the paper. When disturbed and forced to take to
the water, they were found in thirty min\ites reassembled in the char-
acteristic gregarious fashion behind the filter paper against the
glass.

For convenience in study, a majority of the larvae were transferred
when 1 day old to individual glass jars one-third filled with clean wet
sand from the lake shore and provided with strips of filter paper
soaked in muck from the creek bottom. The jars, which were the
common half pint jelly glasses recommended by Hine, were kept
covered with filter paper, held in place by the tin lid which had a disk
cut from its top to admit air. By renewing the moisture on the
strip of filter paper in the jars, the filter paper cover serves ideally
to control the humidity.

A considerable number of the larvae were not separated, but were left
together for observation in a glass dish with a few strips of paper
saturated with muck from the creek. The young larva is briefly
described as follows:

"The larva one hour after hatching [Plate 5, Fig. 80, a, b] is 1.5 mm. in length.
The following day several were found to measure 1.8 mm. The general color
is a dirty white with a tracheal system of waxy white, the abdominal contents
pale green, and the Malpighian tubules of a lilac color. There are 2 black
eye-spots located midway on the head capsule. The latter tapers to a sharp-
pointed mouth with a prominent pair of great hooks or mandibles. The seg-
ments are provided with typical, conical, truncated, prolegs, each armed with a
chaplet of medium long, brown hairs. The siphon which is carrot-shaped at
this stage is a prominent feature."

Food in a variety of forms was furnished the larvae. They thrived
from the start on minute crustaceans, larvae of Stomoxys, mosquito
larvae, and young angleworms. Full grown angleworms were found
unsuitable, and larvae of the blow-fly and flesh-fly were not satisfac-
tory unless killed previously, as they were capable of killing or in-

150 THE EARLY STAGES OF TABANID^

juring even well developed Tahanus larvae. As soon as the insect
becomes aware of the presence of food, the claw-like mandibles are
protruded from the head capsule, and bury themselves in the live
food like meat hooks. With a slight curve dorsally, the larva's body
is brought forward, and a small portion of the food is lacerated. This
is aided by a twisting of the head and a pulling with the extended
jaws. The mandibles are brought together with a rapid clawing
action, the parts working in apposition. When prehension is effected,
the jaws move alternately upwards and downwards and laterally,
and the bolus is swallowed in fibrous strands.

Seeking and devouring food is not a continuous operation as it is
in the case with Stomoxys and the dung-flies. The Tahanus larva
requires a long rest after a sufficient meal is taken. A 2 day old
Tahanus is capable of devouring two half grown larvae of Stomoxys
in twenty-five minutes. In one instance a full grown Stomoxys larva
was destroyed in exactly twenty minutes. Here the attack on the
Stomoxys was made through accidental collision, the Tahanus in-
stinctively thrusting out its mouth and tentatively taking a bite.
It apparently became greatly excited (this was its first meal), and
thrusting its head into the body of the Stomoxys larva, commenced
to probe by twisting its head rapidly. In less than a minute the
cuticle was broken through and an ample slit was made through
which the entire head was buried in the body of the victim, where-
upon an energetic gouging took place. The Tahanus worked through
the cephalic third of the body upwards to the head, then worked in
the other direction on the lower two-thirds. This gouging was con-
tinued until the Stomoxys had become completely eviscerated, during
which time the head of the Tahanus kept steadily probing, twisting
its pharynx from side to side, and pushing forward with its rostrum
until the Stomoxys larva was completely devoured with the exception
of the cuticle.

The full grown Tahanus larva does not wait for its food, as is the
tendency in the young stage, but actively pursues its prey. When
an angleworm is seen, perhaps 2 mm. distant, the elastic head cap-
sule of the larva darts forth, curves its claw-like hooks about the
worm's body, and with its head curled under its struggHng prey, re-
treats quickly into the sand until all but its cephalic end is hidden.

WERNER MARCHAND 151

It begins to feed then, devouring in twenty minutes an angleworm
fully four times its own length.

The intestinal tract seen through the hyaline cuticle soon par-
takes of the color of the food ingested. The color is pale brown when
the food consists of the wet muck in which crustaceans and minute
forms are sought. As a result of feeding on blow-fly larvas and angle-
worms, the young Tabanus assumes a variegated color. The intes-
tinal tract then appears tinted with green, yellow, brown, and red
particles of the food.

In one set of larvas, as Mitzmain says, "The origin of cannibalism
as an acquired habit was observed." This was seen in larvae which
had been kept together for four days since their birth. Until that
day no food was offered them except that v;hich they might have
obtained from the surrounding creek water. Apparently they lived
together amicably with their bodies compressed against the glass
dish and the bit of filter paper. A live angleworm was placed in the
glass dish while the resting larvae were observed with a lens. The
worm was not placed in the immediate vicinity of the mass of larvae,
but nearly 4 cm. distant. The presence of the food appeared to act
as a stimulus. No movement was made towards the worm, but each
larva appeared to become greatly excited and began to prod the
larva nearest to it and to nip its neighbor's appendages; several
instances of laceration were noted. This doubtlessly marked the
beginning of systematic cannibalism. From this cause, 39 of the
365 larvas kept in the large glass dish were destroyed within four
days. Four dead bodies were recovered. Upon another occasion
the extent of cannibalism was much more marked. A lot of 415
larvae which hatched on November 12, 1912, was placed in a deep
glass dish with moist lake-beach sand, and fed daily on angleworms.
Each morning it was observed that only about one-half of the worms
supplied the previous day were eaten, so that with the daily fresh
supply more than enough food was present. Another lot of 300
larvae, the same age as the preceding, was kept in individual glasses
under similar conditions. On December 6, counts were made of the
survivors in the large glass dish. Thirty-five larvae remained, of
which eighteen were the maximum size, eleven were a little more than
one-half this size but equal to the largest found in the individual

152 THE EARLY STAGES OF TABANTD^

jars, and the remaining six larvae were so small as to be easily over-
looked. The count of the larvae from the individual jars showed a
loss of twelve, or less than 5 per cent. Allowing 5 per cent for loss
from other causes, it appeared that above 85 per cent of the larvae
kept together in the large jar was destroyed through cannibalism.

It has been observed by Hine in other species that a Tabanus larva
is enabled to survive for a few days in the absence of food. In this
species, likewise, there seems to be a decided resistance to starvation,
two instances showing periods of ten and twelve days.

Mitzmain found the movements of the body to be in general similar
to those of larvae of the muscid type. There is a general progressive
peristaltic movement, invariably accompanied by a decided telescop-
ing of the segments. The head is raised as the prolegs of the anal
end push the body forward, then it is lowered. The mouth is pro-
jected when the head capsule is extended, but recedes quickly when the
glass sides of the container or any obstacle is encountered. The larva
can easily move backwards for a considerable distance. This it does
if wedged in a tight place or in capturing food when it retreats into a
channel previously made in the sand.

The larvae readily adapt themselves to a watery medium. They
can remain submerged for several minutes at a time without apparent
discomfort. When placed in deep water the movements of the
body consist of a general struggling without apparent definite pur-
pose. At any rate, there is little or no progression, the body doubles
like a bow, the head and tail meeting, then straightens with a whip-
ping action. In swimming, the body is held along the surface of the
water and the syphon is extended towards the air in a manner sug-
gestive of the larva of an anopheline mosquito. The principal move-
ment observed is that of simple telescoping of one segment into an-
other. When speed is required or an obstruction is to be passed,
there is a vigorous whipping movement of the syphon laterally,
towards and away from the head. This latter movement is also
noted when the insect is disturbed.

When a young larva is placed in water containing entomostracans
or other minute animals, a barely perceptible churning of the liquid
occurs in the region of the mouth. This disturbance is no doubt
caused by the movements of minute tentacles which assist in procur-

WERNER MARCHAND 153

ing food. These tentacles form the armature of the stomal disk,
consisting of a process arranged like a turnstile mounted on a pitted
chitinous plate at the base of the great hook or mandible. In the
very young larva the stomal disk appears as a chaplet of delicate
chitinous rods. When a larva is treated with strong caustic potash,
the stomal disk appears to be the only structure which resists its ac-
tion, the other chitinous structures, even the heavy pharyngeal
apophyses, are bleached. In common with the other chitinous
portions of the head capsule the stomal disk is shed at each of the
three ecdyses.
An account of the general development of the larva follows:
The young larva shows in its form and behavior its adaptability
to an aquatic life. This is well illustrated when a larva is placed in
an aquarium containing mosquito wrigglers. The Tahanus has no
difficulty in keeping afloat with them and foraging at will upon the
active culicid larvae. Tahanus larvae have been observed capturing
wrigglers, holding them by their jaws under the water, and actually
killing the cuHcid through drowning. In one instance a Tahanus
larva held its victim, which was fully five times its size, suspended
beneath it in such manner that the culicid was unable to project its
syphon for breathing purposes, while that of the Tahanus was func-
tional. The Tahanus, obtaining a secure perch by dragging itself
and the prey above the water, devoured the mosquito wriggler in a
few minutes. In another instance the weight of the culicid pulled
its captor under the water to the sandy bottom a distance of nearly
30 cm. Here the Tahanus showed its superior vitality by remaining
attached for nearly two minutes until apparently assured of the
immobility of its prey, then, releasing its hold, the Tahanus larva
struggled to the surface where it rested with syphon extended. The
mosquito larva meanwhile moved feebly several times, and suc-
cumbed within a few minutes.

This adaptability is lost, however, in the developed larva which
becomes more slothful in movement and grub-like in superficial ap-
pearance. Both extremities, the head and the syphon, become ob-
tuse in form, and the ventral protuberances functioning as prolegs
become more truncated. Growth after the second molt becomes
noticeably less in length and more in thickness. The greatest growth

154

THE EARLY STAGES OF TABANID^

observable was shown to be between the periods of the first and the
second molts.

The following table is given by Mitzmain to show the normal growth
of a larva. The measurements and the critical stages of life are
indicated.

TABLE I.
Progress of Development of a Larva.

Date.

Length.

Stage of development.

mm.

Sept. 15

1.5

At birth.

" 16

1.8

1 day old.

" 20

3.0

" 21

4.0

" 22

5.0

" 23

6.5

" 26

11.0

After first molt.

" 30

20.0

Oct. 8

22.0

" 9

25.0

Second molt.

" 12

27.5

" 17

29.5

Mature larva.

" 21

27.5

" 24

17.5

Pupa, third molt.

In all biological accounts of the Tabanidae there appears to be one
phenomenon which is uniformly noted. This is the remarkable dif-
ference in growth shown by individual flies of the same species. The
only process in the development which seems to be synchronous is
the hatching of the eggs. After that the variations in time of develop-
ment are extreme. In Tahanus striatus, for example, some larvae
twelve days old measured 3 mm., while others under precisely the
same conditions measured fully 11 mm. In another instance two
flies emerged as well developed imagos October 31, while twenty-
seven of the same brood still remained apparently healthy in the
larval stage December 20.

Mitzmain is the first to mention molting or ecdysis in tabanids,
stating that he has been unable to find, "in the very meager litera-
ture available," any reference to the molting process in Tabanidae.
It is referred to indirectly by King at Khartoum, who found in

WERNER MARCHAND 155

Tahanus higuUatus the shed larval skin adhering to the "puparium."^^
Mitzmain had, however, not seen Paoli's work on the organ of
Graber, in which also mention is made of moltings.

By Mitzmain the process of shedding the skin was observed in
Tahanus striatus in a great many instances. The time of molting
of a brood of larv£e is extremely variable, which is consistent with the
great variations noted in the time of development in general. The
process has been accurately noted in two individual larvae, although
observed superficially in numerous others. The three molts are simi-
lar in their general aspects, the main distinction being the more
profound changes produced in the insect at the later molts.

The usual preparations for molting were observed in this species.
The premonitory signs were the refusal of food, uneasiness when
exposed to light, desire to find a remote corner, and finally the stiff-
ening of the cuticle. In one instance the larva was found in one
spot pressed against the glass for three days. Here, between the
sand and the glass of the jar, an excrementous cement was used to
fasten the end of the abdomen. This material holds the end of the
body very securely, although the remainder of the body requires free
lateral movement. By the time the ecdysis is completed, the head
has moved 3 mm. from the spot where preparations for the process
were made, while the anal end has retained its original position.

The shedding of the skin usually requires several hours; in one in-
stance, due no doubt to interference by the observer, the time was
nearly twenty-four hours. In the first and second molts, splitting
of the cuticle begins at the thorax, resulting in the tearing out of the
entire head capsule which adheres to the molt during the remainder
of the process. The anal segments are molted finally and the larva,
emerging in its new skin, crawls its length on the cast skin and rests
alongside it for two or more hours.

The first molt begins with larvae 7 days old, the majority molting
before the tenth day. The second molt usually occurs after an in-
terval of at least four days, and in some larvae as late as eight days,

^^The use of the term "puparium" for the tabanid pupa should not be en-
couraged. The tabanid pupa is a true pupa, in contrast with that of muscids,
or syrphids, where the true pupa is enclosed in the last larval skin, justifying
the name puparium to dififerentiate it from a free pupa.

156 THE EARLY STAGES OF TABANID^

when fifteen to eighteen days old. The time of the third molt pre-
cedes immediately the appearance of the puparium. This period, as
has been noted, shows the greatest diversity among individuals of the
brood. The third molt in twelve instances was shed between the
ninth and twelfth days of life.

In other individuals the process was not completed within three
months, yet the adult fly was an apparently healthy insect.

Certain unimportant changes in morphology, dependent on the
molting process, are noticeable. The loss in size due to contraction
of the cuticle preparatory to ecdysis is usually compensated by a
substantia extension immediately following the process. The extent
of shriveling of the cuticle is represented by 1 mm. in the first molt-
ing, 1.5 to 2 mm. in the second stage, and 2 to 3 mm. preparatory to
the third stage. There is a notable increase in length resulting from
the second ecdysis. A larva, measuring 22 mm. on the day previous
to the shedding of the skin, measured fully 25 mm. the following
day. In measurements of this sort one must make allowance for the
extraordinary amount of telescoping of segments. As much as 5
mm. may be involved in this process.

The structures mainly involved in the ecdysis are the tracheal sys-
tem and the appendages of the head. The anal ring of the trachea
constituting the syphon is drawn off in each molt in a perfectly
cyhndric form. The body trachea is torn from its connections in
irregular strands. The entire head capsule, including the chitinous
pharyngeal framework, .the great hook, and other mouth structures,
are found in perfect form in the various exuvias. These parts upon
renewal in the larVa become more heavily reinforced.

The exuvia (Plate 5, Fig. 81, a, b, c) is usually in a good state of
preservation; crumpled, to be sure, but it can be extended in alcohol
to three-fourths the length of the larva. Following each ecdysis,
the larva is invariably leaden gray with tracheal strands of waxy
white. Three anal segments including the syphon become lead-
colored and stiffened in structure. They are at this stage more trun-
cated, with an anal band of cuticle 1 mm. in depth, making the
syphon appear somewhat atrophied. This is no doubt consistent
with its restricted function. The color of the viscera has changed
from the brilliant red and yellow to an indeterminate white, and the

WERNER MARCHAND 157

lilac tint of the Malpighian tubules has changed to a salmon color.
The latter changes are due probably to a clearing process, in which
the larva indulges during the quiescent stage preceding each ecdysis.

After the second molt the fleshy protuberances functioning as pro-
legs become reinforced with a slight cuticular ring at their bases.
The mouth-parts at this stage are heavily chitinized. The great
hooks or mandibles show a marked serration of the biting edge. The
head projects more, exposing the dark brown ocelli, which, prior to
the second molt are seen only through the cuticle of the thorax situ-
ated nearly on the middle of the concealed head capsule.

The signs characteristic of the final molt which results in pupation
are refusal of food, restlessness, attempted migration, and finally
burial in the sand at the bottom of the jar. The body decreases
slightly in length, but the thickness remains the same.

On the extremity of the abdomen, tiny tubercles appear which
project more from time to time, becoming tapering and spike-like.
Near the caudal end of the abdominal segments, roots of hairs ap-
pear. These at first resemble brown spots of pigment and gradually
lengthen into stiff brown hairs. The cuticle on the body becomes
stiffened and shingle-like at the joints of the segments. The latter
telescope less, and one can see numerous particles of sand embedded
in the joints of the segments. These sand particles have been car-
ried in during the telescopic movements of the abdomen.

After the fully developed larva passes through a period of semi-
dormancy buried in the sand, the skin is seen to be ridged with cutic-
ular plates. The head region is reinforced by stiffened cuticle, and
the mouth orifice is closed by a plug of hard rose-colored cuticle.
This pigmented material lines the entire pharyngeal sinus, plugging
the mouth and the cephalopharynx. The cuticular plug has a sub-
stantial fold which forms a slit for the passage of the molting mouth.
Caudally a similar impervious mass closes the opening of the syphon.
A cuticular collar strengthens the base, and the connective tissue
surrounding the trachea of the tract of the syphon tends to contract.
Then the supports of the central trachea are gradually cast loose by
a gentle wriggling of the insect's body. About this time there is a
general wrinkling of the epidermis, the folds telescoping upon
each other, and the surface becomes like parchment.

158 THE EARLY STAGES OF TABANID^

Synchronous with the primary contraction of the segments, a light
pea-green suffuses the last three segments of the body. The re-
mainder of the larva changes to this color over night. By morning
the abdominal segments have changed from green to ocher, when
the molting of the cuticle ensues. The shedding takes place in sec-
tions. The chitinous framework of the head is thrown off like a
hood. This portion is everted upon the body, and remains dangling
from the exuvia during the process. One-half the length of the skin
is loosened on the side opposite that to which the chitinous framework
of the head is attached. This is shed by a peculiar auger-like move-
ment of the tail end which is not attached to the glass or other ob-
ject in the container, as in the previous molts. The skin is virtually
unrolled from the detached head to the anal end, where it lies in a
crumpled heap. Then the skin of the other side of the body begins to
be shed. The chitinous framework constituting the former head
capsule of the larva becomes rolled up in the exuvia, while the skin
is torn slowly from the new membrane. When the first half of the
skin is peeled off to the anal tip, the cast skin becomes attached to
some object. In this instance the glass of the jar served as an anchor-
age during the remainder of the ecdysis.

The upper half of the body of the newly molted larva is encased
as in an armor in pouches and pads of integument, outlining in a
gauzy film the future appendages of the fly.

Mitzmain's description of the full grown larva (Plate 5, Fig. 80,
c) is given below:

"The length is 28 to 29.5 mm.; the width, 3 to 4 mm. The anterior half
of the body is a greenish yellow, the remainder is a dirty white. At this stage
the form is grub-like."

