Full text of "Bulletin of entomological research"

See other formats

Mote e 2
FR OT teat orhtiecin

ted eSe-gies!
Shir ier 8 APE AD I Melt «

Ce lh dal tin be MS oh:
PO SIN tri es ehcp ia > the A Sindh ded
ra°pia lata ttebalerarare-avirgretcate.

PUT MOREE eral el oie! viehripte sy
ere eR Twit ree NP rig ety iphate by tpt gag re
- errr n a FOS wlytacerptetens:

2 : 2 iabamendy pores Vy ewes Pre eree wre
ile a i ie alk | : bw PeVY rr

t?
i A tee
wea ee eae oe eee Sy"
ee ewe ee FW ce F ole ew wee we”

EEE
vr
‘

ie ‘ *: a a te! ete sie cl " weet reer ete
: ceo we era

6 ‘ 4 rch naw ? : : ‘rr ce
Ceres HEE tes gy ; de adhe d a ae we a hey i ae cerry ss ee vn ie ee PEP OO CRETE ENE Pe ee Py
‘vo ere wt + 2 ‘ee ? ve 2 vw & pln 4 Pad . , ri ‘ psy at - ° a . we} . 7 oe ireve eva Teese gee es
eA Ae eI ee : we ere ; wna /
evepane
‘

trere
vy

reer ee ee

22 6s ese yh. tote byt ‘ : rr i — w'w's ere , 4 & ——

7 ‘ves pte a hs - ; 5 : 3 — . —_ ¢ "ew ee i> + * . a . “ . ue ov : ‘ errs
od % . cin e®

Pepe ey wee ote
ve’ Gwetwelel ere ers
ee ee ee

SMITHSONIAN. DEPOSIT

Ps a

Py
Lt

A.
‘ r
i
4 k
Vv ; a) 4
% * wy
te :
Z af
YY °

ENTOMOLOGICAL RESEARCH

ISSUED BY THE IMPERIAL
BUREAU OF ENTOMOLOGY.

EDITOR: THE DIRECTOR.

VOL. X.

LONDON:

1919-1920.

IMPERIAL BUREAU OF ENTOMOLOGY.

Honorary Committee of Management.

VISCOUNT HARCOURT, Chairman.

Lieutenant-Colonel A. W. Ancocx, C.LE., F.R.§8., Juondon School of
Tropical Medicine.

Major E. E. Austen, D.S8.O., Entomological Department, British Museum
(Natural History).

Dr. A. G. Bagsnawe, C.M.G., Director, Tropical Diseases Bureau.

Major-General Sir J. Ross Braprorp, K.C.M.G., F.R.S., Secretary,
Royal Society.

Major-General Sir Davip Brucz, K.C.B., F.R.S., Chairman, Governing
Body, Lister Institute.

Mr. J. C. F. Fryer, Entomologist to the Ministry of Agriculture and

Fisheries.

Sir Sipney F. Harmer, K.B.H., F.R.S., Director, British Museum (Natural
History).

Professor H. Maxwsiit Lerroy, Imperial College of Science and Tech-
nology.

Hon. E. Lucas, Agent-General for South Australia.
Dr. R. Stewart MacDoveaut, Lecturer on Agricultural Entomology,
Hdinbureh University.
Sir Joon McFapynan, Principal, Royal Veterinary College, Camden
Town.
Sir Parrick Manson, G.C.M.G., F.R.S8., Late Medical Adviser to the
Colonial Office.
Sir Danret Morets, K.C.M.G., Late Adviser to the Colonial Office in
Tropical Agriculture.
Professor R. Newstreap, F.R.S., Dutton Memorial Professor of Medical
Entomology, Liverpool University.
Professor G. H. F. Nurratz, F.R.S., Quick Professor of Protozoology,
Cambridge.
Professor E. B. Pouuton, F'.R.S., Hope Professor of Zoology, Oxford.
Lieutenant-Colonel Sir Davip Pratn, C.M.G., C.I.H., F.R.S., Director,
Royal Botanic Gardens, Kew.
Sir H. J. Reap, K.C.M.G., C.B., Colonial Office.
The Honourable N. C. RotuscuHip.
Dr. Hueu Scort, Curator in Entomology, Museum of Zoology, Cambridge.
Sir ArntHur H. Suretey, G.B.E., F.R.8., Master of Christ’s College,
Cambridge.
Mr. R. A. C. Speruine, C.M.G., Foreign Office.
Sir Stewart Stockman, Chief Veterinary Officer, Ministry of Agriculture.
Mr. F. V. TuHeopatp, Vice-Principal, South Hastern Agricultural College,
Wye.
Mr. C. Warzsurton, Zoologist to the Royal Agricultural Society of England.
The Chief Entomologist in each of the Self-governing Dominions
is an er officio member of the Committee.
General Secretary.
Mr. A. C. C. Parxrnson (Colonial Office).
Director and Lditor.
Dr. Guy A. K. Marsnatu, C.M.G.
Assistant Director.
Dr. 8. A. NAVE.
Head Office—British Museum (Natural History), Cromwell Road, London,
5.W.7.
Publication Office. — 88, Queen’s Gate, London, 8.W.7.

: BFE " wit evi
‘ j & |
. fern *s, e . Wiberg eg
’ ¥ Tae * Ws te Seid Tis 4 | Li
yon PRLHASTS | Taye
| * iy a yate : AAe ‘ hid ‘y
Aeeae ; :
if ,
\
sj | ia
; am , ia
cee : : A i : ee j par a
ays ‘ ef ¢ (vale 1 ri y
: . ns
pe 1% ; . : m
cay aha] Bes | e
: a jt i sedate
; | : rau P jad : #« i" ae hey ey rN ‘|
Ab weft Gt i ay ny ed ih an So el Aire im es
ae hs Sa fain | re aa nat fe Vii paugdr ae ne
‘7 ; i es , j wis . . ; ;
rv aya ae i WELT tt dick Y OW A aad ) oe a io r
¥ 3 - a
ae | j ‘eit ii } an a es Pe | ete
: bi it Tk ae aWaliite® he he hin TS Yuta ; Aaa the |
shld heey ules nih Lig TiN i beans ne iW iuaites, ae
ose esl ey 7 l
re ik a GA orn itii v eon ‘pb Niglae a
er tt AL Rg rt LB iii Aste Bil beh
s
ats : : '
; ens, law
i eu
‘
; ¥:

CONTENTS.

ORIGINAL ARTICLES.

ADERS, W. MANSFIELD-. PAGE.
Insects Injurious to Economic Plants in the Zanzibar Protectorate.... 145

Notes on the Identification of Anophelinae and their Larvae in the
AFA Zany Aled toate GE C UOT UCI) oy 0 so) wc a0 eevee! «8 hao] s+ o> Gow dh> atch opegae = enprstass 329

Austen, Masor E. E.
A Contribution to Knowledge of the Tabanidae of Palestine........ 277

Baker, Dr. A. C.
Wicaranaprena wiolae, Lheo,, and its Allies 7.520). 4552462525 c4e5 oes 45

BaRRAUD, Lizut. P. J.
Notes on some Culicidae collected in Lower Mesopotamia.......... 323

Brzzi, Pror. M.
A new Australian species of Reora with a remarkable Life-habit (Dipt. ;
TURF JEIOCI CES) ages a 3 eae rr ee eree ABE.

Notes on the Ethiopian Fruit-flies of the Family Trypaneidae, other
1H SEI g JOCKS Ul aie tick args re ie. Aa ee tee Beare aera 211

Bratn, Cuas. K.
ites Coccidae om South Adricas—ivs SG) See, AA GE 95

Cresrm, i.’ D., Jnr.
Acuschubenrcedimor Caress. pow cont Sob uisite kale oll wad al. 43

Distant, W. L. .
A new Lygaeid Bug found among Stored Rice in Java.............. 41

Duke, Dr. H. LynDHURST.
An Enquiry into the Relations of Glossina morsitans and Ungulate
Game withiepectal reterence to imderpest.. 2.02 02.25..-.02-)--- 7

Epwarps, F. W.

Pios@ui vo NO@vesh spate PLE COR tote ME a oi wine's cee aie, « 129
Fisker, W. F.
Investigations into the Bionomics of Glossina palpalis.............. 347

Incram, Dr. A.
The Domestic-breeding Mosquitos of the Northern Territories of the
Cre Gln Eee erect tS o's) Se us cucievatt Blelaline eco ea Y si wle-ers aye e ede 47

(693) Wt.P2/154. 1,000. 9.20. B.&F.,Ltd. G11/14.

Vi. CONTENTS.

InecraM, Dr. A., & Macriz, Dr. J. W.S.
The Early Stages of West African Mosquitos—iv.................. 59

JACK, RUPERT W.
Tsetse Ply im Southern Rhodesia, AGS 20... ... 22. eee 71

Macrir, Dr. J. W.S.
The Chaetotaxy of the Pupa of Stegomyia fasciaia................ 161

Macriz, Dr. J. W. S., & Ineram, Dr. A.
The Karly Stages of West African Mosquitos—iv................. 59

MacGrecor, Cart. M. E.

On the Occurrence of Stegomyia fasciata in a Hole in a Beech Tree in

Epping Borest.\ . sles oa oe ia Ge eee ee ee een eee een eee oe
MarsuHat., Dr. Guy A. K.
Some Injurious South Atricam Weevil Saye. eyelet eee ee ere 273
Mauvtikx, S.
A new Hispid Beetle injurious to the Oil Palm in the Gold Coast...... al
Mutr, F.
On ‘some, African Delphacidae (Homoptera)--\ 5.22.0 epeee ee 139

NEWSTEAD, ROBERT.
Observations on Scale-Insects (Coccidae)—vi...................... 179

Sranton, A. T.
The Mosquitos of Far Eastern Ports with special reference to the
Prevalence of Steqomyva fasciata, B.S sce See ee 333

TRAGARDH, Ivar.
On the Use of Experimental Plots when studying Forest Insects..... 157

TURNER, R. E.
On a New Mutillid Parasite of Glossina morsitans.........0+-+-+00- BAN

VeItcH, RoBERT. oar olbte of
Notes on the more important Insects in Sugar-cane Plantations in Fyji 21

MISCELLANEOUS.

Collections received: : 20 ii.0. 2. oe ee eee 93, 209, 345, 465

Vi.

PLATES.

PAGE.

ule Wings of Neotoxopiera violae and its allies.......... facing 46
II.—IV. Views of habitats of Glossina morsitans in 8. Rhodesia ,, 90
V.—XII. SOM na AeaMCat COCCIMACK le. sie ents sd oles oe wks we ts
XIIT.—XV Views showing ravages of Oryctes in Coconut Palm.. a LOG
2OWil SOIC NS eL CLS Biv | ko en eee a 208
eV avail Wines of Trypetidae. 32 oii... eee eee ge
XIX. Injurious South African Curculionidae............ ap AS:

MAPS.

PAGE.

PiearGot hx periment OW GIOSSUNG. ca... wee eke bee ee eee es ees to face 17
Charts showing density of Glossonain Uganda.......... 372, 388, 411, 423, 458
Map showing increase of G'. morsitans in Sebungwe District......... io face 90

Pepeol Soummern ty Nodesian (2.00 Nese
Mie menenel SUL ay ORD cote) cove, we) < ol aceu cue, sche 0/4 cirn pd: Sielnce eda) olen «ma, 6 es a. BO

Nout?
ae

ILLUSTRATIONS IN THE

‘

Details of :—

Rioxa termitoxena, Bezzi,sp.n. ..
Rhabdocnemis obseura, Boisd. .
Pupa of Anopheles pretoriensis, Theo.
eae . rufipes, Gough
Ochlerotatus hirsutus, Theo.
Bey gi Be nigeriensis, Theo.
» » Culex ager var. ethiopicus, Edw.
Pe hiie's » guasigelidus, Theo.
» wunivittatus, Theo.
Male cenitalia, Culex aurantapex, Tidw.

39 39

cs 49 Inicranotroms nigeriensis, Muir, Sp. n.

a Ps bridwellt, Muir, sp. n. ay
ae Ae ibadanensis, Muir, SI. Tsny'
li fs Delphacodes bridwelli, Muir, Sp. n.

we Meqamelus furcifer, Horv. ve

is ue Bs kolophon, Kirk as

i x Be nigertensis, Muir, var. n.

flavolineatus, Muir, a n.
Oly peococe cus hempeli. CKIL, g .
Pseudococeus mquilinus, Newst. , Sp. n., on

se perniciosus, N. & W., var., 2

Yilamentosus, CkIll., Q s

Pseud lophilippia inguilina, Newst., sp. 0., ce
Antonina waterstont, } Newst., sp. n., e 183;
Pulvinaria brevicornis, Newst., sp. n.,

- broadwayt var. ae Newst., nov. 4
Lecaniwm subacutum, Newst., sp. n.,
DLecanium (Eucalymnatus) decemplex, Newst. ae sp. 1,

a inguilinum, Newst.,spn., Q . Ke

yh deformoswm, Newst., sp. n., on a

nigrum var. nitidum, Newst., Q ae

Platysaissetia montrichardiae, Newst.,sp.n., Q .
Aspidiotus (Chrysomphalus) apicatus, Newst., sp. Dn. cs

Be umbontferus, Newst. s Be

re articulatus var. magnospinus, Newst.,

A (Odonaspis) rhizophilus, Newst., sp. t., 9

on florineides, Newst.,sp.n., 2 ..
Chionaspis madiunensis, Zehnt., Q

nd fenundieutus. Newel: Spe ile 9

5 praelonga, Newst., Sp. n., 2

a auratilis, Newst., sp. n., @

“a dura, Newst., sp. n., ©

laniger, New st., sp. Tie a
H aematopota sewellt, Aust..sp.n.. ¢& o

Chart of effect of Rinderpest Blood on Glossina on
Trochorrhopalus strangulatus, Gyl. Ar
Rhopaea subnitida, Arrow ; R. pee Arrow
Adoretus versutus, Har: ne
Simodactylus cinnamomeus, Boisd.
Lacon eiehcollis, \QeriACe ee
Monocrepidius pallipes, Esch.
Ampera intrusa, Dist., sp. n.
Portable breeding Gace...

_ OAR),

B40,

TEXT.

x. ILLUSTRATIONS IN THE TEXT.

Map of Northern Territories, Gold Coast :

Map of experimental plot in pine forest, Sweden

Diagram showing distribution of pine pests
Cephalo-thorax of pupa of Stegomyta fasciata

Abdomen of pupa of Stegomyia fasciata .. a ns
Coelaenomenodera elaeidis, Maulik, sp. n., 172; larva..
Aspidoproctus gowdeyi, Newst., sp. n., g genital armature
Aspidiotus longispina, Morgan, 9 puparium

Head of :—
Chrysops punctifera, Lw., 2 4
Haematopota sewelli, Austen, sp. D., J 3 wing of 0
x minuscularia, Aust., sp. Nn. :
Bs minuscula, Aust., sp. 0.
innominata, Aust. a Ml.
Tabanus alexandri nus, Wied.,

i insecuior, nist, sp. n., ie i yt
> rupinae, Aust., Sp. ., 3 Q, 304 : ‘imago, OTe
> arenwagus, Aust. NSD. Dent) a 308 ; imago, ©
>» accensus, Aust., Sp. n., co 311; imago, ¢

5% leleant, Aust., Sp. n., $2 O, 315; imago, 2

>» pallidipes, Aust., Sp. 0, Q, and i imago

dalei, Aust., Sp. D., 3 ; 31 9; wing
Mutilla auxiliaris, Turner, sp. n., 2 ¢
Anopheles funestus, thorax of larva, plumose hair enlarged
Sketch-map of the Kome-Damba Group of the Sesse Islands
harts re Glossina on islands and shores of Victoria Nyanza

158

159

SiG 163
165, 168
173, 174
ss Liregl
. 194

279
283
286
286
286
298
298)
303
306
310
312
318
329
32

330
352

"370, 371, 429. 431, 445

Xj.

ERRATA.
Page 20, line 42, for Sleep read Sleeping
i ee Sky San NACL » world
» 135, fig. I, ,, ame >) | Same
ee AON a 1 ,, Dicranotropis ,, Delphacodes
el] ene 34." 5: Ehyncophorus ,, Rhynchophorus
Pts Se Cyelas » Cylas
* 183, LOO LACES » facies of
he POU ea MMe Saag) So ae)
> 200, 5, 39, »,, pernicillaris » perspicillaris
29 321, ) a, Dy) Mie Oe M
an 465, » 42, 5 Jawaera » Goldberg
Page 64, line 15, for “6 mm.” read ‘‘0°6 mm.”

29, Transpose ‘‘ 8—9 hairs’ and ‘ 5—7 hairs ”’

69, fig. 7, The upper enlarged pecten spine should

bear no secondary spines.

161, line 7, For “ Circum-ocular’’ read ‘“‘ Post-ocular ”’

993

Lol, » “Superior”

28, 29 99 £ yr) 99

29, ,, ‘“* Posterior’”’ » ‘Median ”’

ole 4) “oC Amterion=’ », ** Lower anterior ”’

» ‘** Upper anterior”

33, 5, ‘*Antero-inferior”’ ,, ‘‘ Lower posterior ”’

34, ‘‘ Postero-inferior”’ ,, ‘“‘ Upper posterior ”’

36, After “these setae” insert ‘<(previously referred
= as the inner lateral row.—Bull. Ent. Res.

x, p. 59.)”

41, After «these setae” insert “ (previously referred
to as the outer lateral row.—Bull. Ent. Res.
x) py O94)”

19, For “‘ seta single ”’ read ‘‘seta; single ”’

Si es (iP) 99 Bs 66 (ee) 39

ee ee

Er a A SN nat

IMPERIAL BUREAU OF ENTOMOLOGY.

BULLETIN
OF

ON TOMOLOGICAL. RESEARCH.

Won, X:. 1919.

A NEW AUSTRALIAN SPECIES OF RIOXA, WITH A REMARKABLE LIFE-
HABIT (DIPT.; TRYPANEIDAK).
By Prof. M. Bzzzz,
Turin, Ttaly.

Dr. Guy A. K. Marshall has recently submitted to me a Trypaneid, discovered in
Northern Australia by Mr. G. F. Hill. The fly is said to have been bred from larvae
having habits very different from those of the related members of the family ; and
being moreover interesting from a morphological and biogeographical standpoint,
it forms the object of the present note.

The Oriental species of CeRaTITINAE with 6 bristles on the scutellum, with a
complete thoracic chaetotaxy, with a bristly third longitudinal vein, with a not wavy
second longitudinal vein, and with a wing-pattern consisting of an extended brown
patch with hyaline spots and hyaline indentations, have been ascribed by me* to
the two genera Diarrhegma and Riova. While the first of these genera is less
numerous and is well characterised by the form of the head, the second contains
heterogeneous elements, which have been divided by me into three groups. Fol-
lowing on this, Prof. Hendel} erected the new genus Hexacinia for the peculiar
group of stellata, Macq., and divided the remaining species into the genera Rvoxa
(with a pectinate arista) and Rioxoptilona (with a plumose arista). But this last
distinction seems to be a premature one, as it is impossible to draw a sharp line of
separation between the species in which the arista is plumose on the upper side alone,
and those in which it is plumose on both sides.

I have before me two species of Australian TRYPANEIDAE which are referable to
the genus Rioxa, s.l.; they closely resemble one another both in body coloration
and wing markings, but may be distinguished as follows :— 4

1(2). Two upper and two lower pairs of orbital bristles present; arista regularly
pilose, with some short hairs only on the under side; mesonotum and
scutellum without dark spots, the latter with the middle pair of bristles
weaker than the others; discoidal cell with two hyaline spots, which are
confluent with the hyaline indentations of the hind border

musae, Froggatt

* Mem. Ind. Mus., Calcutta, 1913, iii, pp. 53-175, pl. viii-x (vide pp. 108 and 111

t Wien. Ent. Zeit., 1914, xxxiii, pp. 73-98 (vide pp. 78 and 82).

(C572) P4§140. 1,000. 10.19. B.&F.Ltd. G11. A

PROF, M. BEZZI.

2(1). Only a single upper and a single lower pair of orbitals; arista pilose on both
sides and in the male with a tuft-like pilosity at the end; mesonotum in
front of the scutellum with two blackish spots, which are extended over
the scutellum itself; middle scutellar bristles as strong as the others ;
discoidal cell with but one isolated hyaline spot .. termitoxena, sp. n.

Rioxa musae, Froggatt, 1899.

oe musae, Froggatt, Agric. Gaz. N.S.W., Sydney, 1899, x, p. 501, pl. u, figs.
-3; id., “ Austral. Insects,” Sydney, 1908, p. 308; id. ae on Fruit-fly
sad ihe Pests in various Countries, 1907-8, Dept. Agric. N.S.W., 1908, p. ai
pl. vu, figs. 1-3.
Rioxa (Trypeta) musae, Froggatt, Proc. Linn. Soc. N.S.W., 1911, xxxv (1910), p. 872.
Rioxa musae, Bezzi, Mem. Ind. Mus. 1913, i, pp. 76 and 114.
The present species was originally obtained from bananas brought to Australia
from the New Hebrides; but it has been found subsequently breeding in both culti-
vated and native fruits in Queensland and New South Wales.

2. Rioxa termitoxena sp. nov. (fig. 1).

Differing from all the other known species in having only a pair of upper orbital
bristles, and very distinct on account of the peculiar arista of the male.

42. Length of body, 65-7 mm.; of the ovipositor, 155 mm.; of the wing,
7-75 mm.

Head (fig. 1, A) entirely reddish-yellow ; occiput quite unspotted and shining,
chiefly near the borders ; it is paler on the sides beneath, and the lower swellings are
not developed. Frons opaque, with the middle stripe more dark reddish (sometimes
brownish on the fore half in the female), with more yellowish sides and with a black
ocellar spot; it is clothed with rather dense, short and equal, black hairs; lunula
yellowish. Face shallowly but distinctly concave above the rather prominent
mouth-border ; the broad keel between the antennal grooves is flat; cheeks very
narrow, linear, whitish; jowls broader than the third antennal jot, unspotted.

Fig. 1. Rioxa termitoxena, sp. n.
a, head of male ; b, wing of fully coloured female.

Eyes rather broad, but with the vertical diameter longer than the horizontal one.
Antennae inserted near the middle of the eyes, shorter than the face, and entirely
reddish ; first joint very short, with black hairs at the sides ; second joint short and
globular, with short black hairs; third joint about twice as long as the first two
joints together, almost bare, poe at the end, only slightly concave above:
arista. with rather long and scattered hairs from the base to the end on the upper
side, and on the apical half alone beneath; in the male the terminal hairs of the

A NEW AUSTRALIAN SPECIES OF RIOXA. 3

arista are closer together on both sides, forming thus a kind of apical tuft, which is
not clearly indicated in the female. Palpi broad, yellowish, reddish at the outer
border, with yellowish hairs and some black bristles; proboscis dirty yellowish,
with pale hairs; the hairs of the lower part of the head are likewise pale yellowish.
All the cephalic bristles are black, those of the occipital border being numerous and
pointed ; the postvertical are long; outer vertical half as long as the inner one ;
no ocellar; only a single superior orbital,* the basal one; only a single inferior
orbital, the apical one, which is placed very near the antennae, and is rather weak
and short ; genal bristle well developed, black.

Thorax on the back entirely reddish and rather shining, the very short pubescence
black and the whitish dust very sparse ; the humeri, the sides above the notopleural
line, and a broad postsutural stripe above the root of wings, are of a more yellowish
colour, but not sharply defined; along the dorsocentral lines there is sometimes in
fully coloured specimens a brownish, ill-defined longitudinal stripe ; the two blackish
spots in front of the scutellum are rounded and lie between the praescutellar and
inner pair of the posterior supra-alar bristles, not extending in front beyond the
praescutellar bristles. The pleurae are entirely reddish and shining, quite unspotted,
even the hypopleural spots being not paler than the surrounding parts; the short
hairs are mainly black. Scutellum triangular and of rather large size; it is
flattened above, with well developed lateral keels; it is entirely yellowish, but
on each side there is a triangular blackish or brownish spot in contact with that at
hind border of the mesonotum ; it is clothed on the sides with short black hairs,
and bears six equally strong, black bristles, even those of the middle pair being as
strong as the others, and those of the apical pair being parallel or diverging. Post-
scutellum and mesophragma shining black (in fully coloured specimens), with a
narrow, reddish, middle longitudinal stripe. Thoracic chaetotaxy complete; all
the bristles are black, even the scapular ones, the middle pair of which is as strong
as the lateral ones and has its bristles rather distant ; praesutural bristles rather
weak and short; dorsocentral bristles placed much behind, about on a level with
the praescutellar ones; one strong mesopleural bristle, and 2-3 more weak ones
below it; pteropleural bristle as strong as the sternopleural one. Squamulae and
halteres whitish, the former with white hairs at the border.

Abdomen elongate, slightly constricted at base, and about as broad as the thorax
at the end of the second segment; it is shining reddish, with broad, shining black
bands at base of segments 3-5, the bands being in the female broadly interrupted
in the middle, while in the male they are entire or nearly so, the last segment being
entirely black; the hind borders of segments 2-5 are pale yellowish with whitish
dust, thus contrasting strikingly with the black basal bands. Male genitalia black,
rounded, prominent below; ovipositor shining black, depressed, with the basal
segment as long as the last two abdominal segments together. Venter entirely
reddish-yellow, unspotted. The short abdominal pubescence is black, like the
bristles at the sides and end; the two basal segments are clothed with soft, long
whitish hairs at the sides.

* In the allied genus Pitlona there is likewise but a single s. or., but it is the apical one.
In the recently described genus Ortaloptera, Edwards (Trans. Zool. Soc., 1915, p. 419,

pl. xxxvii, fig. 9), from New Guinea, the cephalic chaetotaxy is very like that of the
present species, while that of the thorax is much more reduced.

(C572) | a2

4. PROF. M. BEZZI.

Legs entirely and evenly reddish from the coxae to the end of the tarsi; their
short pubescence is pale, but it is black on the outer side of the femora; front
femora with long black hairs above, and with 3-5 black bristles below at the end;
middle tibiae with two apical black spurs, and with a short row of 2—4 short and
stout black bristles at the base on the hind sides ; hind tibiae with a complete row.

Wings (fig. 1, B) long and broad, with a well developed costal bristle; the stigma
is of usual length; the second longitudinal vein is straight; the third is bristly
throughout, moderately bent backwards beyond the middle of its last portion, and
parallel with the last portion of the fourth; small cross-vein below the end of the
first longitudinal vein and beyond the middle of the discoidal cell; hind cross-vein
long, as long as or even a little longer than its distance from the small one, reaching
below the fifth vein at an obtuse angle ; point of the anal cell acute and longer than
the second basal cell. The base of the wing is hyaline to a little before the end of
the second basal cell and to the base of the stigma; but in fully coloured specimens
the second costal cell, the basal half of the first basal cell, the base of the second
basal and more narrowly that of the anal cell, are pale yellowish. The stigma is
entirely dark brown, without any hyaline part. The brown patch extending across
the greater part of the wing is darker on the upper half, lighter on the lower, and
distinctly yellowish in some parts towards the middle. At the fore border there is
a single hyaline indentation of triangular shape just beyond the stigma, reaching
with its point the third longitudinal vein, a little beyond the small cross-vein. The
three rounded hyaline spots in the middle are as follows: one at the end of the first
basal cell, before the small cross-vein; one a little before the middle of the first
posterior cell, above the upper end of the hind cross-vein; and one of greater size
before the end of the discoidal cell, in contact above with the fourth vein, and
distant beneath from the fifth vein. At the hind border there are two hyaline
indentations ; a smaller one of more triangular shape in the second posterior cell,
pointing with its inner corner to the upper basal angle of the same cell, without
reaching it ; a larger and broader one towards the middle of the third posterior cell,
with its inner, broad end in contact with the fifth vein, but without entering the
discoidal cell. The axillary lobe is mainly hyaline, with a brown spot at the end of
the anal cell, not extending beyond the axillary vein; the alula is hyaline. The
narrow extreme base of the marginal cell is hyaline, like a small elongate spot into
the first basal cell, just before and above the end of the second basal cell; the
whitish oblique stripe into the brown base of the third posterior cell is very striking.
The indentation of the fore border and the three hyaline spots in the middle are
distinctly whitish, like the inner or superior part of the broad indentation of the
third posterior cell.

Type 3 (British Museum) and an additional specimen of female sex from North
Australia, Port Darwin, 14. x. 1914 (G. F. Hill), bred from galleries of a termite,
Mastotermes sp., in tree-trunks ; the specimens of this couple are not fully coloured,
as is frequently the case with bred flies. Type 9 (British Museum), and an addi-
tional specimen of the same sex, from same locality and same collector, 8, vii, 1913 ;
these specimens are fully coloured, having been taken on the wing.

Macquart (Mém. Soe. Sci. Lille, 1855, p. 144 [124], pl. vu, fig. 7) has described
Urophora bicolor from Adelaide (type in Bigot’s Collection), which is evidently a

A NEW AUSTRALIAN SPECIES OF RIOXA. 5

Rioxa, but differs from both the preceding species in the wing pattern. Froggatt
(Australian Insects, p. 308) reports it as a Trypeta as follows: “A larger native
species with reddish brown head and thorax; with black body; with beautifully
mottled black wings having the base and sides unclouded. I have taken it on the
trunks of wattle trees near Bathurst, N. 8. Wales.” It is probable that this species
is different from that of Macquart; the habit of settling on trunks of trees is very
suggestive of that of the new species here described. In his report of 1909 Froggatt
has briefly compared (p. 114) this 7. bicolor with his own species 7. musae.

Macquart in the above-cited work (1847, p. 109 [93], pl. vi, fig. 7) has also described
from Australia (coll. Bigot) Tephritis lugubris, which is likewise possibly a Rioza,
but has a very different wing pattern and body coloration.

The Trypeta por nia of Walker (List, 1849, p. 1039), from Port Stephenson, is also
probably a species of Rzoxa.

It would be interesting to know if the reduced cephalic oe (by the want
of the second upper orbital bristle) is a character of the true Australian species of
Riora ; in this case it may be considered of generic value, chiefly when supported
_ by such a different kind of life-history as that observed for the new species 7’. term-
toxend.

Of the numerous species recorded or described from New Guinea, only the
following seem to belong really to Rioxa: formosipennis, Walker, 1861; lateralis,
Kertész, 1901; sexguttata, De Meijere, 1913; and seriata, De Metjere, 1915; but
they all have a wing pattern very different from that of the species here described.
The other species, like debeauforii, De Metjere, 1906, nigra, De Meijere, 1906,
msigms, De Meijere, 1913, nigripennis, De Meijere, 1913, and flava, Edwards, 1915,
seem to belong to other genera ; quadrifera, Walker, 1861, was placed by Prof. Hendel
in his new genus Themaroides in 1914.

Pa,

seb laa
mee ce 4. We
y Ay ayy

e
we

Wee ie Wi
ba « 5 sy Re:

AN ENQUIRY INTO THE RELATIONS OF GLOSSINA MORSITANS AND
UNGULATE GAME, WITH SPECIAL REFERENCE TO RINDERPEST.
By H. Lynpuurst Duxe, M.D., D.T.M. & H. (Camb.),

Government Bacterrologist, Uganda Protectorate.

In the course of a conversation in November 1917, Mr. E. Hutchins, the Chief
Veterinary Officer, Uganda Protectorate, drew my attention to an outbreak of
rinderpest in the Northern Province of the Protectorate, which first manifested
itself in the buffalo of the Chopi Country. The disease was then rapidly spreading
through the game in the Glossina morsitans area between Masindi Station and the
Kafu River. He pointed out the excellence of the opportunity for investigating
the alleged relationship between rinderpest and fly. As I had worked from May
till September 1914 in the Northern Province fly belt in association with Mr. W. F.
Fiske, I was in a particularly advantageous position to estimate any alteration
in the fly distribution which might have occurred as a result of the epidemic. The
question had a special local significance, as, should the disease prove fatal to the
fly, the introduction of rinderpest among the situtunga on the Islands of Lake
Victoria would simultaneously remove the reservoir and the carrier of the mammalian
trypanosomes. My release from military duties was duly sanctioned in November
1917, but unfortunate circumstances over which I had no control postponed its
fulfilment until April 1918. By this time the disease had almost worked itself out
in the fly area, and, what is more to be regretted, the long-deferred rains had com-
menced throughout the district. The heart of a unique opportunity was thus lost.

The Masindi fly belt had been studied in 1914 for several months by Fiske, a
highly trained entomologist with considerable experience in the scientific handling
of insect pests on a large scale. His careful observations on the bionomics of the
fly in this belt are of the greatest value in arriving at conclusions regarding the present
problem.

There is a strong consensus of opinion among Huropeans of experience in South
Africa that the great rinderpest epidemic of the nineties resulted in a marked
decrease in the tsetse in the fly areas through which the disease passed, amounting,
in some districts at any-rate, to the immediate or eventual disappearance of the fly.
The explanation of this phenomenon most generally advanced is that the fly dis-
appeared as a direct result of the great destruction by the disease of the wild game
upon which it depended for food. The fly were, in other words, starved to death.

In this paper it is proposed first to review the somewhat meagre evidence bearing
on the interrelation. of game, fly, and rinderpest, and to discuss the same with the
aid of personal experience and observation. Then to proceed to the actual experi-
ments which were undertaken to decide whether rinderpest blood, per se, is or 1s not
destructive to Glossina ; and, finally, to endeavour to arrive at some kind of under-
standing regarding the relations of fly and game and the conditions which lead to a
disappearance or diminution of Glossina. My apology for presuming to attack .
so complicated a problem is that I have recently been privileged to witness the

8 DR. H. LYNDHURST DUKE.

occurrence in Uganda of a marked and independent diminution in the numbers
of two of the best known species of Glossina, and also an outbreak of rinderpest among
the game of a morsitans area. It is to be hoped that the observations made on the
spot may be of value in throwing additional light on the bionomics of Glossina.

My thanks are due to His Excellency the Governor, Sir F. J. Jackson, who has
throughout evinced an interest in these investigations, and has kindly given me
the benefit of his wide experience of game and fly in South Africa. Also to Mr.
Hutchins, C.V.O., whose valuable advice and assistance I have found of the greatest
help. I am indebted also to the Administrative Officers for their assistance in
obtaining food in the face of a general scarcity caused by the famine.

Historical and General Evidence from the Literature.

As regards the disappearance of Glossina morsitans from various districts in
Rhodesia and the Transvaal following on the rinderpest epidemic of the early
nineties three explanatory theories demand special consideration :—

(1). That a fly which absorbs the blood of an animal suffering from rinderpest
dies as a direct consequence. Death might be immediate, and due to poisonous
by-products circulating in the blood; or it might result subsequently as a result
of the development of the rinderpest organism in the tissues of the fly. Again,
rinderpest blood might exert an injurious effect on reproduction.

(2). That the fly disappeared owing to loss of its normal food supply, the wild
game. A special development of this theory supposes that certain species, especially
buffalo, are more responsible than others.

(3). That the phenomenon had no direct connection with the rinderpest, being
due to some obscure coincident cause.

With regard to rinderpest blood, per se, killing fly, we have reliable evidence that
in certain districts no difference was observed in the numbers of the tsetse, in spite of
the passage of the disease through the game of the fly belt. Sir Alfred Sharpe (1)
writes that when the rinderpest visited Nyasaland, there was a noticeable decrease
in the numbers of big game in some of the fly belts, especially of buffalo, “but I have
never noticed that this had any effect on the Glossina.” Val Gielgud (2) writing from
N. E. Rhodesia says: “ Since the almost total destruction of buffalo by rinderpest
these flies have not disappeared, although, perhaps, not so numerous as formerly.
Stevenson- Hamilton (3) remarks, ‘“‘ while there is no doubt that @. morsitans (?) abso-
lutely disappeared from considerable areas during or immediately after the epidemic,
we have it on reliable authority that elsewhere it was in no way affected.” Sir David
Bruce (4), in his evidence before the Sleeping Sickness Committee i October 1913,
states that, in response to his request, Montgomery in British East Africa fed “ a lot
of flies on rinderpest animalsand it didthem no harm.” I recollect sending the pupae
from Mpumu to Montgomery for these experiments, the report of which I have not
been able to obtain. The pupae were those of G. palpalis. I have, however, always
understood that owing to difficulties of transport of the pupae, etc., it was found
to be impossible to pronounce an opinion as to the effects of the rinderpest organism
on the fly, except in as much as the blood of the sick animals was apparently not
immediately fatal, Further evidence on this important point will be produced in a
later section.

THE RELATIONS OF GLOSSINA MORSITANS AND GAME, 9

Regarding the second explanation, that fly disappeared because their food supply,
the game, was destroyed by the rinderpest, there is considerable difference of opinion.
While the majority of the observers hold that the fly are more or less intimately
dependent on the game for their existence, others, and among them Stevenson-
Hamilton, Sir F. J. Jackson and Neumann, point to the existence of fly areas in
which, they state, game is either absent or in numbers inadequate to constitute
the sole food supply for the existing tsetse.

If it were possible to discover an area in which tsetse exist year after year and
from which the presence of ungulate game and their attendant carnivora could be
completely excluded, we should then know that the fly could subsist entirely on
baboons, monkeys, small mammals, birds and reptiles. But, so far as I can discover,
no such game-free fly belt has been found. Even if such did exist, however, it
would not necessarily follow that the sudden or even the gradual disappearance
of the ungulates from a tsetse area rich in game would not ultimately be detrimental
or even fatal to the fly. Apparently rarity of the game is a different proposition ;
the human standpoint may well differ from that of the fly in such matters. Thus,
little is known at present regarding the length of time wild fly can exist without
blood, or the relation between food supply and the production of larvae. I have
kept wild captive palpalis alive without any food for 10 days, and it is highly pro-
bable that fly under natural conditions can exist much longer without blood. The
hardiness of the fly, coupled with the local habits of such game as duiker, dikdik,
warthog and bushbuck, must be borne in mind in dealing with an apparent absence
of game from a fly belt.

One of the best authenticated examples of the occurrence of tsetse in a locality
almost totally devoid of ungulate game is the fly belt mentioned by Stevenson-
Hamilton (3) in the Nyassa district of Portuguese East Africa. His personal
observations as to the rarity of the game at the time of his passage, the dry season
of 1908, must be accepted as correct as coming from a careful observer and a keen
hunter. It is, however, plain from his reports that big game did exist in the
neighbourhood, though in very smal! numbers. He himself mentions a “ few
kudu ” in the narrow strip of country between the fly belts; and Howard (5) re-
ferring to Hamilton’s trip mentions one waterbuck having been shot, the presence
of a few elephant, and the finding on one occasion of buffalo spoor. The party
traversed the bush along a main native track, and their opportunities for studying
the fauna of the belt were limited to the vicinity of this track. Howard produces
evidence showing that 5 or 6 years before Hamilton’s visit this district was full of
big game, and it is evident from the account of Barnett’s trip along the Rovuma
River (6), that in 1910 both game and fly were present in fair numbers in the
country immediately to the north of the belt. Reference to Howard’s map of the
fly distribution in Portuguese East Africa shows that in 1910 the first of the two
belts was a mere focus, possibly a disappearing remnant of the greater numbers
seen by Hamilton. Hamilton describes the fly as in ‘‘ swarms ” and “‘ extremely
troublesome, often biting ourselves and our natives after sundown.” It is, indeed,
usual to find references to areas where game is rare couched in similar terms. But
it is possible, indeed probable, that under these circumstances, their number and
persistence testify rather to the desperate vigilance of flies assembled from a large

10 DR. H. LYNDHURST DUKE.

radius than to a similar prevalence throughout the belt. Hamilton notices that
the fly were concentrated along the main track in the neighbourhood of camps
and halting places, localities in which experience had taught the insects to expect
their supply of blood. They are even reduced to so far departing from their usual
habits as to bite at night time. Fiske, among other reliable observers, has drawn
attention to the following propensities of G. morsitans. ‘‘In the Masindi fly belt,
the only morsitans area in which I have seriously studied this fly, I have noticed
that the passage of a ‘ safari’ of porters or, better still, of a cyclist, will to a very
large extent clear a road of its waiting tsetse, with the result that an observer
following on a short time afterwards may find few or no fly until arriving near the
edge of the belt. The swarm of morsitans which sweeps along in the wake and round
the person of a cyclist passing through a fly belt gives him an entirely false impression
as to the true prevalence of the fly.” I labour this point because the fallacy of in-
ferring from the numbers of tsetse buzzing about the observer a similar infestation
of the belt as a whole is by no means rare.

A consideration of all these points reveals an alternative or, at any rate, a modified
explanation of the facts noted. (Game was very rare in the part of the belt traversed,
and the local fly were probably hard-pressed for mammalian blood. Hence their
voracity and persistence, and their concentration at spots where, in addition to
their natural liking for the edge of a path, they could best satisfy their hunger.
Further, it being the dry season, these fly would tend to concentrate in certain
favoured localities (cf. infra). Probably game such as duiker, oribi and warthog
existed scattered throughout the belt ; also a few large game remained. The exis-
tence of fly in the country immediately to the north of this belt in 1910 may mean
that, in consequence of a migration of the game, the tsetse were gradually shifting
northward to merge with the Rovuma belt, in a richer game country. |

Hamilton himself realises the danger of dogmatism on so important a question
as the result of a single trip ; but none the less, he has laid considerable stress on the
apparent independence of these fly. His facts show that, in certain localities,
morsitans can temporarily adapt itself to the almost total absence of ungulate game.
Natives and their domestic animals, monkeys, birds, reptiles, and probably hyaenas,
were available to the fly in the present instance. Whether the survival under
these gameless conditions 1s permanent, or merely pending an alteration of range,
is at present an unsolved question.

In the majority of the tsetse areas of British East Africa it appears from the
writings of Jackson, Neumann, and others (7) that game occurs only in very small
quantities. Neumann (8), describing the country south of the Athi River above its
junction with the Tsavo, writes : ‘“‘ Here are great stretches of uninhabited bush
country with a perennial river running through it, and hardly any animals, though
plenty of birds and of ‘fly,’ Tsetse.” But in his description of this scanty
game he says “‘a very few waterbuck at intervals, an occasional lesser kudu or two,
a little lot of impala, sometimes at one particular point, and rarely an odd bushbuck,
are all that are seen, without counting the diminutive paa, and now and again a
rhino, giraffe or hippo spoor.”

Here we have a fairly comprehensive menu, however barren the district may seem
to the hungry hunter; yet it is just “sufficiently ‘restricted to render the fly

THE RELATIONS OF GLOSSINA MORSITANS AND GAME. 11

appreciatively attentive to a human visitor. Such conditions are, I imagine, fairly
typical of many of the so-called gameless fly belts. A less keen observer than
Neumann might very well have overlooked some of the species mentioned.

Competent observers such as Woosnam, Yorke, Bevan, Lloyd, Shircore, etc.,
are unanimous that they know of no instance where morsitans occurs in the total
absence of big game (9).

It is possible, however, that the tsetse of a belt are adapted to the special con-
ditions of that belt as regards the distribution, species and numbers of the game.
In a district where buffalo and other big animals are numerous these will be most
easily accessible to the fly, especially in the long grass season. The tsetse will there-
fore learn to depend on these bigger animals. In another belt the regular food supply
will be the smaller antelopes, baboons, warthog and carnivores. Any cataclysmic
alteration in the game distribution of a belt, especially if it occurred at the critical
time when the grass is up, might well be seriously detrimental to the tsetse, whereas
a more gradual alteration would give time for a varying degree of adjustment to the
new conditions.

G. palpalis on the islands of Lake Victoria often shows a deliberate preference
for reptilian over mammalian blood. In other regions, such as the neighbourhood
of the crater lakes near Lake Albert Edward, crocodiles are absent and Varanus
rare,and the palpalis are plainly dependent on the numerous hippo and other game
animals. Evidence of a similar independence in G. morsitans is wanting. The
smaller mammals in morsitans country, though doubtless occasionally fed upon

by the fly, will, by their habits, be difficult to find, especially when the grass is up.

I have found Varanus in the middle of the Masindi morsitans belt, though there
were no fly around at the time. The extraordinary partiality of G. palpalis for this
saurian suggests the possibility of the more fastidious morsitans feeding upon it on
occasion, and this fly has been found to contain non-mammalian blood in nature.
In the present state of our knowledge, however, these creatures must be looked upon
as accessory sources of food, and it is the general opinion of experts that mammalan
blood is necessary at any rate for the propagation of the fly (Lloyd).

My own experience of morsitans, apart from the Masindi belt which will be dealt
with later on, is founded on visits to the following belts in “ German ” Hast Africa
under active service conditions :—(1) The great fly belt on the Tabora and Mwanza
road ; (2) the Buzinza area to the west of Mwanza ; (3) the Namirembe-Biharamulo
belt, possibly continuous with (2), around the south-west corner of Victoria Nyanza ;
(4) between Ikoma and the Mara River and KE. towards Olgoss; (5) the area along
the Kagera River, near the south frontier of Uganda.

In all these belts game is common, excepting the first, in which I spent about 14
days at Shinyanga, well within the fly zone, at a time when game, excepting dikdik,
was rare, having recently migrated in search of water. In none of the above belts
was buffalo found within many miles of my camps. The fauna observed in each
case respectively comprised :—(1) Zebra, Thomson’s gazelle, giraffe, duiker, dikdik
(plentiful), hyaena, roan, impala, carnivores; (2) sable, topi, reedbuck, kudu (few),
hyaena, duiker, roan, impala, zebra, dikdik, giraffe, carnivores ; (3) the same as
in (2), plus rhino and dikdik (very common); (4) Thomson’s and Grant’s gazelles,

12 DR. H. LYNDHURST DUKE.

oribi, rhino, hartebeest, topi, roan, wildebeest, duiker, zebra, impala, dikdik, carnivores ;
(5) roan, impala, oribi, duiker,eland, zebra, topi, impala, wild dog and other carnivores.

In these belts fly persisted throughout the dry ‘season and the burning of the
grass. They appeared much addicted to the neighbourhood of paths, and, in places
where motor transport was employed, illustrated to a remarkable degree the
“following ” propensity of morsitans. On the whole, the fly were not particularly
troublesome, except in connection with motor vehicles, around which they collected
literally in swarms.

In the North-East Transvaal Stevenson-Hamilton (3) states that the fly com-
pletely disappeared immediately after or during the rinderpest of 1896-1897, and
has never come back. Apparently it disappeared also from the adjoining Portu-
guese territory. As regards the effects of the disease on the game he writes,
“after the rinderpest it was found that the last of the eland in the N. EK. Transvaal
had disappeared, and that the buffalo were reduced to a herd of about 20, which
remained in the densest part of the Sabi Bush in the heart of the former fly country.
A small number of kudu and bushbuck survived in the same locality. Impala,
wildebeest, and other species, native to the district, appeared much as formerly.” In
this district the rinderpest arrived among the game at the season of the long grass.

Gibbons (10) writes of North Barotseland: “I may mention that between 1896
and 1897 the tsetse has disappeared in many districts where it was previously
rampant, and is so decimated in others as to be scarcely perceptible.” Again Jack
(11) speaking of the Lomagundi district of Southern Rhodesia says: “It would
seem that the fly in this part survived the rinderpest in very small numbers in
scattered localities, failed to increase to any extent, but persisted until recent years,
although on the verge of extermination.”” Hamilton contends that “‘ in the absence
of any other destructive cause, the fact that even a few bufialo, kudu, and bushbuck
survived, supposing these animals to provide the favourite food of the morsitans,
must surely have induced a partial survival of the latter, a nucleus which in the
course of years would have increased with the increasing herds.”’

From a consideration of the above evidence the following points emerge :—

(1). That there is no known instance of the prolonged existence of G. morsitans
in an area where game is entirely and permanently absent.

(2). That the fly is generally found associated with game in considerable numbers
and variety. It may, however, occur in localities where wild ungulates are
apparently scarce. Under these circumstances the tsetse are probably pressed for
food and their appreciation of the human visitor is apt to give their exasperated
victim an exaggerated impression of the degree of infestation of the belt as a whole.
This impression will be enhanced by the “ following ” propensities of the fly.

(3). That during the rinderpest epidemic of the nineties the tsetse of South Africa
were affected differently in different localities. The fly were either (a) apparently
unaffected ; or (6) markedly reduced innumbers, temporarily or permanently; or
(c) reduced to the point of ultimate extinction. Generally, however, the passage
of the rinderpest throughout the fly country was marked by a great diminution in
the numbers of the tsetse.

(4). That the rinderpest affected some species of game very severely and left others
almost untouched, with the result that a large number of animals escaped destruction.

THE RELATIONS OF GLOSSINA MORSITANS AND GAME. 13

(5) That the theory regarding buffalo as essential to the subsistence of the fly
is untenable. It is possible that the fly of an area where buffalo are very numerous
may resort to this species as its chief food supply. This would, however, apply
equally to whichever of the larger ungulates predominated, and probably depends
on the ease with which these big animals are accessible to the fly, especially at the
season when the grass is high. The arrival of the rinderpest at this season would
enhance its effect on the fly through its food supply.

(6) That much of our information regarding the relations of game and _ tsetse
is based on superficial observations by untrained observers on isolated occasions.
In such circumstances the persistence of small isolated foci may easily be
overlooked.

Description of the Masindi Fly Country.

The Masindi fly area comprises an extent of bush country in which tsetse has
always been fairly numerous and evenly distributed in the wet season. In the dry
season, after the annual grass fires, a diminution in number has been noticed both by
Viske and by the native inhabitants of the belt. During this dry season Viske says that
the fly appear to be concentrate in certain localities, which show special characteristics
and resemble Shircore’s ‘‘ primary centres” (12). Game is fairly plentiful throughout
the fly belt. The main motor road from Masindi Port to Butiaba traverses the belt,
which towards the River Kafu is limited by open plains swarming with kob and
reedbuck. A few small villages are scattered about here and there. The fly scrub
itself consists of scattered small trees and bushes rising above the coarse grass, which
grows in tufts and, in July and August, reaches a height of 6 or 7 feet. Long low
ridges are separated by valleys of varying extent, sometimes narrow and thickly
covered with bush, sometimes wider with thmner bush or open stretches of shorter
grass. In the rainy season these valleys hold swamps. In the dry season there
are two types of waterhole found in the belt :—

(1) Muddy depressions or holes containing filthy water and pitted with game tracks.
These holes are occasionally found on the ridges, but more generally lie along the
bottoms of the valleys. This type of waterhole dries up slowly, and the gaping
eracks in the sun-baked bottom will absorb many showers without leaving any
standing water on the surface.

(2) Punched out holes or ponds often of clear water, situated on the ridges. Of
these holes Fiske says “ They fill and overflow during the rains and usually dry
quickly following their cessation.” Also, and this is probably important, they are
replenished by any chance shower of rain. Such a waterhole is a feature of Fiske’s
“foci,” i which the fly concentrate during the dry season. Other characters of
these foci are freedom from bush, and shorter finer grass; they are surrounded by
bush which is often unusually thick. Game is also always present. Fiske assumes,
as a result of his work in the belt, that morsitans finds particularly favourable conditions
for breeding within these foci, at all events during the least favourable season of
the year, and that it disperses from them freely, especially during the wet season.

No fly are found in the open plains about the Kafu except such as may follow
passengers along the roads. As regards pupae, I have found these in fair numbers
in the crevices of ant-hills and under fallen tree-trunks. Fiske is of opinion that

14 DR. H. LYNDHURST DUKE.

throughout the belt, when the grass is up, large numbers of pupae are deposited
indiscriminately in the grass-tufts, and that the crevices and tree-trunks are more
especially a resort for the dry, short-grass season. He finds that the fly breeds less
freely during the dry season; an observation also made by Lloyd in Northern Rhodesia.
During my stay in the belt in 1914, in May and June, fly were fairly numerous
throughout the area, especially along the various roads traversing the belt. It was
practically impossible to bring a dog along the Nakasongola-Masindi road where
this traverses the fly scrub without 1t becoming mfected with trypanosomes. Cycling
and even marching along this and the other roads was unpleasant to a degree on
account of morsitans ; whilst trekking through the bush in search of game the fly
were in evidence, and often a nuisance.

Observations in the Masindi Belt during April and May, 1918.

On my arrival in the fly area in April 1918, I was at once struck by the marked
decrease in the numbers of the fly met with. I visited all the localities seen in 1914,
including some of Fiske’s foci, and am confident, from a comparison with the number
of flies caught in May 1914, that there has been a very marked reduction. I had
with me fly boys who had previously worked with me in this belt, and they were
unanimous that fly were much diminished everywhere and often unobtainable where
they were formerly fairly abundant. On only one occasion did I find fly at all
numerous, and this was in the immediate wake of a large herd of elephant at a
waterhole. The fly had disappeared from this spot the next morning. Cycling
along the main motor road was no longer a penance, and on trekking about from
point to point in what used to be the most thickly infested parts of the belt the fly
were few and far between. I found no evidence of a corresponding local increase
in any part of the belt north of the Kafu River. The grass was growing rapidly,
though, perhaps, not so long as it was in places during my visitin May 1914. There
is reliable evidence that fly were still worrying cyclists along the motor road throughout
the earlier part of 1917. In May of that year two independent and reliable observers
report that the fly were numerous. In November 1917 Mr. Bain, the District
Engineer working in the Masindi district, tells me that he visited the fly belt along
the main road and was struck by the apparent absence of tsetse.

A considerable number of buffalo were killed by the rinderpest in this fly area.
Bushbuck, warthog and duiker also suffered severely, waterbuck and hartebeest
being little affected. The kob and reedbuck on the Kafu plains also died in
considerable numbers. A certain number of bufialo survived the epidemic.

I next visited the Bugungu fly country, through which the disease had spread
irom Chopi. These parts I had visited in August and September 1914. Owing to
the famine of 1918 it was impossible to “ safari’ through this big area. I therefore
went by sailing boat along the north-east shore of Albert and up the Victoria Nile
to Fajao, camping at intervals along the bank and striking inland to visit the scenes
of my former tour. As a direct result of the epidemic, the bufialo moved south
along the Bugungu plain to the point where the escarpment joins the lake near the
outfall of the Waki River. They left a train of dead along their course, a striking
feature of these bodies being that most of the skins remained intact and the bones
were not generally pulled about by carnivores. Bodies of kob and warthog were
~ also found.

THE RELATIONS OF GLOSSINA MORSITANS AND GAME. 15

Throughout the parts visited, G. palpalis was still found where it occurred in 1914,
and in large numbers. On several occasions I found it abundant, as testified by
actual catching in the bush surrounding the dried-up remains of a buffalo, the sick
animals almost always making for water. G. morsitans was found in various parts
of the area, but here the reduction in numbers was much less marked than in the
Masindi Port belt. In one place, however, I found (and caught) considerable numbers,
and here the fresh tracks of a herd of some 20-30 buffalo were observed. In the
Bugungu region numbers of waterbuck have survived, and elephant are common
and frequent visitors. Bufialo have survived in small numbers out of the numerous
herds which existed throughout the district before the rinderpest. Kob and reedbuck
also occur ; bushbuck and pig used to be common, but I saw none alive during my
visit.

In order to ascertain whether the fly had migrated southwards across the Kafu
River with the buffalo, I sent two reliable fly boys for a three weeks’ trip through the
Buruli district. In 1914 VWiske and I found fly exceedingly rare in this area, but
its existence was known. The boys obtained G. morsitans from several localities,
sparsely distributed. At one spot, some 15 miles south of the Kafu, they described
the fly as numerous. A single specimen of G. pallidipes was taken. It appears,
therefore, that fly are more numerous in Northern Buruli than they were in 1914.
There is no known focus of morsitans in the south from which these fly could have
spread. Whether this increase is a direct result of the rinderpest, it is difficult to
say. The buffalo are in the habit of visiting and crossing the Kafu in normal times,
and it is almost certain that, in their panic of the rinderpest, considerable numbers
must have trekked south into Buganda. The disease has at the time of writing
reached within 40 miles of Kampala, and the main agent in this spread is beyond
doubt the game, especially buffalo.

To sum up, there has been a marked diminution in the numbers of G. morsitans
in the Masindi fly belt, comciding with the passage of the rinderpest through the game
of the district. The fly is not exterminated, but the reduction in numbers is much
more marked than that which ordinarily results from the annual hot season. In
Bugungu the reduction, if such has occurred, is less apparent. There is no apparent
reduction in the numbers of G. palpalis in either district.

In both areas considerable quantities of waterbuck have lived through the
epidemic. Bufialo have also survived in some numbers, though, on the Bugungu
plains especially, the number of dried carcases testifies to a heavy death roll.

Experimental Investigation of Effect of Rinderpest Blood on Glossina.

It was found impossible to obtain G. morsitans in sufficient numbers to carry out
feeding experiments in the actual belt. 1 was therefore obliged to utilise G. palpalis,
and, to avoid all danger of introducing the rinderpest into the populous districts
around Entebbe and Kampala, it was decided to select a suitable island on Lake
Victoria for the scene of the experiments. Nsazi Island some 15 miles south-east
of Hntebbe was chosen, and Mr. Hutchins supplied me with rinderpest blood from
Singo, where immunity inoculations were in progress. He most kindly performed
the inoculations and twice came out to the island to assist me with his advice.
regarding the progress and diagnosis of the disease in the calves. The temperature
charts will be found below. Calf B. R. was seen by Mr. Hutchins, who confirmed

DR. H. LYNDHURST DUKE.

Ui) Daily

G,
2

'
'

27'28|29'30

16.718.

24'25|26

22123
inoculated

1; OO

gd
car (2
=|

+
N

24610 burinp s31g eal]

i}
!

I
{

ai l22

'
'

JULY

(op) -

= —

G ———-—

N] pa ia

a Fil Sa ala gael

0 | —— ©

E Reece

o : — rates ee

= pea at

= Re

2 a se - °

: aa :
=

JULY.

} cies
pee

Pf apace

Se eared aMbh ake sew 1h

bahay vr ow

Wii siaiienecneorsat Ame adi se nor
: Ty bg ee ai “

nee Piety peecnuntnet shi we

L S i ‘ a ¢
< i + hee ete go ng
PORES As ed eto iF doce tal Bela Reten ar
’ i Se ‘

aren aes 8

BULL.

——

Ee
<
o.
2
=
O
>
4
ri
Z|
Wi

~
rm
«
5
G
SO
:
<

be)
S
G
e
*
by
®
<<
b SS)
i

omimencement

THE RELATIONS OF GLOSSINA MORSITANS AND GAME. 17

the diagnosis. The other two rinderpest animals, calves C. R. and D. R., showed
typical symptoms, both before and after death. The following animals were used
in these experiments :—

Calf A—not immunised against and not infected with rinderpest. Kept isolated
from rinderpest animals on the island. Used for control feeding.

Calf B. R—inoculated with citrated rinderpest blood on 6.vui.1918, 34 hours
after removal from sick animal. Died with typical symptoms of rinderpest on
19.vu.1918.

Calf C. R—inoculated on 6.vi.1918 with 20 c.c. of pooled citrated blood from
three rinderpest animals, 30 hours after removal from body; this apparently failed
to infect. Stalled in closest contact with Calf B. R. and died on 24,vu.1918 with
symptoms of rinderpest. Post mortem showed mouth ulcers and acute inflammatory
exudation into lower bowel.

Calf D. R—inoculated on 16.vu.1918 with pooled citrated blood from three
cases of rinderpest, 24 hours after removal from body. Killed on 1.vin.1918 after
showing typical rinderpest symptoms for some days, confirmed by post mortem
examination. |

Sheep E, F, G, H, and K—healthy animals kept for control feedings, and stalled at a
distance from the rinderpest animals. None of these sheep showed any signs of
illness.

Calf A—developed symptoms strongly suggestive of “ amakebe,” 1.e., great en-
largement of lymphatic glands, transient diarrhoea, and anaemia. The blood
proved negative to Piroplasma a few days before it was killed. The animal also
showed extensive skin lesions, which Mr. V. F. Richardson, to whom I described
the case, said were very suggestive of a blastomycotic infection. This animal was
killed on my returning to the mainland, together with all the other animals then alive.
There was no sign of any inflammation of the gut. The animal may, therefore, be
considered as clean as regards rinderpest, and was rapidly recovering condition when
it was killed.

On 17th and 22nd July the dead flies in the experiments were entered on the day ©
following in each case, except in Experiment 22.

It will be seen from this table that rinderpest blood exercises no apparent effect
on the fly.

It was impossible to pronounce decisively regarding the effect on larva production.

A certain number of larvae were produced during the course of the experiments,
of which a number were prematurely aborted. No difference in this respect was
noted between flies fed on rinderpest blood.and those fed on clean animals.

Meteorological Conditions in the Masindi Belt in 1918 and Previous Years.

Masindi is 29 miles along the main motor road to Butiabwa from Masindi Port.
The fly belt extends up to about 12 miles from Masindi Port.

The following table shows the monthly rainfall for the years 1914-17, and
January—May 1918. In regard to these figures, which were taken at Masindi Station,
it must be recollected that the actual fly scrub receives less rain per annum than
Masindi, which is surrounded by high hills. The right-hand figures in each annual

(C572) B

BULL. ENT. RES. VOL.X. PART |.

[Ot H.L. DUKE}

5 3
Number of Flies x DAILY TI AND DEATHS OF MALES we =
in the Experiment pee N = wy REATMENT AN. HS OF FLIES, MALES AND FE: . REMARKS.
di x § £8 ale ie =a] ill ipa > ie alae] ca] =
= & a5 wos > | 3
comms | 8 &2| 8 5
© S$ See =
flies Bg & & & 9 |lo| 7 \/2 | 13 | 74115 | 16 \ 717 | re | 19 @5|26|27|26|29|30)a/ |/ |}2 13 |*#|5)6)7 |e | 9 |10 | 17 | 12 | 12 | 14 | 15'| 16 | 17 | 78 | 19 |20 | 27 | 22 | 28 | 24 | 25 | 26 | 27
= iS
NaS 39
=
Kimmi % ape PEN ew Ps) I) eh NIDA a on A lA “ll
Island 8-1 |8-1|0-0]0:0|2-0}2-0}1-0/2:0]0-0]1-0]0-oj0- : : -110-010-0
x AIlAITAITA[ATASA;)ATATATS Sains = es
“ 0-0|0-0|0-0|0-0|0-0|0-0|/0-0] of 1-0lo-0 7111-0 Control to Expt, 8.
by. A|A[AIAIAIATATA [A TATA AIA eee
a es | 8-0 | 1-0}2-1 |5-1|4-0]2-1 |3:0] 2-0} O-1}1 |0:0 |
x AJlAI[ATA AIAITATATATS Fr
ue 1-0 |0-0]0-0]0-0]1-010-0|0-0]0:0| __|0-1)0-0]0:0|0-0 -0|0:0]0-0
ci R July 16 BSBA RACERS 78S AS BR Es En Sea Si Pee (eS Sait | Fed on calf for the last four days
3-0] 3-0|0-0/0-0|0-0/0:0|0:0/0:0 0 | 0-1 |0-0] 0-0 :0| 1-0 |0:0 of its disense,
R July [5 A A A a A [A BR/BR /BR|BRIBR/| S DOR|S DR Fed on two rinderpest
a Y 17-2| 3:1 |3:2}0-0|0-0|0:0 |0:0|0:0 |0-0] 0-0}0-0|0-0|0-0|0-0 0-0 | animals.
R Meee A|A |A /BRIBRIBR|IBR/IBRIBR|S |[DR|DR]| Ss 2 (DRI
ry uly
6: 0-0 0-0 ” |
A Ss Ss Ss 1 lpoeee
5 A Has rindergest blood any weakening
_|z: 0:0 2-1 |I- : effect on fly? of Ex.2
R A A |S Gicisfelelcf{ec s [kK |[s me fs fis fs] Fed on calP from day after First
u, re 7-0|4-2|1- : 0-0}0-0 0-0/0-0/0-0]0-1]0-0|1-0]0-0 o|0-0]4-1 | 1-0} 5-1 |3-1|0-1 |4-0]2-0|3-0 |2-0|1-0 |2-0 CHEESE!
R ATA s Als G/Gi{s|{c/Glclca S|His [RH |[H|sS [Hs [a [Ss [H Is
pee | ” 4-0/8:1| 1-1 |0-0]2:1 | 2-1|1-0|0-0/0-0]0-0/0-0|0-0]0-0 0-1 |0-0|0:0|4-0|4-0| 1-0] 0-1 -0|1-0}3-1|1-0 | 4-1 |3-2|0-0)0-0]1-0|0:0|1:0 |3:0)4-0 IL ii
R A |A/A/A /BR/BR/BR/BB/SR/(BRIBR] S [DR] S DRIDR|S |/DR/DR|DR|DR H|TH|S |H [H|S [H/|S/TH {STH |S Fad on tno Hinder aasE aE
a a 1-2 |9-4|1-1 |1-0]0-0/0-:0]2-1]0-0 2:0|0-0 | 0-0] 1-0 0:0 |0-0]0:1 |2-0}1-0 |2-:0}0-0 10/71 |6:1 |o-0]2-1 [1-1 | 12 [oo] 04 | F1 |3-0)2-0|4-0
i A |A |e |CRICRICRICRICRICRIBR| S IDR|S DR|DR|S |DR|DR|DRIDR TK STS te els fers fe tS ik 1s
Nsazi
ES nS Island R July 18-19 Z | Fed on three rinderaest animals
Sem 2:010:0|/2:0/0-0|20|3-0/0:1 2:2|0-0/0-0]1-0 0-1 !0-0/2-0/5-0]1-0|0-0]0-0 -0/0:0]0-0] 1-0 | 6-1 |4-0/4-0|0-0/2-0]0-2|0-0]1-0 1-0 | 1-0 | 1: +0) 0- :0|3- : :
ao fETE/ElE/eE/E;SETE!E lpRis lets G | .Gujisn | Gh lnc iicmG KiOofmiys fms ficlS lw lS tx 1s Ip ; 5
2zg|a } Kimmi July 19 Single feed on rinderpest. srimel
| Istand | mle i2rol3-0|3-0 [5-0 1-0 |1 2139. 1-0 |0-1 |0-0|0:0}0-0 9:0 |0:0|0-0|0:0|0-0]0:0] 0:0 0:0} 0-0 |4-0/0:0/0-0/0-0|5-0| 1:1) 11 |3:0|0-0|1-0 {1-1 |: : : : on last day of disease
Baus | R ies | E E | £ [DRIDR|DR]|S DR/IDR|S /DOR|S |F | F KolSele Kol SmpKal Sulake) Sal Kul Salat iS s i
al ees waahs = antes : | __|12:3]3:0} 0-0} 9-0 1:0 |0-0] 0-0] 1-0 |0-0/0-1 0-0} 7-0 |2+1|1-0]2-0|0:0|o-1 12-0 |6:2|1+1 [2:2 |0-0| 1-1 | 3:1 2-0) 0-0]0-0/0-0}1 0 j2'3 on [on
el | 4 see Eaateeaees es zie pS ite) oie S/H |s |) [s) fs (| js |) (ss) ffs i |
| al 2 ‘ [2: _| 1:0 | 0-0] 0-0} 0-0 0-0/0-:0]0-0|1-0 |o-0 0-0|0-0 0:0|2:0|4:0|0-0|1-0 |2:0|0-0| 0-0} 1-0] 0-1 |0-0] 1-0 |0-0/0-0 | 1-0 i:
19 | 2 | Bulago x Ee ETEeVTE;LETeE|S Py Se eee S|H|s | 1S | 1s |/ls [His [Ss [5 |S /s
Island 6-0| 6:0 |0:0]0-0] 1-0 |0-0/2-0|0-0/0:0/0:0|3:0/0-0} 0-0 |0-:0|0-0|0-0|0-0|0-0/0:0 2:0|/0-0|1-0 0-0 |0-0 |2-0|3-0 [2-1 |1-0|4-0|1-1 |0-0]0-0] 1-0 |1-0|0-0/1:0) [
R uly 18 or23 E/E;E;/E;SE,JE Creu or Ey) SojDR DRIDR[S |DR|S |G |G SiepS belo iis fst) oS loa 1S le i Siiw)sS lace ls Fed on two rinderpost animals.
: 1
R NS Sasen 9:0}2:0|4-0|1-0|3-0 |0.0|0:0/0.0) 0.0|0-0/0-0/0-0] 0-0 |0-0|0-0|0-0|0-0/0-:0/0-0 0} 1-0/a-0)0-0 | 1-1 |3-1|1-2 |o-0|5-2| 1-1 |o-1|0-0]0-0|3-0|0-0|0:0 Nees}
r R ali ion Eye Pe jer CRICR/CR|CR|CR/CR CRICR]|S E |e Sas Ere |e Sees 1S lis PS packs K |S K/[s |K
ss ins 0-0|1-0|2-0|0-0|0-0|0-0| 0:0/0:0| 0:0|0-0]0-0|0-0|0-0 Q-0}0-0]0-0]|1-0 |0-0]0-0]0-0 0-00-01: |0-0] 0-1 |o-0/0-1 |o-0|0-0] 0-0 0-0|0:0|0-0/0-0 |o-0|1-0 _|l
R E|E |e IcR|CR|CRICR|CR|CRICRICR|CR| 5S El[Fis/El/eE/fE/e Riso (eS (S'S |s 1s jes alt |
sf | 2 1-0] 8-1 |0-1 |0-0/0-0|0-0| 0-0/0: 0)0-0j0-0]0:0 |0-0 5-0|2-2|2-0|4-0] 2:1 |0-0|0-0 0|0-0|0-0|0-0| 1-0 |6-0|1-0 |2-0|9:4|3-2|/0-2| 0-1 ale alt | I I)
a R efe|/sfelele le Hi{s/H |s /H/s |sS|s ls 7
Island He o-o|2 1 |o-0 2-0}0-0/0-0/0-0 *2|5-4/0-O| 1-3 |3-0 |3:0|}0:0]0-0/1-0|1-0|1'0 at Bia _| ;
R EES (ele Te |e jH |S) |s |S |H 1s 1s [Ss [ |
aL ny i : e {O20} (020111020: /]G70)(Ox0)}t020)/0x01/020 0-0 | 0-0] 1:0 |O-0|0:0|0:0|0-0|0-0|1-1 | 0:0 ll a + ale =
Kimmi awe | ae Fie rye ee | eye SP (Een S| EO SN ton (RS [US PLE ESO || |S eee ESI ES) Fed once on rinderpest wnimal
Island ay, | - | 0-0|4-0|13-:2| 6:0 1-0 | 1-0] 0-0 : o|2-2\2-1|4-0|3-4 |2-2) 1-1
ene )| 0-0 |2-0| 0:0/ 0-1 |3-1 | 0-0] 0:0/2-0/2-1 | 3-4 |3-0]1-3 | 9-0] 1-1 |0-0/0-0]/ 0-0 |2-0|0-0|1-0 |3-0|2
se | I A A[A[S|[A|[ATATA|[S [KITK/[S|TK]¢ [ki s]{xK ris }mis feisi=s fois ise Is im
ry |
| | I J 0-0 | 2-0] 0-0] 0+1 | 1-1 | 0-0!.0:0}0-0] 0-0] 0:0] 0:0] 1-0 | 2-0 | 5:0] 1:0 |0-0| 2+) | 2-0] 0-0] 0-213-0] 0-1 |2:0| 1:0] 70| 42) Ihe] 6-31 75 [21
Bred cut betwee R July 23 (eal fa S [pris |orR|s |pRioR| HIS |S |HIs |H 1S | )S |H|SlH|S/fHlslHIisis|is]s|/sis]s
ned cut Between u ,
SE TE247, 2 0:0|0:0]0:0|0-0|0:0|0:0|0:0]0:0]0:0 |0-0!0:0/0:0|0:0|0-0| 1-0 | 0-1 |o-4 | 0-0|0-0]4-2| 2-0] 0-0] 0-3 | 0-0] 0-0 0-1 | 1:1 | 0-0] 0-2 | 83 8-6] 2:1 1 9°S | 1S
{ = ——|_82' g LOZIEOSO))
Kimmi xX ‘a ETS)TE Ve | ele lm lS isis fx Ss ll Ssa) 1S fxs th papier TS (Ss (TiS TST “|| I; |
Isd. 23-7 : _I| -0|1:0 | 1:0 | 1-0 | 1:0 | 0-0] 0-0] 1-0 | 5-2| 0-0] 0-0] 0-0 |0-0] 2:0] 1-0 | 7-0 | 6:0] 2:1 |0-.0|2:0/3-0| 0:0] 2-0] 3:0| 4-0] 6-0] 4:0] 6:0) 9:1 | $0) 10 SsOi test a =
x ie e/elsie@ irFlse |eolixnxls |s lis |nNis iH | SlH/|S/H|/S [HIS [H[ sis
,
3 1:0 10-0] 0:0} 9-0] 9:0] 0-0/0: |0:0]0:0] 0:0] 0:0] 1:0 | 0-0} 4:0] 6:1 4:0} 9-1 | 2-0] 1-1 2-0|2:0| 2:0| 3:0} 1-0] 0-1 1-0 S Ss Ss S =
in DRIDRIDRIDR|DRIDRIORIS |H|H|S|H|]S |H|/S |H1S /"1S/4/S|H lal Sys isisys
Brad out between R July 26 2:0| 4-0| 3-7| -0| 5-1 | 2-1 |o-4
ca toleh J 4 | 0:0 0:0] 0-0]0-0 | 0:0] 0:0|.0:0|0:0|0:0 |0-0] 0.0] 0-0| 0-0/ 0-1 0-1 | 0:0] 0:1 |O-0/ 0-0 | 0-0] 0:1] 0-0 | 0-1} 0-0] 0-0 }2-1 |e + H ST
Bred oul be July 30 if DRIDR|ORIS |K |S /S]|K]/siwKiS [Si[K{/STS {sys s {sists iS Tris b
red out between sed.
77 + 307 R ees ai} 0:0] 0-0] 0:0] 0:0] 0:0/0:0]0:0] 0:0| 0:0 | 0-0] 0:1 | 0-0] 1-1 |0-0 0.0] 08 | a1 | 41] 2-1] 48 }41{13js0t tt =k
iE sie * | Gap | Sei |ns | S| Kilesmlsels (si ]'s) ls: Siisais isis | ie
red out on “ele 0: 00:0 |0:0] 0:0] 0-0 | 0-:0/ 0:0] 0-0/0:0}01 10-0) 1-7 |2:0 | 3:0} 4-6 | O12}2-2 | 1- hk
29-7 |. Ke I [aa } | LL io) OOO alt | Lo 5 | seal 1 Malby &Sons Lith

9772.P4.140, 1055 O19

In the “Daily Treatment and Deaths” columns the ocourrénce of capital letters A-G, show that the flies fed that day on the calf or sheep corresponding to that letter (see list. above). ‘fhe addition of R shows that the animal had rinderpést.