"The head capsule, which occupies one-fifth the length of the larva, is a
cylindrical bulb, formed by the invagination of the thoracic ectoderm. It sup-
ports the eyes, the antennae, and the mouth parts. It is bound by a framework
of chitinous rods, the cephalopharyngeal apophyses. This structure, observed
through the thorax when the insect is in action, is composed of four black, me-
dium-thick, skeletal pieces running the length of the three cephaHc segments in
the form of a pyramid, with its apex provided with the external mouth parts.
It terminates in the claw-like mandibles which are similar in color and texture."

"The mandibles [Plate 5, Fig. 79] are heavy, powerful structures, slightly ser-
rated on their inner surfaces. The musculature of these appendages permits the

WERNER MARCHAND 159

two elements working in apposition. At rest they are held horizontally, and can
be projected suddenly and thrust vertically downward, which is obviously of
great assistance in grasping the prey."

"The palpi and antennae in this species are silvery white, and usually found
glistening with moisture."

"The eyes are oval in shape, with the long axes parallel. When the larva is
prepared to molt, the pigmented spots are usually distorted. In this species
the eye spots or ocelli are very prominent, especially in the younger stages of
the larva. They can be first seen in the embryo where they appear as dark
beaded structures through the chorion of the well developed egg. In the young
stage of the larva the eyes appear in the pharyngeal cavity midway between the
mouth and the cephalopharynx, and as growth continues they are located nearer
the distal end of the head capsule; so that when the larva is full grown and the
mouth structures protrude in locomotion or prehension the eye spots are seen
to project on the head capsule with the mouth parts."

"The trachea which terminates in the conical tubular syphon is lead gray in
contrast to the dense white of that portion anterior to the anal segment. Anterior
to the syphon there is a cuticular cellar of slightly darker shade."

"The prolegs are formed by truncated projections, six in number, three on
each side of the midventral line and extending laterally. Each protuberance is
provided with a tuft of short, fine, brown hairs. These hairs appear to be sur-
rounded by a secretory substance, which is slimy in character."

"At the base of the syphon, beneath the cuticle on the dorsal side opposite
the anal capsule, is a tiny structure which attracts attention on account of its
movements and peculiar arrangement. In the newly hatched larva it is a process
composed of four lustrous black disks arranged in two pairs, one in front of the
other, and set in a mass of fat bodies. The larger of the disks, the anterior
pair, is less than 0.1 mm. in diameter. The movement of the process is similar
to that of a pendulum, and is active only when the larva moves. With each
molt these disks become smaller and increase in number. In the full grown
larva the process becomes a triangular mass of loosely arranged beaded disks.
They appear to be mere specks of pigment beneath the skin, but their structure
and action are so constant that either the process is characteristic of the species
or investigators have overlooked or ignored them in other species."

The description of the pupa (Plate 12, Fig. 152, a, b; Fig. 143, a,
h) after Mitzmain is given below:

"The average length is 18 mm., and width, 3.5 mm. The color is pale brown,
the last 2 segments of the abdomen being slightly darker. The head tubercles
are not clearly defined; color, dark brown. The prothoracic spiracular tubercle is
slightly elevated, oblique; rima, salmon-colored and crescentic in form."

"The first abdominal spiracle is perfectly round and larger than the others,
which are slightly ovoid; the rima of all the spiracles curves from above
posteriorly."

160 THE EARLY STAGES OF TABANID^

"The terminal abdominal segment [Plate 12, Fig. 143, a, b] shows a sexual
distinction in the arrangement of the short spines midway on the ventral side,
anterior to the terminal teeth. In the male ten to twelve of these spines form
a continuous serrated border. In the female the spines occur in two groups of
four to six spines similar to those of the male, but separated by space equal in
width to that of one of the groups."

"The terminal teeth of the posterior segment are arranged with two pairs
close together on the dorsal side and one pair on the ventral side. These teeth
are black-tipped and acute; all of them are directed slightly outward. The
lateral teeth of the two dorsal pairs are the longest. The ventral pair is smaller
and is set slightly in from the periphery of the segment."

"After the final ecdysis which results in the formation of the puparium, the
nymph, at first a light green, gradually changes to yellow. Upon the second day,
the eye spots change from yellowish to pale brown, then to a chocolate color.
Beginning with the third day the pads of the wings and the legs, at first light
brown, assume the same color as the eyes. The chitinous pad enveloping the
wing is densely opaque, so that only the plications of the developing wing can
be discerned. Upon the penultimate day, the fifth or sixth usually, the abdomen,
which heretofore has been a uniform yellow-brown, becomes striped with light
orange and brown, which colors gradually deepen until the time of emergence."

The process of emergence from the puparium is also described by
Mitzmain (Plate 12, Fig. 152, b):

"In emergence, the puparium which lies buried to some depth in the sand is
invariably dragged to the surface where the final acts of emergence are com-
pleted. Two or three days prior to the act of emergence, the puparium shows
considerable mobility when disturbed by handling or stimulated by light. Cer-
tain movements, which one learns through numerous observations to be charac-
teristic, can be considered as actually premonitory. These occur usually from
ten to twenty minutes prior to the breaking of the cuticle, and serve the observer
as warning signs. If during this interval a low power lens is focussed on the
compound eye, the epidermis of the fly separating from its connective tissue fas-
tening of the puparium can easily be seen. This action resembles strikingly a
wave of water moving between the walls of the puparium and the epidermis of
the fly. It may be considered as the movements of a semiliquid layer between the
fly and its puparium. Another movement, which can be observed within a few
minutes after that previously described, is the momentary contraction and ex-
pansion of the sides of the abdomen between the two lateral ridges. This too, no
doubt, is effective in tearing the connective tissue lining to facihtate emergence.
A few minutes later the anal end of the abdomen is torn loose from its fastening,
and emergence of the fly begins."

WERNER MARCHAND 161

Since the puparium is unrolled from the head, the compound eyes
are soon exposed to view, so that the sex of the fly may be distinguished.
The appendages, antennae, palpi, and mouth-parts are dimly visible.
The head appendages are freed primarily by the spasmodic wriggling
of the abdomen, but the labellum, which is seen to become turgid and
flaccid in turn by the injection of air into, and withdrawal of air
from, the "extensive tracheal sacs which lie in the cavities," and the
erectile stomal disk through the pressure downwards against the walls
of the puparium, assist also in the process. That these head ap-
pendages assist effectually in the emergence is evident from the
lines of cleavage in the enveloping membranes.

The puparium splits on the median line of the thorax; simul-
taneously the hood enveloping the head drops by a sternal hinge.
The labellum can be seen still pressing upon the interior of the hood
as the head emerges. In a minute the wings are rent from their
envelopes by the sturdy pressure of the legs, which have slid out of
their sheaths simultaneously with the cleavage of the thorax. The
legs directly assume their normal position, and the fly walks forth
bodily, spreading its plicated wings. The liberated wings show a
clear expanse of unwrinkled membrane which at first is soft in texture
and clear lead-colored throughout. Finally the inflated abdomen
appears in the dorsal slit, and at once is drawn clear of the encumber-
ing puparium.

The time from the appearance of the head to the evacuation of the
puparium requires less than two minutes. This time is increased a
minute or two whenever the wing sticks to the lining of the puparium,
resulting usually in a torn wing.

Directly after emergence the wings are shorter than the body, but,
constantly vibrating, they gradually lengthen, whereupon they be-
come hardened and prepared for flight. The fly does not spend any
time preening itself, as is the case with some of the Muscidae at this
stage. The time prior to flight is spent, however, in a clearing
process. This begins with a copious discharge of meconium within
three to five minutes after emerging. At first the defecation is per-
formed at least five times per minute, then once per minute for a
period of twelve minutes. At the end of this time the excretions be-
come more watery in character. In the meantune the fly walks

162 THE EARLY STAGES OF TABANIDiE

about in a restless manner, constantly vibrating the balancers and
flapping its wings, while the distended abdomen becomes reduced to
more normal proportions.

The meconium, which is deposited in large quantities, is pale brown
in color, rapidly changing to amber, then becoming clear. The pri-
mary, heavier excretion appears decidedly oily in nature, when
examined with the microscope.

In from fourteen to twenty minutes, voluntary flight takes place.
This is at first tentative, the insect alighting upon the floor about a
meter distant. After a minute of rest, flight is resumed, the fly
escaping through the open window.

The puparium left behind shows certain points of cleavage which
prove to be very constant. There is a dorsal slit on the median line
of the thorax which extends nearly the length of the notum. Another
slit extends midway across the orbital region through the genai to
the wing pouches. A third slit extends between the two wing en-
velopes, and a slight one behind the prothoracic spiracular tubercle.

In the thirty- two emergences recorded, the males preceded the
females by an average of half a day. The males spent from three
to seven days in the pupal stage, averaging five and one-half days,
while this period required four to nine days with an average of six
days in the female flies.

Bainbridge and Fletcher (1914) observed Tahanus striatus near
Madras, and give the following statements which may serve to com-
plement Mitzmain's work in as far as it seems based on field obser-
vation :

"Eggs are laid in a large mass, usually on a leaf or twig overhanging water
into which the young larvce drop on emergence, thence-forward leading an
aquatic life burrowing in the mud at the water's edge and feeding on worms or
living or dead insects. The full grown larva is 40 to 50 mm. long, dull whitish,
elongate, tapering at each end with protuberances at the edges of the segments.
When full fed it leaves the damp mud at the water's edge and after a quiescent
period pupates in the earth above water-level."

An egg mass of Tahanus striatus on a paddle leaf is figured (Plate
2, Figs. 22 and 23).

Patton and Cragg state that Tahanus striatus oviposits as a gen-
eral rule on blades of grass, pieces of stick, etc., at the edge of a river,
stream, or pond.

WERNER MARCHAND 163

The larvae are said to be powerful swimmers having air sacs con-
nected with their tracheal tubes, so that they can float or sink at
will.

Large numbers of egg masses are regularly destroyed in Madras
by a small chalcid parasite which was not identified.

The eggs are brownish white when deposited (Patton and Cragg).

Tahanus stygius Say. — A species of the Middle and Southern States,
recorded from Massachusetts, New Jersey, Illinois, Ohio, etc., the
life history of which is comparatively well known through the
observations of Hart (1895) and Hine (1906).

On its oviposition and egg Hine is the first to give us information.
The species oviposits principally on the leaves of Sagittaria stand-
ing in shallow water, habitually placing the eggs just above the point
where the petiole meets the expanded part of the leaf (Plate 1, Figs.
3 and 4).^^ The precision with which this habit is followed becomes
a matter of interest. Out of hundreds of masses of eggs observed,
only a few were placed on other species of plants or in a different
position on the leaf (Plate 1, Fig. 5). The female is occupied for a
half hour or more in placing the several hundred eggs composing a
single mass, and during this time the observer can take a position
close by and watch the proceedings without frightening her away,
but species of Tahanus are more particular about the approach of
intruders than are various Chrysops.

The egg mass is white when first placed but soon turns brown;
it is convex, and is composed of about five layers, one above the other.
Individual eggs are nearly the same size as those of Tahanus atratus,
and are similar to them in form. Hatching as observed, occurred in
seven days after oviposition. From a careful study of microscopic
sections of eggs killed as soon as laid, it was concluded that develop-
ment does not begin until after oviposition, consequently the time
given is the entire incubation period.

When first hatched the larvae contain a considerable amount of
unused yolk, which furnishes them food for a time; it is therefore
unnecessary for them to eat anything for a few days. This is ad-

^^ The leaf represented in Fig. 4, is a leaf of Peltandra, similar in shape to the
Sagittaria leaf (Plate 1, Fig. 3).

164 THE EARLY STAGES OF TABANID^

vantageous no doubt, for food is not always at hand, and in case it
is not, the fact that nourishment is furnished naturally gives them
an opportunity to investigate their surroundings.

At hatching time nearly all the larvae that come from a single mass
of eggs appear at the same time and when they have freed themselves
from the shells, go tumbling down into the water, scattering more or
less and sinking to the bottom, where it is difficult to observe their
further actions.

Hine found that small catiishes (Ameiurus mclas) are enemies of
these larvae, and he observed two of them devour 200 young larvae
of Tahanus within a few hours.

Hine's attempt to rear the larvae from the egg to the adult was not
successful. On July 21, a number of larvae just hatched were placed
in a breeding jar containing damp sand covered over the top with fine
plant material, and small crustaceans were put in for food. The
larvae took kindly to the surroundings, accepted the food offered,
and began to grow at once. After about two weeks, as angleworms
were much easier to obtain, these were substituted for the crustaceans,
with no bad effect on the larvas, which continued to grow, though
rather slowly. The largest attained a length of about 10 mm. by the
beginning of winter, when they ceased eating. They appeared to be
in good condition in the spring, but for some reason died without
further increase in size.

On August 2, of the same year, Hine took a large larva of this
species in Summit County, Ohio, from under a flat stone along a
brook that ran from a spring. When taken this specimen measured
over 40 mm. in length and had every appearance of being mature,
but it continued to eat the anglewonns given it until late in the
fall. It then ceased feeding until the following spring, when it took
a small amount of food and entered the pupal stage about the middle
of May, the adult, a male, issuing June 14.

Hine thinks that it is hardly possible that this species passes all its
transformations in a single year, for the larvae reared from eggs were
not over 8 mm. long when the specimen over 40 mm. long was col-
lected; and as the latter did not produce the adult until about the
normal time for adults to appear under natural conditions, it does
not seem possible that the first mentioned larvae could have reached
maturity and produced adults before the second year.

WERNER MARCHAND 165

The young larva is briefly described by Hines as follows:

"Larva, when first hatched, 4 mm. long; entirely light colored; form as in
older specimens. As growth continues size is the only noticeable change."

The mature Urva has been figured and described in detail by Hart
(Plate 3, Fig. 40). This species was the most abundant tabanid
larva found in the vicinity of Havana, Illinois, in the spring of 1895.
It first appeared in Hart's collections September 14, 1894, when a
number were noted swimming among vegetation near the margin at
Station B, on the wet springy shore of Quiver Lake, with sandy
and muddy ground, grass, and coating of alg£e. In the spring they
were, however, found at nearly all of the stations, but more particu-
larly in connection with tipulid, muscid, and Er is talis larvae in matted
accumulations of dead stems and leaves over mud. They were es-
pecially abundant on March 30 in Flag Lake, where large plump
larvae appeared at every turn. It was a surprise to find a few of them
upon the bottom in the open shallow water, far from shore, in the
middle of Quiver Lake at Station A, for which locality Hart gives the
characteristics: "Shallow, mud and sand, grass and floating vegeta-
tion, variable." Young larvae have been common in connection with
larvae of Bittacomorpha and Limnophila, at Station I (see Tabanus
atratiis) since March 17. At others of Hart's stations they have been
common in moist drifts of fine rubbish washed up by waves. Pupae
were formed in the breeding cages May 10 and 23. One emerged
May 27, and another tried to emerge June 2, but died and was re-
moved from its case.

The larvce of Tahaniis stygius resemble, accordhig to Hart, those
of the lineola group in their striation and coloration, but differ in
their short lateral prothoracic areas and larger size. They are like
atratus in size, but may be readily separated from it by their coarser
lateral striation, straw-yeliow tint, slender lateral pigmented stripes,
and usually projecting terminal stigmatal spine.

Hart's description of the mature larva follows:

"Larva. — [Plate 3, Fig. 40.] Length 45-55 mm., diameter 6-7 mm. Bright
straw-yellow, varying in some young larvae to nearly clear white; marked with
light fuscous brown microscopic pubescence, usually paler at each stage than
atratus."

166 THE EARLY STAGES OF TABANID^

"Lateral prothoracic striated areas are not more than half as long as the dorsal,
striation not finer than that of the middle and lower lateral areas of the mesothorax,
striated portion shining; a small smooth spot adjoining the impressed line below;
remaining upper lateral thoracic areas a little less coarsely striated, but not
strongly different from that of the prothorax; abdominal lateral areas a little more
finely striate; dorsal and ventral areas with margins striated, disks nearly smooth
in adult larvae, last segment more strongly striate, especially beneath.''

"Dark annuli distinct, broad, including false feet, a dislinct transverse dorsal
and ventral pale spot in front of the false feet; abdominal annuli often with a
small triangular backward prolongation on median line above. Prothoracic
lateral space occupied by a pale brownish fuscous quadrate spot in front of the
striated space. Meso- and metathoracic lateral stripes usually distinct, but
slender, scarcely dilated posteriorly, lateral edges of dorsal areas diverging; lateral
stripes of abdomen almost wanting, except on last two or three segments. In
these stripes the punctures of the upper and lower rows are indicated by rounded
pale dots, and those of the inner rows by elongate dots. Last segment with bases
of respiratory tube and anal prominence encircled with dark rings; joined by a
lateral connection, its dorsum with at most a short basal line or pair of dots on
each side. Coarser pubescence of false feet tipped with pale brownish."

"Main internal tracheae thick and noticeable, especially in young larvae,
lustrous, subparallel, not strongly sinuate, nearly straight posteriorly; terminal
stigmatal spine dark reddish brown, smooth, usually protruded."'

The pupa (Plate 11, Fig. 125; Plate 13, Figs. 153 and 161) has
been described by Hart as well as by Hine, both of whom reared the
adult from pupas obtained from adult larvae taken in nature. The
following is Hart's description.

" Pupa, female. — Length about 30 mm., diameter about 6 mm. Light brownish
fuscous, thorax paler, shining, abdomen roughly transversely WTinkled, and sub-
opaque. Palpal sheaths distinct, as far apart as are the setae borne by the
larger tubercles at the center of the anterior surface of the head [Plate 13, Fig.
153]; surface between them rounded, bearing a small wrinkled tubercle at middle;
antennae and tubercles darker than surrounding surface; ocellar tubercles dis-
tinct; prothoracic spiracular tubercles slightly but evenly elevated in a plane
parallel to that of the surrounding surface; rima nearly straight in its outer half,
inwardly curving strongly forward, and ending in a conspicuous hook; free mar-
gin of tubercle rounded at tip."

"First abdominal with two distinct setae each side above the spiracles; ab-
dominal spiracular tubercles rounded, broad behind, low-subhemispherical, rima
long, following posterior border of tubercle, sHghtly curved at middle, more
strongly curved forwards at each end; on anterior surface a transverse groove
extending across the tubercle, but not as long as the rima. Fringes of unequal

WERNER MARCHAND 167

spines, often tipped with blackish, all but two of the long spines wanting in a
broad space above on seventh segment. Terminal teeth [Plate 13, Fig. 161]
nearly equal, tipped with blackish, their points marking the angles of a hexagon,
slightly wider than high. Ventral fringe of last segment not webbed together;
lateral tufts high — on a level with upper lateral line."

Hine's description is shorter:

"Pupa [Plate 11, Fig. 125] 29 mm. long; color dark, approaching fuscous;
prothoracic spiracle strongly bent at the middle; rima oblique and straight for
the outer half of its length, remainder gradually curved, with a broad hook at
the inner end. Teeth at the end of the abdomen [Plate 13, Fig. 161] six in num-
ber, nearly equidistant from one another, of nearly the same size, with the
extreme tips slightly turned inward."

Tlie pupa of stygius is, according to Hine, mucli lilce tliat of sulci-
frons, but tliere is some difference in tlie prothoracic spiracles and in
the abdominal teeth.

Tabanus sulcifrons Macquart. — A North American species inhab-
iting Pennsylvania, New Jersey, Ohio, Illinois, and Louisiana.
Closely related to Tabanus exul and abdominalis , and like the two
latter, appearing late in the season, in Ohio abundant in the latter
part of July and all of August, in Louisiana recorded from September
7 to October 16.

Hine has made many interesting observations concerning the habits
of the adult flies of this species, but was unable to worli out its life
history. Although the eggs were procured in many stages of de-
velopment by dissecting the females, the habits of oviposition have
not been observed. The form of the eggs and the number produced
by a single female are the same as in other species of its size. Speci-
mens containing eggs almost fully developed were taken in various
places, but Hine could not get any clue as to where oviposition oc-
curred by dissecting the females where they were collected, as he had
hoped to do.

The pupal case of the species was procured by locating a female
which had just emerged. The place where this pupal case was taken
was on a side hill, about 75 feet above the bed of a small stream.
The description follows:

168 THE EARLY STAGES OF TABANID^

" Pupa.— [Flate 11, Fig. 126; Plate 13, Fig. 165.] Length 26 mm., diameter
6 mm. Color yellowish brown, the thorax being nearly the same color as the ab-
domen. Tubercles of the head region well marked and distinctly darker than
the surrounding parts. Prothoracic spiracular tubercle brown in color, ele-
vated, narrow, ventral half oblique, dorsal half turned directly forward, thus
forming a distinct bend near the middle of the length; rima nearly straight from
outer end to the middle and evenly curved for the remainder of its length, inner
tip curved backward, thus forming a well-defined hook. First abdominal spiracle
nearly round; its rima following the posterior curvature, very narrow, but a little
widened above; remaining abdominal spiracles a little smaller than the first one,
each with a short, slightly curved or straight rima. Terminal abdominal segment
[Plate 13, Fig. 165] with several small spines near the middle of its length and
six larger spines at its apex. These spines are all brown in color, with the apex of
each approaching black. Six apical spines of nearly the same size; the dorsal
pair point upward, outward, and slightly backward, the lateral one on each side
outward and backward, while the ventral pair extend almost directly backward.
These six spines mark the corners of a hexagon with nearly equal sides, but the
ventral pair are a little nearer together than the dorsal pair."

Tliis appears to be all tliat we Icnow about tlie early stages of this
species. From the place where the pupa was found, tlie larva appears
to be terrestrial in habit, lilce that of Tabanus fronto, but one should
bear in mind that possibly the young larva is aquatic and leaves the
water at a later stage of its development.

Tabanus tceniola Palisot de Beauvois. — This species is, according to
King, the most common and most widely distributed tabanid found
in the Anglo-Egyptian Sudan, and tlie most frequently accused of
causing the death of camels. Occurring on the White Nile as far
north as Dueim, stray specimens have occasionally l3ccn taken in
Khartoum. King was able to secure oviposition in capti\dty in this
species as well as in Tabanus par, and has described the egg and
larva. The pupal stage is not known.

In order to secure eggs, gorged females were taken in May, on cattle
grazing near Bor, and placed in a breeding cage with a dish contain-
ing grass and weeds growing in mud and water. They were fed on
sugar and water, and a few batches of eggs were obtained. A single
egg batch was taken in May on a blade of grass overhanging a dried
up water pool near Kanissa wood-station, and a number of batches of
eggs were collected early in July from grasses and weeds overhanging
rain pools at Gebelein.

WERNER MARCHAND 169

The eggs (Plate 1, Figs. 8 and 9) are placed by the female fly on the
upper side of a blade of grass or some similar plant, and, with the excep-
tion of the single batch taken at Kanissa wood-station, all those found
were overhanging water. An unfinished batch of eggs resembles an
arrow-head. The eggs are closely applied to each other and left bare,
so the batch can easily be seen when freshly laid, owing to its shin-
ing white to yellowish white color. Prior to hatching, the egg mass
becomes darker. According to King, "the egg is spindle-shaped,
about 1.75 mm. in length and, when first laid, is white in colour.
It becomes darker as the embryo within develops."

The eggs obtained in the breeding cage were laid on May 24 to 25,
and hatched on May 29. This would correspond to the very short
incubation period of four to five days. The larvae (Plate 1, Fig. 9, a)
were placed in glass basins containing mud, growing grass, and water,
and were offered the expressed stomach contents of female ticks —
Rhipicephalus simus — taken from a dog. They fed readily on this
until June 1 1, when they were placed in clean river sand and water, and
their diet was changed to mosquito larvae. These mosquito larvae
were either killed or laid living on the wet sand out of reach of the
water, in which position the tabanid larvae were able to kill them.
In water the mosquito larvae were too active to be caught. On July
16 their food was changed again to freshly killed and bruised earth-
worms, and these they also ate readily. While still young they be-
came vicious cannibals, and consequently each larva had to be given
a separate dish. They were brought to Khartoum on July 19 and a
few days later it was noticed that the majority were not taking their
food. They were then nearly if not quite full grown, so it was thought
that they had buried themselves in the sand prior to pupating. A
thorough search, however, revealed the fact that they had disappeared,
and it was not until later that mice were identified as the cause of the
loss. The two remaining larvae were then killed and preserved. It
is possible, therefore, that the larva described by King is not quite
mature.

The larvae of Tahanus tcEuiola are more active and ferocious than
those of Tahanus par, vigorously attacking any other larva with
which they may come in contact. They have not, however, the
power possessed by Tabanus par of lying dormant in the soil for at

170 THE EARLY STAGES OF TABANID^

least fifty-seven days if the conditions are unfavorable for their
development.

King describes the larva somewhat as follows:

"The larva [Plate 3, Fig. 49], when fully extended, measures about 29 mm.
Color white to grayish white, mandibles black. On the anterior third of each
abdominal segment, except the eighth, is a ring of pseudopods, eight in each ring
— two dorsal, two lateral, four ventral — except on the first abdominal segment,
where the dorsal pair is wanting. On the second abdominal segment the dorsal
pair is very strongly developed. The ventral pseudopods^^ are always larger
than the dorsal. Each pseudopod bears a crown of colorless spines or hooks,
and between the pseudopods there are also spines or hooks, often darker in color,
and forming a continuous ring. The anus is situated ventrally at the base of the
eighth abdominal segment and is edged with dark hairs. On either side of the
anus is a patch of dark hairs, roughly kidney-shaped, and beyond each patch,
laterally placed on the segment, are two small round spots of dark hair. The
syphon tube consists of two segments, and when exserted is shorter than the
eighth segment. The whole surface of the larva is more or less shiny, with vary-
ing longitudinal striation, the areas bearing very fine striae being markedly duller
than the rest. The prothorax has the dorsal area smooth in the anterior two-thirds
and rather coarsely striate posteriorly; the ventral area is almost entirely smooth
and divided in two by a medium furrow; the two lateral areas are finely striated in
the basal third and more coarsely so in the anterior parts. The mesothorax has
the dorsal and ventral areas smooth and shining in the anterior two-thirds, and
rather coarsely striate posteriorly, the ventral area having no furrow; the lateral
areas are a little more finely striate than those of the prothorax, and there is
a rather broad dull non-striated band at both the anterior and posterior margins.
Similar dull bands occur on the metathorax and the abdominal segments, but com-
pletely encircling the segments. The abdominal segments 1 to 7 have the dorsal
and ventral areas moderately shining, and the striation is rather coarse and irreg-
ular; the lateral areas appear much duller, owing to the extreme fineness of the
striation. On the eighth abdominal segment the striae are moderately well
marked and of similar appearance on all the faces."

The pupa of Tabanus tceniola is not known. From the egg mass a
natural enemy parasite was obtained by King, adding hereby one
more example to the cases of egg parasitism observed in tabanids.
These small Hymenoptera were bred from an egg mass of Tabanus
tcBfiiola, taken at Gebelein. The species had not yet been identified
at the time when King's report appeared, but figures were given, to-
gether with the parasitized egg mass, showing the exit hole of the

•^8 See p. 108, foot-note 18.

WERNER MARCHAND 171

parasites. To judge from the illustration it might be a species of
Phanurus, but as the figures are life size, not much can be seen of
structural details.

As elsewhere in tropical Africa, this is probably the commonest
species of the genus in the Mlanje district, where Neave bred it also
from larvae.

The larva is, according to Neave, chiefly remarkable for its white
color and lack of pigment, and for the presence of a row of bristles
immediately anterior to the anus. It is one of the most active and
restless species Neave had to deal with.

Tabanus {N eotahanus"^^) triangulum Wiedemann. — This Brazilian
species was reared by Lutz, at Manguintros, Brazil, in February,
1914, from full grown larv£e, which were obtained by sifting the mud
from the edges of a small brook (together with those of Neotahanus
ochrophilus), and breeding in damp moss. No morphological dif-
ferences were observed between this species and Tabanus ochrophilus.

Tabanus trimaculatus Palisot de Beauvois. — A species occurring in
the middle and southern states west to Kansas (Hine). We possess
very few data on its early stages.

Hine mentions (1903) that he has been in possession of the eggs,
in connection with his studies on extermination of tabanids by col-
lecting and destroying the eggs. By counting it was found that
twenty egg masses of Tabanus trimaculatus averaged over 500 eggs
each.

Brimley found, on April 15, 1909, two larvae of Tabanus, under the
bark of a soggy log which was an inch or two above the water. One
died; the other was put into a bottle with some wet dirt and rotten
wood, and "from this a male of Tabanus trimaculatus was bred on
May 18 of the same year. The larva that died and which was
presumably the same species was preserved in alcohol. It measures
37 mm. in length and is white without markings. "^^

Tabanus tropicus Linne. — A species found in Europe (Laibach and
Trieste, according to Schiner), but also in India (same species?) and
here said to be a carrier of surra (Neveu-Lemaire) .

^^ In the meantime, I have repeatedly found the larvae in Princeton (1917)
and bred the adult.

172 THE EARLY STAGES OF TABANIDiE

Scholtz (1850) reports that he has found the pupae, together with
those of Tabanus autumnalis and of Hccmatopota pluvialis in the
neighborhood of Breslau in June, 1850, at the edge of a pond cov-
ered with Zewwa, the water of which was polluted from manure piles
surrounding it. The pupae were found near the edge under a thick
mass of moist Lemna, together with Stratiomys — and Syrphus —
pupae ;^° the flies hatched after a few days.

Tabanus ustus Walker. — An African species, observed by Neave in
enormous numbers around a pool in a nearly dry stream bed, on the
plains in Portuguese territory, southern Nyasaland.

One male and three females were bred in the laboratory at the end
of October apd beginning of November. The larva, of which the
terminal segments are figured (Plate 5, Fig. 66), resembles that of
Tabanus biguttatus, but is less pigmented.

The upper hooks of the pupal aster are considerably larger than the
remainder. The spines of the dorsolateral comb are much reduced,
especially in the male.

The pupal aster of the male and the dorsolateral comb of both sexes
are figured (Plate 14, Fig. 177, a, b, c).

Tabanus variabilis Loew.^ — An African species, not rare near Mt.
Mlanje, southern Nyasaland, on wooded streams in the neighborhood
of the mountain in October and November, occasionally later.

The larvae were found in this locahty by Neave, in abundance.
They are entirely different from those of Tabanus atrimanus, being
almost colorless, though in quite mature individuals the base of the
syphon and the syphon itself are of an orange color. The most strik-
ing peculiarity of this larva is, however, the presence of a distinct
papilla of a dark color on each side of the anal segment. This is
easily recognizable in life and distinguishes this species from any
other seen by Neave. The anus is also unusually prominent.

The pupa is also remarkable for its dark coloration, especially on
the dorsum of the thorax. The aster is characterized by the large
horizontally extended middle pair of hooks, and its outline is there-
fore entirely different from that of the closely allied Tabanus atri-

30 See also pp. 72 and 91.

WERNER MARCHAND 173

manus. The dorsolateral comb consists of a few short and rather
stout spines.

Figures are given of the syphon of the larva (Plate 5, Fig. 70), the
pupa (Plate 12, Fig. 138), pupal aster, and dorsolateral comb of both
sexes (Plate 14, Fig. 176, a-d).

Tabanus virgo Wiedemann. — ^A small species, recorded for Bengal
and South India. On its early stages we possess notes through the
work of Patton and Cragg.

As in all the small tabanids observed by these authors, the eggs are
laid on blades of grass just at the edge of a shallow stream, or on the
leaves of the lotus plant at the edges of small ponds, but never over
deep water. The larvae, as in the other small species, are said to
have no air sacs, and to die when falling into deep water.

The mature larva of Tabanus virgo is figured by the same authors
(Plate 4, Fig. 63). Attention is called to the openings of the tracheae
being flush with the body. In fact the syphon is extremely short
when compared with that of Tabanus hicaUosus, while as shown in the
figures, in Tabanus ditcBniatus it is of medium length. On the figure
given, the larva of Tabanus virgo shows no striation or color pattern,
but the mouth-parts appear larger than in the other species, and the
prolegs are large, fleshy, and somewhat modified.

A small egg mass of this species is figured (Plate 1, Fig. 16) as laid
on a dry twig, and consisting of only five eggs, attached to a twig
which in fact is much smaller in diameter than the eggs themselves.

The pupa is figured (Plate 11, Fig. 134) by Patton and Cragg, and
in addition, an enlarged figure of the eighth abdominal segment of
the pupa is given (Plate 12, Fig. 146), showing the six terminal teeth
and their arrangement.

Tabanus vivax Osten Sacken. — A North American species which
apparently is never very plentiful, but has been taken in a number
of the eastern states (New York, Maine, Ohio, etc.). In Ohio it is on
the wing during the last half of June.

The life history has been worked out by Hine. The eggs are de-
posited in masses composed of several hundreds, on stones that pro-
ject above the water in riffles of streams. The egg mass is nearly
round in outline, only slightly convex, composed of about three

174 THE EARLY STAGES OF TABANID^

layers one above the other. The color of the whole mass is brown,
mottled over the top with white. In these respects they do not
differ in particular from the eggs of other species of the genus, but the
masses observed were not so convex as those of Tahanus atratus, and
being placed on stones of a color similar to themselves are rather
difficult to see. Females have been observed ovipositing as early as
June 8, but most often eggs are deposited after this date.

The larvae (Plate 3, Fig. 43) occur in the streams in the fall. Hine
states that in September and October each year they collected the
larvae of the dobson-fly (Corydalis corntUa L.) for study in the lab-
oratory. By turning stones at the edge of swift riffles, or by means
of a net stretched across the riffles to catch such specimens as are
dislodged by turning stones behind the net in the stream, in addi-
tion to Corydalis larvce, a large number of the larvae of this horse-fly
were found. Though having done much collecting in streams, Hine
asserts that the larva of Tahanus vivax is the only tabanid larva
taken in riffles so far. It was not found difficult to rear these larvas.
Larvae taken late in the fall were placed in damp sand and fed on
angleworms. As winter approaches they refuse to eat and remain
quietly in the sand until the following spring; then they feed actively
for a few days and change to pupae. They reach the adult stage in
late spring or early summer. Like other tabanid larvae, the larvae of
Tahanus vivax are not particular as to their food; all that appears to
be necessary is that they obtain small soft bodied animals. Crus-
taceans serve them as well as insects and their own species as well as
some other species — whatever, in fact, is in the sand of the breeding
cage.

Hine has never observed the larvae in nature in the spring; conse-
quently their habits at this time of the year are not exactly known,
but Hine supposes that they leave the water and pupate in the
earth near at hand.

Hine's description of the larva and pupa follows:

''Larva. — [Plate 3, Fig. 43.] When full grown, about 25 mm. long. Gen-
eral color yellowish white, anterior margin of each thoracic segment and a narrow
band, including the prolegs, on the anterior half of the first seven abdominal
segments opaque, and appearing darker than the other parts, which are more or
less shining and usually finely striate longitudinally. Prothoracic segment di-

WERNER MARCHAND 175

vided by longitudinal grooves into four nearly equal parts, which may be called
the dorsal, ventral, and lateral areas. The lateral areas are shining and finely
striated on the posterior third and opaque on the anterior two-thirds; the dorsal
and ventral areas are opaque on about the anterior fourth and distinctly shining
on the remaining parts. The ventral space is plainly divided into two equal parts
by a longitudinal groove. In order to see the character of this segment, it must
be fully extended. The mesothoracic and metathoracic segments have a num-
ber of longitudinal grooves, some of which are very narrowly bordered by opaque
darker coloring, which proceeds backward from the narrow anterior border of
these segments. Each of the first seven abdominal segments has on its anterior
part a transverse row of eight tubercles which encircles the segment. These all
bear short spines or claws at the apex, excepting a dorsal pair on each of the first
three or four segments. They may be called prolegs, since they have the parts
necessary to such organs and, what is more, are used as prolegs. On the poste-
rior dorsal border of most of the abdominal segments there may be a narrow,
irregular, opaque marking of the same color of the narrow band in the region of
the prolegs; eighth segment on each side with two narrow, curved markings which
have the appearance of being composed of contiguous punctures. These mark-
ings are of the same shade of color as the other darker areas, and the lower one
is more than twice as long as the upper."

"Pupa. — [Plate 11, Fig. 128.] 18 mm. long and 4 mm. in diameter. Light
brown in color, thorax somewhat paler than the abdomen. Antennal and other
tubercles of the head and thorax prominent and darker than the surrounding
parts. Prothoracic spiracular tubercle slightly elevated, reniform, oblique; rima
uniformly curved for nearly its whole length; but just before the anterior end the
curvature is stronger, although no hook is formed. First abdominal spiracle
nearly round; rima almost uniformly curved, posteriorly very slightly widened
just at the end, anteriorly slightly narrowed and curved so as to form a short
hook. The other abdominal spiracles agree with the first one in general, but
there is slight variation in the enlargement and curvature of the extreme ends.
Terminal teeth [Plate 13, Fig. 160] prominent, shining brown in color, darkest
at the extreme tips. Dorsal pair of teeth smallest and closer together than
the ventral, lateral teeth longer and larger than the ventral and located much
beneath the dorsal, in fact they are nearly midway between the dorsal and
ventral."

UNIDENTIFIED SPECIES OF TABANUS.