=fed throughout on clean control animals. The first figure gives the male, the second the female deaths. S—starved that day. j

18 DR. H. LYNDHURST DUKE.

column of the table represent the rainfall taken some 15 miles north-west of Masindi
in a still more hilly and much fcrested district with a proportionally heavier rainfall.
In consequence of the exceptional drought in November 1917 the grass fires of the
1917 hot season began early in December, somewhat sooner than usual. A severe
famine prevailed during the early months of 1918 throughout the whole Province
of Bunyoro, |

Masind: Area Rainfall.

| 1914. | 1915. | 1916. | 1917. | 1918.

Pn ANS: oe da ie B. iW Ber ans AC B. AY B
January 2-31} 1:00} 1:53} 0°84} 0:00] 0:00 4:82] 3°79 | 0:00 | 0-00
February 1:45 | 2°97 | 2:99) 1:14] 3:96] 4:71] 559! 7281 0-00 | 0-42
March | 4°17 | (4:61-] 697 | 3°98 5:12 | 5:01) 71-23 | 1-71 |) 9-6) ae
April | 380] 380] 4371 603| 748] 7:16 | 4:961 6:87 | 10-06
May | 675 | 825] 645 | 810] 993] 576) 865] 7:09 |
June 4°36 | 7:09 3°88 | 4°54 | 2:44] 4:56) 7:04 | 4:45 |
July | 3-46 | 6°70 | 2:57] 2°00 | 0-56} 0-88! 1:49] 1-33
August 3-68 | 3:82 | 2°05 | 2:26! 520! 552] 3:38 | 5:00
September | 7°40] 6°74] G14] 675] 5:94 | 10°79 | 11°85 | 6:80
October 6:00 | 5:10] 689} 9:39 | 8291 10°58} 7:51 | 7°75 |
November | 9-07 9°82] 3°58} 600! 3:80} 4:50/| 0-80] 1:46 |
December | 1:33 | 0°78 | 292| 250.) 2:19] 200| 4:17] 1:99 |

| i

The column A shows the rainfall at Masindi, and B the rainfall at a spot 15 miles _
north-west of Masindi.

In normal years the grass fires begin at the end of December and extend through
January into February. The big rains begin in March. The grass is burnt some-
what irregularly, but the result of the hot season and its fires is that most of the
grass in the fly belt is burnt to the ground and the smaller shrubs severely scorched.
Growing trees are hardly affected ; fallen trunks are more or less destroyed, often
smouldering for days. The grass grows rapidly after the onset of the rains, and
in July and August has reached a height of 6-8 feet in many parts.

Conclusions.

Of the game species available to the fly in the Masindi belt, warthog, bushbuck,
duiker and carnivores are undoubtedly the most important, considering the belt
as a whole. EKlephant and bufialo come and go; hartebeest and waterbuck are
local in their range, especially the latter; the kob and reedbuck of the open plains
sometimes encroach upon the edge of the fly scrub, but are probably seldom bitten ;
and there are, I believe, no dikdik in this area.

In Bugungu the conditions are similar, except that elephant and waterbuck are
much more numerous, and hippo from the Nile and Lake Albert roam far inland.

Fiske (13), discussing the food supply of the Masindi belt fly, applies the principle,
established almost beyond doubt for G. palpalis, that the percentage of females
among the active flies caught is a direct index to the abundance of food animals.
He writes that ‘ The flies in this belt are well fed during April, May, and a portion
of June, but during the latter part of June and in July begin to find difficulty mn
obtaining sufficient food. The percentage of females caught rises, and coincidently

THE RELATIONS OF GLOSSINA MORSITANS AND GAME. 19

the flies become more persistent in their desire to feed upon man. This change in
their behaviour is also coincident with the coming to maturity of the grass, which
in early May would not suffice to hide even a duiker, but by the middle of June would
completely conceal bushbuck and pig in most localities, and larger animals in
some.” :

Now the rinderpest commenced in Chopi a few miles to the north of the scene of
Fiske’s and my work in 1914. The disease spread rapidly through the buffalo, which,
as usual, migrated, some going towards the Bugungu escarpment, some making
south for the River Kafu and Buganda. Right through the long-grass season the
game was dying in numbers throughout the fly area. The elephant, which descend
towards the River Kafu in the rains, were still in the thick elephant-grass country
of the Chopi reserve. Bushbuck and warthog were dying in numbers. The food
supply of the fly was thus severely reduced just at the time when food is, under
ordinary conditions, most difficult to find. The drought, beginning in November,
led to an early drying of the grass and commencement of the fires. Similarly the
intensity and prolongation of this drought led to an unusually complete burning of
the bush. The fly were thus doubly handicapped. Firstly, their food supply was
diminished at the most critical season of the year; and secondly, the adverse
conditions of deprivation of shelter and destruction of breeding grounds were inten-
sified by the prolonged drought.

I consider that these phenomena afford a reasonable and probable explanation
of the remarkable diminution in the numbers of tsetse in the Masindi fly belt. The
Bugungu country, being uninhabited, is less subject to burning in the dry season,
as these fires are started and kept up by natives.

It remains now to be seen whether the principles elucidated in connection with
the Masindi morsitans belt are applicable to the problem of the North-East Transvaal.

In this latter district the rmderpest arrived about June in 1896, and the game were
still dying fast in October, the hot season. The grass reaches its full growth in June
and in some years the fires may not commence until September. I can find no data
regarding the particular climatic conditions obtaining in 1896-7 in this part of the
Transvaal. It is therefore impossible to say whether or not the dry season that
year was unusually prolonged.

One of the two factors already discussed in dealing with the Masindi fly is, however,
common to both cases, i.e., the arrival of the disease in the district at the season of
the year when the grass is up and game most difficult to find.

Regarding the complete and permanent disappearance of the tsetse from a large
strip of country, described by Hamilton as coincident with the passage of the rinder-
pest, it would appear from his paper that he arrived in the district in 1902, some
five. years after the epidemic. From the date of his arrival he is satisfied that no
recurrence of the tsetse in their former haunts has taken place. That the fly
disappeared completely in the season [ think improbable.

The truth probably is that, in consequence of the combination of adverse conditions
afforded by the arrival of the disease in the cold weather, together with a prolonged
dry season with consequent more complete burning of the grass, the fly were very
markedly reduced between June 1896 and April 1897. Small isolated foci probably

(572) B2

20 DR. H. LYNDHURST DUKE.

persisted which, for reasons undetermined at present, instead of increasing,
diminished to the point of absolute disappearance before Hamilton’s arrival in 1902.

This explanation is compatible also with the apparent irregularity in the effects ©
of the disease on fly distribution in different areas. To exert its maximum effect
the epidemic should arrive during the season when the grass is long. In its passage
through the country it may reach certain fly belts when the grass is yet short and the
game easy to see. In such case the effect on the fly might be less severely felt. The
tsetse would be better able to follow the migratory movements of the susceptible
game, and to locate the whereabouts of such relatively immune species as waterbuck,
impala, and hartebeest, whose distribution is local.

The practical outcome of these observations is that morsitans should be attacked
during the dry season of the year, when natural conditions are most unfavourable
to it. The two main indications are :—(1) Destruction of the primary centres
or foci referred to by Shircore and Fiske by filling in or drainage of the water supply ;
destruction of game in the vicinity, and careful burning of the grass at the most
suitable time. (2) Careful control of bush burning so as to obtain the maximum
effect of the fire ; indiscriminate and patchy burning by natives to be stopped, and
the fires started and maintained in a systematic manner so as to ensure universal
burning of the grass.

An observation made by Dr. Sander (14) in German Kast Africa in a pamphlet
on tsetses in 1905 is of interest in this connection. He says that both Europeans and
natives attribute the increase and spread of tsetse in the area between Tanga and
Kilimanjaro to the cessation of annual grass fires. The game in this area had
diminished very greatly. bj

References.

1. Bull. Ent. Res., 1, p. 174.

2, Austen, E. E. Monograph of the Tsetse Flies, p. 290.

3. Bull. Ent. Res., ui, p. 113 et seq.

4. Minutes of Evidence of Departmental Committee on Sleep Sickness; Question

321, |

5. Bull. Ent. Res., 1, p. 40.

6. Bull. Ent. Res., u, p. 44.

7. Austen, EK. E. Monograph of the Tsetse Flies.

8. Neumann, O. ‘‘ Klephant Hunting in East Equatorial Africa,” pp. 141-142.
9. Minutes of Evidence of Departmental Committee on Sleep Sickness.
10. Gibbons, Major A. St. H. ‘Africa from South to North, through Marotse Land.”
11. Jack, R. W. Bull. Ent. Res., v, pp. 97-110.
12. Shircore, J. O. Bull. Ent. Res., v, pp. 87-90.
13. Fiske, W. F. Report on Preliminary Investigations into the Bionomics of
| Glossina morsitans in Uganda.

14. Minutes of Evidence of Departmental Committee on Sleep Sickness. Mr. G. A. K.
Marshall; Quest. 2020.

NOTES ON THE MORE IMPORTANT INSECTS IN SUGAR-CANE
PLANTATIONS IN FYJT.
By Rosert Veitcx, B.Sc., F.E.S.
Entomologist to the Colonial Sugar Refining Company, Lid.*

The writer of the following notes has been engaged on entomological investigations
in Fiji, Australia and Hawaii since 1914 ; the work has been entirely confined to the
pests of sugar-cane and has been carried on mainly in the first-mentioned country,
where insect depredations are very serious and materially reduce the profits on
a considerable number of plantations. The investigation of these Fijian cane
pests open up a field that is practically new, and so the work will probably continue
for some considerable time. The length of the period that is likely to elapse before
the publication of final results is such that it seems desirable to place on record a
few of the main points in the life-histories of the more important insects of the
sugar-cane plantations of that country, and so the writer has compiled these notes
containing summaries of their life-histories in so far as they are at present known
to him.

The Sugar-cane Industry m Fj.

The six sugar-cane mills operating in the Fiji Islands are all modern well-equipped
factories handling large quantities of cane. Four of the mills are the property
of the Colonial Sugar Refining Co. Ltd. (Fiji & New Zealand), the remaining two
being owned by other interests. The writer is not conversant with the conditions
at the latter mills, and so his remarks on the status of the various insects are based
mainly on observations made in the districts supplying the mills of the Colonial
Sugar Refining Company, namely, Lautoka, Rarawai, Nausori and Labasa, the
first three being situated on the island of Viti Levu, and the fourth on the island
of Vanua Levu. Nausori is the only one situated on a windward coast, and 1s
generally spoken of as a “ wet” mill on account of the rainfall there being so much
heavier than at the other three, which are usually referred to as “dry.” The
use of the term “ dry ” does not mean that any of the mills so designated suffer from
an insufficient rainfall; on the contrary it is generally ample at all the mills, and
indeed in some seasons it is excessive. The average rainfalls for the eight seasons
ended December 1916 are as follows :—Nausori 120 inches, Labasa 87 inches, Rarawai
82 inches and Lautoka 66 inches. Most of the rain falls during the hot season,
which lasts from December to April.

In Fiji sugar-cane is grown on comparatively small isolated blocks of land, which
generally form a narrow interrupted strip along the coast, or along the banks of
the more important rivers, large compact areas such as exist in Hawaii and other
sugar-cane countries being quite unknown. The smallness and isolation of the

* Published by permission of the General Manager,

22, ROBERT VEITCH.

cultivation, the interior being one mass of mountains, which not infrequently run
down to within a mile of the coast line. The most profitable soil for the growth
of sugar-cane is found on the alluvial flats along the banks and near the mouths
of the rivers; between these cultivated river flats and the mountains there is a
variable quantity of undulating land and foothills on which some cane is grown,
but only a small percentage of that land can be profitably put under sugar-cane,
and consequently very little cane is grown at an elevation of more than one hundred
feet above sea-level.

The cultivated soils may be classified under the following headings: (1) alluvial
flats, (2) red hill soils, (3) sands, (4) stiff clays and (5) “ tiri” or reclaimed salt-
swamps. The alluvial flats are very rich and give heavy crops of cane, while the red
hill soils are deficient in humus and give a light but sweet yield; the sands give
good sweet crops in wet years, but in dry seasons the results are frequently poor. The
stiff clays are difficult to work, and consequently they rarely give good returns.
The reclaimed salt-swamps vary greatly in physical properties, but are all alike
in giving only a medium or poor yield of cane.

The season for replanting fields that have been under fallow is in the early
months of the year, while the harvesting season usually commences in June and
ends in December.

The leading varieties grown are Badila, Malabar and Pompey. Badila is the
best all round variety, and in 1917 it constituted 97 per cent. of the crop at Lautoka,
97 per cent. at Labasa, 88 per cent. at Rarawai and 32 per cent. at Nausori. At
the latter mill Malabar is the leading variety, as it is believed to be the most suitable
for the poorly drained flats surrounding that mill| Pompey is now being exten-
sively grown on the second-class lands at Rarawai, and on such soils it seems to give
a better financial return than Badila.

The general rotation 1s a 25 per cent. one, which means that 25 per cent. of the
land is under plant cane, 25 per cent. under first ratoons, 25 per cent. under green
fallow and 25 per cent. under preparation for replanting. Mauritius bean (Stzzo-
lobium aterromum) is the green fallow crop usually ploughed in, although cowpea
is also used to a certain extent.

The Cane Beetle Borer (Rhabdocnemis obscura, Boisd.) (fig. 1).

The beetle borer has long been known as the most destructive pest of sugar-cane
in Fiji, where it now constitutes a very serious entomological problem. The available
evidence indicates that its original habitat is New Guinea, whence it has spread
to North Queensland, Fiji, Samoa, Tahiti, Hawaii and other islands of the Pacitic.
It is believed that the pest reached Fiji in a consignment of seed cane imported
from Hawaii.

The borer shows a marked preference for soft rank cane, and as this occurs
most commonly in the heavy crops on rich alluvial flats, it follows that this pest is
more destructive on such soils than elsewhere. The amount of damage does not
depend solely on the type of soil on which the cane is grown, for it has been found
that the susceptibility of the different varieties varies greatly. Badila is a soft
sweet cane liable to fall and rot, while Malabar is a hard cane with a low sugar
content and with a less marked tendency to fall to the ground when heavy tonnages

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 23

are grown, and as the borer prefers soft sweet stalks of cane to tunnel in, it almost
invariably happens that on similar types of soil Badila suffers much more severely
than Malabar.

The white elongate oval egg of this pest is about one-sixteenth of an inch in length,
and is laid in the rind of the stalk behind a half-loosened leaf-sheath, and a week
after oviposition the grub hatches out and bores into the cane, becoming full-grown
in eight to eleven weeks. In boring the cane the grub destroys a very large amount
of tissue, partly by eating and swallowing it, and partly by chewing it to extract
the sap. The tunnels made are not infrequently two feet long, but generally they
do not extend to more than six inches; in tunnelling the cane the grub frequently
eats too close to the rind, and in doing so it makes small rupture holes about the
size of a pin’s head. The borer grubs can be readily located by the presence of these
holes, which stand out somewhat conspicuously as a result of the rind im their
immediate vicinity being of a rather lighter shade than elsewhere. The full-grown
grub measures somewhat more than half an inch in length, and is legless, wrinkled
and yellowish grey in body colour, the head being light reddish brown. The fifth
and sixth adbominal segments are greatly swollen ventrally.

Fig. 1. Rhabdocnemis obscura, Boisd., x 3.

The full-grown grub prepares for pupation by eating a hole in the rind of the
cane somewhat less than a quarter of an inch in length, and at a short distance
above this hole it pupates in a cocoon of fibres separated from the tissue in which
it has been tunnelling. The cocoon is extremely strong, and as it is very difficult
for anything to penetrate it, the pupa is generally immune from the attacks of the
small brown ant (Pheidole megacephala, F.); were the cocoon less strongly made
a large proportion of the pupae would undoubtedly perish, for this ant is extremely
voracious and a very formidable predator.

24 ROBERT VEITCH.

The newly formed pupa is cream-coloured, but as the pupal stage advances the
the darker markings of the future beetle appear and eventually show up distinctly.
The pupa is half an inch in length and the pupal stage lasts for about two weeks.
The newly formed beetle rests in the cocoon for a considerable number of days before
venturing forth to feed and mate, and when it does so, it leaves the cane by the hole
made by the grub prior to pupation. The period that elapses between the laying of
the egg and the emergence of the fully hardened beetle is between three and four
months. The beetle is half an inch in length and is a typical Curculionid ; the colour
is dark brown, broken by several black patches on the thorax and elytra. The sexes
can be distinguished by the fact that in the male the rostum is shorter, straighter,
thicker and also rougher on the under surface than in the female (fig. 1).

The damage caused by this pest is often very severe, for it 1s no uncommon occur-
rence to harvest fields in which it is difficult to find a single stalk unattacked by it.
The percentage of stalks damaged by borers at the Fiji Mills during the 1917 crushing
season was 14, a figure that is sufficiently high to cause considerable anxiety. The
activities of this pest lead to a very considerable reduction in the crop, for the sugar
yield is affected in several ways. In the first place, there is the loss of sugar in the
cane tissue swallowed or chewed by the borer grubs, and in badly attacked fields
this is very considerable, because many of the stalks are so extensively tunnelled
that they contain little or no sound cane. The loss of sugar is not confined to
the tissue actually eaten, for it has been shown that the boring of the grubs
materially reduces the sugar content of the untunnelled portions of infested
stalks. Another source of loss in attacked fields is the increased lability
of bored stalks of cane to break and fall to the ground, where they become still further
infested and rot rapidly. The general effect of the attacks of this pest is to depreciate
the value of a large proportion of the crop on the richer lands, and to render much
of it absolutely valueless for milling purposes.

The following control measures have been found to be of great assistance in Fiji :-—-

(1) Seed cane for replanting fallow fields should be cut in areas known to be free
from borer ; where this is not possible the seed used should be carefully inspected,
and all infested stalks should be rejected and destroyed.

(2) The trash on badly infested fields should be burned as soon as possible after
cutting, as by so doing numbers of beetles underneath the trash and most of the eggs,
stubs, pupae and beetles in the discarded cane will be destroyed. Cane stalks that
are badly diseased or badly attacked by the borer are generally left on the field,
and unless the trash is burned off it finds an excellent breeding ground in such
discarded cane. Trash should not be burned on poor land, because on such soils 16
is necessary to preserve all humus-forming material ; fortunately, however, cane on
poor land is seldom badly infested by the weevil, and so there is rarely any necessity
for burning trash on such soils.

(3) Fields to be replanted should be ploughed out as soon as possible after
harvestins, thus depriving the pest of such an excellent breeding ground as neglected
volunteer ratoons.

(4) The beetles should be collected by means of traps of split canes laid in suitable
fields. These traps attract large numbers of weevils which can be collected and
destroyed, i

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 25

A prolonged effort was made to increase the efficiency of the control of this pest by
the introduction of a Tachinid fly parasite, Ceromasia sphenophori, Vill., which
was introduced by Mr. F. Muir from New Guinea to Hawaii, where it soon brought
about a reduction in the losses on the plantations. As the conditions in Fiji appeared
to be favourable to the introduction of the parasite, it was thought possible to repeat
the success, and accordingly Dr. J. F. Ilhingworth, of the College of Hawaii, was
commissioned by the Colonial Sugar Refining Company to introduce the parasite
to this country. He arrived on 30th May 1913, and succeeded in rearing the parasite
in breeding cages, from which he liberated large colonies, all of which failed to become
established. He then returned to Hawaii after spending two or three months in
Fiji, and from the date of his departure until the end of 1914 the work was carried
on by Mr. D. 8. North (Mycologist to the Company) ; from January 1915 to March
1917 the work was superintended by the writer. Mr. North was successful in estab-
lishing the parasite in several fields on Natova plantation, and it looked as if success
was in sight, but unfortunately this field colony died out shortly after the breeding
cage on the plantation was closed. The writer’s experience while in charge of the
work was similar ; so long as the breeding cages were maintained, it was possible to
find considerable numbers of parasitised borers in the fields adjoining the cages, but
when these cages were closed down the field colonies dwindled and eventually died
out ; 1t is thus evident that these field colonies depended for their continued existence
on a steady stream of new parasites from the breeding cages ; when that steady stream
ceased the rate of multiplication in the field seemed unable to counterbalance the
very heavy mortality rate, and so the colonies gradually died out. The effort to
establish the parasite lasted almost four years, and ended in failure in spite of the fact
that very large colonies were liberated, and that at one time breeding work was being
carried on at five different centres in order to give the fly a trial under all the different
conditions prevailing in the sugar districts. This failure illustrates the uncertainty
of all parasite work ; for the success in Hawaii seemed to assure an equally favourable
result in Fiji. Climatic conditions cannot be held responsible for the failure, for the
parasite was bred in very large numbers in captivity, clearly showing that the climate
was favourable. The writer is of opinion that the large jumping spiders and the
small brown ants, combined with peculiarities in agricultural methods and in the
growth habits of the leading variety of cane, have been responsible for this disappoint-
ing and unexpected result.

The ease with which this parasite establishes itself when conditions are favourable is
shown by the fact that it became established in North Queensland from a few flies
that escaped from an intermediate breeding cage used by Muir and Kershaw when
they were transporting the parasite from New Guinea to Hawaii.

The Small Cane Beetle Borer (Trochorrhopalus strangulatus, Gyl.) (fig 2).

This weevil is frequently found in cane. It is much smaller than the more des-
tructive species, from which it can be easily separated in the larval, pupal and adult
stages. The larva of the larger species has the 5th and 6th abdominal segments
distinctly swollen ventrally, a feature which is absent in this insect ; and in the pupal
stage the two can be readily separated by the fact that the smaller species does not
make a cocoon. The adults can be distinguished by their colour, the smaller species
being unicolorous, while the larger one has its brown colouring broken by black

26 ROBERT VEITCH.

marks on the elytra, the thorax and elsewhere. Other less easily observed
differences exist, but the above are quite sufficient to separate the various stages.

So far the writer has never observed this species attacking perfectly sound cane,
and his experience is that it generally breeds in rotten fallen cane, and less frequently
in sickly standing stalks. Were it to turn its attention to sound cane it would un-
doubtedly become a serious pest, as its rate of multiplication is very great ; but at

Fig. 2. Tvrochorrhopaius
strangulatus, Gyl., + 5.

present its importance is very slight. The control measures recommended for the
larger cane beetle would also be effective with this species.

The White Grub of Sandy Soils (Rhopaea vestita, Arrow) (fig. 4).

White grubs frequently occur in large numbers in certain types of soil, and are
often the cause of very severe damage, so serious in fact that several tracts of coastal
land have been thrown out of cultivation because of bad infestation. Two species
have been found attacking cane on an extensive scale,—and being new to science,
they have been described by Mr. Gilbert J. Arrow (Ann. Mag. Nat. Hist. (8) xvi, p. 319)
under the names of Rhopaea vestita and Rhopaea subnitida ; both species are natives
of Fiji.

The local distribution of these grubs is very interesting; for on no occasion has
the writer found the two species together in the same type of soil. A. vestita has
always been found in sandy fields situated close to the coast, while A. submitida
occurs in rich alluvial lands and in red hill soils, but not in sandy lands. Although
the latter species is found in alluvial soils, it is never present in large numbers
and has never been observed doing appreciable damage ; for although it undoubtedly
attacks the cane, its numbers in such soils are so small that it has no apparent
effect on the crop. On several occasions, however, this same species has caused
considerable damage on red hill soils. &. vestiia is responsible for extensive injury
year after year on sandy soils, and is by far the more destructive of the two species.

The life-history of R. vestita has been more extensively studied than that of
R. subnitida, and the main features of it are outlined in the following paragraphs.

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. ii

The finely reticulated, elongate oval eggs are on an average about 2°5 mm. in
length and 1:25 mm. in breadth when newly laid, but the egg increases in size as
the incubation period advances, and just before hatching it is spherical rather
than elongate oval, the diameter being approximately 3°5 mm. The eggs are laid
singly at a depth of 15 to 20 inches, having an average incubation period of 31
days, the maximum and minimum being 34 and 28 days respectively. Dissection
shows that each of the paired ovaries consists of six tubes each producing three
eggs, the maximum number a female lays being 36, although in captivity it is
usually considerably less. The eggs are laid from May to September inclusive,
the chief egg-laying months being June and July.

The grubs feed on both living and decaying vegetable matter in the soil, very
large quantities of soil being swallowed in order to extract this ; but living vegetable
matter, such as cane roots, is also attacked, and in this respect the damage is
particularly severe in January, February and March, ie., when the grubs are well
grown. When full-grown the white grub averages 27 mm. in length and 6 mm.
in breadth; the body colour is creamy white, with the exception of the last
abdominal segment, which appears greyish black as a result of the contents of
the alimentary tract showing through the skin. The great majority of the grubs
are found within six or seven inches of the surface of the soil all the year round,
and so they are always within reach of the plough.

In April, or, even earlier in some cases, the grubs enter the pre-pupal stage which
lasts for a couple of weeks, at the end of which time the pupa is formed in an earthern
cell at a depth of 6 to 18 inches below the surface. The pupa averages 19 mm.
in length and 8 mm. in breadth; the pupal period lasts for 31 days. Pupae are
found from April to September, but are most plentiful in April and May.

The beetles do not leave the soil immediately after the transformation from
pupa to beetles, but lie resting in their earthern cells for two or three weeks before
emerging to feed and mate. Beetles have been found in the soil on 7th February,
but this is unusually early, for they do not reach their maximum abundance until
June. During the swarming season they are found in thousands on their food-
plants, the chief of which are acacias, guavas, coconuts and cane, but for the
purpose of mating they will also swarm in large numbers on any prominent object
such as windmills, telephone poles or large trees. In the swarming season the
beetles emerge about twenty minutes after sunset, and mating proceeds for two
hours, after which they settle down on their food-plants, where some of them feed.
while others remain quite inactive; shortly before sunrise they return to the soil
or to their hiding places under cane trash and fallen leaves. The most prominent
feature of this return to the soil is the rapidity with which it is accomplished, as
not a single beetle is to be seen a couple of minutes after the first one has taken
flight.

The beetle is reddish brown in colour, and is clothed with very small yellowish
setae, except on the breast, which is covered with long yellow hairs. The average
length is 17 mm. and the average maximum breadth 9 mm.

The life-history may be summarised as follows :—The female lays twenty to
thirty eggs singly in sandy soil in June and July at a depth of 15 to 20 inches, and
after an incubation period of 28 to 34 days the grubs emerge and feed on humus and

28 ROBERT VEITCH.

living vegetable matter, being most destructive in January, February and March
of the next year, by which time they are nearly full-grown. They pupate in April
and May in earthern cells at a depth of 6 to 18 inches, and after 31 days or so in
the pupal stage the beetles appear. The adults do not leave the soil immediately
after their transformation, but rest in the earthern cells for a week or two, and then
emerge to feed and mate and repeat the life-cycle. There is only one generation
in a year.

This pest is very destructive on sandy soils, in which it attacks cane of all ages.
Fields replanted in January, February and March almost invariably suffer severely,
as the grubs at that time are well grown and immediately attack the young plants,
destroying the roots, or eating the young shoots at the base and thus severing them
from the set and the root system. Many of the attacked plants die, but others
survive, although more or less crippled, the older cane being better able to with-
stand the attacks, as its recuperative powers are greater, but even such crops are
frequently severely checked.

The following natural enemies take a heavy toll of the white grubs of this species.
The wireworms of Monocrepidius pallipes, Esch., attack and destroy the grubs
in large numbers, and the Scoliid wasp, Discolia ovalauensis, Sauss., parasitises
very many ofthem. An unidentified species of mite attacks the grubs in considerable
numbers when kept in captivity, and the attacked individuals generally die. In
the field conditions do not seem to be so favourable to the mite, and it is only rarely
that one there finds grubs as badly infested as in the laboratory. Minah birds
destroy large numbers of grubs during cultivation operations, and are undoubtedly
of great assistance in minimising the losses from this pest.

Fields known to be badly infested by white grubs should not be planted until
April, as by that time most of the grubs will be in the pupal or prepupal stage,
and so the crop will be free from serious attack until December or January, when
the cane will be well established and so better able to resist the grub attacks. The
collection and destruction of beetles and grubs is also of assistance in controlling
this pest, while the frequent ploughing of infested fields will expose many grubs
to destruction by minahs.

The White Grub of Alluvial and Red Hill Soils (Rhopaca subnitida, Arrow) (fig. 3).
The cane grub of sandy soils has been dealt with in considerable detail in the
preceding paragraphs, but a much briefer reference will suffice for the less destructive
species, R. subnitida, partly because it is of less importance, partly because less
detail is required owing to the similarity between the life-histories of the two
species, and partly because it has been less closely studied than RA. vesiata.

The seasonal occurrence of the two species is practically the same, and so far
the writer has not observed any difference in feeding habits. The beetles eat
acacia, coconut, cane and guava leaves, but the amount of food consumed is small.

The beetle, pupa, grub and egg of R. submitida are each larger than the corres-
ponding stages of R. vestita. The beetle of the latter species can always be readily
identified by the presence of a smooth median area on the pygidium, which in the
former is setose all over. The writer cannot distinguish between the pupae and

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 29

eggs of the two species, but the well grown grubs are not difficult to separate, as
that of R. subnitida is more hairy and has longer legs than the grub of R. vestita.

The egg-laying records of females kept in captivity and the dissection of ovaries
show that in reproductive powers the species are similar. It is therefore difficult
to explain why this grub should be so much less numerous than the other. The
beetles are just as hardy, and in the laboratory the grubs are not nearly so easily
injured as those of the sandy soil species. The explanation of the relative scarcity
of this grub will probably be found in the presence of some enemy which attacks
R. subnitida only and has so far escaped observation, or in the existence of some
other adverse factor in the soil types frequented by this grub. This is a point
worthy of further investigation.

The control measures recommended for R. vestita will also be of agsistai:ce in
checking the destructive activities of this species.

ql er i
a A
a
Fig. 3. Rhopaea subnitida, Arrow, X 4; Fig 4. Rhopaea vestita, Arrow, x 4:
a, portion of elytra highly magnified @, portion of elytra highly magnified
to show the scaling. to show the scaling.

The Rose Beetle (Adoretus versutus, Har.) (fig. 5).

The rose beetle is not generally regarded as a pest of cane, but on a recent plantation
inspection the writer observed the grubs of this species attacking the germinating eyes
of cane sets planted some two or three weeks previously.

The grubs of this Rutelid beetle have been found in all types of soil, and almost

invariably occur within six inches of the surface. The beetles are found all the year
round and are a great nuisance to all rose growers, as they attack and destroy the

20 ROBERT VEITCH.

leaves of the rose bushes, and in many districts it is impossible to grow roses suc-
cesstully unless they are screened every evening before sundown.

Fig. 5. Adoretus versutus,
Har, <4.

The beetle feeds on the leaves of sugar-cane, guava, acacias, cacao and rose bushes,
This species is fortunately of slight importance in the cane-fields, and no control

measures are required.

The Introduced Parasite of White Grubs (Scolia manilae, Ashm.).

While on a recent visit to Hawaii the writer was much impressed by the degree
of success obtained in the introduction of Scola manilae, Ashm., to control the
white grubs of Anomala orientalis, Waterhouse. This pest was accidently imported
from Japan in quite recent years and is now firmly established on two plantations
on the island of Oahu, where it is seriously reducing the cane crops. A considerable
number of parasites were introduced to control it, and of these the Scolia shows most
promise ; at the end of 1917 it had become extremely numerous, and showed such
splendid prospects of success that the writer determined to take a colony back to
Fiji, in the hope that they might attack the Fijian cane grubs and rose-beetle grubs.

The officers of the Hawaiian Sugar Planters Association gave all possible assistance,
and supplied laboratory accommodation while the writer was breeding the colony, and
also much helpful advice based on their own experience in handling this parasite.
The colony was successfully transported to Fiji, and six hundred fertilised females and
over two hundred males were liberated in an area that is always badly infested with
stubs. It is still too early to say whether or not success has been attained, but the
laboratory trials showed promising results, for the parasite attacked the grubs of
R. vestita and A. versutus and laid eggs on them ; adult parasites were subsequently
reared from some of these eggs. A distinctly unfavourable factor was introduced a
few weeks after the parasites were liberated, for a heavy flood occurred, resulting
in a large portion of the grub-infested area being under water for a few hours, and if
larvae and pupae of the parasite were present in the soil, it is almost certain that
a considerable number of them perished.

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. Pl |

The Sugar-Cane Wireworm (Simodactylus connamomeus, Boisd.) (fig. 6).

The following species of wireworms have been found in the soils of the Fijian cane-
fields : Simodactylus cinnamomeus, Boisd., Lacon stricticolis, Fairm., and Monocre-
pidius pallipes, Esch. The first two species, but more particularly S. cownamomeus,
are very destructive to young cane plants, while the third is distinctly beneficial, as
it is a very formidable enemy of the white grubs of Rhopaea vestita.

S. cinnamomeus is by far the commonest of the three species, being found in all
five types of soil, but reaching its maximum abundance in the rich low-lying flats,
where it frequently occurs in enormous numbers.

The full-grown wireworm is sub-cylindrical in shape, and measures 25 mm. in length
and 2mm. in breadth. The dorsum varies in colour from a very light brown to a
blackish brown, an intermediate shade being typical ; the pleura and venter are pale
brown. The newly hatched wireworm measures 2 mm. in length, and for a few days
is of a semi-opaque white colour, but rapidly assumes a faint brown tint. It grows
very slowly, the available evidence pointing to a larval period of not less than three
years. The older wireworms under laboratory conditions moult every ten or twelve

weeks.

Fig 6. Simodactylus cinna-
momeus, Boisd., x 3.

This wireworm pupates in a small earthen cell with smoothly plastered inner walls,
and ten or eleven days later assumes the adult form. The beetle does not leave the
soil immediately after its transformation, but generally rests in its cell for a fortnight
or longer before emerging to feed and mate.

The beetles are inactive during the day, and may be found in large numbers hiding
behind the leaf-sheaths of cane, maize and various weeds, and in maize cobs, they
also hide under lumps of earth on the surface of the soil. Towards dark they become
more active and leave their hiding places.

The females lay their eggs in clusters or strings in small cavities formed in lumps
of moist friable soil, the number of eggs per cluster or string varying from five to
fiitty. The elongate oval eggs are pearly white in colour and when newly laid measure
‘7mm. in length and 3 mm. in breadth ; in the cool months of June and July the
mcubation period varies from 16 to 19 days.

32 ROBERT VEITCH.

The ravages of this pest are particularly severe in newly planted fields of cane.
Stools of cane of all ages are attacked, but it is only in the earlier stages that the plant
succumbs to the injury, for once it is properly established it is well able to survive
the attacks of the wireworms, being then sufficiently vigorous to send out new eyes
and roots to replace those destroyed by them. The wireworms in newly planted fields
commence the attack by eating the butts or the ends of the sets or cuttings of cane,
and then five or six days later they turn their attention to the eyes and roots, which
by that time, provided germination conditions have been favourable, have started
to swell and come away. The small roots are frequently destroyed, and in a bad
attack only 20 or 30 per cent. of the eyes escape injury. The wireworm attacking
the eye eatsa hole at the base and then devours the whole of the softer inner tissues,
thus killing it. Eyes that have germinated and sent up shoots nine or twelve inches
long are also attacked and frequently so badly damaged that they die off. The
greatest damage to the eyes and young roots occurs in the second and third weeks
after planting. In a bad attack as many as 75 per cent. of the sets or cuttings are
completely destroyed.

' Heavy additional expenditure is incurred when a newly planted field is badly
injured, for the numerous misses or failures have to be replaced by new sets or trans-
plants, and this is a costly operation, especially if the latter are used.

Preventive and remedial measures are urgently required for the control of this
beetle, but unfortunate:y it seems to be particularly healthy, being apparently free
from parasitic and predaceous enemies and also immune to the most virulent poisons ;
this latter fact was demonstrated in a poisoning trial in which the wireworms feeding
on the poisoned cane seemed if anything even healthier than those feeding on the
control sets. The following measures have, however, been found very beneficial in
minimising the losses :—(1) In fields known to suffer from wireworm attacks it is
well to make provision for failures by the continuous planting of a certain proportion
of the rows ; usually there is a space of at least fifteen inches between the sets or
cuttings in the rows, but in continuous planting no such space exists and hence the
number of sets per row is much greater than in the ordinary rows. When required
the surplus stools in the rows planted continuously should be dug up to fill the blanks
or misses in the ordinary rows. This transplanting should be done only in wet weather.
(2) Transplants only should be used to fill the gaps, as new sets are at once attacked
by the wireworms and many of them are destroyed ; the transplants being older are
better able to resist the attack, and if good rains fall shortly after transplanting,
very few of them fail, and so a full stand of cane is obtained. (3) The cane on alluvial
flats should be second ratooned, as by so doing the annual planting area is reduced,
and thus the annual loss from wireworms is minimised. (4) Drainage should be im-
proved wherever possible, and the importance of clean cultivation cannot be over-
estimated.

The Yellow Wireworm (Lacon stricticollis, Fairm.) (fig. 7).

This species 1s not nearly so destructive as S. cumnamoneus, being much less
numerous and not nearly so voracious; it is occasionally met with im numbers in
alluvial soils and it also occurs in red hill and sandy soils, but 1s never so abundant |
as the common species.

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 33

The pearly white egg is more nearly spherical than that of S. connamomeus, and
measures “6 mm. in length and ‘4 mm. in breadth; the mcubation period is 12
or 13 days. The larva or wireworm is pale clay-yellow, except for the head, the
first thoracic segment and the last abdominal segment, which are brown; the
full-grown wireworm measures 18 mm. in length and 2°5 mm. in breadth, and
the duration of the larval stage is probably three years or longer. The pupa is

Fig. 7. Lacon stricticollis,
Fairm., X 4.
typical of the ELaTeRIDAE, and the blunt posterior angles of the thorax serve to
identify the pupal stage of this species, as in the other two the posterior angles
are backwardly produced as sharp teeth. The pupal stage lasts for 13 or 14 days,
The feeding habits of this species seem to be similar to those of S. cinnamomeus,
and the control measures recommended for that species would be just as effective

for this one.
The Predaceous Wireworm (Monocrepidius pallipes, Esch.) (fig. 8).

This species has been found only in sandy soils in association with the white
stubs of Rhopaea vestita, and has never been observed attacking cane. The larva

Fig. 8. Monocrepidius pallopes,
Esch., X 4.

is very similar in appearance and size to that of Lacon stricticollis, from which it
can be distinguished by the shape of the caudal notch, which in Monocrepidius
(C572) C

34 ROBERT VEITCH.

pallipes. is smaller and carries smaller teeth on the margin. The larval stage
probably lasts for three years or longer and the pupal stage for nine or ten days.

The most interesting point in the life-history of this species is the fact that it
feeds voraciously on the white grubs of Rhopaea vestita, and it is undoubtedly one
of the most important factors holding that pest in check. It has been observed
eating grubs in the field, and in a prolonged laboratory experiment it was found
to consume on an average one grub a week; in the experiment referred to, the
wireworms were placed in tins containing rich soil, a white grub and a piece of
fresh cane, thus giving them the choice of animal food, living vegetable matter
and decaying vegetable tissue or humus. The animal food was almost invariably
preferred by this species.

The Sugar-Gane Army-Worm (Cirphis unipuncta, Haw.).

The caterpillars of this Noctuid moth frequently appear in considerable numbers,
and are responsible for an appreciable amount of damage to the younger crops.
The caterpillars can be found all the year round, but generally are most abundant
in the cooler months. This pest is a notorious enemy of maize and other cereals
in the United States, but fortunately the damage in Fiji has never reached serious
dimensions.

The most extensive outbreak of this insect in the writer’s experience in Fiji
occurred in May and June 1918, in young plant fields six to ten weeks old. The
caterpillars were in large numbers on ten plantations in the Lautoka and Rarawai
Mill districts ; no damage was observed in cane older than nine months, and with
one exception the attacks were confined to plant crops.

The army-worms shelter during the day in the folded leaves of the central shoot
of the cane plant, and after dusk they move out on to the tips of the leaves, where
they feed ; when they are present in large numbers they frequently strip the leaves
right down to the midrib, thus severely retarding the development of the young
plant. The young caterpillars feed on the leaves of tender grasses, until they are
able to attack the harder tissue of the cane leaves. There are usually five moults
at intervals of three to five days, the caterpillar being full-grown in three weeks.
The full-grown caterpillar is 36 to 39 mm. in length, and is greenish brown in colour.
When ready to pupate it enters the soil and forms an earthern cell at a depth of
two or three inches, and inside this it pupates a few days later. The pupal stage
lasts ten or eleven days.

This pest is well held in check by its enemies, and its attacks rapidly die down,
for parasites and predators take a heavy toll of it. The minah bird destroys
numbers of the caterpillars, and a hornet (Polistes macaensis, F.) kills even more ;
the imago of the latter is very fond of the caterpillars, which also form an important
item in the food given to its grubs. The Tachinid fly, Sturmia bimaculata, Htg.,
and a Braconid (Apanteles sp.) have been bred from this species of army-worm, and
the larvae and adults of various CaraBipaE have been found to feed on the cater-
pillars. Natural enemies are usually sufficient to hold this species in check, but
in the event of a serious outbreak resulting from a combination of circumstances
unfavourable to the natural enemies it could be controlled by the use of poisoned

sprays.

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. gO

A Second Species of Sugar-Cane Army-Worm (Cirphis loreyi, Dup.).

The army-worms of this species are occasionally destructive in the cane areas
although generally their attacks are confined to a few stools along the edges of
fields. During a residence of several years in Fiji the writer has seen only one
attack in which the damage was at all serious; this occurred in an eleven-month-
old crop of plant cane, and was so severe that many of the leaves were stripped
down to the midrib. This species occurs in the old word tropics, and in Africa
it is a serious pest of maize.

Minahs, hornets and the Tachinid, Sturmia bimaculata, Htg., keep this pest
in check, and the use of artificial control measures is unnecessary.

The Mauritius Bean Army-Worm (Prodenia litura, F.).

Mauritius bean is extensively used on all the plantations, it being part of the
normal rotation to long-fallow land once every four years with Mauritius bean
as a covering crop, which is eventually ploughed in as a green manure. The
Mauritius bean army-worm occurs in every fallow field and is responsible for a
very appreciable check to the growth of the crop.

The eggs of this moth are laid in clusters or nests, each of which contains from
500 to 700 eggs, the usual spots chosen for egg-laying being the leaves of the food-
plant. The caterpillars hatch out after an incubation period of three days, and
immediately commence eating small irregular patches on the under-surface of the
leaves, but as they grow older they eat through the leaf tissue, destroying epidermis,
ground tissue and small veins. The caterpillars also attack cotton, tobacco and
cacao, and the writer has recently bred them from the fruits of the “ivi” or
Tahitian chestnut. They are full grown in three weeks, and then pupate in earthern
cells two or three inches below the ground, the moth emerging after a pupal period
of eleven or twelve days.

The enemies of this insect are the minah, the hornet and the small brown ant
(Pheidole megacephala, F.). The ant destroys large numbers of the eggs. Remedial
-measures are not at present required for the control of this pest.

The Moth Borer of Cane (Zrachycentra chlorogramma, Meyr.).

The larvae of this Tineid moth are occasionally found tunnelling cane, and
causing a slight amount of damage. Their work has usually been most conspicuous
in low-lying reclaimed tiri fields, which frequently contain a considerable proportion
of sickly canes. Some of these moth borers have been found in half-rotten cane,
but others have been found in soft light-coloured Badila stalks which were standing
up well and were far from being rotten, although they were of a somewhat sickly
appearance. No stools of perfectly healthy cane have been attacked by this insect.
The full-grown larva is about 25 mm. in length, and lives in a case which it drags
about from place to place; this case is a spun web to which numerous small
particles of cane, earth and vegetable debris are attached. The case affords a
considerable amount of protection to the larva, which is able to wriggle into or
out of it with astonishing rapidity.

The damage caused by this species bears some resemblance to the work of the
beetle borer, but it differs from it in several important respects. The tunnels of the
moth are wider, shorter and more irregular than those of the beetle borer and they

(C572) a2

36 ROBERT VEITCH.

contain a much rougher, darker coloured frass. Other points of importance in the
mode of attack of this pest are firstly the fact that, unlike the beetle borer, it swallows
all the cane it chews, and secondly that it frequently leaves its tunnel and moves up
the cane-stalk or even to another stalk, where it commences a new tunnel. Otherwise
the larva destroys cane in a manner somewhat similar to the beetle borer
by eating tunnels along the stalks, destroying large quantities of tissue and lowering
the sugar content of the untunnelled portions of the affected stalks.

The full-grown larva pupates in an extremely tough case made of small pieces of
cane fibre overlapping each other like roofing tiles ; it is securely closed at both ends
and must afford very good protection to the pupa. The pupa is formed sometimes
in the old larval tunnel, and at other times behind the leaf-sheaths or under debris,
the pupal period lasting for three weeks.

The burning of the trash after harvesting and the exposure of tunnelled stalks to
the heat of the sun will largely check the increase of this insect should it ever show
signs of attacking sound cane.

The Cane Leaf-Miner (Cosmopteryz sp. N.).

The larva of this very small and beautiful moth mines in the midribs of cane, being
particularly abundant in young plant fields, where it does a slight amount of damage
by tunnelling along the midrib, interfering with the supply of sap to the leaves and
discolouring the tissues surrounding its tunnels. This species is well controlled by
parasites.

The Hornet (Polistes macaensis, F.).

The common hornet of the cane-fields is far from being a pest, but it is so mtimately
connected with the sugar plantations that no reference to the insects of the cane-
fields would be complete without an outline of its life-history.

This hornet is only a recent introduction, but it has already become one of the
commonest insects in the islands, and has probably attracted more general interest
than any other species in Fiji. It is widely distributed throughout India and the
Kast, and appears to have reached these islands some fifteen years ago, having in
all probability been introduced on a labour ship from India. It showed up first at
Suva and has now spread all over Viti Levu; it has also gained a footing in Vanua
Levu, and in a few years time it will doubtless be generally distributed throughout
the islands.

The life-history of this species may be summarised as follows :—A fertilised female
that has hibernated during August and September commences nest-building at the
beginning of October. The cells of the nest are made of the usual fibrous material
mixed with cement, and in each the female lays a small greyish white egg from which
the young larva hatches in afew days. At first the larva is fed on sugary substances,
but after it is a few days old the diet is changed to an insect one. The larva pupates
14 to 18 days after hatching, and the imago emerges after a pupal period of 15 or 16 days;
after the emergence of the first new generation progress on the nest is very rapid, and
by April it may be as large as 12 inches in diameter. Towards the end of April males
and females appear in equal numbers whereas previously the new hornets emerging
were nearly all females. This last generation leaves the nest and the hornets hide
in sheltered spots, emerging on sunny days in June and July to mate, but by the

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 37

end of the latter month mating has ceased and all the males are dead. The females
hibernate for seven or eight weeks during August and September, and then they
reappear and commence nest-building.

This species is a desirable addition to the Fijian fauna, because it acts as a useful
check on many pests. The writer has observed the hornets feeding pieces of cater-
pillars, flies and grasshoppers to their larvae, while the hornet itself has been seen
eating full-grown crickets, young grasshoppers, army-worms, jumping spiders, flies
and small beetles.

The sting of this species is very painful, and therefore many neaple. losing sight
of its value as a predator, have passed a verdict of unqualified condemnation on it.
While it is true that labourers and others, including the writer, are sometimes
badly stung, yet it is fortunate that during the height of the cane-cutting season
the hornets are hibernating, and so the labourers can work without fear of being stung.
At the beginning of the season a few nests are met with and towards the end of the
season the small new nests begin to show up, but even then the danger is not nearly
so great as it would be were the cane cut from January to April, when the nests are
large and defended by numerous individuals, During the early months of the year
there is little occasion for going into big cane at Lautoka, Rarawai and Labasa Mills,
and so the labourers do not generally encounter the hornet at its worst; but at
Nausori Mill the position is different, for there the usefulness of the hornet is distinctly
lessened by the fact that the cane is trashed during the season when the nests are at
their maximum size, and considerable expense is incurred in sending special labourers
(in canvas suits, gloves and veils) into the cane to destroy the nests before the ordinary
labourers can enter the field to trash the cane. The trashing of cane has now been
abandoned at the other three mills, being regarded as an unprofitable measure under
the climatic conditions of the dry districts.

The Leaf-Hopper (Perkinsiella vitiensis, Kirk.).

This pest can be found in any cane field, but fortunately it does little damage,
as it is kept in check by numerous natural enemies, the most effective of these being
the ege-parasites Ootetrastichus, Paranagrus and Anagrus. The Stylopid parasite,
Elenchus tenurcornis, Curt., attacks both young and adults, and the numerous spiders
in the cane-fields also take a heavy toll of this pest.

The eggs of this leaf-hopper are laid beneath the epidermis of the cane plant, usually
along the midribs of the leaves, but also in the internodes of the stalk, or even in the
leaf-sheaths ; generally three eggs are laid in each cavity, the position of which is
indicated at first by the piece of white wax covering the opening of cavity, and later
by the reddish discoloration of the tissues surrounding it.

The young leaf-hoppers hatch out and feed on the cane, sucking the sap and ex-
creting honey-dew, on which a black fungus grows. In Fiji the writer has never seen
the leaf-hopper present in numbers large enough to stunt the growth of the crop
appreciably, and control measures are unnecessary.

The Mauritius Bean Bug (Brachyplatys pacificus, Dalli.).

This little. black Pentatomid bug occurs wherever Mauritius bean is grown, and it
can also be found feeding on a number of the leguminous weeds that flourish along the .

38 ROBERT VEITCH.

drains and roads on the plantations. If the upper leaves of any bean plant be pushed
aside the bean bug can often be found in such numbers that the stem of the plant is
literally black with them. :

The bug is 5 mm. in length and belongs to the sub-family ScUTELLERINAE. The
eggs have been found on the leaves of Mauritius bean, on the leaves of the commoner
weeds of the cane-field, on pieces of timber lying in fields and even on the doors and
window-sills of bungalows. The eggs are laid in a double row, the number in each
set varying from ten to thirty. Hach egg is about one millimetre in length and is
an extremely neat white cylinder with a little lid that is opened when the young bug
emerges ; in most cases the lid falls off when the bug leaves the egg, but sometimes
it remains hanging half open as if it were hinged to the cylinder. The eggs are laid
end to end in a double row with the lid ends directed outwards. They hatch out
after an incubation period of six or seven days.

The newly emerged bugs immediately start sucking the plant sap. They moult
frequently, but the exact number of the moults and the intervals between them have
not yet been definitely ascertained. The numbers of this pest are often so great that
they considerably retard the growth of the green crop, but the damage is not sufh-
ciently serious to warrant the adoption of control measures. The multiplication of
the bugs is checked by a Chalcidid egg-parasite described by Mr. J. Waterston as
Ooencyrtus pacificus, and an important fungous enemy is an unidentified species of
Isaria which attacks the nymph and imago.

The Gane Aleurodid (Alewrodes comata, Mask.).

This Aleurodid is occasionally found in extremely large numbers on the under
surface of cane leaves, but as a rule it is not common, and even in its worst attacks
it is only a minor pest.

The pale yellow eggs are attached to the leaves and are usually found in clusters
containing sometimes as many as 150 eggs. The various stages are all typical of
this family.

The damage to the cane is twofold ; for in the first place, it is stunted by the
loss of sap sucked out of the tissues by the Aleurodid, and in the second place,
a sooty fungus grows on the excreted honey-dew and interferes with the process of
photosynthesis.

The most effective check to the increase of this pest is the maggot of a Syrphid
which feeds on all its stages. The life-history of this predator has been worked out
and is briefly as follows. The white cylindrical egg of the Syrphid is firmly cemented
to the underside of the cane leaf, and is almost invariably laid alongside an actively
ovipositing Aleurodid. The young maggot hatches out in three days, and imme-
diately starts to feed voraciously, attacking all stages of the pest, but seeming to
prefer the eggs and the adults to the larvae and pupae. The quantity of food
consumed is enormous, as is illustrated by the fact that a single maggot has been
observed to destroy as many as one hundred eggs in a day, as well as a considerable
number of adults; while another was seen destroying thirty eggs and four adults
in twenty minutes. The maggot is typical of the Syrphid family, and is full-grown
i seven or eight days ; it forms a greenish pear-shaped pupa, from which the adult
emerges six days later. This Syrphid is an unidentified species of Xanthogramma.

INSECTS IN SUGAR-CANE PLANTATIONS IN FIJI. 39

The Cane Mealy-Bug (Pseudococcus bromeliac, Bouche).

This mealy-bug is often found at the nodes of the cane-stalks behind the leaf-sheaths,
where it punctures the rind and sucks the sap, thus lowering the vitality of the plant.
It rarely occurs in large numbers and is of only slight economic importance.

The older females are quite active, but the young nymphs move about freely
until they find a suitable spot at which to pierce the rind of the cane ; when once
that is found they become very sluggish, and settle down to feed and reproduce.

Locusts (Locusta danica, L., and Cyrtacanthacris guttulosa, Walk.).

These two species sometimes become numerous along the edges of fields, stripping
the cane leaves down to the midribs, but the damage is rarely serious. The minah
bird is a very efficient check on the locusts, which form its principal food; whenever
trash or any rubbish is being burned off in a field the minahs gather in large flocks
round the edges and reap a rich harvest from the insects that are driven before the
fire and smoke ; these consist very largely of grasshoppers.

Conclusion.

The three most important pests in the cane fields are undoubtedly the beetle borer,
the wireworm and the white grub, each of which constitutes an entomological problem
of the greatest importance. Their life-histories have been as fully studied as the time
at the writer’s disposal permits, but there are still many points of the greatest im-
portance that require investigation in the immediate future. Although much
research work remains to be done, a number of satisfactory control measures have
been evolved, and their enthusiastic adoption will undoubtedly lead to a great
reduction in losses ; while there is every reason to believe that with an increasing
knowledge of the habits of these pests further and still more effective measures will
be devised,

The other pests more briefly referred to in these notes are of minor importance only,
and although they have to be recognised as injurious on the plantations, the damage
they cause is at present so slight that control measures are unnecessary.

In conclusion, the writer has to acknowledge the courtesy of the Imperial Bureau
of Entomology in identifying the species referred to in these notes.

Babloography.

(1) Muir, F. “ Notes on some Fijian Insects,” Bulletin No. 2, Hawanan Sugar
Planters’ Experiment Station, 1906.

(2) Jepson, Frank P. “ Report of Economic Entomologist,’ Annual Report of
Department of Agriculture, Fiji, pp. 53-54.

(3) Ilhngworth, J. F. “ Further Notes on the Breeding of the Tachimd Fly
parasitic on the Cane Beetle Borer,” Jl. Econ. Entom., Concord, vii, no. 5, October
1914, pp. 390-398.

(4) Ilingworth, J. F. “A New Pest of Cane in Fiji,” J]. Ecom. Entom., Concord,
vu, no. 6, December 1914, pp. 444-445.

(5) Veitch, R. “ The Sugar Cane Wireworm in Fiji,” Agricultural Report No.1
of the Colonial Sugar Refining Company, 1916.

(6) Veitch, R. “ The Hornet in Fiji,” Agricultural Report No. 2 of the Colonial
Sugar Refining Company, 1917.

hub

rare

ie, ‘ wy, ve bt i
Py) ee rey i oy ee

cieth yy) bagtt 4 ts

aa 484 ceatiag rere

f Ppa pein, § el inhigs id. alt ye

hhaXe MRD pie
onal:

Vise

A NEW LYGAEID BUG FOUND AMONG STORED RICE IN JAVA.
By W. L. Distant.

Dr. Guy A. K. Marshall, Director of the Imperial Bureau of Entomology, has
requested me to indentify four specimens of a Lygaeid sent to him by Dr. C. J. J.
Van Hall from Java, where it was found among stored rice. It is a very distinct
species and represents a new genus of which I here give figure and description.

Fam. LYGAEIDAH.
Subfam, APHANINAE.
Div. Myopocuaria.
Ampera, gen. nov.

Head about as long as anterior lobe of pronotum ; first joint of antennae a little
shorter than head; pronotum longer than broad, transversely constricted before
middle, collar narrow, anterior lobe distinctly globose, posterior lobe strongly centrally
longitudinally incised, the basal lateral angles subnodulosely prominent ; scutellum
broadly subangulate ; rostrum with the first joint slightly passing base of head—
imperfectly seen in carded specimens; anterior femora moderately incrassated ;
other characters generally as in Pamera, from which it principally differs by the
structure of the pronotum and scutellum.

Ampera intrusa, sp. nov. (fig. 1).

Head, anterior lobe of pronotum and scutellum, black or blackish, posterior frontal
lobe more castaneous ; corium more or less stramineous, the apical area more or less
infuscate ; membrane black or piceous, body beneath blackish ; legs pale stramineous ;

Fig. 1. Ampera intrusa,
Dist., sp-n.

second joint of antennae considerably longer than first; membrane not or very slightly
passing abdominal apex ; rostrum stramineous, imperfectly seen in carded specimens.
Var. Head and pronotum and body beneath, dark ochraceous,
Other structural characters as in generic diagnosis.
Long, 33-4 millim.
Java (Dr. Van Hall—from stored rice).

A USEFUL BREEDING CAGE.*
By Laurence D, CiEare, Jnr., F.E.S.,
Biological Division, Dept. of Scvence and Agric., British Gurana.

Entomological investigations in the tropics usually entail a certain amount of
difficulties not experienced in temperate regions, and it is often necessary to devise
various kinds of apparatus to overcome these. The breeding cage described here
was devised by myself as the result of a number of difficulties experienced in br2eding
experiments. Cages of this type have been in constant use in this laboratory for the
past four years and have been found to meet almost every requirement. The ease
with which they are constructed, together with the possibility of varying their size
to suit one’s needs, as well as their portability, are all highly commendable features.

Fig. 1. A portable breeding cage

The cages are composed of cylinders of brass mosquito-proof wire netting, 20
meshes to the inch, held together by brass paper-fasteners. To the top and bottom
of these cylinders covers are fitted. In this laboratory large glass petri dishes, between
8 and 10 inches diameter, are used for this purpose. Galvanised iron pans may,
however, be used instead of these dishes, and if made in nests would be a decided
advantage when travelling ; the wire being carried in rolls. It will be seen that the
diameter of these cages is limited only by the size of the pans, while the height can be

* Published by permission of the Director of Science and Agriculture, British Guiana.

44 LAURENCE D. CLEARE, JNR.

varied from 3 feet, the width of the netting, down to a few inches. If it is desirable
to use cages taller than 3 feet, and this will only be very rarely, the height may be
increased by joining on an additional strip of netting by means of paper-fasteners.

Originally, fine muslin, held in place by rubber bands, was used for the tops of these
cages ; but after some large larvae were attacked by rats the dishes came into use.
Obviously muslin can still be used in certain places with advantage, especially when
weight has to be considered.

The wire netting is usually fitted inside the dishes both top and bottom, but in
cases where earth is necessary for pupation it is advantageous to fit the cylinder
outside the bottom ; the advantage of this will be found when lifting off the wire to
replace the food-plant. Food-plants, of course, have their stems placed in small
bottles or other receptacles containing water.

Cages of this type have been used in tick breeding experiments by Mr. G. KE. Bodkin,
and when thus employed were placed in larger dishes containing water and kerosene
oil to prevent the escape of the ticks. This method could also be used to prevent
attacks of ants, which, in this Colony at least, have always to be considered in insect
breeding,

It may here be mentioned that in this laboratory all tables used for breeding ex-
periments are kept standing in bowls containing water and kerosene oil. Were the
bare wooden legs of a table placed in this mixture the oil would penetrate the wood
and ruin the legs. Professor Harrison has devised a method to avoid this ; small
iron rods about 6 inches long and 1 inch in diameter are driven perpendicularly into
the legs and it is these that come in contact with the mixture of oil and water.*

*[As an alternative to water and kerosene, dusting une saucers occasionally with
pyrethrum powder has been found very effective in keeping off ants; this has the
fdaeekare that the legs of the table require no special adaptation.—ED.]_

NEOTOXOPTERA VIOLAE, THEO., AND ITS ALLIES.
By A. C. Baker, B.S.A., Ph.D., US.

Bureau of Entomology, Washington, D.C.
(PuateE I.)

In 1915 (Bull. Ent. Res. vi, p. 1381) Theobald described an Aphid taken on violets
in Africa as Neotoxoptera vielae. He has also recorded a statement by Davis that it
probably is Rhopalosiphum violae, Pergande, 1900 (The Entomologist, xlix, p. 149).

A study of Pergande’s species as it is to be found in greenhouses in the United
States and Canada shows that the form described by Theobald is really an aberration
from Pergande’s species. This is shown clearly by the figures accompanying this note.
Figure | is the typical forewing of wolae, Perg., and from examples of this type aber.
rations have been found as shown in figures 2 to 6. It will be noticed that there are
several cases in which the media is once branched, the condition met with in Theo-
bald’s form. Figures 4 and 6 show a much greater departure from the type, and yet
these forms as well as all the others figured can be obtained from one colony and in
one “ family ” of the species. The hindwing as figured by Theobald is quite typical
for Pergande’s species (fig. 7), but here also figures 8 and 9 represent in each case an
aberrant form.

The writer is unable to state that similar conditions would be met with also in
Africa and that all of these types would occur. From the fact that they occur in
_ America, and bearing in mind that nearly all the allies of this species present similar
conditions, one would expect to find them.

In order to understand the proper generic designation for this species it is necessary
to review briefly the small group to which it belongs. In the subfamily APHIDINAE
there is small group, the PENTALONINA, the members of which are mostly tropical
or subtropical. This is based on Pentalonia nigronervosa, Coquerel, 1859 (fig. 10), an
Aphid with most remarkable venation. This venation, as far as the writer’s expe-
rience goes, is constant, although no large rearings have been made. To understand
this venation it is only necessary to look at Idiopterus nephrolepidis, Davis, 1909,
and some aberrant forms of this species, which forms, by the way, are not at all rare.
Figure 11 shows the typical fore-wing of this insect ; here the radial sector is quite
distinct. In figure 12 this is seen to be coalesced with the media in such a way as
to produce a closed cell comparable to the closed cell in Pentalonia. Figure 13 is
one of another aberration in which the media does not exhibit its typical forked
character. :

‘In 1915 Del Guercio (Redia, vii, p. 463) erected the genus Fullawayella for Macro-
syphum kirkaldyz, Fullaway, 1909. This insect is closely related to violae, Perg., but
differs in the dilation of the cornicles and other details. In describing the species
Fullaway (Ann. Rep. Haw. Agr. Exp. Stn. 1909, p. 22) stated that the venation is
normal, that is for Macrosiphum, but he figured what evidently is an aberration
similar to those found in other and related species (fig. 15). It seems evident that
violae, Perg., and kirkaldyi, Fullaway, belong to the same genus, and it is equally
evident that it is not Rhopalosiphum, the type of which is nymphaeae, Linn. Essig
(Pomona College Jl. Ent. iii, p. 541) has stated that he believes kirkaldyi and

46 A. C. BAKER.

nephrolepidis to be identical and treats the species as a native of the Hawaiian Islands
imported into the United States. While this view may be correct, the figures given
by Fullaway do not agree, as far as the cornicles are concerned, with any specimens
of nephrolepidis. They do agree, however, in being swollen somewhat like those of
Micromyzus nigrum, V. d. Goot, 1916. We have been unable to study the type of
the species, and since Essig’s view is a surmise, we consider korkaldyi to be distinct
from nephrolepidis as indicated by Fullaway’s figures..

The genus Micromyzus was erected by Van der Goot for his new form, nigrum
(fig. 16) (Cont. 4 la Fauna des Indes Neérland. 1, fasc. 3, p.52, 1916). M. nigrum is
undoubtedly closely related to kurkaldyz, if Fullaway’s figures are correct, and is
congeneric with it. Mzcromyzus must therefore fall for Fullawayella.

One other genus which evidently belongs to this group has been described (Patch,
Ent. News, xx, p. 338, 1909). This is Microparsus, with variabilis Patch (fig. 14) as
type, a form which was first figured and described by Sanborn but not named by him.
This species, hke the others in the group, shows aberrations.

The different genera and species may be separated as follows :—
1—Forewing with a constant closed cell Pentalonia.

a—vVeins heavily bordered ..

a'—-Veins not heavily bordered
1'—Forewing typically without such cell
2——Hind wing with one vein only

a—One species only . We
2'-_Hind wing typically with more Shar one vein

nogronervosa, Cql.
caladu, V. d. Goot.
2.

Microparsus.
varvabilis, Patch.

3—Cornicles cylindrical Idiopterus.
a—OQne species only .. nephrolepidis, Davis.
3'—Cornicles more or less swollen Fullawayella.
a—Cornicles broadly swollen 4g ie en? Ok
a+—Cornicles slightly swollen - Be

6—Cornicles very broadly swollen and distinctly”
reticulate near distal extremity
b'—Cornicles moderately swollen and not reticu-
late near distal extremity vrolae (Perg.)
c—Segment iv of antenna without sensoria kirkaldys (Fullaway).
eo Semment i iv of antenna with 3-5 sensoria .. nigrum (V. d. Goot).