Tahanus sp. Nos. 1 and 2. — Following Mann's observations, Kollar
succeeded in finding freshly laid eggs of another species of Tahanus,
which was not determirted, in a damp meadow at Dornbach near
Vienna, of which, after nine days, similar larvae hatched as were
obtained from the egg masses from Wippach, of Tahanus quatuorno-
tatus. Kollar also found a parasitic wasp ovipositing in the Tahanus
eggs; this wasp was found to be specifically different from that ob-
tained from the eggs of quatuornotatus .

At about the same time eggs of another undetermined species
were found by Hbfmann, quoted by Kollar, No descriptions are
given.

Tahanus sp. No. 3. — Riley and Johannsen (1915) give a photo-
graphic illustration of the egg mass of an American species of Tahanus,
the name of which is not given (Plate 1 , Fig. 6) .

Tahanus sp. No. 4. — (Hart's sp. a.) Two examples of larvas
of this peculiar species have been collected by Hart, from diverse
situations. One was taken under bark in woods near Urbana, Illi-
nois, April 6; the other, from a prairie ditch, in Kane County, Illinois,
which was swollen by a heavy rain. Hart's description is as follows:

"Larva. — Length 19 mm., diameter 2.5 mm. Last antennal joint short and
very slender, epistoma not sulcate anteriorly, but with an elongate puncture.
Whitish, lateral pubescent stripes wanting, annuli much reduced and pale except
upon false feet. Prothoraxshining, with anterior opaque annulus; lateral areas as
long as the dorsal, their upper and lower thirds rather coarsely striate, middle third
smooth, with several punctures; ventral area smooth, middle groove with three
striae, dorsal area nearly smooth. Striae of upper lateral spaces of meso- and meta-
thorax and of dorsal and ventral areas of abdomen moderately coarse; those of
lateral area of abdomen somewhat finer; dorsal and ventral areas of mesothorax
with a few striae; of metathorax rather sparsely striate. All areas more or less
shining. On the anterior side of each dorsal false foot, at its outer end, an
opaque light brown elongate fleck. False feet shining and rather finely striated
on each side. No projecting spine posteriorly; only a narrow pale annulus on
last segment, at base of breathing tube."

176

WERNER MARCHAND 177

Tabanus sp. No. 5. — (Hart's sp. b.) Hart reports that, in col-
lecting the larvae of Limnophila and Bittacomorpha in the swampy
slough of Station I, many small tabanid larvae were found in the mud
and debris, and among them occurred, on April 15, two examples of a
distinct very white form with faint markings like those of stygius,
but laterally striate more like atratus, and with a conspicuous isolated
smooth spot in the lateral striated area of the pro thorax. He de-
scribes the larva as follows:

"Larva. — [Plate 3, Fig. 42.1 Length 22-23 mm., width 2.5 mm. Very pale
whitish, markings like those of stygius, but pale yellowish fuscous and incon-
spicuous. Head pale brownish."

"Lateral prothoracic areas not more than half as long as the dorsal, striation
scarcely visible, microscopically fine and opaque, much finer than that of the
middle and lower lateral areas of the mesothorax, which are somewhat shining; a
rather large smooth spot included in the striated lateral area of the prothorax,
not far from the shining ventral area but entirely isolated from it; remaining
upper lateral thoracic areas distinctly more coarsely striated than the areas be-
low them, and quite shining; abdominal lateral areas more finely striate, feebly
shining, microscopically striate on the posterior portion of each area. Dorsal
and ventral areas shining, with sparse marginal striae interrupted on the disks,
those of thorax especially smooth."

"Dull annuli broad, including the false feet, a distinct transverse dorsal and
ventral pale spot in front of the false feet. Lateral prothoracic area occupied by
a very pale fuscous opaque quadrate spot in front of the striated area. Meso- and
metathoracic lateral stripes usually visible, but slender, not dilated, lateral edges
of dorsal areas diverging; lateral stripes of abdomen almost wanting, except on last
two or three segments. Last segment with bases of respiratory tube and anal
prominence ringed with opaque fuscous, that around anal prominence sending
up an indistinct stripe, with posterior extensions. Pubescence of false feet
whitish or pale fuscous. Respiratory tube slender, no spine protruding."

Tabanus sp. No. 6. — Malloch (1917) describes briefly, and without
illustrations, an unknown tabanid larva, which differed from other
larvae he had before him, in being entirely white and without lines or
patches of pubescence, as well as in being more robust, and less
tapered at the extremities. In general appearance it very closely
resembled an asilid larvae, the resemblance being accentuated by the
small size of the locomotor organs; and it stood as "Asilidse" in the
Illinois State Laboratory collection. The specimen was obtained in
Pulaski, Illinois, June 1, 1910, in a pit-cage used in rearing white

178 THE EARLY STAGES OF TABANID^

grubs. This is a surprising occurrence, as most of the species of
Tabanus are confined to damp ground or to aquatic surroundings.

In his analytical key, Malloch gives the following characteristics of
this larva, "Body without markings either of color or hairs. Body
very closely and finely striated, entirely white." The size of the
larva is not given.

Tabanus sp. No. 7 . — Unknown European species, observed by
Surcouf and Ricardo (1909). Only the pupa is figured (Plate 12,
Fig. 150).

Larva. — Surcouf and Ricardo have reported on the capture and
rearing of a tabanid larva, of which, however, the species was not
determined. Roubaud had given them "a young larva of Tabanus^'
which he had collected on October 3, 1905, at Meudon in the muck
of a pond; this larva was reared for eight months in water, feeding
not on living prey but on organic matter introduced with a bunch
of moss which covered the bottom of the crystallizing dish in which
it was kept. On May 13, 1907, it was seeking a protected place in
this moss, lost its mobility, turned more translucent than usual,
and, on May 16, had transformed into a pupa, which, unfortunately,
perished in consequence of an accident. The larva is described as
elongate, whitish, very active, pointed at both ends, and provided
with a sort of crested ring on each of its segments; these rings are
equipped all along their course with retractile tubercles serving as
means of locomotion.

Pupa (Plate 12, Fig. 150). — The pupa which Surcouf and Ricardo
obtained from this larva measured 17 mm. in length and 4 mm. in
diameter. This pupa has the aspect, according to the authors, of a
lepidopterous pupa; the upper anterior part was smooth above, com-
prising the thorax and head, the lower anterior parts bearing under
their chitinous coverings, the antennae, the palpi, the eyes, and the
first pair of legs. The wings are contained in a case which reaches
the top of the scutellar segment, unarmed. The posterior region of
the pupa comprises seven segments which have at the tip a crown of
rigid hairs intermingling with pointed tubercles with enlarged bases.
The last segment has two tubercles, each one consisting of three
irregular more or less curved spines. The third thoracic segment
and the first six abdominal segments all have a spiracle on each side.

WERNER MARCHAND • 179

These spiracles appear slightly projecting above the surface.

Tahanus sp. No. 8. — An unidentified African species. No de-
scription but good illustration of ventral and dorsal aspect of larva
is given by Brumpt (1910). The larva shows dorsal prolegs well
developed which indicates an aquatic mode of life (Plate 4, Fig. 55,
a, b, c).

Tahanus sp. No. 9. — Unidentified African species of which a poor
illustration is given by Griinberg (1907), (Plate 4, Fig. 58).

Tahanus sp. No. 10. — A young specimen of an unidentified Euro-
pean species is figured by Graber in his work on the chordotonal
organs (1882-83) (Plate 8, Fig. 99); it is supposed to belong to
Tahanus autumnalis, and may belong to Chrysops. (See also page
23.)

Tahanus sp. No. ii.— Henneguy (1904) gives the illustration of a
young Tahanus larva (Plate 10, Fig. 109), with typically inflated
tracheal trunks, in Europe. (See also pages 36 and 38.)

Tahanus sp. No. 12. — Paoli (1907) gives an illustration of the larva
(Plate 10, Figs. Ill to 116) in which he studied Graber's organ.
Though this larva is supposed to belong either to Tahanus cordiger or
to Tahanus autumnalis, it is not unlikely that it is another species.
The two species mentioned happened to be, with Tahanus hovinus
and Tahanus quadrinotatus, and one or two others, the only European
species which had been found in the larval state up to that date,
but the descriptions are all insufficient.

The larvae which were studied by Paoli live in slow flowing rivers,
especially in the soft mud containing decomposing organic matter,
among the algae, especially CharacecE, which grow in the water, in
places where they can easily utilize the atmospheric air as they are
not provided with organs adapted to locomotion in water.

The body consists of thirteen distinct segments, of which one is the
head, three belong to the thorax, and nine to the abdomen; the tenth
segment of the abdomen is much reduced and fused with the ninth;
the eleventh is absent.

180 THE EARLY STAGES OF TABANID^

Graber, as Paoli mentions, gives his larva (in the figure) erro-
neously, only ten segments, while it is known that the larvae of Diptera
have fourteen segments, of which the last one is vestigial.

The largest specimens of the larvae examined are, when completely
extended, 30 to 32 mm. in length. The body is fusiform (Plate 10,
Fig. Ill), the color a dirty yellowish white, the head is elongate,
retractile; the three thoracic segments do not present any peculiari-
ties. The first seven abdominal segments are furnished each one
with eight papillae, arranged all around the anterior third of each seg-
ment, which consequently in cross-section presents itself with eight
prominences arranged like the corners of a regular octagon. These
papillae are retractile, and each of them is furnished at the tip with
numerous strong booklets. The larva moves in the mud and among
the algae, in addition to making worm-like movements due to the con-
tractions of the subcutaneous muscular stratum, also by protracting
and retracting the fifty-six papillae which in this way form a support
and assure to the larva its locomotion. The eighth segment bears
on its ventral side a thick sulcate prominence, in the middle of which
the opening of the anus is situated. Anterior to this prominence a
large lip-like expansion is formed, with numerous strong booklets
which are curved forward at the free margin. Behind the anal
opening the eighth segment is narrow and curved upwards; he
ninth segment is short, cylindric, thin, more or less retracted into the
inside of the eighth, bearing at its extremity the rudiments of the
tenth, in which the two stigmata open.

Paoli, after thus briefly describing the larvae, adds that when dis-
turbed by contact or pressure, they undergo contortions in every di-
rection, making a sound like the crackling of small electric sparks.
They feed on animal matter and ferociously attack small larvae of
other aquatic insects, and also devour one another; when they are
attacking one another the crackling sounds are heard repeatedly.
For the description of Graber's organ in these larvae see page 29 and
following.

Tabanus sp. No. 13. — Maxwell-Leffroy and Hewlett (1909) give, on
a colored plate, illustrations of the egg mass, young larva, full grown
larva, pupa, and an ichneumonid parasite of an unidentified Indian

WERNER MARCHAND 181

species of Tabanus. The egg mass (Plate 1, Fig. 7) is yellow in the
original and not very convex in structure. The young larva (Plate
5, Fig. 76, a), is whitish and semitransparent. The full grown larva
(Plate 5, Fig. 76, b) is yellow, its natural size 47 mm. The pupa is
yellow in color, the thorax grayish (Plate 12, Fig. 139). The hymen-
op terous parasite figured (Plate 5, Fig. 76, c) is the only parasite
of this (ichneumonid) type observed so far as a tabanid parasite,
the other being cases of Proctotrupidae.

Tabanus sp. No. 14. — Only the pupa of this Indian species (no
name given) is figured by Maxwell-Leffroy and Howlett (1909) (Plate
11, Fig. 132).

Eggs and young larvae described by Lutz (1914) of an unknown
Brazilian species which he supposes may belong to the tabanids,
probably belong to a leptid species.

PARASITES OF THE EARLY STAGES OF TABANIDiE.

Hymenopterous Egg Parasites; Early Discoveries.

Kollar (1854) is the first to mention an egg parasite of Tahanus.
From egg masses of Tabanus quatuornotatus collected at Wippach,
Austria, by Mann, June 25, 1854, there were observed to hatch not
only young Tahanus larvae but "also another completely developed
insect belonging to an entirely different order and family, a little
animal which belongs to the extraordinarily large and still inextric-
able army of the parasitic wasps (Ichneumonidae)." It was con-
cluded that, immediately after the Tahanus had deposited its eggs, a
parasitic wasp had appeared and deposited its eggs in those of the
Tahanus. Parasitic wasps were then known to develop in the eggs
of Lepidoptera, chiefly Bombycidae, and of various Hemiptera, but in
Diptera such egg parasites had not been observed (Kollar). The
length of one of these parasites is given as two-thirds of a line; the
size of the Tahanus eggs as one line in length, and one-seventh of a
line in diameter; Kollar thinks it probable that several wasps develop
from a single Tahanus egg, which is probably not the case.^^

On egg masses of a second species of Tahanus, collected by himself
at Dornbach, near Vienna, Kollar found a similar but specifically
different parasitic wasp, which was just ovipositing its eggs in those
of the Tahanus. In order to learn how much time was necessary for
the development of the wasp, Kollar kept the Tahanus eggs in a
small bottle closed by means of perforated paper, and obtained on
the fifteenth day several hundred wasps. From the number of
parasites he believed he was justified in assuming that from one
Tahanus egg more than one wasp may develop. However, the num-
ber of eggs in the cluster had been said, for Tahanus quatuornotatus
Meig. to be 350 to 400, and it is not said whether the number of
wasps hatching exceeded the number of eggs.

The two sexes of the parasites could be differentiated ; a description
and figures were promised but apparently have never been published.

^' See Hart (1895), Phanurus tahanivorus, p. 272.

182

WERNER MARCHAND 183

Hart (1895) is the next observer who noticed the egg parasites of
Tabanidas. Two egg masses of Tabanus atratus which had already
produced larvae were placed in a dry vial, and a little later it became
evident that both masses had been parasitized by Hymenoptera,
minute black images emerging freely in the vial. An examination
of one of the masses showed that about one-half of the eggs had been
infested. Examples of the imago were sent to Mr. W. H. Ashmead,
who found the species to be a new one; and it was described by him
as Phanurus tabanivorus (Ashmead). An egg of the Tabanus is fig-
ured, containing one imago of the parasite. ^^

The description of the parasite by Ashmead is given here:

Phanurus tabanivorus (n. s p., female). — "Polished black, impunctate, the head
and thorax clothed with a fine sparse pubescence. Head subquadrate, roundly
emarginate behind, a little wider than the thorax; eyes oval, faintly pubescent;
antennae eleven-jointed, black, if extended backwards not quite reaching to the
apex of thorax, and terminating in a long fusiform five-jointed club, the first
joint of which is not quite as wide as the second; ob-trapezoidal, twice as wide
as long, the second, third and fourth joints transverse quadrate, a little wider
than long; the fifth or last joint conical and a little narrower than the preceding
joint; the scape is about as long as the funicle with the pedicel, the latter obconi-
cal; joints of funicle a little narrower than the apex of the pedicel, the first
joint scarcely longer than thick, the second and third small, transverse
moniliform."

"Thorax subovoid, not twice as long as wide, the mesonotum scarcely longer
than wide, the scutellum lunate, polished, without pubescence; wings hyaline,
ciliated, the cilia on the anterior and posterior margins long, much shorter at
apical margin; tegulae black; venation brown, the marginal vein a little shorter
than the stigmal, the latter only slightly thickened at tip, the postmarginal vein
very long, fully two and a half times as long as the stigmal; legs fuscous, the
trochanters, knees, tips of tibiae and tarsi honey-yellow or testaceous. Abdomen
elongate, pointed fusiform, about twice as long as the head and thorax united,
polished, the first segment not longer than wide, with an elevation above at base,
the second segment the longest, twice as long as wide at apex, the suture be-
tween it and the first striated, the third segment hardly half as long as the second,
the fourth about two-thirds the length of the third, the three following forming a
cone of which the fifth is very short, its apical margin with a median sinus, the
sixth twice as long as the fifth, the seventh very short, scarcely discernible;
sheaths of ovipositor a little prominent."

*^ Ashmead's illustrations (Figs. 55 and 56 of his paper) have been omitted.

184 THE EARLY STAGES OF TABANID^

"Male. — Length 0.8 mm. Black, but with the head, presternum, and legs
testaceous; the antennae twelve-jointed, brown- black, with all the joints of the
flagellum, except the pedicel and the last joint, small, moniliform, joints three to
five increasing in size but smaller than pedicel, joints six to the penultimate a
little wider than long; abdomen not, or only slightly, longer than the head and
thorax united, the genitalia long, exserted, curving downwards."

"Habitat. — Havana, Illinois."

"Types in collections of the Illinois State Laboratory of Natural History and
in my (Ashmead's) collection."

"Described from eleven male and nine female specimens bred from the eggs of a
common horse-fly, Tahanus atratus Fabr."

Hine (1906) also has observed Phanurus tahanivorus and reports
that he reared more than a hundred specimens from a single cluster
of eggs.

Phanurus tabani Mayr. — In his monograph of the N. A. Procto-
trypidas,^' Ashmead characterized four species belonging to this
genus, none of which, however, are closely allied to Phanurus tahani-
vorus. On the contrary, it appears to resemble more closely three
European species described under the genus Telenomus; viz., Teleno-
mus othus Hal., Telenomus laricis Hal., and Telenomus tabani Mayr.

Dr. Gustav Mayr, *4n his excellent revision of the European
species of Telenomus" did not recognize the validity of Thomson's
genus Phanurus.

Ashmead, however, believes that, as defined in his monograph, the
genus can be readily separated from Telenomus, although both Hali-
day and Walker have described Telenomi which should now be rele-
gated to Phanurus; while Thomson, in his definition of Telenomus,
has included species that really belong to the genus Hadronotus
Forster.

These errors, according to Ashmead, probably induced Dr. Mayr
to reject the validity of Phanurus in his work (cited above).

It is interesting to note that Phanurus (Telenomus) tabani Ma)n:
has habits similar to Phanurus tabanivorus, the species having been
bred by Brauer from a European horse-fly, Tabanus sp.

Phanurus tabanivorus, although evidently related, is somewhat
larger in the female sex, smoother, with the joints of the antennae

''Ashmead, N. A. Proctotrypidce, Monograph, pp. 140 and 141.

WERNER MARCHAND 185

and the segments of the abdomen relatively different, while the male
is much smaller, differently colored, and with the antennal Joints
totally dissimilar.

Dr. Mayr's species, Phanurus tabani, approaches nearest to Tele-
nomus laricis Hal.,'* with which he makes comparison; while Phanurus
tahanivorus agrees more nearly with Telenomus othus Hal., repre-
sented on the same plate, Fig. 4 (Mayr's paper).

Phanurus emersoni Girault. — A new Phanurus egg parasite of
Tabanidae was discovered by Girault quite recently (1916), Phanurus
emersoni Girault, and was reared from tabanid eggs at Dallas, Texas.
Girault gives descriptions of three allied species.

1. Phanurus opacus Howard. — Both sexes are black; the thorax
above is subglabrous.

2. Phanurus floridanus Ashmead. — The head and thorax are pol-
ished, the tibiae and knees pale brown; segments one and two of
abdomen have very short striae at base. The club is stouter than
in ovivorus.

3. Phanurus ovivorus Ashmead. — The club is slenderer than in the
preceding, the tibiae dark, the thorax above showing faint reticulation,
cephalad but mostly glabrous. The first two segments of the abdo-
men do not have striae at base, or else these are extremely minute and
short. In flavipes the vertex and scutellum are uniformly finely
reticulate. The species ovivorus is very close to opacus, if not identical
with it.

4. Phanurus emersoni Girault. — Female, length 0.90 mm. Black,
the wings subhyaline, the venation pale dusky, the tarsi yellow.
It differs from Phanurus opacus Howard in that the male is vari-
colored in the latter, and from the female opacus, floridanus, and
ovivorus in that the vertex and scutum of the latter are densely reticu-
lated. It differs from tahanivorus in that the abdomen of the latter
is only somewhat longer than the rest of the body, the third segment
is not a fourth the length of the third (second?), the thorax above is
reticulated, and the male has the entire thorax honey-yellow, also the
antennae (besides the legs and head as in tahanivorus). It is closely
related to the female of ovivorus, which it resembles. The stigmal

'* Mayr, Entomological Magazine, ii, Plate XIII, Fig. 2,

186 THE EARLY STAGES OF TABANID^

vein is nearly twice as long as the marginal, and about half the
length of the postmarginal. Funicle 1 is half as long again as it
is wide, two-thirds the length of the pedicel. No. 2 a little shorter
than No. 1, No. 3 still shorter. No. 4 globular, smallest, No. 5 cup-
shaped. No. 6 the same, larger, wider than long, Nos. 7 and 8 sub-
quadrate. No. 9 ovate, longer than wide. Short distinct striag at
base of segment No. 2 of the abdomen.

In the male, Funicles 1 to 3 are somewhat longer than in the fe-
male, while Nos. 4 to 9 are moniliform, wider than long, small; the
club joint is ovate and as long as Funicle 1 and stouter.

Described from a large number of both sexes reared from tabanid
eggs at Dallas, Texas (Bishop),

Types. — Catalogue No. 19,664, U. S. N. M., one male, eight females,
on two tags and a slide bearing one male, four females.

Types of opacus, ovivorus, floridanus, and fiavipes examined.

Telenomus benefactor Crawford. — Patton and Cragg (1913), (page
298), say that Telenomus benefactor Crawford is another chalcid
which parasitizes the eggs of tabanids in the Sudan.

Telenomus kingi Crawford. — This species has been reared from eggs
of Tabanus kingi.

Unidentified Parasites Recorded.

Chalcid sp. 1. King (1910) has bred, from an egg mass of Tabanus
tceniola, taken at Gebelein, numbers of a small Hymenopteron, which,
in 1910, had not been identified, but was figured, together with the
parasitized egg mass, showing the exit hole of the parasites. As the
insects are figured in natural size, not much can be said about struc-
tural details (see Plate 1, Fig. 8, and text, p. 170).

Chalcid sp. 2. In Madras, a similar insect, which has not been iden-
tified (Patton and Cragg, 1913), regularly destroys large numbers of
egg masses of Tabanus albimedius and Tabanus striatus.

A parasitized egg mass can be recognized, according to Patton and
Cragg, by the almost black color which it assumes when the devel-
opment of the embryos of the Hymenopteron is almost complete.

Chalcid sp. 3. A hymenopterous parasite has also been found to
infest the eggs of Goniops chrysocoma (McAtee). Consequently the
phenomenon seems to be quite general.

WERNER MARCHAND 187

Ichneumonid sp. An unidentified ichneumonid parasite of an uni-
dentified Indian species of Tahanus is figured (Plate 5, Fig. 76, c) by
Maxwell-Leffroy and Howlett (1909), apparently the only known
instance- of a true ichneumonid parasitic in Tabanidae.^^

Parasites of the Larvce.

On parasites of the larvae we possess only a few remarks by Hart
(1895), who found a larva covered with small whitish scales which
were possibly the eggs of a parasite. Nothing further is known.^^

^^On an egg mass of Tabanus atratus F., found in 1915, I noticed a small
Hymenopteron which I took for an egg parasite and which I found to belong to
Tricho gramma. However, no Trichogranima hatched from the egg mass, but
instead of it, a large number of the well known Phannrus, apparently tabanivorus
Ashmead.

^^ To these should be added the author's own observation of Nematode para-
sites which were repeatedly found in larvae of Chrysops.

DISCUSSION OF TABLE II.

Of 91 species data are available;" but this is not much if we con-
sider that more than 2,000 species of Tabanidae have been described,
and 76 species have been recorded for the State of New Jersey alone.
Moreover, in a number of these 91 species from all parts of the
world, we have only very fragmentary data, as for instance in the
twelve species of Chrysops listed; of five of them we possess only
notes on their oviposition ; the whole life history of none of them has
been worked out. There are also comparatively few data on European
species.^^

From the list it is evident that the majority of Tabanidae of which
the life history is known, are aquatic in habitat: of 91 species noted,
65 are aquatic or probably so, 18 are terrestrial or probably so; of 8
species nothing definite can be said about their habits. The terres-
trial habit is then to be considered the exception. It is worthy of
note that all the known species of Chrysops are aquatic, in as far
as the eggs are deposited above water, or the larvae have been found
in the mud. The limit between an aquatic and terrestrial mode of
life is, of course, not always very sharp when the habitat is in fact
the soft mud at the border of ponds, brooks, and streams, as seems
to be usually the case, but most authors agree that the larvae live in
mud saturated with water, an environment which differs physio-
logically very little from water itself. I have therefore listed all
species as aquatic the larvae of which have been found in the mud
near water, or where the eggs are deposited above water. But there
is no doubt that some tabanids live in the larval condition at con-

^^ Of two or three of the 91 species discussed here, we have in fact only in-
dications about their breeding habits, no actual knowledge of the stages, so in
Gasir oxides, Tabanus fuscipes, and Tabanus nigerrimus. Leaving these out, we
still have about 88 species on which some data are available.

^^ Recently, I have obtained larval stages of 9 other North American species,
bringing the total number of species on which data are known close to 100.
These species are: Chrysops niger, obsoletus, univitlatus, Tabanus reinwardtii,
pumilus, orioii, two undetermined species of Tabanus, and one gemis incertum.

188

WERNER MARCHAND 189

siderable distance from water and probably a greater number than is
apparent from this table, because terrestrial larvae will be found
much more rarely being less accessible and spread over wider areas.
Since the genus Tabanus is on the whole phylogenetically the
younger, in comparison with Chrysops, it would appear that in some
species of this genus the larvae have become secondarily terrestrial, a
possibility which is of importance for the understanding of the group
as a whole. This may also apply to Hcematopota, a genus absent
in Australia and apparently of more recent origin — while retaining a
rather primitive wing-pattern — showing aquatic tendencies in the
oviposition of some species, yet in its common and widely spread
species, HcBmatopota pluvialis, completely terrestrial in the larval and
pupal stages.

Practically nothing is known, with the exception of the few in-
stances listed, of the early stages of the numerous remaining genera
of the family Tabanidas other than Chrysops and Tabanus proper,
and here is a splendid opportunity especially for workers in tropical
countries.

In regard to continents, the data in our possession are distributed
as follows:

(Two identical with European; viz., African
species.)

It is seen that Africa ranges highest, and concerning this particular
subject, has ceased to be a "dark continent," as compared with
North America or Europe.

Africa

29 species,

North America

24 "

Europe

21 "

Asia

15 "

South America

4 "

Australia

0 "

190

THE EARLY STAGES OF TABANIDiE

ri. (V. ru

rv. r^ ck- Bj

3 ;

a*

<

-III I 2S I

:z: ^

00 lo S — '
'^ -' — On

"3 a) =2 u

m W S ^

I I

.O 12 .r, ^ g O JO

g g S 2 2 2 2

- N l7 _

.s
^ ^ ^ -^ d fe "S

;z; :z; Iz; cAi pq CL, pq

to

On I I I I On 0>

:z; ^

h-1 z:^

=0 2

»— 1 '"' On On On

ID lO lO

^ On O "^

5n On X) 22 =» 2

cj

4J~ > >

w ^ :^ ^

(U <u u

" ^ ^ ^^ ^ >^ ^ 2 fei;

fe^

-73

T-i ro

On lO ro

C

- O 1

1 On On O

c3

M On

1 00 OO On

n

M"^

r-( s— * *— (

03

I-. aj

(U -i-i D

<-^ .S

.s s .s

Oh

W

ffi W K

I I

°o 2

o +->

MM

e3

<3

to

<:

o i:
^ <

S
of

< ^

W

■2,

I^ s. <. <<

u s^ a r^

< e

a "«

C5 CJ3

<^

. . 3

< tS ^ "^

c^ ^ -S .5

.,-1 CS rO ■>*

lO O t^ 00 On

NO t^ 00 On

WERNER MARCHAND

191

00 fO
00 oo

S P3

^ Pi ffi W ^ ^ c^

dn W ;z; P « C/D S

00 <^
oo 00

ui {- bC

2 ^ O ^ ^ „

_ OO g; ON o ^
22 °° °^ ,i3

Ov

fco

u

s«

"2 JCJ 00 o\

u :3 rS 53 .S 3 s

^ Pi ffi ffi s ^

00 00 <u

>> vT

5 :0

m ■ C O C bp

"2 2 CO ^

M 2; '"'

u, g g fl

t^ 22
00

^ § 1^

° j: bo >

Q m

-0

Ov

TJ

UN

cJ

hf)

rt

hr

ri

U)

n

bO

0

2

0

2

u

u

£ i I

"O _i 10 NO

C ^ On O

ra bC 00 On

j- bC '^ '^

3 o .y

M ,3

b ^

CO t* On

S3 bo ■^

S 2 bB

^ ^ -S

Ph fc<i

«

£ ^

hi; <

12;

oi

^^

192

THE EARLY STAGES OF TABANIDiE

6 -C

3 >-

2 a 3 ii = 1-

< H < H

2 " 3 =

•^ ^;:.^

0\

c 2 "^

oT t3 S

C30 ™ JJ On

O T3 S 2 is

5 «s

f> *^ j5 *-? F^^ *^

<u .y

PP

rt o

1^ S K pq

"2 U o

^ (S

. >.

pq

"S

J-" o ^o r;" lo ^

Ov »-( o Ov Ox

■^ OS Os '"' 00 r<"

3

o

Q

Neave,
King, 1
Hine, 1
Neave,
Hart, 1
Mallocl

so

?1 ■^ S O^ M - —< O

Z « PL, K m M Oh

o NO r* lo

1

y,

On — 1 o •'^ Os

1

i)

'-I On On '-* 00

3

o

> be iT > tT

2 .S .S S S

Q

'i^UiK^K

<^

On O
bo 4;

O NO

1 Sg 1 1

r^

be 3

ro

o

2 O

bo 4)"

o-s

.S .3

Q^

Q

w a

ffi

fl W)

I I I I I I I

•5 -5

^ .<«;.«• .

a>

< <

«« J

jartjGirt&irt-:

a

ea

:S s-a

♦juTjOUOU.s

o

<: i-H

Oi!o3vt2 3i!a

:2;<:z;W<W<:m

3

w

•C '

<

Nor
Afri
Nor

^^

cr
o

8

•SP

o »

.bp
'53

^

Pi

3 ^

S?-2

«3 a 1/2 rt
1^

*» «c:,st,^— -nC, «>« <«)-«:

•S ^ ^ ^ i2

NO l^ 00 On O

^* ^* ^* ^t^ 'O

WERNER MARCHAND

193

n.

(V.

o-

[b

(^

_«

[o

.y

1

'C

U

(

^t.

'E

_(.

'E

L

u

^j

t/

«

rt

1

<u

nJ

1

c

?^

a

<u

n

3 2

s

tl

3

^

2

3

S

fc!

3

2

S

3

3

ty

HJ

C

O*

a>

cr

u

O'

<

H <

<

H

<

H

<

r<~

0\ o

On

O

00

O

^

1

c

1

1

a
a

c

15

NO

2

OC

1

1

O

»D r^

On O

> >

N

a

00

u
o

>

__ O

1

be

1

a;
3

. 1

1

00 On

p

03 oj

c

3
1-1

u

m

„

f<i

On O

s

On

r

1

On

■*

1

c

ID

On

ID
OC

•D

o
o.

a
, _o

C

C

NO

2

1

OC
00

o

c

ID fO

On O

oT 4)

> >

1

00
M

nj

O

3

1

bo >

.S §3

"o

p

(U

1

c
c

bB

CO

u

3

2

1

00 On

13.2

so
p

!z; ^

m

w

1^

ui^'^

h-I

m

P.

w

'^

w a

O

c

c
a
a

ID~

o

OS

t:

2

o

, T3

On

c

1^

<~o

1 1

1

1

1

1

1

1

1

*-

_j 1

1

c

1

c

a

bO

ii

o

1 1

1

1

1

1

1

1

ID

00

_o
'3

no"

1

C

s

bO

2

1

S

a
o

bO
bO

2

."2

(5

J3
o

On

pq

"c

fc4

N^^

kJ

p

Ph

^

u

'3
pq

S

i

d

(1

0

d

•iH

'C

>^

'C

'E

M

a;

(U

OJ

<u

V

E

<

<u

1

<

<u

1

«j

<u

1

ei

^

0

a-c

p

^

.2

rt

O,

^

a. ^

a ^

^

.y "•

c

o

_t

_C-

o

o

o

4->

<

o

<

3
O

c

3

O

3

1

3
W

C

1

c

r

>

3

a

_b
'Z

o

c4

W

-^

SI

a <
^ s

■2 <a
a a

to

i

8

s

£
s

1

"J

s;

a
S

8

4J

H
<o

s

<o
S

t.

H-1

-2

3

i

a

Q
1 C)

<-

i

(J

•2

tl

_bi

■ft

)
1^

/■S

«1

1

0

8 ^

«

13 S

N*

-O

e5

■^ CN

rn

Tt

ID

NO

r--

OO

O

tN

r'i

•>*

u-

NO

«--

00

1/-

t/>

tn

ir

>D

lO

ID

>D

ID

NO

■o

NO

NO

VC

NC

VO

<o

194

THE EARLY STAGES OF TABANID^

i

ru

—

rt

fv-

.2 ^

rt-

.3

cJ n.

"o

'S ^

■^ 1

1

t-

_t

■£ 0

« 'Z!

tn

1

a

tn 'S

OJ oi

(U 1

i)

rt

aj (Tj

•s

^ ^ ^

^ S ^

- t!

3

fcl 3 2

ffi

a^ CT

a;

(U

O* aj o- 1

H <

H

H

<:

H <

Ov

0

fO

0

Os

10

10 10

0\

"O '"'

VO O

Tf OC

■.-1 -H T3

■.-1 f^

1 rt -2" 1

<«'

O "-1

On CN d

0 1 1

1

1

1

Q.

0\ Ov

ol

'-H r-l nJ

bB OS 1 1

1

1

1

1 J! ^ I

■ S

1-H •r-l

iS

aT oT c

Sf '^

=3 r?

iT U) kT "c

> > 2

0 .y

1 -s

Hi

'^ .3

5^

Ov

ro

t^

S

t^ t^ Os 0

^

o

Ov ^ -^ ,

"^ "^ -rt

Ov OS C

S ro ^
0 00 1

1

OS

Os
0\

Os Os "5 Os

King, 19
Neave, 1
Lutz, 19
Brimley,

Neave, 1
Neave, 1
Patton a

Cragg,
Hine, 19'
KoUar, 1

I

Hart, 18'
Malloch,
Hart, 18'
Malloch,
MaUoch,
Surcouf i
Ricard
Brumpt,

10

fr>

10 '

Os

o

ro

TS

X, ro b2 0

s^':-

bfi

1 r'

1 0 1

1 1 S

.0 00 ffi
bC Ov ■" Cl

- -c

S

1

1

1 1 1

a

1 2;

1 Oj 1

1 1 «>

'?s

tn

1

1

1 1 1

M

a

2f '-' uT I.

d

bJB iT

i

2 oT J2 iS

1 ^ s

c

a

a

U .S ^ ^

f5

-d

5

d

"C

•c -

•c

•c

*M

<u

4) ■•

u

<u

«J

'.3

►2

a
< ^

1.

a

<

1. .

S

-rl 0

45 J3 O- M rt

M (X

^

ja 0. M

1-" y

4J *-> 0

.^ 0 s. 2

4-. 0 -

*j

*J 0 <J

52; <

3 fc *-

1— 1 <. hH

1

Nor
Eur

Afri

>

>

3
cJ
aj

m

5

aj
-0

.ec

T3 — •

cr

aj nJ

8

0 0!

8^

1 1 1
■S 5 1

-c
_a

a;

'"'
C

c

c

c

C

c

c

OC

c

8 i -S

6 1^ :2

Si <o

:2

r>-.

a

8 ^

« 2

w ^ S

« S ^

^ ^ ^

^

^

5

S S S

.0

a

f-i

Ov C

^ CN) r''

Tt LO so

t^ 00 0

C

,_H

CN

PO ■<* 10

o
2;

•c

i-«

t^

t^ t-^

t^

t~-

r^

r~

t--

t--

OC

OC

OC

OC

OC

OC

)

WERNER MARCHAND

195

1

i

^

sa

O

!tt

O

a;

'T-' -^

<u

i-pi

I 1

1

1

ii

13

K

^ i^"

X

>.

j-T

rt

p OJ

nS

o

1)

§

M • — 1

^

»H

Tt<

O ^

O

OS

1

^

0\ tvi

to

o

■^ 00

- 00

^

(^

MH (^

hoi-i

>.

o

-o o

1

3

Ch

c 2

rt ^

1

5 uT

W)

^-H

o

XJ a

(U

„

><

Griin

Grab

'83

C

r3
o

ri"

~

1
&

!tl

o

1 1

1

1

Ss

1

1 1

1

1

=)

C CK

1

(U

(Sj '"'

^

t«!

>. tf

c3

2 aj

^

>H ,_^

NOTES ON METHODS OF REARING AND STUDYING TABANIDS IN

EARLY STAGES.

Collecting the Larvce. — The natural habitat of tabanid larvae is the
wet mud or sand in the immediate neighborhood of water, hence
they are rarely caught by the usual method of collecting aquatic in-
sects by means of a net. The lack of a practical collecting method is
probably the main reason why these animals while very common
have passed almost unnoticed in an epoch of fresh water biological
investigations. Neave seems to have employed large nets, as did
Hart also, but Neave was able to send out a collecting staff of ne-
groes to obtain his material. The use of an ordinary sieve, first ap-
plied by Hine to collect the larvae of Tahanns vivax in rapid streams,
is more advisable for collecting the larvae from the mud. Patton
and Cragg suggest placing lumps of mud and sand with water in a
pail; by stirring the mass the large larvae appear floating at the sur-
face. The smaller ones are caught when the muddy water is passed
through a sieve.