Of the species in the key only two, violae, Perg., and nephrolepidis, Davis, have been
studied with a view toward determining the commonly present aberrations. If the
opportunity to study other species in this way should arise, it is not improbable
that other names would sink in synonomy. ‘This brings up again the question of
the advisability of retaining a name, in a lower category than a species, for an aber-
ration or of giving names to such forms if undescribed. In some groups of the
APHIDIDAE these forms are remarkably abundant and their naming would place in the
literature an almost endless series of designations. If, however, certain of these forms
were found constantly 1 in certain regions (as for example if the violae of Theobald,
as he believes, is a somewhat fixed African form) it would seem advisable for the sake
of convenience to use some designation for them.

tulipaella (Theo.)

Ce me eee ee Ge

2 i is

ema ee ae Bioysanlh

ta Se aie J a 28 oe

ae. , i
aes aa ‘
as yi ph er a8 um
ne ear Gui aveR mIONS cues, Ai
<a iis enh abe ai ae
x i euunae ; ah a) Tae
we easing ener, bik

12-13

EXPLANATION OF PLATE].

Fullawayel’a violae, Perg., forewing, typical.

9 09 - aberrations.
99 93 hindwing, typical.

» » 33 aberrations.
Pentalonia nigronervosa, Coq., wings.
Idiopterus nephrolepidis, Davis, forewing, typical.
aberrations.

Microparsus variabilis, Patch, wings.
Fullawayella kirkaldyz, Full., wings.

sk nigra, v.d. Goot, wings.

BULE ENh) RESEARCH VOL OxXe, | DART 1. IPL AGrS Me

Neotoxoptera violz, Theo., and its allies.

THE DOMESTIC BREEDING MOSQUITOS OF THE NORTHERN
TERRITORIES OF THE GOLD COAST.
By A. Ineram, M.D., C.M.,
West African Medical Staff.

The work of Dr. J. J. Simpson (Bull. Ent. Res. v, pp. 1-36 and Bull. Ent. Res. vin,
pp. 193-214) has added largely to the knowledge of the bionomics of the larger blood-
sucking insects of the Gold Coast, especially of the Northern Territories of the Colony,
yet up to the present time, with the exception of the records obtained from one or
two stations, singularly little is known with regard to the distribution in the Northern
Territories of the smaller blood-sucking insects such as mosquitos and “‘ sandflies.”’

It was therefore determined, when instructions were received from the Hon.
Principal Medical Officer to accompany him on a tour round the Northern Territories
with a view to determining the prevalence of S. fasciata, F., to take the opportunity
of recording the distribution of as many other species of mosquitos and “ sandflies ”
as possible.

The tour occupied about three months, the Northern Territories being entered on
the 3rd May and left on the 2nd August 1918. It must be confessed that the results
obtained were disappointing ; this is to be attributed (a) to the exceptional dryness
of the season, (b) to the short halts—less than 24 hours—made at the majority of
places, and (c) to the difficulty of hatching adult mosquitos from larvae and pupae
*‘en route,’ few larvae and fewer pupae surviving the continous jolting to which the
water in the breeding jars was submitted when borne on the head of a carrier. So
far as the commoner mosquitos were concerned the death in the pots of their larvae
or pupae was unimportant, as identities could generally be established from an
examination of the dead larvae or pupae ; but unfortunately on one or two occasions
larvae or pupae which were not known were collected and as these succumbed without
hatching the identity of their imagines 1s yet to seek.

A routine plan of working was adopted and adhered to as closely as possible
throughout the tour. A visit was made to the native compounds each evening
between 5 and 6 o'clock to search for collections of water containing larvae or pupae
of mosquitos ; when these were found, samples of the contents were transferred to
glass jars and the top of the jars covered with fine-meshed gauze. These jars were
carried along from day to day till adult mosquitos had hatched or the larvae and
pupae had died. In the case of small villages every compound was visited, but in
the case of the larger villages or where the individual compounds were several
hundred yards apart, as in the Frafra, Grunshi and Lobi countries, only half a dozen
of the compounds nearest to the Rest House were examined.

Ttees growing in close proximity to the Rest Houses were scrutinised carefully
for “rot” holes, which are known to form a favourite habitat of larvae and pupae
of certain mosquitos, e.g., Stegomyia. The walls of the Rest Houses were examined
each day for Phlebotomus and every evening when sitting out in the open air a look
out was kept for Culicoides.

48 DR. A. INGRAM.

The accompanying map shows the places where S. fasciata, F., was found, and
it will be seen that its distribution is practically universal throughout the districts
traversed. In the diary attached its presence is not recorded at several of the
halting places, but probably this is due to insufficient observation; with the
exception of Yeji it is recorded from all the principal stations. The returns from the

=
Tish ee AslB3wkal | |

a PB ag a ; eC) ce Sieh i
Jefisi © fe} faite 7iana re! fe Tit a
© © _ © |Nacon Sambolugu Nangudi (6 7 ?
Nandaw, B82atiasan Zz . origoire,
Na q 404¥aragu ?
Dawen! | [e] Gambaga

| @
NORTHERN

y ONadawle | :
kalleo Yagaba
fe) Wa a
© Janina
Patenge ® ‘
© Aulmasa ae
Tuna. || |
| [°] Tamale
Tien aoa
@Nankuma' sae e Yamalaga y
* [e)} Bole dd © 3°
\ sigan : | Turu
\ ee anpa A | © y;
\ Wer, ae | . Dogenkede <
\ ye | | Sa/aga &
| | A Ban ae ieee | 1 Goel v4 sae ce
! S \ | ©Mak
a Jugbe a= | f ongo %
| LF Gere | | Waerte rt ee?

8 = { fy

Z - A na
# © Sitkassiko NORAIHERN
ye ae
/
A WEN Sa ee aed "

i °Wkoratiza
° 2

3 Kaperleuon & |Z
if Banda : | \ mie ee
° 2 eAintampo \ rang

io}

© Indicates places where Stegomyia fasciata has been found.
[el Indicates present or former European stations where S. fasciata has heen found.

majority of places stopped at merely show the presence of such mosquitos as
breed in the collections of water in the immediate neighbourhood of;human dwellings
and, as has been pointed out elsewhere (Bull. Ent. Res. vii, pp. 171-172), returns
of mosquitos bred from domestic utensils convey a false impression as to the actual

DOMESTIC BREEDING MOSQUITOS OF THE GOLD COAST. 49

indigenous ‘‘ domestic ” mosquitos of any place. Where a halt of three or four
days was made and it was possible to go further afield in search of larvae, see
were invariably found.

Larvae of six species of mosquitos were commonly found in the various collections
of water encountered in the native compounds and houses, namely :—S, fasciata, F.,
S. vittata, Big. (sugens, Theo.), Culiciomyia nebulosa, Theo., Culex duttoni, Theo.,
C. decens, Theo., and C. tegripes var. fuscus, Theo. Larvae of A. costalis, Loew,
and Culex aoe. Theo., were also taken in domestic utensils, but they were far
from common.

As regards the habits of these mosquitos, it was noticed that larvae of C. duiton:
were usually found in water in pots that had been used for the brewing of “ peto ”
(native beer) or had contained that hquid. “ Peto ” has a characteristic sour odour
which is very persistent and clings for long to any vessel that has contamed it.
Larvae of C. duttoni were never found at any distance from human habitations,
except at Tumu, where they were discovered in a small pool near the swamp at a
distance of about 500 yards from the nearest hut; this particular pool it was
afterwards ascertained was made use of by the natives for the purpose of cleansing
their “‘ peto”’ pots. Larvae of C. decens and of Culaciomyra nebulosa were generally
discovered in utensils holding foul water, often associated with larvae of €. duttoni
and larvae of C. tigripes and occasionally with larvae of S. fasciaia. “‘ Peto” pots
apparently offered no attraction to the adult females of these two species. Larvae
of C. tigripes were found indifferently in vessels containing fresh or foul water ;
possibly the adult female lays her eggs only in water which she knows instinctiv ely
to contain eggs or larvae of other species of mosquitos.

Larvae of S. fasciata and S. vittata were frequently found associated, usually
in clean water. The commonest receptacles contaiming these larvae were small
earthenware pots or portions of calabashes used for holding the drmking water
for the fowls belonging to the compound. It is to be noted that S. vittata 1s not
solely a ‘“‘ domestic’ breeder; larvae of this species may be found in enormous
numbers in shallow pools formed by the weathering of outcrops of gneiss (*) and
laterite far from any human habitation. Apparently, as has been recorded by
Bacot, this species passes through its cycle of development at a very rapid rate,
and perhaps it has been compelled to do so by the rapid evaporation of the water
from these pools under the influence of the tropical sun. Larvae of Culiciomyia
nebulosa were also occasionally found in pools in the “ bush” at great distances
from the villages.

Larvae of C. invidiosus were quite common in swampy pools in the “ bush,” while
they were only rarely found in collections of water in the native compounds. On
the other hand, larvae and pupae of C. decens, which so closely resemble those of
C. mvidiosus, were not met with in swamp pools, but abounded in domestic utensils.

Larvae of C. ager var. ethiopicus, Kdw., were taken only in swamp pools in
which filmy algae (Spirogyra?) were growing; the tint of these larvae is almost
the same as that of the algae and is doubtless an instance of protective colouring
(cf. larvae of Culex annulioris, Theo., and C. consimilis, Newst.). Larvae of
Anopheles pretorrensis, Theo., were found at Winduri in rock pools in association
with larvae of Stegomyia vittata.

(C572) D

50 DR. A. INGRAM.

In many of the trees in the neighbourhood of European quarters and Rest
Houses “‘rot’’ holes were found; with the exception of three, which contained
water and mosquito larvae, these holes were either absolutely dry or showed very
slight moisture, but as many of them proved capable of holding water, this dryness
was probably due to the dearth of rain. The debris collected in 25 of these holes
was scraped out and, after thorough drying in the shade where there was any
indication of moisture, was preserved with a view to testing it for the presence of
‘resistant’ eggs. Twelve of these samples were experimented with whilst the
tour was in progress, but in no instance did any mosquito larvae hatch. As it
was thought that the conditions obtaining “ en route ” were not ideal for the hatching
of larvae, the remaining samples were brought back to Accra, where after being
immersed in water they were kept under observation for a period of one month
without appearance of larvae. It seems open to doubt whether Stegomyia fasciata
and other species of this genus tide over the dry season in the Northern Territories
by means of “ resistant” eggs stranded in ‘“‘ rot” holes of trees.

Portions of the leaves and twigs of those trees near the dwellings which seemed
to show ‘rot’ holes most commonly were collected for identification purposes
and were forwarded to the Director of Agriculture at Aburi. Pownciana regia, the
flamboyant (often planted for ornament near Kuropean houses), wild fig trees,
mangoes, silk-cotton trees, “ dawa-dawa”’ trees (Parkia biglobosa) and tamarinds
were frequent ofienders in this respect.

With regard to the above-mentioned specimens of trees the Senior Curator of the
garden at Tamale, to whom they were submitted for identification by the Director
of Agriculture, writes under date 23.xu1.1918 as follows :—“ Specimens received ;
I will do all I can to obtain full botanic names, but it will take some time. Most
of the trees grow in districts remote from Tamale.... As regards the point raised
by Dr. Ingram about these trees developing “rot” holes, (this) might be applied
to any tree growing in the Northern Territories large enough to afford firewood.
Live branches are hacked off to be dried for firewood, causing jagged wounds which
soon develop into ‘‘ rot’ holes; rather the treatment than the tree.” The trees,
however, from which the specimens in question were obtained were those growing
close to the quarters of the European officials in the stations or those in the com-
pounds of the European Rest Houses; such trees are not usually made use of for
firewood by the natives, but are carefully preserved for shade or for ornamental
purposes.

The compounds of 44 villages and towns in the Northern Territories were visited
and samples of mosquito larvae were obtained from 42. In 34 of the villages and
towns larvae of S. fasciata were found (77 per cent.), in 35 larvae of Culex duttons
(79 per cent.), in 23 larvae of C. decens (52 per cent.), in 21 larvae of C. tegrupes
(47 per cent.), in 21 larvae of Culiciomyia nebulosa (47 per cent.), in 20 larvae of
S. vittata (45 per cent.), in 4 larvae of Anopheles costalis (9 per cent.) and in 3 larvae
of C. envidiosus (6°8 per cent.). It may be of interest to compare these figures with
the records of domestic breeding larvae obtained at Accra during the past two
years. Of 1,121 samples of larvae collected by the sanitary inspectors and forwarded
to the Laboratory for examination, S. fasciata occurred in 876 (78 per cent.), Culex
fatigans in 221 (19°7 per cent.), A. costalis in 43 (3 per cent.), C. decens in 20 (1°7

DOMESTIC BREEDING MOSQUITOS OF THE GOLD COAST. 5]

per cent.), C. invideosus in 18 (1°6 per cent.), C. duttons in 5 (0°4 per cent.), C. tagripes
var. fuscus in 3 (0°2 per cent.) and Culicoomyia nebulosa in 1 (0°08 per cent.). It
will be noticed that with the exception of two, namely C. fatigans and S. vittata,
the species found in Accra and in the towns and villages of the Northern Territories
are identical. C. fatigans has yet to be recorded from the Northern Territories
and S. vittata from Accra. An erroneous impression is apt to be given by the
above figures, as although S. fasciata was found in 77 per cent. of the villages and
towns visited, it is to be remembered that it was specially looked for and in many
instances was not found without prolonged and careful search. It is a much rarer
mosquito in point of numbers in the Northern Territories than in Accra and in
the coast towns generally. The possible effect upon disease and its transference
of the higher proportional representation of the other domestic Culicines in the
Northern Territories, as compared with Accra, is an interesting subject for
speculation.

Filariasis and diseases usually associated with Filarva bancroft: are stated by
Medical Officers who have been stationed on the northern boundary of the Northern
Territories to be common ; if this be the case the réle played by Culex fatigans must be
taken by some other species. A. costalis, Mansomordes uniformis and S. fasciata are
recognised potential carriers, and Culex duttona ‘““ was found to be one of the hosts of
Filarva nocturna by Dr. Dutton” (Theobald, Appendix to Report of the Malaria
Expedition tothe Gambia 1902, p. vi, Liverpool School of Tropical Medicine,
Memoir X). |

Phlebotomus was taken on the walls of most of the Rest Houses; and as these walls
are usually composed of swish mixed with a certain proportion of dung, they offer,
when they become seamed with fissures, ideal breeding places for these insects.
Possibly the washing of the mud walls of the Rest Houses with extract of ‘“ dawa-
dawa ”’ trees (Parkia biglobosa) has some efficacy in diminishing the breeding places of
Phlebotomus, as it certainly seems to prevent the walls from cracking ; washing with
wood ash is not nearly so effectual. As Phlebotomus is not partial to light or breezes,
this is the probable explanation of its rarity upon the walls of the Rest Houses in the
Lorha District, which are the largest, best lighted and best ventilated of any in the
Northern Territories.

Pappataci fever has been reported from Nigeria (Yellow Fever Commission Fourth
and Final Report, pp. 13-17); it is almost certain that it also occurs in the
Gold Coast.

Culveordes were taken in the evenings at several places and were felt and seen at
others where they were not captured. Fortunately these little pests are less preva-
lent in the open country of the north than in the forest region of Ashanti. Culicordes
grahami, Aust., which is the commonest species in Ashanti, is rare in the Northern
Territories ; it appears to be replaced by other larger species with less conspicuously
spotted wings (Forcipomyia). Larvae and pupae of Forcipomyia ingrani, Cart.,
were found in the moist debris scraped from the “ rot” holes in a flamboyant tree
at Gambaga.
~ With reference to the greater prevalence of S. fasciata throughout the Northern
Territories during and shortly after the close of the wet season, attention may be
directed to the fact that all recent cases of yellow fever amongst Huropeans resident

(C572) D2

52 DR. A. INGRAM.

in the Northern Territories have occurred between July and the end of November,
most of them indeed between September and the end of November.

It remains for me to express my indebtedness to the Hon. Principal Medical Officer
for help rendered and interest shown in the work, and to thank Dr. Marshall, of the
Imperial Bureau of Entomology, for kindly identifying certain of the mosquitos
mentioned in the lists of the diary.

Diary of Entomological Findings during the Tour.
GOLD Coast CoLony.

15.iv.18. Tafo. Culicoudes tgraham, taken on arms in evening; Phlebotomus
(P. *minutus var. africanus, Newst.) seen on walls of Rest House.

16.iv.18. Anyinam. Heavy rain from 6-8.30 p.m. No mosquitos or sandflies
in evidence.

17.iv.18. Jyagate. Culicoides *grahami seen on arms.

18-19.iv.18. Mpraesu. No mosquitos or sandflies captured in District Onan
sioner’s quarters. Hippocentrum trimaculatum very plentiful on the
top of the Bauxite Hill.

20.iv.18. Kwahu Prasu. Culicoides ?grahami exceeedingly troublesome; several
taken. Musca domestica unpleasantly numerous. “ Bush” too close
to Rest House.

ASHANTI.
21.iv.18. vJoasu. Culicoides? grahami taken.
22-30.iv.18. Goomassie. Culicoides grahami and Forcipomyia sp. taken. Culiciomyia
nebulosa taken in the Medical Officer’s quarters.
l.v.18. vJato’s Zongo. A few mosquitos biting, but none captured.
2.v.18. Attabubu. Many mosquitos biting, none captured.

NORTHERN TERRITORIES.

3.v.18. Prang. A few mosquitos heard, none taken ; a Chironomid (? Cubes
seen on walls of latrine. |

4.v.18. Kaperleum. No roof to latrine; latrine bucket contained ram-water in
which were larvae of Culictomyra nebulosa, also several egg-rafts of
Culex on surface of water. Village about a mile from Rest House.

5-6.v.18. Yeji. Mosquitos seen emerging from an old water-hole (well) near
boys’ quarters; hole too deep to procure larvae. Stegomyta vittata
larvae found in water contained in a disused canoe on the river bank.
A fig tree close to the District Commissioner’s quarters showed several
“rot” holes. 3

7.v.18. Macongo. Anopheles funestus and A. costalis taken on walls of latrine ;
Phlebotomus sp. also taken on walls.

8-9.v.18. Meriki’s or Maliki’s Zongo. Native compounds visited. ee of
S. fasciata found in a disused kettle. Larvae of Culex duttoni in a
“fufu’’ mortar, larvae of C. tigripes var. fuscus and C. invdiosus
found in a water-hole.

DOMESTIC BREEDING MOSQUITOS OF THE GOLD COAST. 53

10-13. v.18. Salaga. Phlebotomus taken on walls of M. O.’s quarters. Larvae of
S. fasciata found in calabashes. Larvae of C. duttona and C. decens in
a shallow water-hole in town. ‘‘ Rot” holes seen in an akee tree (Blighia
sapida), in a fig treeand in a silk-cotton tree (Hriodendron anfractuosum).

14.v.18. Dogankade. Phiebotomus sp. and Forcvpomyia sp. taken in Rest House.
Larvae of C. duttoni, C. tigripes var. fuscus and S. vittata breeding in
a water-pot in the Rest House compound. Native compounds not
visited.

15.v.18. Turu. Philebotomus sp. on walls of Rest House; Culicoides sp. felt and
seen. Larvae of S. fasciata and C. invidiosus found in a hole in a felled
tree close to Rest House. A single specimen of Tabanus subangustus,
Ric., taken.

16.v.18. Palbe. A few Phlebotomus sp. taken on walls of Rest House. No collec-
tions of water found in native compounds.

17-18.v.18. Yamalaga. Phiebotomus taken on walls of Rest House. No collec-
tions of water found in native compounds, ‘“‘ Rot” holesseen in a
“‘dawa-dawa,”’ tree (Parkia) close to the Rest House. The inhabitants
of Palbe and Yamalaga at the present season have to go long distances
for their water (2 miles ?), consequently water is at a premium and none
is left for mosquito larvae to breed in.

19-24,v.18. Tamale. Native town not visited, as the record of mosquitos occurring
at this station due to the work of Dr. C. H. 8. Watson and Dr. J. J.
Simpson proves the presence of at least 15 species, amongst them
S. fasciata. A “rot” hole in a flamboyant tree near the cemetery
contained water in which larvae of S. ?unilineata, Theo., were breeding.
No Phlebotomus or Culicoides seen or captured.

25.v.18. Savelugu. Phlebotomus on walls of Rest House. Larvae of mosquitos in
pots and calabashes in nearly every compound inspected, the species
being C. dutioni, C. tugripes var. fuscus, C. decens, S. fasciaia and
S. vittata. Larvae of A. costalis found in water-holes, which are
numerous within 400 yards of the Rest House.

26.v.18. Diari. A single Phlebotomus taken on Rest House walls. Utensils in
native compounds harbour larvae of C.duttons and C. tugripes var. fuscus.
Only three compounds visited, as an approaching tornado precluded
further search.

27-29.v.18. Nasia. Larvae of C. dutioni, C. decens and of S. fasciata collected from
utensils in native compounds. A single specimen of Anopheles funestus
taken on the Rest House wall.

30.v.18. Kpalgu. Phlebotomus on walls of Rest House. Larvae of C. dutiona,
C. decens, Culicoomyia nebulosa and S. fasciata found in native com-
pounds,

3i.v.18. Maitbindiga. Phlebotomus on walls of Rest House. Larvae of C. duttoni
found in earthenware pots. No larvae of Stegomyia found.

DR. A. INGRAM.

1-5.vi.18. Gambaga. Philebotomus on walls of latrine and on Rest House verandah ;

6.v1.18.

7.v1.18.

8.vi.18.

engorged Phlebotomus also taken within mosquito net. C. duttoni,
Culiciomyia nebulosa and S. fasciata bred out from larvae collected from
the native compounds. Larvae of C. ager var. ethiopicus, C. invidiosus,
A. costalis and A. funestus, found in ofif-sets of the stream running
through the European garden. S. vittata found breeding in rock pools.
“ Rot ’’ holes numerous in flamboyant and other trees near Rest House.

Zongoire. A solitary Phlebotomus seen on the walls of the Rest House,
but it eluded capture. Larvae of C. duttoni, C. tigripes var. fuscus and
Culiciomyra nebulosa, collected from native compounds. No Stegomyia
larvae found.

Kugri. No Phlebotomus seen. Larvae of 8. fasciata, C. duttoni, C. tagripes
and C. invidiosus, collected from utensils in native quarters.

Binduri. No Phlebotomus seen ; dry atmosphere, strong breeze prevalent.
Larvae of S. fasciata, C. duttoni, C. decens and Culiciomyia nebulosa,
collected from native compounds. “ Rot” holes in a tamarind and
in an unknown tree close to Rest House.

9-12.vi.18. Bawku. Phlebotomus on walls of M.O.’s quarters ; engorged Phlebotomus

13.vi.18.

14.vi.18.

15.vi.18.

caught inside mosquito net. Town clean, very few pots or calabashes
im native compounds in a position to hold water, most of them “ up-
ended.” Larvae of S. fasciata found after diligent search ; larvae of
C. duttoni and C. decens rather more common. Larvae of A. funestus,
A. costalis, A. mauritianus, A. rufipes, C. ager var. ethpicus,
C. quasigelidus, C. invidiosus and Mimomyia mimomyiaformis, Newst.,
collected from pools near garden and down course of main stream.

Tishi, Phlebotomus on walls of Rest House. Larvae of C. duttoni,
C. tagripes, C. decens, Culiciomyia nebulosa, S. fasciata and S. vittata,
collected from native compounds and surroundings.

Tili. Phlebotomus taken on walls of Rest House. Larvae of S. fasciata
and C. duttont found in native compounds; village small, consisting
of about five scattered compounds.

Nangudi. Phlebotomus plentiful on Rest House walls. Larvae found in
native compounds are those of S. fasciata, S. vittata and C. duttone.

16-17.vi. 18. Zouragu. Phlebotomus on walls of District Commissioner’s quarters.

18.vi.18.

In collections of water from the native compounds are larvae of
S. fasciata, S. vittata and C. decens. ‘Larvae of C. invidiosus and
A. costalis collected from pots in European garden. In pools down
course of stream (at present not flowing) larvae of C. wmvidiosus,
C. tritaeniorhynchus and A. costalis found. ‘“ Rot” holes in several
trees around District Commissioner’s quarters.

Winduri (Tong Hills). No Phlebotomus seen. Native compounds not
visited. Larvae of S. vittata very numerous in rock pool and three
larvae of A. pretorvensis collected from these pools. A large water-
hole near Rest House had a few clumps of Pistia stratiotes growing
in it, but the only larva collected from it was that of Mimomyia
splendens, Theo.

DOMESTIC BREEDING MOSQUITOS OF THE GOLD COAST. 55

20.vi.18. Sambolugu. No PAlebotomus found on walls of Rest House, walls treated
with extract of ‘“dawa-dawa.” Larvae or pupae of S. fasciata,
S. vittata, C. duttoni, C. decens and C. tigripes in water contained in
earthenware pots in native compounds.

21.v1.18. Mayoro. Walls of Rest House washed with “ dawa-dawa” ; no Phlebo-
tomus found. Earthenware pots in native compounds harboured
larvae or pupae of S. fasciata, S. vittata, C. decens and C. tigripes.

22-25.vi.18. Navarro. Phlebotomus on walls of M.O.’s quarters. Collections of
water in the native compounds furnished larvae of C. decens, C. duttoni,
C. tigripes var. fuscus, Culiciomyra nebulosa and A. costalis. Larvae
of S. fasciata and S. metallica were found in a watering-pot beneath
a large fig tree close to District Commissioner’s quarters; this fig
tree showed numerous rot holes in roots, which had been filled by
rain on 22.vi.18 ; no larvae appeared in these holes during the ensuing
three days. No “resistant” eggs present? SS. fascoaia taken on
arm in M.O.’s quarters at 4 p.m., 25.vi.18. Larvae of the following
mosquitos collected in pools along course of stream (not flowing at
present) C. ager var. ethiopicus, C. quasigelidus, C. onvidiosus, C. unwit-
tatus, Ochlerotatus caliginosus, Graham, A. costalis, A. funestus and
A. mauritianus.

26.vi1.18. Tiana. No Phlebotomus found on walls of Rest House, which were treated
with “ dawa-dawa.” Earthenware pots in native compounds con-
tained larvae of C. decens, C. tugrupes (killed to prevent their devouring
the other larvae), Culiciomyia nebulosa and S. fasciata.

2.7-28.vi.18. Nacon. No Phlebotomus taken; inside walls of Rest House show
ornamental designs, but are not treated with “ dawa dawa”; double
doors and large windows (27” x 27”), room well ventilated and well
lighted. Larvae of S. v2ttata numerous in rock pools, borrow-pits
and earthenware pots. S. fasciata larvae in small collections of clean
water in native compounds. Culrcordes felt and seen but not captured.

29.vi.18. Batiasan. No Phlebotomus found; walls of Rest House not treated in
any way, but a breeze blowing. Nearest native compound at least
1 mile distant from Rest House. Larvae of C. duttom, C. decens,
C. tagripes (killed to prevent devouring the other larvae) and S. fasczata,
in pots in native compounds. S. vittata breeding in small collections
of water provided for hens.

30.v1.18. Pinna. Culccordes seen on arm about 4 miles on the Batiasan side of
Pinna. No Phlebotomus taken on walls of Rest House; walls un-
treated, but showed no cracks. Usual series of larvae collected in
native compounds, namely of C. duttoni, C. decens, S. fascrata and
S. vittata. 7

1-3.vu.18. Tumu. Culicoides felt and seen, but none captured. No Phlebotomus
found. Larvae or pupae of C. duttoni, C. tigripes var. fuscus, Cul-
coomyra nebulosa, S. fasciata and S. vittata, breeding in native compounds.
Larvae of C. duttont found breeding in a pool in swamp where utensils

4 vu.18.

5.vi. 18.

6.vu.18.

7.vi. 18.

8.vii. 18.

DR. A. INGRAM,

which had contained “ peto” are washed. Larvae of A. costalis
collected from foot-prints of cattle at the edge of the swamp. In the
swamp larvae of A. funestus, A. mauritianus and C. muvuidiosus
discovered ; also an unknown pupa with deeply infuscated paddles
(this pupa died). Swamp pools contained very few larvae, having
recently been filled with storm water.

Lilixia. Phlebotomus on walls of Rest House; walls washed with wood
ash. Larvae of S. vittata found in a portion of a calabash in the Rest
House compound. In the native compound several egg-rafts of
Culicine mosquitos (probably Culex duttona and Culiciomyia nebulosa)
were seen, but the only larvae collected were those of C. tagripes var.
fuscus. “Rot” holes were visible in the roots and trunk of two
“dawa-dawa ” trees in the Rest House compound.

Jefisi. No Phlebotomus taken; walls of Rest House treated with wood

ash. Earthenware pots in the compounds of the native huts con-
tained larvae of Culiciomyia nebulosa, C. duttoni, C. decens, S. fascrata
and S. vittata. Rest House well lighted on western side with good
air space between roof and wall.

Nandaw. No Phlebotomus taken; walls of Rest House untreated, but
room airy, with ample window accomodation and good air space between
roof and wall. Larvae of C. duttom, C. decens, Culicuomyra nebulosa,
S. fasciata and S. vittata in native compounds.

Ulu. No Phlebotomus found. Culicoides taken on arm. Collections of
water in native compounds contained larvae of C. duttoni, C. decens,
Culiciomyra nebulosa and S. fasciata. Larvae of S. vittata and Ochlero-
tatus nigervensis, Theo., in small pools near native huts.

Daweni. No Phlebotomus taken. Larvae and pupae of C. dutions and
C’. tugripes var. fuscus and larvae of C. decens found in pots in native
compounds. Larvae of Culiciomyra nebulosa found in a disused
“fufu ” mortar some distance away from village.

9-12.vi.18. Lorha. No Phlebotomus taken. Calabashes and pots in native

quarters contained larvae of C. decens, Culsciomyia nebulosa and
A. costalis. Larvae of A. funestus, A. mauritianus, C. invidiosus,
C. ager var. ethiopicus and C. quasigelidus, and a pupa of Mimomyra
mimomytaformis found in pools below the European garden. ‘‘ Croco-
dile ” pool and an adjacent pool covered with Pistia stratiotes ; larvae
of Mansomoides collected from the roots, but unfortunately died,
so that species could not be determined. It remains to be discovered
if the larva of M. uniformis, Theo., differs in detail from that of
M. africanus, Theo. M. unsformis, but not M. africanus, has been
recorded from this station and from Wa. Dr. Corson while resident
at this station made a collection of mosquitos (determined by Mr. Carter
of Liverpool) containing 13 different species, including S. fasciata.

13.vu.18. Burifu. No Phlebotomus taken. Native compounds not visited. Larvae

of A. costalis found in swamp.

14.vi.18.

15.vii. 18.

16.vu. 18.

DOMESTIC BREEDING MOSQUITOS OF THE GOLD COAST. atl

Yawa or Yaga. No Philebotomus taken. Larvae of Culicoomyia nebulosa
discovered in an earthen pot in Rest House yard. Larvae of Culi-
ciomyta nebulosa, C. duttona, C. decens, C. tugripes var. fuscus and
S. fasciata, collected from native compounds. Almost all the larvae
were devoured by the larvae of C. tagripes in the course of the next
two days.

Nandawli. Phlebotomus taken on walls of Rest House, which is not of
the “ Lorha” type; individual houses close together, yard divided
by retaining walls, ventilation impeded. The usual species of mosquito
larvae found in native quarters, namely C. dutioni, C. decens, C. tigripes
var. fuscus, S. fasciata and S. vittata.

Kaileo. No Phlebotomus taken; walls of Rest House untreated, but
plenty of window accomodation and two doors giving a thorough
draught. Larvae of S. fasciata, C. duttons, C. decens and C. tagripes
in pots in native quarters.

17-20.vii.18. Wa. Phlebotomus found on walls of Rest House. Larvae or pupae

21.vii. 18.

22.vu. 18.

23.vu.18.

24. vi. 18.

25.vu.18.

of C. duttoni, C. tigripes var. fuscus, C. decens and S. fasciata secured
from earthen pots in native compounds. Larvae or pupae of C. ager
var. ethiopicus, C. mvidiosus, A. costalis, A. funestus, A. mauritianus
and Mimomyia mimomyraformis, in pools near garden. One or two
larvae of Mansonioides (these pupated but died as pupae) taken from
roots of Pistia stratiotes growing on surface of “Crocodile” pool.
Dr. Corson formed a collection of 11 species of mosquitos while resident
in Wa in 1914.

Tanina. No Phlebotomus found in Rest House. C. duttoni, C. decens,
S. fascoata and S. vitiata, found in larval form in native compounds.
Culicoides abundant.

Kulmasa. Phlebotomus (scarce) found on walls of Rest House. Larvae
of C. duttom, S. fasciata and S. vittata taken from collections of water
in the native compounds. Culicoides taken.

Tuna. No Phlebotomus found; walls of Rest House showed no cracks.
Culicoides taken. Larvae of C. duttoni and Culiciomyia nebulosa
discovered in earthen pots in native quarters. No larvae of S. fasciata
seen,

Mankoma. No Phlebotomus taken. Larvae of C. dutton, C. tigripes
var. fuscus, Culiccomyra nebulosa and S. fasciata, found in water vessels
im native compounds, Culicoides taken.

Bole. Phlebotomus found on walls of District Commissioner’s quarters.
Native compounds not visited, nor any active search made for larvae
of mosquitos, as a collection of the mosquitos of this station, numbering
thirty species, was made in 1911-1912 by myself. 8S. fasciata was
included in that collection. An unknown larva with basal hair tufts
on its siphon tube was taken in one of the large borrow-pits, but
unfortunately died ‘‘en route.” Culicoides taken (probably Forci-
pomyra mcomptifeminibus, Aust., which was taken here in 1912).

58 DR. A. INGRAM.

28.vu.18. Sakpa. No Phlebotomus taken. Glossina palpalis, R.D., and Glossina
morsitans, Westw. (submorsitans, Newst.) taken in Rest House.
Larvae of C. tugripes var. fuscus and Culiciomyia nebulosa found in
native compounds. No S. fasciata larvae seen. Culicoides taken.
G. morsitans taken on road between Suripe and Sakpa. G. palpalis
taken on arm in Suripe swamp.

29.vi.18. Malowe. Numerous G. morsitans on road from Sakpa. Phlebotomus
on walls of Rest House. S. fasciata, C. tagripes var. fuscus and
Culiciomyia nebulosa, found as larvae in native compounds. Culicoides
taken.

30.vu.18. Banda Nkwanta. G. morsitans on road between Malowe and Wasipe
fairly numerous, scarce on road between Wasipe and Banda Nkwanta.
Phlebotomus found on walls of Rest House. Larvae of C. duttoni,
C. tigripes var. fuscus, S. fasciata and Culaciomyia nebulosa, in earthen
pots in native quarters. Culicoides numerous.

l.vii.l8, Jugbe. Phlebotomus taken on Rest House walls. Collections of water
in native compounds contained larvae of C. duitoni, C. tagrvpes var.
fuscus and Culiciomyia nebulosa. No Stegomyia larvae seen. Culicoides
taken.

ASHANTI.

2.vii.18. Buere. Mosquito larvae not searched for. Culzcordes taken.

3-5.vili.18. Kintampo. Several flamboyant trees on side of road from Provincial
Commissioner’s bungalow to Court House showed “rot” holes, m
one of which larvae of Culicoomyia nebulosa were found. Phlebotomus
previously recorded here, also S. fasciata.

6.vi.18. Kukuma. Phlebotomus on walls of the Rest House. A single specimen
of Mansonioides africanus seen entangled in a spider’s web in Rest
House. Culicoides taken.

7.vin.18. Nkoranza. Philebotomus (previously recorded from this place) on walls
of Rest House.

8.vi.18. Sekodumase. Phlebotomus (previously recorded from this place) on
walls of Rest House. Culicoides ? grahami taken.

9,vii.18. Ghechewere. Phlebotomus on walls of Rest House. Culicoides grahama
numerous.

10.viii.18. Kwaman. Phiebotomus on walls of Rest House. Culicoides grahanu
exceedingly numerous.

THE EARLY STAGES OF WEST AFRICAN MOSQUITOS—IV.
By A. Ineram, M.D., C.M., and J. W. 8. Macriz, M.A., D.Sc.,
West African Medical Service. |

a
oo

In three previous papers (Bull. Ent. Res. vii, pp. 1-18; vii, pp. 73-91; and
vill, pp. 135-154) we have described early stages of a number of West African
mosquitos collected at Accra and at other places in the Gold Coast. The present
paper deals with a few more early stages of mosquitos, for the most part collected
by one of us (A.I.) in the Northern Territories of the Gold Coast, and thus fills in
a gap or two in the knowledge of the life-histories of these insects.

The majority of the descriptions which follow are of pupae. As in our previous
paper we have taken for diagnostic purposes more particularly the characters
of the paddles, and the hairs or setae at the posterior angles and on the
dorsal aspect of the abdominal segments. On the dorsal aspect and at
the posterior margins of the segments of the abdomen there are important
hairs which form two distinct rows on each side of the body from usually the 3rd
to the 7th segments. The one row is situated midway between the posterior angles
and the middle line of the body, the other about midway between the posterior
angles and the first row. These two rows of hairs do not appear to have received
names, we therefore propose to call them respectively the “inner and outer lateral
rows.” :

We have found it advisable to introduce into the descriptions of pupae certain
ratios, such as the ratio of length to greatest breadth of the paddles or respiratory
trumpets. The respiratory trumpet 1s composed of two parts, a closed or tubular
part which is proximal, and an open part which is distal; the former we propose
to call the meatus, and the latter the pinna. The relative length of the meatus
varies greatly in different species of mosquito, and this fact we think may be most
accurately expressed as the ratio of the length of the meatus to the total length
of the respiratory trumpet.

Anopheles pretoriensis, Theo.

Larva.-_The larva of this mosquito is included by Edwards (Bull. Ent. Res.
in, p. 374) in his key to the larvae of the genus Anopheles.

Pupa (fig. 1).—The pupa is small, measuring about 4 mm. when extended,
and feebly chitinised.

The paddles, which are about ‘7 mm. long, are oval, the ratio of length to greatest
breadth being about 1°6 to 1; they are furnished with a midrib, and an external
buttress which does not seem to extend beyond the proximal third of the blade.
The hook-shaped hair on the distal edge of the paddle is well developed and about
half the length of the paddle; proximal to it, near the end of the midrib, is a
short straight hair which is forked almost from its base, and is about one-sixth
the length of the hook-shaped hair. The paddles carry a fringe of rather delicate
hairs, longest on the distal border; this fringe appears to begin at a point beyond
the middle of the external border.

60 DRS. A. INGRAM AND J. W. S. MACFIE.

At the posterior angle of the 8th abdominal segment is a stout seta giving off
a number of branches; it is about one-quarter the length of the paddle, and its
spread of branches is only moderate.

At the posterior angles of the 7th, 6th and 5th segments are long spines, curved
and sharply pointed; on the 4th and 3rd segments the corresponding spines are
short and blunter: and on the 2nd segment very small and feebly chitinised.

In addition to the spines already described there are numerous hairs on the
body of the pupa, the most important being those on the dorsal aspect at the
posterior margins of the 3rd to the 7th segments which are arranged in two rows
on each side of the abdomen. The hairs of the one row, the outer lateral row,
are situated a little internal to the angle ; they are branched. The hairs of the other
row, the inner lateral row, are situated about mid-way between the posterior
angle and the middle line ; they are long single hairs on the 7th, 6th and 5th segments,
and branched hairs similar to those in the outer lateral row on the 4th and 3rd
segments. The dendritic hairs on the Ist segment are well developed.

Fig. 1. Pupa of Anopheles pretoriensis, Theo.

The respiratory trumpets are about ‘4 mm. long; they are, as usual, composed
of two parts, a closed part or meatus which is proximal, and an open part or pinna
which is distal. In the pupa of A. pretoriensis the meatus is short; the ratio of its
length to the total length of the trumpet being about 1 to 3. The pinna as it
expands forms a slight angle a little beyond the distal margin of the meatus; this
angle is much less acute than that shown by A. cosialss.

Diagnosis.—The pupa of A. pretoriensis has to be separated from the other
known pupae of this genus which have long, sharp and usually curved spines at
the posterior angles of the 7th, 6th and 5th segments, that is, from those of
A. costalis, A. funestus, A. marshalli and A. pharoensis. A. costalis is distinguished
by having on the paddle a well developed buttress and a fringe which begins at
a point proximal to the middle of the external border, and by the characters of
the respiratory trumpet to which reference has already been made; A. funestus
by having a spine of considerable size at the posterior angle of the 4th segment,
and by hairs in the inner lateral row on the 7th, 6th and 5th segments, which are
often sub-divided; A. marshalli (if the specimen examined by us was typical) by

EARLY STAGES OF WEST AFRICAN MOSQUITOS. 61

having a curved, but not hook-shaped hair at the end of the paddle, and by the
hairs in the inner lateral row on the 7th, 6th and 5th segments, which are often sub-
divided ; and A. pharoensis by the large hair on the end of the paddle not being
hook-shaped. |
- Habitat.—The larvae of this mosquito were captured in shallow pools in outcrops
of quartzite at Winduri, Tong Hills, in the Northern Territories of the Gold Coast,
18.vi.1918. They were associated with larvae of Stegomyia vittata, Bigot (sugens,
Theo., nec Wied.).
Anopheles rufipes, Gough.
Larva.—The larva of this mosquito is included by Edwards (loc. cit., p. 374)

in his key to the larvae of African species of the genus Anopheles.

Pupa (fig. 2).—The pupa is small, measuring about 3°5 to 4 mm. when extended,
and very strongly chitinised.

Fig. 2. Pupa of Anopheles rufipes, Gough.

The paddles, which are about ‘6 mm. long, are oval, the ratio of length to greatest
breadth being about 15 to 1; they are supported by a well-developed midrib and
an external buttress, which is not so well developed as that of A. costalis. The
hair on the distal edge of the paddle, which in some other species resembles a boot-
hook, is unfortunately missing from the two pupae examined. On each paddle,
near the end of the midrib, there is a short hair divided towards its tip. This
splitting towards their tips of the small hairs of the paddles is not a character of
specific importance, as it occurs occasionally in A. costalis. The paddle bears a
fringe which begins at a point a little proximal to the middle of the external border.

At the posterior angle of the 8th abdominal segment is a stout seta giving off
branches, which resembles the corresponding seta of the pupa of A. funestus; it
measures about one-quarter the length of the paddle.

At the posterior angle of the 7th segment is a curved and pointed spine; on the
6th, 5th and 4th segments the corresponding spines are straight and_ shorter.
These angle spines are smaller the more anterior the segment to which they are
attached ; those on the 3rd and 2nd segments being minute.

In addition to the spines already described there are numerous hairs on the
body of the pupa, especially along the posterior margins of the segments; the

62 DRS. A. INGRAM AND J. W. S. MACFIE.

most important are those which form the outer and inner lateral rows on each
side on the dorsum of the abdomen. The hairs forming the outer lateral row are
branched. Those forming the inner lateral row are long single, double, or triple
hairs on the 7th, 6th and 5th segments, and branched hairs, similar to those in
the outer lateral row, on the 4th and 3rd segments. The dendritic hairs or tuits
on the Ist segment are well developed.

The respiratory trumpets are about 36 mm. long; the pinna, the open part,
extends nearly the whole length of the trumpet, the ratio of the length of the
closed portion, the meatus, to the total length of the trumpet being nearly | to 5.
There does not appear to be any angle formed by the pinna a little above the
distal margin of the meatus, such as there is in A. costalis.

Diagnosis.—The pupa of A. rufipes has to be distinguished from that of A. mauri-
tuanus (neither having long sharp spines at the posterior angles of the 7th, 6th
and 5th segments), and this may be done by the spine at the posterior angle of
the 7th segment, which is long and curved in A. rufipes, but short and straight in
A. mauritianus. The harp-shaped form of the dendritic seta on the 8th segment
of A. mauritianus is not a good character, as it is inconstant.

Habitat.—The two pupae on which the above description is based were found
in pools of clear water collected in the dry bed of a stream near the garden at Bawku
in the Northern Territories of the Gold Coast, 11.vi.18. With them were associated
pupae of A. funestus.

Ochierotatus hirsutus, Theo.
Larva.—The larva of this mosquito has not yet been identified.

Cie Dy 5 oe |

Fig. 3. Pupa of Ochlerotatus hirsutus, Theo.

Pura (fig. 3).—The pupa is large, measuring about 6 to 7 mm. when extended,
and well chitinised.

EARLY STAGES OF WEST AFRICAN MOSQUITOS. 63

The paddles, which are about 1 mm. long, are broad, the ratio of length to
greatest breadth being about 1'2 to 1; they are supported by an external buttress
and a moderately well developed midrib. At the end of the midrib is the usual
long single hair, which is about one-eighth the length of the paddle. The paddles
are devoid of a fringe.

At the posterior angle of the 8th abdominal segment is a tuft of 7 to 8 subplumose
hairs, of which one or two may be branched. This tuft is short, about one-fifth
the length of the paddle.

A little above the posterior angle of the 7th segment is a tuft of 3 or 4 pubescent
hairs; this tuft is about one-sixth the length of the paddle. In a similar position
on the 6th, 5th and 4th segments there is a single hair, sometimes divided.

In addition to the tufts and hairs, already described, there are on the dorsal
aspect of the abdomen numerous small and delicate hairs, most of which are quite
inconspicuous ; those situated in the positions corresponding to the outer lateral
row on the 4th to the 6th segments are, however, better developed, and are long
double or triple hairs. On the same segments in the positions corresponding to
the inner lateral row are small tufts of delicate hairs. The hairs on the posterior
margins of the 7th segment are all small. On the posterior margin of the 3rd segment
is a small tuft in line with the inner lateral row, and a single hair of moderate length
a little above and external to it. The dendritic hairs or tufts on the Ist and 2nd
segments are moderately well developed.

The respiratory trumpets are about ‘7 mm. long; they are rather broad, and
have wide apertures. The ratio of the length of the closed portion, the meatus,
to the total length of the trumpet is about 1 to 1°5.

Diagnosis.—_See O. nigeriensis (p. —).

Habitat.—The two pupae on which the above description is based were found
in a collection of rain-water in a disused kerosene tin at Accra, Gold Coast, 11.xu.17.
The tin was partly concealed by vegetation.

Ochierotatus nigeriensis, Theo.

Larva.—the larva of this mosquito is included by Edwards in his “‘ Revised Keys
to the Known Larvae of African Culicinae ” (loc.cit., pp. 376 and 377), and is figured
by him.

Pupa (fig. 4).—The pupa is of moderate size, measuring about 5 to 6 mm. when
extended, and is very strongly chitinised.

The paddles, which are nearly 1 mm. long, are broad, the ratio of (eaioeh to greatest
breadth being about 1:2 to 1; they are supported by a midrib and by an external
buttress, which, however, is not conspicuous. Near the end of the midrib is the
usual long single hair, measuring in this species about one-eighth the length of the
paddle. The paddles have no fringe, but may show a few minute teeth, especially
about the middle of the external border ; these teeth are too inconspicuous to show in
the figure.

. At the posterior angle of the 8th abdominal segment is a tuft of about six (5 to 7)
hairs which are subplumose at their bases and sometimes branched towards their
tips. This tuft is short, its length being about one-fifth the length of the paddle.

64 DRS. A. INGRAM AND J. W. S. MACFIE.

Near the posterior angle of the 7th segment is a tuft of three or four (2 to 5) hairs
which are subplumose towards their tips. This tuft is a little shorter than the tuft
on the 8th segment, measuring about one-sixth or one-seventh the length of the
paddle. Near the posterior angles of the segments anterior to the 7th there are,
as a rule, single hairs, occasionally double hairs.

In addition to the tufts and hairs already described, there are on the dorsal aspect
of the abdominal segments a number of other hairs which are for the most part feebly
developed; the most notable are those occupying positions corresponding to the hairs
of the outer lateral row, which on the 4th to the 6th segments are long and usually
single or double. The hairs corresponding to those of the inner lateral row on the
4th to the 6th segments are quite insignificant tufts or hairs ; a similar tuft is present
on the 3rd segment, with a single hair of moderate length above and a little external
to it. The dendritic hairs or tufts on the lst and 2nd segments are rather feebly
developed, those on the 2nd segment being small and poorly chitinised.

Fig. 4. Pupa of Ochlerotatus nigeriensis, Theo.

The respiratory trumpets are about 6 mm. long; they have wide mouths. The
ratio of the length of the closed portion, the meatus, to the total length of the trumpet
is about 1 to 1°5.

Dagnosis.—The pupa of 0. nigeriensis resembles very closely that of the preceeding
species, O. hirsutus, and both fall into the group in our key which includes 0. minutus,
O. punctothoracis,,and O. caliginosus, species hitherto described in insufficient detail
to be differentiated. O. nigeriensis may show a few minute teeth on the external
border of the paddle which might be regarded as a fringe, but the tufts at the posterior
angles of the 8th and 7th segments are sufficient to separate this species from
O. albocephalus. No very satisfactory means has been found for distinguishing the
pupae of O. hirsutus and O. nigeriensis, but perhaps the hairs of the inner lateral
row on the 7th and 6th segments would serve, although they are inconspicuous.

Habitat.—The pupae of this mosquito were found in a borrow-pit containing
slightly muddy water outside the village of Ulu, Lorha District, Northern Territcries
of the Gold Coast, 7.vii.1918.

EARLY STAGES OF WEST AFRICAN MOSQUITOS. 65

Culex ager, Giles, var. ethiopicus, Edw.

Larva.—The larva of this mosquito has been figured by Edwards (loc. cit., p. 380)
and has been included by him in his “ Revised Keys to the Known Larvae of
African Culicinae.”’ One feature, however, which is not referred to by him and which
is not shown in the figures, may be mentioned because it is unusual, that is, the shape
of the mental plate. In the larva of C. ager var. ethiopicus this structure is shaped
like an equilateral triangle, and its margin when viewed at a moderate magnification
(1/6 objective and No. 3 ocular) appears to be crenated but devoid of teeth (fig. 5, a) ;
when examined with higher powers, however, the crenated margin is seen to be
composed of a large number (about 60) of minute teeth, which increase in size slightly
as they recede from the apex.

Fig. 5. Pupa of Culex ager var. ethiopicus, Edw.; a, mental plate of larva.

Pupa (fig. 5).—The pupa is large, measuring about 6 mm. when extended, and
well chitinised, the sockets of the hairs, even minute ones, being marked by a thick-
ened darkened ring. The paddles are infuscated. The trumpets have wide apertures,
which are directed anteriorly in the living pupa.

The paddles, which are nearly 1 mm. long, are broad, the ratio of length to greatest
breadth being about 11 to 1 ; they are supported by an external buttress and a well
developed midrib. Near the distal end of the midrib are two minute hairs, the one
considerably larger than the other ; the larger hair is sometimes divided towards its
tip. The paddle is infuscated at its distal end and over almost the whole of the
inner lobe, and has no fringe.

At the posterior angle of the 8th abdominal segment is a tuft of 5 to 7 subplumose
hairs, which are usually branched. This tuft is about one-quarter the length of the
paddle.

(C572) EB

66 DRS. A. INGRAM AND J. W. S. MACFIE.

A little above the posterior angle of the 7th segment is a tuft of 3 to 6 subplumose
or pubescent hairs, which are usually branched. This tuft is almost as long as the
one on the 8th segment, measuring about one-quarter the length of the paddle.
Near the angles of the segments anterior to the 7th are single hairs, those on the 6th
and 5th segments being strongly developed and the others more feebly.

Of the other hairs on the abdominal segments the most important are those forming
the inner and outer lateral rows: On the 4th to the 6th segments the hairs in both
these rows are long single, double, or triple hairs, those on the more anterior segments
being usually but not invariably more sub-divided than those on the more posterior
segments: on the 7th segment the hair belonging to the outer lateral row is small,
that belonging to the inner lateral row a small double or single hair. These hairs,
which are rather variable, are shown in the figure ; it will be observed that they
differ from the corresponding hairs of the pupa of C. quasigelidus. On the 8rd seg-
ment the most conspicuous hairs are a small tuft in line with the inner lateral row,
a single hair above and a little external to this tuft, and a second small tuft above and
external to the single hair. The dendritic hairs or tufts on the 1st segment are fairly
well developed and have about 8 or 9 primary branches.

The respiratory trumpets are long, averaging nearly 1 mm. in length, and have
wide apertures. The ratio of the length of the closed part of the tube, the meatus,
to the total length of the trumpet is about 1 to 1°7.

Diagnosis.—This pupa may readily be distinguished from the pupa of C. quasige-
lidus (see p. 67), the only other Culex (sens. str.) pupa known to have infuscated
paddles, as follows.

CO. ager var. ethiopicus. C. quasigelidus.

Paddles. Infuscation most notable on|Infuscation not extending
the inner blade, andextend-| more than half way down
| ing almost to the base on| the paddle at any point.

Habitat.—Larvae and pupae of this mosquito were found during June and July
in swamps or in pools of clear water containing algae (Spirogyra) at many places in
the Northern Territories of the Gold Coast, ¢.g., at Gambaga, Bawku, Navara, Tumu,
Wa, Bole, ete.

Culex quasigelidus, Theo.

Larva.—tThe larva of this mosquito has been described by Wesché (Bull. Ent.
Res. i, pp. 38 and 39), and Edwards (loc.cit., p. 383) has noted some age differences.
Two fully developed larvae examined by us showed the following characters, which
supplement the previous description : the mental plate small and composed of a
median toothwith about six (5 to 6) teeth on each side; the comb of about six (6 to 7)

EARLY STAGES OF WEST AFRICAN MOSQUITOS. 67

teeth arranged in a single row or in two short rows; the pecten of about nine (8 to 10)
teeth ; and the hairs on the dorsal border of the anal segment a fan-like arrangement
of about five hairs above and a long hair below on each side.

G2
ZA

Pete eae 3) H)

Wi itizae
iy Le

'
t
Lis 4,

Fig. 6. Pupa of Culex quasigelidus, Theo,

Pura (fig. 6).—The pupa of C. quasigelidus has been described very briefly by
Wesché (loc.cit., p. 39), with figures of the respiratory trumpet and of the terminal
segments of the abdomen. So long as this Culex pupa was unique in having a
“ distinct dark cloud ” on the paddle Wesché’s description sufficed for identification,
but as the pupa of C. ager var. ethiopicus (see p. 66) has also an area of infuscation
on the paddle it is now necessary to amplify the earlier description. The pupa is
of medium size, measuring about 5 mm. when extended, and is well chitinised.

The paddles, which are about ‘8 mm. long, are broad, the ratio of length to greatest
breadth being about 11 to 1 ; they are supported by a well developed midrib and an
external buttress. Near the end of the midrib are two small hairs, the one larger than
the other; the larger hair, which measures only about one-twelfth the length of the
paddle, is sometimes divided towards its tip. There is a patch of infuscation at the
distal end of the paddle which involves both the inner and outer lobes but does not
extend far down towards the insertion of the paddle. There is no fringe on the
paddles.

At the posterior angle of the 8th abdominal segment is a tuft of 8 or 9 subplumose
hairs, which are also branched. This tuft is about one-quarter the length of the
paddle.

Near the posterior angle of the 7th segment is a tuft of 4 or 5 subplumose hairs,
which are also branched. This tuft is almost as long as the one on the 8th segment,
measuring about one-quarter the length of the paddle. Near the posterior angles
of each of the anterior segments, the 6th, 5th, 4th, and 3rd, is a delicate tuft of 3
or 4 hairs.

(C572) E2

68 DRS. A. INGRAM AND J. W. S. MACFIE.

Of the other hairs on the abdominal segments the most important are those forming
the inner and outer lateral rows. The outer lateral row is composed of a minute
divided hair on the 7th segment, a long single or double hair on the 6th and 5th
segments, and a tuft of three or more hairs on the 4th segment. The inner lateral
row on the 7th to the 4th segments is composed of tufts which are larger in the more
anterior segments. These rows are shown in the figure. On the 3rd segment the most
conspicuous hairs are a tuft in line with the inner lateral row, a double hair above
and a little external to this tuft, and another tuft above and external to the double
hair. The dendritic hairs or tufts on the Ist segment are well developed and have
about a dozen primary branches.

The respiratory trumpets are long and narrow, about ‘9 mm. in length. The ratio
of length to greatest breadth is, in a mounted specimen, 5 to 1. The trumpet is
banded, the proximal end being pale and the distal end dark, and the middle zone is
divided into two portions, the upper being pale and the lower dark. The ratio of
the length of the closed portion, the meatus, to the total length of the trumpet is
about 1 to 1:16, that is the meatus is very long.

Dragnosis.—See Culex ager var. ethiopicus (p. 66).

Culex univittatus, Theo.

Larva (fig. 7).—The head is large, nearly as wideas the thorax. The brushes are
well developed. The antenna is dark and covered with prominent spicules, the
hair-tuft being situated just before the middle point of the shaft. The mid-frontal
hairs are multiple, the constituent hairs being lightly subplumose ; the ante-antennal
tuft is also composed of lightly subplumose hairs. The mental plate is triangular
in shape ; it has a central tooth with about ten to twelve teeth on each side, the
lateral teeth being small, but gradually increasing in size towards the base of the plate.

The thoracic plumes, which are formed of subplumose hairs, are well developed.

The lateral abdominal hairs are multiple on the first two segments, triple on the
third segment, and thereafter single on each segment.

All the plumes on the 8th abdominal segment are formed of simple or only slightly
subplumose hairs. The comb consists of about 8 spines arranged irregularly. The
siphon is about seven or eight times as long as its width at the basal ring, and tapers
regularly from base to apex. The pecten extends about two-fifths of the length of
the siphon from the basal ring ; it consists of 13 to 16 teeth of which the two or three
furthest from the base are more detached and have no secondary spines. There is
a slight tuft of a few simple hairs some distance beyond the last tooth of the pecten,
namely, at a point about three-quarters of the length of the siphon from its base.

The anal segment carries very long papillae, three times the length of the segment,
the dorsal pair being slightly longer than the ventral. The beard is well developed.
The hairs on the dorsal end of the segment are peculiar, consisting of a fan-like
collection of 8 or 9 hairs above and an extremely long hair below on each side. The
long hairs are nearly twice as long as the anal papillae.

Dragnosis.—EKdwards includes the larva of this mosquito in his key to the larvae of
the genus Culex (Bull. Ent. Res. iii, p. 381), but states that there are six teeth in the
pecten, whereas our specimens had 13 to16. This discrepancy does not, however,

EARLY STAGES OF WEST AFRICAN MOSQUITOS, 69

confuse the identification of this species or its differentiation from those whose larvae
have a siphon about half as long as the abdomen.

Fig. 7. Head and terminal segments of larva of
Culex univitiatus ; a, mental plate.

Habiiat.—The larvae of this mosquito were found in a grass-grown pool containing
clear water at the edge of a swamp at Navaro, Northern Territories of the Gold
Coast, 23.vi.18.

I es | |
eye ey +n * ; ig 7 pe ‘

ie a ofib i fon be ents

4,6 :
be ia 8)
S ree:
ty
“2 Bae
i
4
a
ee et ,
4
+ a :
a
m i
\ “ i mt f
v
/
4
)
oa
k
>
2
e)
2)
&
Ay
oe «
—
i
uF :
& <
ts na
i oy 5
‘
1
Ve ‘i
‘. my
th Noe :
mi
y
7 F ’
Tot
°
ith.
-
* .
4
j ~ : ,
; ¢ hut
«
a
, bi
‘ 45
.
ies r
<P 7
I
- ”
t
my.
f os ee jar
~ “

,
Cee a
Ne Le f j }
a if a
Te ee LT wih hen ihr
ss ie |

. oe i a me Pit é

Miige WOVEN

4 i aE
Fil BRADNER
4
P

TSETSE FLY IN SOUTHERN RHODESIA, 1918.
By Rupert W. Jack, F.E.S.,

Entomologist, Department of Agriculture, Rhodesia.
(PuatEs IJ-IV, Mars I-III.)

The history of tsetse-fly in Southern Rhodesia up to the present year (1918) con-
tinues on the whole to be one of expansion, although on the other hand one small,
but important, belt appears to have become extinct. In addition, a fly area in the
Mocambique Company’s territory has extended up to our eastern border, with the
result that serious losses of cattle from trypanosomiasis have occurred on the farms
in that region. Detailed reports on the advance or retrogression of Glossina morsitans
in other parts of Africa appear to be lacking, and it is felt that in view of the position
of this territory in relation to tsetse-fly a review of the situation to the present day
may not be without general interest.

The “fly areas” in Southern Rhodesia may be referred to as follows :—
(1) Sebungwe Belt, occupying the greater portion of the western half of the Sebungwe
district ; (2) Umniati Belt, lying astride the Umniati river, partly in the Hartley
and partly in the Sebungwe districts ; (3) Northern Belt, occupying the greater part
of the country between the escarpment and the Zambesi River in the northern
portion of the Lomagundi district, and extending eastward into the Darwin district ;
(4) Jetjenini Belt, in the neighbourhood of Jetjenini Mountain, Lomagundi district ;
(5) Suri-suri Belt, in the vicinity of the Suri-suri river, between Hartley and Gatooma,
in the Hartley district ; (6) Melsetter Border Belt, lying almost entirely in Portuguese
Hast Africa, but affecting certain farms on the eastern border of the Melsetter district.

(1) Sebungwe Belt. Owing to its extent and the fact that sleeping sickness is
endemic in a certain portion of it, this is at present the most important fly-area in
the territory. The extension of the infested country since it shrank to insignificant
proportions in 1896-7 until the year 1913 has been described in a previous paper.*
During the past five years this progress has been continued and the new country
involved is shown in Map1. The present area of this belt is estimated at about 4,500
square miles.

The value of careful notes on the channels along which the fly has spread lies chiefly
in the accumulation of information concerning factors which favour the pest. It
is unfortunate that circumstances have not permitted of more detailed work in this
connection, but such observations as have been made were carried out with this
express object in view. The practical point lies in the possibility of rendering
threatened areas unsuited to the pest or of interposing a barrier and thus checking
the further extension of the belt.

Broadly speaking, the contentions of the early hunters concerning the association
G. morsitans with big game are fully supported in this area, although this is not in-
variably the case elsewhere. The increase of the fly and the extension of its bound-
aries have corresponded with a notable increase of the larger antelopes, and to a. less
extent of buffalo. The new territory occupied by the fly is all game country. The
most rapid extension has taken. place through country where game abounds; fly is

12 RUPERT W. JACK.

most abundant at points where game congregates during the latter part of the dry
season, and the advance appears to have come to a standstill at more than one
point where the country beyond is sparsely inhabited by game.

These statements call for some amplification. A glance at Map 1 shows that during
the years immediately following the rinderpest epizootic in 1896, the fly was confined
to the vicinity of Manzituba Vlei.* From this vlei in a direction roughly 8.8.W. lies
a practically unbroken stretch to the Shangani River, in which game is now present in
almost primitive abundance, and as the map indicates, a notable extension of the
fly has taken place along this channel, and is still in progress. In a direction more
S.W., however, the edge of the belt has been at a standstill in the neighbourhood
of the old Native Commissioner’s camp at Kariyangwe since 1907. West of this
station the country is notable for its rocky character and the scarcity of antelope,
and it appears probable that the area between this point and the Zambesi River
never has been fly country. At the point marked Manjolo on the Nogola River
game, especially eland, is moderately abundant.

It is possible that the reason for the lack of fly in this region may be due to unsuit-
able vegetation or other causes ; but there are numerous spots where the vegetation is
to all appearances quite suitable, and there is on the other hand reason to believe that,
where game is sufficiently abundant, tsetse can exist in great numbers although the
vegetation does not seem to afford ideal conditions. The country lying immediately
to the north of the headwaters of the Mzola River, for instance, is very open, being
in fact known as the Matobolo Flats, the native name ‘“ Matobolo”’ signifying an
open plain. Dotted about the plain are a number of small termite mounds, which
bear stunted trees of an evergreen habit. In crossing a portion of these flats in 1916
the writer was surprised to find G. morsitans abundant, sheltering in the shade of
the stunted trees and darting out to attack the passer-by. Game is plentiful, and
certainly constitutes the only apparent source of blood to the flies, which in hot
weather appear to be extremely confined in their range. It would appear therefore
that a very limited extent of shade will serve G. morsitans, provided that living blood
js sufficiently plentiful in the immediate vicinity. Where the food supply is more
uncertain, greater freedom of movement is doubtless a necessity. We cannot
therefore attribute the unsuitability to tsetse of certain tracts of country exclusively
to small differences in the type of vegetation, which are not apparent to the eye. We
must look for some combination of factors in which facilities for more or less regular
meals are probably very important.

The extension of the Sebungwe belt to the S.E. and E. calls for little comment,
as the country involved is all well stocked with game.

The writer is not familiar with most of the country to the north of the belt. Fly
was located on the lower Sengwe River in 1913, and this belt was supposed to be
isolated. It is possible, however, that it may in reality have been connected with
the main belt in the vicinity of Sinatchungwe’s kraal. In any case, reports of
# : a the authority of the late Mr. Val Gielgud, formerly Native Commissioner for this

ISUI1Ct.

{ In 1868 James Chapman remarked on the absence of fly on the south bank of the

Zambesi between the Gwaai confluence and a point a few miles below the Sebungwe

River confluence, a distance of about 30 miles. This strip contains no game except a
small variety of duiker.—‘‘ Monograph of the Tsetse Flies,” E. A. Austen, p. 143.

TSETSE FLY IN SOUTHERN RHODESIA, 1918. 73

Government officials and hunters indicate that this is the position at the present day ;
but whereas the region of the Sengwe is heavily infested, the narrow neck joining
the two belts contains but little fly, and mules, atleast, still pass through with apparent
impunity. A considerable extension of the belt in the region of the lower Sengwe
and on the Zambesi is also reported.

(2) Umniati Belt. The history of this area is of peculiar interest. In 1896-7
it appears to have shrunk to very small dimensions, but the exact position at that
time is somewhat obscure. Subsequently the belt extended, in common with other
fly areas in the territory and, it may be added, in the presence of considerable quan-
tities of big game. Its present area is judged to be about 1,000 square miles. The
writer first visited this part of the country in November 1910, when the southern
limit of fly on the Umniati was about seven miles further north than it is to-day. At
that time game was plentiful on both sides of the river but especially on the west,
and included elephant, rhinoceros, eland, kudu, sable, waterbuck, impala and wart-
hog, whilst bushbuck was, and remains still, moderately abundant along the river
itself, In addition to game, large and numerous troops of baboons frequented the
river, and have not decreased in any way to the present day. These animals are,
as a matter of fact, more abundant in this region than in any other part of the
territory with which the writer is familiar. The adjacent portion of the Hartley
district has been open to free shooting, with one break, since 1905, and this has had
an adverse effect on the game on the Sebungwe side of the Umniati river. Pro-
fessional hunters, camping in the “open area,” have poached regularly beyoud
the Umniati and of late years game has become very scarce where it was abundant
as late as 1910 and fairly abundant up to 1912. Between 1910 and 1912 the fly
extended about seven miles up the Umniati, but since that year it seems to have
come to a standstill at this point. The belt now extends westward as far as the
base of the Mafungabusi plateau. The writer has never visited the extreme northern
limit of this belt, but is quite familiar with that portion which lies south of the
latitude of the Umniati-Sakugwe confluence. In this part it may be stated that fly
is everywhere scarce away from the vicinity of the Umniati River, but at one point
on the western side four or five miles above the junction with the Njongwe (Dumbwi)
it is present in great numbers. According to native evidence this is the spot where
the fly survived the adverse conditions of 1896-7, and it is certainly the locality
where the marked reduction of the game at the present time has had no appreciable
effect in reducing the numbers of the pest. The persistent presence of large numbers
of baboons is, however, a complicating factor, and the association of the fly with the
vicinity of the river is consistent with the theory that these animals may be acting
as hosts to the insect. It is not necessary to enter at present into the conflicting
evidence as to whether these animals ever allow thenselves to be fed upon by tsetse.
It is desired merely to record the facts regarding the Umniati fly belt as far as they
aré known at present. It may be added that baboons do not enter as a serious factor
throughout the great bulk of the fly-infested country in the territory.

It is to be noted that the fly has shown no tendency to spread from this belt in
a south-easterly direction towards the railway line, although the country is apparently
suitable. The game has been kept down very effectively by hunters in this area for
a number of years past.

T4 RUPERT W. JACK.

(3) Northern Belt. The writer has not visited this area since 1912 and is un-
acquainted with the conditions prevailing there at present. The region is untouched
by civilisation, and since the year of the rinderpest epizootic the fly has, according to
numerous reliable reports, increased and spread very greatly. The latest reports
from the’ Native Commissioners indicate the present extent of tsetse as shown in
the map. The infested country is estimated to include about 2,500-3,000 square
miles at the present day.

(4) Jetjenim Belt. The infested country in this region remains untouched by
civilisation as far as the forest is concerned, but owing to its accessibility it is a
favourite shooting area amongst hunters, and game is less plentiful than it has been
in the past. Tsetse occurs, but not in great numbers, and of late years has shown
little or no tendency to spread. In 1914 an outbreak of trypanosomiasis amongst
some cattle which had been running for eight or nine months on the Chumsenga
River, which was supposed to be outside the fly belt, gave the impression that a south-
ward movement was in progress. This may indeed have been the case at the time,
although the evidence that this river had only recently become infested was incon-
clusive, and fly was certainly present when the cattle were taken there, as proved by
the report of a trooper of the B.8.A. Police the previous year. It may be mentioned
that game, especially kudu, sable and wart-hog, was found to be moderately abundant
in the neighbourhood in September 1914. It is also significant that the northern
limit of native-owned cattle in this region has apparently remained the same for a
number of years past, namely, about the headwaters of the Ridziwe River and along
the Mvume. Capt. Thornton, of the B.S.A. Police, found tsetse at Doma Hill mm
1909 and this still marks the southern limit of fly in this direction. The stationary
nature of the belt is all the more significant in the light of the fact that in pre-rinder-
pest days the fly used to extend considerably further south. The southern limit
of fly in this district in 1896 according to Mr. R. T. Coryndon, C.M.G., now Governor
of Uganda, is marked on the map. It may be added that there is a fairly wide extent
of primitive country between the present limits of the fly and the limits given by
Coryndon, although a portion of the intervening area has been settled. The unsettled
area doubtless contains a moderate amount of game at certain seasons, though the
writer has seen very little in the course of some three journeys ; but the region is
too near to settlement for large game to be really plentiful. The Jetjenini fly belt is
about 450 square miles in extent.

(5) Surt-surt Belt. From all appearances tsetse has now died out in this area.
The history of the fly in this belt up to the year 1913 has already been published, but
in order to make the sequence of events clear it is necessary to recapitulate. Tsetse
in the Hartley district shared in the great shrinkage of 1896-7 and indeed has not
reappeared in many parts which were formerly infested, notably in that portion
lying east of the railway. The late Mr. Scott, of the Native Department, formerly
resident in the Hartley district, supplied information to the Director of Land Settle-
ment, dated 12th June 1909, of which the following is a condensed version. “ In
1895 there used to be a fly belt between a point four miles from Old Hartley and the
headwaters of the Ngombe or Mombe river. In September 1897 I never saw or
heard of fly during a trip from Fort Martin, to the Beatrice Mine, to the Mgezi at the
Manizi range, down between that river and the Msweswi to the Singondo junction

TSETSE FLY IN SOUTHERN RHODESIA, 1918. 75

with the Umniati, and then to Hartley. Natives stated that fly had disappeared on
game being destroyed by rinderpest. Subsequently to that year I criss-crossed the
whole district, except the piece of country bounded on the east by the railway, north
by the Hastings road and south by the Msweswe, and never found fly at any time.
In 1898-9 I heard of it at the water-holes and headwaters of the Shagari river. In
1901 I heard of its being in the country between the Beri and Umfuli rivers, near their
junction, but messengers sent out did not bring any in nor did they find any. In
the latter part of that year or early in 1902 I caught tsetse on the railway line on
the headwaters of the Suri-suri River. In August 1902, J. McAdams brought in a
number that he had caught (some six or eight miles, I think) west of the Golden
Valley Road on the Suri-suri River.” (Note.—The last statement should probably
read ‘‘ west of the Suri-suri River on the Golden Valley Road,” as the latter runs
approximately east and west.)

It is evident therefore that from 1898 onwards tsetse began to make its presence
felt again in the Hartley district, and that by 1902 it was present in considerable
numbers in the neighbourhood of the Suri-suri and Shagari Rivers. There is ample
evidence to the effect that during the years following the fly was abundant in
certain parts of this area and continued so until the year 1908.

The course of events in the Hartley district constitutes such a valuable record in
connection with the question of the relation of tsetse-fly to the larger mammals that
at the risk of being accused of labourimg the various pots the writer is impelled to
place as much of the available evidence as possible on record. From a strictly
scientific point of view it is rather unfortunate that the old fly area in this locality
has been deforested, as there is a tendency at present to ignore the events, obvious
at the time, which preceded the deforestation. The points it is desired firmly to
establish are (1) that there was a notable reduction of fly in the basin of the Suri-
suri, Umswezwe and Shagari Rivers between the years 1908 and 1913; (2) that there
was a coincident reduction of large game ; (3) that a considerable area surrounding
the upper portions of the three rivers mentioned still remained virgin forest as late
as 1913, when the wood contractors commenced to work this area in connection with
the supply of timber and firewood to the Cam and Motor mine. |

The year 1908 seems to have been marked by an unusually large number of cases
of trypanosomiasis around the Suri-suri fly area, and the losses continued during the
early part of 1909. This may have been due to a culmination in the numbers of
tsetse, to the increased hunger of the flies due to the reduction of game up to that
year, or to increased agricultural and industrial activity in the neighbourhood. A
number of farms on the railway line near the headwaters of the Suri-suri were taken
up in that year, and considerable losses were experienced in this region. In addition
the Veterinary Department records losses at the Dreadnought and neighbouring
mines north of Gatooma, at Hippovale farm on the Umfuli River, on the Eiffel Flats
east of Gatooma, and numerous cases amongst transport oxen working from Gatooma
and Hartley. |

Direct evidence as to the abundance of the flies themselves is also not lacking. Mr.
W. EH. Masters, of the Dreadnought Mine, in a letter to the writer states that in a well-
Jenown. vlei slightly to the east of the Dreadnought Mine on the Hartley-Golden
Valley road, fly about this time was present in hundreds on every warm day. Mr.

76 RUPERT W. JACK.

Ll. E. W. Bevan, Veterinary Bacteriologist, records that in this year a local resident
caught fly for him “ by the matchboxful” at the foot of a range of hills running
N.E. from Gatooma, parallel to and not far from the railway line. Dr. Alex. Macken-
zie’s experience on the Mowiri River in this year and other evidence in regard to the
abundance of fly near the Suri-suri have already been recorded. There appears
therefore no room for doubt that tsetse was really abundant in certain portions of
this area in 1908.

When the writer commenced investigations in this region in the latter part of
1909 he naturally sought to profit by the experience of local residents in regard to
the spots where fly was to be found in most abundance. Early in August Dr. Mac-
kenzie very kindly accompanied him to the Mowizi River, where fly had caused him
so much inconvenience the previous year, and the writer revisited the area almost
monthly in the latter part of 1909 and in 1910. The highest number of flies seen in
any one day was nine, notwithstanding the fact that the whole object of the visit
was tosearch for them. On severaldaysnoneat all were encountered. Early in1910
a visit was made to the vlei near the Dreadnought Mine and no tsetse were seen in
the course of two days. The range of hills near Gatooma was also visited about
the same time, with completely negative results so far as finding tsetse was concerned,
and as a matter of fact farms along the north-west side of the railway line, including
the hills mentioned, began to be taken up from 1909, and cattle were introduced,
as they would not have been had the fly been much in evidence at that time. Some
of these cattle contracted trypanosomiasis, but the losses were not sufficient to cause
the abandonment of the farms, and cattle have been present continuously ever since.
It should, however, be mentioned that Mr. Ll. HE. W. Bevan, employed an injection
as a cure for trypanosomiasis in cattle with considerable success at this time.

Reference to the annual reports of the Chief Veterinary Surgeon for the years 1909
to 1912 is very instructive, although the record is to some extent complicated by the
fact that Gatooma is the base for certain mines in or near the Umniati fly belt and for
hunters taking advantage of the free shooting, who have been in the habit of camping
annually on the Umniati river, frequently in the fly. However, references to try-
panosomiasis in the Hartley district for the four years mentioned are as follows :—
1909: “In Hartley district the mortality during the year was very considerable.
Accurate figures are not available, as many animals showing symptoms of illness
were at once disposed of.’’ Incidentally, the increased mortality was attributed to
the increased number of cattle employed for the mines and on farms in the district.
1910: “ The mortality from this disease has, especially in the Hartley dostrict,* shown
a marked decrease. Whether such decrease is due to the effects of the suspension
of the game laws or not, I am not prepared to discuss here, but an effort will be made
to obtain further information from members of the staff and residents in the affected
districts.” 1911: “ Only a few cases of this disease occurred, in the districts of
Hartley, Lomagundi and Mafungabusi, which contain various areas in which tsetse-
fly exists.” 1912: “In Hartley district fly are still to be found in small numbers
in the farming and mining areas, but only a few deaths of cattle were reported.”

Although, owing to special circumstances, which will be dealt with shortly, the
month of December 1913 marked the commencement of a sharp increase in the

*Author’s italics.

TSETSE FLY IN SOUTHERN RHODESIA, 1918. ri

‘number of cases in the Hartley district, the impression up to that time is summed
up in the following passage from the report of the Director of Agriculture for that
year :—

“The free shooting of game allowed in the neighbourhood of Hartley, with the
object of eliminating tsetse-fly from that populous area, appears to have achieved
its object in a very large measure, as the fly is by no means as abundant as formerly ;
and though domestic stock are more widely distributed, very little is heard of their
being fly struck. The suppression of trypanosomiasis amongst cattle, consequent
on the reduction of fly by driving away the game, cannot be proved to demonstration,
but certainly, as in the destruction of dogs to eliminate rabies, everything points to
this result having been obtained.”

It will be seen therefore that not only the writer’s personal observations but the
whole veterinary record for the district prove the great reduction of fly that took
place in the area between the years 1908 and 1913.

The fact of a great reduction of game having taken place during the same period
scarcely needs proof, seeing that free shooting was allowed in a comparatively populous
district. The writer noted a steady diminution in both game actually seen and fresh
spoor from August 1909 onwards, and the increasing difficulty of obtaining a “ bag ”
was a common complaint as time went on, the scarcity of wart-hog, which was at
one time abundant, being especially noted. The marvel is that any game survived
the constant persecution ; but so attractive was the area at certain seasons of the year
that big buck were to be met with at times even after the deforestation commenced
and are probably not altogether absent at the present day. Professional hunters
shot largely in this area up to about 1909, as the remains of their camps, littered with
bones and refuse, in that year indicated. They apparently did not find it worth
while in the years following, but residents, at Hartley especially, made hunting ex-
cursions to the Suri-suri an agreeable week-end recreation as long as there was a
reasonable chance of getting a shot. The reduction of game is actually a more
difficult matter to prove, apart from personal observations, than the reduction of
fly, but the Hartley residents who were in the habit of using this area as a happy
hunting ground are not likely to call the statement into question.

The fact that a large area remained untouched by civilisation, with the exception
of the reduction of game, at the time cutting on the wood contract commenced, is
easily established. The writer has in his possession a copy of a large scale map
drawn up by the contractors showing the area blocked out for wood cutting, the areas
reserved for other mines and the remaining untouched forest in the neighourhood.
There was a continuous area of virgin forest not far short of a hundred and fifty
square miles, embracing the bulk of the old fly belt, at the time the wood cutting
concession was granted, dating from the Ist January 1913 (see Map III).

As it is impossible that the presence of European settlement could have any direct
effect on tsetse-fly four or five miles away in the shady depths of its native forest,
there are only two alternative explanations of the Hartley phenomenon. Either
the reduction of the game was the direct cause of the reduction of fly or we must
appeal again to natural causes and coincidence.

In connection with the Cam and Motor wood contract, a light railway was run
down into the fly area in 1913 and in October of that year a considerable number

78 RUPERT W. JACK.

of working oxen were introduced. A number of these oxen contracted trypanosomia-
sis. The Veterinary Bacteriologist in his report for 1914 states :—“‘ It is estimated
that nearly 25 per cent of these transport animals became infected, but it is impossible
to state definitely how many were suffermg from trypanosomiasis alone, and how
many were the victims of starvation during the drought of the past season.”

In view of the heavy losses from trypanosomiasis that have occurred of recent
years amongst cattle living outside the known fly areas, these figures are far from
suggesting any great abundance of testse in the Suri-suri at the period involved.

The apparent effect of the clearing of the forest, possibly combined with the intro-
duction of an abundant food supply in the shape of oxen amongst the lingering tsetse,
was interesting. The forest was cut from the main railway line back towards the
Mowiri River, skirting the edges of the alienated farms adjoining the railway, and
what little fly remained would presumably follow the receding forests. Away from
the edges of vleis and water-courses, however, the forest becomes leafless after the
end of July and is not frequented by tsetse at this time, hence the fly may be judged
to have moved back along the Mowiri to the Hartley-Shagari road. This road runs
through the farm Ameva close to Hartley and on this farm for the first time there
was an outbreak of trypanosomiasis in 1914. At the same time two cattle on
Hippovale were suspected of being struck, but blood-smears did not confirm this.
The traffic between Shagari and Hartley would account for the fly being carried in
the direction of the latter, and it may be judged that the first effect of the clearing
of the forest was to scatter the remnants of the fly. A few fly were still to be found
in 1914 to the 8.W. of the cleared area, associated with the Nswunzwe river, but in
June of that year the writer encountered only a single specimen, when looking for
fly in this locality. The scattered remnants of the fly apparently died out in a com-
paratively short time, and no cases of trypanosomiasis associated with the Suri-suri
area have been recorded during the past two years or more.

(6) Melsetter Border Belt. Since 1914 cattle on certain farms close to the Portu-

guese border in the Melsetter district have become infected with trypanosomiasis
during the summer months. This has been due to the extension of a belt of fly in
Portuguese Territory towards and up to the Rhodesian border. In this case it is
not G. morsitans that is implicated, but Glossina brevipalpis, Newst., and G. pallidipes,
Austen. These two species are intermingled in the strip of country lying across
the border from the affected farms. In this region the border practically follows the
division between the high and low veld, although Spungabera, the seat of government
for the Mossurize district of the Mocgambique territory, is situated on the high veld.
Most of the farms on the Rhodesian side are of an open character and certainly not
suited to become permanent fly haunts. The low veld on the Portuguese side of
the border is heavily afforested, and the forest extends into Rhodesian territory up
the river valleys, several of which are very deep, the water-level beg sometimes
2,500 feet below the surrounding country. In other places the forest extends up the
side of the escarpment across the border ; whilst one or two farms bear open forest,
distinct from the sub-tropical growth found in the river valleys. Experience indicates
that cattle on the Rhodesian side are only subject to infection during the wet season,
and it is judged that the flies extend their range at this season, as is the case with
G. morsitans. Practically all cases are traceable to stock having been in contact

-TSETSE FLY IN SOUTHERN RHODESIA, 1918. 79

with the forest during the wet season. They appear to graze on its outskirts with
impunity in the winter. No specimens of tsetse have as yet been taken within the
Rhodesian border in this region, and the flies in 1917 were present only in small
numbers in the neighbourhood of the border on the Mogambique side. On the other
hand the losses in cattle have been severe in some cases, aS many as eighty-six head
during the past season. There is thus reason to think that in Melsetter, as elsewhere,
some agency other than tsetse may serve to transmit the infection from an infected
beast to others herded in its vicinity.

The situation in the Melsetter district is probably unique in its way, constituting
an instance where a very considerable degree of settlement and development lies
immediately adjacent to practically primitive conditions across a boundary which
is both political and natural.

Breeding Haunts of Glossina morsitans.

Very little work has been carried out in this connection, the only area searched
for pupae since the writer’s previous work on this subject being Sipani Vlei, lying
east of the Sengwe River and north-west of Gokwe in the Sebungwe district. This
Vlei is intensely infested with fly late in the dry season and bears clumps of evergreen
trees on termite mounds, as in the case of Manzituba Vlei, which has already been
described.* In one hollow tree alone some forty live pupae were secured and
several hundreds of empty pupa-cases, constituting probably a record for any
single location. Pupae could be found by searching in almost any sheltered
situation at the bases of the evergreen trees near the Vlei and a considerable
number were there secured. It may be stated, as bearing on the question of
the season during which G. morsitans breeds most freely, that the large haul men-
tioned above was secured in November 1914, and that at the base of the same tree
in August 1916 less than ten live pupae were secured, although fly was noted as being
“extremely numerous and attentive.’ This observation is of little importance
in itself, but is mentioned here as it happens to be in accordance with the conclusions
of Lloyd from careful notes made in Northern Rhodesia.t It is to be remarked also
that in the writer’s previous work in this connection, in August 1911, only four live
pupae were secured amongst a total of ninety-one, the remaining cases being empty.
Lamborn also supports Lloyd’s observations in this particular,{ and it would seem that
the diminution in the breeding rate and the prolongation of the pupal period during
the colder months of the year is more or less established. Lloyd, however, came to
a further conclusion, namely that breeding practically ceases during the wet season§
and that the latter part of the dry season, from July to October or November,
would therefore comprise practically the whole of the active period of reproduction.
This is exactly the season when G. morsitans in Southern Rhodesia is concentrated
in the permanent shade bordering on vleis, rivers, etc., and it would certainly be
of considerable interest if Lloyd’s deductions in this respect were correct. The
statement, however, needs confirmation, and whilst the writer is unfamiliar with
conditions north of the Zambesi, it would appear that the deduction is based upon

* Bull. Ent, Res., ii., p. 357.
} Bull Ent. Res. v, p. 58.

{ Bull. Ent. Res., vii., pp. 29-50.
§ Bull. Ent. Res., v., p. 58.

80 RUPERT W. JACK.

a questionable premise, namely, that the distribution of the fly is the same in both
wet and dry seasons. If this is the case in the locality where the observations were
made, the results obtained have great value; but it is certainly not the case in
Southern Rhodesia, nor, according to Lamborn, in Nyasaland. The vast bulk of the
forest in this territory consists of some three species of Brachystegia, producing the
type of forest known to the Matabele as “ gusu,” and the well-known mopani. The
habit of “ gusu ” forest varies with the soil and subterranean water-supply. Broadly
speaking, the trees concerned appear to become leafless all over the territory at
about the same time of year, namely late in July. In many places, as at Salisbury,
the new leaves appear again almost immediately, and if this were the case everywhere,
it is to be presumed that no marked seasonal change in the distribution of fly in
gusu country would take place. In all the fly country known to the writer, however,
the forest does not as a rule come into leaf again until the middle of November or
later, bemg apparently dependent to some extent on the commencement of the
heavy rains.* As a striking instance of the effect of different soil conditions, in
August 1916, the writer crossed the Mafungabusi plateau towards the Umniati River.
Water is well held on the plateau, and permanent streams pour down the gorges into
which the edge is broken, only to lose themselves during the dry season in the deep
sand which extends from 15 to 20 miles between the plateau and the Umniati River.
On the plateau the whole forest was in full leaf, whilst between the plateau and the
river it was quite leafless. With respect to mopani, this type of forest has never
been noticed in foliage between the end of July and the beginning of November and
has frequently been noted as leafless after the middle of the latter month.

During the leafless period tsetse congregates in those parts of the fly belt where
there are evergreen trees, such as Ficus, Acacia, Parinarvum mobola, Diospyros
mashuna, Kigilia pinnata, etc. Such trees occur on the banks of water-courses and
the edges of vleis, even though these dry up on the surface during the dry season.
Some of them are also found on “ ant-heaps”’ thinly scattered through the forest.
In some spots, as on the edges of vleis, where there is abundant underground water
the Brachystegias come into leaf again in August, even though the vast majority of
the same species in the vicinity remain leafless till the advent of the rains.

The concentration of G. morsitans from August to November and its general dis-
tribution during the wet season and first half of the winter renders observations
bearing on the question of its most prolific breeding season very difficult indeed.
There is a general opinion that the fly is most abundant in October and November,
and this may be the case. It is certainly most in evidence during these two months,
but, being much more widely scattered at most other times, it is difficult to draw
comparisons. Nevertheless, the establishment of the fact that the hot months pre-
ceding the rains are of maximum importance to the welfare of the fly would be of
considerable value. It may well be that, apart from the breeding rate, the summer
months show the greatest mortality amongst the flies from the attacks of msect
enemies and that they generally become reduced in numbers during this period ;
but as the earliest month with which we can compare the prevalence of fly at any

*An exception occurred during the present year (1918), apparently due to the pheno-
menally heavy rains of last season, the gusu forest being generally two months ahead
of its usual time. The mopani, however, was not affected.

TSETSE FLY IN SOUTHERN RHODESIA, 1918. 81

particular spot with its prevalence in November is August, between which month
and the past wet season the cold months of May, June and July have intervened,
we are in a similiar position with regard to judging whether the fly loses ground in
the wet season or not, as in estimating the comparative breeding rates. Although
the connection may not be apparent, such facts might well shed some light on the
question of the dependence of G. morsitans on the larger fauna of the forest. The
matter will be referred to again under its own heading.

Transmission of 7. »ecorum in the Absence of Tsetse.

In spite of the fact that direct experimental proof is still apparently lacking,
evidence indicating that Trypanosoma pecorum is commonly spread amongst asso-
ciated animals in the absence of Glossina has accumulated im different parts of
Africa during recent years to such an extent that the fact appears now to be
widely accepted. (Bull. Ent. Res., viii, pp. 35-41.) In any division of the
continent where fly belts occur this danger must necessarily be of considerable im-
portance, and not least in Southern Rhodesia with its comparatively large European
population. Not only is there conclusive evidence with regard to this form of trans-
mission in the territory, but fresh instances demonstrating its practical importance
occur almost every season. [If all cattle owners within measurable distance of the
fly areas were aware of the danger of allowing cattle which have been exposed to
fly to mingle with healthy cattle, it is apparent that the losses from trypanosomiasis
in the territory would be materially reduced. Unfortunately, transport riders and
others who are most likely to penetrate fly country with their waggons are not easily
_ reached through the usual mediums of publication.

It may be pointed out that considerable confusion concerning the occurrence of
tsetse In various parts, particularly in the Hartley district, has been occasioned by
the spread of “ fly disease’ amongst animals which have never been in a known
fly belt, and that in drawing conclusions as to the former extent and abundance
of tsetse in this district the possible réle of other vectors must not be ignored.

Relation between G. morsitans and Big Game.

Whilst it is obvious that there are still considerable grounds for a contrary
opinion, accumulated evidence in Southern Rhodesia constitutes a case for a vital
association between G. morsitans and the larger fauna of the forest, which it is
impossible to ignore. As already stated, it is possible that baboons play a more
important part in some areas than was previously recognised.

The weakness of the contrary theory appears to the writer to lie in the fact that no
alternative explanation has been forthcoming of the extremely striking phenomena
that support the assumption of a vital association between the two forms of life,
at least in South Africa. It is not the writer’s intention to detail the arguments in
favour of the theory, as these have already been dealt with in a previous paper,
but it may be as well to consider such facts as have been made public which weigh
on the other side. That there are difficulties in accommodating all the known facts
is freely admitted, but these, surely, sink into insignificance in comparison with the
difficulties of explaining a vast amount of evidence on any other basis. It is possibly
true that G. morsitans is not everywhere dependent upon big game, in the restricted
sense, but that the fly is and has been dependent upon the larger mammals through-
out the greater portion of its range in South Africa is more than probable.

(C572) F

a2 RUPERT W. JACK.

The main arguments to the contrary are apparently as follows :—

(1) The occurrence of fly in numbers in certain tracts where large game is
apparently absent. |

(2) The fact that a small proportion of the game survived the rinderpest in 1896,
so that the fly could not have been absolutely starved.

(3) The discovery of nucleated corpuscles in the stomachs of captured flies, imdi-
cating an avian or reptilian diet.

(4) The fact that fluctuations in the distribution of morsitans have been noted
apart from any wholesale destruction of game.

(1) Observations of this nature on excellent authority have been published
relative to Northern Rhodesia, Nyasaland, East Africa and elsewhere, and it would
be idle to dispute that such observations carry very considerable weight. On the
other hand, in regard to Northern Rhodesia Lloyd, whilst noting that game was not
always abundant where tsetse were very much in evidence, nevertheless found
excellent cause to conclude that large game is at least highly important to the
welfare of the fly.* Again in Nyasaland, whereas certain casual observers have
disassociated fly and game, Lamborn, specially engaged in tsetse-fly investigations,
came to the opposite conclusion. The well-known statement of Sir F. J. Jackson
relative to a dry stretch of country lying between the Msologeni and Tsavo Rivers in
British Hast Africa in 1892 was apparently due to a single and presumably rapid
journey in the driest part of the season. It appears that the investigators who are most
inclined to associate morsitans and the larger mammals are those who have continued
their investigations over a considerable period in one territory, and have also paid
special attention to the bionomics of the fly. The bulk of the evidence to the contrary
has not emanated from entomologists, but from the comparatively superficial and
disconnected observations of men of other professions.

With reference to Lloyd’s observations on the influence of the prevalence of game on
the ratio of the sexes caught, the following may be of interest :

November 1910, Gorai River, Zambesi Valley, Lomagundi district. Game, except
warthog and duiker, practically wanting ; 37 males and 45 female flies caught.

December 1910, Umniati River. Game and baboons plentiful; 57 males and 20
females caught.

April 1911, Gorai River, Zambesi Valley, Lomagundi. Game moderately plentiful ;
53 males and 3 females.

August 1911, Manzituba, Sebungwe district. Game abundant; 143 males and
74. females.

October 1911, Umniati River. Game present but not very abundant, baboons
abundant ; 85 males and 50 females.

June 1914, Thirty-one flies sent in by hunter, probably taken on Umniati River
where game was by this time scarce, although baboons abounded ; 25 males and 6
females.

November 1914, Umniati River. Game scarce but baboons abundant ; 100 males
and 28 females.

August 1916, Sipani Vlei, Sebungwe district. Game abundant; 32 males and
14 females.

*Bull. Ent. Res. v, p. 60. and vii, p. 67-71.

TSETSE FLY IN SOUTHERN RHODESIA, 1918. 83

These figures are more or less in accordance with Lloyd’s results, and they also
suggest that baboons may constitute an efficient substitute for game as a food-
supply for tsetse. It should be noted that all the collections, except that near the
Gorai river in April 1911, were made during the season when the fly is concentrated,
and that the collection referred to is the only one showing extreme disparity in the
proportion of the sexes. During the period of concentration there is no scope for
the females to separate from the males, whilst during the remainder of the year
they have a considerable extent of forest at their disposal. This fact may have
some bearing on the result.

(2) The fact that a material percentage of the game survived the rinderpest is
an obstacle in the way of accepting the theory that the reduction of the game at this
period was the sole cause of the immense reduction of the fly. We are, however,
in the same difficulty with regard to the disappearance of fly from many belts in
South Africa, following the wholesale destruction of game by hunters, for in neither
case was the removal of the animals complete. It is in this connection that further
evidence as to the question whether the hot months preceding the rains are of
maximum importance to the fly or not is required. Supposing this to be the case, the
change in the habits of game consequent on persecution might have a very consider-
able effect, The most favoured haunts of the fly during the season of concentration
are the edges of vleis, and under natural conditions this places them in a position
to secure more or less regular meals ; for, as is known to all hunters in Africa, the
game comes regularly to graze on the green grass in these vleis in the afternoon during
the hot dry months and must necessarily pass through the haunts of the fly. Under
persecution game develops a habit of visiting the vleis at night and leaving at dawn,
so that the fly does not get the same opportunities for feeding. It is conceivable
that the great losses amongst the game during the rinderpest may have produced a
similar shyness, which, supplementing the undeniably immense reduction in numbers
of the animals, might account for a practical cessation of breeding, during what may
be the most important period. During the wet season the presence of only a fraction
of the usual food-supply scattered throughout the forest would in any case have a
very deleterious effect.

Some of the more vigorous opponents of the game and fly theory appear to think
it necessary to suppose that every fly was starved at the time of the rinderpest if the
lack of food is to account for the fly’s disappearance, but this is obviously quite unneces-
sary. A practical cessation of breeding would produce the same result and this might
conceivably be brought about by a greatly reduced and uncertain food-supply. We
know little enough concerning the factors that induce the fly to breed, but we do
know that regular meals are necessary to the insect in confinement, and it seems
almost probable that a state of semi-starvation in nature would either cause the
insects to cease breeding, or that pregnant females, with growing larvae draining
their vitality, would tend to die off or abort in such circumstances. Unfortunately
it is impossible to verify such facts under artificial conditions.

The most striking point in connection with the controversy is, however, that not
a single alternative suggestion of any weight has been made. It has been vaguely
suggested that climatic conditions may have been accountable or that the fact of

84. RUPERT W. JACK.

feeding on animals suffering from rinderpest may have actually killed the flies.
With respect to the latter, observations might have been made in Kast Africa durmg
recent years, but it is on the face of it an almost incredible hypothesis. Moreover,
if every fly was killed in this way does it not imply a very close association between
animals subject to rinderpest and tsetse-fly ? With respect to the climate the writer
has been at some pains to obtain particulars of rainfall and temperature in Southern
Rhodesia at the period involved and there is nothing whatever to suggest any remark-
able deviation from the normal. The following returns were furnished through the
courtesy of the Rev. E. Goetz, 8.J., M.A., F.R.A.S., of Bulawayo.

A set of observations is on record from the Zambesi from 1891 to 1897, but some
of these were taken at Baroma and some at Zumbo, a complete set for either station
being lacking. These two stations are, however, only about fifty miles apart and the
climates are more or less similar. It may be mentioned that fly apparently
disappeared from this vicinity in the 1896-7 season and has not since reappeared.
The records for three seasons ending 1896-7 are given on the opposite page.

If climatic conditions are to account for the widespread dying out of tsetse-fly at
this period, one would expect either an intense general drought, an excessively pro-
longed period of dull weather and rain, an excessively high maximum or an excessively
low minimum temperature, but there is no record of any such occurrence. As a matter
of fact the seasons of 1889-90 and 1890-1 showed far heavier rainfall, as also did last
season, 1917-18. Dull weather was so prolonged last season that the maize crop
was largely a failure and no effect is apparent on the tsetse-fly. Seasons of much
lower rainfall than 1896-7 had also no effect on the pest, and, as may be seen from the
record, the maximum and minimum temperatures recorded did not actually vary
as widely as those of the preceding years.

These figures would seem to dispose effectively of the theory concerning exceptional
meteorological conditions, and at present we are altogether without any plausible
explanation to account for the immense reduction of fly at that period except the
slightly antecedent immense reduction of wild ungulates.

(3) The presence of nucleated corpuscles in the stomach of G. morsitans has
been recorded by several investigators, but there is apparently only one record of
any considerable percentage of flies showing this evidence of a non-mammalian
diet,* and the circumstances may have been exceptional. There is no doubt that a
hungry morsitans is quite willmg to feed upon any warm-blooded creature, and
possibly any vertebrate, that comes within its notice, and in the case of certain large
birds, the operation should not be difficult. Amongst those which no doubt con-
tribute an occasional meal to the fly may be included the ostrich, cranes, herons,
ground hornbills, vultures and various birds of prey, including eagles, owls, etc.,
and doubtless several other types. Such birds are obviously far too scarce or come
far too rarely within range of the fly, to constitute an efficient substitute for the usual
mammalian diet. That the vast majority of birds with their active dis-
positions and untiring pertinacity in pecking at msects serve to support morsitans
is extremely unlikely, and there appears to be no evidence at all to suggest such a
possibility.

*(Lloyd, Bull. Ent. Res. iii, p. 236.—Of 52 flies containing blood, the nature of the

blood cells could be recognised in only 20; of these, 5 contained non-mammalian
nucleated cells.—Ep.]

TSETSE FLY IN SOUTHERN RHODESIA, 1918.

L90GG F0:0 1¥-0 SP-C 6L-T LO+L vP-P G0-¢ 66- TP-0 69-0 00-0 00-0 66-868T
G0-L6 F0-0 1€-9 LV-0 69-7 86:0 GL-L G8-L 66:9 00-0 00-0 00-0 00:0 "86-L68T
L6-LG €9-0 00-0 00-0 GEG 8é-7 PL-8 GLP G8-1 (80-1) | (0-1) 00-0 00-0 °L6—-968T
(€6-93) | 00-0 00-0 (II-1) | (6T-§) €9-& 9F-€ 19-01 PL-P 06-0 00:0 00-0 00-0 ‘96-G68T
VL:PG 00-0 00-0 €¢-0 10-% 93-6 €9-¢ 98:6 v9-1 66-0 00-0 06-0 00-0 “S6-F68T
[ej07, | “oun “ACT qudy | ‘yorvyy Areniqogq) rene e|1,qur,oo(q|",qut.AON|*10q0990 |1,U1,4deg) ysnsny “ANE "It9 XK
‘punjyaqnio py ‘umjyunofadopy iv pjnfumy

16-L8 | 0-0 90-0 L6-T PE-9 61-81 PLT 18-9 | GL-& LL-0 90:0 16-0 00-0 "66-868T
8P-LE 00-0 LG-0 9L-T VES 86: CLP 98-6 LL-€ FO-T vL-0 00-0 00-0 "86—-L68T
90-66 a es vE-1 6L:P LO-L SL: ¥8- GE-9 10-0 i cs = *L6-9681
¥6-7E oT = 18:0 F6-1 66-9 69:8 88-61 S6-T 8P-0 a ed a "96-S68T
[evo], | ‘oung | “LOW judy | ‘yorey Arenirqo | Arenue pl )t.qur 00)". qur,A0 N/19q0490 ‘I,UL.jdeg) ysnsny | “Ame "189

‘punjnuoysnyy ‘hingsyvg yo sayour ur yofuny

JACK.

RUPERT W.

“‘NWOLU “G-FGES[T wosvay

9G 0 oe i) 68 G¢.é1 8o-FL v8 G:VG P-8 0 0 *(uUn) [pesUTe y
8g 9¢ = 89 8 ae OL oP cy oP 6g PQ | (UveTT) “UNA PW
9.F1 eel = S16 G.0Z 6-33 6-43 0-93 P&S 1-33 €-03 o-91 "((O) “UL “8q Vv
G.8Z ece = 6.98 G88 3-9E L-FE TL G.Gg €-€8 8-83 b-9G ('O) XB ‘sqVv
= = = -s = = 6-63 P-1E F-08 6-L3 9-63 P13 ‘(‘Q) ‘dure, weep
‘oun “LEW Tudy | ‘qoreji ‘qaq |Arenuve) ‘seq “AON '10q0300 | ‘gdog |‘gsnsony | ‘Ane
‘puMuolwg ‘1-968. wosnag
= Oa eG 6F | OL-IZ | 66-29 | FO-EE o¢ 0 0 0 0 | “(unu) [TeFALe A.
= 69 09 69 PL Lal Lh Ph 9€ ug oF g¢ | “((uvey{) “WN “ou
<< 0-61 b.0G 38S 3 b-0G Gg. 1-33 DLT PLT 8-1 9-61 CO) “WHA ‘SQV
ars G:G& 0-VE 6-VE 6-€& €-P& G-GE v-OF 0-17 L-L& G-6E L-G€ “(O) “XBW “84V
a 6-83 G.LG 8-L3 6-93 F-96 9.9% 8-18 €-08 §-LG SSS 8-83 ‘(‘O) ‘duo, weep
0 “LOIN | ‘Tudy | ‘qor1eyl ‘qo,y |'Arenuve) ‘ood “AON |'10q0300 | ‘gdeg | ysnsny| ‘Ane
‘OQUNT “9-GGES[ wWosnay
6-91 6-06 OT-LG 6-98 9-64 6-89 0 0-1 0:6 Or *(UnUd) T[VyUre yY
G9 GL +8 89 29 og ig og 8¢ 19 | ‘(WeoyT) “Wn “Tou
T-61 L-16 6:06 G.GG G.1G 9-€6 L-E6 ¢.81 L-9T g.€1 "(O) “UTA “SQV
€-GE V-GE L-€& €-9é G-GE 0-66 G-I[P 0-€& G6 L-86 ‘((O) “XBW 8qV
0.96 6-96 G96 L-LG V-86 v-0€ 6-18 0-96 6-86 1-16 ‘((O) ‘dwoy, ‘wees
oun ‘KEW ‘dy | ‘yoreyy qo,q | Arenuep]) ‘o0q "AON |'10qo109 | “deg | 4ysnsny| “Ame

TSETSE FLY IN SOUTHERN RHODESIA, 1918. 87

With regard to the possibilities of an extensive reptilian diet, the larger reptiles
are very scarce in most morsitans areas, and those that depend upon water are fre-
quently altogether absent. A reptilian diet in many areas would necessarily be
confined to small active lizards and chameleons. There is no evidence to suggest
that the former would allow themselves to be fed upon to any extent, but chameleons
are no doubt an occasional source of sustenance. Where the fly occurs along the
banks of a fairly large river, however, there would appear to be no obvious reason
why the same food-supply as that which serves palpalis should not suffice. G. mor-
sitans, however, not only shows no predilection for such situations, but appears
largely to avoid them. In the portion of the Umniati area already referred to, fly is
certainly to be met with on the banks of the river itself, but is far more abundant in
the adjacent forest. Its distribution therefore does not suggest a dependence on
water-loving reptiles or amphibia. The pest actually lives in the old game haunts,
now greatly depleted, but its habitat is constantly crossed by large troops of baboons
moving to and from the river, and the paths of these animals are quite a conspicuous
feature in this locality. Whatever diet is serving the fly along this stretch, however,
it does not appear to produce the same increase in the pest as an abundance of game,
as is shown by the lack of tendency to expansion since the reduction of the latter.

(4) The fact that fluctuations in the distribution of morsitans occurred previous
to any wholesale reduction of game has really little bearing on the controversy, as
nothing comparable with the phenomena which accompanied game reduction
either locally or over a vast area is on record. It would be curious if tsetse-fly were
not subject, at least in some degree, to similar influences to those which induce
enormous fluctuations in the occurrence of many other insects, such as excessive
multiplication of parasites and other enemies in a given area. Glossina, on account
of its slow rate of reproduction, has less ability for recovery than the vast majority
of other insects and so must be more liable to die out entirely under persecution.

To sum up, the case for the vital association of G. morsitans with the larger mammals
appears to the writer to be extremely strong, in South Africa at least, and the evi-
dence brought forward in opposition comparatively insignificant. There appears
to be no direct evidence at all that morsitans feeds to any great extent on the smaller
mammals under natural conditions, and the failure to find trypanosomes in the blood
of such mammals by Montgomery and Kinghorn, and the Royal Commission in
Nyasaland, constitutes distinct evidence to the contrary. Whatever may be the truth
concerning the vital dependence of morsitans on the larger mammals in all areas,
there is every reason to believe that the pest feeds upon them by preference and
thrives best in their presence, and apart from any other consideration this constitutes
a powertul argument for the suppression of the larger game animals, which are, in
any case, bad neighbours to farmers and stockowners. The writer yields to none
in his love of wild life, but when it comes to a choice between developing colonial

territory economically or maintaining it as a zoological garden, there can only be one
decision.

Explanation of Fly-belts.

Austen on page 4 of his “ Handbook of the Tsetse Fles” sums up the
question of fly belts as follows:—“ We are still somewhat in the dark as to

88 RUPERT W. JACK.

the factors that determine the limits of these ‘ belts,’ but, although tsetse are
undoubtedly dependent upon the blood of vertebrates for their continued existence,
all recent evidence goes to show that the most important element is the physical
character of the locality, and probably its suitability as a breeding ground.” These
remarks, of course, refer to tsetse-flies generally, and not morsitans in. particular, .
but they leave out of consideration one rather obvious fact, namely, that the limits
(of morsitans at least) are not necessarily permanent. Uninfested country adjacent
to a fly area may be just as, or even more, favourable to the pest than the country
it inhabits, the fact bemg demonstrated beyond question by its becoming heavily
infested later. This has been occurring constantly in Southern Rhodesia since the
rinderpest, the limit of the Sebungwe belt in certain parts, for instance, year by year
traversing country obviously favourable on both sides of the line. The character
of the locality, probably intimately combined with the incidence of a suitable food-
supply, appears to account for the permanent limits of a belt ; but what determines
the transitory limits? Why is it that morsitans does not spread thinly over the
whole of its potenial area instead of inhabiting only a part? A fly belt has, in fact,
a coherent quality of its own which needs explaining. Except during the season of
concentration from July to November or later, the insect is as a rule by no means
confined by physical conditions, as is shown by the fact that it more or less deserts
its dry season haunts and spreads for some miles through the surrounding forest.
As far as the writer has been able to ascertain its dispersing range does not exceed
two or three miles, although there would seem to be no obvious reason why it should
not be considerably greater. Again, the males are undoubtedly carried in all directions
up to ten miles, or possibly more, by the movements of animals and human beings,
but as Lamborn has shown in Nyasaland, they have both the mstinct and the power
to return to the point whence they were carried. The females seem only to seek
animals and human beings for the purpose of feeding, and they are probably rarely
carried any considerable distance. A strong ‘‘ homing instinct ” has been proved
in respect of the males, and is to be inferred, at least over a comparatively short
distance, in respect of the females.

The instinct of a fly to return to its haunts if it strays or is carried into unfavourable
country, and the instinct of the males to return to the common haunt, where the
females are to be found, are quite comprehensible ; but they do not explam why
favourable country adjacent to fly areas is frequently free of the pest. The fact of
the matter seems to be that the fly is absolutely unadventurous, if the term be
permissible, not ranging very far from its provedly favourable dry season haunts,
nor allowing itself to be carried beyond the range of its power to return. It appears
to the writer that this is to be explained, not by the lack of any tendency to
wander, but by the possession of a definite instinct to avoid wandering ; as it is to be
supposed that in the absence of such a controlling instinct the flies under the stimulus
of hunger would tend to range far and wide in search of food, as they well might do
without encountering unfavourable conditions during nearly eight months of the year.

Assuming the existence of a strong instinct to avoid undue wandering, an explana-
tion of its development on the basis of benefit to the species must surely be possible.
One of the results is a more or less gregarious habit, but it is difficult to perceive what
benefit the fly derives from this. Other things being favourable a single fly. is

TSETE FLY IN SOUTHERN RHODESIA. 89

undoubtedly quite capable of looking after itself. The writer formerly inclined to the
opinion that owing to the slow rate of breeding a gregarious habit, in enabling the
sexes to meet and mate with the maximum certainty and regularity, might be of
considerable value ; but this theory would seem to be untenable in the light of the
fact that there is no evidence to show that separate matings are necessary for each
act of reproduction, the probability being that one act of copulation suffices for the
female’s life-time. The following suggestion is put forward for consideration,
namely the value of parental experience. If there is one thing proved in connection
with tsetse-fly it is its lack of adaptability. Under favourable conditions of forest
and food-supply it frequently occurs in great numbers, but it is altogether absent
from large tracts of country not affording its very particular requirements as to con-
ditions, and tends to vanish entirely from its habitat if conditions be modified to its
disadvantage. Given a wandering habit without a “ homing instinct ” a considerable
proportion of the flies would tend to stray into unfavourable country and perish,
and it appears conceivable that this fact may have led to the development of a strong
instinct against wandering, which functions whether the surrounding country be
favourable or not. In the vast majority of cases the fact of a fly being born at any
particular spot proves the suitability of the locality to the species, and, as in the
case of a certain conservative type of human being, what is good enough for the
parents may be judged good enough for the off-spring.

The Manner in which Fly-Belts extend.

This is a matter which has received some attention in this territory for a
number of years past. The following rules are im accordance with the evidence
accumulated, but must be regarded at present as merely tentative :—

(1) The advance does not take the form of disconnected offshoots at a distance
from the main belt, but the whole movement is analogous to that of a rising flood,
flowing along favourable channels and gradually extending the flooded area ;
(2) when the line of advance is interrupted in regard to permanent shade, as in
crossing a watershed, the movement is confined to the wet season; (3) large
numbers of flies are necessary at the previous dry season limit before a moderately
wide region affording only summer shade can be crossed; (4) the fly can only
spread in the presence of considerable quantities of the larger mammals.

The observations leading to these conclusions have been confined to Southern
Rhodesia, but with the exception of the last they appear to be due to the same cause,
which is independent of locality, namely, that the flies tend to remain in or return to
one particular spot. The recently observed facts in this territory are altogether
opposed to the theory that fly moves into and establishes itself in new districts with
game, as has been stated by hunters and explorers in the past. Newly infested spots
may be disconnected by a few miles from the former limits of the belt during the
season of concentration, but the belt is continuous during the season of dispersion.

The fact of fly bemg only able to advance from one dry season haunt to another
in the wet season is easily understood on the basis of the seasonal dispersion of the
insect, which enables larvae to be deposited in the intervening forest until the neigh-
bourhood of the next river, or other suitable dry season haunt is reached. When
the next season of concentration arrives, the pest presumably concentrates in both

(C572) rel

90 RUPERT W. JACK.

directions, a portion of the flies going back to the previous dry season haunt and a
portion to the new, probably whichever is the nearer. The term “ wet season ”’ is
not used loosely in this connection, notwithstanding the fact that the forest retains its
foliage for some months after the cessation of the rains. The wet season lasts until
April, and the three months following are, with the occasional exception of August,
the coldest in the year, when the breeding rate appears to be at its lowest ebb and the ~
tendency to extension is apparently checked by this fact. It is to be remarked that
whereas advance along a well shaded river may be a more or less continuous or re-
gular process, advance across interfluvial areas tends to be irregularly spasmodic.
One dry season haunt having been attained, a pause of several years sometimes inter-
venes before the next step, and observations indicate that the next step, if it be a
fairly long one, say five or six miles, is only possible when the fly has become really
abundant at its previous halting place. As an instance, the passage from the Mzola
to the Kana River in the Sebungwe district, a distance of about six miles, took
place in the wet season of 1915-1916, but the fly was present on the Mzola in small
numbers in 1912-13. Once the pest obtained a footing on the Kana, it increased
rapidly in the presence of large quantities of game and late in 1916 had become fairly
abundant, showing that it was no lack of favourable conditions ahead that caused
the delay. The necessity for great numbers would appear to be explicable as fol-
lows :—At the beginning of the rains the flies scatter at once for a short distance in
all directions from their narrow dry season haunts. A certain proportion:of the
females deposit larvae near the limit of their spreading range towards the new objec-
tive. Flies emerging from these also scatter in all directions, a proportion of the
females again larvipositing in the required direction. The repetition of this process
may result in the neighbourhood of the next river or vlei being reached before the
next season of concentration ; but it is clear that, the rate of reproduction being so
low, only a small proportion of the numbers that originally scattered would extend
so far in a few months, hence the necessity of large numbers to start with.

These remarks may seem somewhat academical, but they have a distinctly practical
side. If the writer’s deductions are correct, it would appear that the clearing of a
comparatively narrow strip of forest might serve to check the advance of the fly.
notwithstanding the fact that the males would constantly be carried across to the
other side. The cost and maintenance of such a clearing would be prohibitive on
a great scale, but by taking advantage of local conditions it might prove practicable
to check the advance of a belt in some particular direction. An experiment of this
nature was actually commenced at the south-western extremity of the Sebungwe belt
in October 1918, but has had temporarily to be abandoned owing to the influenza
epidemic.

Bure EN RESEARCHER EVOL. 2G RART Ie reAa ele

Fig. 1. Gusu forest in full leaf near Gwaai River,
Southern Rhodesia, October 1918.

Fig. 2, Leafless Gusu forest near Sipane Vlei,
Southern Rhodesia, August 1916.

lie

Rea oe ee

ee re er

SW, JENA, IRESBARGHE Wl, XX. JPA i, Ieee JIU.

Fig 1. Mopani forest in full leaf in the Zambesi Valley, April 1914.

Fig. 2. Mopani forest in leafless condition, Wankie’s District,
near Gwaai River, October 1918.

>
Aden’

ied

Weer ore

=f ae

Whee ake A005)

BuLvt. Ent. RESEARCH. VoL. X. Part lI. Peas IW,

Fig. 1. Large numbers of pups of Glossina morsitans were found in a hollow
in the trunk of this tree; Sipane Vlei, Southern Rhodesia, November 1914.

Fig. 2. Enlarged view of the hollow in which the G/ossina morsitans
pupee were found.

“7
=
es ; ; E
. Ze "7
e
G >
=
R me
> =
4 ¥
:
J
ae
= =
[pa 1
Gore a
: Pi = a
HA
ah
= Es 5 ee rt
3 ys
j
1
ri 3
K
=
a ’
Ea
a
i
j t
r
J
:
te hs ;
e
\ E
J =
a t
e
5 us
i ‘
oo
a =}
Eee = iy
; pit ah

- a
98) ap en

SP ae

i os |

ier
i. i es aga es 5

COMPILED & CRAWN IN TH
S8Auw

ane a
a
eh

Vy ‘
hr

t
pid |

Butt. Ent. Research. Vo. X., Part 1. Map I.

Part inhabited by tsetse immedietely after the rinderpest.
Limits of tsetse in 1904...
Limits of tsetse in 1907._..__

Limits of tsetse in 1910, _
Limits of tsetse in 1913.
Limits of tsetse in 1918...

NORTHERN

RHODESIA

Bale & Danielsson,

Map showing the increase in recent years of GLOSSINA MORSITANS in the Sebungwe District.
Southern Rhodesia.

Mae ax Wa ty
ta ine ow hinge 2

: Laden hs

BULL. ENT. MESEAnKN

INVYANGA

{INYANGA

COMPILED & ORAWN IN TH
SAL!

Burt. Ent. Researcn. Vor. X.. Parr |,

MAP OF

SOUTMERN RHODESIA

ISIS

Showing belts of Glossina morsitans.

NORTHERN] RHODESIA

CTDRIA FAL

UMVU AS,

ILI aNzif

COMPILED & DRAWN INTHE SURVEYOM GENERALS OFFICE
BALIEBURY RHODESIA

feted il; ee:
Pia Te cy ae ng i ia tg

few! ’
‘ ) e > ?
o Pe iar " -
2 Lite
— 4
, “
‘ / .
Fi ae ri
Le —
r ;
AD, -
> a Sa
> y ,
wx) - > I
y Fut f i
Vy. 2m
A és Yim
on J
‘ Y.
1 m = ji oF
+= ae ie - *

4 ;
y r : 3
7 ’
. ‘ >
4 t
> ‘ :
5 . me ‘
4 i f M $ 14 Fi ; ey
: 4 sex : ‘ 4 Bs ; :
= ? ~ > 7 5 P *
: Be ¢ \ a > -s o “a an ay www dp vane a <P © beet 44 gee ~
= ¥ a yo e i * 5 < P
» ae te 7 a ; * , a
‘ te rs * “yak of =
' 4 ee es POEL sy | 5 “eA ; Sn ak
P r ’ see aul " . 3 v E cee ar
Pee pe baw bial ry i 9 fr} y er pie ae _
3 + i 4 as 3% i .
¢* — q tan we vy 3 es rey
— y , . i 7 Ma rer > :
ae € "
ay . 5 i . . « ‘
: ie ' ¥ 7 ae
ee ' a ‘
vin ?
< 3 f
eam eee a Mw
fam ’

se ey ee ee

; ;
|

.
.
~

ie >
aS tog iets pepo
i

Po ‘

Sa teidap hr Eh 5 sire a cl

oe eileen
r
{

Silda ih apes on i i | = re se) eT
* 2 7 “
cap eae ih ; i
we : : kK: feat s he
5 oe eer ens .
ie Satie wh ;
4 aac vi ras
ns * tk
een : fs
. iS 4 of
' *«
. it, ——_ wy sy Aas etx be he

. de Pa Tel ae
Ce ets Be el ee

Aven aU
Butt. Ent. RESEA

- ee : ri ‘ i ; ;
i i) i ae :
' )
ie
}
. Ms
i
2 }
iy
% j
Vg, :
tt
4
Ce aoe a
a
if
}
i
i }
j Sy
fl kK
AS
\
i
{
i)
1

: fs
Ji fall

‘ J

Yi 1

' , on

ue
ee me t Ny 7 , i

Butt, Ent. RESEARCH. VoL. X., Part I.

Map III.

Dreadnought

Scale of Miles.
2 & G G&G

ee ee eee eeeerionn

Gatooma ¥

Forest more or less untouched in 19138.

ail)

Forest cleared by wood contractors between 1913 and 1915.
Forest being cut at the end of 1915.

THE SURI-SURI FLY BELT IN THE HARTLEY DISTRICT, SOUTHERN RHODESIA.

aie echaptaee eae

es ees

Mees seein seg oem deren
eR ee

ands 8 a ¢
at pe

Fi ty

ON THE OCCURRENCE OF STEGOMYIA FASCIATA IN A HOLE IN A
BEECH TREE IN EPPING FOREST.

By Captain M. E. MacGrecor, R.A.M.C.,
Officer in charge of the Entomological Laboratory, Sandwich, Kent.

For the last few months I have been using a beech tree-hole in Epping Forest
as a source of supply of Anopheles plumbeus ; and it has been our custom to collect
larvae from the hole and transfer them to the laboratory at Sandwich, where the
development is continued under artificial conditions. In this way we have been
able to obtain large numbers of Anopheles plumbeus and Ochlerotatus geniculatus,
together with the recently discovered Orthopodomyia albionensis * as an associate.
Not long after finding the Orthopodomyia, I was surprised to find yet another species
from the same tree-hole, two male Stegomyia fasciata emerging from the tank
containing the mixed larvae from Epping Forest. The specimens were of normal
size, and we now have them preserved in our collection here.

As is well known, Stegomyia fasciata has been brought to England for experimental
purposes, and has been kept breeding in captivity by several workers for some
years. It is therefore probably accounted for in nature in England by stray
individuals which have accidentally escaped from the laboratories. It is, however,
very surprising that the first recorded specimens to be found under natural conditions
in England should have occurred in a forest to a large extent removed from human
habitation, since Stegomyza fasciata abroad is essentially a domestic mosquito. The
insect may therefore be indigenous.

The larvae were collected, and bred out in the laboratory, under my personal
supervision, so that I can vouch for the absence of any error in the observation.

* Journal of Royal Army Medical Corps, Nov., 1919.

rae ee ||

i ,' Wise erties v

COLLECTIONS RECEIVED.

The following collections were received by the Imperial Bureau of Entomology
between Ist January and 3lst March, 1919, and the thanks of the Managing Com-
mittee are tendered to the contributors for their kind assistance :—

Dr. W. M. Aders, Government Economic Biologist :—-23 Diptera and 2 Lepidoptera ;
from Zanzibar.

Mr. T. J. Anderson :—9 Diptera, 137 Chalcids, 267 other Hymenoptera, 3 Lepidop-
tera, 27 Coleoptera, a number of Coccidae, and 12 other Rhynchota ; from British
Kast Africa.

Mr. B. N. Bandyopadhyay :—11 Culicidae, 1 Tabanid, and 1 Glossina ; from
British Hast Africa.

Lieut. P. J. Barraud :—315 Culicidae, 131 Culicid larvae, 17 Culicid preparations,
44 other Diptera, 36 Slides of Siphonaptera, 5 Hymenoptera, 11 Coleoptera, 2 Coleop-
terous larvae, 29 Lepidoptera, 10 Orthoptera, 26 Odonata, 5 Odonate nymphs, and
2 Mites ; from Mesopotamia.

Mr. G. KE. Bodkin, Government Economic Biologist :—3 Diptera, 49 Hymenoptera,
18 Coleoptera, about 30 Mallophaga, 2 species of Coccidae, 14 other Rhynchota,
8 Orthoptera, 14 Odonata, and 3 Ticks ; from British Guiana.

Capt. P. A. Buxton :—1 Simulium, 14 Tabanidae, 29 other Diptera, 9 Coleoptera,
and 23 Rhynchota ; from Mesopotamia.

Dr. G. D. H. Carpenter :—9 Diptera, 1 Chalcid, 64 other Hymenoptera, a number
of Coleoptera, 2 Planipennia, 2 Lepidoptera, 4 Cimicid bugs, 34 other Rhynchota,
and 26 Orthoptera ; from Portuguese East Africa.

Mr. W. C. Crawley :—81 Coleoptera ; from Hampshire.

Mr. J. Sydney Dash :—5 Lepidoptera, 1 Hymenopteron, 31 Coleontees and.
6 Rhynchota ; from Guadeloupe.

Mr. D. d’Emmerez de Charmoy, Government Entomologist :—17 Opiliones ;
from Mauritius.

Mr. P. R. Dupont, Curator of the Botanic Station :—12 species of Coccidae and
a number of fungi; from Seychelles.

Mr. C. C. Gowdey, Government Entomologist :—2 Tabanus, 4 Hippoboscidae,
37 other Diptera, 10 Chalcids bred from Coccidae, 205 Hymenoptera, 34 Coleoptera,
6 Mallophaga, about 100 Aphids, 18 species of Coccidae, 30 other Rhynchota, 3
Orthoptera, 3 Ticks, and 1 Scorpion ; from Uganda.

Mr. H. Hargreaves :—212 Culicidae, 1 Haematopota, 47 other Diptera, 2 Hymenop-
tera, 19 Coleoptera, about 150 Anoplura, 27 Rhynchota, and 2 Orthoptera; from
Taranto,. Italy.

Mr. J. C. Hutson :—86 Coleoptera ; from Ceylon.

Imperial Department of Agriculture :—9 Hymenoptera, and 7 Coleoptera ; from
the West Indies.

Mr. Nigel K. Jardine :—278 Diptera, 38 Lepidoptera, 34 Hymenoptera, 144 Coleop-
tera, 29 Rhynchota, 10 Orthoptera, 9 Odonata, and 2 Arachnida ; from Ceylon.

94. COLLECTIONS RECEIVED. ;

Capt. Malcolm E. MacGregor :—4 Diptera, 3 Hymenoptera, 16 Mallophaga, about
80 Ticks, and a number of intestinal worms ; from Portuguese Hast Africa.

Mr. K. D. Shroff, Assistant Entomologist :—a collection of Crustacea, Mollusca
and worms ; from Burma.

Mr. R. Veitch :—8 Diptera, 16 Hymenoptera, 3 Thysanoptera, 42 Coleoptera,
14 Lepidoptera, 29 Orthoptera, 7 Rhynchota, and 6 Spiders ; from Fiji and Queens-
land.

Capt. Jas. Waterston, R.A.M.C.:—6 Phlebotomus, 13 Siphonaptera, and a large
number of lice; from Macedonia.

Dr. W. G. Watt :—100 Culicidae and about 50 Culicid larvae; from the Gold

Coast.

i
\
&
{
f y
t -
{
. \
z.
: »
re
,
j
%

THE COCCIDAE OF SOUTH AFRICA —IV.*
By Cuas. K. Bratn, M.8c., M.A.,
Division of Entomology, Pretoria, South Africa.
(Puates V—XII.)

CONTENTS.
PAGE.
Genus Chionaspis (continued) .. He be ap - wn . 95
,, Lepidosaphes ai ig ae wig ot ih * a 105
oy . Lschnaspis? 5: sah as ou Ae si ae ist 107
Subfamily ASTRROLECANINAE ie st s ‘ei i. A o4 108
Genus Asterolecanvum , 0) iy cs , a He an 108
», Lecaniodiasprs i i! bt u. iis 4 ay, KK 115
» Cerococcus .. J; sf LM a bi ve ng oe 119
Subfamily TACHARDIINAE .. mM ms ¥, ee ue Mi He 122
Genus Tachardia .. Ws at HG a4 8 a2 ail a 122
Subfamily ? .. i ee a if - " a « ‘2: 126
Genus Halimococcus aN e aT un KS ae i a 126
Baccacoccus .. Bi a a ny ne af $f re 127

147. Chionaspis scutiae, sp. n. (Plate v, fig 176).

Scale of adult 2 about 2 mm. long, moderately broadened behind, convex, usually
curved, white, not glossy, but with =: conspicuous growth-lines; exuviae orange-
brown to dark brown. The colour of the second exuviae is only slightly obscured by
a faint secretionary covering.

Puparium of g comparatively large, white, non-carinate, with yellow or orange
exuviae.

Adult 2, when mounted, regular in outline, about 1°3 mm. long, elongate, widest
behind the middle, hyaline, except the median lobes which are very dense, and the
middle part of the pygidium which is yellowish. The segmentation of the body is
unusually distinct but is not strongly marked at the margins, nor are the abdominal
segments produced.

The pygidial margin is simple. There is only one pair of lobes present, which are
short and broad, densely chitinised, brown, with the rounded outer margin once
notched. Plates are apparently absent, but there are two stout spines on each side.
Pygidium as illustrated (fig. 176). Circumgenital glands in 5 groups:

94
GAO 0219
12299 iid 3299

Habitat: “ Wacht-een-beetje” (Scutia indica, Brohn), eee C.P. ; collected
by C. P. Lounsbury, September 1907.

Collection No.: 182.
*For Part II, see Bull. Ent. Res. ix, p. 107; Part III, Bull. Ent. Res’ ix, p. 197.
(C605) P4/140. 1,000. 1.20. B.& F.Ltd. G.11. A

96 CHAS, K. BRAIN.

148. Chionaspis (Pinnaspis) chionaspitiformis (Newst.) (Plate v, fig. 175).

’Draspis chionasprformis, Newst., Bull. Ent. Res. i, p. 198, 1910.

Hemachionaspis chionaspitiformis, Lindinger, Jahrb. Hamb. Wiss. Anst. xxvii,

Beih, 3,191:

Scale of adult 2 about 2 mm. long, white, appearing very loosely constructed and
soft in texture, moderately broadened behind, with yellowish or brownish exuviae.
The second exuviae are covered with a thin layer of secretion.

Puparium of ¢ elongate, non-carinate, with yellow exuviae.

Some specimens collected by Claude Fuller on twigs of a native tree at Busi,
Portuguese East Africa, appear as matted, dark brown to blackish scales, with brown
exuviae. This, I think, is due to an accumulation of sooty fungus, as the insects
agree in all other respects with those on African mahogany from Rhodesia.

Adult 2 mounted, about 14 mm. long and 0°7 mm. broad at the widest point,
which is situated at a considerable distance beyond the middle. The insect is
narrow, rounded in front, gradually broadening to the free abdominal segments, from
which it suddenly narrows to the hind extremity, which is pointed. The whole body
is hyaline except the mouth-parts and median portion of the pygidium, which are
yellow. The median lobes are densely chitinised and brown. The antennal tubercles
are rather large with one very long flagellum and a few short spurs. Parastigmatic
glands present, 5 to 8 at each anterior spiracle.

The pygidium is characterized by a single pair of lobes, which are broader than long
and have their inner margins in close contact or fused ; their distal end is evenly
rounded and the outer margin deeply notched. From the outer edge of each lobe a
short spine arises. The plates are longer and more curved than in C. cassiae—as
illustrated in fig. 175. Circumgenital glands in 5 groups, as in cassie, or with the
posterior lateral glands slightly more numerous.

The most striking differences between this species and C. cassiae are :—that the
scale is white, without the transverse brownish ridges, and the pygidium appears
more acute, with the plates longer and more curved.

Habitat: On African mahogany, Zomba, Nyasaland ; collected by Ross-Townsend,
December 1908 (Cape Collection 2100). On native tree, Busi, Portuguese Hast Africa ;
collected by C. Fuller, May 1915.

Collection No. : 322, 322a.

149. Ghionaspis simplex, Green, var. (Plate vi, fig. 178).

Chionaspis simplex, Green, Cocc. Ceylon, ii, p. 160, 1899.

Scale of adult ? large, 3 mm. long, and 1°5 mm. broad, broadest about middle,
white, dull, without distinct growth-lines. First exuviae colourless to yellowish ;
second exuviae large, roundly convex, glassy, usually -+ buffin colour, with a depressed
median area bordered on each side by a prominent beaded ridge. In living material
the second exuviae are apparently covered by a thin, dull white layer of secretion,
which is easily flaked off. In the majority of dry specimens this layer and the first
exuviae are missing.

Puparium of ¢ not observed.

Adult 2° contain many well developed embryos with 5-jointed antennae.

THE COCCIDAE OF SOUTH AFRICA. Sif

Adult 9, when mounted, elongate, narrowly rounded in front, broadest immediately
behind the middle, and somewhat broadly pointed behind, hyaline, with the pygidium
slightly denser ; abdominal segments well marked and flatly rounded at the margin.
Pygidium without lobes or plates, coarsely corrugate, with conspicuous spines and
dorsal glands as illustrated (fig. 178). Circumgenital glands in 5 groups, which are
often almost contiguous in a wide bow :

18—25
24—-32 24—32
48—60 48—60

Remarks.—This species is very much like C. semplex Green, and is in all probability
the variety he mentions from Mauritius ; it has but 5 groups of circumgenital glands
instead of 7 as in C. simplex.

- Habitat: On bamboo, Durban, Natal ; collected by A. Kelly, May 1915.

Collection No. : 175.

150. Chionaspis caffra, sp. n. (Plate v, fig. 174).

Scale of adult 9 about 2 mm. long, narrow, but gradually widening to near the
posterior end, where it is flattened and broadly rounded, straight or slightly curved.
Exuviae brown ; second exuviae covered.

G puparium small, white, with orange exuviae ; generally non-carinate, sometimes
faintly tri-carinate.

Adult 9°, mounted, about 1:4 mm. long, narrow in front, then gradually widening
to some distance behind the middle ; anterior part of body and median parts of the
anterior abdominal segments considerably chitinised ; abdominal segments not
rounded, nor produced at the margin. Antennal tubercles small, each with one very
stout spine. L, close together, inner margin slanting slightly outward, crenulate.
L, small, roundly conical. Circumgenital glands in 5 groups:

5—T
8—14 8—l4
19—25 19—25

Habitat : On Acacia sp., probably A. caffra, The Thorns, Pretoria; collected
by C. P. Lounsbury, September 1915.

Collection No.: 150.

151. Chionaspis ambiguus, sp. n. (Plate v, fig. 177).

Scale of adult 9 elongate, mytilaspiform, narrow in front, broadly rounded behind,
usually + curved, arched, 2°3 mm. long, somewhat covered by outer bark tissues.
Colour of scale pale to dark brown, with a greyish surface covering ; exuviae yellow.

¢$ puparium not observed.

Body of adult 9 elongate, broadest just behind middle, anterior end tapering slightly
and broadly rounded in front, posterior end tapering abruptly to the two densely
chitinous median lobes. These, together with the chitinous thickenings extending
into the pygidium, and the mouth-parts, are yellowish ; remainder of body hyaline.
Free abdominal segments (4) slightly rounded, not much produced. Antennal
tubercles moderately large, with two stout spines. Parastigmatic glands 0.

(C605) a2

98 CHAS. K. BRAIN.

Pygidium (fig. 177) with one pair of lobes, which are close together on their inner
surface, their outer edges being crenulate. Between the median lobes there are two
large spines but no plates. At the outer margins there is also a large spine, then a
gland from which arises a small sharply pointed plate, then a very small rudimentary
lobe, which is pointed and chitinous and in some cases seems to be composed of two
lobules of which the outer is the smaller. These are followed by a long simple dagger-
shaped plate, then a large gland opening and another plate. Beyond this the margin
appears thickened lke a minute rudimentary lobe, beyond which there are two
additional pairs of plates. There are a few plates on the margin of the first abdominal
segment from the pygidium in addition to the glands. Extending into the pygidium
are three pairs of chitinous thickenings. Circumgenital glands in 5 groups, of which
the median and anterior laterals are often almost united into a bow.

7—l1
10—14 10—14
6— 9 6—9

Remarks. A large percentage of the @ scales showed the circular exit holes of
some Hymenopterous parasite.

Habitat : On twigs of lilac, Fort Beaufort, C.P.; collected by C.P. Lounsbury,
June 1913.
Collection No. : 263.

152. Chionaspis leucadendri, sp. n. (Plate v, figs. 166, 171).

Scale of adult 9 about 2°6 mm. long, white, smooth, slightly glossy, not very convex,
long and narrow, somewhat widened and flattened behind. Exuviae brownish ;
second exuviae covered by a very thin layer of secretion.

2 scales chiefly on stems ; ¢ puparia most common on leaves.

g puparium comparatively large, white, flat, shghtly roughened, not carinated,
with pale exuviae, which are sometimes + greenish yellow in colour.

Adult 9, mounted, about 1:4 mm. long, narrow in front, but quickly widening
to about middle, whence it gradually narrows again to the median lobes. Abdominal
segments well indicated but not prominently produced. Body moderately chitinised,
the mouth-parts, lobes, and margin of pygidium denser. The four abdominal seg-
ments adjoining the pygidium have their margins with a few gland-pores and a number
of short, stout, very acute spine-glands. These, seen in optical section, appear like
rose-spines, but the points are often long, slender and curved. They occur chiefly
on the margin, but extend in scattered formation some distance within the edge.

P, small, spine-like. lL, rounded, striate, arising from the thickened margin of the
pygidium. P,—6 long, dagger-shaped. Circumgenital glands in 5 groups, which are
almost contiguous, in the form of a horse-shoe :

4—7
11—18 11—18
26—32 26—32

Hatitat : On silver-leaf tree (Leucadendron argenteum), National Botanic Gardens,
Kirstenbosch, Capetown; collected by the late Professor H. H. W. Pearson,
August 1914.

Collection No. : 145.

THE COCCIDAE OF SOUTH AFRICA. 99

153. Chionaspis (Phenacapsis) lounsburyi, Cooley (Plate v, fig. 169).
Chionaspis lounsburyi, Cooley, Can. Ent. xx, p. 87, 1898.
Phenacaspis lounsburyi, Fernald, Catalogue, p. 238, 1903.

Adult 92 scale about 3 mm. long, white, glossy and pearly, sides almost parallel
when at the margin of a leaf, but usually broadened behind the middle when on the
blade ; evenly rounded behind, often with faint transverse ridges. Ventral scale
absent or represented only by the incurved margins of the dorsal scale and an
extremely delicate layer which remains on the leaf. Hxuviae orange-yellow or orange-
brown; first exuviae paler, second covered by a pearly white layer of
secretion.

S puparium about 1 mm. long, similar to that of 2 but smaller, non-carinated.

Body of the adult @ elongate, all colourless and hyaline, except the mouth-parts
and lobes which are faintly yellow. The free abdominal segments are not produced:
at the margin, but are slightly rounded. oe

L, low, broad, and very widely divergent, usuallv appearing as though they pro-
jected from behind the thickened median area of the pygidium. J, composed of two
small lobules which are distinctly separated ; the inner lobule is the larger. Lobes
distinctly striate. Plates simple, thin. |

The antennae are placed a long way back near the mouth-parts and consist of minute
tubercles each with one long thin curved hair. Parastigmatic glands present, usually
3 to 5 at anterior spiracles. Circumgenital glands in 5 groups :

4—T
WI—15 11—15
20—26 20—26
Habitat: On Rhus thunbergi, Ceres, C.P.; collected by C. P. Lounsbury, , 1896.

On a native Proteaceous plant, Wolseley, C.P.; collected by C. P. Lounsbury. June
1915.

Collection Nos. : 166, 170.