Rearing Vessels. — The methods of various authors in rearing taba-
nid larvae have been treated in connection with the species. Thus in
the paragraphs on Tahanus lasiophthaknus and vivax, details of Hine's
method in rearing these species are found, and Mitzmain's valuable
contributions to breeding technique are found under Tahanus striatus.

Hine proposes to keep the larvae in jelly glasses the covers of which
are perforated by a few holes, and which are half filled with wet
sand. Mitzmain uses dishes with some mud and wet filter paper.

Patton and Cragg have advised the use of very large trays, several
feet long and six inches high, in which large numbers of larvae may
be reared under almost natural conditions. The bottom of the tray
has to have a hole closed by a cork stopper, to make it possible to
change the water in the tray from time to time. Pupae should be
taken out, as the larvae may injure them, and placed in separate
cages, in little vertical holes. Near each pupa a little flag is fastened,
bearing the number of the pupa, or other data.

196

WERNER MARCHAND 197

Neave, who was raising many different tabanid species in southern
Nyasaland, found that the best receptacles which he could obtain
locally were the small basin-shaped vessels made of hard clay by the
natives. These were of various sizes, from six inches to a foot in
diameter. They were placed separately under cages made of mos-
quito netting on a wood framework. Neave recovered the larvae
for examination or other purposes merely by washing them out of
the mud or sand in which they were placed. This of course requires
great care in the case of the smaller species.^^

Feeding. — While most authors give earthworms as food to the larger
larvae, Neave advises the use of immature larvae of muscid flies,
collected from the carcasses of rats, etc., trapped for the purpose.
These larvas buried themselves at once in the mud, where they were
apparently consumed by the tabanid larvae, which thrived under these
conditions. The larger species also greedily attacked the freshly
killed bodies of small tadpoles, mollusks, and bits of fish, placed on the
surface of the mud, though they were seldom actually seen to do
this unless examined at night. It was found in most cases that the
tabanid larvae did best in mud or sand, this point being usually de-
cided by the conditions under which they had been found, which was
very wet, but without standing water on the surface.

Transportation. — If it is necessary to transport for any distance
larvas which have not reached the resting stage, it is important that
the jars, etc., in which they are placed should contain only wet mud
or sand and that there should be no standing water on the surface.
Some of Neave' s earhest captures, which had to be transported 50
or 60 miles from the Shire River to Mlanje, were nearly all lost from
this cause, the larvas being apparently drowned by the movements
of the water on the surface.

Treatment of Pupal Stages. — ^Neave found that the soil in which the
pupae are kept should be considerably drier than that which suits
the larvae.

Patton and Cragg have proposed to remove the fresh pupae from
the vessels in which the larvae are kept and to place them vertically

'^ For rearing tabanid larvae in test-tubes, see Marchand, /. Econ. Entomol.,
1917, X, 469-72.

198 THE EARLY STAGES OF TABANID^

in open holes in sand, where they wriggle themselves soon into the
desired position. Each pupa receives a label in the form of a little
flag which is fastened in the sand. As soon as a pupa has hatched, it
is taken out with its label.

According to Neave, it is difficult to obtain good specimens from
bred material, as the imagos begin to fly about and to injure them-
selves before they are hard enough to be killed. They should there-
fore be placed in larger cages and kept there until they are sufficiently
hardened.

Points Important in Description. — Specific differences in tabanid
larvse are not always easy to detect, especially in those of Hamato-
pota. They are generally found, according to Neave, in the distri-
bution of the pigmented areas on the last segment around the base
of the syphon and the anus. These so called pigmented areas are
really areas of pigmented hairs (Waterston) in which are entangled
small foreign bodies. Their actual color therefore varies to some
eictent with that of the medium in which the larvae have lived. The
amount of pigmentation, though not its distribution, also varies with
the age of the larvae.

Hart considers the striation of the lateral areas of the prothorax of
great systematic importance in the description of the larvae.

For practical purposes of description the following points should
be noted (Patton and Cragg): In the larvae, abdominal markings,
presence or absence of striae,^" characteristics of the pseudopods
(prolegs), length of the syphon tube, and structure of the antennae.
In the pupae, length of the antennal sheath, character of the thoracic
spiracle, particularly its inner margin, length of the hairs on the
abdomen, structure of the abdominal spiracles, and shape and size
of the spines and teeth on the eighth segment.

Mr. James Waterston, of the Imperial British Bureau of Ento-
mology, is quoted by Neave as having suggested the convenient
name *' aster" for the group of hooks at the termination of the last
segment of the pupa. The form of this differs a good deal in the
various species, and another characteristic which seems of some
specific value is the nature of the uppermost section, often isolated,

^ Especially on the thoracic segments.

WERNER MARCHAND 199

of the series of combs on the anterior part of the last segment; this is
called by Neave the dorsolateral comb. It is not always present
and varies much in form.

Preserving. — To preserve the larvae and pupae, for purposes of dis-
section, etc., our methods are still undeveloped. Harf^ (1895) says:
"The best results with most larvae of any size were obtained by
heating them in water, not too rapidly, to about 200° F., and setting
them aside till cool. A small percentage of acetic acid will prevent
the collapsing of very soft larvas." The principal trouble with this
method arises from the expansion of the air within, but a slight infla-
tion is sometimes desirable. This method is not suitable for pupas
generally, and Hart advises the use of 80 per cent alcohol and water
for their preservation. Experiments with formalin indicate that it
will satisfactorily preserve small and easily penetrable forms.

Malloch (1917) has given further methods for preserving the larvse
and pupae.

*i Hart (1895), pp. 159 and 160.

BIBLIOGRAPHY.*

Ashmead, W. H., Descriptions of three new parasitic Hymenoptera from the
Illinois River. Phanurus tabanivorus , Bull. Illinois State Lab. Nat. Hist.,
1895-96, iv, 274-276. (See Hart (1895).)

Bainbridge, T., and Fletcher, Some South-Indian Insects, Madras, 1914.

Baldrey, F. S. H., The evolution of Trypanosoma evansi through the fly: Tabanus
and Stomoxys, J. Trap. Vet. Sc, 1911-12, vi-vii, 271-282.

Beling, T., Beitrag zur Metamorphose der zweifluegeligen Insecten, Arch.
Naturges., 1875, xli, 35-39.

Beling, Beitrag zur Metamorphose zweifliigeliger Insecten aus den Familien
Tabanidae, Leptidae, AsiHdae, Empidae, Dolichopodidae, and Syrphidae, Arch.
Naturges., 1882, xlviii, 187-240; C. relictus, pp. 187-190.

Bodkin, G. E., and Cleare, L. D., Notes on some animal parasites in British
Guiana, Bull. Entomol. Research, 1916, vii, 179-190.

Brauer, F., Beobachtung Marno's xiher Hexatoma, Verhandl. zool. hot. Ges., 1868,
xviii, 74-75.

Brauer, Kurze Characteristik der Dipteren-Larven, Verhandl. zool. bot. Ges.,
1869, xix, 448.

Brauer, Beitrag zur Verwandlungsgeschichte der Regenbreme {Hcematopota
pluvialis L.), Verhandl. zool. bot. Ges., 1869, xix, 921-922.

Brauer, Die Zweifliigler des Kaiserlichen Museums zu Wien. III. Die Tabanus-
Arten der europaeischen, mediterranen und sibirischen Subregionen, Denkschr.
kaiser. Akad. Wissensch., Mathem. Naturw. Classe, 1880, xlii, 151 (pupa
of Tabanus solstitialis) .

Brauer, Die Zweifliigler des Kaiserlichen Museums zu Wien. Ill, 1883, p. 24,
Tabanidai. (Contains description of larva of Hexatoma pellucens, illustra-
tions of the latter and of larvae of Tabanus spodopterus and cordiger.)

Brimley, C. S., Notes on the life-histories of Tabanidse (Diptera), Entomol. News,
1909, xxii, 133.

Brumpt, E., Precis de parasitologie, Paris, 1910.

Degeer, C, Memoires pour servir a I'histoire des insectes, 1752-58, 7 vols.

Degeer, Bromsarnas ursprung, uptaeckt af Carl Degeer, Konigl. Vetenskaps
Academiens Handlinger for Ar 1760, xxi, 276-291, Plate IX, Figs. 1, 2,
3, and 4. (The illustrations of Tabanus bovinus are identical with those
given in the Memoires, but the illustration of the pupa is more exactly re-
produced in the Memoires.)

Degeer, Memoires pour servir a I'histoire des insectes, 1776, vi, 5th Memoire.
German edition (Goeze), 1882.

Dufour, L., Notices entomologiques. I. Consultation sur une larve aquatique,
Ann. Soc. entomol. France, 1862, Plate II, Fig. 2, a and b.

* Reference is made to the larval stages of Tabanidae in a number of text-
books or other papers all of which were not mentioned in this Bibhography if
they contained no original information.

200

WERNER MARCHAND 201

Fabricius, J. C, Entomologia Systematica, 4 vols., Hafniae, 1772-94; Supple-
ment 1798.
von Friedenfels, E., Ueber Artemia salina und andere Bewohner der Soolen-

teiche in Salzburg, Vcrhandl. Mitt. Siebenbiirg. Vereins fiir Naturwissen-

schaften in Herrmannstadt, 1880, xxx, 112-178 (p. 168 Tabanus autumnalis) .
Girault, A., A new Phanurus from the United States, with notes on allied species,

Canad. Entomologist, 1916, xlviii, 149-150.
Goldi, E. A., Die sanitarisch-pathologische Bedeutung der Insekten und ver-

wandten Gliedertiere, namentlich als Krankheits-Erreger und Krankheits-

tJbertrager, Berlin, 1913.
Graber, V., Ueber neue otocystenartige Sinnesorgane der Insecten (Neues Organ

einer Fliegenmade) , Arch, mikroskop. Anal., 1878, xvi.
Graber, Die chordotonalen Sinesorgane und das Gehor der Insekten, Arch.

mikroskop. Anat., 1882-83, xx, xxi.
Griinberg, K., Die blutsaugenden Dipteren, Jena, 1907.
del Guercio, Le tipuli ed i tafani nocivi nelle risaie de Molinella (Bologna), Redia,

1913, 299-345.
Hart, C. A., On the entomology of the l^lHnois River and adjacent waters. Bull.

Illinois State Lab. Nat. Hist., 1895-96, iv, 220-247. (Gives photographic

pictures of the larvje. Same volume contains original description of

Phanurus tabanivorus Ashmead.)
Henneguy, L. F., Les insectes, Paris, 1904. (Graber's organ, p. 488.)
Herms, W. B., Medical and veterinary entomology. A textbook for use in

schools and colleges as well as a handbook for the use of physicians,

veterinarians and public health officials. New York, 1915.
Hine, J. S., On the life history of Tabanus vivax, Ohio Naturalist, 1903, iv, 1-2.
Hine, Tabanidae of Ohio, Ohio State Univ. Contribution from the Dept. of

Zool. and Entomol., Ohio Acad. Sc, No. 10, Special papers. No. 5, 1903.
Hine, Insects injurious to stock in the vicinity of the Gulf Biologic Station,

U. S. Dept. Agric, Division oj Entomol., Bull. 44, 1904, 57-60. (Mentions

eggs of T. costalis.)
Hine, Same paper reprinted in Second Rep. Gulf Biol. Station, 1903, Bull. 2, 1904.
Hine, Some economic considerations with reference to the Tabanidae, Proc.

XVIth Annual Meeting Assn. Economic Entomologists, U. S. D. A. E.,

No. 46, 1904, 23-25; Gulf. Biol. Station Rep., Bull. 2, 1904, 86-87, mentions

eggs of trimaculatus.
Hine, Habits and life histories of some flies of the family Tabanidae, U. S. Dept.

Agric, Bureau Entomol., Technical Series 12, 1906, pt. 2, 19-38.
Howard, L. O., The insect book, 1901. (The family Tabanidae is treated on

pages 131-133. Riley's figures of Tabanus atratus are reproduced.)
Keilin, D., Sur certains organes sensitifs constants chez les larves de Dipteres

et leur signification probable, Compt. rend. Acad., 1911, cliii, 977-979.

Quoted in Brauer (1883). No tabanids mentioned.
King, H. H., Report on economical entomology, Rep. Wellcome Research Lab.

Gordon Memorial College, 1908, iii, 212-214. (Life history of Tabanus

biguttatus.)

202 THE EARLY STAGES OF TABANID^

King, Third Rep. Wellcome Research Lab. Gordon Memorial College, and Bull.

Entomol. Research, 1910, i, pt. 2. (Bionomics of Tahanus par and tceniola.)
King, Some observations on the bionomics of Tahanus ditceniatus Macquart,

and Tahanus kingi Austen, Bull. Entomol. Research, 1910, i, pt. 4.
Kollar, M. V., Beitrag zum Haushalte der sehr lastigen Viehbremsen (Tabanidae),

Sitzungsb. math, naturwissensch., Classe d. k. Acad. Wissensch, 1854, xiii,

531-535.
Krauss, H., Otocyst-enartiges Organ bei Tahanus autumnalis L., Zool. Anz.,

Cams, 1879, ii, 229-230.
Lecaillon, A., Sur la ponte des ceufs et la vie larvaire des Tabanides, particu-

lierment du Taon a quatre taches {Tahanus quatuornotatus Meig.), Ann.

Sac. entomol. France, 1905.
Lecaillon, Sur I'organe de Graber de la larve de Tahanus quatuornotatus Meig.,

Compt. rend. Assn. anat., 1905, vii, 130-131, 3 figs.
Lecaillon, Sur quelques points de I'histoire naturelle des tabanides, en particulier

de Tahanus quatuornotatus Meig., Compt. rend. Soc. biol., 1906, Ix, 459.
Lecaillon, Deuxieme note sur I'organe de Graber, Compt. rend. Assn. anat., 1906,

viii, 65-67.
Lecaillon, Sur la structure de la couche chitineuse tegumentaire et sur les in-
sertions musculaires de las larve de Tahanus quatuornotatus Meigen, Compt.

rend. Assn. anat., 1906, viii, 68-70.
Lecaillon, Nouvelles observations sur la ponte des oeufs et la vie larvaire du

Tahanus quatuornotatus Meig. (Dipt.), Ann. Soc. entomol. France, 1911, Ixxx.
Lutz, A., Dipteren. Mem. Inst. Oswaldo Cruz, Rio de Janeiro, 1914, vi, 43-49.
Macquart, Histoire naturelle des insectes. Dipteres, 1834, i, 193. (Degeer's

figure reproduced.)
Malloch, J. R., A preliminary classification of Diptera, exclusive of Pupipara,

based upon larval and pupal characters, with keys to imagines in certain

families. Bull. Illinois State Lab. Nat. Hist., 1917, xii, pt. 1.
Marno, E., Zur Biologie von Hexatoma, Verhandl. zool. hot. Ges., 1868, xviii, 74-75.
Maxwell-Leffroy, H., and Hewlett, F. M., Indian insect life, Agric. Research

Inst., Pusa, India, Calcutta, and London, 1909. (Tabanidae, pp. 592-595;

Tahanus striatus, etc. ; larval habits of Gastroxides ater.)
McAfee, W. L., Facts in the life history of Goniops chrysocoma (Diptera, Tabani-
dae), Proc. Entomol. Soc. Washington, 1911, xiii, 21-29.
Mitzmain, M. B., The biology of Tahanus striatus Fabricius, the horsefly of the

Philippines, Phil. J. Sc, Section B, 1913, viii, 197-221, Plate 7.
Neave, S. A., The Tabanidae of southern Nyasaland, with notes on their life

histories, Bull. Entomol. Research, 1915, v, pt. 4, 287-320. 4 plates.
Osburn, R. C, Tabanidae as inhabitants of the Hydrophytic Area, J. New York

Entomol. Soc, 1913, xxi, 63-65.
Packard, A. S. Textbook on entomology. New York, 1898, gives on page 627,

Fig. 520, an illustration of the pupa of Midas clavatus which is erroneously

referred to Tahanus atraius. Fig. 520 c probably represents atratus (after

Riley?, very imperfect), 520 d is Proctacanthus. The pupa of Tahanus

lineola is represented on p. 630, Fig. 585.

WERNER MARCHAND 203

Paoli, G., Intorno all' organo del Graber, nelle larvae di ditteri tabanidi, Redia,

1907, iv, 247-258.
Patton, W. S., and Cragg, F. W., Textbook of medical entomology, Christian

Literature Society for India, London, Madras, and Calcutta, 1913.
Ferris, E., Histoire des insectes du pin maritine, Ann. Soc. Entomol. France,

1870, X, 135-232. (Description of larva of Hcematopota, pp. 196-198; pupa,

pp. 198-199; tabanid early stages in general, pp. 199-201.)
Picard, F., and le Blanc, G. R., Sur les moeurs lignicoles de la larve de Ta^awM^

cordiger Meig. (Dipt. Tabanidae), Bull. Soc. Entomol. France, 1913, 318-321,

Fig. 1.
Portschinsky, Biologic, Beschreibung, Bekampfung von Tabaniden, (Russ), 4th

edition, Trd. B. entom. Ucen. Kom. M. Zeml., St. Petersburg. (Travau\ de

Bureau entomologique de Comite scientifique du Ministere d'Agriculture et

des Domaines. St. Petersbourg, Russe), 1908, i, 1-52, 21 figs. (This paper

contains an illustration of the egg mass of the European species Tabanus

montanus which is black and placed on the leaf of a grass plant above

water, and two illustrations of the pupa of Tabanus tarandinus. These

two species have to be added to the list of species on which data are

available. The illustrations are not reproduced because the paper which

I have seen on one occasion, became unavailable.)
Riley, C. V., The black breeze-fly — Tabanus atratus Fabr., Second Ann. Rep. on

the noxious, beneficial, and other insects of the State of Missouri, made

to the State Board of Agriculture, 1870, 128-132.
Riley, W. A., and Johannsen, O. A., Handbook of medical entomology, Ithaca,

1915, (gives illustrations of the larva of Tabanus atratus and of eggs of a

common unknown species).
Scholtz, H., Ueber den Aufenthalt der Dipteren wahrend ihrer ersten Stande,

Z. Entomol., 1848, 25-34. (p. 28, tabanid larvae.)
Sharp, D., The Cambridge Natural History, Insects, 1901, vi, pt. 2, 483.