153a. Chionaspis (Phenacaspis) lounsburyi ekebergiae, var. n. (Plate v, fig. 168).

Parastigmatic glands present, 5 or 6 glands in a close crescent. Pygidial margin
as figured (fig. 168). Circumgenital glands more numerous, ¢.g.,

6—9
11—20 11—20
29—46 29—46

Plates more distinct and last row of dorsal glands more numerous than in
C’. lounsburyt, usually 9 : 4.

Habitat : On Ekebergia sp., Durban ; collected by C. Fuller, July 1915.

Collection No. : B. 186.

100 CHAS. K. BRAIN.

154. Ghionaspis (Phenacaspis) natalensis, Ckll. (Plate v, fig. 165).

Phenacaspis natalensis, Ckll., Ann. Mag. N.H. (7) ix, p, 25, 1902.

Phenacaspis natalensis, Fernald, Catalogue, p. 238, 1903.

Professor Cockerell’s description is as follows :—

“2 scale white, about 3 mm. long, pyriform ; exuviae pale orange-brown.

“Five groups of circumgenital glands; median of 10, anterior laterals 22-26,
posterior laterals 19-26. Anal and genital apertures opposite. Median lobes large,
widely diverging, broader than long, the long inner margin strongly serrulate ; their
bases well apart, the space occupied by the usual pair of short spines. Second lobe
represented by three elongated and rounded lobules, the first of which is largest and
bears a spine. Third lobe represented by a very long narrow lobule bearing a spine
followed by a broad and much shorter lobule, and then a very broad serrulate lobule,
having its outer slope much the longest. The fourth lobe is represented by a triangu-
lar lobule bearing a spine and two slight swellings of the margin, too slight to be called
lobules.

“ 2 scale feebly tricarinate or barely keeled at ail.”

Habitat: On mango, Durban, Natal ; collected by C. Fuller, 1901. On palm,
Durban ; collected by A. Kelly, July 1915.

Collection Nos.: 171, 171a.

155. Chionaspis (Dinaspis) imbricata, sp.n. (Plate vi, fig. 179).

in material seen by the writer the scales invariably occupy cracks in the bark on the
stems of the host-plant.

Scale of adult 2 small, elongate, almost parallel-sided, white, with orange to brown
eXUV1ae,

Puparium of 3 similar but smaller.

Adult 9, when cleared and mounted, small, averaging 0°75 mm. long and 0:48 mm.
broad. Theanterior endis broadly rounded ; the sides of the body are almost parallel
to a little more than half the length, when they gradually taper to the pointed
pygidium. The anterior half of the body and the pygidial area are more densely
chitinised than are the free abdominal segments and appear yellowish. The chitin on
these parts is thickened in minute ridges, giving the impression of a finger-print, a
character only found in a few Coccids.

It is impossible to describe the pygidial margin as having definite lobes, and plates
are absent. The chitin is here folded so as to form numerous rounded more or less
imbricated prominences, some of which reach, or extend beyond, the margin, which
thus appears festooned. There are a few spines of moderate length at intervals
around the margin. The anal opening is set well back from the margin and around it
are a number of conspicuous thin spots in the chitin, appearing as perforations. In
the second stage 9 there are about three simple plates on each side of the pygidium.

Antennae set well forward, tubercles moderately prominent with one or two curved
setae. Parastigmatic glands 0. Circumgenital glands O.

Habitat : On stems of Huclea natalensis, Point Road, Durban ; collected by C. P. v.
d. Merwe, July 1916.

Collection No. : 157.

THE COCCIDAE OF SOUTH AFRICA. 101

156. Chionaspis (Dinaspis) diosmae, sp.n. (Plate v, fig. 167).
Insects -- clustered on leaves of host-plant, especially on the upper surface.

Adult 9 scale about 2-2 mm. long, comparatively broad, moderately convex, silky
iN appearance, with conspicuous growth lines. First exuviae yellow ; second exuviae
brownish, covered by a dense layer of secretion similar to remainder of scale. The
layer covering the second exuviae is very easily removed and carries the small first
exuviae with it. Owing to this fact many of the specimens appear to have brownish,
shining pellicles.

Adult © viviparous ; mounted specimens contain many well-developed larvae.

Adult 9, mounted, about 1:6 mm. long and 0°8 mm. broad ; widest about middle
and narrowing to each end. Body moderately chitinised. The pygidial margin is
quite different from that of any other species of Dinaspis known to me. The median
notch is angular, with narrow straight thickened margins, below which the median
lobes project. Owing to this fact the appearance of the lobes varies greatly according
to how the specimens are mounted, L, in some cases appearing quite distinctly, in
others being almost invisible. There are apparently no plates. Pygidium as
illustrated (fig. 167). Circumgenital glands O.

Habitat: On buchu (Diosma crenata), Wellington, C. P.; collected by C. P. v. d.
Merwe, November 1904 (Cape No. : 1554).

Collection No. : 147.

157. Ghionaspis (Dinaspis) lounsburyi (Leonardi) (Plate v, fig. 172).

Dinaspis lounsbury2, Leon., Bol. R. Sc. Agr. Portici, p. 216, 1914.

Chonaspis capensis, Newst., Bull. Ent. Res. vii, p. 378, 1917.

Scale of adult 9 about 2mm. long, elongate, generally straight, but varying in shape
according to the position on the plant, condition of crowding, etc. On straight stems
the sides of the scale are + parallel from the end of the second exuviae to beyond the
middle, where it becomes slightly broadened with the hind margin broadly rounded.
Other specimens situated in the angles of thorns, etc., are much shorter and broader,
sometimes almost as broad as long. Colour of scale white ; first exuviae bronze to
bronze-brown ; second exuviae brown, slightly pointed behind.

Puparium of § about 0°8 mm. long, white, non-carinated, with orange pellicles.
Posterior margin of puparium broadly rounded, opening by the upper part splitting
from the lower so that the dorsal flap is exactly like the lower one.

Adult 92 viviparous ; when mounted, elongate. Anterior part suddenly narrowed
and rounded in front, paler yellow ; median portion of body darker yellow, wider,
with almost parallel sides ; posterior extremity broadly rounded. The free abdominal
segments are broadly rounded at the margins. Antennal tubercle small, with two
setae of medium length and thickness. Pygidium as illustrated (fig. 172). Circum-
genital glands O.

Habitat: On stems and leaves of native spiny plant (Gymnospora buaxrfolia) ;
extremely common around Pretoria and also received from Umtali, Rhodesia (sent by
R. Lowe Thompson, of Salisbury).

Collection No.: 251.

102 CHAS. K. BRAIN.

158. Chionaspis (Dinaspis) distincta (Leonardi) (Plate v, fig. 170).

Dinaspis distincta, Leonardi, Bol. Sc. R. Agr. Portici, p. 213, fig. 33, 1914.

Chionaspis distorta, Newst., Bull. Ent. Res. vii, p. 377, 1917.

Scale of adult 2 about 2-2 mm. long, moderately broad, convex and roughened,
dull white or greyish white in colour, with exuviae dark orange-brown; second
exuviae covered. ,

Puparium of ¢ robust, bufi-coloured and non-carinate, with yellowish or orange
exuviae. Very few gj specimens are present in the large amount of material before me
and in a number of cases the § puparium does not lie close to the stem in the usual
manner, but seems to be attached at the anterior end and projects outward between
several female scales.

Adult 9 viviparous ; elongate, narrow in front, gradually broadening to behind
the middle, to which point the body is highly chitinised ; posterior to this it is thin,
hyaline. There are three pairs of lobes, which are more heavily chitinised than the
remainder of the pygidium : L, broad, inner margins converging at the base, faintly
trilobed or broadly crenulate ; median lobe small, L, and L, small, + triangular.
Antennae with one long flagellum and four short stout spurs. Circumgenital glands O.

Habitat : On stems of Protea hirta, Pretoria ; collected by the writer, October 1914.
Professor Newstead’s reference “‘ Windersboom, Transvaal,’ should be Wonderboom,
Pretoria.

Collection No. : 163.

159. Chionaspis (Poliaspis) carissae (Ckll.) (Plate v, fig. 164).
Poliaspis carissae, Ckll., The Entom. xxxv, p. 112, 1902 ; Fernald, Catalogue,
p. 243, 1903.

Scale of adult 2 about 1:8 to 2 mm. long, usually straight, widest shortly behind
the second exuviae and somewhat abruptly narrowed and attenuated posteriorly ;
white, glossy, with brown exuviae.

Puparium of 3 white, very long, distinctly tri-carinate ; exuviae almost colourless.

Specimens on “ umkavoti”’ at Durban had the 9 scales faintly marked by trans-
verse ridges and the ¢ puparia with a pronounced median ridge.

The body of the adult 9 is long, narrow in front, then almost parallel-sided to the
free abdominal segments, with the posterior margin regularly and broadly rounded.
The median portion of the body is highly chitinised and appears brownish-yellow in
mounted specimens. The anterior end is paler in colour, and the free abdominal
segments and the pygidium are almost colourless. Circumgenital glands in 8 groups :

3— 4 3-4 3— 4
1l—17—_s 24s W1—17
17-—27 17—27

Remarks. Professor Cockerell’s original description is as follows :—

‘“ © seale similar to that of P. cycadis, but perhaps narrower. Second skin paler,
as in cycadis.

“ similar to P. cycadis, but the strongly serrulate reddish-brown median lobes are
wide apart, the interval being nearly as great as the breadth of a lobe; the second lobe
consists of two lobules, of which the inner is the larger, and its tip projects a little

THE COCCIDAE OF SOUTH AFRICA. 103

beyond the level of the tips of the median lobes ; the margin just beyond the second
lobe bears two large dorsal glands, like those of the series on the next segments
anteriorly ; there are only four dorsal glands in the short rows nearest the anal orifice.
The middle of the body is red-brown, and strongly chitinised. Circumgenital glands
in eight groups ; the posterior laterals 19, middle laterals (cephalolaterals of other
genera) 11, median 5, in a transverse row, and the anterior groups characteristic of
Poliaspis form a transverse series broken into three linear groups of three or four,
which are widely separated.

“ § scale tricarinate.

“ On Carissa (? C. grandiflora, D.C.) plant, which belongs to the Apocynaceae,
Durban, Natal (Fuller). In these species and P. cycadis the anterior group of glands
are in transverse lines; in P. media, and the species described by Fuller from
Australia, the groups are circular.”

Habitat: On Carissa grandiflora (amatingula), Durban; collected by C. Fuller
(part of original material). On “umkavoti” (Chaetachme aristaia), Durban ;
collected by C. Fuller, October 1914.

Collection Nos. : 181 and 167.

160. Chionaspis (Pinnaspis) cyanogena (CkIl.) (Plate vi, fig. 180).
Hemichionaspis cyanogena, Ckll., The Entom. xxxiv, p. 226, 1901 ; Fernald,
Catalogue, p. 240, 1903.

Scale of adult 2 about 2 mm. long, narrow, moderately convex, white, but appear-
ing slightly greyish in massed specimens. Exuviae pale yellow to orange ; second
exuviae covered by a thin secretionary layer which is easily flaked off and therefore
often missing.

Puparium of ¢ white, non-carinate, or with a distinct median ridge. HExuviae pale
yellowish. |

The body of the Qis elongate, often rather narrow in front and suddenly broadening
to about the middle and suddenly narrowing again to the pointed pygidium. The
margins of the free abdominal segments are not conspicuously produced, but are
supplied with numerous gland-openings and a few short, pointed projections. The
pygidium has the median lobes moderately produced, their inner margins straight and
in close contact, their outer extremities sloping gently backward and twice notched,
forming three sub-equal rounded crenulations. The median lobes and the posterior
median part of the pygidial area are yellow, and the yellowish coloured spots on the
margin may indicate three other pairs of rudimentary lobes. At these points gland-
openings are apparent, and small rounded projections may be seen on the margin ;
the plates area httle longer than the median lobes and are distributed as shown in
fig. 180. Antennal tubercles + globular, with one curved flagellum. Circumgenital
glands in 5 large groups :

12—18
15—19 15—19
13—17 13-17

Habitat: On Alternanthera sessilis, R. Br., Durban; collected by C. Fuller. On
Alternanthera sessilis, R. Br., Scottsburg, and on native weed, Natal coast; collected
by C. Fuller, July 1915.

Collection Nos.: 177 and 177a.

104 CHAS. K. BRAIN.

161. Chionaspis (Pinnaspis) aspidistrae, Sign. (Plate vi, fig. 181).

Chionaspis aspidistrae, Sign., Ann. Soc. Ent. Fr. (4) 1x, p. 443, 1869.

Choonaspis brasihensis, Sign., Ann. Soc. Ent. Fr. (4) 1x, p. 444, 1869.

Choonaspis latus, Ckll., Psyche, vii, Suppl. 1, p. 21, 1896.

Chionaspis aspidistrae, Green, Cocc. Ceylon, ii, p. 110, 1899.

Hemichionaspis aspidistrae, Cooley, Spec. Bull. Mass. Exp. Sta. p. 45, 1899.

Chionaspis aspidistrae, Newst., Mon. Brit. Coce. i, p. 187, 1901.

Hemachionaspis aspidistrae, Fernald, Catalogue, p. 239, 1903.

Pinnaspis aspidrstrae, Lindinger, Die Schildlause, p. 79, 1912.

Scale of adult © variable, largest specimens about 2°5 mm. long, more or less pear-
shaped, thin, usually semi-transparent, whitish or yellowish to reddish brown.
Kixuviae colourless or faintly yellow. :

Adult 9 elongate, broadest across the abdominal segments, which have the margins
produced into rounded conical processes in the young adult stage. Pygidium as
illustrated (fig. 181). Circumgenital glands in 5 groups:

5—15
15—23 15—23
17—23 17—23

Habitat: On ferns, Aspidistra, etc., Cape, Natal, and Transvaal.
Collection No. : 176.

162. Chionaspis (Pinnaspis) proxima (Leonardi).

Hemichionaspis proxima, Leon., Bol. R. Sc. Agr. Portici, p. 193, 1914.

A translation of Professor Leonardi’s description is as follows :—

“Female. Body elongate, with the extremity rounded and having the greatest
width about the middle. The abdominal segments narrower than the cephalothorax
and projecting laterally in well defined lobes, which have their free margins rounded.
The lobes of the last abdominal segments have 2-3 long and robust hairs along the
free margin. The buccal apparatus with the maxillary-mandibular bristles radiating
and extending beyond the posterior extremity of the body.

“Antennae tuberculiform and surmounted with a long flagellum. Posterior
stigmata without glands.

“ Pygidium with two pairs of lobes, of which the middle pair, which is deeply
coloured, has the inner margins approximated and the outer more deeply incised.
Those of the second pair are separated from the first by a robust hair, and by the
mouth of a large wax-gland. They are very small, slender, somewhat hyaline, with
the free margin rounded and entire. On the outer side of each of the second lobes
there is first a hair, then the mouths of two large wax-glands, and at a certain distance
- from these another hair followed by the mouths of two other wax-glands, and lastly
near the pre-anal segment a fourth hair.

** The circumgenital glands are in five groups, e.7.,

10 8 12
21 1 17 19 gl 21
17 1g 16 16 22 23

THE COCCIDAE OF SOUTH AFRICA. 105

“The sexual aperture is situated in the middle of four groups of lateral glands. The
anal aperture is almost the same height as the sexual opening, but rather more towards
the pre-anal segments.

“ Colour of the body in specimens treated with acetic acid, yellowish, except the
pygidium which is yellow-ochraceous.

“Female scale. Oval, flatly convex, with the secreted portions scant, whitish grey,
and the exuviae at the apex and narrower than the scale, and of an ochroleucous
colour.

“* Dimensions of the scale :

Length of scale a ‘e Me a e 1,280u
Width of scale .. a an a bee ee 800u
Length of larval exuviae if ms o 350u
Width of larval exuviae ig Me me im 220u
Length of nymphal exuviae .. Fi me me 750u
Width of nymphal exuviae .. 470u

“Male scale. Elongate, sides parallel, terete or sliehtly sarzed, rounded behind,
with the dorsum deeply grooved, and the resulting ridges well marked. Larval
exuviae vellowish, small and situated at one end. Waxy scale white. Size of scale :
length 960u ; width 320u.

“ Habitat: Collected at Thies (Senegal) and at Mamu on mango, at Konakry on
Anona, at Hann (Dakar) on Calotropis procera and on an undetermined plant ; at
Dodowa, at Lagos, at Cotonou (Dahomey), at Quifangando (Angola) and at Pretoria
on undetermined plants.

“ Notes: This species is near Hem. orlandz, Leon., from which it can readily be
distinguished by the following characters. It has glands at the anterior stigmata
and the glands around the sexual aperture are always more numerous; there are
present, although not very apparent, a second pair of lobes, and further, the last four
abdominal segments are supplied with hairs, not only the last two as is the case in
Hem. orlandi.”

This species is not represented in this collection and although a thorough search
has been made around Pretoria it has not been found by the writer.

Habitat : On an undetermined plant, Pretoria ; collected by Professor Silvestri.

Collection No. : 178.

Genus Lepidosaphes, Shimer.

Lepidosaphes, Shimer, Tr. Am. Ent. Soc. i, p. 373, 1868.

Mytilaspis, Sign., Ann. Soc. Ent. Fr. (4) viii, p. 841, 1868 (no description).

Mytulasprs, Sign., Ann. Soc. Ent. Fr. (4) x, p. 91, 1870.

Scale of adult @ elongate, usually narrowed in front and gradually broadening
behind. The colour is very variable, but often orange-brown to dark brown. The
exuviae are terminal, with the second exuviae covered. The female scales are usually
+ curved. Ventral scale variable, similar to those of Chionaspis spp.

The ¢ puparium is similar to that of the 2, but smaller and narrower; the larval
exuviae are terminal. The posterior part of the puparium is often separated by a
thin transverse band of secretion, which acts as a hinge, allowing the extremity of the
scale to be raised to permit the adult J to emerge.

106 CHAS, K. BRAIN.

Adult 2 elongate, with the margins of some of the segments often produced.
Circumgenital glands in 5 groups, but each group usually consisting of fewer glands
than is normal in Chionaspis.

163. Lepidosaphes pinnaeformis (Bouché) Kirk. (Plate vi, fig. 183).

Aspidiotus pinnaefornus, Bouché, Stett. Ent. Zeit. xu, p. 111, 1851.

Coccus becku, Newm., The Entom. iv, p. 217, 1869.

Aspidiotus citricola, Pack., Guide to Study of Insects, p. 527, 1869.

Mytilaspis fulva, Targ., Bol. Soc. Ent. Ital. p. 131, 1872.

Mytilasprs flavescens, Targ., Ann. R. Min. Agr. p. 84, 1876.

Mytilaspis citricola, Comst., Rept. U. 8. Dept. Agr. p. 321, 1881.

Mytulaspis citricola, Lounsbury, Rept. Ent. C. G. H. p. 67, 1896.

Mytilaspis citricola, Green, Coce. Ceylon. i, p. 78, 1896.

Mytilaspis pinnaeformis, Newst., Mon. Brit. Coccidae, i, p. 204, 1901.

Mytilaspis beckiz, Ckll., Ent. News, xiti, p. 17, 1902.

Lepidosaphes pnnaeformis, Kirkaldy, Fauna Haw. iii, 2, p. 110, 1902.

Lepidosaphes pinnaefornus, Fernald, Catalogue, p. 313, 1903.

Lepidosaphes beckui, Fernald, Catalogue, p. 305, 1903.

Common Name: Mussel Scale.

Scale of adult 9 elongate, usually + curved, narrow in front and broadened behind,
with a flattened paler marginal area. Colour varying greatly with host-plant, etc.,
from pale greyish or yellowish brown to reddish- or olivaceous-brown. In size it may
reach 3 mm. long and nearly 1 mm. in breadth.

The © puparium smaller, narrower, and more delicate in texture, with sides =
parallel, and colour usually a little paler.

Adult 9 elongate, narrow in front, widest at the first free abdominal segment, 1.¢.,
behind the middle. The colour is generally whitish, or faintly yellowish with the
terminal segments darker. The antennae are represented by a small tubercle with
two stout curved hairs. The pygidium as illustrated (fig. 183). Circumgenital
glands in 5 groups : ,

5—8
9—17 9—I17
7—1l1l 7—Ill1

Habitat : On croton, Capetown, Port Elizabeth, and Uitenhage (C.P.), Johannes-
burg (Tr.), and Durban (Natal). On orange, Cape Province, Natal and Transvaal.
On Murraya, New Guelderland, Natal.

Collection Nos. : 262—262e.

164. Lepidosaphes gloveri (Pack.) Kirk. (Plate vi, fig. 182).
Coccus glovert, Packard, Guide to Study of Insects, Ed. 1, p. 527, 1869.
Mytilaspis flava, Targ., Catalogue, p. 44, 1869.
Mytilaspis glover, Comst., Rept. U. 8. Dept. Agr. p. 323, 1881.
Mytilaspis glover, Lounsbury, Rept. Ent. C. G. H. p. 71, 1896.
Mytilaspis pallida, Green, Ind. Mus. Notes, iv, p. 5, 1896.
Mytilaspis glovervi var. pallida, Green, Cocc. Ceylon, i, p. 85, 1896.
Leyidosaphes glovert, Kirkaldy, Fauna Haw. Is, 1, 2, p. 111, 1902.
Leyidosaphes glover, Fernald, Catalogue, p. 309, 1903.
Lepidosaphes gloverr, Lind., Die Schildlause, p. 106, 1912.
Common Name: Long Scale.

THE COCCIDAE OF SOUTH AFRICA. 107

The scale of the adult 9 is very long and narrow, 2°5 to 3 mm. long and about 0°5
mm. broad, long comma-shaped, usually reddish or greyish-brown with paler margins ;
on some host-plants the colour is paler. Hxuviae terminal, yellowish. Ventral scale
well developed, whitish.

& puparium much smaller and more delicate than the Q scale, and paler in colour.

Adult 2 elongate, sides + parallel, abdominal segments slightly wider, but without
prominent lateral projections. The colour when alive is pale yellow. Pygidium as
illustrated (fig. 182). Circumgenital glands in 5 groups :

4
6—8 6é—S8
4—5 4—5

Habitat: On citrus, Cape Peninsula; collected by C. P. Lounsbury. Also at Port
St. Johns, Stanger, and Warmbaths.

Collection No.: 261.

Genus ischnaspis, Douglas.

Female scale extremely long.and narrow, with the first exuviae extending beyond
the anterior margin. Male puparium elongate, similar to that of the second stage
female, without the central hinge of the 3 of Leprdosaphes.

Adult 2 very long and narrow, posterior margin of pygidium forming a regular
shallow concave depression, from which the median lobes project. Dorsal surface of

pygidium with a peculiar lattice-work design. Circumrgenital glands in 3 or 5 small
groups. |

165. Eschnaspis longirostris (Sign.) (Plate vi, fig. 184),
Mytilaspis longirostris, Sign., Bull. Soc. Ent. Fr. (6) ii, p. xxxv, 1882.
Ischnaspis filrformis, Douglas, Ent. Mo. Mag. xxiv, p. 21, 1887.
Ischnaspis filiformis, Newst., Mon. Brit. Coc. i, p. 210, 1901.
Common Name: Black Thread Scale.

Scale of adult 2 3 to 35 mm. long, very narrow, with sides parallel ; posterior
extremity slightly broader ; colour shining black, with greyish margins. Scale very
convex, often showing slight indications of transverse ridges, usually straight, but
where a number are crowded together the scales often make a regular bend to avoid
other scales and then continue again in the original direction. Larval exuviae
terminal, brownish in colour; second exuviae occupying about one-quarter of the
length of adult scale, covered with shiny black secretion like the remainder of the
scale. Ventral scale complete, robust, white or greyish in colour.

Adult 9 very long, sides parallel, broadest at the free abdominal segments. Colour
yellow. Pygidium as illustrated (fig. 184). Circumgenital glands in 3 groups:
3—4
4—6 4—6

A few large (to 4 mm. long) specimens in the Cape collection, otherwise typical,
have 5 groups of glands. Three of these have:

4 5 4

108 CHAS. K. BRAIN.

Habitat: On citrus, Warmbaths (Tvl.). On Chaetachme aristata (umkavoti),
Durban, Natal; palms, Kast London, Durban and Pretoria; Aucuba, Hast London.
On Dracaena australis, Kast London ; collected by C.P. Lounsbury, November 1906.

This species is particularly abundant on citrus at Warmbaths, Transvaal.

Collection Nos. : 277-2776 and 278.

Subfamily AsTEROLECANIINAE.

Adult females fixed to the stem of their host-plants ; sometimes causing pits in the
stems and thus becoming slightly or almost entirely embedded in the outer tissues.
Adults usually enclosed in a more or less horny or glassy cyst, which is most often
yellow in colour and, in one genus, has a marginal fringe of glassy filaments. The
legs and antennae are most often rudimentary or absent in the adult stage. Figure-8 ©
glands are present in one or more stages.

This subfamily is represented in South Africa by four genera which may be
distinguished as follows :—

Test of 9 usually yellowish, + transparent, with marginal fringe .. Asterolecanium.
Test of Q dense, almost ligneous, without fringe .. : .. Lecanodiraspis.
Test of Q = = waxy and without fringe, 2 without piemaee spines Cerococcus.
© naked, causing shallow rounded galls in stem* By et .. Amorphococcus.

Genus Asterolecanium, Targioni-Tozzetti.
Asterolecanium, Targ., Inter. 2nd Mem. Studi Cocc. Catalogue, p. 41, 1869 ; Sign.,
Ann. Soc. Ent. Fr. (4) x, p. 276, 1870.

Planchoma, Sign., Ann. Soc. Ent. Fr. (4) x, p. 282, 1870.

Asterodiaspis, Sign., Bull. Soc. Ent. Fr. (5) vi, p. ecix, 1876.

‘Female insect completely enclosed within a thin but compact glassy or horny test,
characterised by a continuous marginal fringe of glassy filaments. The test may be
hemispherical, flat, or even somewhat concave above. It varies in outline from
circular to linear. The surface is usually smooth, with, in some species, erect or
curling filaments on the disc, similar to those of the marginal series. . . . At the
posterior extremity there is a small opening through which the larvae escape. This
aperture is sometimes placed at the end of a tubular extension, which is often more or
less elevated. The colour of the test is invariably of a greenish or yellowish tint,
but the-fringe and dorsal filaments are sometimes tinged with red... .

“The adult female insect, denuded of its covering, is at first approximately of the
same form as its covering. After oviposition it shrivels up and hes at the anterior
extremity of the test, the remaining cavity being packed with ova. The antennae are
rudimentary, consisting of a single short joint with a few curved hairs at its extremity.
The limbs are totally absent... . No stigmatic spines. Anal lobes absent or minute,
usually represented by a pair of small tubercles and each bearing a stout seta.
In a few species, both tubercles and setae are wanting. Between the setiferous
tubercles are usually from two to four smaller tubercles, each bearing a small spine.
The anal ring is sunk in a tubular pit which sometimes opens on to the extreme margin,
but usually terminates shortly before the margin on the dorsal surface ; it normally
carries six stout hairs, but is sometimes hairless. . . .

* See Bull. Ent. Res. 1x, p. 112,

THE COCCIDAE OF SOUTH AFRICA. 109

“Male puparium of similar structure to that of the female, but smaller and pro-
portionately narrower. The fringe is simpler, consisting only of the nymphal and
larval elements. The winged adult emerges from beneath the posterior margin
without the aid of any hinged operculum such as occurs in the allied genus
Lecaniodiaspis.

“Adult male with distinct neck. Antennae ten-jonted, with three or four
knobbed hairs at apex. Ocelli large, in two pairs—dorso-lateral and ventral
respectively. Rudimentary eyes small, colourless and inconspicuous. Scutellum
ample. Halteres absent or obscure. (Genital sheath long and sharply pointed.
Two long caudal filaments.

“Larva of typical form. Antennae distinctly six-jomted. A marginal series of
(usually twenty-eight) paired glands. Posterior extremity with a pair of longish
caudal setae” . . . (Green).

Key to South African Species of Asterolecanium.
A. Anal ring hairless.
a. Antennae rudimentary.
Anal extremity of test turned up, tubular ; caudal setae of adult 9 short, stout
brevispinum, sp. D.
Anal extremity of test not prominently upturned or tubular ; caudal setae long
variolosum, Ratz.
aa. Antennae long, 10-jointed, caudal setae long wu .. borboniae, sp. n.
AA. Anal ring large, with 6 long spine-like hairs.
a. Marginal fringe of adult test reddish.

Test small, flattish, usually naked when olds: ie iN pustulans, Ckll.

Test small, flat ; on acacia a ” - .. conspicuum, sp.n.

Test large, rounded ; on bamboo ss net S, .. bambusae, Boisd.
aa. Marginal fringe ae adult test white.

Test yellow, globular, enveloped in very long ney ae .. euryopsis, Fuller.

Test greenish or creamy white, fringe short . £ stentae, sp. n.

166. Asterolecanium brevispinum, sp. n. (Plate vu, fig. 188 ; Plate vim, fig. 197).

Test of adult 2 about 3 mm. long, 2°2 mm. broad, very convex, with the caudal
extremity produced into a narow upturned process. The colour is bright yellow and
somewhat transparent, except at the anterior end which is bright brown in the few
specimens at hand, probably owing to the presence of the shrivelled female body.
Of the eight specimens in this collection five of the tests are entirely naked. The
other three show the presence of long large curved glassy filaments (fig. 188), which
were apparently equally long and equally numerous over the whole dorsal and
marginal areas. The integument of the body is very thin ; the antennae are small,

-+ cordate, with two short spurs. The caudal extremity is simple, with a small hair-
less anal ring ; the caudal setae are represented by short stout spines. The figure—8
glands are comparatively large (fig. 197).

Remarks. This species is somewhat like A. coffeae, Newst., but the fringe is white
instead of yellow, and there are other distinctive differences to be seen in the mounted
specimens, particularly the hairless anal ring and short caudal spines.

Habitat : On veld bush, Ceres, C. P.; collected by C. P. Lounsbury, October 1908.

Collection No. : 21

110 CHAS. K. BRAIN.

167. Asterolecanium variolosum, Ratz. (Plate vii, fig. 190 ; Plate viii, fig. 193).
Asterolecanwum quereicola, Sign. et Auctt. (nec Lecaniwm quercicola, Bouché).
Coccus variolosus, Ratz., Tharander Jahrbuch, xx, p. 187, 1870 (not seen; fide

Judeich & Nitsche).

Lecanvum quercus, Altum (nec Linné), Forstzoologie, iii, Insecten, p. 365, 1881.

Coccus variolosus, ‘“‘ Ratzeburg MS.,” Hagen, Can. Ent. xix, p. 60, 1887 (no

description).

Planchoma fimbriata (Boyer) Maskell, Extract, Trans. N. Z. Inst., 1894, p. 63.

Coccus quercicola, Sign., Judeich & Nitsche, Lehrb. Mitth. Forstins., ii, p. 1252, 1895.

Planchona quercicola (Bouché) Maskell, Trans. N. Z. Inst. xxviii, p. 396, 1895.

Asterodiaspis variolosus, Boas, Dansk. Forstzoologie, p. 395, 1896.

Planchonia (Asterolecanium) quercicola (Bouché) Froggatt, Dept. Ag. N.S.W.

No. 75; poi eo7,
Asterodiaspis quercicola (Bouché) Newstead, Coccidae of Brit. Is. (Ray Soc.) 1, pp.
1, 14, 34, 35, 36 and 39, 1900.

Asterolecanwum variolosum, Newst., Mon. Brit. Coce. 11, p. 156, 1902.

Ova pale yellow to light brownish yellow according to age ; about 190u long and
115 broad.

Larva, just emerged, pale yellow, with legs and antennae hyaline, about 260 long
and 145u broad. Antennae obscurely six-jointed. Caudal lobes small, tuberculate,
each with one long seta, averaging about the same length as the antennae. LHyes
prominent, lemon-yellow.

In the half-grown female the colour of the insect is clearly visible through the thin
transparent test. It is then 0°7 mm. long and 0°55 mm. broad, caramel-brown, with
a distinct white fringe and distinct transverse ridges on the dorsum (fig. 190).

At maturity the adult 2 completely fills the test, which has become stouter and
yellower but is still transparent enough to allow the greenish brown colour of the
adult insect to show through plainly. At this stage the tests vary in size from 1°2
mm. long by 0°9 mm. broad and 0:26 mm. high to 1-5 mm. Jong by 1°3 mm. broad
and 0°6 mm. high. As the eggs are matured and laid the body of the female shrinks
and the posterior end of the cyst serves as an ovisac. At this stage the test is
regularly domed and practically smooth, while the fringe has usually more or less
worn away. The line indicating the end of the shrunken body is often very distinct,
so that the front half of the test appears deep brown and the hind half pale yellow.

After clearing, staining and mounting, the adult 9 is almost circular, with the anal
extremity slightly narrowed. The anal orifice is very small and hairless, with two
small spines, one on either side. The caudal lobes are obsolete, but the two long setae
persist as shown in fig. 193. The figure-8 glands are small, in a single series all round
the body. The antennae are rudimentary, with one long and one short spiny hair
(fig. 193a).

Habitat: On Quercus sp., Cape Peninsula, Elsenburg (C. P.), Johannesburg, Irene,
Pretoria and Vereeniging (T'vl.).

Collection No. : 18.

THE COCCIDAE OF SOUTH AFRICA. lll

168. Asterolecanium borboniae, sp. n. (Plate vii, fig. 198).

Test of adult 2 small, about 1 mm. long and 0°75 mm. broad, rounded in front and
rather pointed behind, slightly convex, with the margin rather thickened and slightly
crenulate. The extreme posterior extremity appears tubular and is upturned or
recurved over the back. There is apparently no marginal or dorsal fringe. The
colour is pale greenish yellow and the test is almost transparent, showing the dark
body of the female at the anterior end.

When mounted the body is broad pear-shaped, about 0°9 mm. long and 0°6 mm. at
greatest width, with the posterior extremity slightly produced,and broadly rounded.
The characters are very indistinct. The anal ring is small ; the caudal setae stout,
about 65 to 70u long, and the margin has but a single row of figure-8 glands, which
are small, measuring approximately 8u across the pair. They are rather wide apart
(about 30u) and are not associated with simple glands (fig. 189). The mouth-parts,
with the rostral loop, average about 185, long.

A g mounted on the type slide from the Cape collection, measures 0°77 mm. long
without the antennae and genital spike. The latter is broad and strong and measures
150u. The antennae are 10-jointed, measuring approximately : (1) 20, (2) 26, (3) 60,
(4) 64, (5) 74, (6) 70, (7) 68, (8) 58, (9) 50, (10) 46.

Habitat : On leaves of “ gorse” (Borbonia cordata, Linn.), Ceres, C. P. ;- collected
by T. F. Dreyer, November 1906. :

Collection No. : 302.

169. Asterolecanium pustulans (Ckdll.) (Plate vu, fig. 187 ; Plate vii, fig. 200).

Asterodiaspis pustulans, Ckll., Jl. Inst. Jamaica, 1, p. 143, 1892.

Planchonia pustulans, Ckll., Science Gossip, xxix, p. 77, 1893.

Asterolecanium pustulans, Ckll., Bull. Bot. Dept. Jamaica, p. 8, 1893.

Test of adult 2 about 2 mm. long and almost as broad, nearly circular, with the
posterior extremity slightly produced. Dorsum convex, with a rather conspicuous
rounded median ridge, but without transverse ridges. When young the test is flat
and entirely covered with pinkish filaments, which appear to be divided on the
marginal area. When a little older the dorsal filaments are missing, but there is a
distinct marginal fringe of rather long pink processes. When full-grown the
marginal fringe also is usually lacking. The colour of the test is greenish yellow,
except when it contains the shrivelled female or eggs, which show through the semi-
transparent test. On apple the insects do not make pits in the bark, nor is there a
ridge around them. Marginal row of figure-8 glands single, with the addition of a
single row of simple glands.

The figure-8 glands are of two sizes and are scattered over the body surface ; the
smaller are about equal to the marginal series in size, the others almost twice as large
(fig. 200 a). Antennal tubercles prominent, with two curved spines and a minute spur
(fig. 200 6). Caudal lobes moderately prominent. Caudalsetae 75u. Anal setae 40u.

Habitat : On apple, Lourengo Marques ; collected by C. B. Hardenburg. On stems
of oleander, Lourengo Marques ; collected by C. P. Lounsbury, October 1914.

Collection Nos.: 300 and B.300.

(C605) B

112 CHAS. K. BRAIN.

170. Asterolecanium conspicuum, sp. n. (Plate vii, fig. 186 ; Plate vui, fig. 196).

Test of adult 9 about 1°5 mm. in diameter, almost circular, with a small, narrowly
rounded posterior extremity. On some species of Acacia the insects cause distinct
pits in the bark of the twigs (A. robusta ?), but on A. horrida this is apparently not the
case. Occasionally the bark has been seen to cause a low rounded mound around the
insect, producing an effect on the twig very similar to that caused by Amorphococcus
acacwue.

The test is at first flat, yellowish, with short reddish hairs over and around it. Later
it becomes moderately convex, smooth on top, semi-transparent and greenish yellow,
with only the marginal fringe persisting. When cleared, the integument is extremely
delicate and quite hyaline.

The figure-8 glands in the marginal series are comparatively eae and are accom-
panied by small circular glands of two sizes. The caudal extremity has low rounded
tubercles which are scarcely produced, each with one stout spine of moderate length
and a shorter one on the inside. The anal ring has six spines almost as long as the
longer ones of the lobes (fig. 196).

Remarks. This species is often observed to be heavily iaestea with a Hymenopterous
parasite, and in many cases the majority of old tests show the exit holes of such insects.

Habitat : On Acacia spp. (native species only), Pretoria, Marikana, and South-West
Protectorate.

Collection Nos.: 17 and 303.

171. Asterolocanium bambusae, Boisduval (Plate vii, fig. 185 ; Plate vii, fig. 195).

Asterolecanium bambusae, Bdv., Insectologie Agricole, 1869.

Asterolecanium bambusae, Sign., Ann. Soc. Ent. Fr. (4) x, p. 280, 1870.

Asterolecanium bambusae, Newst., Mon. Brit. Coce. 11, p. 151, 1903.

Asterolecanium bambusae, Green, Cocc. of Ceylon, iv, p. 328, 1909.

Mr. Green’s description, without reference to figures, is as follows :—

“ Test of adult female oval, convex above, the posterior extremity slightly produced
into a blunt point where is a small terminal aperture ; smooth, glassy ; colourless,
or tinged with pale green or yellow ; transparent, revealing the form of the insect and
eggs beneath. Dried examples assume a more definite ochreous colour, a brown patch
at anterior extremity representing the dead body of the insect. Marginal fringe
pinkish, consisting of a double series of glassy filaments springing from the margin in
pairs, each pair contiguous at the base and for the greater part of the length, but
diverging laterally at their free extremities which meet the ends of the adjacent .
filaments, thus forming a series of narrow loops irregularly crossed by fine web-like
threads; the outermost series continuous, except at anal extremity, and longest;
the second (nymphal) series less than half the length of the outer, and interrupted at
more or less regular intervals ; an innermost (larval) widely spaced series of crook-
shaped filaments. Length (without fringe) 2 to 25 mm. Breadth 1°10 to 1°60 mm.
Fringe, outer series 0'1 mm. ; inner series 0:04 mm.

‘* Adult female insect at first more or less filling the test. After the deposition of
the eggs, the abdomen becomes shrunken and the insect occupies the anterior part of
the test only, the remaining cavity being packed with ova. Colour of insect dull

THE COCCIDAE OF SOUTH AFRICA. lal

green, the dorsum more or less mottled with reddish brown. Dried insect uniform
reddish brown. Rostrum conspicuous, pyriform, approximately central. Antennae
submarginal, rudimentary, consisting of an irregular tubercle surmounted by two
longish stout bristles and one short spiniform hair. Spiracles conspicuous,
subglobular, at a considerable distance from margin, with scattered series of minute
pores connecting them with the margin. Abdominal extremity slightly cleft; anal
tubercles broad and stout, but not very prominent, each with a long seta at its
apex and one or two short stout spines near the base. Anal ring stout, with six
stout hairs which just project beyond the margin. Margin with a continuous
series of large approximately circular thick-rimmed pores, in pairs, and a double
ventro-marginal series of minute pores. There are some scattered paired pores of
a slightly larger size on the dorsum, and many simple circular pores on the ventral
surface of the abdomen. Length of extended insect 1:25 to 175 mm. Breadth
0-80 to 1:25 mm. After oviposition the insect becomes greatly shrunken and does
not resume its original proportions during maceration.

“ Karly adult female at first pinkish, later dull green, finely maculated with reddish
brown. ‘The outermost series of the fringe is at first the shorter, but soon outgrows
the nymphal fringe. ‘Traces of the larval fringe can often be distinguished at intervals
within the margin.

“In the nymphal stage the fringe is in a single series, but within the margin are
remains of the divaricating filaments of the later larval stage.

‘““ The newly hatched larva is of an oblong oval form, very pale pinkish or reddish
brown. Limbs well developed. Antennae six-jointed, the sixth much the longest,
with truncate apex. Margin with fourteen 8-shaped spinnerets on each side, and a
median dorsal pair of similar pores immediately above the rostrum. Length 0:2 mm.

“ The larva subsequently secretes a fringe of short glassy filaments which are paired
and diverging on the cephalo-thoracic area, but single on the abdominal margin
(although the spinnerets giving rise to these latter are of the same form as the others).

‘“ Ova very pale pinkish or yellowish.

“Male unknown. A single example was observed, in which the dorsum was
distinctly tricarinate and bore a marginal and three dorsal series of curling filaments.
This may possibly be the male puparium.”

The caudal extremity and part of the marginal series of figure—-8 glands of the é adult
female are illustrated (fig. 195).

Habitat : On stems of large bamboo, Durban ; collected on Aina occasions by
Claude Fuller, C. . cnn and A. aes

Collection No.

172. Asterolecanium euryopsis, Fuller (Plate vii, fig. 192 ; Plate vin, fig. 201),
Asterolecanwum euryopsis, Fuller, Agr. Journ. C. G. H. xiv, pp. 93-95 and 818, 1899.
As no fresh material of this species has been seen by the writer and all the numerous

specimens of this collection are somewhat rubbed the following particulars are taken

from Mr. Fuller’s description (l.c.) :—
‘“ The infested twigs have the appearance of being covered with a fine soft floss of a
yellowish or white colour which is easily rubbed off and the insects exposed. The
(C605) B2

114 CHAS. K. BRAIN.

scales within which the females live very much resemble small glass beads, or the drops
of resin which the plants exude and for which they have no doubt often been mistaken.
They are thin and transparent with a yellow tinge of colour, smooth and shining, and
very convex, being nearly spherical. Generally the yellow colour of the scale is not
noticeable as it appears either brown or dark green, according to the different stages of
development of the female inside. In size they are about one-fourteenth of an inch
in diameter and a little less than that in height.

“Tf these female scales are carefully examined with a magnifying glass it will be
noticed that the floss is a production of the insect and that it 1s secreted through the
scale at all parts, but mostly around the margin, and curls up in various directions,
so that in fresh specimens the scale is quite hidden. If one of the scales is open at
about this time of the year, mid-January, the eggs, small pinkish brown particles, will
be found inside. Scales which contain eggs are generally parti-coloured, one half
being yellow and the other brown or black. It isthe yellow half which contains the
eggs, and the yellow colour is due to the scale, whilst the darker portion is the dried
remains of the female showing through it.”’

Puparium of g about equal to the diameter of the female test in length, shghtly less
than half as wide, flat, with a slight median keel, yellow and transparent, likean empty
female test. Thereis no definite operculum, such as isfound in many species, but the
extreme hind margin is split to allow the exit of the ¢.

Adult 9, cleared and mounted, broad pear-shaped, 1°5 mm. Jong and 1:2 mm. broad.
Body hyaline, with figure-8 glands very large and uniformly scattered except at the
posterior extremity, where they are practically absent. The caudal extremity (fig.
201) is deeply excavate, with the part around the anal opening densely chitinised.
The caudal setae are replaced by two stout spines, and the spines of the anal ring are
comparatively very long and stout.

Remarks. This insect became so numerous on the harpuisbosch in certain parts of
Cape Colony about the years 1898 and 1899 and appeared to have such a toxic effect
upon the host-plant that it was suggested that the insect be spread as widely as
possible in an endeavour to kill off the bush, which rendered large tracts of the country
around Tarkastad useless. It was recorded in the Agricultural Journal (l.c.) that the
bushes grew to a height of 8 to 10 feet and that no grass would grow under them.

Habitat : On harpuisbosch (EHuryops tenuissimus, Less.).

Collection No. : 20.

173. Asterolecanium stentae, sp. n. (Plate vu, fig. 191; Plate vii, fig. 199).

Test of adult 9 about 3 mm. long, regularly oval and very convex, with the extreme
posterior extremity slightly upturned. Colour greenish yellow suffused with brown,
which is more intense at the margin. Dorsal surface scantily supplied with short
white glassy processes.

The young female is similar in form, but is pale translucent green with a white
marginal fringe and with numerous glassy filaments on the dorsum. The median line
is not pronounced, but bears a large number of long glassy filaments in a longitudinal
crest. The marginal fringe is about one-fourth the greatest width of the body in
length (fig. 191).

puparium not observed.

THE COCCIDAE OF SOUTH AFRICA. 115

Cleared and mounted the adult 9 is broad pear-shaped, hyaline, except for the
mouth-parts, spiracles and anal portion. The figure-8 glands form a continuous
single row around the margin, except at the sides in the vicinity of the spiracles, where
a double row is present in association with single glands of two sizes. Figure-8 glands
are also present in scattered form over the dorsum. A little removed from the
posterior end is a transverse series of large opaque glands reminding one of the grouped
glands of the Diasprn Az (fig. 199). Theantennae are small tubercles with two or three
small spines (fig. 199, 6).

Remarks. This insect is obviously very similar to A. thes, Douglas, both in general
appearance and microscopic characters. It is, however, slightly larger and could be

distinguished by the slight difference in colour and the slightly upturned posterior
end of the test.

Habitat: On stems of Caralluma caudata (Asclepiadaceae), sent in by Miss 8S. Stent,
July 1916. Also on Huerma transvaalensis, Stent, and Stapelia sp., in Rockery at
Division of Botany, Pretoria.

Collection No. : 29.

Genus Lecanicdiaspis, Tare.

Lecaniodiaspis, Targ., Bol. Soc. Ent. Ital. i, p. 261, 1869.

Lecaniodiaspis, Ckll., Can. Ent. xxxi, p. 267, 1899.

Prosopophora, Douglas, Ent. Mon. Mag. xxviii, p. 207, 1892.

Burchippia, Green, Ann. Mag. Nat. Hist. (7) vi, p. 450, 1900.

The adult 9 in this genus is entirely enclosed in a compact tough papery test, which
is generally broad oval in shape and buff or yellow in colour. Its upper surface may be
flat or convex, and may be + smooth or ribbed or carinated. Occasionally there are
small waxy processes on the dorsum, and the insects form shallow pits in the bark of
the host-plant. In the South African species only are there waxy plates on the
dorsum. There is a small, + circular aperture at the posterior extremity of the
test to allow the larvae to escape, but this is usually obscured until the eggs have
hatched. The test, when first formed, fits closely to the body of the female, which
gradually shrinks with oviposition until the test becomes in reality an ovisac.

The adult © is sometimes apodous ; in other cases there are rudimentary legs or,
more rarely, they are well developed. The antennae are usually well developed, seven
to nine segments being most common. The anal extremity is slightly cleft, with the
outer angles generally rounded and bearing a stout seta and a few spines. The base
and inner angles of the cleft are supplied with a bilobed thickening, which reminds
one very strikingly of the anal plates of the Lecanimaz. The dorsal anterior margin
of the cleft has a transverse chitinous band, in front of which is situated the anus.
The anal ring bears 8 to 12 hairs, usually 10. The derm is plentifully supplied with
8-shaped glands, and on the dorsal surface of the abdomen there are often two
longitudinal rows of cribriform plates.

Key to South African Species of Lecaniodiaspis.
A. Legs absent.

Test smooth, flat, not divided into il m, at .. mimosae (Mask.)

Test divided into plates i a o .. natalenis, sp. n.
AA. Legs rudimentary. 3

Test smooth, rounded above .. + oe re AG magna, sp. N.
AAA. Legs well developed.

Shorter than antennae K 3 et chy sf - braber, sp. n.

Longer than antennae ue BS ci ae it tarsalis, Newst.

116 CHAS. K. BRAIN.

174. Lecaniodiaspis mimosae (Mask.) (Plate ix, fig. 202 ; Plate x, fig. 215).
Prosopophora prosopidis var. mimosae, Mask, N. Z. Trans. xxix, p. 316, 1897.
Lecaniodiaspis mimosae, Ckll., Check List Suppl. p. 392, 1899. ;
Test of adult 9 about 4°5 to 5 mm. long., 3°5 mm. broad and 1°7 mm. thick, with

the dorsum almost flat, the upper and lower surfaces almost parallel, with the margins

rounded. When not crowded together the specimens are glued flat to the bark,
button-like, but when a number are massed together they are often much distorted.

The colour of the young is creamy, but later becomes suffused with brown, with a

more distinct median line. The dorsal surface is flaky, without keel or transverse

ridges.

3 puparium 1:8 mm. long, 1 mm. broad, elongate oval, rather more pointed in front,
flat, with a median keel and faint transverse ridges, pale brown, with a distinct semi-
circular operculum. |

When boiled in KOH and cleared, the derm of the adult 9 is thin and hyaline, and
the numerous gland-pores are very conspicuous. There are two dorsal rows of
cribriform plates, 5 in each row. These are + circular, small, 24u in diameter and
minutely perforate (fig. 215, b). The antennae are long (about 180), with seven or
eight joints, sometimes appearing 9-segmented owing to a pseudarticulation in
segment 6. The terminal segments bear several long stout processes with blunt
extremities (fig. 215). Legs absent; in two cases indications of extreme rudiments of
legs were found, in one case represented by leg i. and in the other by leg ii. The two
pairs of spiracles are broad, and the stout tracheae generally remain after treatment
with potash. The anal lobes are short ; each bears one very strong spine and 'a few
short conical ones. The stigmatic spines are long, stout, slightly curved, and
clubbed (fig. 215, a).

Habitat: On Acacia horrida and other species, Fort Beaufort ; collected by C. P.
Lounsbury, September 1900. Port Alfred ; collected by A. Kelly, March 1915.
Namaqualand (Cape No. : 1254).

Collection No. : 22.

175. Lecaniodiaspis natalensis, sp. n. (Plate ix, fig. 205; Plate x, fig. 213).

Test of adult 2 about 2°5 mm. long and 1°6 mm. broad at the widest part, which is
about the middle, flat, somewhat elliptical with the two ends narrowed. In some
cases the anterior end is broadly rounded and the posterior extremity pointed. The
dorsum is flat and covered with a layer of white material, which is distinctly divided
into, three series of + rectangular plates, the appearance of which suggests an
Orthezia. The median series is not quite so broad as the two lateral ones and consists
of nine patches, the number which is apparently constant for each of the two lateral
series also (fig. 205).

When :placed-in boiling KOH the body becomes pinkish but the liquid is not
appreciably coloured. |

The body of the adult @ is flat, about 1°8 mm. long, and 1:2 mm. broad at the widest
part, which is just in front of the middle. The anterior end is often suddenly
narrowed at the level of the antennae and the anterior margin rounded. - The
posterior portion tapers to the deeply cleft extremity. The mouth-parts are

THE COCCIDAE OF SOUTH AFRICA. a7

comparatively small, with a long rostral loop. The antennae (fig. 213, a) are short,
annular, of four or five segments. Legs entirely absent. The whole dorsal surface
has numerous scattered, minute 8-shaped glands. Cribriform plates (fig. 213, b)
small, inconspicuous, few in number. The anal cleft is as shown in fig. 213. The
caudal setae are strong, and shorter than those of the anal ring (60u). In addition
to the two caudal setae there are five or six short spines on 1 each caudal lobe. Anal
ring with 8 long hairs (100,).

Habitat : On stems of Hibiscus, Durban ; collected by C. P. v. d. Merwe, July 1916.

Collection No. : 301.

176. Lecaniodiaspis magna, sp. n. (Plate ix, figs. 206,209 ; Plate x, fig. 214).

Adult females ae on on crown of the host- eee at just about ground-
level.

Test of adult 2 about 6 mm. long, 4°5 mm. wide, and 3 mm. high, regularly broad
oval, or slightly narrowed in front and with the hind margin very slightly flattened,
with a faint median indentation. The dorsum is very convex, ventral surface
slightly rounded. The test is entire and homogeneous in texture, smooth or very
faintly roughened, without ridges, but occasionally with very faint ribbed corrugations
atthe sides. The colour is of a uniform biscuit tint (fig. 206).

When received, the largest specimens removed for mounting contained many eggs.
The body entirely fills the test, which is thin, tough and papery. In boiling KOH the
test breaks down and the liquid is stained a rich brown colour, while the 92 takes on
a rose-pink hue. When boiled for some time the body wall is clear, hyaline, and very
delicate.

Adult 9°, cleared and mounted, about 3:7 mm. long, almost circular, hyaline, with
the antennae and anal armature dense. Legs extremely rudimentary, represented
by a small (50x) conical tubercle, with a minute process at the tip and several delicate
hairs which appear to represent the digitules (fig. 214, a). Antennae comparatively
short (170), of nme segments, which are annular and slightly tapering to the apex
(fig. 214). Anal ring with 10 stout hairs. Cribriform plates small, with comparatively —
large pores. Caudal spines short (27), two on each lobe. The anal plates are broad,
each with deep wrinkles and two stout spines, and possibly two more slender, one
above and one below (fig. 214,6). These latter are missing in the slides, but the pores
indicate their presence.

Remarks. Two plants were submitted, the one crown bore seven large females
ranging from 5 to 6 mm. in length, the other about 50 specimens identical in
appearance but smaller, all the tests being about 3 mm. in length (fig. 209).

Habitat : On crowns of native bush with small narrow leaves ;_ collected at Groot
Drakenstein and sent'in by C. W. Mally, June 1916.

Collection No.: 27.

177. Lecaniodiaspis brabei, sp. n. (Plate x, figs. 218-218, a).

Test of adult 9 about 3:2 mm. long, 2 mm. wide, and 1°5 mm. high, oval, convex,
ochre-yellow, with a thin covering of greyish secretion which is easily flaked off. The
dorsum is not quite smooth, but has faint rounded rib-marks and occasionally a faint

118 CHAS. K. BRAIN.

median ridge. With the roughened secretion removed the colour and general appear-
ance of this species is very like the figure Green gives of his L. azadwrachtae.

§ puparium of the usual type, pale bufi-coloured, not yellow as is the Q test.

Larva yellow, 0°5 mm. long, and 0:24 mm. wide. Antennae and legs well developed.
Antennae 5-segmented ; basal segment cylindrical, hardly as wide as ii. The
posterior extremity of the body is narrowed and deeply and roundly indented, so that
there appear to be two large caudal lobes. The inner margins of these have chitinous
plates similar to those of the adult ; the transverse chitinous bar in front of these is
not very conspicuous. The caudal lobes each bear one very long seta, about 0°2 mm.
in length and one or two short spines.

Second stage 2 about 1'2 mm. long. In this stage the body has become more
rounded, especially at the posterior extremity. The caudal lobes are comparatively
shorter and their extremities are curved together, and the long caudal setae of the
larva are replaced by two stout spines about 35 in length. The antennae and legs
are situated further from the margin and appear proportionately smaller. The
antennae are 6—jointed.

Adult 9, cleared and mounted, about 3 mm. long, almost circular, with the derm
colourless, hyaline. The legs are well developed, but small, shorter than the antennae
(180u: 250u). The three pairs of legs are about equally developed (fig. 218, c). The
antennae are normally 8—jointed, but occasionally appear 9-jointed through the
pseudo-articulation of the 5th segment (fig. 218). There are scattered stout spines,
about 27u long, on thick bases, at intervals around the margin (fig. 218, 6). The
stigmatic spines are usually unequal, one long and one short, curved, slightly clubbed
(fig. 218, a). Figure—8 glands small, uniformly scattered over the surface.

Remarks. This species is similar in many respects to Lecaniodiaspis africana, Newst.,
but is readily distinguished by the smaller size and the fact that the antennal segments
ill, iv, v, and vi, are longer than wide and the tarsus is considerably longer than the
tibia (fig. 218, c).

Habitat: On wild almond (Brabeium stellatifolium, Linn.); collected by C. P.
' Lounsbury, Newlands, 1896 (Cape No.: 1274); also collected by C. W. Mally,
December 1915. |
Collection No. : 298.

178. Lecaniodiaspis tarsalis, Newst. (Plate ix, fig. 203 ; Plate x, fig. 217).

Lecanwum tarsalis, Newst., Bull. Ent. Res. viii, p. 16, 1917.

Test of young 9, about 1:5 mm. long, flat, with a distinct median keel, purplish grey
in colour.

Test of adult 9 bufi-brown, about 3 mm. long, 2 mm. broad and 1 mm. high, with
the dorsum roughened by a series of low tubercular projections. There is a distinct
median series and a sub-dorsal row on each side, the marginal areas being irregularly
roughened,

S$ puparium creamy white, about 1°3 mm. long, and 0°9 mm. wide; similar in form
to the 9 test but more flat and much lighter in colour.

Adult 9, mounted, about 1°6 mm. long and 1‘1 mm. broad, regularly oval, with a
complete regular marginal row of stout spines 30u in length, about 15 on each side

THE COCCIDAE OF SOUTH AFRICA. 119

(fig. 217, 6). Legs and antennae present. All three pairs of legs about equally
developed, all longer than the antennae. Antennae 9-jointed (fig. 217). Caudal
setae stout, about 80u in length. Cribriform plates small (124 diameter), yellow
(fig. 217, a). On the venter, a little anterior to the position of the anal ring, are two
long straight slender setae, about 115 in length, pointing backward.

Habitat: On native tree (Dombeya rotundifolia), Pretoria; collected by the writer,
September 1914. On Hibiscus in a nursery, Pretoria; collected by A. Kelly, June
L1G: !

Collection No. : 23.

Genus Geroceccus, Comst.

Adult 9 enclosed in a dense waxy test without a marginal fringe as in
Asterolecanium, and generally without waxy processes. The test 1s usually dense,
opaque, not semi-transparent, and simulates the test of a Tachardia more than that
of an Asterolecanium. The posterior end of the test is sometimes slightly produced
and uppurned, with the anal aperture at the extremity.

Adult 92 with the caudal extremity produced into a rounded protuberance which
bears the anal armature. There are two stout lobes, each with a long seta and
several spines ; the space between them is + chitinised, with a prominent rounded
tongue-shaped plate. Antennae and legs rudimentary. Derm with conspicuous
S-shaped glands. Stigmatic spines absent.

G puparium elongate, tubular, with a large oval or circular operculum above the
posterior extremity.

Key to South African Species of Cerococcus.
A. Test smooth, + globular, brown .. a ig be .. ovordes (Ckll.).
AA. Test rough, ornate, brick-red ipl be n ey, .. ornatus, Green.
AAA. Test yellowish or brownish.
Marginal 8-shaped glands much larger than those on dorsum passerinae, sp. 0D.
Marginal 8-shaped glands little if any larger _... we .. royende, sp. Nn.

179. Gerococcus ovoides (Ckll.) (Plate x, fig. 219).

Pollina ovoides, Ckll., The Entom. xxxiv, p. 225, 1901.

Professor Cockerell’s description is as follows :—

“© scale a rounded conical object much like a Lepidopterous egg, about 14 millim.
diam., roughened radiately, pale brown with four longitudinal stripes of white
secretion converging to the top of the scale, which is usually reddish.

“ § scale elongated, about # millim. long, roughened, yellowish or pink, with an
oblique terminal cap.

“ 2 scales, soaked in liquor potassae, give a deep orange-brown colour, and the
insects themselves turn deep crimson; Q adult globose ; skin with many simple round
glands and tubular glands, and some figure-of-8 glands; anal ring with numerous hairs ;
caudal lobes prominent, conical, about 45u long, with a few small spines, and ending
in stout bristles about 90u long ; mouth-parts well developed, but small; labium
short and broad, dimerous, the last joint with bristles on its margin; antennae

120 CHAS. K. BRAIN.

represented by small rounded tubercles about 15u long, with a little terminal
prominence which appears to represent a second joint, and about six stout bristles
about 18 long ; spiracles small but distinct ; legs wanting.

“ Embryonic larva with a row of figure—of-8 glands down each side, and dorsal and
subdorsal rows of small round glands, the latter failing caudad, the last five glands of
the dorsal rows being absent in the subdorsal ; labium very short and broad, cup-
shaped ; form of insect elongate-pyriform ; antennae thick, 6-jointed, last joint not
very greatly longer than the one before, and notched asin P. pollini.”

The antennae and caudal extremity are figured (Plate x, figs. 219 and 219, a).

Habitat: On stems of undetermined native tree, Back Beach, Durban; collected
by C. Fuller. On fig, Pietermaritzburg ; collected by A. Kelly, 1911.

Collection No. : 26.

180. Gerococcus ornatus, Green.

Cerococcus ornatus, Green, Cocc. Ceylon, iv, p. 306, 1909.

Mr. Green’s description, omitting figure references, is as follows :—

“ Test of adult female irregularly oval, the posterior extremity (in fully matured
examples) produced into a short upturned tube. In earlier examples this caudal
extension is in the form of a trough, partially closed above by stout curling filaments.
Dorsum strongly convex, the sides sloping inwards to the comparatively narrow
ventral area, which is thin and easily ruptured. Dorsum with three prominent
transverse waxy ridges produced, at the middle and two sides, into stout pointed
processes ; the lateral processes longest, tapering to fine points, and projecting (in
fresh examples) considerably beyond the margin, which is itself thickened and
tubercular. Posterior margin with two or more tapering processes directed
backwards. There are several pairs of short white curved filaments on the median
line, behind the third transverse ridge. In older examples the waxy processes are
gradually reduced until they practically disappear, and old worn tests are more or
less globular in form. Colour of early test purplish brown ; the ridges, processes, and
thickened margins bright coral-red or crimson. Older examples become at first
uniform pale red, and finally dull reddish brown. There are two white waxy ridges on
each of the sloping sides, extending from the first and third lateral processes,
respectively, to the stigmatic openings. Length 2 to 2.5 mm. Breadth (exclusive
of processes) 1°12 to 1°50 mm.

‘“ Male puparium oblong, narrow, rounded at the extremities. An inconspicuous
median and two prominent transverse ridges, the latter terminating laterally in
pointed processes. A smaller pointed process, on each side, close to anterior extremity.
Posterior third occupied by a large circular operculum with raised margins. Colour
pale red tocrimson. Length,1mm. Breadth, 0°45 mm.

“ Adult female insect of a delicate greyish tint, in life. Broadly oval: the terminal
segments abruptly narrowed. Abdomen terminating in two prominent irregularly
conical lobes, with a bluntly conical median plate between them ; each lobe with a
triangular chitinous plate on its inner side and a longish stout seta at its apex.
Ventral lip of anal aperture with four or more long stout spines. Anal ring with
eight hairs. Rudimentary antenna with from eight to twelve stout hairs on its apex.

THE COCCIDAE OF SOUTH AFRICA. 121

Limbs entirely wanting. Dorsum with numerous large and conspicuous paired
glands, disposed principally in three broad transverse bands across the middle of the
body, leaving the anterior and posterior areas comparatively free. The first and
third bands divide near the margins, where they enclose a series of small circular
glands. ‘There is a straggling group of the large paired glands within the anterior
_margin; a short marginal series on each side of the abdomen, just before the
constriction ; and another short series, on each side, close to the anal lobes. Amongst
the larger paired glands, and scattered over other parts of the dorsum, are many of a
much smaller size. Cribriform plates small, each with a broad densely chitinous
border and areolate centre; in two groups of four, immediately anterior to the
narrowed part of the abdomen. Venter with small circular glands in loose transverse
series across the abdominal segments, and some scattered paired glands of the smaller
size. Length, 2 mm.

* Adult male uniform dark brown. Wings hyaline ; nervures colourless. Genital
sheath very broad at base; sharply pointed at extremity. Although the single
example examined showed no waxy caudal filaments, there is a distinct glandular pit
on each side, giving rise to fine paired setae such as usually support caudal filaments.
Length,1 mm. Expanse of wings, 1°75 mm.”

- Habitat: On stems of Aberia caffra and Calodendron capense, Thunb.; collected at
Pietermaritzburg by A. Kelly, April 1917.
Collection No.: 95.

181. Gerococcus passerinae, sp. n. (Plate x, fig. 216).

Test of adult 2 about 2°4 mm. long, elongate, convex, with the posterior end
slightly produced, thin, brittle, dull straw-yellow. The dorsum is convex, with a wide
shallow groove from about the middle to the caudal end. The remainder of the test
_ is uniformly arched, but in a few specimens there appears to be a submarginal series
of low tubercles.

Puparium of ¢ usual, elongate, paler in colour than the @ test.

Cleared and mounted the adult 9 is broad pear-shaped, with the prominent anal
portion produced. The integument is thin and hyaline, with 8-shaped glands
numerous but scattered. Those of the marginal series are large, and mostly in a
single row, but at the sides the series becomes double or treble for a short distance.
The 8-shaped glands scattered over the dorsum are smaller, being only about half the
diameter of those in the marginal series. The rudimentary antennae consist of small
conical protuberances each with about 6 spines (fig. 216, a). The caudal lobes are
prominent, with their inner faces chitinised and the spines strongly developed (fig. 216).

Habitat : On native shrub (Passerina ericoides, L.; Thymelaeaceae); collected by
C. P. v. d. Merwe at Montague, C. P., October 1914.

Collection No.: 24.

182. Gerococcus royenae, sp. n. (Plate ix, fig. 210).

Test of adult 9 varying in size to 3 mm. long and almost as broad and high, very
roundly convex, with a short caudal prominence around the aperture. The test is
comparatively thick and brittle, dull orange-yellow to orange-brown in colour,
without fringe or processes of any kind.

122 CHAS. K. BRAIN.

The g puparia are elongate, tubular, open behind, straw-coloured, with a brownish
secretion, which often appears in + rectangular patches on each side of the median
line (fig. 210).

The mounted @ is almost circular, with the caudal extremity slightly produced.
The integument is clear and has a marginal series of 8-shaped glands of medium size.
Other similar glands are scattered over the dorsum and these are of about the same
size as those of the marginal series, not much smaller as in C. passerinae. The
antennae are small conical protuberances with several curved spines, and are longer
and narrower than in C. passerinae. The caudal lobes are moderately long, but are
not thickened on their inner edges, and a short distance in front of the anal ring are
two chitinised discs.

Habitat: On stems of blauwbosch (Royena pallens, Thunb.; Ebenaceae),
Fauresmith, O. F.§.; collected by J. C. Faure, March 1915.

Collection No.: 96.

Subfamily TACHARDIINAE.
“ Insects enclosed in a resinous cell with three orifices. Adult females apodous,
with the terminal segments produced into a tail-like organ bearing, at the extremity,
the anal orifice, which is surrounded by a broken setiferous ring ”’ (Green).

Genus Tachardia, R. Blanchard.

Professor Cockerell (‘“ The Entomologist,” xxxiv, p. 249, 1901), before describing
Tachardia albida from Natal, gives the following synopsis of this genus :—

“ Tachardia, Blanchard (Lac Insects).

This genus contains some very diverse elements, which will no doubt eventually
be treated as genera. We may for the present recognise three subgenera :—

(1) Tachardia proper. Type, T. lacca, the East Indian commercial lac. Female very
elongated, vasiform ; the individuals enclosed in masses of lac surrounding
the twigs, never separate. I know of only one species of this group.

(2) Tachardiella, subg. nov. Type, 7. cornuta, Ckll. Female more or less globular ;
individuals often separate. This includes the species of America and
Australia.

(3) Tachardina, subg. nov. Type, 7. albida, with the characters given below.”

It will be noticed from the description of 7. albida given later that the main points
of difference between this and the other species mentioned by Professor Cockerell are
(a) the tests are collected in large masses ; (b) the caudal process is peculiar ; (c) the
spine is absent. Of these characters none is peculiar to the African species, and
T. albida is the only species I know of in which the spine is absent. 1 therefore follow
Newstead in disregarding the subgenera for the present and including the six South
African insects in the genus Tachardia.

I might mention that the insects in this sub-family are by far the most difficult to
make satisfactory microscopic slides of, as nearly all the old dry specimens seem to be
eaten out, and after dissolving the lac test there usually remain but a few fragments
which are quite useless for determination. Fresh material is therefore most desirable
in this group.

THE COCCIDAE OF SOUTH AFRICA. 123

Key to South African Species of Tachardia.

- A, Antennae present.
a. Test crimson or deep red, with pale radiating lines .. actinella, Ckll. & King.
aa. Test deep shellac-brown,
b. Rudimentary legs represented by small conical spines .. munor, sp. n.
bb. Legs entirely absent .. me a a i -» ». karroo, sp. n.

AA. Antennae absent.
B. Spine present.

a. Test deep purplish brown, with radiating ridges .. .. decorella (Mask.)
aa. Test yellowish brown .. +. Me a ie ne affluens, sp. n.
BB. Spine absent.

a. Test whitish or pale yellow... ae Ee as) re albida, Ckll.

183. Tachardia actinella, Ckll. & King (Plate ix, fig. 208; Plate xi, fig. 221).

Tachardia actinella, Ckll. & King, The Entom. xxxiv, p. 342, 1901.

The original description is as follows :—

“0, Scale about 3 mm. long, 3 broad, and scarcely 2 high, rounded and depressed,
dark crimson, with about sixteen strong, but obtuse, radiating yellowish white ridges ;
~ centre of scale formed as in 7. decorella. The scales are mostly separate, but some-
times two or more coalesce.

“Q. Dark red, 24 mm. long, 14 broad ; bright pink when boiled in taustic potash ;
this colour is due to the internal juices, the skin being perfectly colourless. Antennae
stout, cylindrical, pale, about 140u long, 52 broad at base, 36 in middle, 28 at end,
obscurely four-jomted. Lac glands with over 60 orifices. Mouth-parts about 141
broad. Spine well-developed, 120u long, rapidly enlarging 35u from tip to a very
broad (95) base. Anal ring with 10 long bristles; chitinous anal plate roughly
semicircular, posteriorly with a deep linear incision 60 long ; on each side of the anal
plate is an elongated process terminating in two sharp spines, the structure being
apparently the result of a fold in the plate. Anteriorly, the anal plate is tuberculated.

“ 3. Scale cylindrical, elongated, of the usual form ; dark red.”

A large collection of 3 material was received from Grahamstown in March 1915
when the insects were just emerging. The following particulars are therefore added
from notes made at the time :

& test deep madder-red, glassy, about 1 mm. long and 0°5 mm. broad ; flat, upper
surface slightly arched, with a median flatly-rounded keel with transverse striae.
The anterior end is rounder but narrower than the posterior extremity, which is
somewhat elevated, so that the opening from which the § emerges appears dorsal.
Prior to emergence this orifice is closed by a glassy cap. When the insect is ready to
emerge the cap is raised from behind and the two white caudal filaments protrude.

Body of ¢ adult (freshly mounted) 1:2 mm. long, and 0°33 mm. broad across the
thorax. Genital spike 0°24 mm. long. Wing 0°95 mm. long. The body and
scutellum are of a beautiful madder-red colour; the eyes are black; the legs and
antennae pale, almost colourless ; and the wings are clear but iridescent.

124 CHAS. K. BRAIN.

Caudal filaments (2) long, from 1} to 14 times the length of the body. These are
dense white, slender, and in hfe widely divergent. The antennae are 10-jointed,
the joints, when freshly mounted, measuring approximately :—(1) 30, (2) 58, (3) 100,
(4) 92, (5) 85, (6) 68, (7) 58, (8) 68, (9) 50, (10) 65.

Habitat : On orange, Qumba, C. P. On pomegranate, Ladysmith, Natal. Also
on native trees at Kast London and Grahamstown, C. P., Natal and Transvaal.

Collection Nos. : 71, 75, 325.

184. Tachardia minor, sp. n. (Plate xii, fig. 227).

© test 15 to 2 mm. in diameter, rich castaneous brown, with the larval cast
conspicuous. Viewed from above the test is broad heart-shaped ; edges of caudal
orifice shghtly elevated ; spiracular openings obsolete. In the material examined the
© tests always appeared singly on the twigs, never massed as in 7’. actinella.

¢ puparium about | mm. long, varying in colour from pale yellowish brown to deep
castaneous ; dorsal area strongly ridged.

When cleared and mounted the adult ° is almost circular and very transparent ;
the mouth-parts are comparatively large, averaging about 170u long by 82 at their
greatest width. On each side, extending backward from the middle of the mouth-
parts, is a series of three small conical spines, which appear as though they might
indicate the extreme rudiments of legs. The antennae average 60u in length and are
obscurely 3 or 4-jointed. Lac plates varying from 57 to 68u in length; about 48u
broad ; with few glands, the number varying from 14 to 28 in specimens examined.
Spine about 68 long and 54 across the base, distinctly funnel-shaped, narrowing at
about half its length. Chitinous anal plate almost semi-circular, with the basal
portion rugose and its apex four lobed. Anterior spiracle with 16 simple glands.

Habitat: On rhenosterbosch (Elytropappus rhinocerotis, Less.) ; collected by C. P.
v. d. Merwe, Montague, C. P., November 1914.

Collection No. : 25.

185. Tachardia karroo, sp. n. (Plate xii, fig. 230).

© test 2°5 to 3°5 mm. in diameter, sometimes sing!e, but usually massed together
on the thicker stems of the plant. Somewhat similar to 7. minor but more spherical,
larger and paler in colour. In a number of cases the test is semitransparent, deep
amber-yellow, with three distinct pale lines radiating from the median dorsal ridge to
the margins in the form of a broad Y.

S puparium comparatively short, deep chestnut-brown.

When cleared and mounted the following characters may be determined :—The
antennae are obscurely 3-jointed, about 105u long and 27 wide at the base. The
apical joint has at its tip 3 or 4 short spines. The anterior spiracle has a few (2 or 3)
simple glands. Lac plates about 120u long by 110u broad, with 62 to 84 gland
openings. Spine about 90u long and 78w across the base; deep funnel-shaped,
narrowing about one-third to one-fourth its length from the tip. Anal plate broader
than long ; basal portion coarsely wrinkled, apical part with 4 small pointed lobes.

Habitat: On rhenosterbosch (Elytropappus rhinocerotis, Less.) in association
with 7. minor; collected by C. P. v. d. Merwe, Montague, C. P., November 1914.

Collection No. : 324.

THE COCCIDAE OF SOUTH AFRICA. 125

186. Tachardia decorella (Mask.) (Plate ix, fig. 212 ; Plate xi, fig. 222).

Carteria decorella, Mask., N. Z. Trans. xxv, p. 247, 1892.

Tachardia decorella, Newst., Bull. Ent. Res. vin, p. 127, 1917.

The. ° tests generally coalesce so as to form a mass completely surrounding the thin
twigs of the host-plant, often for a length of 30 to 40 mm. The individual tests, at
maturity, are almost globular, but flattened above, 3 to 4 mm. in diameter, deep
purplish brown, or almost black, with dull grey speckles arranged in radiating ridges.
The intermediate spaces are somewhat glossy.. The lacis very hard and brittle.

The larval tests are deep red. Those of the young females are flat and button-like,
brown, with the reddish larval cast superimposed and with the grey ridges more
pronounced than in the adult.

When mounted the body is broad ovate and transparent. The anal plate is densely
chitinised, about 190 broad and 170u long, with its basal portion embossed with very
small “ grease spot ” design. The mouth-parts are about 185 long and 126 across
the middle. The lac-gland plates are approximately 170u long and have numerous
gland openings, all the specimens examined having over 100. The spine is shallow
funnel-shaped, about 120u long and 112u across the base, becoming abruptly
narrowed at about two-thirds its length from the apex. Near the margins of the
segments are more or less circular groups of simple glands.

This insect differs in some small details, such as the colour of the test, from Maskell’s
description of 7’. decorella, but the marginal gland groups and other characters are so
close that, for the present at least, 1 ascribe his name to the species.

Habitat: On Acacia karroo, Linn., Crocodile River, Transvaal. On Acacia
melanoxylon, Zoological Gardens, Pretoria.

Collection No. : 323, 326.

187. Tachardia affiuens, sp. n. (Plate xi, fig. 224),

Test of adult °° usually found singly on the stems of host-plants ; occasionally in
groups of two or three, never in large groups.

Test + globular, almost as deep as wide, sometimes slightly tapering to the top,
about 3 mm. in diameter, yellowish to dull shellac-brown with a reddish spot in the
centre. The test is generally smooth, but may show indications of faint ridges to
the margins, which are prominent in young specimens.

d puparium exceptionally long and narrow, pale yellowish brown, with reddish
larval cast ; posterior aperture closed by a rough buff flap.

When mounted the characters are indistinct, but the following measurements
have been supplied from the examination of a range of specimens. :—Mouth-parts
about 1704 long and 100u wide. Spine about 100u long and 85y at base, rapidly
tapering so that the apical half is almost parallel-sided. Lac plates small, about
85 long and 75 broad, with approximately 40 gland openings, all in a compact
group. The anal plate is more straight-sided than usual and ends in four almost
equal lobes. Its base is slightly wrinkled aud has scattered transparent spots, giving

it the grease-spot appearance something like the design on the dorsum of a
Pseudaomdia.

126 CHAS. K. BRAIN.

Remarks. This species resembles 7’. longisetosa, Newstead, but is easily distinguished
by the fact that the hairs of the anal ring are of normal length, and the glands of the
lac-plates are fewer in number and not scattered.

Habitat: On Huclea sp. and other native plants ; very common around Pretoria
and apparently widespread throughout the Union.

Collection No.: 76.

i88. Tachardia albida, Ckll. (Plate ix, fig. 211).

Tachardia albida, Ckll., The Entom. xxxiv, p. 249, 1901.

Professor Cockerell’s description 1s as follows :—

“ Forming smooth yellowish-white masses on the twigs ; the extremely dense and
hard lac of the several individuals running together ; masses up to 10 mm. diam.,
and 30 in length. The individuals are marked externally by orange patches, each
presenting a small corrugated or segmented ridge, and an aperture. Cavities for
females globular to subpyriform. Male scales of the usual elongated form, red,
with a very short dorsal segmented ridge, about one-third of total lengths of
scale.

“ ©. After boiling in liquor potassae globular, giving a very fine crimson colour.
Skin after boiling transparent, truncate and caudal processes remaining ferruginous.
Mounted female on slide about 5 mm. diameter. Truncate processes (or ‘lac
tubes’) very short, orifices very small and numerous. Spine apparently absent.
Caudal process peculiar ; transversely oblong or subreniform, with a deep posterior
notch, on each side of which are two lobules ; lateral hind margins bearing a sharp
spine ; surface finely reticulated ; anal ring hidden, only the ends of the numerous
bristles projecting. Spiracles large and well-developed. Mouth-parts well-
developed, but small, about 135u broad; “lobes oraux ” (as figured by Targioni-
Tozzetti in T. lacca) very large.

‘“ Larva in female fusiform, narrow, tapering posteriorly, about 560u long and
240 broad ; caudal bristles very long.”

Habitat : On Acacia karroo, Linn., Verulam, Natal (Fuller No. 5), and Pienaars
River, Transvaal ; collected by the writer, January 1917.

Collection No. : 72.

Subtamily ?
Genus Halimococcus, Ckll.

Halvmococcus, Ckll., The Entom. xxxv, p. 15, 1902.

Only one species has yet been described 7.e., H. lampus, Ckil. Protessor Cockerell’s
original descriptions of the genus and type species are as follows :—

‘““ A Dactylopiine Coccid enclosed in a horny sac shaped like that of Solenococcus,
without legs or antennae in the adult. Larva with no rows of dorsal spines, no
hairs on anal ring, and no caudal tubercles, but four long caudal bristles as in
Phoenicococcus. Closely related to Phoenscococcus (which lives in Algeria), but
distinguished by the form of the sac, which exactly imitates that of Solenococcus.”

THE COCCIDAE OF SOUTH AFRICA. 127

189. Halimecoccus lampas, Ckll. (Plate ix, fig. 204 ; Plate xu, fig. 225).

Halimococcus lampas, Ckll., The Entom. xxxv, p. 15, 1902.

“©. Enclosed in a dark brown horny sac (which is not dissolved by liquor
potassae), which is shaped like a Terebratula shell, 2.e., oval, with the end raised
and terminating in an orifice. Length of sac 510u, breadth 300, breadth of orifice
about 66u. The orifice is closed by a reticulated plate, except basally, where there
is a semilunar opening. In immature examples the sac is prominently segmented
on the ventral side.

“ @. Scale small, cylindrical, horny, ferruginous, of the same texture as that of
the female, but usually somewhat paler. Length 350u, breadth 140. The end
comes off, leaving a round opening, as in Muscid pupae.

“ ©. A mere bag, with well-developed mouth and spiracles.

‘“ Larva. Rather narrow ; legs and antennae present. No caudal tubercles,
but two pairs of long caudal bristles ; two small bristles close to these. Antennae
about 45 long, six-jointed, last jot much the longest. No dorsal spines. Last
antennal joint with two long bristles. Femur remarkably stout, about 15u broad ;
length of femur + trochanter about 30u.

“ Older specimens have actually shorter (36u) antennae, with joint 6 longer than
4 +5; 5longer than 3, 3 longer than 4, | large.

“ A few white curled waxen threads protrude from beneath the sacs of the females.”

Habitat : On leaves of palm (Hyphaene crimta), Durban, Natal ; collected by
C. Fuller about 1901.

Collection No. : 42.

Genus Baeeacoceus, nov.

Alhed to the Lecanttmnaz. Body of adult 9, at maturity, naked, without test or
waxy covering, almost globular, appearing like a berry on the twigs of the host-
plant. In the type species the colour is yellow to orange, often with a faint
metallic bronzy sheen, the whole insect simulating a Margarodes cyst. Antennae
well-develoved, 7-jointed. Legs well-developed, long, upper and lower digitules
present, clubbed. Anal ring large, with numerous (16 2) hairs. Stigmatic clefts,
two on each side, with 4 or 5 broad conical protuberances supphed with glands.

Type, B. elytropappi, sp. n.

190. Baccacoccus elytropappi, sp.n. (Plate ix, fig. 207; Plate xii, fig. 226).

Adult female almost globular, to 3 mm. in diameter, yellow to brownish, glossy,
hike a Margarodes cyst, with the small brown anal plates occupying the position of
the caudal pore of an Asterolecanium test. In the middle of the dorsum there is a
prominence, like a minute keel, occasionally with two rounded prominences in close
proximity but a little posterior to it. There is no fringe in the material at hand,
but four short white bands, two on either side from the stigmatic clefts at the point
where the insect is attached to the thin twig or leaf of the host-plant.

When cleared and mounted the derm is almost hyaline, appearing yellowish,
with the antennae and legs of about the same density. The integument is without
gland-pores and hairs except immediately around the anal plates (fig. 226, d). The

(C605) C

128 CHAS. K. BRAIN.

legs are long and slender, with long slender digitules, which bear spherical knobs.
The lower digitules of the claw are stouter and shorter. The antennae are 7-seg-
mented with joints 3 and 4 long, and almost equal. The stigmatic clefts are broad
and shallow with 4 or 5 low conical glandular protuberances (fig. 226, c). The anal
ring has numerous (10 to 12) long hairs and is closed above by two plates as
illustrated in the figure.

Habitat : On thin twigs of rhenosterbosch (Hlytropappus rhinocerotis, Less.),
French Hoek, C. P. ; collected by C. P. Lounsbury, 1904.

Collection No. : 297.

ee
x ita

F a 1a ary iy 4 rigs yy eH) i a vy

, sheep nih Poa

hh “a
ie ie, ah

uae Sh
‘ El Veit oe =

rie y
mt
i

| a
vie ie
v

EXPLANATION OF PLATE V.

. Chionasprs (Poliaspis) carissae, Cll.

(Phenacaspis) natalensis, Ckll.

39
i. leucadendri, sp. n.

(Dinaspis) diosmae, sp. n.
a (Phenacaspis) lounsburyt var. ekebergiae,

var. Nn.

. Chionaspis (Phenacasprs) lounsburyr, Cooley.

ig (Dinaspis) distencta, Leon.
A leucadendri, sp. n. (see fig. 166).
‘ (Dinaspis) lounsburyi, Leon.

ss sp. on shrub (name of host-plant and
locality not known).

. Chonaspis caffra, sp. n.

35 chionaspitiformis (Newst).
ES scutuae, sp. 0.

a ambiguus, sp. N.

ISWiLjt, JSAP, IRESENAGE, Wel, OX, IBAsGe Ws PLaTE V

del. cx.8

South African Coccide.

AO Morr AVALIX?

yy soi igang | :
‘ fbnastou) rely Paina va Moke ik
stent) Biaevely aa } i

EXPLANATION OF PLATE VI.

Fig. 178. Chionaspis simplex, Green, var.

179. a (Dinaspis) embricata, sp. n.
180. hs (Pinnaspis) cyanogena (CkIL.).
181. * (Pinnaspis) aspidistrae, Sign.
182. Lepidosaphes glovert (Packard).

183. ss pinnaeformes (Bouché).

184. Ischnasyis longirostris, Sign.

years, W/Al.

MOE Me ART 2:

Bucci. ENT. RESEARCH.

South African Coccide.

Aad

6
‘
\ ‘
bart
)
oh
f
%
i
t
fe
i
(

€ A
(ete = {
d
\ 1
ne
7 , 5
, i
=
A %
y int Y i
MS
a)
»
iva =

“i

ad a is A
Vi ivi i i
oe Ati i f

EXPLANATION OF PLATE VII.

Hig. 185. Asterolecanium bambusae, Bdv.

186. et conspicuum, Sp. L.

evs ¢ pustulans, Ckll.

188. s brevispinum, sp. N.

189. - pustulans var. brachylenae, nov.
190. i variolosum, Ratz.

191. = stentae, sp. 0.

192. euryopsis, Fuller.

ISUILIL, JEW, IRissianyecw, Wot. O<, Ieee YD, Piynieis, WAGE.

IS). Size

South African Coccide.

_aeypoloat ‘ rnin ssbopteh "

‘ bil sini pa!

cf ity tel ne Peeters

H siti aa a) f i

i bil allan: gh sal kay
even iat mr
Ne ry i Daa Ny uM

198.

200.

201.

EXPEANATION OF PEATE VIE

. Asterolecanvum variolosum, Ratz.; a, marginal gland

series.
- pustulans var. brachylenae, nov. ;

5
a, marginal gland series.
bambusae, Bdv.; a, marginal gland
series. |
a conspicuum, Sp. N.; a, marginal gland
series.
brevispinum, sp.n, caudal extremity
of 2; a, antennal tubercle.
Fe borboniae, sp. n., caudal extremity of
2; a, 8-shaped glands of derm;
b, antennal tubercle.

» stentae, sp.n.; a, marginal gland
series; 6, antennal tubercle.

e pustulans, Ckll., caudal extremity of
Q,

euryopsis, Fuller, caudal extremity of
©; a, dermal glands.

BULL, JEG, IRESENR@s. WoL, IX, [Paina Ws yas WAUEL

° o®
© Oy, rok eae a © A) CO ROMNG WON CHONamcntS es)
° f
6 @ G&D gy GG g@ a @ 8 @D ED GD GD GP oer i
® 0 ty OGRE Os a

South African Coccidee,

PPT)
Gi
ue

feta

ie
By

oy
an

OY ah

}

rf
i hat
vo

ay il
rage
Wie

oy
ais iss
Vee
Feet!

EXPLANATION OF PLATE IX.

. Lecaniodiaspis nimosae (Mask.)

i tarsalis, Newst.

. Halimococcus lampas, Ckll.
. Lecanodiaspis natalensis, sp. n.

a magna, sp. n.

. Baccacoccus elytropappi, sp. n.

. Tachardia actinella, Ckll.

. Lecaniodiaspis magna, sp. n. (smal! form).
. Cerococcus royenae, sp. Nn.

. Tachardia albida, Ck.

decorella (Mask.).

PMA VX

ENT. RESEARCH. VoL. X. Parr 2.

BONL IC,

South African Coccides.

yey, : ¥ ad a
Pe MR el ah ebm al oy a AP inpay

1dis9 o oe he RON,

od 8 pligunnatins pod iene ie

i. sek nea ln eee
ng aie cuait Pie oa
aes re ce f ( v yi : { ms re x
ov ate py a wires)’ Ole

pe i igh ng * Papnin “ | ©
$i: nich , a ate Oe Ui ee tthe
Hees Iobsies (IED)’ sthevo. dyoninbelh | Lar
ae Ae abe coal ba ee fab |

a '
a
j
ue }
es ; aa
iv - Y
é

apis | AIS yi |

i” ‘
ee
nt
"i
n
2
=
oh
ial
“]
‘
<
¢
th
=
-
Wa fod

were,

Fig. 218.
214.

215.

216.

Zils

218.

219.

220.

EXPLANATION OF PLATE X.

Lecaniodiaspis natalensis, sp. n. ; a, antenna of Q ;
b, eribriform plates.
magna, sp. MN. antennae of 9;
a, rudimentary leg ; b, anal plate, 9.
mimosae (Mask.), antenna of 9;
a, stigmatic spines; 6, cribriform
plate.
Cerococcus passerinae, sp.n., caudal extremity ;
a, antennal tubercle of 9.
Lecaniodiaspis tarsalis, Newst., antenna; a, cribri-
form plates; b, tarsus; c, mar-
ginal spines; d, stigmatic spines
OFS:
brabei, sp.n., antenna ; a, stigmatic
spines; 6, marginal spine; ¢, leg
1, of 9;
Cerococcus ovoides (CkIl.), caudal extremity ;
a, antennal tubercle of 9.
Amorphococcus acaciae, Brain, caudal extremity ;
a, antennal tubercle of 9.

JBiGiijL, JSG, JRIESIe VAG, Wolk, X. IPA B, PLATE X.

South African Coecide.

EXPLANATION OF PLATE XI.

Tachardia actinella, Ckll., lac-gland plate ; a, spine
of 9.
,, decorella (Mask.), lac-gland plate ; a, anal
plate of 9.
ss decorella (Mask.), spine of 9.
a affluens, sp.u., spine of 9.

eyes DX

RESEARCH. VoL. X. ParT 2.

ENT.

SHOVEL.

South African Coccide.

ee STAIN IO ViOITAVAIIKS

Ibe

i Re Ric) barbers pide - 16. Nel
an ’

Be: e wos 8 aft tye abe viyths Mv ive eh: \
‘Avirs Ly 7 Aol ‘ontaa nite i Fh A
a hie Ki A t Ho yg x

(tenlt an sibiulial
ar, ae 7,
i wl

@ os thet ieya lbh Ailoatiton
q any: rus! > , pais fl pay never i

A % bt

pate oy balgy

he bo mLALCTR ef GS Bato TALE
en a ae hata pense ol if aay Nog)
al

wet? Lieb . TAD Miia Rios lo Th,

8 on pouch becca etd fe evi thing: at

’
vt
oe oe ee
sb 143 |
ne on
, a .
mo , :
yy)
tJ
vars
|
‘ / Ee
¥ 1 Se t
AY
ai
rl! 4)
folk wa
itp y' 2
aj veo i
,
Me
Sees 3
. Sas
"eat i
stk We im
= ‘ en

EXPLANATION OF PLATE XII.

Halimococcus lampas, Ckll., dorsal view ; «a, lateral
view; b, stigmatic cleft; c, antenna
Ob 2,

Baccacoccus elytropappi, sp.n. ; a, antenna ; J, leg
i.; ¢, stigmatic cleft; d, anal lobes

of 9.
Tachardia minor, sp.n., lac-gland plate of 9.
. decorella (Mask.), spine; a, anal plate
of 9.

actinella, Ckll., spine of 9.

karroo, sp. n., spine; a, lac-gland plate
of 9.

affluens, sp. n., spme of 9.

decore la (Mask.), lac-gland plate of °.

PEATE) XI,

Wile, 2X, Jena 2.

BuLc. ENT. RESEARCH.

South African Coecide.

ary

ne ae:
ry

oo ne

——— a_i

~ apy Ww
re Ni e 4) ns
Pi. ¢
?
\

’
‘

129

MOSQUITO NOTES.
By F. W. Epwarps.

Anopheles hyrcanus, Pallas.
Culex hyrcanus, Pallas, Reise durch Russischen Reichs, }, p. £75 (1771).
Anopheles sinensis, Wiedemann, Auss. Zweitl. Ins. i, p. 547 (1828).
Anopheles pictus, Loew, Dipt. Beitr. 1, p. 4 (1845).
Anopheles pseudopictus, Grassi, Atti R. Ac. Lincei, vii, 1, pp. 102 & 560 (1899).

The salient points in Pallas’s brief description are as follows :—“‘Cinereus .
Frons hirtella. Antennae triarticulatae ; extremum trinode, basis vero pilis nigris
subpennata. Palmi [? palpi] nulli: ensis nudus . . . basi pilosus. Pedes

longissimi, grisei, posticis corporis fere tripla longitudine, praesertim tarsi, elongati.
Alae lanceolatae, cinerascentes, venis hirsutis ad crassiorem marginem nigro maculatae
Comes prioris [C. caspius, ? =O. dorsalis, Mg.], rarior red ferocior.”

If we make the highly probable assumption that Pallas described the palpi as the
antennae, and that the latter were either broken off or overlooked in his specimen,
the above description becomes understandable, since the sentence referring to the
“antennae ” might very well be used to express the appearance under a low-power
lens of the shaggy palpi of an Anopheles of the Myzorhynchus or Nyssorhynchus
groups. Although no collections of mosquitos have been made since Pallas’s time
in the North Caspian region, it is unlikely that any species, other than A. sinensis,
to which the description could possibly apply, occurs there. The name hyrcanus
will therefore have to be substituted for senensis. |

Anopheles subpictus, Grassi.

Anopheles subpictus, Grassi, Atti R. Acc. Lincei, vin, 1, p. 101 (Feb. 1899).

Anopheles rossi, Giles, J. Trop. Med. 11, p. 63 (Oct. 1899).

Grassi’s short description clearly indicates A. rossi, and the fact that he states that
his description was drawn up from a specimen sent him by Ross from Calcutta
removes any possible doubt which might exist. Itis to be regretted that the operation
of the rule of priority will prevent the commemoration in zoological nomenclature
of Sir Ronald Ross’s work.

Stegomyia fasciata* var. atritarsis, nov.

Tarsi of fore and mid legs almost entirely black, two or three white scales at the ©
bases of the first two joints ; hind tarsi black, with very narrow white rings (narrower
than the diameter) at the bases of the first three and the last joints. White rings
on the male palpi narrower than usual. Markings of thorax, abdomen and femora,
and structure of male genitalia, normal.

Gotp Coast: Accra, vi. 1919,1g 192 (Dr. J. W. Scott Macfie).

Perhaps the most remarkable colour variation yet recorded in this or any mosquito.

* Dr. Guy Marshall urges the retention of this name, rather than the adoption of the
earlier argenteus, Poiret, owing to its wide use in medical literature ; a course which has
been sanctioned in afew cases by the International Commission on Zoological Nomencla-
ture. With this suggestion I readily concur. The names Stegomyia and Ochlerotatus
though used here as genera are to be understood in a subgeneric sense.

(605) D2

130 F. W. EDWARDS.

Ochlerotatus caspius, Pallas.

Culex caspius, Pallas, Reise d. Russ. Reichs, i, p. 475 (1771).

Culeaz dorsalis, Theobald, Mon. Cul. ii, p. 16 (1901) (2 nec Meigen).

Grabhamia dorsalis, Theobald, Mon. Cul. 11, p. 251 (1903).

Grabhamia subtilis, Kd. & Et. Sergent, Bull. Mus. Paris, xi, p. 240 (1905).

Grabhamia willcockst, Theobald, Mon. Cul. iv, p. 294 (1907).

Grabhamia longisquamosa, Theobald, Ann. Mus. Nat. Hung., i, p. 102 (1905).

Mansonia arabica, Giles, J. Trop. Med. p. 130 (1906).

Culex arabicus, Becker, Denkschr. k. Ak. Wiss. Wien, Ixxi, p. 140 (1910).

? Culex maculwentris, Macquart, Dipt. Exot. Supp. 1, p. 7 (1846).

Culex pulchripalpis, Theobald, Mon. Cul. 1, p. 13 (1901) (nec Rondani).

Culex penicillaris, Rondani, Bull. Soc. Ent. Ital. iv, p. 31' (1872) ; Ficalbi, Bull.

Soc. Ent. Ital. xxxi, p. 160 (1869). }

Recent discoveries (see below, under O. currier) make it appear doubtful whether
Meigen’s C. dorsalis has been correctly determined by Theobald and others, but:
as it is an abundant species over a wide area of Hurope and Asia it is desirable to
find a name for it which is not likely to be upset, and I believe that C. caspius, Pallas,
is such a name. Pallas’ original description is of course short and vague, but as
he states that the species was abundant and very vicious in the marshes near Guriev
(North Caspian) it is probable that he met with one of the common European salt-
marsh breeders. The salient phrases in the description, “ Color subgriseus, thorace
cinereo-fasciato . . . pedes subannulati” would seem to indicate the present
species rather than O. currer, while definitely excluding O. salinus. Some confirm-
ation of this conclusion has recently been obtained by Capt. P. A. Buxton’s discovery
of O. dorsalis, Theo., at Resht, South Caspian. Pallas’ statement that the antennae
are ‘‘ utrique sexu filiformes ” may be safely disregarded ; and whatever “ vaginae
multae ” may mean, Theobald’s translation of “ sheath of proboscis snowy white ””
has obviously no justification whatever.

Large series of this species have recently been received at the British Museum
from Italy, Macedonia, Palestine, Egypt and Mesopotamia, which show every
gradation in the variation in the colour of the abdominal scales. In some (as in
most British specimens) the prevailing colour of the abdomen is black ; in the
majority the abdomen has pairs of black spots on each segment, or on the anterior
segments; while in the palest the black is entirely absent, leaving only the ochreous
or whitish ochreous ground-colour. The thoracic coloration varies less ; the two
white longitudinal lines are nearly always present, though sometimes very narrow ;.
in two specimens from Italy, and in the type of G. longisquamosa, they are absent,
the mesonotum then being uniformly fawn-coloured. The amount of dark scaling
on the wing varies very considerably, but the dark scales are always fairly evenly
distributed. No variation is discernible in the male genitalia. The comb-scales
of the larvae may or may not have a differentiated terminal spine, a variation which
has also been noted Ly Dyar in O. currier.

Ochlerotatus caspius var. hargreavesi, nov.

Among a series sent from Taranto, Italy (H. Hargreaves), are six females which
have the whole of the central area of the mesonotum covered with whitish scales,

MOSQUITO NOTES. 131

though in regard to abdominal markings they agree with moderately dark specimens
of the species. Nothing approaching this variation has been seen from elsewhere,
and it therefore seems justifiable to distinguish them under a separate name. It
will be interesting to find whether any difference exists in the male.

Ochlerotatus curriei, Coq.

Culex curriei, Coquillett, Can. Ent. xxxiil, p. 259 (1901).

Culex onondagensis, Felt, N. Y. State Mus. Bull. 79, p. 278 (1904).

Aédes curriet, Dyar, Insec. Inscitiae Menst. v, p. 122 (1917).

? Culex punctatus, Meigen, Klass. 1, p. 6 (1804).

? Culex dorsalis, Meigen, Syst. Beschr. vi, p. 242 (1830).

* Grabhamia broquettu, Theobald, Entomologist, xlvi, p. 179 (1913).

A small series recently received from Wareham Heath, Dorset (H. Scott), proves
on close examination to represent a species distinct from the one we have usually
known as dorsalis, though this latter was taken in company with it and O. salinus,
Fic. The two forms differ as follows :—

O. caspwus, Pallas (dorsalis, auct.). Scales of mesonotum ochreous brown or
fawn-coloured, with two narrow white lines running the whole length. Scales of
proboscis mostly pale except towards base and tip. Abdominal scales usually of
three colours, black, white and ochreous. Wing scales mostly rather broad, dark
ones being evenly spread over all the veins, though in varying proportions with
the light ones. Dark scales more numerous on the femora. Apical part of basal
lobe of side-piece of male genitalia not prominent.

O. curriei, Coq. Scales of mesonotum mostly greyish ochreous ; a dark brown
band of varying width in the middle, extending only about half way from the front ;
patches of the same colour on each side in front. Scales of proboscis mostly black.
Abdominal scales of two colours only, black and greyish white. Wing scales mostly
quite narrow ; at the base of the wing they are all pale ; dark ones predominate on
the apical half of the costa, on the first longitudinal vein, and on the forks of the
fourth, while on the third and fifth longitudinal veins the scales are almost all dark ;
elsewhere they are mostly pale. Femora, except towards tips, almost. all pale.
Apical part of basal lobe of male genitalia prominent.

Meigen’s descriptions of C. punctatus and C. dorsalis do not apply very well either
to O. caspius or to O. curried, though they almost certainly must refer to one or the
other, and perhaps with more probability to the latter. An examination of the
types would be essential to settle the point.

The British examples of O. curriet only differ from North American specimens
in having the dark scales of the wings rather more numerous, especially on the
costa. The male genitalia are identical.

Ochlerotatus rusticus, Rossi.
_ Culex rusticus, Rossi, Fauna Etrusca, 1, p. 333 (1790).
Culex pungens, Robineau-Desvoidy, Mém. Soc. Hist. Nat. Paris, 11, p. 407 (1827).
Culex quadratumaculatus, Macquart, Suit. 4 Buffon, 1, p. 34 (1834).
Culex dwersus, Theobald, Mon. Cul. 1, p. 73 (1901).
Culex quadratumaculatus, Villeneuve, Bull. Soc. Ent. France, p. 55 (1919).

132 F. W. EDWARDS.

On a careful comparison of fresh specimens with Robineau-Desvoidy’s description
I entirely agree with Dr. Villeneuve in his identification. He has however overlooked
the fact that the name quadratimaculatus was an unwarranted substitution by
Macquart for Desvoidy’s pungens ; moreover I consider it is equally evident from
Rossi’s description that his C. rusticus refers to the more strongly marked form of
dwersus with a continuous median dorsal yellow line. This being the oldest name
must replace all the others.

Ochlerotatus antipodeus, sp. nov.

2. Head with narrow golden scales in the middle ; on each side a patch of narrow
black ones, and external to these a small patch of flat white ones ; upright scales
and bristles blackish, except for a pair of golden bristles directed forwards over
the eyes. Proboscis and palpi black-scaled, the latter about one-seventh as long
as the former. Thorax dark brown; mesonotum with blackish bristles and dark
reddish-brown scales ; narrow golden scales arranged in five rather definite lines :
one median, extending from the front margin to just before the scutellum, where
it forks ; a subdorsal pair on the posterior half of the mesonotum ; and a lateral
pair on the anterior third ; some scattered narrow golden scales about the middle,
and a patch of the same above the bases of the wings. Prothoracic lobes with a
few golden scales ; area behind them (proepimerum) with small flat blackish-brown
ones. Mid lobe of scutellum with narrow golden scales in the middle, narrow dark
brown ones laterally ; lateral lobes with a few narrow dark brown scales. Pleurae
with patches of flat white scales. Abdomen clothed mainly with blackish brown
scales ; small white lateral basal patches on each segment and narrow yellowish
white basal bands on segments 2-5. Cerci elongate, blackish. Legs black-scaled ;
femora whitish beneath ; narrow white rings on the bases of the first three tarsal
joints, broadest on the hind legs, where there are also a few white scales at the bases
of the last two jomts. Claw formula 1.1:1.1:0.0. Wangs with blackish scales,
those in the lateral series almost linear. First fork-cell fully twice as long as its
stem, its base a little nearer the wing-base than that of the second. Cross-veins
separated by quite twice the length of the posterior one. Halteres yellowish with
brown knob.

Length (without proboscis), 5 mm.

New ZeauanD: Karikari Bay, 5.1.1916, 149 (incl. type) (Albert E. Brookes) ;
Te Horo, near Wellington, 39 (received through R. F. L. Burton).

Ochlerotatus lepidonotus, sp. nov.

6S. Head with a small area of yellowish narrow curved scales in the middle,
mixed yellow and black upright forked scales on each side of these, and whitish
flat scales more outwardly. Proboscis black. Palp: longer than the proboscis
by the last joint. Long joint mainly pale scaled, but also with many black scales
which are specially numerous towards the base and apex; its apical fourth much
swollen, with long dense hairs on its outer side, which are orange basally, black
apically. Penultimate joint about the size of the swollen portion of the long joint ;
with whitish scales basally, black-scaled apically ; internally and ventrally with long
dense orange hairs with black tips ; a few shorter stiff black hairs on the outer side

MOSQUITO NOTES. — 133

at the apex. Terminal joint slightly longer and distinctly thicker than the
penultimate ; with whitish scales at the base, otherwise black, black-haired.
Thorax with black integument ; mesonotum with pale yellowish scales and bristles
(denuded in middle). Pleurae (except sternopleura) densely clothed with flat
whitish scales. Postnotum with a tuft of pale yellowish narrow curved scales
apically. Abdomen clothed mainly with pale greyish ochreous scales, with some
blackish brown ones intermixed to a varying extent; in the darkest specimen
the pale scales are confined to rather narrow bands across the bases and apices
of the segments, the remainder being all black ; venter mainly whitish. Genitalia :
side-pieces less than four times as long as broad, with rather dense dark brown
hairs on the apical half, but none overhanging the harpagones; external margin.
slightly curved; internal practically straight, except for an indentation just
posterior to the basal lobes.. Basal lobes very small, with a tuft of hairs, none
differentiated ; apical lobes barely distinguishable, with a few short hairs. Claspers
with a straight and rather stout terminal spine. Harpagones just over half as long
as the side-pieces, the stems pubescent, four times as long as the appendage, which
is rather narrow, slightly curved and frayed at the tip. Harpes undivided, a little
shorter than the stem of the harpagones. Legs: femora and tibiae mainly with
pale ochreous scales except towards the tips, where the scales are black. Tarsi
black, except for numerous pale scales towards the base of the first joint. Claw
formula 2.1:2.1:1.1. Wangs very scantily scaled; the scales brown, except
for those on the subcosta, and the base of the costa and first, second, fourth, fifth
and sixth longitudinal veins, which are pale yellowish. Fork cells about as long
as their stems, the upper with its base much beyond that of the lower.

. Palpi nearly a third as long as the proboscis, black, with scattered whitish
scales. Scarcely any black scales on the abdomen. Claw formula 1.1:1.1: 1.1.
Wing-scales rather more numerous, and mostly pale yellowish, except those on the
costa and first longitudinal vein.

Length (without proboscis), 7-8°5 mm.

Maceponia: 43 (including type), 10 9, swept in meadow near Galiko R., Salonica,
26. iv. 1918 (Capt. J. Waterston).

A very distinct species, with one remarkable characteristic, the possession of
scales on the postnotum. In some respects, notably the form of the male palpi,

it resembles O. rustocus, which, it is interesting to note, was taken at the same time
and place.

Ochierotatus (Finlaya) echinus, sp. nov.

Ochlerotatus lateralis, Edwards, Bull. Ent. Res. ii, p. 250 (1911) (nec Meigen).

Ochlerotatus ornatus, Hdwards, Bull. Ent. Res. iii, p. 21 (1912) (? nec Meigen).

Ochlerotatus geniculatus, Edwards, Bull. Ent. Res. iv, p. 49 (1913) (nec Olivier).

Difiers as follows from the common. O. geniculatus :—The four dark stripes of the
mesonotum are brown instead of black, and are separated by narrow golden, instead
of whitish hnes; the sides of the mesonotum are almost pure white, instead of
yellowish white ; the scutellar scales are broad, flat and white; the abdominal seg-
ments have narrow basal yellowish bands in addition to the white lateral spots ;.
and the hind femora have a dark dorsal line on the basal half.

134 F. W. EDWARDS.

Maceponi4 : Stavros, near Salonica, 1918, 19 (type) reared from larva (Capt.
J. Waterston). Morocco: 19 caught on horse and mule outside Fez, v. 1909
(Major C. E. P. Fowler). Aucreria: 19 (Dr. Sergent; no data).

In the absence of a male, the best distinctive character which can be adduced
is the presence of flat scales on the scutellum; the larva also is very different from
that of O. geniculatus, hence this form must undoubtedly be regarded as specifically
distinct. The Moroccan and Algerian specimens have the flat scutellar scales, but
in thoracic markings appear to agree with typical O. geniculatus; hence they are
somewhat doubtfully conspecific with the type.

Culex apicalis, Adams.

Culex apicalis, Adams, Kans. Univ. Sci. Bull. ii, p. 26 (1903).

Culex territans, Howard, Dyar & Knab, Monogr. iti, p. 293 (1915) (nec Walker).

Culex territans, Schneider, Verh. Natf. Ver. Bonn, lxx, p. 45 (1913).

Culex hortensis, Edwards, Ent. Mo. Mag. (3) i, p. 167 (1915) (nec Ficalbi).

Culex saxatilis, Dyar, Insec. Inscitiae Menstr. vii, p. 36 (1919).

Culex pyrenarcus, Brolemann, Ann. Soc. Ent. France, 1918, p. 427 (1919).

Specimens of C. pyrenaicus sent me by Dr. Villeneuve from ‘Rambouillet, France,
agree in almost every detail with North American specimens in the British Museum,
the only distinction discernible being that the harpagones of the male genitalia
are more noticeably serrated on their tips. This difference is so minute that I con-
sider the specific identity of the European and North American forms to be
unquestionable, particularly in view of the fact that the male genitalia vary slightly
among American specimens.

The species can be readily separated in both sexes from C. hortensis, Fic., by the
white spots at the tips of the hind femora and tibiae being very small or absent ;
the wing scales also are perceptibly broader. Capt. J. Waterston has recently
discovered the larvae of both C. apicalis and C. hortensis in the neighbourhood of
Salonica, and from an examination of his material it can be positively stated that
the larva figured by Schneider is that of C. apicalis and not of C. hortensis as I
suggested in 1915.

Dyar, in the paper quoted, discusses the synonymy of the species, and from the
evidence he adduces I should certainly conclude that C. apicalis is the correct name
for the species, C.. testaceus, v. d. Wulp, being too doubtful to be made use of.

Culex aurantapex, Edw.

Culex aurantapex, Edwards, Bull. Ent. Res. v, p. 74 (1914).

Taenwrhynchus domesticus, Leicester, Cul. of Malaya, p. 169 (1908) (nec Culex
domesticus, Germat).

I described this species from a single female from Nairobi. Recently a female and
three males reared from larvae have been received from Dar-es-Salaam (A. W. J.
Pomeroy). Although these specimens differ from the type in having slightly broader
wing-scales and more numerous pale scales on the wings and legs, there seems little
reason to doubt their specific identity with the Nairobi example. They appear to
agree in every respect with specimens of Leicester’s T. domesticus. As Leicester

MOSQUITO NOTES. 135

points out, the wing scales are decidedly narrower than in C. bitaeniorhynchus, a
difference which, together with a slight distinction in the male genitalia, will probably
suffice to distinguish the two forms specifically.

The genitalia are rather peculiar (fig. 1). The apical lobe of the side-pieces has the
flat plate much reduced, little more than a flattened bristle ; in addition there are
four or five undifferentiated bristles and two stout spines. The harpes have only a
minute basal appendage. The “ harpagones”’ are divided into two pairs of strong

Fig. 1. Male genitalia of Culex auraniapex, Edw. ; a, apex of side pieces ;

b, harpes and harpagones, side aes “ the ame, dorsal view (one side
sickle-shaped plates directed dorsally, the ventral pair is the longer, and gives off
ventrally from its base an appendage which divides into a rounded knob and two
strong teeth. In the typical C. bitaeniorhynchus of the Oriental region, as well as
in the var. ethiopicus, Kdw., this ventral appendage of the second harpagonal plate
is much reduced and differently formed.

The var. ethiopicus differs from the typical bitaenrorhynchus chiefly in having
the upper (or basal) sickle-shaped plate of the harpagones much broader than the
lower (or apical). The other points of distinction given (Bull. Ent. Res. ii, p. 30,
1912) do not hold good. Specimens of this form were also reared by Mr. Pomeroy
at Dar-es-Salaam. The larvae seem to be indistinguishable.

Culex watti, sp. nov.

g. Head scales mostly whitish. Palpr longer than the proboscis by nearly the
length of the last two joints, which are densely hairy. Long joint with a narrow
whitish ring before the middle, and another broader one in the middle of the apical
half. Last two jomts with narrow pale basal rings, creamy above, white below ;
hairs at tip of last jomt also pale; penultimate joint with a short white streak on
underside just beyond middle. Proboscis brown, with a narrow, distinct though not
sharply. margined pale ring beyond the middle. Thorax brown-scaled, without
special ornamentation, except that the scales on and in front of the scutellum and
in front of the wing-bases are paler. Abdomen dark brown, the segments with dull
whitish basal bands of about even width, those on segments 6 and 7 considerably
expanded laterally. Genitaiwa (four specimens mounted): side-pieces normal,
rather closely resembling those of C. trifilatus, Edw. (Bull. Ent. Res. v, p. 64),
but the clasper is a little more narrowed apically, and on the lobe the modified
bristles in each set of three (anterior to the plate) are of about equal length; the
one adjoining the plate has a more flattened tip. Uncidistinct, pomted. Harpagones

136 F. W. EDWARDS.

with only two incompletely separated divisions, the first sickle-shaped, the point
directed ventrally and touching the second, which is broad and flattened horizontally.
Harpes with the usual crown of spines, but without any trace of a basal thumb-like
projection ; their sides straight and parallel. Legs brown ; femora whitish beneath
almost to the tips; a distinct pale spot at apex of hind tibiae, which are also obscurely
pale along the inner side ; the faintest suggestion of pale rings at the tarsal articu-
lations, most distinct at the junction of the first and second hind tarsal joints.
Wings with linear-lanceolate brown scales; bases of fork cells practically level.

Q. Resembles the male, but the pale band of the proboscis is much more evident
on the underside than above, and is much broader and very ill-defined basally.
Middle tibiae in one specimen with a very indistinct pale longitudinal stripe in front.

Length (without proboscis), 5 mm.

GOLD Coast: 4¢ (including type) 29, Kumasi, 1919 (Dr. W. G. Wait).

Hast Arrica: 4g 29, Dar-es-Salaam, 30.1.1918 (A. W. J. Pomeroy).

The most obvious characters of this species are the banded proboscis and palpi
of the male, together with the practically unbanded legs. The specimens agree in
most respects with Theobald’s description of C. ataenata ; I have however examined
the type in the Liverpool School of Tropical Medicine, and consider it to be only
C’. univittatus, with which indeed the present species might easily be confused.

The genitalia are most distinct, and clearly show that OC. watti is more nearly
related to C. duttont than to any other African species ; O. duttoni differs chiefly
in having the harpagones undivided, elongate, and bent at right angles in the middle.
In adult coloration C. duttoni differs obviously in the ringed tarsi and striped tibiae.

Theobaldia arctica, sp. nov.

§. Differs from 7. annulata, Schrank, as follows :—Palpi rather more slender
and darker, both hairs and scales on the last two joints being almost all black.
Femora without any trace of pre-apical pale rings; hind tibiae without pre-apical
pale patch on the outer side, but with a narrow longitudinal pale streak along the
inner side ; first tarsal joint of all the legs almost entirely black, only a very few
white scales at the base.

Genitalia: seventh sternite pomted, with an irregular row of ten stout close-set
spines at the tip. Basal lobes of side-pieces small, with two stout bent setae and a
number of hairs. Side-pieces also with a distinct pre-apical prominence on the
inner side which is densely hairy. Harpes rather slender, more swollen at the tip,
where they have four or five small serrations. Harpagones* rather slender, a little
shorter than the harpes, their tips sharply pomted and hooked dorsally. Unci
small, pointed, half as long as the harpagones.

N. Russia: 1g, Archangel, 25.1x.1918 (Capt. A. G. Carment).

In coloration and genital structure this insect agrees almost entirely with
T. alaskaensis, Ludlow, and may in fact be the same. Judging from Dyar and Knab’s

* These are the structures called ‘‘unci’’ by Howard, Dyar & Knab, but as they
are undoubtedly homologous with the “‘ harpagones’”’ of Culex I use this term for them.
The same names are used in quite different senses in different genera of CULICIDAE, and
the terminology of the parts of the male genitalia of these and other Diptera is badly
in need of revision and unification.

MOSQUITO NOTES. LST

figure, however, the harpes are differently toothed and the subapical prominences
and hair-tufts of the side-pieces are more strongly developed than in alaskaensis.
The genitalia of 7. annulata differ in the entire absence of stout spines on the seventh
sternite and of any preapical prominence or hair-tuft on the side-pieces, and in the
differently toothed harpes and less distinctly hooked harpagones.

Theobaldia indica, sp. nov.

Coloration entirely as in T. arctica.

Genitaha : seventh sternite with a group of about seven short stout spines
on its pointed tip. Side-pieces over three times as long as broad ; basal lobes
and apical prominences practically as in 7. arctica. Harpes very slender, very
little expanded apically, the tip being slightly cleft into two, but otherwise
untoothed. Harpagones slightly longer than the harpes and very broad, nearly halt
as broad (vertically) as they are long; a small sharp, backwardly directed hook on
the dorsal margin at some distance before the tip, which is rounded.

PunsaB, Inpia: 1g (type), Bakloh, 28.11.1900 (Capt. Lindesay) ; 19, Umballa,
9.111.1905, and 19, Dalhousie, 4.v.1906 (Col. H. J. W. Barrow).

The specimen recorded by Giles (Gnats, ed. i, p. 206) from Bakloh is probably
this species ; the male recorded by Theobald (Mon. Cul. i, p. 335) is the type described
above. No specimens of the true 7. annulata from India are in the British Museum
collection.

Pe A 3 PRN NaS bev Pe ee ie ac ng jenny j
| zy | | EA ’ Rain tal
ie . ra pm : Rp’ . 4 ue : ie Mi iy a see vin rsatron mr wi migrtall, ‘
aaa mal tps 2 ey, fae hey in ee i | hala a aginey
ms be ‘ pre etd ra! ‘tet e
a) svi wis a bait penne take Ain Nh ik SN
ee re elie, i i yi ie i: Rye Mae: ti aaa Rowe
POC manet bonne at Cabhidst @ a aie Bi nh is te iy | ae ae Ap
Weg i 2 TREY i TA Wie ire Shale: tin bi ae
wis 6 i il asin

Metal ¢

Me mie 3 ra | nih aK i)

ng ae ° . ‘ DAEs Re aan ve he in ua pea ‘yee Hit, oe ae

es Waa Vie 5 i, See Sone. | Galt ee ae ae ‘i, Maley f ;

ail ia Pee Leer ot in!
MS da apa ire ig sia Lis ts AM " mow pon

fetid TaN SR ON Ve AR, ROT Raa Oe aie _ebctiee ar

awa ot! eC ah A AR ine cc nc inane ‘iid ita uate aD.
ee ey Ch AS tig i i ee rg yt yyy s

a I te | “ig a

" er me H
ye i

i a hit a ny diay.

co rp

139

ON SOME AFRICAN DELPHACIDAE (HOMOPTERA).
By F. Murr.

Very little is known about the African DELPHACcIDAE, especially the tropical
forms. For this reason it is worth recording the few species represented in a small
collection made by Mr. J. C. Bridwell at Oloke-Meji, Ibadan, Nigeria, in 1914.
All the specimens were swept from grasses and reeds. It is to be hoped that larger
collections will soon be available and that those who work at them will examine
and give figures or adequate descriptions of the genitalia, especially of the aedeagus.
Owing to the wide distribution of some species and the great difference in colour
of some of the macropterous and brachypterous forms, and to the variation in colour
of other species, and the paucity of good specific distinctions in the chrodtic
characters, it is almost impossible to identify many species of the Delphacides group
without making use of the phallic characters.

The types of the new species are deposited in the collection of the Hawaiian Sugar
Planters’ Association, Honolulu, Hawaii. Measurements are from the apex of
vertex to anus and from the base to apex of one tegmen.

Tropidocephalus brunnipennis, Sign.
One male and one female from Oloke-Meji. The male genitalia of this specimen
are identical with those of specimens from Australia (7. eximeus, Kirk.)

Perkinsiella sp.

One female, which is closely allied to P. bakeri, Muir, from the Philippine Islands,
but I fully expect the male to be distinct. This genus in the Pacific is always attached
to sugar-cane. It would be of interest to know if this is so in Africa, and to ascertain
what parasites keep them in check.

Delphacodes disonymos (Kirk).

Delphax disonymos, Kirkaldy (1907), Haw. Sugar Planters’ Assn. Ent. Bull.
i, p. 151, line 10, p. 155, Pl. xviii, fig. 4.

Delphacodes disonymos (Kirk.), Muir (1917), Proc. Haw. Ent. Soc. iu, 4, p. 333.

Delphax matamtu, Kirkaldy (1907), op. cit. p. 151, line 13, p. 155, Pl. xvi, figs.
4,5, Pl. xvi. fig. 15.

Delphacodes matantu (Kirk.), Muir (1917), op. cot. p. 333; Muir (1918), op. cit.
ii, p. 427, fig. 3. |

I consider these two species of Kirkaldy’s as being the long and short winged forms
of the same species. His two figures of the pygofers are drawn from different positions
and so appear slightly different ; the aedeagi are the same. A similar case of difference
in colour of the tegmina is found, I believe, in D. erectus nigripennis (Crawiord),
which I consider to be the brachypterous form of D. erectus (Crawford).

146 F. MUIR.

Deiphacodes nigeriensis, sp. nov. (fig. 1).

G. Macropterous; head slightly narrower than pronotum, vertex as long as base
is wide, which is slightly broader than apex, base slightly behind the middle of eyes ;
first joint of antennae more than half the length of second (1 to1‘7); lateral carinae
of pronotum slightly divergingly curved, not reaching to the hind margin; first
joint of hind tarsus equal to the other two together.

fee rate EN |
2205 ms
>

Fig. 1. Dicranotropis nigeriensis, Muir, sp. n.;
a, aedeagus, lateral view; b, right genital style.

Face and clypeus between carinae black, carinae light brown; antennae, vertex,
thorax and legs ochraceous tawny, abdomen dark brown. Tegmina hyaline, very
light yellowish, veins darker, with sparse minute granules.

Pygofer similar in outline to that of D. disonymus (Kirk.), and the genital styles
somewhat similar, aedeagus distinct, figured from the right side (figure 1,a); the
left side with a row of small spines from a dorso-apical point to the middle; anal
styles not so large as in D. disonymus, curved and slightly flattened, bases moderately
wide apart.

Length, 1:8 mm. ; tegmen, 2 mm.

NicErta : Oloke-Meji, Ibadan, 1914 (J. C. Bridwell).

Described from one male specimen. I consider this to come near to D. disonymos
(Kirk.), and it is very probable that there is a short-winged form in which the
tegmina are dark shiny black.

Delphacodes bridwelli, sp. nov. (fig. 2).

3. Macropterous; head nearly as broad as pronotum, width (including eyes)
twice the length ; vertex square, base about middle of eye; length of face slightly
more than twice the width, slightly narrowed between eyes and at apex; antennae
reaching slightly beyond the base of clypeus, first joint about half the length of

20.

= er

Fig. 2. elphacodes bridwelli, Muir, sp. n.;

a, aedeagus, lateral view; b, right genital style.
second ; lateral pronotal carinae divergingly curved, not reaching hind margin ;
hind tibiae equal in length to tarsi, first joint of tarsus equal to the other two together.
light brown; black or dark brown between carinae of face and clypeus and over
thoracic pleura, coxae and abdomen. Tegmina hyaline, very light brown, veins
darker ; wings hyaline, veins brown.

ON SOME AFRICAN DELPHACIDAE. 141

Opening of pygofer about as broad as long, dorsal emargination large, anal angles
well produced, slightly curved ; genital styles straight, outer edge slightly sinuous,
inner edge produced into a quadrate process near base, apex truncate ; aedeagus
flattened laterally, broad, apex rounded, an irregular row of spines from near apex
on dorsal aspect across left side to near the medio-ventral area, a somewhat similar
row on right side ; anal spines large, curved, their bases near together.

Length, 2 mm. ; tegmen, 2°2 mm.

NicErta: Oloke-Meji (J. C. Bradwell).

Described from one male.

Dicranotropis bridwelli, sp. nov. (fig. 3).

g. Macropterous; head as broad as prothorax ; width twice the length, including
eyes ; vertex square; length of face nearly twice the width (1 to 1-8), narrowed
between eyes and slightly at apex, median carina furcate about middle; length
of the first joint of antennae slightly more than half the second, terete. Length
of first joint of hind tarsus equal to the other two together; spur as long as first
joint of tarsus, thin, concavo-convex, with many small teeth on hind margin.

Fig. 3. Dicranotropis bridwelli Muir, sp. n.;
a, pygofer, back view; b, do., side view.

Dark chestnut-brown or black; antennae, carinae on head and thorax, rostrum
and legs lighter brown ; base of dorsum of abdomen and middle of apical tergite
hght. Tegmina hyaline, fuscous, a clearer hyaline area over subcostal, radial and
median apical cells, veins of the same colour as membrane, granules small, bearing
small black hairs ; wings hyaline with brown veins.

Pygofer as figured (fig. 3).

Length, 2°9 mm. ; tegmen, 3°8 mm.

NicErRIA : Oloke-Meji, Ibadan, 1914, on coarse grass (J. C. Bridwell).

Described from one male specimen. There are two female specimens in the

collection which may be this species ; they are much lighter, being nearly uniformly
light brown.

Dicranotropis ibadanensis, sp. nov. (fig. 4).

3S. Macropterous; head as wide as pronotum, width 2:2 times the length
including eyes ; vertex square or very slightly wider than long, base slightly behind
middle of eye; length of face twice the width, sides slightly curved, narrowest

142 F. MUIR.

between eyes and at apex, furcation of median carina one-third from base ; antennae
reaching to near middle of clypeus, first joint slightly shorter than second ; first
joint of hind tarsus as long as the other two together, spur slightly shorter than
first joint of tarsus, with many small teeth on hind margin.

/ : /
oe

\
= oe ple aie!
a 4 ¥

Fig. 4. Dicranotropis ibadanrensis, Muir, sp. n.
a, pygofer, back view; 6b, do., side view.

Light brown, nearly black over clypeus, coxae and abdomen ; carinae of thorax
lighter, traces of lighter spots on face. Tegmina hyaline, a curved, fuscous mark
from cross-veins over cubitus extending along veins to the apex of media and
cubitus, veins basad of cross-veins yellowish, distad of cross-veins darker, granules
few and small; wings hyaline with brown veins.

Pygofer as figured (fig. 4).

Length, 1‘°9 mm. ; tegmen, 2.8 mm.

NicEria: Oleke-Meji, Ibadan, 1914 (J. C. Bradwell).

Described from one male.

Genus Megamelus, Hieber.

There are a number of closely allied species of this genus having similar facies,
with a wide distribution. Only a close comparison of series from various localities
will enable one to decide where one species ends and another begins. I sank
M. kolophon (Kirk.) into M. furcifer (Horv.), but I believe it will be better to retain
the name as a variety until more material has been studied. I now recognise three
varieties of this species, viz. :—

1. M. furcifer (Horv.) (fig. 5). The fork at apex of the genital styles small ;
clavus with a small black mark at apex.

Fig. 5. Megamelus furcifer, Horv., (eee style.
6. ie kolophon, Kirk ae #
Tis Ps nigertensis, Muir, var. n.; a, aedeagus ; b, genital
style.

ON SOME AFRICAN DELPHACIDAE. 143

2. M. kolophon (Kirk.) (fig. 6). Fork at the apex of the genital styles much larger ;
no black mark at the apex of clavus ; aedeagus slightly smaller, with smaller spines.

3. M. nigeriensis, var. nov. (figs 7a, 7b). Fork at apex of genital styles wider
and the emargination between them shallower ; a dark mark at the apex of clavus.

These forms come near to Kelisia kirkaldyx, Muir, and Kelisia fieberi, Muir, and
eventually will be placed together in the same genus. Until I can revise these
genera I prefer to let them remain as they are for the present.

Megamelus furcifer, var. nigeriensis, nov. (figs. 7 a, 7 5).

gS. Macropterous; length of head, prothorax and mesothorax about 1°5 times:
the width of the head including eyes, vertex shghtly longer than wide ; length of
face 2°5 times the width, narrowest between eyes, furcation of median carina at
extreme base ; antennae terete, reaching beyond the base of clypeus, second joint
1-4 times the length of first ; first joint of hind tarsus as long as the other two together,
spur longer than first joint of tarsus, large, with many small teeth on hind margin.
Carinae of face and clypeus, vertex, middle and lateral portions of pronotum,
middle of mesonotum, tegulae and base of abdomen yellow; face and clypeus
between carinae, medio-lateral portions of pronotum, lateral portions of mesonotum.
and greater portion of abdomen black ; front and middle legs dark brown, hind legs
lighter brown. Tegmina hyaline, a dark mark at end of clavus and fuscous over
apical veins.

The pygofer is similar to that of M. kolophon, Kirk.,* but the genital styles are
wider at the apex and not so deeply indented, the aedeagus is slightly larger and
the spines stronger.

Length, 2 mm. ; tegmen, 2°3 mm.

Niceria : Oloke-Meji, Ibadan, 1914 (J. C. Bridwell).

Described from two males.

Megamelus flavolineatus, sp. nov. (fig. 8).

3. Macropterous; width of head including eyes 1°6 times the length, nearly
as wide as the pronotum ; length of vertex 1-4 times the width ; antennae reaching
beyond the base of clypeus, first Joint more than half the length of the second.

ms 8b.
oe a \°

Fig. 8. Megamelus flavolineatus, Muir, sp. n.; a, aedeagus; b, anal
spines ; ¢, genital style.

(1 to 1°7); length of face 2°5 times the width, shghtly narrowed between eyes,.
sides subparallel, median carina forked at middle of face ; first joint of hind tarsus.
as long as the other two together, spur nearly as long as the first joint of tarsus,
large, thin, with many small teeth on the hind margin.

* Haw. Sugar Planters Assn. Ent. Bull. 1, pl. xv, figs. 9-10.
(C605) E

144 F. MUIR.

Dark brown or black; vertex, carinae of head, middle and lateral margins of
pronotum, middle of mesonotum, tegulae, antennae and legs yellow or light brown.
Tegmina hyaline, veins yellow or light brown, a dark mark at apex of clavus, fuscous
over apical portion of cubital veins.

Aedeagus short, deep at base where it is laterally compressed, a row of three or
four spines on the right side and another somewhat similar row on the left; anal
spines large, shghtly curved and pointed ; genital style as figured (fig. 8, c).

Length, 1-8 mm. ; tegmen, 2°5 mm.

. Macropterous ; similar in colour to the male.

Length, 2°0 mm. ; tegmen, 3°0 mm.

Nigeria: Oloke-Meji, Ibadan, 1914 (J. C. Bridwell).

Described from six males and eleven females. Some of the specimens are lighter
in colour than the type, especially between the carinae of clypeus and genae.

Phyltlodinus badius, sp. nov.

©. Macropterous; head as wide as pronotum, width 2°4 times the length;
vertex wider than long; length of face double the width, narrowest between eyes
and at apex, median carina furcate slightly distad of middle; antennae terete,
reaching shghtly beyond the base of clypeus, first joint slightly shorter than second ;
femora and tibiae of first and second legs and femora of hind legs flattened and
expanded ; lateral pronotal carinae divergingly curved, not reaching the hind
margin. First joint of hind tarsus equal in length to the other two together ; spur
as long as first joint of tarsus, many small teeth on hind margin.

Dark chestnut or nearly black; carinae of head and thorax, second joint of
antennae, middle and lateral margins of pronotum, middle of mesonotum, four or
five pairs of small spots on face, middle and hind femora and hind tibiae lighter
brown. Tegmina hyaline; before the cross-veins the veins are light narrowly
bordered with brown, with small hght granules; beyond the cross-veins the veins
are of the same colour as the membrane, which is brown over the cubital area and
over the radial apical veins. spreading into the membrane ; wings hyaline. with
brown veins.

Length. 2°9 mm. ; tegmen, 4°0 mm.

. Brachypterous; n colour similar to the macropterous form. As in most
brachypterous Delphacids the reduction is mainly in the apical cells. The middle
of the membrane between the veins is raised in longitudinal ridges, giving the
tegmina the appearance of having numerous veins.

Length 2°7 mm.; tegmen. 2°00 mm.

NicgERIA: Oloke-Meji, Ibadan, 1914 (J. C. Bridwell).

Described from two macropterous and one brachypterous females, the type being
one of theformer. The second macropterous specimen fs darker than the type.

145

INSECTS INJURIOUS TO ECONOMIC CROPS IN THE ZANZIBAR
PROTECTORATE.
By W. MANSFIELD-ADERS,
Government Economic Biologist, Zanzibar.
(Purates XIII-XV.)

The following notes on the various insects injurious to the economic crops of the
Zanzibar Protectorate are based on investigations made during the last few years.

The chief economic products of the two islands (Zanzibar and Pemba) are cloves
and coconuts, the former being of greater importance.

INSECTS AFFECTING CLOVES.

Fortunately no insect pests of the clove tree (Hugenia caryophyllata) have been
discovered, but it is apparent in many plantations that a number of trees are
unthrifty and die. The cause of death still remains uncertain, but most of the
evidence points to some physiological condition, such as waterlogging of the roots
near swamps, denudation of alluvial soil on slopes, or the tap-root reaching an
impervious layer of stone. The bark of certain unhealthy trees is attacked by
termites (Termes bellicosus) and dead branches are infested with various species
of Bostrychid beetles. Some young clove trees from Pemba had been killed by
T. bellacosus feeding on the bark of the lateral roots.

The adults of a large Tenebrionid beetle, Pycnocerus passerinw, Bertol., have
been found on several occasions lurking under dead bark on clove trees, while
larvae and pupae occur in deep tree-holes filled with decomposing vegetation ;
I am of opinion that they do no actual damage, but only feed on dead bark, etc.

Several experiments have been undertaken to test whether various omnivorous
Lepidopterous larvae, such as those of Huproctis producta, would feed on clove
trees when deprived of other food, but all died.

INSECTS AFFECTING COCONUTS.

Oryctes monoceros and O. boas. These two rhinoceros beetles are common through-
out the two islands and are a serious menace to the coconut industry, the former
being the commoner species. The damage done to young trees is often very serious ;
in some plantations more than half of the trees from 23 to 3 years old are killed,
while others are stunted and their productiveness much reduced. A certain number
of trees survive initial attacks, but as their growth is retarded, they take a long
time to reach maturity. Mature trees rarely succumb to the attacks of the beetles,
although a number of felled trees were found to harbour the insects.

The following record of the life-history of Oryctes monoceros has been obtained :—

Eggs found in dead coconut trunk .. tne beg lee Vai HOTT,
Kegs hatched .. oe si " PY Hee [Os Viti LAs.
Larvae started to build pupal cell .. i sal) AGe xan LOM Ge
Pupal cell completed and pupa formed * JAY BOyaa.) VOUT:
Adults emerged ue ae My Pee. Mule ho tT

(C605) E2

146 W. MANSFIELD-ADERS.

It will be seen from this that the length of the larval life is 113 days, that of the
pupa 18 days.

After the mature larvae had built their pupal cell they shrunk considerably in
size and became very sluggish before pupating ; this period of inertia lasted 5 days.

Pupae are often found in association with larvae in old coconut trunks, generally
among the coarse fibres where disintegration has not yet taken place; the larvae
prefer the centre, where the fibres have been reduced to the consistency of a warm
moist debris.

The larvae of Oryctes boas are more usually found in rubbish heaps composed
of decomposing vegetable matter and manure. I have found manure heaps riddled
with larvae of all ages ; horse and donkey droppings seem the most attractive.

The following preventive measures have been tried :—

(1). Traps have been set in various localities filled with rotting coconut trunks,
decaying vegetation, etc.; the addition of a little horse manure enhances their
attractiveness and they have given excellent results.

(2). The collection of adults and larvae by natives has been tried in Hast Africa,
but the results were not encouraging.

(3). In small plantations the daily examination of all young trees (from 24 to.
3 years old) for adult beetles should be quite efficacious. The easiest method is to
insert a small piece of wire with harpoon points into the entrance hole, transfix
the beetle and withdraw it. After a little experience the palms harbouring beetles
are easily recognised by the moist tow-like frass protruding from the entry holes.
Sprinkling dry earth or sand in the crowns of the young trees may possibly act
as a deterrent to adult beetles. There is no doubt that bare plantations and isolated
trees are more severely attacked, probably because the beetle being a heavy sluggish
flier is able to alight more easily on its food-plant under these conditions. The
planting of some tall bushy catch-crop, such as Cajanus wdicus, around the
plantation might be efficacious.

Some imported larvae infected with a fungus (Metarrhizwum amsopliae) were
received, and it was proved that our local larvae are susceptible to the disease ;
further, in some control experiments one case of actual indigenous infection was
found. A long series of experiments with a view to introducing this fungus was
planned, but owing to unforeseen circumstances all work in this direction was
abandoned.

I have on several occasions seen young trees suffering from what might be called
bud-rot. The whole of the crown is easily pulled out leaving a hollow depression
full of dark fluid with a most foetid odour. It may perhaps be proved later that
bud-rot is secondary to the attacks of the beetle.

To summarise the above :—The rhinoceros beetles (Oryctes monoceros and O. boas)
are common throughout Zanzibar and Pemba, and are the most serious pests of the
coconut industry. Young trees from 24 to 3 years old are most usually attacked,
many are killed, others greatly delayed in reaching maturity. Trees in isolated
positions and on bad soil are more often attacked. Fully matured trees harbour
adult beetles, but the damage to them is not serious. The most useful preventive
measure is the trapping of larvae in pits filled with rotting coconut trunks, vegetable
debris and a little manure.

INSECTS INJURIOUS TO CROPS IN ZANZIBAR. 147

No insect parasites have been found preying on the eggs, larvae, pupae or adults.

Among mammals the most useful in destroying larvae are the MAcroscELIDAE
and SoricipaEz. The following list includes all known to me.

Rhynchocyon adersi. A large elephant shrew, which is somewhat local in its
distribution and chiefly found in scrub on the east coast of the island. It has not
been observed feeding on larvae under natural conditions, but partook of them
greedily in captivity.

Petrodomus sultan. A rare species, concerning the habits of which no observations
have been obtained.

Pachyura murina. These small shrews are widely distributed throughout the
island and have been found in beetle traps on several occasions.

Two species of crows (Corvus scapulatus and Corvus splendens) have been noticed
feeding on the larvae of Oryctes boas in manure heaps. Many species of birds have
been collected and their stomachs carefully examined for larvae or fragments of
adults, but none have been found.

The following are minor pests of coconut palms :—
Rhynchota.

Aspidiotus destructor, Sign. This dangerous coconut scale-insect has been found
occasionally on husks and leaves, but at present does no appreciable damage.

Aspidiotus cyanophyllr, Sign., A. lataniae, Sign., and Hemichionaspis minor, Mask.,
have all been found on the husks of coconuts in association with Aspodiotus destructor.

Cerataphis lataniae, Licht. This cosmopolitan Aphid has been found frequently
on the leaves of coconut palms but never in great numbers. Ornamental palms
of various varieties are heavily infested.