{Tabanus larva, supposed to be that of Atylotus fulvus).
Surcouf, J. M. R., and Ricardo, G., fitude Monographique des Tabanides

d'Afrique (Groupe des Tabanus), Paris, 1909 (Pupa of Tabanus bromius,

etc.). These figures reproduced in Brumpt (1910),
Wahlberg, P. F., Bidrag till Svenska Dipternas Kannedom. Koningl. Wetenskaps

Academiens Handlingar, for ar 1838, Stockholm, 1839. (Though Tabanus

glaucopis imago is described, nothing is said about the larval stage.) Quoted

by Brauer (1883).
Walsh, B. D., On certain remarkable or exceptional larvae, coleopterous, lepid-

opterous, and dipterous, with demonstrations of several species injurious

to vegetation which have been already published by Agricultural Journals,

Proc. Boston Soc. Nat. Hist., 1863, ix, 302-306.
Walton, W. R., Notes on the egg and larva of Goniops chrysocoma O. S., Entomol.

News, 1908, xix, 464-465.
Westwood, J. 0., Introduction to the modern classification of insects, 1840, ii,

538.
Zetterstedt, Diptera Scandinaviae, i, 105 {Tabanus larvae), and 123 {Chrysops).

Plate 1.

Fig. 1. Female of Chrysops moerens, ovipositing on Typhale^i. After Hine
(photograph from life).

Fig. 2. Egg masses of Chrysops moerens, on Sparganium leaf. After Hine
(photograph) .

Fig. 3. Egg mass of Tabanus stygius on Sagittaria leaf. After Hine.

Fig. 4. Egg mass of Tabanus stygius on Peltandra leaf. After Hine.

Fig. 5. Egg mass of Tabanus stygius in unusual location. After Hine.

Fig. 6. 'EggmdiSsesoi Tabanus sp.} After Riley and Johannsen (photograph).

Fig. 7. Egg mass of Tabanus sp.? Natural size 24 mm. After Maxwell-
Leffroy and Howlett (colored original yellow).

Fig. 8. Egg mass of Tabanus tczniola, with chalcid parasites. After King
(colored plate) .

Fig. 9. a, Egg mass of Tabanus tceniola; b, young larva. After King
(colored plate) .

Fig. \Q. 'Egg nxdissoi Tabanus par. After King (colored plate) .

Fig. 11. Young larva of Tabanus par. Natural size 10 mm. After King
(colored plate).

Fig. 12. Egg mass of Tabanus bicallosus on a blade of grass. After Patton
and Cragg.

Fig. 13. Two single eggs of Tabanus bicallosus. After Patton and Cragg.

Fig. 14. Egg mass of Chrysops dispar, on a blade of grass. After Patton and
Cragg.

Fig. 15. Single egg of Chrysops dispar. After Patton and Cragg.

Fig. 16. Egg mass of Tabanus virgo, on a dry twig. After Patton and Cragg.

Fig. 17. Single egg of Tabanus ditczniatus. After Patton and Cragg.

Fig. 18. Egg mass of Tabanus biguttatus. Natural size 14 mm. After King
(colored plate) .

Fig. 19. Egg mass of Tabanus speciosus. After Patton and Cragg.

Fig. 20. Egg mass of Tabanus kingi. After King.

King's illustration of a rock situated above water, showing the location of
these egg masses, has been omitted.

MONOGRAPH NO. 13.

PLATE 1.

iy 18

(Marchand: The early stages of Tabanidoe.)

Plate 2.

Fig. 21. Egg mass of Tabanus ditceniatus spread out in a single layer on a
blade of grass. After Patton and Cragg.

Fig. 22. Egg mass of Tahanus striatiis on paddy leaf. Natural size 13 mm.
After Bainbridge and Fletcher.

Fig. 23. Same egg mass, magnified. After Bainbridge and Fletcher.

Fig. 24. Female of Tabanus striatiis, ovipositing. Natural size of tly 19 mm.
After Mitzmain (photograph).

Fig. 25. Egg masses of Tabanus striatiis. After Mitzmain (photograph).

Fig. 26. Empty (?) egg mass of Tabanus striatiis. After Mitzmain (photo-
graph).

Fig. 27. Egg mass of Tabanus striatiis in process of hatching. After Mitz-
main (photograph).

Fig. 28. Female of Goniops chrysocoma, ovipositing. After McAtee (photo-
graph).

Fig. 29. Egg mass of Goniops chrysocoma, seen from above. After Walton.

Fig. 30. Egg mass of Goniops chrysocoma, lateral view. After ]Mc.-\tee (pho-
tograph).

Fig. 31. Egg mass of Goniops chrysocoma. Natural size 14 mm. After J\lc-
Atee (photograph).

Fig. 32. Empty egg mass of Goniops chrysocoma after hatching. After
McAtee.

Fig. 33. Double egg mass of Tabanus quatiiomotatus. After Lecaillon.

Fig. 34. Two single egg masses of Tabanus quatuoniotatiis. Natural size 9
mm. After Lecaillon.

Fig. 35. Group of egg masses of Tabanus quatiiomotatus on dried twig. After
Lecaillon.

Figs. 36, 37, and 38. Egg masses oi Tabanus quatuomotatus, magnified. After
Lecaillon.

MONOGRAPH NO. 13.

PLATE 2.

4 . A^^'^v^

^.M^

■r^' ;^

f/

28

22

33

34

35

w

W 24

i

27

T*Jj

(Marchand: The early stages of Tabanida:.)

Plate 3.

Fig. 39. Larva of Chrysops viUatus. Natural size 15 mm. After Hart (pho-
tograph).

Fig. 40. Larva of Tahanus stygius. Natural size 52 mm. After Hart (pho-
tograph) .

Fig. 4L Larva of Tabanus atraius. Natural size 53 mm. After Hart (pho-
tograph) .

Fig. 42. Larva of Tahanus sp.? Natural size 24 mm. After Hart (photo-
graph).

Fig. 43. Larva of Tabanus vivax. Natural size 36 mm. After Hine.

Fig. 44. Larva of Tabanus lasiophthalmus. Natural size 23 mm. After
Hine.

Fig. 45. Larva of Tabanus atraius. After W. A. Riley and Johannsen (pho-
tograph).

Fig. 46. Larva of Tabanus atraius. Natural size 52 mm. After C. V. Riley.

Fig. 47. Larva of Tabanus ditceniatus. Lateral view of immature larva.
After King (colored plate).

Fig. 48. Larva of Tabanus par, almost full grown. Natural size 52 mm.
After King (colored plate).

Fig. 49. Larva, full grown, of Tabanus tceniola. After King (colored plate).

Fig. 50. Larva of Chrysops wellmani. Natural size 13 mm. After Neave
(Terzi).

Fig. 51. Larva of Tabanus gratus. Natural size 23 mm. After Neave
(Terzi).

Fig. 52. Larva of Tabanus atrimanus. Natural size 32 mm. After Neave
(Terzi) .

Fig. 53. Larva of Tabanus insignis. Natural size 34 mm. After Neave
(Terzi) .

Fig. 54. Larva (supposed larva) of Tabanus pcrtinens. Natural size 36 mm.
After Neave (Terzi).

MONOGRAPH NO. 13.

PLATE 3.

39

40 W 41

42

H

44

47

48

(Marchand : The early stages of Tabanidse.)

Plate 4.

Fig. 55. Larva of Tahaiius sp.? a, Ventral view; b, cross-section of segment;
c, dorsal view. Natural size 32 mm. After Brumpt.

Fig. 56. Larva of Tabanus kingi. After King.

Fig. 57. Larva of Tabanus ditcenialus. a, Immature larva; b, mature larva,
lateral view; c, 6th and 7th segments of immature larva, lateral view; d, caudal
end of immature larva. After King.

Fig. 58. Larva of Tabanus sp.? After Griinberg.

Fig. 59. Early stages of Tabanus bovinus. a. Larva; b, pupa; c, anal end of
larva; d, anal end of pupa. After Degeer.

Fig. 60. Larva of Tabanus (Atylotus) fulvus (?). a, Entire larva; b, head and
mouth-parts; c, one proleg; d, segments 10, 11, and 12 in lateral view, .\fter
Sharp.

Fig. 61. Mature larva o{ Tabanus ditceniaius. After Patton and Cragg.

Fig. 62. Mature larva of Tabanus bicallosus. After Patton and Cragg.

Fig. 63. Mature larva of Tabanus virgo. After Patton and Cragg.

Fig. 64. Larva of Tabanus ditceniatus. a, Segments 3, 4, and 5, lateral view;
b, segments 11 and 12, lateral view; c, segments 11 and 12, ventral view. After
King.

Fig. 65. Larva oi Tabanus kingi. a. Lateral view of two abdominal segments;
b, one proleg; c, ventral side of anal segment. After King.

MONOGRAPH NO. 13

PLATE 4.

59

i 63

(Marchand: The early stages of Tabanidje.)

Plate 5.

Fig. 66. Caudal end of larva of Tabanus ustus, lateral view. After Neave
(Terzi).

Fig. 67. Caudal end of larva of Tabanus corax, lateral view. After Neave
(Terzi) .

Fig. 68. Caudal end of larva of Tabanus medionotatus, lateral view. After
Neave (Terzi).

Fig. 69. Caudal end of larva of raio;»<5 Z)/^«//a;2/5-, lateral view. After Neave
(Terzi).

Fig. 70. Caudal end of larva of Tabanus variabilis, lateral view. After Neave
(Terzi) .

Fig. 71. Caudal end of larva of Tabanus maculatissimus, lateral view. After
Neave (Terzi).

Fig. 72. Caudal end of larva of Chrysops longicornis, lateral view. After
Neave (Terzi).

Fig. 73. Caudal end of Chrysops. Dorsal view; diagram. After Neave.

Fig. 74. Caudal end of Hcemalopota. Dorsal view; diagram. After Neave.

Fig. 75. a, b, Larva of Tabanus biguitatus. Natural size 42 mm. After King
(colored plate).

Fig. 76. a, Young larva of Tabanus sp.7 Colored original greenish white.
After Maxwell-Leffroy and Howlett (colored plate), b, Full grown larva of
Tabanus s p.} Same species as Fig. 16, a. Colored original yellow. Natural
size 47 mm. After Maxwell-Leffroy and Howlett (colored plate), c, Hymen-
opterous parasite of same larva. After Maxwell-Leffroy and Howlett (colored
plate).

Fig. 77. Head and first segment of larva of Tabanus atralus. After Malloch.

Fig. 78. Mandibles of larva of Tabanus biguttatus. After King (colored
plate).

Fig. 79. Mandibles of larva of Tabanus strialus. After Mitzmain (photo-
graph) .

Fig. 80. Larval stages of ra/^a;^^^ ^/r/a///^. a, Newly hatched larvae; i, young
larva magnified; c, two full grown larvae. Natural size of the latter 34 mm.
After Mitzmain (photograph).

Fig. 81. a, b, c, Shed skins (exuviae) of three different larval stages of Tabanus
atratus. After ]\Iitzmain (photograph).

MONOGRAPH NO. 13.

PLATE 5.

68

73

74

75

76 P

80

vo^t

(Marchand: The early stages of Tabanidx.)

Plate 6.

Fig. 82. Early stages of Ilcewalopola pluvialls. a. Larva, natural size 13
mm.; b, head and mouth-parts, dorsal view; c, head and mouth-parts, lateral
view; d, ventral view of anal segment; e, terminal spiracle;/, pupa; ,;,', pupal
structure. After Perris.

Fig. 83. Stages of Ilcematopota plnvialis. a. Larva, lateral view, natural
size 14 mm.; b, terminal spiracle of larva; r, pupa; d, terminal segment of pupa.
After Braucr.

Fig. 84. Head and mouth-parts of HcEmalopota pluvial is. ant, antenna;
Ir, labrum; lb, labium; md, mandible; p, palpus. After Brauer.

Fig. 85. Larva of H(cmatopola criidcUs. Natural size 13 mm. After Xeave
(Terzi) .

Fig. 86. Caudal end of larva of Hccmatopota insaliabilis. After Neave (Terzi).

Fig. 87. Larva of Tabaniis spodoptcnis. a, Larva magnified; b, dorsal view,
natural size 45 mm.; c, lateral view; d, caudal end of larva; e, terminal
spiracle;/, head of larva, lateral view, magnified. After Brauer.

Fig. 88. Larva of Tabaims cordigcr. a, \'entral view of larva, natural size
35 mm.; b, caudal end; c, cross-section through segment; d, caudal end with
tracheae.

Fig. 89. Larva of Hexatoma pdliiccns. a, Dorsal view of larva; b, dorsal
view of last three segments of larva, showing pigmentation. Natural size 35 mm.
After Brauer.

Fig. 90. Head and mouth-parts of larva of Hexatoma pclhccciis. a, Lateral
view; b, lateral view of mouth-parts; c, dorsal view of head. Ir, labrum; vid,
mandible; p, palpus; ant, antenna; oc, eye-spot; lb, labium. After Brauer.

Fig. 91. Head of larva of Hccmatopola plnvialis, dorsal view. Parts as in
Figs. 84 and 90. After Brauer.

MONOGRAPH NO. 13

(Marchand: The early stages of 'J'abanicia;.')

Plate 7.

Fig. 92. Full grown larva of Goniops clirysocoma. a, Dorsal; b, ventral;
c, lateral view. Natural size 17 mm. After McAtee (photograph).

Fig. 93. Young larva of Goniops chrysocoma. a, Entire larva, dorsal view;
b, mouth-parts, magnified, from above; c, mouth-parts, lateral view. After
Walton.

Fig. 94. Full grown larva of Go}iiops chrysocoma. After McAtee (photo-
graph).

Fig. 95. a, b, Structures of head and mouth-parts of the full grown larva of
Goniops chrysocoma. anf, base of antennae; Ibr, labrum; mx, maxilla; mxp,
maxilla palpus. Other details as in Figs. 84, 90, 91. After McAtee.

Fig. 96. a, Larva of Tipula olcracea and b, Tabanus ignotus?, drawn for com-
parison. After del Guercio. (To judge from the figures, both drawings seem
to represent tipulid larvae.)

Fig. 97. Head of Tabanus albimcdius, lateral view, md, mandible; mx,
maxilla; ant, antenna; sp, spines; //', labrum; p, palpus. After Patton and
Cragg.

Fig. 98. Intestinal tract of the full grown larva of Tabanus albimcdius. ces,
Esophagus; pv, proventriculus; slg, salivary glands; mg, midgut; mpt, Malpi-
ghian tubes. After Patton and Cragg.

MONOGRAPH NO. 13

PLATE 7.

92

?«/" AnJ.

anf Ikr

94

" mJ

(Marchand: The early stages of Tabanidae.)

Plate 8.

Fig. 99. Newly hatched larva of Tabauus aiitumnalis (?). 5 1 to 5 13, somites
(segments) of the body; w, prolegs armed with hooks; tb, touch-tactile bristle;
ph, pharynx; md, midgut; ed, hindgut; wa, Malpighian vessels;^, anus; tr, tracheal
trunks; / sp, terminal spine; g sup, upper, ^g ////, lower cephalic ganglion; bg\ to
bg \\, ventral chain of ganglia; ch 1 to ch i, chordotonal organs; ot, otocyst-Kke
organs on posterior border of anal segment (Graber's organ). After Graber.

Fig. 100. a. Right side of 2nd and 3rd body segments of young Tabauus larva
highly magnified.

2nd segment: cu, cuticle; ma, matrix; m 1, circular muscles of pharynx;
in 2, muscles originating on the posterior border of the first segment, and ex-
tending forward and inwards, possessing strongly marked transverse striation
and great contractility; hn, nerve of the skin; ekg, mkg, ganglion cells connected
with this nerve and possessing one nucleus ekg, or several mkg; tb, tactile
bristle; s 2 c/i 1 monoscolopic, s 2 ch 2 triscolopic chordotonal organ; //, liga-
ment; st 2, rods; cfs, terminal filament; e I, e 2, c 3, attachments of filaments;
mn, nerve going to muscle; /r, branch of tracheal system; hg, ganglion; nu,
nuclei.

3rd segment, ;;/ 3, system of sagittal muscles, freely anastomizing, and in
fresh condition smooth (unstriated) in appearance; bl, region of integument cov-
ered with bristles, in anterior part of segment; s 3 ch 1 to s 3 ch 6 chordotonal
organs (taken from different individuals); {ch 1 and ch 2, and again ch 4, 5, and
6 were observed in the same individual) ; e i, e 5, e 6, points of attachments of the
terminal filaments of 5 3 ch 5; e 7, e S, c 9, those of .? 3 ch 2; g Inf, lower pharyn-
geal ganglion. After Graber. Zeiss oil immersion lens.

Fig. 100. b. Tactile bristle of young larva of Tabajius aulumnalis (?), with
its ganglion, cu, cuticle; tb, tactile bristle, articulating on its place of attach-
ment; ma, matrix (hypoderm), in fresh condition showing no cell limits, with
reddish nuclei; starting from it a fine filament torn out of the sheath surround-
ing the bristle; k, nuclei; n, nerve; g, ganglion. After Graber. Zeiss oil
immersion lens.

Fig. 101. Triscolop chordotonal organ of the 2nd segment of the same
larva. », nerve; g 'i^, g 2, g 3, the three ganglion cells, lying one above the other,
with reddish nuclei; xfs, neurit; st, rods; ko, heads of rods; mk, nucleus of the
proximal end; dk, of the distal end of the rod\ fs 1 to fs 3, the three terminal
filaments. After Graber. Zeiss oil immersion.

Fig. 102. Monoscolop chordotonal organ of the 2nd segment of the same larva
xfs, neurit visible in the interior of the rod. After Graber. Zeiss oil immersion.

MONOGRAPH NO 13.

PLATE 8.

100

z=' ruu

101

102

(Marchand: The early stages of Tabanidae.)

Plate 9.

Fig. 103. Larva of Tahanus autumnalis (?). g, Graber's organ. After
Graber.

Fig. 104. Posterior end of larva, magnified, comprising 9th, 10th, 11th, and
12th segments. (In the original the numbers are given erroneously as 8, 9, 10,
and 11.) tr, tracheal trunks; dv, dorsal vessel; g, Graber's organ;/, terminal
tube; m, two muscles extending forward from Graber's organ. After Graber.

Fig. 105. Graber's organ in situ, dv, dorsal vessel; tr, lateral tracheal
trunks; fm, muscle of dorsal vessel; m, muscles; /, terminal tube; n, nerves (?);
/, fat body; cu, cuticle of integument; hp, hypoderm. X 60. Zeiss oil immer-
sion. After Graber.

Fig. 106. The organ isolated, c 1 to c3, chitinous capsules; sp, extremit}^ of
capsule; ep, its epithelium; cp 1 to cp 4, pedunculate bodies, (in the original
only the first capsule bears this name, while the inner ones are treated as
"internal sacs"); ml to ml, muscles; n to « 1, nerves (?) ; ga, ganglion-like
swelling of first nerve; /, terminal tube; op, operculum. X 133. Zeiss oil
immersion. After Graber.