Coleoptera.

Rhina amplicollis, Gerst. Aduits, larvae and pupae of this large weevil were
found in a dead coconut tree in Pemba.

MNocalandra frumenti, F. These small Curculionids were found innumbers in crevices
on coconut husks. The infested nuts showed in the majority of cases cracks and
fissures from which gum was exuding. Attention had been drawn to the trees
on account of the shedding of immature nuts and peculiar deformities in the shape
of the nut, but it is by no means certain that the injuries were due to the weevils,
which were probably secondary visitors.

Rhyncophorus phoenicis, ¥. This large weevil has been recorded from Pemba.
Larvae, pupae and adults were found in an old stump of a coconut tree, others

in a dead oil palm (laeis guineensis). As far as can be ascertained at present it
is not a serious pest of coconuts.

Isoptera.

Termes bellicosus, Smeath. Termites often attack seed-beds and cause great
havoc among the nuts, in some instances 50°% of them having been destroyed.
The soaking of nuts in Cooper’s dip before planting gave good results. Mr. T. J.
Anderson, Government Entomologist in British East Africa, advised watering
the seed-beds with sea-water, but as the beds attacked were situated inland a
considerable distance from the sea, the method has not been tested.

148 W. MANSFIELD-ADERS.

INSECTS AFFECTING COTTON.

A few experimental plants of the following varieties have been grown, viz.,
Egyptian :—Abassi, Nubari, Yannovitch, Mit Affifi, Asil, Sakellarides ; American :—
Allen’s Long Staple and Sunflower.

All the plants were heavily infested by a variety of insects of which the following

proved the most injurious.
ae

Lepidoptera.

Pectinophora (Gelechia) gossypiella, Saunders. This has proved to be the most
serious pest, as every variety of cotton was attacked. All stages of development
were taken throughout the year. Other food-plants were the silk-cotton tree
(Eriodendron anfractuosum) and Ladies’ Fingers (Hibiscus esculentus), the latter
plant being always heavily infested.

Pyroderces simplex, Wism. Some adults emerged from stored seed.

Earias insulana, Boisd. A few specimens were obtained from all varieties of
cotton. A very common pest of Hibiscus esculentus.

Sylepta derogata, F. This leaf-roller occurred abundantly on all varieties and was
responsible for serious damage.

Prodenia litura, F. Remarkably few larvae were found. Other food-plants
are Hibiscus esculentus, young tobacco plants, castor plants, and on one occasion
a ripe tomato fruit.

Acrocercops bifasciata, Wism. Cotton plants of all ages were attacked by this
leaf-miner. The attacked leaves have a very characteristic appearance, the whole
of the upper epidermis being raised to form one large blister.

Bucculatrix lozoptila, Meyr. Larvae were taken feeding on the upper epidermis
of Abassi cotton ; they construct a tough white cocoon shaped like an upturned
boat on the leaf.

Orgyia vetusta, Hmp.* One small swarm of larvae appeared, completely defoliatmg
the attacked plants.

Euproctis producta, Wlk. Larvae occasionally swarm on cotton; they are
omnivorous, moving from one food-plant to another, castor oil plants seeming to be
their favorite food.

Rhynchota.

Dysdercus fasciatus, F. This stainer was prevalent during the bolling season
and caused much damage to the lint ; all varieties of cotton were attacked, especially
Caravonica. These insects were observed plunging their proboscis into young
oily seeds in bolls which had opened prematurely. They were also seen on many
occasions feeding with avidity on fresh mammalian carcases, skins and skulls.

Dysdercus superstitiosus, F. This has been taken occasionally, but is by no means
a common species.

Oxycarenus albidipennis, Stal. All instars are common in well matured open
bolls. In neglected crops they occur in vast numbers, although the actual damage
done seems to be negligible.

INSECTS INJURIOUS TO CROPS IN ZANZIBAR. 149

Pseudococcus obtusus, Newst. At times this is a serious pest. It is one of the
commonest Coccids of the island and feeds on a variety of hosts. Many experimental
plants of cotton were completely covered, the main stem being smothered.

Pseudococcus virgatus, Ckll. Found occasionally on leaves and beneath the
sepals. This Coccid is very abundant on a number of ornamental plants and shrubs.

Pseudococcus citri, Risso. A rare species on cotton.

INSECTS AFFECTING CEREALS.

The principal cereals cultivated are rice, maize, and millet (Sorghum vulgare),
none of them in any great quantity.

Lepidoptera.

Cirphis loreyi, Dup. The larvae of this Noctuid are common in all maize
fields. They bore and tunnel through the main stalk, attack the base of the cob,
and occasionally feed on unripe seeds. Pupation takes place either in the stalk
orin the cob. The moist white excrement, very typical of the presence of the larva,
is generally found at the base of the leaves or cob. The large exit holes of the imago
are easily detected. The species also attacks millet.

Sesamia calamistis, Hmp. Very similar in its habits to Curphis. In young cobs
the tassel is often attacked, the larvae working downwards into the cob ; it likewise
feeds on millet.

Busseola fusca, Hmp. Larvae and pupae were found in the main stalk of millet,
but never in great numbers.

Chilo suppressalis, Wik. This species is the commonest borer of maize and the
most harmful, as in some instances it will destroy a full half of the crop; it also
attacks millet.

Parnara mathias, F. Larvae of this butterfly have been found occasionally
feeding on the leaves, but as a pest it is insignificant.

Rhynchota.

Peregrinus maidis, Ashm. A field of imported maize (Hickory King) was found
heavily infested with this cosmopolitan pest ; but the common native variety of
maize is rarely attacked.

INSECTS INJURIOUS TO VEGETABLES.
CABBAGES.

Crocidolomia binotalis, Z. Cabbages are always heavily infested with the larvae
of this moth and several valuable crops have been completely ruined. Pupation
takes place on the leaf under a light silken web or in the rolled end of a leaf. The
pupal stage averages 10 days. Other food-plants are turnip, watercress, radish
and nasturtium.

Pluiella maculipennis, Curt. The small brilliant green larvae of the diamond-
back moth were taken in conjunction with those of the former species. Pupae
are found on the leaf in a silken web.

Phytometra signata, F. This moth is a minor pest.

150 W. MANSFIELD-ADERS.
Eeag Piant (Solanum meiongena).

Acanthocoris fasciculatus, F. This bug is a serious pest. In many market-
gardens these useful plants are smothered with it ; the tips of the voung shoots wilt
quickly when heavily infested. It occurs as a minor pest of the chilli (Capsicum sp.).

CUCURBITACEAE.

Dacus vertebratus, Bezzi. The larvae of this fruit-fly are extremely common
in all varieties of native gourds and cause a large amount of damage. Other food-
plants are Luffa acutangula and various squashes. Market-gardeners have attempted
to grow melons and cucumbers, but very few reach maturity owing to the ravages
of this fly.

Two other species of Dacus also occur in gourds, D. brevistylus, Bezzi, and Dacus
punctatifrons, Karsch.

PicEON PEA (Cajanus indicus).
Rhynchota.

Pseudococcus obtusus, Newst. This insect, one of the commonest Coccids of the
island, is found in enormous numbers on this useful plant.

Coleoptera.

Tragocephala variegata, Bert. Adults of this Longicorn are prevalent during the
hot season from October to March. Eggs found in the field in April hatched in
10 days ; larvae have been kept in captivity for 3 months without pupating ; pupae
taken in the field produced imagos in 12 days. The female gnaws a small hole
in the bark of a lateral branch and there deposits an egg. The larvae feed for a
short time in this branch, then tunnel down the main stalk, making many vent
holes during their transit and eventually pupating in a plugged cell low down in
the trunk or even in the root. A large exit hole denotes the escape of the adult.

a Lyprops breviusculus, Gerst. Numbers of these beetles have been found feeding
“ on the peas. All instars are present in the pods. They are generally found in
neglected fields where old pods have been left on the trees.

Several species of MrtorpAr feed on the leaves but are only minor pests. The
following have been identified ;—Mylabris dicincta, Bert., Mylabris amplectens,
Gerst., and Coryna ambigua, Gerst.

Lepidoptera.

Marasmarcha atomosa, W\sm. These small delicate moths have been taken
towards dusk hovering around the pods. The characteristic larva, studded with
spines and hairs radiating from tubercles, feeds through a hole on the pea. The
pupa closely resembles the larva and is generally found on the outside of the pod.

Deudoriz antalus, Hopfi. The large flat conspicuous larvae of this butterfly are
an occasional pest of the pigeon pea. They are voracious feeders and attack the
seeds in young green pods with avidity, at times causing a considerable amount
of damage. Pupation takes place in the pod.

INSECTS INJURIOUS TO CROPS IN ZANZIBAR. 151

Diptera.
Agromyza sp. The larvae of this fly attack the young green seeds in the pod,
making a ring-like track around the seed ; the pupae are formed within the pod.

Lapigs’ Fincers (Hibiscus esculentus).

These useful plants harbour a number of important insect pests of which the
following have been recorded :—Pectinophora gossypiella, Harias insulana, Sylepta
derogata and Prodenia litura ; as previously mentioned all of these occur on cotton.

Nisotra weiset, Jac. Numbers of these small flea-beetles swarm on the leaves,
giving them a typical shot-hole appearance. The same species has been found on
roselle (Hibiscus sabdariffa).

Sweet Potato (Ipomoea batatas).

Cyclas formicarius, F. This cosmopolitan weevil is a common pest throughout
the two islands, many tubers being completely riddled and destroyed. All instars
are found in the tubers.

Asyidomorpha puncticosta, Boh. The leaves are often severely attacked fe the
larvae of this Cassidid beetle, which are conspicuous on account of their lone anal
process.

Cassava (Mambhot utihssema).
This useful plant, the chief food supply of the natives, is exceptionally free from
‘pests.
Pseudococcus (Dactylopius) virgatus var. madagascarensis, Newst. A few plants have
been observed to be infested with this Coccid.

Tetranychus sp. This small red mite is abundant in some plots, and causes
‘wilting and curling of the leaves.

INSECTS INJURIOUS TO FRUIT TREES.
CITRUS.
‘Rhynchota.

Lepidosaphes becku, Newman. This scale-insect is the worst enemy of Citrus,
and in many plantations the trees are thickly encrusted. The main stem and laterals
are the favourite sites, but leaves and fruit are also attacked. I have seen numbers
of oranges packed ready for export infested with this scale.

Icerya purchast, Maskell. Fortunately very few trees are attacked by this Coccid,
which at present is quite a minor pest.

Icerya seychellarum, Westw. Has been observed on lemons.

Pseudococcus obtusus, Newst. This Coccid has been taken on all varieties of
orange, in addition to its other food-plants.

Pseudococcus citrt, Risso. A minor pest which has never been found in abundance.

Ceroplastes rubens, Maskell. One record from a young orange tree.

Coccus viridis, L. This scale is very prevalent on young trees in some plantations
and shows a marked partiality for imported species of Citrus.

1.52 W. MANSFIELD-ADERS.

Aspidiotus (Pseudaomdia) trilobituformis, Green. Occasionally met with; this
Coccid has a variety of hosts.

Cerataphis latanae, Licht. This is another of the commonest citrus pests; the
leaves of many trees both old and young were found to be heavily infested. This
Aphid is also abundant on many varieties of ornamental palms.

Aphis tavaresi, del G. The black cotton aphis is extremely seasonal in its
occurrence and is generally found on the leaves and young tender shoots.

Lepidoptera.

Argyroploce leucotrzta, Meyr. (Citrus Codling Moth). Larvae have been found
on several occasions in mandarins, but not in oranges.

Papilio demodocus, Hsp. A very common pest on all young citrus trees. Many
voung plants are completely defoliated, and the larvae are occasionally found feeding
on tender shoots of mature trees.

Coleoptera.

Tragocephala variegata, Bert. The habits of this Longicorn have been described
under pests of the pigeon pea. It attacks all varieties of Crtrus.

Porphyronota maculatissima, Boh. A few adults of this unimportant Cetonud
beetle have been taken from time to time feeding on the leaves.

Gyponychus cervinus, Gerst. A number of adults of this weevil have been
obtained, they attacked the leaves of young orange trees ; a minor pest.

Diptera.

Ceratitis capitata, Wd. This serious pest does not seem to be very widespread.
Oranges and mandarins of imported varieties were found to be infested, but
indigenous trees (oranges, mandarins and lemons) are far less susceptible to the
attacks of this fruit-fly.

| MANGOs.

Sternochetus (Cryptorrhynchus) mangiferac, F. Adults are commonly found in
the stone of the fruit, larvae and pupae in fallen immature fruit.

The following Coccrpar have been found on the leaves and fruit :—Asprdrotus
(Chrysomphalus) dictyospermi, Morg., Aspidiotus destructor, Sign., Pseudococcus
obtusus, Newst., Lecanium adersi, Newst., L. (Saissetia) nigrum, Nietn., L. (Sacssetia)
punctuliferum, Green.

Of these the only one of importance is Pseudococcus obtusus, which heavily infests
both the leaves and fruit of many trees.

BANANAS.

Aspidiotus destructor, Sign. Has been found occasionally both on the fruit and
leaves.

SOURSOP.

Ceratitis rosa, Karsch. One fruit was found to harbour larvae of this fly.

INSECTS INJURIOUS TO CROPS IN ZANZIBAR. 153:

INSECTS AFFECTING SHADE TREES.
Lepidoptera.

The African almond (Lerminalhia catappa) is common throughout the two islands
and is much appreciated for its ornamental foliage and shade properties. In certain.
situations, especially in the town, it is severely attacked by bag-worms (PSYCHIDAE).
Unthrifty trees are heavily infested, most of the leaves being almost skeletonised.

Miresa melanosticta, Baker. A minor pest of the African almond. The large
green slug-like larvae with lateral tubercles are not easily detected in spite of their size.
They have the peculiar habit of arranging themselves in circles on the leaves, their
heads facing inwards. Pupation takes place in the soil in a hard round cocoon,
with a distinct lid.

Curia forda, Westw. The larvae feed on young casuarina trees (Casuarina
equisetifolia) and are easily detected by their conspicuous yellow markings. Pupation
takes place in the soil at a depth of about 3 inches. Another food-plant is the
naseberry.

Asura saginaria, Wik. The larvae of this moth have have been taken occasionally
on cinnamon trees, but itis quite a minor pest.

Rhynchota.

Pseudococcus obtusus, Newst. When in poor soil casuarina trees are often
heavily infested with this scale-insect, the lower branches being principally attacked.

Aspidiotus (Pseudaonidia) trilobitiformis, Green. Many rubber trees (Ficus elastica).
are heavily infested with this Coccid.

Dysdercus swperstitiosus, F. The open bolls of the silk-cotton tree (Hriodendron
anfractuosum) are often attacked by this common cotton-stainer, and D. fasciatus
is found in conjunction with it.

INSECTS AFFECTING TIMBER.
Coleoptera.

Macrotoma palmata, F. The larvae of this Longicorn are a serious pest in the
mangrove wood rafters in native huts. On one occasion several larvae were
extracted from large tunnels which they had made in a piano case. Numbers
have been found in the field in old dead mango stumps.

Dinoderus minutus, F. Extremely common and destructive to dried bamboo
and a very serious pest. Adults have been found in Madagascar teak, a shght
attack. On one occasion the bark of an avocado pear (Persea gratissima) was found
to be heavily infested with it.

Isoptera.

Termes bellicosus, Smeath. Ubiquitous in its distribution, the white ant is the
worst timber pest of the two islands. Imported European soft woods are reduced
to ruin in a few years, though Indian teak is rarely attacked. The structural timber
of the native huts (mangrove, Ceriops candolliana) and the roofing made of plaited
coconut leaf are often attacked.

Termites have been reported as injuring the following living trees :—Seed and
seedling coconuts, young avocado pears (the roots badly attacked), and cloves.

154 W. MANSFIELD-ADERS.
INSECTS INJURIOUS TO MISCELLANEOUS PLAN®SS.

Lepidoptera.

Thalassoides digressa, Walk. The larvae occasionally feed on the leaves of the
castor plant.

Duomitus capensis, Wik. The larvae have been found tunnelling in the main
stem of the castor plant.

Brithys pancratu, Cyr. The caterpillars are voracious feeders on all species of
hhes ; when young they are gregarious and feed under the epidermis of the leat,
later they separate and devour the whole plant. Pupation takes place in the soil
or at the base of the leaves, the pupal stage averaging from 8 to 10 days.

Euchromia formosa, Guér. The larva is clothed with tufts of hairs, dull-coloured
and inconspicuous, and forms a cocoon of silk and hairs on the branches of the
food-plant. 1tis common on all species of creepers belonging to the genus [pomoea.

Glyphodes sericea, Drury. The larvae are common on gardenias. Adult larvae
are of a transparent green colour with four black spots on the dorsum. They are
leaf-rollers and live and pupate in twisted leaves, their pupal stage averaging 9

days.

Coleoptera.

Entyposis impressa, Kolbe. Larvae of this weevil have been recorded as feeding
on the roots of castor plants and caladiums.

Brachycerus atrox, Gerst. Great numbers of adult weevils have been taken in
the soil around the tubers of lilies (Amaryllis sp.). During the day they hide in
burrows near their food-plant and come out to feed on the leaves at night. Larvae
feed on the tubers, eating into the centre.

Mausoleopis amabilis, Schaum. Adults of this Cetonid are common in the
blossoms of various ornamental plants, especially roses.

Rhynchota.

Aspidiotus (Chrysomphalus) aonidum, L. A common scale-insect on rose stalks.

Pseudococcus virgatus, Ckll. Has been recorded from various ornamental
creepers, etc.

Orthoptera.

Chrotogonus hemrpterus, Schaum. A very troublesome grasshopper; one
experimental field of castor was ruined by its depredations.

INSECTS INJURIOUS TO STORED PRODUCTS.

In the tropics, where food is abundant throughout the year, insects attacking
stored grain are more in evidence than in Europe. Owing to the custom of storing
grain in loose receptacles and exposing it in open bins for sale nearly every sample
showed evidence of insect attack.

INSECTS INJURIOUS TO CROPS IN ZANZIBAR. 155

Coleoptera.

Calandra oryzae, L. This cosmopolitan pest of gram is extremely common in
rice. and is also found in maize and sorghum.

Laemophloeus pusillus, Schénh. Very abundant in maize.

Tribolium castaneum, Hbst. Common in maize and occasionally found in rice.

Silvanus surinamensis, L. Abundant in maize.

Tenebroides mauritanicus, L. An occasional pest of maize.

Bruchus chinensis, L. A serious pest of all beans and pulses. Numbers are
always to be found in Phaseolus mungo and Cajanus wmdicus.

Bruchus ornatus, Fhs. Not so common as B. chinensis, but has been recorded
from beans.

Carpophilus humeralis, F. Very common in maize cobs in the field, generally
in those attacked by fungus.

Cossonus suturalis, Boh. An occasional pest of stored sweet potatoes.

Dermestes vulpinus, F. Abundant in dried fish.

Latheticus oryzae, Waterh. An occasional pest of maize.

Rhizopertha domanica, F. Occasionally in maize.

Necrobia rufipes, de G. Both larvae and adults are abundant in dried copra,
feeding on the kernels; the former bore long cylindrical tunnels into the dried
copra.

Lepidoptera.
Ephestia cautella, Wik. Larvae very common jn rice and various flours.

atl

pees ;

iSesqp ly
am Le er 4 z

ate

: ‘
My An

th
oy

, ie i an
2h

: ae a [ato pee. a ern “ase,

|
5
oe 4, it 4
i A e
, aad x,

i" wie ‘ i

' aw

fs yy a

Ph i2 a

Vicoitanis wis as
‘waening ae phils

Rik My

yt ak «Hie pune |

%
vy ~
f= i cy :
J J ' ot
: ‘ a oe DO
J
.
} f }
i
{
4
A

Geena. ern, ee a

ny a ‘ enn .

8-No

2 Aha AL Heaney ON
a sire afi Bacar”

bana he
vy ae Loar: ie es

oe eh laitagls WP: Wh aibtodh

e. NOW a vellacaenbtersinh he

fay ut fads, cay Pym noc
me Hino! gabe txeep al

D iy . A ry nae oye a

é % Se lei 7 k :
‘tite alle Sivien sivas Bes a

= 7 ay PY v
vee, deri hy aa

dan ee heyy

r 7 eR TP Oy.

od
- iat

‘

Pree aa if a
Lah

BULL, IEW, IRESENRGH. WOilbe 2X. JPART 2. Pete XO

A young Coconut Palm with stunted leaves badly attacked by Oryetes, showing
the entry hole of an adult Beetle at the base of a leaf-stalk.

Bui. JBiwae, RESEARCH, WO, XM, IPasar Ze PLATE XIV.

Leaf of Coconut Palm broken off just above the Bore-hole of an Oryetes.

fant ty =,
<m, &
af eS tah

BW, JEN, IRESEAaRGE, Wo, 2X, Jari DY IPTG OW

An Oryetes boring into the heart of a young Coconut Palm.

ON THE USE OF EXPERIMENTAL PLOTS WHEN STUDYING
FOREST INSECTS.
By Ivar Tracarpa, D.8c., F.E.S.,

Swedish Institute of Experimental Forestry, Entomological Department,
Expervmentalfdltet, Sweden.

The problems of how to obtain accurate records of the infestation of injurious
insects has during recent years received considerable attention; and when we
take into consideration that it is quite as impossible to study the course of an
outbreak during its different stages as it is to compare two outbreaks occurring in
different localities or at different times, without possessing methods of recording
in an accurate way the degree of infestation and injury, it is evident that every effort
must be made to solve this difficulty and devise means of obtaining accurate records
in these respects.

In the case of the insects injurious to forest trees this need of accurate information
is, for various reasons, perhaps still more urgent than in the case of herbivorous
insects, but at the same time the problem seems to present fewer difficulties. The
necessity for possessing accurate methods of recording the infestation and the injury
done is obvious, because the attacks made on the trees by injurious insects often
continue for several years, which means that the degree of infestation and the amount
of injury done must be continually observed and accurately recorded during several
years, if we are to obtain a reliable idea of the course of the outbreak and to be able
to predict the probable seriousness of the attack during a succeeding year. It is
also, in the case of forest sects, necessary to investigate the consequence of the
injury to the health of the trees and to observe the infestation of secondary insects
following upon the primary ones, in order to know what measures are to be adopted
to prevent the spreading of an outbreak or the deteriorating of the timber if the
trees are killed.

But on the other hand, trees undoubtedly present fewer difficulties to such
investigations than do herbaceous plants, because, at least in many instances, the
degree of infestation may easily be fairly estimated without in the least interfering
with the trees or causing any disturbance in the prevailing conditions at the period
when the insects have temporarily migrated from the trees and are hibernating
in the ground.

When investigating an outbreak of the pine-tree looper (Bupalus piniarius, L.),
for instance, the degree of infestation is comparatively easy to ascertain by counting
the pupae in the ground in May in different parts of the forest ; all that is required
being a sufficient number of girls or boys to examine the ground thoroughly. The
amount of injury done, viz., the degree of defoliation, on the other hand, is more
difficult to estimate, especially when the same trees are to be examined during
several years, which makes it impossible to fell them and count the number of
devoured or half-devoured needles.

During the recent outbreak of Bupalus piniarvus in Sweden in 1916-1917 the
author had occasion to devote some time to the studying of these problems, and in
the following paragraphs the method employed will be briefly described.

158 IVAR TRAGARDH.

We find in the literature regarding Bupalus piniarvus the statement that the
trees are able to withstand a defoliation, provided it is not repeated and does not
occur early in the autumn. If, on the other hand, the defoliation is repeated during
two years in succession the trees are said to be past recovery. As this conclusion
was based on observations made chiefly in Germany and as it did not seem safe
to assume its applicability in Sweden, investigations on the state of health of pine-
trees after a defoliation repeated during two years were carried on in 1917 and 1918.

At Sorby in Sédermanland the degree of infestation in May 1917 was found to be
97°8 pupae per square metre, 18 per cent. of which were parasitised. By counting
the needles it was found that the average defoliation at the top of the trees was
75%, the injury decreasing gradually downwards, so that in the lower part of the
crown about 60% of the needles were intact.

| 0) D CD)
| i O O © Kees a
| Or) eng @o QO
ie) O
| ‘@ 12) O (9) \
on a ° iit of
O O o |
2 po 3 o
) (ur }
" i O . o) e © |
O O O |
=) / @) @) '
oman © ®
O 5 as 5 a
o ©
1O (e) m O ois ;
ees re)
| 4D) 3 - O ®
) ® @ )
x fe) nee ®D IO) i oO
fo) iD o @
oar O @ : Oo o)
| D © = Qierra | O 3 !
Fig. 1. Map of experimental plot in pine forest attacked by Bupalus piniarius,
L., at Sdérby, Sweden; 0 = felled trees, | = average defoliation, I] = severe
defoliation, — = attack by the pine beetle.

At the same time an experimental plot was laid out in the part of the forest most
seriously injured. In this plot the trees were marked with numbers; their height,
diameter at breast height, and the distances between them were measured; and a
map was drawn (fig. 1), the scale used for the plot being 1: 100 and for the trees
1: 400. Further, the degree of defoliation of each tree was recorded, only three
degrees being distinguished, viz., no injury, average injury, severe injury. The
plot was re-examined in May 1918, when also the degree of infestation was ascer-
tained, the number of pupae being found to have been then reduced to 8°8 per square
metre, 60°/ of which were parasitised.

The examination made in 1918 revealed the fact—as astonishing as it was
satisfactory—that in spite of the attacks during the two previous years not a single
tree had yet succumbed. In 4% of the trees, however, pine-beetles were breeding

THE USE OF EXPERIMENTAL PLOTS AND FOREST INSECTS. 159

and in 10% they had in vain attempted to enter. The supposition at once presents
itself that the primary cause of the attack of the pine-beetle was the defoliation
brought about by Bupalus piniarius, which had weakened the trees to such an
extent as to render them appetising to the beetles. A closer examinations of the
trees on the experimental plot showed, however, that this cannot be the case in.
this instance, at least not to any great extent.

According to the estimates made in 1917, 71 of the 126 trees of the plot had an
average injury, 14 being severely defoliated and the rest having escaped any visible
damage. Only one of the 14 severely injured trees had in the spring of 1918 been
attacked by the pine-beetle, and that in vain. Moreover, we notice that one of the
trees in which the pine-beetle was breeding had not previously been attacked by
Bupalus, nor had five other trees into which the beetle had in vain tried to enter
been previously injured by the moth. Fig. 2 shows how the attacks of the moth

15,4,-20cr
ZO PAO
25,,-30¢m

20% | ithe spe
i ~ d
i a
~ |
a) ‘

Fig. 2. Diagram showing the distribution of attacks of
Bupalus piniarvus (solid line) and the pine beetle (dotted line)
on trees of different dimensions.

and beetle are distributed in trees of different dimensions, from which it is evident
how little the attack of the pine-beetle is connected with that of Bupalus. We
notice that only 24°% of the trees measuring less than 15 cm. at breast height have
suffered from the moth, whereas all trees measuring more than 30 cm. have been.
attacked. The pine-beetle, on the other hand has preferred the smaller trees,
those exceeding 30 cm. not being attacked at all.

From this it is evident that the two insects, the pine-tree looper and the pine-
beetle, work on two different lines, one selecting the largest trees, the other
preferring the smaller ones. The investigation has also revealed the fact that
during the outbreak in 1916-1917 the forest was able to withstand the defoliation.
and that the subsequent attack of the pine-beetles was in no way the consequence
of the previous defoliation, but would have happened in any case.

(C605) F

160 IVAR TRAGARDH.

It is of course impossible to draw any conclusions from these data as to
what is likely to happen during other outbreaks, among other things, because we
have no data regarding the parasites of the eggs and the younger larval stages ;
but if other outbreaks are studied in the same way, it seems likely that we shall
by degrees be able to prognosticate after the first year the fate of forests attacked
by Bupalus.

During such investigations, as well as others when special attention is paid to
the influence of the attack on the health of the trees, I venture to think that the use

of experimental plots as outlined above will be found very expedient or even
indispensable.

161

THE CHAETOTAXY OF THE PUPA OF STEGOMYIA FASCIATA.
By J. W. 8. Macriz, M.A., D.8c.,
West African Medical Service.

The pupa of Stegomyra fasciata is furnished with 200 setae, 100 on each side,
which may be conveniently described according to their situations as follows :—
Cephalo-thoracic setae,
Circum-ocular 3
Antero-thoracic a: ui: ‘. ony a
Dorsal is a at a a4 es “i
Supra-alar L
Postero-thoracic 3

Abdominal setae,
Dorsal (including the lateral setae and the setae on
the paddles) * be ae of peny6]
Ventral Ms ss rE s. @ Lilia t

100 on each side.

The pupa is bilaterally symmetrical, that is, setae occur in similar situations on
each side of the body, so that it will suffice to describe the arrangement on one side
only. The setae on the two sides of the same pupa, however, often vary as regards
their sub-divisions, and similar variations occur between different individuals ;
as an example, in Table I are shown some of the variations that were found in ten
pupae taken at random. An examination of a larger number would have revealed
a wider range. As a rule, a seta which is sometimes single, sometimes divided,
is longer when single. Hor example, in one pupa the seta at the posterior angle of

TABLE I.
Cephalo-thoracic Setae, showing the Variations found im ten Pupae.

Seta. 1 2 3 4 5 6 | a 8 9 | 10

‘Cireum-ocular, Superior sie ee 2 a me Lh Bil eee al OR i heat
Posterior cme med bos Da) — De Le | QDs Del Qa 2 2.2 52

Inferior sige 2S Wie ees he ee ey LD gk oF ia

Antero-thoracic, Anterior eee ae a ee a Le Stee | 2 2a oe
Superior sil ype la | ED DD | Bexculroe Dulce 2

Antero-inferior Sieh nen ee eh be) LS Loe

Postero-inferior | LEE 21 VSB a7 292) 2 2S 22) Bere?

Dorsal See De Shs 2: 2h | 2 2h 2 ea ee Or soe
Supra-alar : ae et ee el) te le) Lo We Poet
Postero-thorazic, Internal Oe eo SOD OR LRP 26D QO 2S Diane
Median eget ee: aa, ek Madey Teeth bss hele de rd: Lede pT dhe eed

External Ammer ae el tee WS Me Ted ys tN ted Dee eT

ry
bo

(605)

162 J. W. S. MACFIE.

the seventh segment was single on the right side, double on the left; the former
measuring 266u, and the latter only 159u in length. This fact is not specifically
mentioned in the descriptions which follow, but should be understood.

It has been stated that ‘‘ As the pupa increases in age additional hairs are
developed.” In Stegomyia fasciata this is not the case, the pupa immediately after
it has shaken itself free from the larval skin having exactly the same number of
setae as pupae one or two days old, and as pelts left behind after the emergence of
the adult mosquitos. The setae on male and female pupae are also identical.

The description which follows is based on the comparison, seta by seta, of ten
specimens of the pupa of Stegomyra fasciata.

Cephalo-thoracic Setae.
The setae on the cephalo-thorax are few in number, and relatively inconspicuous..
There are altogether twenty-four, namely twelve on each side of the body.

Post-ocular setae. (fig. 1).

These setae are situated close to the posterior margin of the eye, and are some-
times difficult to see because, being directed forwards and outwards, they are apt.
to be obscured by the dark ocular pigment behind them. In a pelt they are also.
difficult to find, as they become displaced when the mosquito emerges. They are
best seen in pupae which have been macerated in caustic potash. There are three.
post-ocular setae, namely :—

Supervor seta (fig. 1,1). A delicate seta of moderate length situated in the angle
formed between the upper posterior margin of the eye and the lower border of the
antenna as it sweeps backwards ; usually single, sometimes double.

Median seta (fig. 1, *). A small delicate seta situated just behind the eye near
its lower margin ; usually double, sometimes single.

Inferror seta (fig. 1, *). A delicate seta of moderate length situated a little below
the median seta and at about the level of the posterior margin of the eye ; usually
single, but in one specimen triple on one side.

Antero-thoracic setae (fig. 1).

These setae are situated at the anterior margin of the thorax, or rather in the
triangular area between the anterior margin of the thorax and the upper border
of the antenna as it sweeps backwards. They are directed forwards, upwards,
and outwards. There are four antero-thoracic setae, namely :—

Lower anterior seta (fig. 1, *). A delicate seta of moderate length situated a little
above the antenna and a little behind the anterior margin of the cephalo-thorax ;
double or triple.

Upper anterior seta (fig. 1, °). A delicate seta of moderate length situated above
and slightly posterior to the lower anterior seta ; single, double, or sometimes triple.

Lower posterior seta (fig. 1, °). A stouter and longer seta than any of the preceding,
situated just above the eye close to the upper margin of the antenna as it sweeps
backwards ; single.

Upper posterior seta (fig. 1, "). A seta of moderate length and strength situated
close to, and immediately behind the lower posterior seta; generally double,
sometimes single.

CHAETOTAXY OF PUPA OF STEGOMYIA FASCIATA. 163

Dorsal seta (figs. 1,°; 2, °).

This seta is situated on the thorax a little posterior and internal to the respiratory
trumpet, and is directed upwards. It is small and delicate; usually double, but
sometimes single or triple.

{ ys

Fig. 1. Cephalo-thorax of pupa of Stegomyia
fasciata, lateral view.
1-3. Post-ocular setae: J, superior; 2,
median ; 3, inferior.
4-7, Antero-thoracic setae: 4, lower anterior;
5, upper anterior; 6, lower posterior ;
7, upper posterior.
8. Dorsal seta.
9. Supra-alar seta.

Supra-alar seta (figs. 1,°; 2, ”).

This seta is situated on the thorax above the root of the wing, and is directed
outwards and forwards. It is rather long and slender ; single.

Fig. 2. Cephalo-thorax of pupa of Stegomyia
fasciata; dorsal view.

1-4. The ends of the antero-thoracic setae:

1, upper anterior; 2, lower anterior ;

3, lower posterior; 4, upper posterior.

5. The end of the superior post-ocular seta.
6. Dorsal seta.
7. Supra-alar seta.

8-10. Postero - thoracic setae: 8, internal ;

9, median ; 10, external.

164 J. W. S. MACFIE.

Postero-thoracic setae (fic. 2).

The posterior part of the cephalo-thorax is readily detached as a separate piece
which is concave both anteriorly, where it fits on to the rest of the cephalo-thorax,
and posteriorly, where it joins the first abdominal segment. It consists of two
lateral triangular plates connected by a narrow central bridge. The bridge is
composed of a small quadrilateral area in the middle, and on each side a short
prolongation, from the anterior margin of which arise the postero-thoracic setae.
These setae are directed forwards and outwards. There are three postero-thoracic
setae, namely :—

Internal seta (fig. 2, 8). A seta of moderate size and strength situated close to
the outer margin of the quadrilateral area ; usually double, sometimes single or
triple.

Median seta (fig. 2, °). A stouter and longer seta situated a little external to
the internal seta ; single.

External seta (fig. 2, 1°). A seta of moderate size and strength situated a little
external to the median seta ; single or double.

Abdominal Setae.

The majority of the setae of the pupa are attached to the abdominal segments :
those on the ventral aspect being for the most part small and inconspicuous, those
on the dorsal and lateral aspects being more highly developed, and including the
setae of greatest systematic importance.

Dorsal abdominal setae (fig. 3).

The dorsal abdominal setae are numerous, and are mostly situated near the
posterior margins of the segments. With them may be ranked the lateral setae,
which are of importance because they provide convenient means for specific
determinations, and the setae on the ends of the paddles. With the exception
of the setae on the first segment all the abdominal setae are directed backwards.

Certain of the dorsal abdominal setae fall naturally into series, and it will be
convenient to describe all the setae in relation to them. The following series can
readily be recognised :—

1. The setae composing the lateral row (fig. 3, A). These are especially wel.
developed setae situated at, or slightly anterior and internal to, the posterior angles
of the segments. On the most anterior segments they are slightly ventral.

2. A series of delicate setae situated close by and internal to the setae of the
lateral row (fig. 3, A’).

3. The setae composing the sub-median row (fig. 3, C). These setae are
situated near the posterior margin of the segments at a point about half-way between
the middle line and the posterior angle.

4. A series of setae situated near by and anterior to the setae of the sub-median
row (fig. 3, C’, C”).

5. The setae composing the sub-lateral row (fig. 3, B). These setae are situated
near the posterior margins of the segments at a point about half-way between the
sub-median row and the posterior angle.

CHAETOTAXY OF PUPA OF STEGOMYIA FASCIATA. 165:

6. A series of setae situated near by and anterior to the setae of the sub-lateral
row (fig. 3, B’).

7. A series of very minute setae, the anterior dorso-central setae (fig. 3, D), situated
near the anterior margins of the segments almost in line with the setae of the sub-
median row.

The dorsal abdominal setae may now be described as they occur on each segment,
making use of the abbreviations A, A’, B, B’, C, C’, C", and D, as defined above.
SEGMENT I.

There are eight setae on each side of the first segment ; unlike those on the more
posterior segments they are not directed backwards.

Dendritic tuft. In the middle of the segment, and at a short distance from the
middle line, is a large seta developed into a dendritic tuft, which arises from a
delicate membrane and is directed upwards. It has a triangular basal portion

Fig. 3. Abdomen of pupa of Stegomyia fasciata, dorsal view ; for
explanation see text.

which gives rise to a number of stout primary branches, seven to sixteen in number,
which in turn divide and subdivide repeatedly. This seta may correspond to C on
the more distal segments.

Antero-internal seta. A long, stout seta situated immediately anterior to the
dendritic tuft, and directed forwards ; single. This seta may correspond to C”
on the more distal segments.

Antero-external seta. A stout seta of greater length situated close to the antero-
internal seta, but a little external and posterior to it ; directed forwards and outwards;
usually single, sometimes double. This seta may correspond to C’ on the more
distal segments.

Medio-internal seta. A small tuft of two to four hairs situated near the anterior
margin of the segment about half-way between the antero-internal seta and the
lateral border; directed forwards and outwards. This seta may correspond to
B’ on the more distal segments.

Medio-external seta. A tuft of two to five hairs, rather longer than the medio-
internal tuft and situated close to it but slightly posterior and external ; directed
forwards and outwards. This seta may correspond to B on the more distal segments.

Postervor-internal seta. At the point where the lateral border of the segment
begins to curve inwards there is a group of three setae: the innermost of these is

166 J. W. S, MACFIE.

the postero-internal seta. It is a rather long delicate seta, directed outwards or
forwards and outwards ; usually single, sometimes double. This seta may correspond
to A” on the second segment.

Postero-external seta. A rather long seta situated close by and external to the
postero-internal seta, directed outwards or forwards and outwards ; usually single,
sometimes double. This seta may correspond to A’ on the second segment.

Lateral seta. A small seta external to the postero-external seta and situated
just posterior to the angle made by the ‘border of the segment curving inwards ;

directed outwards; single. This seta seems to correspond to A on the more distal
segments.

SEGMENT II,

A—short, fairly stout but not very highly chitinised; single. A’ and A”—there
are two long, delicate setae situated close by and internal to the lateral seta, the
one (A”) slightly anterior to the other (A’); both usually single or double. B—
delicate seta of moderate length; single, double, or triple. B’—small delicate
tuft of two to five hairs situated a little anterior and internal to B. C’—a rather
long seta situated internally to B and B’, and near the posterior margin of the
segment; single, occasionally double. C”’—a seta of moderate length situated at
about the same level as B’, but internal to it, and slightly either internal or external
to C’; single. Nearer the middle line than C’, and near the posterior margin of
the segment, is a seta of moderate size which may represent C displaced mwards ;
usually double or triple, occasionally smgle. D—a minute seta.

SEGMENT ITI.

A-—short and stout, rather longer and stronger than on the second seg-
ment; single. A’—a long delicate seta situated anterior and internal to A; single
or occasionally double. B—seta of moderate length; single, double, or occasionally
triple. B’—small tuft of two to four hairs or a single seta, anterior and internal
to B. C—a seta of moderate size situated near the posterior margin of the segment
a little mternal to the line of the sub-median row; usually single or double, but
occasionally with as many as five branches C’—a rather long seta situated at some
distance externally to C, and anterior to it; single or double. C’—a small seta,
relatively stout, placed more anteriorly but less exteriorly than C’; single.
D—a minute seta.

SEGMENT IV.

A— short and stout, rather longer than on the third segment ; single. A’—a long
delicate seta situated anterior and internal to A; usually single, occasionally double
or triple. B—a long seta reaching half-way or further across the fifth segment ;
single. B’— a small seta anterior and internal to B; single, double, or triple. C—
a delicate seta of moderate length ; single ordouble. C’—a seta of moderate length
external and anterior to C; double or triple. C’— a small rather stouter seta
internal and anterior toC; single. D— a minute seta.

SEGMENT V.

A— stout, longer than on the fourth segment; single. A’— a long delicate seta
anterior and internal to A; single, occasionally double. B— a long seta reaching

CHAETOTAXY OF PUPA OF STEGOMYIA FASCIATA, 167

about half-way across the sixth segment; single, occasionally double. B’—a small
seta anterior and a little internal to B; usually double, sometimes single or triple.
C— a moderately long delicate seta; single or double. C’— a moderately long delicate
seta external and anterior to C; single or double. C”’— a rather short, relatively
stout seta internal and anterior to C; single, occasionally double. D— a minute
seta.

SeemeEnt VI.

A— stout, longer than on the fifth segment; single, end occasionally forked.
A’— a long delicate seta anterior and internal to A; single, occasionally double.
B— a long seta reaching nearly half-way across the seventh segment; single. B’—
a delicate seta of moderate length anterior and a little internal to B ; usually double,
occasionally single. C— a rather small delicate seta ; single or double. C—a seta
of moderate length external and a little anterior to C; single, occasionally double.
‘C’—a small seta, rather stouter, anterior and internal to C’ ; single. D—a minute
seta.

Segment VII.

A—a long strong seta, sometimes pubescent ; single or double. A’—a delicate
seta of moderate length close by and internal and posterior to A; single or double,
occasionally triple. B—arather long seta single. B’— a rather long delicate seta
external and anterior to B; single. B’—a smaller delicate seta internal and anterior
to B; single, occasionally double ; this seta probably corresponds to C’ on the more
anterior segments. C— a rather small delicate seta; single. C’— a small seta,
rather stouter, anterior to C; single; this seta probably corresponds to OC” on the
more anterior segments. D— a minute seta.

Secment VIII.

A— a well developed tuft of two to five, usually four, strong hairs, sub-plumose
and sometimes branched. D— a minute seta. There is one other seta, P, on this
segment, situated near by and external to the root of the paddle ; it is rather long,
slender, and single. At the distal end of the midrib of the paddle there is a
moderately long strong seta (P); single.

SEGMENT IX.

No setae.

A typical segment may be regarded as having the following setae: a lateral
seta with a seta a little internal to it; a seta belonging to the sub-lateral row with
a seta a little anterior to it ; a seta belonging to the sub-median row with two setae,
the one internal to the other, a little anterior to it; and a minute anterior dorso-
central seta.

Ventral Abdominal Setae (fig. 4).

The ventral abdominal setae are relatively inconspicuous. They are directed
backwards, and most of them are situated near the posterior margins of the segments.
Some of them are arranged in series, and it will be convenient to describe them
before enumerating the setae found on the different segments. The following series
can be recognised :—

1.—A series of setae, the medzo-lateral setae (EK), situated a little posterior to
the middle points of the segments a short distance internal to their lateral borders.

168 J. W. 8S. MACFIE.

2.— A series of setae, the postero-lateral setae (D), situated near the posterior margins
of the segments a little internal to the posterior angles.

3.—A series of setae, the enner ventral setae (B), situated near the posterior margins
of the segments at points about half-way between the middle line and the posterior
angles.

4.—A series of setae, the outer ventral setae (C), situated near the posterior margins
of the segments between the inner ventral setae and the postero-lateral setae.

5.—A series of minute setae, the anterior ventro-central setae (A), situated close to
the anterior margins of the segments and very near the middle line of the abdomen.
These setae, owing to their size and position, are often impossible to see.

The ventral abdominal setae may now be described as they occur on each segment,
making use of the abbreviations, E, D, C, B, and A, as defined above.

SEGMENT I.
No setae.

Fig. 4. Abdomen of pupa of Stegomyia fasciata,
ventral view ; for explanation see text.

SEGMENT II.

The only seta detected on this segment is a delicate single seta of moderate length
situated in a position corresponding to B. This seta is inconstant; that 1s to say,
in some pupae it was absent, and no socket could be found from which it might
have been detached.

SEGMENT III.

H— small, delicate; usually double or triple, occasionally single. D— a short
tuft of two to four hairs, sometimes branched. C— a delicate seta of moderate
length ; single. B— a small, relatively stout, seta; single. A— a minute seta.
SecMEnT IV.

E— a small delicate seta; usually double, sometimes single or triple. D—a
small tuft of two or three hairs, sometimes a single seta. C— a long delicate seta ;
single. B— ashort, relatively stout seta; single. A— a minute seta.

SEGMENT V. ;

E— a small delicate seta; usually double, sometimes single or triple. D—a
small tuft of two to four hairs, or a single seta. C— a long delicate seta, rather
longer than on the fourth segment; single. B— a short, relatively stout seta ;
single, sometimes forked at the end. A— a minute seta.

CHAETOTAXY OF PUPA OF STEGOMYIA FASCIATA. 169

SEGMENT VI.

E— a small delicate seta ; single, double, or triple. D— a delicate seta; usually
single, sometimes double. C— not represented. B— a small stoutish seta; single.
A— aminute seta. A little internal to B there 1s a delicate single seta of moderate

leneth (B’).
SEGMENT VII.

E— a small delicate seta; single, or more frequently double or triple). D— a
delicate seta ; single. C—notrepresented. B—asmall stoutish seta; single. B’—
a delicate single seta, similar to that on the sixth segment, a little internal to B.
A— a minute seta. The correlation of the setae near the posterior margin of this
segment and the sixth is not quite clear, and the difficulty is increased on this
segment because, owing to the reduced width, the relative positions are misleading.
The interpretation suggested is based on the facts that the seta D is in line with the
corresponding setae on the more anterior segments, and the seta B is in line with,
and similar in appearance to the seta B on the other segments.

SEGMENT VIII.

The only seta found on this segment is a very small single seta situated close to
the anterior margin of the segment at a point nearly two-thirds of the distance
from the lateral border to the middle line. This seta may represent A, which
appears to be absent, but if so it is displaced outwards and a little backwards.

SEGMENT IX.
No setae.

A typical segment may be regarded as having the following setae : a medio-lateral
seta, a postero-lateral seta, an outer ventral seta, an mner ventral seta, and a minute
anterior ventro-central seta.

ae Bei Aa Fed eee
fa z be tem i of ee ue o agi led ee ah
| we st ie saan sible Me :
nt de a eee hadi met Uh icirogad (aie
ee. Hace ttseja. Chia tid oh The
as ai ee see avid aae igh ies silrteldh craw |
ih ti Sn AES 9b) Ra, egins ea ‘laa a
i fot cca Bineatlhie i re ald wt Hibs hoe yok oe, ‘atm |
| deta CE aie ct Fai ‘etonk ady ee Lad 9h Posters, attain

ue Bilas day om ores out icra pba | ter ehiit Peed wht ite pal TH
wilde. tee eee eye je ct, oot air # nad aah’ Napiiie ata

Qa a ! i= ip, ve , . ‘ nem : ‘ cs
~ “ad fi

Pea ihn Sasa DD) ite Wa ‘ini Mint ‘lien: tee Kies ines
Say Py Wadaiedly tf, vont Thies: ju 4) 3
ONE \ eu heii ke Dieser, gion: med wen ‘allt i bis dash anal oa |
sak ome anise on, ee 4 Day dete pee ais i i ti

iu : hati . ' ; } : ps
t Ma ; ieee aie oa @ a : bel ; re a :
* a i le ea j i ‘ " os 4 Hagel

ea Liviotil Ciba i, sabi we aeaal haiti scien vaial
“ita tien w bien i jiey \ethmowt ane aha hdanid vat

2 ri toe ; iy f Vig oe / Sel ee 7
Ree CE Us «ol Ray het ania all ts q : dion site i ts
ie 2 aes aan a (* K, ap A i hte - iy . ‘he . i a We 6 1 ;
ji OM Rin RON, SSE RRR Gig ER TEE ei ata

: t " ‘ ’ i a : e x : st oe * ii po ea y 5 a ‘ rT: . aay we (vp wont by fortrut fa : 7

171

A NEW HISPID BEETLE INJURIOUS TO THE OIL PALM IN
THE GOLD COAST.

By 8. Mautik,
Professor of Zoology in the University of Calcutta.

Through the courtesy of Dr. G. A. K. Marshall I have had the opportunity of
examining specimens of a beetle belongmg to the subfamily HispinarE of the
CHRYSOMELIDAE, which is reported to be mjurious to oil palms (Elaews guineensis,
Jacq.) in the Gold Coast. Concerning this insect Mr. W. H. Patterson, Government
Entomologist in the Gold Coast, writes as follows :—‘ The beetle has suddenly
become a pest, bemg reported as having destroyed all the expanded foliage of
many thousand oil palms in one district. A similar epidemic occurred in
1909, but entirely disappeared during the following rainy season, smce which
time the beetle has been very rare.’ From this it would appear that the
extent of damage is considerable when the insect becomes numerous. Although
Mr. Patterson does not indicate the nature of the damage there is no doubt that
the larvae mine into the young shoots and the adults defoliate the young leaves.
In a recent work (Hispinae and Cassidinae, Fauna of Brit. Ind., London 1919,
pp. 10-18) I have reviewed all the Hispid pests of plants useful to man. Only two
African Hispids have hitherto been noted as harmful, and these attack maize in Kast
Africa.

Only four species of the genus Coelaenomenodera, to which the present insect belongs,
are known from the continent of Africa, viz., speciosa, Gestro, sigmfera, Gestro,
thomson, Gestro, and costulata, Kolbe, the rest of the species, about 28 in number,
are from Madagascar.

The present insect differs from speciosa by the shape of the prothoracic lobe ; from
signifera by its size, though the lobe is similar ; from thomsonz by the fact that it has
the basal six joints yellow to reddish black, a character which also separates it from
signifera ; and from costulata by the shape of the prothoracic lobe and the fact
that the antennae (except the first jomt) and the tarsi are black. It is necessary
therefore to describe the present species as a new one.

Coelaenomenodera elaeidis, sp. nov. (fig. 1).

Body oblong, slightly broadened behind. Colour pale yellow; elytra reddish,
except the basal and apical areas, this reddish colour disappearing when the specimens
are preserved in alcohol ; mandibles, eyes, and the five apical joints of the antennae
black; the basal jomts gradually become red-brown mixed with black from the
yellow of the basal joint to the almost black sixth joint. The frontal lobe of the
prothorax is elongate anteriorly and projects in front, concealing the forehead.

Head: the interantennal space is slightly raised, and the eyes strongly convex.
The antennae pass a little beyond the base of the prothorax and are sparsely covered
with whitish hairs, the six basal jomts being smooth and shining and the five apical
thickened ; the first jomt is small, the second longer, the third, fourth and fifth
equal to each other, and the sixth is intermediate between the thickened apical

172 S. MAULIK.

jomts and more slender basal ones. Prothorax shining, longer than broad, slightly
narrower at the base than the base of the elytra; the anterior angles are ill-defined,
owing to the frontal lobe, the posterior acute and produced ; the sides are undulated.
The upper surface is convex; at the base in front of the scutellum is a deep depression,
and there are two others at each side, a small one just below the frontal lobe and a
larger one posterior to it; in these depressions are a few punctures. The frontal
lobe is semi-elliptical in shape, hollowed, with a median longitudinal ridge; the
front border of the lobe varies, being sometimes rounded and sometimes more pointed ;
its surface is semi-transparent, with a few indistinct large punctures. Behind the
lobe the surface is uneven, being slightly depressed in the middle and sloping
posteriorly. Scutellum small, mconspicuous, deeply sunk im a hollow. Elytra
punctate-striate, each with three costae terminating at the point where the surface
slopes down towards the apex, the lateral margms being slightly expanded. Between

y. .

I DRI? AG (
DiKeKGh Tega, ta ee SS pO
Nay SEN SF

r
DOCK io if SEAR

SIE

Pert

COO OA
SOO ry 6
ew

ee)
=
j

Ta
prepares Hee Secrsecea o* | sens =
SBR LT AASTOR A AG oa
EPR Ne ON i pee

Tole latareve eka
Ay

@
Teo S: Se

iw
©

g
B22
tele ‘

AS COOK
ote)
(ey

he)
“He,

v9 (Uh

Fig. 1. Coelazenomenodera elaeidis,
Maulik, sp. n.
the suture and the first costa there are three rows of punctures, between the first
and second costae two, between the second and third two, and between the third
and the lateral margin three; the punctures are coarse and deep. Underside
uniformly pale yellow, less shining than the upper side. The tibiae are short ;
the mid and hind tarsi are almost as long as the tibiae and broader at the
apex than at the base, so that the basal jomts are smaller than the following
ones; the front tarsi are as broad at the base as at the apex, so that all the jomts ©
are equal in breadth. The venter is slightly hairy.
Length, 4°75-5°75 mm.
Type in the British Museum.
Described from 11 examples.

A NEW HISPID BEETLE. 173

Larva (fig. 2). The full-fed larva is flattish, creamy white, and measures about
five millimeters in length (dried specimen mounted on card). The abdomen hag
ten segments, including the last, on the underside of which the anus is situated ;
the thorax is three-segmented, so that with the head the larva is composed of 14
segments. There are altogether nine pairs of spiracles: one pair on the mesothorax,
and a pair on each of the first eight abdominal segments, each spiracle being placed
on the side of the dorsal surface, except that on the mesothorax, which is ventral.
The prothoracic segment is considerably expanded laterally, the expansions being
rounded, and is deeply emarginate in the middle for the reception of the head. The

Crp a re)

Fig. 2. Larva of OC. elaeidis (ventral view) ;
a, antenna.
meso-and metathorax are not expanded laterally. Each abdominal segment has
a slight rounded expansion on each side bearing two fine hairs at the edge in the
middle, which are homologous with the well developed projections from the sides
of the segments of other Hispid larvae. The larva has no legs.

Morphology of the Larval Mouth.

The head is supported in the prothorax by means of two large elongate-oval
structures with strongly ch:tinised edges and chitinised supports in the middle, which
are inclined to each other at an angle, meeting anteriorly at the mandibles, and
having nearly half their length inside the prothorax. Compared with the size of
these structures, the mandibles, which form their apical part, are minute. The

174 S. MAULIK.

buccal cavity is formed by an upper membranous structure (epipharynx), supported
by two chitmous pieces outwardly bent (fig. 3, B), a lower lip (labium) (fig. 3, A, 1)
and the mandibles. The upper structure has numerous papilliform hairs directed
inwardly and posteriorly. These are probably connected with the pharyngeal
muscles and control their action according to the stimuli received from the food
particles. The condyles of the mandibles (fig. 3, A, md.) are controlled by two bands
of muscles (fig. 3, A, ms.), and the apices of the mandibles are pointed, the mner
border being uneven without any teeth. The labium is a membranous structure
with the apex broadly rounded. No labial palpi are observable in the present
preparation. A pair of maxillae is present, situated ventrally to the mandibles,
each consisting of a base and the main lobe; on the imner border of this lobe are
numerous hair-like structures. Although in the preparation before me I cannot
see the maxilliary palpi, it is possible that they are present. An antenna is situated

“he \
fi NY}
WP ie SN
c WN se
>, WEB:
:|] | A
ay. ip
=} A :
* ( if
aN |) VAN
ANN Rin :
Par Nit HRA
F\Whit Wk be
mh: \ Wipy >
Y \i\\ 4
y
Wh 7
\
Y
\ ‘

Fig. 3. Larva of C. elaeidis, mouth-parts ; md.j, mandibular joint ; mz, maxilla ;
ch.s, chitinous supports ; ms, muscles controlling the movement of the mandibles ;
l, labium; s, setae; B, roof of the mouth.

on each side of the head a little posterior to the mandibular joint ; it is two-jointed,
with a simple lobe at the apex of the second joint and another stouter one bearing
several papillae (fig. 2, a).

The foregoimg observations are taken from preparations made from two larvae,
one full fed and the other younger. They were cleared in potash and washed and
put in 90 per cent. alcohol shghtly coloured with Congo red for about 12 hours; the
chitm was stamed well, showing good differentiation. With more material the
structures may be made out with more precision and in greater detail than I have
been able to do at present. The figure of the larva has been drawn from a balsam
preparation ; m clearing in potash the specimens get stretched to a certain extent.
In drawing the figures of the mouth-parts of the larva the artist was aided by three
photomicrographs taken by myself.

175

OBSERVATIONS ON SCALE-INSECTS (COCCIDAK).—VI.
By Rosert Newsteap, F.R.S.,

The School of Tropical Medicine, the Unversity, Liverpool.
(Prare xX VT")

Clypeococcus, gen. nov.

Female. Dorsum nude and densely chitinised. Marsupium occupying practically
Venter (sternites) remaining attached to the food-plant.

the whole of the interior.
Cephalic margin with a strongly developed clypeus,

Legs and antennae present.
lying prone over the mentum. Abdominal stigmata present.
Larva of the Icerya type, and with several pairs of abdominal stigmata.
Type, C. hempelr, Ckll.
Clypeococcus hempeli, Ckll.
Icerya (Crypticerya) hempelr, Ckil., Can. Ent. xxxi, p. 43 (1899).
Crypticerya hempeli, Hemp., Rev. Mus. Paul., iv, p. 376 (1900).
Female adult (PI. xvi, fig. 1). Sphaeroid and hollow at gestation, the marsupium
filling practically the whole of the interior. Integument densely chitinised, faintly

ribbed and polished ; very dark plum-coloured, with a delicate ‘‘ bloom-like”’ secretion,
the segmentation of the dorsum indicated by faint blackish interrupted lines. Venter

°
(fe
REY

NU Ki
NCA
A:

=
BEEZ ”
ag
= F
> fi
eH
—= =,
cB
are |
PY
“4

Clypeococcus hempeli, Ckll., 9; a, the relative position of the various

Fig. °h.
appendages on the anterior portion of the venter: an, antenna; cl, clypeus ;
m.sh, membranous sheath enclosing the buccal filaments ; b, antenna of larva ;
ce, tarsus of larva; d, gland-pores and hairs on venter of 2; e, dorsal spines of
larva; f, gland-pores of larva.
membranous, resting on a thin layer of hard resin-like secretion, segmentation distinct.

Mentum biarticulate, the “filaments ’’ enclosed in a membranous sheath (m.sh. in
G

(C605)

176 ROBERT NEWSTEAD.

fig 1, a), the latter lying obliquely to one side of the median line in mounted pre-
parations. Clypeus (fig. 1, a, cl.) strongly produced, densely hairy and com-
pletely overlying the mentum. Antennae, legs and thoracic stigmata lying in very deep
depressions. Ventrally the margins and the whole of the thoracic area, together with
the terminal abdominal segment, densely hairy ; the hairs (fig. 1, d) distinctly knobbed
at the tips. Antennae (fig. 1, a, an) robust, of 8-9 strongly chitinised segments.
Legs (fig. 1, a) robust, and strongly chitinised ; claw simple. Abdominal stigmata
‘small and apparently in six pairs. Thoracic stigmata large; the gland-pores mul-
tilocular, the cylindrical ducts long and apparently truncate proximally. Dermal
‘gland-pores (fig. 1, f) large and with a deep cylindrical rim; they are irregularly
disposed on the venter and for a short distance also beyond the margi1; the dorsum
rather thickly studded with minute pores surrounded by dark granular bodies.
Anus large and roughly circular in the outline formed by the dense chitinous walls
of the body; no other exact details observed, but portions of a ruptured
membranous cuticle project into the lumen, suggesting that the true structure is
wanting. Diameter, 6°5-9mm.; antennae, 0°6 mm. long.

Larva. Form elongate. Antennae (fig. 1, 6) clavate, of six segments ; 2nd, 4th,
and 6th longest ; 3rd and 5th very short and subequal ; all the segments with fine
slender hairs, the longer ones less than the length of the terminal segment ; terminal
segment with several slender spinose hairs at the tip ; in the clear membrane which
connects the 2nd and 3rd segments is a relatively very large chitinous ring with a
granular centre; as no hair or spine has been seen attached to this organ it is
suggested that it may be a sensorium or gland-pore; three similar structures also
occur on the coxae. Legs long and slender; claw (fig. 1, c) with a minute,
subapical denticle. Dorsal gland-pores with a broad quoit-like rim and a very short
cylindrical duct, the whole structure when seen in profile reminding one of a very
short-stemmed agaric or broad-flanged stopper to a glass jar; these are arranged
more or less in transverse rows. The terminal segments of the abdomen, more
especially in the region of the anus, with a group of large spines, packed very closely
together in the median area. Terminal abdominal bristles in five pairs ; these are
of great length, or slightly longer than the antennae. The chitinised hind gut with
its flange-like papillae very distinct. Abdomen with at least six pairs of stigmata ;
the external opening minute and membranous; the atrium long, gradually
narrowed distally, strongly ribbed transversely or irregularly moniliform, and the
proximal portion with a deep wide cleft ; the connective tube slender, chitinised,
and bifurcated, each branch having well-formed taenidia; length of the atrium
about twice the diameter of the large quoit-like gland-pores.

Braziu: San Paulo; on the spiny branches of an unknown tree or shrub
(mimosa ?), 1906. (In a small collection of Brazilian Coccidae purchased from Mr.
O. EK. Janson ; no other data).

The determination of the species is based upon an examination of a specimen (ex
coll. T. D. A. Cockerell) kindly supplied by Mr. E. E. Green, to whom I express my
thanks, as in the absence of his material I should most certainly have given this
insect a new specific name.

I have described the morphological characters in some detail as hitherto the true
characteristics seem to have been almost entirely overlooked. The presence of a

OBSERVATIONS ON SCALE-INSECTS. rire

strongly developed clypeus in this Coccid it is unique, and is probably homologous
with that of certain members of the FuncoripEa belonging to the genus Cixius
and other allied forms.

Aspidoproctus gowdeyi, sp. nov.

Female adult. Completely covered dorsally with a thick and densely felted layer
of dusky white and pale yellow wax ; the wax more or less divided into segments
corresponding to the segmentation of the body. Venter pale castaneous, slightly
mealy and in parts covered with white woolly filaments. Legs pitchy red.
Marsupium well developed, the secretionary operculum wanting, but apparently
broken away.

Fig. 2. Aspidoproctus
gowdeyti, Newst., sp. n.,
6 genital armature.

Male. In dry specimens the body is uniformly black or piceous. Wings very
dark smoky brown or blackish ; costa black, subcostal vein dull crimson ; forked
pseudo-vein dull white and very conspicuous. Halteres with 4-5 strongly hooked
bristles. Body mealy. On maceration in KOH the softer parts of the integument
change to dull crimson ; the sclerites and also the legs and antennae, piceous. The
single pair of central processes well developed ; terminal bristles varying in length
and thickness, the longest about four times the length of the process. Genital
armature (fig. 2) shaped somewhat like a flask or ampulla in minature ; the proximal
portion broadest ; near the centre of the dilated proximal portion is a shallow saucer-
shaped appendage suspended to the walls of a large, ovate, chitinous ring ; distal
portion cylindrical and densely chitinised. Intromittent organ not observed.

UcanDA: Kampala, on plumbago and rose, 15.x.1918 (C. C. Gowdey).

The males hatched apparently during transit through the post and were rather
badly crumpled and otherwise injured.

Walkeriana digitifrons, sp. nov.

Female, adult (Plate xvi, fig. 3). Stationary. Somewhat pyriform, narrowest in
front ; highly convex and sloping gradually towards the cephalic margin. Dorsum
more or less covered with an easily deciduous layer of white or pale buff-white,
granular wax ; margin with relatively robust, coalescing plates of white wax, which,
when perfect, are distinctly laminated transversely ; cephalic margin with a short
stout, centrally placed, cylindrical process. Integument very hard, dark castaneous
or piceous; margin, when denuded, with a series of large blunt tooth-like projections
or tubercles, arranged rather widely apart ; dorsum with 3-4 bilateral rows of gland
pits, im which the granular wax generally remains more or less intact; abdominal
(C605) a2

178 ROBERT NEWSTEAD.

segmentation distinct. Venter hollow, but filled with loose white flocculent material.
in which the eggs are laid and the larvae subsequently hatch. Vaginal orifice
transversely linear, and without a secretionary operculum or an invaginated mar-
supium. The integument of the dorsum, after long maceration in KOH, becomes
pale brown in colour; but partly retains its hard and somewhat brittle nature.
Antennae with 10-11 segments; 4th segment the shortest, but its articulation
with the 5th is often incomplete, though the constriction between these segments
is well marked and relatively deep. Eyes in the form of a truncated cone.
Legs rather sparsely hirsute, but normal in shape. Compound gland-pores
of the dorsal pits relatively large, rather narrowly ovate, and collectively forming
a rather coarse and irregular reticulation. Dorsal gland-pores of the usual
multilocular type ; they are fairly numerous and more or less evenly distributed ;
those on the marginal tubercles are surrounded by a radial and somewhat petaloid
pattern, due apparently to the thickening of the integument. Ventral gland-pores
similar to those of the dorsum, but slightly larger and with an ovate central pore ;
they are thickly packed together at the margins and also on the marginal tubercles.
Anus surrounded by a broad concentric band of dark chitin, which is thickly studded
with simple, more or less circular gland-pores. Fine slender hairs are scattered over
both sides of the body ; marginal hairs also slender, but longer than those on other
parts of the body. Length, inclusive of the finger-like process on the cephalic margin,
6-9 mm.

Larva. This is of the usualform. Five of the long hairs on the terminal segment
of the antennae equalling the length of the body. Sides of the abdomen with
numerous long fine hairs ; anal hairs or bristles apparently in three pairs, of which
the central pair are much the more slender and shorter ; outer pairs (2nd and 3rd)
relatively stout and longer than the body.

Uaanpa: Damba Isl., Sesse Islands, Lake Victoria ; on Barkea eminir, 8.x.1912
(C. C. Gowdey).

Pseudococcus inquilinus, sp. nov.

Female, adult. Uxternal covering and lateral appendages destroyed by the
medium in which the specimens were preserved. Form, when denuded, similar
to that of P. longispinus, Targ. With 16-17 pairs of cerarii; anal pair with two,
the others with from 4 to 7 sharply conical spines. Lateral cerarn (fig. 3, c) with
2-3 auxillary setae and numerous, obtusely triangular pores rather closely grouped
together near the cerarn, but scattered and mergmg into the body-pores beyond.
Anal lobe cerarii (fig. 3, a) surrounded by a small circular chitinised area; ventral
surface of each lobe (fig. 3, 6) with a broad and somewhat rectangular chitinised
area. Anal ring well formed; setae (fig. 3, e) a very little shorter than the anal
lobe setae. Dorsal body setae (fig. 3, d) numerous, slenderly spinose, with, in
many imstances, flagellate tips. Obtusely triangular pores very numerous, imter-
mingled with larger simple pores ; and on the last few segments of the abdomen a
few multilocular pores ; tubular ducts very short and scanty (?) the rims ,of which
are chitinised. Dorsal osteoles large, in two pairs. Legs relatively stout.

The young adult female resembles the old adult, but has the gland-pores more
closely packed together.

OBSERVATIONS ON SCALE-INSECTS. 179

BritisH Guiana: “‘ Cattle Trail Survey,’ on an unknown plant; ‘‘ the insects
enclosed by ants (Acromyrmez sp.) in small paper nests,” 1919 (A. A. Abraham,
per G. #. Bodkin).

This species is very near P. comstocki, Kuwana, as defined by Ferris (Leland
Stanford Jun. Univ. Pub. “The California Species of Mealy Bugs,” p. 41)
but differs in having a larger number of cerarian spmes and in the form of
the chitinised area on the ventral surface of the anal lobes.

Taken in association with Lecanium inquilinum, Newst., and L. deformosum, Newst.

Fig. 3. Pseudococcus inquilinus, Newst., sp. n. 9;

a, aorsal and, b, ventral surface of anal lobes ; ¢, lateral

abdominal cerarii; d, body setae; e, hair of anal ring.
Pseudococcus perniciosus, Newst. & Willcocks, var.

Ovisac of female. Arranged m a similar way to those of typical P. perniciosus,*
but the more or less globular masses are smaller. The examples are so badly
weathered, however, that it is not possible to give details of the structure of the
mdividual ovisacs.

Female, adult. Under pressure of the covering glass the form is narrowly ovate.
Antennae of seven segments. One pair of cerari present on the anal lobes
(fig. 4, a), the spines sharp and somewhat slender, with a few obtusely triangular
gland-pores scattered around them. No typical lateral abdominal cerari present,
but their position is indicated by a single, faintly lanceolate spine (fig. 4, c), with
from 1 to 2 simple supplementary setae at some distance away from it: the spines are
traceable m some individuals on the last 3-4 segments, in others on the penultimate
segment only. Body spies (fig. 4, d) mmute, famtly lanceolate, and very scanty
indeed ; hairs small and also very scanty. Anal lobe setae (fig. 4, 6) shghtly longer
than the anal rime setae (fig. 4, e). Integument very thickly set with gland-pores (fig.
4, c, c), more especially so along the margin; these are of three kinds: multilocular,
tubular and obtusely triangular ; the first-named are arranged in narrow transverse
bands on the dorsal surface of some of the abdominal segments, elsewhere they are
regularly disposed on both surfaces. Length of 53 adult 99 varying between 2:1
and 2°7 mm. :

———

* Newstead & Willcocks. Bull. Ent. Res., i, p. 138 (1910).

180 ROBERT NEWSTEAD.

Larva. Anal lobes with a pair of slightly lanceolate spmes; setae longer than
those of the anal ring. Dorsal spies minute, similar m shape to those on the anal
lobes.