Fig. 107. One pair of pedunculate bodies, cp, body; st, hollow stem or
peduncle; cu, scaly cuticle to which they are fastened. X m. Zeiss oil immer-
sion. After Graber.

Fig. 108. A portion of elastic connective tissue from the surroundings of the
organ, after treatment with boiling KOH. X 333. Zeiss oil immersion. After
Graber.

Graber's figures, Figs. 103 to 108, were drawn from fresh material.

MONOGRAPH NO. 13.

PLATE 9.

103 V ^

104

105

107

108

(Marchaiid: The early stages of Tabaiiidae.)

Plate 10.

Fig. 109. Young larva of Tabanus sp.? tr, tracheal trunk; g, Graber's organ.
After Henneguy.

Fig. 110. Graber's organ, enlarged, of larva figured in Fig. 109. After
Henneguy.

Fig. 111. Larva of Tabanus sp.?, showing the location of Graber's organ (the
latter drawn larger than in proportion to the size of the larva). After Paoli.

Fig. 112. Graber's organ in full grown larva of Tabanus sp.?, with muscle
and nerves attached to it. cl to c7, capsules; cp, pedunculate bodies; m,
muscle; n, nerve. After Paoli.

Fig 113. Hypothetical diagram of the development of Graber's organ, c 1,
first capsule; c 2, second capsule; hp, hypoderm; ct, cuticle. After Paoli.

Fig. 114. Terminal tube of Graber's organ. /?/>, hypoderm. After Paoli.

Fig. 115. Last three segments of larva, showing Graber's organ in situ. Dor-
sal view tr, tracheal trunk; m 1, muscles of the first pair; a, anal tubercle.
After Paoli.

Fig. 116. Last three segments of larva, with Graber's organ. Lateral view.
in 2, muscle of the second pair; /, terminal tube. Other details as in Fig. 115.
After Paoli.
Fig. 117. Young larva oi Tabanus quatuonwtatus. After Lecaillon.

Figs. 118 and 119. Posterior end of body of larva of Tabanus quatuornotatus,
somewhat older than that figured in Fig. 117, showing the pedunculate bodies
being expelled through the terminal tube. After Lecaillon.

Fig. 120. Dorsal view of Segments 11 and 12 (syphon) of a young larva
Tabanus corax, showing Graber's organ; the dotted lines show the position of the
anus. After Neave.

MONOGRAPH NO. 13.

PLATE 10.

(Marchand: The early stages of Tabanirlx.)

Plate 11.

Fig, 121. Pupa of Goniops chrysocoma. a, Dorsal; h, ventral; c, lateral view.
After ]\IcAtee.

Fig. 122. Pupa of Tipula oleracea. After Paoli.

Fig. 123. Pupa of Tabanus ignotus. Compare with Fig. 122. After Paoli.
(See also Fig. 96, a, b. No resemblance is found in the pupae which indicates
that Fig. 96, b, does not represent a tabanid larva.)

Fig. 124. Vupaoi Tabanus airatus. Natural size 31 mm. After Hine.

Fig. 125. Pupa of Tabanus stygius. Natural size 28 mm. After Hine.

Fig. 126. Fupaoi Tabanus sulcifrons. Natural size 27 mm. After Hine.

Fig. 127. Pupa of Tabanus lasiophthalmus. Natural size 23 mm. After
Hine.

Fig. 128. VnpK oi Tabanus vivax. Natural size 23 mm. After Hine.

Fig. 129. Vnpa, oi Tabanus ditceniatus. After King.

Fig. 130. Pupa of Chrysops longicornis. Natural size 11 mm. After Neave
(Terzi).

Fig. 131. Pupa of Hccmatopota insatiabilis. Natural size 11 mm. After
Neave (Terzi).

Fig. 132. Pupa of Tabanus sp.? Natural size 18 mm. After IMa.xwell-
Leffroy and Howlett.

Fig. 133. Pupa of Tabanus dita;niatus. \'entral view. After Patton and
Cragg.

Fig. 134. Pupa of Tabanus virgo. Ventral view. After Patton and Cragg.

Fig. 135. Pupa of Tabanus bicaUosus. Ventral view. After Patton and
Cragg.

MONOGRAPH NO. 13.

PLATE 11.

122

123

124

125

126

127

128

129

130

131 132

133

134

135

(Marchaiid: The early stages of Tabaiiidit.)

Plate 12.

Fig. 136. Pupa of Tabanus cordigcr. After Picard and le Blanc (from pupal
shell).

Fig. 137. a, Pupa of Tabanus kingi; b, caudal end, lateral view; c, caudal end,
ventral view. After King.

Fig. 138. Pupa of Tabanus variabilis. Natural size 20 mm. After Neave
(Terzi) .

Fig. 139. Pupa of Tabanus sp.? (see Figs. 7 and 76). Colors of original yel-
low, thorax grayish. Natural size 21 mm. After Maxwell-Leffroy and Howlett
(colored plate).

Fig. 140. Pupa shell of Tabanus atratus. Natural size 31 mm. After C. V.
Riley.

Fig. 141. Caudal end of pupa of Tabanus ditaniatus. a, Lateral view; b,
ventral view. After King.

Fig. 142. Pupa of Tabanus par. a, Entire pupa, lateral view, natural size
22 mm.; b, pupal aster. After King (colored plate).

Fig. 143. Pupal aster of Tabanus striatus. a, Pupal aster of male; b, pupal
aster of female. After Mitzmain (photograph).

Fig. 144. 'Pwpsil SiSttx ol Tabanus bicallosus (f. After Patton and Cragg.

Fig. 145. VxxpaX SiSitr oi Tabanus bicallosus 9. After Patton and Cragg.

Fig. 146. Pupal aster of Tabanus virgo. After Patton and Cragg.

Fig. 147. Pupal aster of Tabanus ditceniatus. After Patton and Cragg.

Fig. 148. Fupal aster oi Tabanus biguttatus. After King (colored plate).

Fig. 149. Pupa of Tabanus lincola, dorsal view. After Malloch.

Fig. 150. Pupa of Tabanus sp.?, natural size 17 mm. After Surcouf and
Ricardo.

Fig. 151. Pupa of Tabanus bromius 6" in process of hatching, natural size 23
mm. After Surcouf and Ricardo.

Fig. 152. Pupa of Tabanus striatus. a, Lateral view of pupa; b, pupa in
process of hatching. After Mitzmain.

MONOGRAPH NO. 13.

PLATE 12.

141

137

138

139 140

142

144

K'Lcf^ !,

143

145

147

146

148

149

150

152

(Marchand: The early stages of Tabanida.)

Plate 13.

Fig. 153. Upper half of thorax of pupa of r<7/)(/»»5 ,v/vg/»5, dorsal view. After
Malloch.

Fig. 154. Upper half of thorax of pupa of Clirysops villains, dorsal view.
After Malloch.

Fig. 155. Thorax of pupa of Tabanus cordigcr, dorsal view. After Picard and
le Blanc.

Fig. 156. Prothoracic spiracle of pupa of Tahauits iiigrcsccns, dorsal view.
After Malloch.

Fig. 157. Prothoracic spiracle of pupa of C/irysops villains, dorsal view.
After Malloch.

Fig. 158. Fupal aster oi Tabanus cord iger. After Picard and le Blanc.

Fig. 159. Vupal aster of Tabamts lasiophlhalmns. After Hine.

Fig. 160. Fupal aster oi Tabanus viva. v. After Hine.

Fig. 161. Vupal aster oi Tabamis sty gins. After Hine.

Fig. 162. Fupal aster of Tabanus lineola. After Hart (photograph).

Fig. 163. Vupal aster of Tabanus atratus. After Hart (photograph).

Fig. 164. Pupal aster of Tabanus atratus. After Hine.

Fig. 165. Pupal aster of Tabanus sulcifro)is. After Hine.

Fig. 166. Pupal aster of Chrysops magnifica, var. inornata. a, Male; b,
female. After Neave (Terzi).

Fig. 167. Pupal aster of Chrysops bimacnlosa. a, Male; /', female; c, female,
side view. After Neave (Terzi).

Fig. 168. Vupal aster of C/irysops wellmani. a, &;b, 9. After Neave.

Fig. 169. a. Pupal aster and b, dorsolateral comb of Hcematopota decora 9 .
After Neave.

Fig. 170. Vnpal aster of Hccmatopola cmdclis. After Neave.

Fig. 171. Vupal aster of Chrysops longicor)iis. a, o^;/', 9. After Neave.

Fig. 172. Pupal aster of Hcematopota insaliabilis 9. a. From behind; b,
side view; c, dorsolateral comb. After Neave.

MONOGRAPH NO. 13.

PLATE 13.

153

154

t -

159

160

162

166

167

168

o-

¥

169

170

tsr

171 l*

172 "

(Marchand: The early stages of Tabanidx.'l

Plate 14.

Fig. 173. Tahanus atrimanus. a, Pupal aster of d^; ft, dorsolateral comb of
9 ; c, pupal aster of 9 . After Neave (Terzi).

Fig. 174. Tabanus medionotatus d'. a, Pupal aster; b, profile of last seg-
ment of pupa. After Neave (Terzi).

Fig. 175. Tabanus higiUtatus. a, Pupal aster of cf and b, oi 9 ; c, dorso-
lateral comb of cf and d, ol 9 . After Neave (Terzi).

Fig. 176. Tabamis variabilis, a, Pupal aster of cf; 6, dorsolateral comb of
d", c, dorsolateral comb of 9 ; d, pupal aster of 9 . After Neave (Terzi).

Fig. 177. Tabanus ustus. a, Dorsolateral comb of cf ; ft, pupal aster o d;
c, dorsolateral comb of 9 (with six spines). After Neave (Terzi).

Fig. 178. Tabanus maculatissimus. a and ft, dorsolateral comb of two dif-
ferent d ; c, pupal aster of 9 ; d, dorsolateral comb of 9 . After Neave (Terzi).

MONOGRAPH NO. 13.

173

PLATE 14.

174

176

177

178

(Marchand: The early stages of Tabanida;.)

Plate 15.

Fig. 179. Tuba ii us uagamiensis d' . a, Pupal aster; b, pupal aster from the
side, showing combs, the dorsolateral comb being absent. After Neave (Terzi).

Fig. 180. a, Pupal aster of Tabaniis descrtus; b, latera' view. After Bodkin
and Cleare (Terzi).

Fig. 181. Tabanus corax. a, Pupal aster of cf ; b, pupal aster of 9 ; c, dorso-
lateral comb of 9. After Neave (Terzi).

Fig. 182. Tabanus fratemus 9. a, Pupal aster; /?, dorsolateral comb. After
Neave (Terzi).

Fig. 183. Tabanus obscuripes d'. a, Pupal aster; b, profile of last segment
of pupa. After Neave (Terzi).

Fig. 184. Tabanus insignis. a, Dorsolateral comb of d^ ; &, pupal aster of cf ;
c, dorsolateral comb of 9 . After Neave (Terzi).

Fig. 185. Tabanus lavcrani 9. a, Pupal aster; b, enlarged view of dorso-
lateral comb; c, pupal aster from the side, showing the combs and the small
dorsolateral comb. After Neave (Terzi).

Fig. 186. Tabanus gratus. a, Dorsolateral comb of d;b, pupal aster of cf ;
c and d, dorsolateral comb of two different 9 . After Neave (Terzi).

MONOGRAPH NO. 13.

PLATE 15.

b 181

(Marchand: The early >tagcs of Tabanida".)

MONOGRAPHS OF THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH

Under the head of Monographs of The Rockefeller Institute for Medical Research are published from
time to time scientific papers which are so extensive, or require such elaborate illustrations, as to
render them unsuitable for current periodical issues. The Monographs are published at irregular
periods, determined by the available material on hand. A number of free copies, which may be SO
or less according to the size of the Monograph, are given to the author. Monographs will be sent
post paid on application, payable in advance. Remittances should be made by draft or check on New
York, or by postal money order, payable to The Rockefeller Institute for Medical Research, Avenue
A and 66th Street, New York, N. Y.

No. 1. Simon Flexner and J. W. Jobling. Studies upon a transplantable rat tumor. Plates 1-16.
J. W. Jobling. The biology of a mixed tumor of the rat. Maud L. Menten. Experi-
ments on the influence of radium bromide on a carcinomatous tumor of the rat. J. W.
Jobling. Spontaneous tumors of the mouse. Plates 17-28. J. W. Jobling. Transplan-
tation experiments in Macacus rhesus with a carcinomatous teratoma from man. (Issued
June 30, 1910.) Price, $1.00.

No. 2, William Travis Howaed and Oscar T. Schxjltz. Studies in the biology of tumor cells.
Plates 1-6. (Issued February 15, 1911.) Price, $1.00.

No. 3, Benjamin T. Terry. Chemo-therapeutic trypanosome studies with special reference to the
immunity following cure. (Issued March 15, 1911.) Price, $1.00.

No. 4, Francis W. Peabody, George Draper, and A. R. Dochez. A clinical study of acute
poliomyelitis. Plates 1-13. (Issued June 24, 1912.) Price, $1.00.

No. 5. J. P. SmoNDS. Studies in Bacillus welchii, with special reference to classification and to its
relation to diarrhea. (Issued September 27, 1915.) Price, $1.00.

No. 6. James L. Stoddard and Elliott C. Cutler. Torula infection in man. Plates 1-9.
(Issued January 31, 1916.) Price, $1.00.

No. 7. Oswald T. Avery, H. T. Chickering, Rutus Cole, and A. R. Dochez. Acute lobar
pneumonia. Prevention and serum treatment. Plates 1-3. (Issued October 16, 1917.)
Price, $1.00.

No. 8. Ernest C. Dickson. Botulism. A clinical and experimental study. Plates 1-10. (Issued
July 31, 1918.) Price, $1.00.

No. 9. James Howard Brown. The use of blood agar for the study of streptococci. Plates 1-34.
(Issued January 21, 1919.) Price, $2.00.

No. 10. William G. MacCallim. The pathology of the pneumonia in the United States Army
camps during the winter of 1917-18. Plates 1-53. (Issued April 16, 1919.) Price, $1.50.

No. 11. Frederick M. Allen, Edgar Stillman, and Reginald Fitz. Total dietary regulation in
the treatment of diabetes. 62 charts. (Issued October 15, 1919.) Price, $4.50.

No. 12. Lewis H. Weed, Paul Wegeforth, James B. Ayer, and Lloyd D. Felton. A study of
experimental meningitis. A series of papers from the Army Neuro-Surgical Laboratory.
Plates 1-17. (Issued March 25, 1920.) Price, $1.00.

No. 13. Werner Marchand. The early stages of Tabanidae (horse-flies). Plates 1-15. (Issued
November 15, 1920.) Price, $2.00.

No. 14. I. J. Kligler. Investigation on soil pollution and the relation of the various types of privies
to the spread of intestinal infections. (In press.)

No. 15. S. T. Darlino. Studies on hookworm infection in Brazil. (In press.)

All inquiries, correspondence, and subscriptions should be sent to the

DIVISION OF PUBLICATION

THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH

AVENUE A AND 66th STREET, NEW YORK, N. Y.

PUBLICATIONS OF THE ROCKEFELLER INSTITUTE
FOR MEDICAL RESEARCH

THE JOURNAL OF EXPERIMENTAL MEDICINE

Edited by
SIMON FLEXNER, M.D.

The Journal of Experimental Medicine is designed to cover the field of experimental medicine.
It is issued monthly, two volumes of over 600 pages each appearing in a year. The subscription
price per year (two volumes) is $5.00, payable in advance. Single copies cost 75 cents. Remit-
tances should be made by draft or check on New York, or by postal money order, payable to The
Journal of Experimental Medicine, Mount Royal and Guilford Avenues, Baltimore, Md., or Avenue
A and 66th Street, New York, N. Y.

THE JOURNAL OF BIOLOGICAL CHEMISTRY
Edited ey
STANLEY R. BENEDICT, New York, N. Y. LAFAYETTE B. MENDEL, New Haven, Conn.
H. D. DAKIN, Scarborough, N. Y. DONALD D. VAN SLYKE, New York, N. Y.

The Journal of Biological Chemistry is designed for theprompt publication of original investigations
of a chemical nature in the biological sciences. It is issued monthly. Each volume consists of as
many numbers as are required to make a total of between 500 and 600 pages. The subscription
price is $3.00 per volume to subscribers in the United States and Canada, payable in advance; $3.25
to those in other countries. The number of volumes issued yearly averages four. The price of single
copies varies; it is given on the back cover of each number. Remittances should be made by draft
or check on New York, or by postal money order, payable to The Journal of Biological Chemistry,
Mount Royal and Guilford Avenues, Baltimore, Md., or Avenue A and 66th Street, New York, N. Y.

THE JOURNAL OF GENERAL PHYSIOLOGY

Edited by
JACQUES LOEB, New York, N. Y. W. J. V. OSTERHOUT, Cambridge, Mass.

The Journal of General Physiology is devoted to the explanation of life phenomena on the basis of
the physical and chemical constitution of living matter. It is issued bimonthly, one volume
of about 600 pages appearing in a year. The subscription price per year (one volume) is $5.00,
payable in advance. Single copies cost $1.00. Remittances should be made by draft or check on
New York, or by postal money order, payable to The Journal of General Physiology, Mount Royal
and Guilford Avenues, Baltimore, Md., or Avenue A and 66th Street, New York, N. Y.

STUDIES FROM THE ROCKEFELLER INSTITUTE FOR MEDICAL

RESEARCH

The published results of investigations conducted in the Institute, or under its grants, are repub-
lished at irregular intervals in volumes designated Studies from The Rockefeller Institute for Medical
Research. The subscription price is $2.00 per volume, payable in advance. A special rate of
$L00 per volume is made to those who subscribe to The Journal of Experimental Medicine, The
Journal of Biological Chemistry, or The Journal of General Physiology. Remittances should be
made by draft or check on New York, or by postal money order, payable to The Rockefeller Institute
for Medical Research, Avenue A and 66th Street, New York, N. Y.

MONOGRAPHS OF THE ROCKEFELLER INSTITUTE FOR MEDICAL

RESEARCH

For information regarding the Monographs see the inside of this cover

SEMIANNUAL LIST

The Institute publishes a Semiannual List of all papers by members of the staff and those work-
ing under grants from the Institute, stating the title and place of publication of these reports.
The Semiannual List will be sent upon application.

All inquiries, correspondence, and subscriptions should be sent to the

DIVISION OF PUBLICATION

THE ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH

AVENUE A AND 66TH STREET, NEW YORK, N. Y.

^^'^hn

SMITHSONIAN INSTITUTION LIBRARIES

3 9088 00597 1312