BritisH East Arrica: Kabete, on coffee, November 1918. ‘‘ Coffee bush infected
in the lab. has been killed by this scale” (F. W. Dry, for T. J. Anderson).

Typical examples of B. perniciosus, N. & W., have from 5-6 pairs of lanceolate
cerarian spies on the distal segments of the abdomen, with 2-3 obtusely triangular
pores scattered near them; and the body spines, though smaller than the cerarian
spmes, are relatively much larger, and also much more frequent than are those in
the variety from coffee at Kabete. Moreover P. perniciosus is much larger, measuring
from 3-4 mm.

-
1
i]
s
‘
8
J
'
1
8
+
‘
'
'
i}
¢

RB Perniciosus, var. FP. filamentosus.

Fig. 4. Pseudococcus perniciosus, N. & W., var., 2; a, b, dorsal and ventral

aspect of the anal lobes; c, lateral abdominal cerarii; dd, gland-pores ; e, hair of

analring. Pseudococcus filamentosus, Ckil., 2; f, gland-pores; gg, lateral abdominal
ceraril.

Brain * has sunk P. perniciosus as a synonym of P. filamentosus, CkiL, without
giving reasons for so domg; this action has led me to re-examime examples of the
latter (part of the type lot kindly presented to me by Professor T. D. A. Cockerell
in 1892) and I find that it differs to a marked degree from P. perniciosus in having
very few gland-pores (fig. 4, f), relatively shorter anal lobe setae and small groups
of obtusely triangular pores round the cerarian spines (fig. 4, g,g). Some of the
tubular ducts of the dorsum are also much shorter and many of them are accompanied
by 2-3 obtusely triangular pores. Clearly therefore P. pernaciosus is specifically
distinct from P. filamentosus, and the former name must be retamed.

* Brain, C. K. Trans. R. Soc. 8S. Africa, v, pt. 2, p. 99 (1915).

OBSERVATIONS ON SCALE-INSECTS. 181

Phenacoccus ballardi, Newst.

The original description * of this rather remarkable insect was given without
illustrations. I have thought it desirable, therefore, to give a photomicrograph
of both old and young adult females (Pl. xvi, fig. 4) m the hope that it will enable
students to determine the species with greater ease.

Pseudophilippia inquilina, sp. nov.

Female, adult. Form short ovate and slightly tumid. Colour in life mauve
pink. Dorsum nude; venter protected. by a thin vesicular glassy scale, which
is firmly attached to the bark of the food-plant. Antennae (fig. 5, @) and Jegs quite
rudimentary; the former, which are much shorter than the stigmata, are composed

Fig. 5. Pseudophilippia inquilina, Newst., sp. n. Q; a, antenna ;
b, leg; c, ¢,, ¢,, stigmata; d, gland-pores. Second stage @: e,
stigmatic cleft; f, anal lobes.

apparently of three segments, the apex having several stiff hairs. Legs (fig. 5, 6)
slightly smaller than the antennae, bare, composed of a single tubercular-shaped
segment, with a relatively large claw; lower digitules stout and bluntly pointed.
Stigmatic clefts (fig. 5, c) clearly defined. The stigmata (fig. 5, ¢, ¢,,¢,), which are
placed close to the cleft, are protected by a well-defined external arch (fig. 5, gl. a7.)
shaped somewhat like a horse-shoe in minature, and closely set internally with
thick-rimmed multilocular gland-pores. Anal cleft nearly four times the length
of the lobes; the latter surrounded by a distinct chitimous arch. Venter rather

* Newstead, R. Bull. Ent. Research, viii, p. 17 (1917).

182 ROBERT NEWSTEAD.

thickly studded with circular gland-pores (fig. 5, d). Dorsum with numerous spines,
especially towards the margin. Rostrum relatively large; the filaments protected
by a very long membranous sheath. Length, 4°2-6°8 mm ; width, 4°3-5°5 mm.

Female, second stage. Short ovate; dorsum flat. Colour and the structural
characters of the antennae and legs as in the adult female. Stigmatic clefts (fig. 5, e)
relatively deep. Stigmatic spmes two, one on either side of the external glandular
arch ; the latter much more extended towards the margm, and with fewer gland-
pores than in the adult; fulcrum to the atrium of the stigmata strongly produced.
Anal lobes (fig. 5, /) with a tongue-shaped sclerite between them at the base.

Larva. Ovate. Antennae and legs well developed ; the former of six segments,
of which the 3rd is much the longest and equal m length to the 2nd, 4th and 5th
together. Anal lobes large; apical hair very long and stout. One large stigmatic
spime in each cleft. Marginal hairs m an irregular double row. Abdommal segments
with a transverse series of hairs, slightly smaller than those at the margin. Rostral
filaments as long, apparently, as the circumference of the body.

JAMAICA: on the banks of the Great River, near Montpelier; attached to the
bark of an unknown tree beneath a large, blackish coloured “‘ paper” nest of
Cremastogaster brevispmnosa, Mayr, var. tumulifera, For. The nest in question
was attached to the bole of the tree about six feet from the ground; 10.xu1.1908
(R. Newstead). ,

I have placed this rather remarkable msect in Cockerell’s genus Pseudophilippia
as it agrees best, m its morphological characters, with Cockerell’s diagnosis ; but
the absence of an ovisac may be thought by other students to preclude its admission
here. It seems to me, however, that the presence of a glassy ventral scale beneath
the body of the female and the curiously protected stigmata do not m themselves
call for the erection of a new genus.

Antonina waterstoni, sp. nov.

Female, adult. Colour, in life, pale buff to dusky buff. Form flat, narrowly
ovate to very elongate, broadly rounded in front, widest generally in the region of
the proximal segments of the abdomen, narrowing rather suddenly behind; last
two segments of abdomen (fig. 6, a) strongly constricted. Antennae quite rudimentary
and apparently unsegmented ; they are placed quite close to the margin. Legs
absent. Mentum very small and unisegmented; just below it there are several
minute tubercles, each with a short stiff hair. Stigmata large and widely separated ;
first pair in a line with the rostrum ; the small group of parastigmatic glands merging
into those at the margin. On the dorsal surface just behind the 2nd pair of stigmata is
a large group of minute spines (fig. 6, b) occupying the whole width of the 1st proximal
segment of the abdomen. Margin, all round, with an almost contmuous band of
relatively large pores (fig. 6, c) and a few minute pointed spines. Anal segment
(fig. 6, a) markedly distinct ; dorsal surface with a pair of forwardly directed bristles ;
ventrally it is almost covered with pores, which are almost as numerous as those
on the preceding segment. Anal lobes quite rudimentary, each bearing a few stiff
hairs. Anal ring (fig. 6, a,) placed in a slight depression; hairs six in number, and
rather stout. Length of young adult, 3-42; width, 2-32 mm. Old adults
measure : length, 5°7-6 mm ; width, 3-3-2 mm.

OBSERVATIONS ON SCALE-INSECTS. 183

The dorsum and venter are very thinly protected with white powdery wax, this
secretion adheres to the food-plant, but readily comes away from the insect ; beyond
the body, especially in the posterior region, the secretion is dense and completely
fills the narrow space between the stem and the leaf-sheath. The general of facies
the female, together with the secretionary matter (ovisac), bears a striking
resemblance to Aclerda berleser, Bufta.

RW.
pL aim

Fig. 6. Antonina waterston, Newst., sp. n., 9; a, terminal seg-

ments of abdomen; a,, terminal segment of abdomen with anal ring,

etc.; b, gland-pores and body spines; c, marginal gland-pores ;
d, dorsal aspect of pupa; e, leg i. of pupa; f, antenna of larva.

Male (fig. 7, a). Rather robust. Head almost as broad as the thorax, articula-
tion faint. Eyes well within the margin, and enclosed by two longitudinal curved
sclerites ; ocelli slightly smaller, placed between the antennae, just within the margin
of the frons, ventrally. Antennae (fig. 7, b) of nine segments; Ist and 2nd very robust;

184 ROBERT NEWSTEAD.

3rd and 4th very slender and markedly narrower than the rest ; 4th much the smallest
and less than half the length of the 3rd ; all the segments with the exception of the
4th and 9th with short stout, bluntly pointed spines ; terminal one with several long
hairs. Leg i. (fig. 7, c) with the tibia and tarsus more robust and much shorter
than the corresponding segments in legs i. and il. (fig. 7, d); all the tibiae with
distal spines, ventrally. Genital sheath short. Caudal bristles long. Wings long,
but rather narrow. Length to end of genital sheath, 0°8.

Pupa (fig. 6, d). Robust. Antennal sheaths short and composed, as far as one
can ascertain, of nine segments. -Leg i. (fig. 6, e) more robust than legs ii. and i.
Claws to tarsi very slender. Length, 0°8 mm.

Fig. 7. Antonina waterstoni, Newst., sp.n., ¢; a, ventral view ; b, antenna ;
c,legi; d,legili; e, vestigial buccal cavity (?)

Puparvum of Male. Three of these were found closely packed together under
a leaf-sheath in association with the females. Collectively they formed an
irregular mass of white, loosely felted and brittle strands of wax, completely enclosing
the pupae.

Larva, adult (¢ Male). Very elongated, parallel-sided. Antennae (fig. 6, f) of six
segments ; 3rd shortest ; 6th longer than the first three together ; all the segments.
with fine hairs, the 5th and 6th with long slender spinose ones. Anal segment of
abdomen with a long stiff bristle indicating the position of the lobes. Anal ring
with six long hairs. Rostrum very broad. Eyes small but prominent. Length,
055 mm.

MacrepontA: beneath the leaf-sheaths of Arundo phragmites, 1917 (Capt. James
Waterston).

OBSERVATIONS ON SCALE-INSECTS. 185

It affords me infinite pleasure to dedicate this newly discovered species to our
esteemed colleague.

The males were all dead, and more or less imperfect ; and although one failed to
trace the long caudal filaments, the presence of these structures is indicated by the
setae which supports them in life. All the examples were lying beneath the leaf-
sheaths and were flattened out as if by pressure of the sheaths.

The male of Antonina australis, Green, has been seen by Froggatt (Agricultural
Gazette, N.S.W. No. 742, p. 3, 1904), but so far as I can ascertain it has not been
described. The discovery therefore of the male of A. waterstona is of interest, as
the members of this sex in all the other species hitherto described are unknown.

The female of A. waterstoni is nearly related to A. socialis, Newstead, but differs
in having much smaller antennae, in the presence of a large isolated group of minute
spines, and in the character of the anal segment, including also the relative position
of the anal ring.

Pseudokermes marginatus, sp. nov.

Female Test. Roughly hemispherical ; glassy white, with a median longitudinal
suture, which renders the two halves easily separable ; dorsal surface with faint traces
of small and somewhat rectangular patches of secretion ; sides with wavy conchoidal
striae ; stigmatic ridges more or less distinct ; margin wavy.

Female, adult. Shape somewhat like that of a soldier’s steel helmet in
miniature, with a narrow mediodorsal ridge, a relatively very broad flat margin
(rim) and prominent anal lobes. Surface faintly uneven but shining. Colour
pale castaneous. Boiled in KOH, the integument of the dorsum changes to pale
straw-colour ; the broad flange becomes quite transparent, and the extreme margin
brownish—the three grades of chitin showing in marked contrast to one another.
Antennae represented by exceedingly minute tubercles bearing 4-5 short stout setae.
Legs entirely absent. Margin wavy and irregular. Stigmata robust, somewhat
cylindrical, and externally obtusely conical. Stigmatic clefts and spines absent.
Marginal spines relatively stout, acutely pointed, and very widely separated. Anal
cleft deep ; lobes somewhat triangular, the proximal and inner margins longer than the
outer ; apices with several fine hairs. Dorsal gland-pores in the median longitudinal
ridge, circular, surrounded by a small pale area, and often divided into two linear
groups. The broad flat marginal flange, in very old and heavily stained examples,
with numerous cell-like clear areas, the inner series forming an irregular dactyliform
pattern; in younger forms these structures are wanting and in their place are seen
a large number of narrow tubular ducts. The extreme margin presents, on its inner
surface, an irregular crenulated appearance, the depressions occupied by a rather
ill-defined duct. Length, 2°1-1°8 ; width, 2°1-2 mm.

British Guiana; Ituni Savannah, on Nectandra sp.," 28.11.1919 (A. A. Abraham
per G. EH. Bodkin).

The test of the female resembles that of the young forms of P. nitens, Ckll., but the
fine vertical striae are wanting and the surface is much more uneven. In the female
of P. nitens the broad margin is wanting; the integument is uniformly membranous
after maceration in KOH, and minute vestigial legs are present.

Male puparia not observed.

186 ROBERT NEWSTEAD.

Pulvinaria brevicornis, sp. nov.

Female, adult. More or less oval in outline and highly convex, or sub-hemispherical ;
generally with two longitudinal rows of rather deep pits, one on each side of the
median line. Colour, in alcohol, varying from pale buff to pale castaneous ; some
are unicolorous, others with two interrupted longitudinal black limes following the
course of the pits, the outer line, in some examples, giving off lateral lmes on the
abdommal segments. Integument thin. Antennae (fig. 8, a) relatively very short
and robust, equal in length to the anterior tibio-tarsal segments together; of 6
segments (the articulations somewhat ill-defined in some examples); 5th and 6th
each with a rather long slender spine. Legs (fig. 8, b) short and very robust.
Stigmatic clefts obsolete ; spines three (fig. 8, c), stout, the central one generally slightly

Fig. 8. Pulvinaria brevicornis, Newst., sp. n., 9; a, antenna;
b, leg ; ¢, ¢,, stigmatic and marginal spines ; d, anal lobes ; e, preanal
gland-pores; f, ventral tubular ducts; g, stigmatic gland-pores.
Male, pronymph : h, antenna; 1, leg.
longer than the laterals, but in some instances it is of the same length as the others.
Marginal spines (fig. 8, c,) simple, pointed and rather widely separated. Anal cleft
short, or two to three times longer than the lobes. Anal ring of 10 hairs. Anal
lobes (fig. 8, d) with the proximal margin much longer than the distal margin ;
apex with several hairs. Dorsum, in heavily stamed preparations, with widely
separated, broadly oval or subcircular cells. Venter with innumerable circular
gland-pores, the tubular connections of which (fig. 8, f), are suddenly truncate on
one side near the proximal end and furnished with a rosette-like extension.
Length, 3-3°75 mm ; width, 1°75-3 mm.
British Guiana: Turkeyn, East Coast, on Avicennia mitida, 22.vi.1917 (G. H. Bodkin).
The mtegumental characters of this msect are unusual, and should serve, together
with the form of the antennae, the anal lobes and stigmatic spmes, to distinguish it
from its allies.

OBSERVATIONS ON SCALE-INSECTS. 187

Pulvinaria broadwayi var. echinopsidis, nov.

Female, adult. Ovate, usually very slightly narrowed in front. Antennae of eight
segments ; 3rd a little longer than the 2nd ; formula 8, 2, 8, 1 (4, 5, 6, 2) or 3, 2, 8, 4
(5, 6, 7). Legs robust and relatively long. Lower digitules strongly merassate
proximally and broadly dilated distally. Stigmatic clefts (fig. 9, a) very shallow ;
spines three, all of them stout, the middle one usually a little more than twice the
length of. the laterals. Marginal spines (fig. 9, a) set rather closely together and
of two types—one relatively short and simple, the other longer and slightly divided

Fig. 9. Pulvinaria broadwayi var. echinopsidis, Newst., nov., 9;
a, stigmatic cleft with marginal and stigmatic spines ; b, anal lobes.

at the tip. Dorsum without glands or cellular structures; venter crowded with
glandular tubes, more especially so m the abdommal region. Anal lobes (fig. 9, 6)
rather narrow and furnished distally with several long hairs. Anal ring of eight
hairs, of which one pair is much smaller than the rest. Length, 2-15; width,
1-1°3 mm.

Ovisac. More or less rounded and formed of loose and somewhat brittle material,
at the side of which the shrivelled body of the female rests. Greatest width,
15-3 mm.

British GUIANA: Botanic Gardens, Georgetown, on Echinopsis latiflora, 1x.1918
(G. EH. Bodkin and H. Morrison).

Lecanium subacutum, sp. nov.

Female, adult (fig. 10, a). Colour of dead examples pale dusky yellow. Flat and
very thin; dorsum wrinkled, the wrmkles at the margin radial. Form long and
narrow ; extremities subacute ; the length three times as great as the greatest width ;
one side of the body is usually more or less straight, the other strongly arched.
Antennae (fig. 10, 6) of six segments, the third very long, and almost equalling
the length of the 2nd, 4th, 5th, and 6th together. Legs long and slender. Stigmatic
clefts (fig. 10, c) shallow ; spines three, all of them very stout and bluntly pointed ;
the central one rather flattened and more than twice the length of the laterals.
Marginal spines (fig. 10, c) simple, rather stout and strongly curved backwards ;
they are placed very closely together, so that the tip of each spme almost reaches
the strongly curved portion of the spine below it. Anal lobes (fig. 10, d) long and
narrow ; the length equallmg that of the 3rd segment of the antennae. Dorsum
with numerous large circular gland-pores (fig. 10, e, e) having strongly chitinised rims
and fine granular centres; they are irregularly scattered over a relatively broad
area between the anal lobes and the antennae. Dorsal spines (fig. 10,e, e) short,
stout, and bluntly pointed. Dermal cells absent. Submarginal tubercles in two

188 ROBERT NEWSTEAD.

pairs: one anterior, the other posterior. Anal cleft relatively short, and from
one-seventh to one-ninth the length of the body. Length, 2-4-3:1 mm.

Young females are relatively narrower than the adults; but do not otherwise
differ from them.

Male puparwm. Glassy white. Very elongate ; outline similar to that of a young
female ; median plate or “ coronet”’ long and very narrow ; stigmatic ridges well
defined ; margin with one division along the line of the lower pair of stigmata.
Length, 1:8-1°9; width, 0°5-0°6 mm.

= Pore
P
f a

\ mY |
| |
\

/
/
f

Sade” \ mt

Fig. 10. Lecaniwm subacutum, Newst., sp.n., 2; a, adult
2; b, antenna; c, marginal and stigmatic spines; d, anal
lobes; e, e, dorsal gland-pores and spines,

Ucanpa: Jana Isl., Sesse Islands, Lake Victoria, on Coffea robusta, 9.x.1918 ;
Butumira Isl., Sesse Islands, on the leaves of an unknown plant, 12.x.18 (C. C.
Gowdey).

In both imstances this species was living in association with Aspidiotus articulatus
var. magnospinus, Newst.

Lecanium (Eucalymnatus) decemplex, sp. nov.

Female, adult. Circular, or more or less so; fiat and very thin. Colour trans-
lucent amber-yellow, often with a tinge of red or pale castaneous. The whole of the
dorsum covered with a thin hard glassy test, the presence of which is extremely
difficult to detect and which is equally hard to detach. Boiled in KOH the female

OBSERVATIONS ON SCALE-INSECTS. 189

(fig. 11, a) presents the following morphogical details: Dorsum divided into five
bilateral plates—two cephalic, three thoracic, and two abdominal; the sutures
separating the cephalic from the first thoracic plates terminating at the stigmata ; the
other sutures are connected with the mesal one (these sutures in the dried examples
appear as well defined narrow ridges). Mesal suture between the anal lobes and the
rostrum with large circular gland-pores (fig. 11, 6). Antennae of six segments;
the third almost as long as the 1st, 4th, 5th and 6th together. Legs well developed ;
lower digitules large and strongly incrassate. Anal lobes (fig. 11, c) somewhat
triangular ; inner margin longest; distal margin shortest; the sclerites beneath
(fig. 11, d,d) stout and somewhat spine-like. Anal cleft deep,and apparently partly
fused but separable. Stigmatic clefts (fig. 11, e, e) small, but deeply invaginated;
spines three, very robust, and blunt at the tips. Marginal spines (fig. 11 e. e) simple,
and set rather widely apart. Submarginal gland pores (fig. 11, f) very small, number-
ing from six to eight on either side. Oval cells can be seen, near the margin, in old
and well stained examples. Length, 3°4-3°7 ; width, 3-2-3-°7- mm.

Fig. 11. Lecaniwm (Hucalymnatus) decemplex, Newst., sp. n.,
2; a,adult 9; b, mesal gland-pores ; c, anal lobe; d, d, sclerites
of anal lobes ; e, e, stigmatic clefts and spines.

Male Puparium. Broadly ovate; divided into eleven plates: two median, one
cephalic, and four bilateral; the lateral plates with partial subdivisions. Length,
1°8; width, 1:4 mm.

British Guiana: Ayaria, Thuraka Lake, Ituribisci Creek, Essequebo, on leaves
of Lecythis sp., 6.x.1918 (G@. H. Bodkin).

This somewhat remarkable species evidently belongs to the subgenus Hucalymnatus,
its distinguishing features being the small number of plates into which the dorsum is

divided.

Lecanium inquilinum, sp., nov.

Female, adult. Ovate or elongate and highiy convex ; more or less circular when
mounted under pressure. Integument pale brown, but thin and transparent after
maceration in KOH. Antennae (fig. 12, a) of eight segments. Legs relatively robust

190 ROBERT NEWSTEAD.

and rather long ; digitules normal. Anal cleft free and a little more than twice the
length of the lobes ; the latter (fig. 12, b) rather elongate and obtusely rounded distally.
Stigmatic clefts very small ; spines three, all robust and blunt, the laterals almost
equal in size to the central one. Marginal spines (fig. 12, c) long and very acute,
the tips in many cases appearing almost flagellate; they are placed very closely
together and are continuous along the stigmatic clefts. Stigmata with a large trumpet.
shaped peritreme, and placed unusually near the margin of the body. There is a
large closely packed group of multilocular gland-pores on the inner walls of the anal
cleft, close up to the lobes. A few minute gland-pores occur in the densely chitinous
patch surrounding the anal lobes. Length, 1°7-2°3 ; width, 1:4-1°8.

Female, young adult. Similar to the old adult, but the marginal spines are much
more bluntly pointed and the chitinous patch surrounding the anal lobes is wanting.

Britisa Guiana: “Cattle Trail Survey,”’ the insects enclosed by ants (Acromyrmex
sp.) in small paper nests 1919 (A. A. Abraham per G. E. Bodkin).

ALA AU

Fig. 12. Lecanium inquilinum, Newst., sp. 1.,
Q; a, antenna; b, anal lobes; ¢, marginal spines.

The form and arrangement of the marginal spines recall those seen in certain
species of Pulvinaria; but the species clearly belongs to Lecanium and somewhat
resembles the hollow hemispherical species. Taken in association with Lecanium
deformosum, sp.n., and Pseudococcus inquilinus, sp. n.

Lecanium (Eulecanium) deformosum, sp. nov. bi

Female, adult (fig. 13, a-d). Dorsum rather flat; sides relatively thick ; general
form very irregular and distorted, some examples being broader than long and others
more or less elongate, but the margins in all cases (12 examples) are asymmetrical
and often distorted to a marked degree. Antennae (fig. 13, e) of six segments ; the
3rd and 6th longest. Legs with the tibio-tarsal segments either distinctly articulated
or partly so, rarely completely fused; leg i. (fig. 13, f) with the tibia generally strongly
curved. Anal lobes (fig. 13, 9) very broadly dilated distally. Anal cleft faintly fused,
but easily separated after maceration in KOH. Stigmatic clefts small, or seated
in faint depressions; spines (fig. 13, h) three, the laterals normally very short, stout,

OBSERVATIONS ON SCALE-INSECTS. 191

and obtusely pointed : in one example (fig. 13, 4,) the lower group on one side has the
lateral spines as long and as stout as the central one. Marginal spines (fig. 13, 2)
simple and hair-like. Dorsal gland-pores minute and widely separated. In well
stained specimens the integument of the dorsum is faintly divided into broad plate-
like radial areas enclosing numerous irregular pigmented markings. Length, 1:2-
1°8; width, 1:1-1°5, mm.

British Guiana: ‘Cattle Trail Survey,” on an unknown plant, the insects
enclosed by ants (Acromyrmex sp.) in small paper nests, 1919 (A. A. Abraham
per G. E. Bodkin).

Fig. 13. Lecanwwm deformosum, Newst., sp. n., 2; a-d,
outline of four adults; e, antenna; f, leg; g, anal lobes;
h, h,, stigmatic spines ; 2, marginal spines.

A small oviparous species, remarkable for its markedly deformed shape, the
presence or absence of the tibio-tarsal articulation, and the apparent variability of
the stigmatic spines. Found in association with Pseudococcus inqulinus, sp. n.,
and Lecaniwmn mquilinum, sp.n.

Lecanium (Saissetia) nigrum var. nitidum, nov.

Female, adult. Usually more or less hemispherical, but some examples are slightly
ovate and narrowed in front; margin markedly flattened, and often with regular
rectangular patches of silvery secretion ; dorsum smooth and shining; anal lobes
usually porrect. Colour varying according to the age of the individual: young forms
pale red-brown, old adults rich dark castaneous. Eyes relatively large, black and
prominent. Antennae (fig. 14, a) of eight segments; the 3rd scarcely as long as the 4th
and 5th together. Legs rather slender. Derm cells forming a closely reticulated
pattern as in Lecamwum nigrum, Niet. Stigmatic clefts very shallow ; spines three,
the central one about five times the length of the laterals, (fig. 14, b). Marginal

(C605) rr

192 ROBERT NEWSTEAD.

spines (fig. 14, c) short and sub-palmate, the distal portion being broadly flattened
and deeply divided. Anal lobes (fig. 14, d) short and obtusely rounded ; inner
margin slightly longer than the proximal; external margin strongly arched. Length,
1-8-2 mm.

Female, young adult. Flat ; straw-coloured when dry. Form when mounted under
pressure somewhat ovate. Antennae (fig. 14, a) with eight segments and similar in
form to those of the adult. Legs i. and ii. much longer than leg i. Stigmatic clefts
(fig. 14, b) mdicated by a very slight indentation of the margin ; spines three, the
laterals minute and acutely pointed. Marginal spines (fig. 14, 6, c) as in the adult.
Anal cleft free, its length slightly less than twice the length of the lobes ; the latter
(fig. 14, d) as in the old adults.

Ucanpa: Bukeke Isl., Sesse Islands, Lake Victoria, on Luzebarziba, 9.x.1918
(C. C. Gowdey). |

Fig. 14. Lecanium nigrum var. nitidum, Newst., 92; a,
antenna; b, stigmatic and marginal spines; c¢, marginal
spines; d, anal lobes. ‘ :
Its small size and generally hemispherical form, together with the highly polished
integument and the sub-palmate marginal hairs, are the distinctive features of this
well-marked variety.

Platysaissetia montrichardiae, sp. nov.

Female, old adult. Blackish or sooty brown. (Generally more or less ovate and
very slightly narrowed in front ; but occasionally the outline is irregular, indented,
or markedly asymmetrical; dorsum very low convex or almost flat, scabrous, the minute
elevations often carrying small particles of the test of the young female. Vertical
sides markedly shallow. Pseudo-margin relatively faintly produced. Tubular
glands at the extreme margin. Stigmatic clefts rarely traceable. Venter hollow,
but more or less covered with a rather thick pellicle of white wax. Dermal -
cells (fig. 15, a, from an unstained preparation) very irregular in outline, each pale
area with a distimct pore at the extreme edge ; the walls of each cell area thick, dark,
and very regular. Other details as in the young adult. Length 4-5:2; width
2°9-3°9. mm.

OBSERVATIONS ON SCALE-INSECTS. 193

Female, young adult. Pale brown to dull castaneous. Dorsum flat and covered
with a dirty white, glassy test, consisting of mmute plates, which collectively present
a faintly imbricated appearance. The test is easily deciduous, and when removed
the integument presents a polished appearance. Antennae (fig. 15, 6,5) of eight seg-
ments, of which the 4th, 5th and 8th are the longest ; formula: (4, 5, 8) 3, 2, (6, 7)
1 or 5 (4, 8, ) 3 (2, 6, 7) 1; terminal hair of great length. Legs relatively slender ;
tarsus of anterior pair (fig. 15, c) with a well defmed constriction ; coxa and trochanter
each with a very long hair. Stigmatic clefts obsolete ; spines (fig. 15, d,d) three, the
central one equal in length to the 4th and 8th segments of the antennae ; lateral
spines poimted and about one-fourth the length of the central one. Marginal spines
(fig. 15, d) slightly shorter and stouter than the lateral stigmatic spines ; they are

Fig. 15. Platysaissetia montrichardiae, Newst., sp. n., 2;

a, dermal pores and “‘cells.””> Youngadult 9; b, b, antennae ;

c, leg; d, d, stigmatic spines ; e, post anal gland-pores ; f, leg
of larva.

separated by a distance equal to three or four times their length. Dermal gland-
pores irregularly ovate, large, placed close together, and most conspicuous towards
the margin. Anal lobes with the outer margin strongly arched ; and surrounded
by a narrow wall of dense chitin. Anal cleft fused ; from one-fifth to one-sixth
the length of the body. Postanal glandular pores (fig. 15, e) relatively large and
circular, forming a broad scattered group extending as far forward as the mentum.
Length, 2°6-3°3 mm.

Female, second stage. Differs from the adult chiefly in the following details :
Antennae of six segments, of which the third is much the longest. Postanal gland-
pores absent ; postanal bristles in three pairs. Length, 0°7-0°8 mm.

(C605) H2

194. ROBERT NEWSTEAD.

Male Puparium. Oblong, opaque, glassy, white; anal cleft distinct; surface
composed of minute rough polygonal plates ; the marginal series forming a roughly
serrated fringe. Length, 1°5 mm.

Larva. Differs from the second stage female in the followmg details: Antennae
with the 3rd and 6th segment equal and longest. Legs (fig. 15, f) with the distal
femoral bristle of great length ; tarsal digitules markedly unequal in thickness and
length respectively, the longer one arisimg from the tarsus some distance behind the
smaller one; digitules to the claw normal. Anal rmg with six hairs. Length,
exclusive of the caudal bristle, 0°5 mm.

Britiso Guiana: Ikruaka Lake, Essequibo, on Montrichardia aculeata, 11. m.1917
(G. EB. Bodkin).

A very heavy infestation, so much so that the insects covered a very large propor-
tion of the branches.

A somewhat remarkable species, distinguishable, in the adult female, by the curious
character of the derm and the relatively short anal cleft (0°7-0°8 mm.); in the second
stage female by the absence of marginal cylindrical ducts to the glands ; and in the
larva by the unusual character and relative position of the tarsal digitules.

Aspidiotus longispina, Morgan.

Aspidiotus longispina, Morgan, Ent. Mo. Mag. xxv, p. 352, pl. v, fig. 1 (1889).

Morganella maskellt, Cockerell, Bull. 6, T.S., U.S.A. Dept. Agric., p. 22 (1897).

Hemaberlesia longispina (Morg.), Leonardi, Riv. Pat. Veg. vi, p. 120, fig. 4 (1897).

Asprdiotus longispina var. ornata, Maskell, Trans. N.Z. Inst., xxx, p. 225 (1897).

Aspridiotus (Morganella) maskelli (Ckll.) Brain, Bull. Ent. Research, ix, p. 136, pl. vii,

fig. 109 (1918).

Female Puparwm (fig. 16).—This, when perfect, is narrowed, strongly produced,
and slightly involute posteriorly, resembling the long curved toe of an Oriental slipper
in miniature. This very remarkable appendage is composed of both dorsal and
ventral portions of the capsulate puparium ; but it is rarely found intact in examples
which have been submitted to even light pressure, as it 1s very brittle and readily
breaks away.

Fig. 16. Aspidiotus
longispina, Morgan,
2, puparium.

Male Puparwum. Elongate, exuviae terminal or sub-termimal. Colour as in the
female puparia.

Male. Not differmg in its morphological characters from typical members of
the genus.

British Guiana: Botanic Gardens, Georgetown, on papaw, 1919 (G. EH. Bodkin).

OBSERVATIONS ON SCALE-INSECTS. 195

The synonymy given above is, I believe, correct, the determmation bemg based
upon an examination of material received from both Morgan and Maskell. A co-type
° from Morgan in my collection was origmally mounted in Canada balsam without
staiming, so that the true form of the strikmgly characteristic squamae on the
pygidium of the 2 could not be detected and was thus overlooked by Morgan in the
first instance, and subsequently also by Leonardi, to whom I sent my example for
examination, and for the purpose also of figuring it m his memoir (I.c.). In 1906 I
had occasion to stain the female given to me by Morgan for comparison with examples
submitted to me by Kotinsky, on Ficus sp., from Honolulu. I then found that the
squamae were strikingly different from what they appeared to be when unstained.
Subsequently I stained and mounted an example of Maskell’s A. longispina var.
ornata and found it to be specifically identical with Morgan’s co-type female. Clearly
therefore the specific names maskella of Cockerell and ornata of Maskell must sink.

Aspidiotus (Chrysomphalus) apicatus, sp. nov.

Female Puparium. More or less circular, convex and very thick ; covered with
a relatively thick epidermal layer of the bark ; colour, when denuded, opaque black,
larval exuviae nude, shining black, forming a well defined nipple ; second exuviae
black ; ventral surface shining black. Ventral pellicle rather stout ; white or dusky
white, with a dark brown or blackish periphery. Diameter, 1°6-1°9 mm.

Fig. 17. Aspidiotus (Chrysomphalus) apicatus, Newst., sp. nz,
2; a,adult; 6, thoracic spine ; ¢, pygidium ; d, paraphyses of do. ;
e, fringe of do.; f, anal plates, etc., of larva.

Female, adult (fig. 17, a). Broadly ovate, with a well marked and rather highly
chitmised, cephalic projection, and immediately below it a minute spiny process
(fig. 17, 6). Margin of body more or less strongly and finely crenulated. No parastig-

matic glands. Pygidium (fig. 17, c) broadly rounded, with a well marked callosity
extending along it proximally, from which there arises on either side a somewhat

196 ROBERT NEWSTEAD.

lanceolate sclerite (fig. 17, d). Margin (fig. 17, e) with nine pairs of well defined para-
physes and a few rather indistinct ones beyond them ; those arising from the distal
lateral portions of the lobes with the sides almost parallel and the ends more or less
suddenly truncate. Lobes in four pairs; the 2nd, 3rd and 4th pairs dentate.
Length, 1-1:1 mm.

Male Puparium. Broadly ovate, with the ends equally rounded ; colour dark
brown or blackish, but covered with the epidermal layer of the bark ; larval exuviae
black and more or less exposed; under surface dark brown or piceous, the central
area partly covered with a white mealy secretion. Uength, 9'5-0°S mm.

Larva. Broadly ovate. Antennae of five segments, of which the 5th is much the
longest and distinctly ringed. Pygidium (fig. 17, f) with two pairs of finely dentate
lobes ; the median pair much the larger and converging towards the middle line.

British GUIANA: Enmore Forest, East Coast, Demerara, on Avicenmia nitida,
1x.1918 (G. H. Bodkin and H. Harrison).

Aspidiotus (Chrysomphalus) umboniferus, sp. nov.
Female Puparium. More or less circular to broadly pyriform; flat, thin, and
somewhat transparent; pale reddish-brown to pale chocolate-brown. Exuviae

Fig. 18. Aspidiotus (Chrysomphalus) wmboniferus, Newst.,
sp. n., a, larval and nymphal pellicles; b, adult 9° after
treatment in KOH; ¢, unrestored 9; d, antenna of ¢ ;
e, pygidium of @.
central or subcentral ; the first nipple-like and black when denuded ; the second,
sooty brown. Ventral pellicle very delicate and composed generally of concentric
rings. Diameter, 1°6-2 mm.
Male Puparium. Elongate ovate, slightly narrowed posteriorly. Colour similar
to that ofthe female. Length, 1 mm.
Femaie (fig. 18, 6, c). Very broadly pyriform; width of cephalothorax slightly
greater than the entire length of the body ; metathorax with a relatively large,
blunt, marginal tubercle, which, together with the margin below it, is strongly

OBSERVATIONS ON SCALE-INSECTS. Us Vie

chitinised. Pygidium (fig. 18, e) unusually narrow and bluntly pointed. Circum-
genital glands in four groups, the posterior lateral groups placed in a line with the
vaginal orifice ; formulae of four individuals :—
8-6 66 7-8 6-6
66 66 56 6-6

Lobes in three pairs ; median pair somewhat quadrate, broader than long; 2nd
and 3rd pairs slightly shorter than the first and more than twice as broad as long.
Squamae between the»median and 2nd lobes, very short; the second of the
proximal pair, immediately beyond the third lobe, trifid. Paraphyses in six pairs,
all rather narrow but clearly defined. Dorsal gland-pores small and few in number,
with the tubular ducts filiform.

British Guiana: Ayaria Creek, Fssequibo, on Lecythis sp., 6.x.1918 (G. H. Bodkin).

In the structurat details of the margin of the pygidium this species very closely
resembles A. perseae, Comst., but it has a much more strongly pointed pygidium,
fewer circumgenital glands, and an extra pair of paraphyses ; these characters, taken
in conjunction with the large metathoracic tubercles and the strongly chitinised
margin below them, form the salient features of this Coccid. Aspidiotus (Chrysom-
phalus) paulistus, Hempel, also possesses very distinct cephalo-thoracic tubercles ;
but in this species there are only two pairs of paraphyses and the pygidium is
relatively shorter and much more rounded distally.

Aspidiotus (Selenaspidus) articulatus var. magnospinus, nov.
This variety (fig. 19, a) differs from typical Aspidiotus articulatus, Morgan, in the
following details : Cephalothoracic margin finely but distinctly serrated. Thoracic

Fig. 19. Aspidiotus articulatus var. magnospinus,

Newst., 9; a, adult; b, thoracic spine; e¢, fringe

of pygidium; d, thoracic tubercle of typical ©
articulatus for comparison.

198 ROBERT NEWSTEAD.

spine (fig. 19, 6) relatively very large, curved and rather acutely pointed, the contour
gradually merging into the cephalo-thoracic margin, and considerably longer than its
greatest width. In arteculatus, the thoracic spine (fig. 19, d), drawn to the same
seale as in fig. b, is suddenly produced, short, and bluntly pointed, and its length
about equal to its greatest width,

Ucanpa: Bufumira Isl., Sesse Islands, Lake Victoria, on the leaves of an
unknown plant, 12.1x.1918 (C. C. Gowdey).

Aspidiotus (Selenaspidus) kamerunicus, Lind., has a similar thoracic spine and a
serrated margin, but in this species the broad palmate squamae, between the second
pair of lobes and the spiny process, are replaced by squamae of a much narrower

type.

Aspidiotus (Odonaspis) rhizophilus, sp. nov.

Female Puparium. Dense, hard and capsulate, but the two halves slightly
separated posteriorly. Form irregular, but old examples are slightly narrowed
and produced posteriorly ; convex above and flat beneath. Larval exuviae generally

Fig. 20. <Aspidiotus (Odonaspis) rhizophilus, Newst., sp. n. ©;
a, pygidium ; b, fringe of pygidium.
towards the anterior margin, greyish in colour and sometimes fissured. Ventral
pellicle with a greyish patch towards the anterior margin. Texture rather rough ;
colour dull black or brownish black. Length, 13-15 mm.

Female, adult. More or less circular, or peg-top-shaped. Rudimentary
antennae with, apparently, only one stout bristle. Stigmata surrounded by
intricate folds of the integument; no gland-pores. Pygidium (fig. 20, a) very
short and broadly rounded distally ; circumgenital glands absent. Dorsal gland-
pores small and few in number, the tubular ducts relatively long and_ slender.
Ventral gland-pores similar to those on the dorsal surface, but more numerous.
Vaginal orifice nearly opposite the anal orifice. Margin (fig. 20, 6) with a series of
irregular, closely adjacent, squamate plates, of which the larger pair are evidently
homologous with the median lobes in typical members of the genus Aspidiotus.

OBSERVATIONS ON SCALE-INSECTS. 199

British East ArricaA: Kabete, on roots of Chloris incompleta, 7.x.1918
(R. H. Deakin, per F. W. Dry).

Aspidiotus fiorineides, sp. nov.

Female Puparium (fig. 21, a). Attached to the edge of the leaf, with equal portions
on both sides. Very elongate; sides compressed ; middle line of dorsum rather
sharply keeled. Exuviae central, bright orange-yellow or pale castaneous.
Secretionary portion very broad, thin, semi-opaque, dusky white. Length, 2-22 mm.

Male Puparium. Similar to that of the female, but much smaller. Larval
exuviae central, and orange-yellow in colour.

Fig. 21. Aspidiotus fiorineides, Newst., sp. n.; a, lateral

view of 9 puparium on edge of leaf; b, adult @; e¢, terminal

portion of body with pygidium ; d, pygidium of @; e, exuviae
(‘‘ pellicle ’’) of second stage 9.

Female, adult (fig. 21, b). Narrowly elongate; the length four times as great as
the width ; extremities narrowed. Dorsum strongly convex ; venter flat. Rostrum
subcentral. Body, with the exception of the pygidium, densely chitinised ; the
chitin at the junction with the pygidium appearing deeply divided. The integument
of the pygidium membranous, and in several specimens turned backwards, looking
as if it were completely invaginated ; but when extended (fig. 21, c) it is very long,
narrowly rounded, and has no circumgenital glands. Margin of pygidium (fig. 21,
d) with three pairs of well developed lobes ; median pair largest and notched on
both sides ; second pair notched on one or both sides ; third pair much the smallest.
Squamae between the lobes branched ; there is also a branched squama just beyond
the third lobe and near it two or three simple ones. Anal orifice large, and placed

200 ROBERT NEWSTEAD.

well forward. Vaginal orifice just below the centre. Parasitised females are some-
times much shorter and broader than what are apparently typical examples ; they
may also be markedly contorted. Length, 1-1‘5 mm.

Female, second stage. The exuviae of this stage (fig. 21, e) are very strongly
arched transversely. Owing to this great convexity it is difficult to determine the
exact structural details of the pygidium ; the margin of the latter appears however
to possess appendages similar to those in the adult; but the three pairs of lobes
and the squamae are all relatively much smaller. Length 0-9-1:1 mm.

Ucanpa: Jana Isl., Sesse Islands, Lake Victoria, on Coffea robusta, 9.x.1918,
n association with Aspidiotus (Selenaspidus) articulatus var. and Lecanvum
subacutum. Newst. (C. C. Gowdey).

The adult female of this very singular species, in its general facies, bears such
a strong resemblance to the second stage females of certain members of the genus
Frorima thatI had provisionally placed it in this genus ; subsequently embryo larvae
were detected within the body of the parent and this at once removed all doubts
as to its true generic position. It might also be mistaken for a Cryptaspidiotus,
but clearly it has no close affinity with this genus. I have found it exceedingly
difficult to orientate the females so that the pygidium lies in its normal position ; in
the majority of cases it either breaks away or lies turned backwards upon the
chitinous integument of the abdomen. In the latter case it is exceedingly difficult
to detect, and the structural details are rendered almost obscure by the dark chitin
of the body wall. Whether the curious, laterally compressed puparium is due
altogether to the exigencies of its position on the edge of the leaf remains to be seen.
No examples were found on the flat surface ; so that it would seem that the creature
had, for some unaccountable reason, acquired the remarkable habit of sitting astride
the edge in a way which seems to be peculiarly its own, at any rate among the members
of the DraspPrNnaE.

Chionaspis madiunensis, Zehntner.*

Female Puparium. White, with a fine smooth texture (5 examples), or dusky
white and slightly roughened owing to the admixture of the epidermis of the food-
plant (1 example). Form moderately convex, circular, broadly ovate, or broadly
pyritorm. Larval exuviae projecting beyond the margin; translucent white or pale
straw-colour. Second exuviae either terminating at the margim or projecting very
slightly beyond it; colour pale straw-yellow or brownish-yellow ; secretionary covering
white. Ventral pellicle very thin, white, and attached to the plant. Diameter of
Jarge circular form, 3 mm ; length of other forms 2°6 ; width, 2 mm.

Female, adult (fig. 22, a). Shghtly elongate, the length of the body bemg from two
to two and a half times the width of the first free abdommal segment. Mesothorax
markedly dilated at the margin proximally; eye placed on a distinct tubercular
extension of the margin. Dorsal surface with a number of relatively large chitmised
patches or tuberosities. Antennae (fig. 22, b) placed closely together immediately
above the rostrum; each consisting of a rather long terminal tubercle and a stout
outstanding bristle. Mouth-parts placed well forward. All the stigmata (fig. 22, c)

*Meded. Proef. Suiker. Java. (3), No. 6, p. 1 (1898).

OBSERVATIONS ON SCALE-INSECTS. 201

with a large group of parastigmatic glands; the first pair placed close up to the
mentum ; the second pair on the metathoracic segment. The first abdominal segment
of the same width as the metathorax; the second much more strongly produced. and
considerably wider than the two others ; third merging into the pygidium. Vaginal
orifice a little below the centre of the group of circumgenital glands ; anal orifice
opposite. Dorsal pores arranged in three well defined but interrupted series. Margm
of pygidium (fig. 22, d) with four pairs of lobes, all of them very similar in
form, but the median pair larger than the rest; second to third pair, inclusive,
duplex; the division between each duplex lobe bemg complete to the base.
Squamae simple : there is a short one between each of the lobes, several on the margin

Fig. 22. Chionaspis madiunensis, Zehnt., 9; a, adult ;
b, antenna; c, stigma; d, margin of pygidium ;
e, parasitised 9.

beyond and also on the two succeeding abdominal segments. The spines are few and
not very conspicuous; there are two minute ones between the median lobes, and a
slightly longer one at the base of each lobe on the outer margin; a very long one on
the dorsal surface in the middle of the first duplex lobe and a smaller one just beyond
the fourth lobe. Three conspicuous marginal pores; one between the third and
fourth duplex lobes and two beyond the fourth ; similar but less conspicuous pairs
can be traced on the dorsal surface over some of the lobes, especially the fourth pair.
In parasitised examples (fig. 22, e) the mesothoracic region is not produced laterally,
and the free abdominal segments are more pomted.

Ucanpa: Kampala, on sugar-cane, 11.11.1918. (C. C. Gowdey).

202 ROBERT NEWSTEAD.

The deeply divided duplex lobes give the margin of the pygidium a stnkingly
characteristic appearance. In its general facies the female resembles that of
Chionaspis herbae, Green ;* but the latter has only three pairs of lobes, whereas
C. madiunensis has four. There are also other slight differences, notably the large
number of chitmised patches on the thorax.

I tender my sincere thanks to Mr. E. Ernest Green for the determination of this
species, and also for comparing it with examples of his C. herbae. The latter I had
not seen.

Chionaspis tenuidisculus, sp. nov.

Female Puparium. Form somewhat oblong, sides almost parallel; highly convex,
the convexity commencing abruptly near the middle region of the second exuviae.
Larval exuviae nude, orange-yellow to golden yellow. Second exuviae bright orange-
yellow; secretionary covering thin, semitranslucent, white. Secretionary portion
pure white, slightly polished and very strongly laminate. Length, 1°4-1°7.

Fig. 23. Chionaspis tenuwidisculus, Newst., sp. n., 93
a, pygidium ; b, margin of do.

Female, adult.- Dead examples dull yellow to red-brown ; cuticle rather strongly
chitmised ; general form fairly well preserved ; segmentation distinct ; margin folded
inwards towards the venter, traces of these folds also visible in the specimens after
maceration m KOH. Form elongate, much narrowed and produced anteriorly.
Rudimentary antennae rather widely separated; tubercle with one blunt spiny
process, and a long and unusually stout spme. Anterior stigmata in a deep horn-
shaped cavity ; with from one to three gland-pores. Pygidium (fig. 23, a) broadly
rounded ; median area with several large clear and more or less oval spaces or vacuoles,
somewhat irregularly disposed. Anal orifice near the articulation of the abdominal
segment. Gland-pores large and rather heavily chitinised. Margin (fig. 23, 6) with
two pairs of lobes ; median pair relatively broad and serrated ; second pair duplex,
the upper lobule very small and generally notched on both sides. Squamae simple
and somewhat slender. Spines minute. Integument rather strongly but finely
reticulated.

* Coccidae of Ceylon, p. 132.

OBSERVATIONS ON SCALE-INSECTS. 203

Ucanpa: Bukassa Isl., Sesse Islands, Lake Victoria, “ on creepers with large fleshy
leaf m forest,” 10.x.1918 (C. C. Gowdey).

The curious cuticular characters should serve at once to distinguish this species
from its allies. The large vacuoles of the pygidium remind one of similar structures
found in certain numbers of the genus Lecanium; but there are no gland-pores
associatéd with the structures in Choonaspis tenurdisculus.

Chionaspis praelonga, sp. nov.

Female Puparium (fig. 24, a). Very elongate ; the length may be nine to ten times
as great a; the width ; more or less straight or markedly contorted ; sides parallel.
Colour pure white ; texture smooth and slightly polished. Hxuviae pale yellowish-
brown, ventral pellicle complete. Length, 3-3-7 mm.

Male Puparium. Pure white and very faintly tricarmate.

PON Me Ne
MM. shy He “| :

Db, 2s “e, antenna ;. d, pycidiumis¢; f, portions of margin of
pygidium of second stage 9.

Big. 24. Chionaspis praelonga, Newst., sp. 0., 2; a, puparium ;

Female, adult (fig. 24, 6). Very elongate ; length about four times as great as the
width ; cephalic margin slightly narrower than the pygidium. Rudimentary antennae
(fig. 24, c) with two long slender spines, one of which is sometimes strongly curved.
Parastigmatic glands 4-5, present at the front pair of stigmata only ; 2nd pair of
stigmata considerably below the middle distance. Pygidium (fig. 24, d) broadly
rounded. Median lobes very slightly recessed, edges serrated ; second pair of lobes
duplex, the mner lobule much the largest, evenly rounded distally, and much more
prominent than the median pair, Squamae slender and spine-like, five on either side
of the median lobes. Circumgenital glands im five groups ; formula of two examples :

5 =
2 8 10 8
29 30 29 23

204 ROBERT NEWSTEAD.

Immediately above the anal orifice, at the junction of the pygidium with the free
abdominal segment, is a transverse linear group of 14-15 circular gland-pores of
a similar form to those surrounding the genital orifice. The large glandular pores
of the dorsum extending as far as the region of the mesothoracic stigmata.

Female, second stage. A little more than one-third the length of the adult female ;
narrowly elongate, sides parallel; cephalic margm shghtly narrower than the
pygidium. The last-named (fig. 24, e, /) with the marginal appendages similar
to those in the adult ; but the serrations of the median lobes more strongly defined,
and the second lobe relatively narrower and longer.

Ucanpa: Bufumira Isl., Sesse Islands, Lake Victoria, on an unknown tree,
12.x.1918 (C. C. Gowdey).

‘Chionaspis auratilis, sp. nov.

Female Puparwm. Moderately elongate; sides generally parallel behind the
exuviae ; strongly convex, the convexity commencing abruptly in the middle region
-of the second exuviae. Sides of secretionary portion narrowly vertical. Colour pure
glistening white, and very finely laminated. Larval exuviae rich dark orange-yellow.
Second exuviae dark orange-brown, with the distal portion (pygidium) much paler ;
-secretionary covering exceedingly thin, translucent, and present on the anterior
half only. Ventral surface open. Length, 1:7-1°8; width, 2-4 mm.

Fig. 25. Chionaspis auratilis, Newst., sp. n., 2; a, 9;
b, antennae; ¢, anterior stigmata; d, pygidium; e, margin
of pygidium.

Male Puparvum. Strongly tricarinate. Larval exuviae rich dark orange-yellow,
secretionary portion of a dull gold colour.

Female, adult (fig. 25, a). Ovoviviparous. Elongate, gradually narrowed
anteriorly ; length a little more than three times the greatest width. Rudimentary
antennae (fig. 25, b) placed very closely together immediately in advance of the
rostrum, and furnished with remarkably long spiny processes two of which are

OBSERVATIONS ON SCALE-INSECTS. 205

strongly chitinised and* longer than the others; and a pair of long slender spines.
Anterior stigmata (fig. 25, c) with three gland-pores, and surrounded by an exception-
ally large peritreme, which latter takes the form of a deep irregular pit, and is
surrounded by a well-defined chitinous thickening of the body wall. Pygidium
(fig. 25, d) broadly rounded. No circumgenital glands. Dorsal gland-pores relatively
few in number. Fringe of pygidium (fig. 25, e) with two pairs of lobes: median
pair much the smaller and strongly tricuspid ; second pair somewhat spathuliform,
and rather narrowly rounded distally. Margin beyond with several minute pointed
tubercles. Squamae simple, three on either side. Spines extremely minute.

Ucanpa : Jana Isl., Sesse Islands, Lake Victoria, on an unknown plant, 9.x.1918
(C. C. Gowdey).

This species possesses two strikingly characteristic features: the bright golden
colour of the male puparia, and the exceptionally large peritreme or depression to
the anterior stigmata. The remarkable form of the antennae of this species bears
a close resemblance to that of Chronaspis coronrfera, Green, but these two species
are otherwise easily separable.

Chionaspis dura, sp. nov.

Female Puparvum. Pure white ; exuviae nude, and of a dull orange colour. Form
elongate ; convexity not greater behind the second exuviae than elsewhere ;
distal portion more or less slightly flattened. Ventral pellicle present at the sides
only. Length, 1°7-2°3 mm.

X41 ats

; 7 b ty @ ‘ mr \ my )

aah all

Fig. 26. Chionaspis dura, Newst., sp. Meg Osh Ay Os, ./D5i0,
antennae; c, pygidium.

Female, adult (fig. 26, a). Colour of dead examples dusky yellow to dull orange.
Form very elongate ; about two and two-thirds as long as the greatest width;
cephalo-thorax, in a line with the proximal portion of the rostrum, about one-half

206 ROBERT NEWSTEAD.

the width of the first free abdominal segment. A little more than two-thirds of the
upper portion of the body more highly chitinised than the rest, this character
disappearing after long maceration in KOH. Rudimentary antennae (fig. 26, 0,b)
relatively robust, with two or three blunt spiny process2s and a pair of long stout
hairs ; they are placed midway between the margin and the rostrum. Parastigmatic
glands (3-4) present at the anterior spiracles only. Pygidium (fig. 26, c) broadly
rounded ; anal orifice placed well forward and just in advance of the vaginal orifice ;
margin with a very large pair of slightly divergent median lobes, with the outer
edges finely crenulated and longitudinally striate; squamae simple and few in
number. Spines minute. No circumgenital glands.

UcanpbA: Bufumira Isl., Sesse Islands, Lake Victoria, on an unknown tree
12.x.1918 (C. C. Gowdey).

The large divergent lobes, the peculiar form of the antennae and the semi-
chitinised upper portion of the body are the distinguishing features of this insect.

Chionaspis laniger, sp. nov.

Female puparvum (Plate XVI, fig. 2). Somewhat mytiliform, but very highly
convex ; composed, externally, of white felted woolly material, which when perfect
is very strongly and coarsely laminated transversely ; beneath the woolly exterior
the puparium is hard and shell-like in texture. Ventral scale more or less complete.
Exuviae completely covered dorsally with dense loose woolly filaments. Larval
exuviae dull yellow; second exuviae dull yellow to pale straw colour. Length,
taken from the ventral surface. 2-2°3 mm.

Female, adult. Somewhat ovate or more or less deltoid ; cephalo-thorax narrowed
anteriorly. Rudimentary antennae (fig. 27, a,a) with a unilateral spiny projection
and two long stout spines, one or both of which may be curved. Parastigmatic
glands to both pairs of stigmata: the anterior pair with about ten; the posterior
pair with four or five. Dorsal gland-pores on all the segments below the lower
pair of stigmata. Pygidium (fig. 27, 6) gradually rounded. Anal orifice opposite
the anterior group of circumgenital glands. Vaginal orifice opposite the space
between the two lateral groups of circumgenital glands. The last-named in five
groups ; the anterior group consisting of about 15; the anterior-laterals of about
30; the lower-laterals of about 60. Margin of pygidium with six or seven dentate
lobes on either side ; the third and fourth and the fifth and sixth are adjacent and
separable only by a narrow line. Squamae simple and very short. Spines small
and slender. Length, 15-19 mm.

Female, second stage. Margin of the pygidium with similar dentate lobes to those
of the adult, but generally narrower and with the dentations more acute.

Larva. Antennae (fig. 27, c) of five segments. Just below the cephalic margin,
on the dorsal surface, is a pair of relatively large glandular pores (fig. 27, c) with
very strongly curved cylindrical ducts ; the integument surrounding these faintly
convoluted. These structures have been observed in the exuviae taken from the
puparia.

Ucanpa: Kampala, on Loranthus entebbiensis, 15.x1.1918 (C. C. Gowdey).

OBSERVATIONS ON SCALE-INSECTS. 207

This species is remarkable for the unique character of the puparia, which, together
with the structural details of the pygidium, should serve as a ready means of
determining it.

All the puparia submitted for examination are arranged transversely upon the
slender branches of the food-plant ; and in one instance there is a double row of
them with the cephalic portions (exuviae) of both rows, respectively, meeting
together in tne middle line.

Fig. 27. Chionaspis laniger, Newst., sp. n.; a, a, antennae of 9; b,
pygidium of 9; c¢, anterior portion of larval exuviae showing antennae
and paired tubular glands.

We tender our congratulations to Mr. Gowdey on his discovery of this extremely
interesting addition to the Coccia fauna of Africa.

(C605) I

ie
1, i hgh

EER, TS Leh K 4
tte ale Seal Soiae
ee ae ne “lia

: 5 7 if t\%
- a cage \ ‘ies i ean iF Ht 4
. is en fe we
*
ye ' ia : ; £
ee Ae stort ie Bat vacen bene Mies Oe
f p
if
ir
Qo = M
q
a, 4 j r
a /
' \
j . J +
‘ ie |
ral iy}
y \ 1 id Ah
/)
{ ‘ }
~ i y
, L- ]
‘) 3
. 2
A P
ly
MG nf
oh ups x
4 4 b .-
wv yy
( P os
.” -s
7
‘ \
Sb a>)
® j
J
. , wt t di
‘i 4 iv
a j
2 y Fi J
rin E } } if
] %
j s t
r ee i
Ph “| - :
f a) es
;
yal
i "2
. a ie '
i m i
} ? 4 }
Ny / = i. 2 8 :
y \ }
‘

Ne Lat & tiled 2h wi” agpegebchenge Baty ‘ ch oid eee wut awit neath .
ay ie were a a ee eT @ a tn hs Raed ay We . ie
eh | ay irom Lalasd aiid al es

+ é ; F ‘ nine 4

et od He Wl -

el SS Bea Pe: Eae V a Y nat me we nod

wy ei
ay fy M

IBULi. JEN, JIRESEWRGEH., Wok, 2X. IJPARr 2. Pini SOWIE

Fig. 1. Clypeococeus nena gel, Ckll., Fig. 2. Chionaspis laniger, Newst.,
adult females, x 234. puparia of females, x 3.

Fig. 3. Walkeriana digitifrons, Newst., Fig.4. Phenacoceus ballardi, Newst.,.
adult females, x 2. young and adult females, x 3.

—

: ~ ts.

209

COLLECTIONS RECEIVED.

The following collections were received by the Imperial Bureau of Entomology
between 1st April and 30th June, 1919, and the thanks of the Managing Committee
are tendered to the contributors for their kind assistance :—

Dr. W. M. Aders, Government Economic Biologist :—13 Culicidae and 38 other
Diptera ; from Zanzibar.

Mr. T. J. Anderson, Government Entomologist :—15 Culicidae, 10 Haematopota,
1 Tabanus, 490 other Diptera, 596 Coleoptera, 10 Planipennia, 4 Isoptera, 189
Rhynchota, 37 Orthoptera, 45 Odonata ; from British Hast Africa.

Mr. B. N. Bandyopadhyay :—23 Culicidae, 5 other Diptera, and 4 Lepidoptera ;
from British East Africa.

Lieut. P. J. Barraud :—2 Chrysops, 4 Tabanus, 1 Asilid, 6 Hymenoptera, and
19 Coleoptera; from Macedonia: 842 Culicidae, 36 slides of Culicid larvae, 3 Chrysops,
1 Stomoxys, 1 Hippoboscid, 116 other Diptera, 24 Hymenoptera, 124 Coleoptera,
5 Planipennia, 50 Lepidoptera, 5 Trichoptera, 6 Isoptera, 50 Anoplura, 9 Rhyn-
chota, 24 Orthoptera, 13 Odonata, 1 Spider, 3 Centipedes, 3 Wood-lice, and
13 Karthworms ; from Mesopotamia.

Mr. G. E. Bodkin, Government Economic Biologist :—8 Diptera, a number of
Ants, 13 Chalcids bred from Coccidae, 12 other Hymenoptera, 27 Coleoptera,
4 species of Coccidae, 22 other Rhynchota, 4 Orthoptera, 4 Ticks, 4 Spiders,
2 Centipedes, 6 Millipedes ; from British Guiana.

Mr. J. H. Burkill, Director of the Botanic Gardens :—4 Agromyzid Diptera and
16 larvae and pupae ; from Singapore.

Mr. J. E. A. den Doop :—20 Weevils, with 17 larvae and 21 pupae; from Java.

Durban Museum :—90 Coleoptera ; from Natal.

The Government Entomologist, Madras :—69 Coleoptera and 6 Rhynchota ;
from South India.

Mr. C. C. Gowdey, Government Entomologist :—24 Ticks ; from Uganda.

Mr. EK. Hargreaves :—146 Culicidae, 1 Pangona, 3 Chrysops, 1 Haematopota,
11 Tabanus, 25 other Diptera, 39 Hymenoptera, 110 Coleoptera, and 3 Orthoptera ;
from Taranto, Italy.

The Imperial Entomologist, Pusa :—3 Stephanidae ; from India.

Imperial Institute :—9 silk cocoons damaged by Anthrenus ; from Morocco.

Mr. 8. Leefmans, Instituut voor Plantenziekten, Buitenzorg :—35 Coleoptera
and 58 Rhynchota; from Java.

Mr. F. Muir :—5 Weevils and 2 larvae, and 29 Delphacidae.

Prof. G. H. F. Nuttall, F.R.S. :—59 Coleoptera, from Burma: and 44 Culicidae,
7 Ceratopogoninae, and 38 other Diptera ; from British East Africa.

Mr. W. H. Patterson :—10 Coleoptera and 2 larvae; from the Gold Coast.

Mr. R. Swainson-Hall :—2 Coleoptera ; from Portuguese Congo.

Mr. R. Veitch, Entomologist, Colonial Sugar Refinmg Company :—9 Diptera,
25. Hymenoptera, 9 Coleoptera, 1 Lepidopteron, and 14 Rhynchota; from Fiji.

210 COLLECTIONS RECEIVED.

Dr. W. G. Watt, Medical Officer of Health :—195 Culicidae and 107 larvae;
from the Gold Coast.

Mr. C. B. Williams :—12 Diptera, 616 Ants, 3 Moths, 306 Rhynchota, 3 Thrips,
1 Centipede, and 1 Spider; from Trinidad.

Mr. Rodney C. Wood :—9 Tabanus, 45 other Diptera, 101 Coleoptera, and

3 Rhynchota; from Nyasaland.

IMPERIAL BUREAU or ENTOMO Loc. a a

; HON ORARY. COMMITTEE OF MAN AGEMENT

a THE VISCOUNT HARCOURT, ‘Chairman, |

Limvr. -CotoneL A. ALCOCK, C.LE., F.R. g.

Masor E. E. AUSTEN, D.S.0.

Dre. A. G. BAGSHAWE, CMG.

Masor-GEnzRAL Sm JOHN R. BRADFORD, K.0.M. G3 F. RS. “ee th
“ Masor-Genzrat Sr DAVID BRUCE, pe BRB? 65 ae ene

Me. J. CO. F. FRYER.

Da S. F. HARMER, F.RS._ | eh ea
Pror, H, MAXWELL LEFROY. — | 8 ie
Hon. E. LUCAS.
’ Da R. STEWART iieptenii

Sm JOHN MoFADYEAN.

_Sm PATRICK MANSON, G.C.M.G., F-.R.S.

Si DANIEL MORRIS. K.C.M.G. |
Prov. R. NEWSTEAD, F.RS. OR. 8 Coa
Paor, G.H. F. NUTTALL, WR O ree
Pror. E. B. POULTON, F.R.S. SER
Lrzot. -CotonEL Str DAVID PRAIN, C. MG, C.I. B. F.RS.

Sm H. J. READ, K.O.M.G:,C.B. i Aa aeaas
ay Hon. N.C. ROTHSCHILD, = Hs PEN ce
Dr. HUGH SCOTT. Ae See Oa Og Hk a en
nt AOE, SHIPLEY, RR Se oF me 5 are ee,
Me. Ry A. C.-BPERBING “OMEGs gibt tc hoe" Bs Aas
Sfe STEWART STOCKMAN, 2:2, 0008 (2 Vas eae
‘Ma Bi Vi THROBALD Se oS gl ee
es C. WARBURTON. ve am mea mR thane tae,

| | ‘Director. Pg pind hy ‘ gee ee)
Dr. GUY AK. MARSHALL Ae AE EN
| _ Besistant Director. NC Ree ane
kas peer CS ae

ies | Secretary. i. a .
. Dat, A. ee ©. FARKINSON.

211

NOTES ON THE ETHIOPIAN FRUIT-FLIES OF THE FAMILY
TRYPANEIDAE, OTHER THAN DACUS.—III.*

By Prof. M. Brzz1,
Turin, Ltaly.

(PLates XVII-XVIIT.)

The new material received from the Imperial Bureau of Entomology through
the kindness of the Director, Dr..G. A. K. Marshall, enables me to prepare this
third paper on the Ethiopian fruit-flies. It is a proof of the great richness of the
African Trypaneid fauna that of over 41 species received only 10 were referable
to previously described species, and that no less than 8 new genera have been
erected, most of them for very strikingly characterised forms.

* * * x x a

The TRYPANEIDAE at present known from the Ethiopian Region, with the addition
of the new forms here described for the first time, are enumerated in the following
systematic catalogue, which numbers 216 species.

I. Subfamily DACINAE.

1. TrRipacus, Bezzi.
lounsburyi, Coq.
sphaeristicus, Speis.
fuscovittatus, Grah.
armatus, F.
fuscatus, Wied.
nebulosus, Walk.
d’emmerezi, Bezzi.
bivittatus, Big. (pectoralis, Walk. ;
bipartitus, Grah.; cucumarius, Sack ;
armatus, Bezzi).
momordicae, Bezzi (bipartitus, Bezzi).
eburneus, Bezzi.
xanthopterus, Bezzi.
humeralis, Bezzi.
disjunctus, Bezzi.
flavicrus, Grah.
scaber, Loew.
stylifer, Bezzi.
punctatifrons, Karsch.

2.;Dacus, Ee.
hamatus, Bezzi.
trigonus, Bezzi.
inornatus, Bezzi (modestus, Bezzi).

oleae, Gmel.
mesomelas, Bezzi.
semisphaereus, Beck.
blepharogaster, Bezzi.
rufus, Bezzi.
annulatus, Beck.
erythraeus, Bezzi.
mochii, Bezzi.
woodi, Bezzi.
macer, Bezzi.
immaculatus, Coq.
brevis, Coq.
fieicola, Bezzi.
bistrigulatus, Bezzi.
brevistriga, Walk.
brevistylus, Bezzi.
vertebratus, Bezzi.
binotatus, Loew.
bistrigatus, Loew.
ciliatus, Loew.
sigmoides, Coq.
africanus, Adams.
sexmaculatus, Walk.
longistylus, Wied.
kingt, Frogg.).

(testaceus, Macq. ;

Il. Subfamly ADRAMINAE.

3. MERACANTHOMYIA, Hendel.
antennata, Hend.

* For Part I, see Bull. Ent. Res. vili, pp. 215-251; for Part Il, ibidem ix, pp. 13-46.
1,000. 3.20. B.&F.Ltd. G.11/4. A

(637) W+.P4/140.

4. SOSIOPSILA, Bezzi.
trisetosa, Bezzi.

Py PROF. M. BEZZi,

Ill. Subfamily CERATITINAE.

5. COELOPACIDIA, End.
strigata, Bezzi.
melanostigma, Bezzi.
madagascariensis, End.

6. CONRADTINA, End.
longicornis, End.
acrodiauges, Speis.
acroleuca, Wied.
suspensa, Bezzi.

7. CELIDODACUS, Hendel.

obnubilus, Karsch (apicalis, {Hend.).

coloniarum, Speis.

8. CARPOPHTHOROMYIA, Austen.
vittata, F.
sceutellata, Walk.
tritea, Walk.
pseudotritea, Bezzi.
superba, Bezzi.

9. LEUCOTAENIELLA, Bezzi.
guttipennis, Bezzi.
trispila, Bezzi.
pentaspila, Bezzi.

10. CHELYOPHORA, Rondani.
magniceps, Bezzi.

11. BISTRISPINARIA, Speiser.
fortis, Speis.

12. CLINOTAENIA, Bezzi.
grata, Wied.
anastrephina, Bezzi.

13. CERATITIS, McLeay.
capitata, Wied.
catoiri, Guér.

14. PTERANDRUS, Bezzi.
anonae, Grah. (pennipes, Bezzi).
colae, Silv.
rosa, Karsch.
fasciventris, Bezzi.
rubivorus, Coq.
volucris, Bezzi.

? penicillatus, Big.

- 15. PARDALASPIS, Bezzi.
morstatti, Bezzi.
punctata, Wied.

brémei, Guér.

melanaspis, Bezzi.

cosyra, Walk. (giffardi,) Bezzi).
silvestril, Bezzi.
senegalensis, Macq.
antistictica, Bezzi.
flexuosa, Walk.

quinaria, Bezzi.

stictica, Bezzi.

aliena, Bezzi.

16. PERILAMPSIS, Bezzi.
pulchella, Austen.
formosula, Austen.

* The genus Cladoderris, Bezzi, is better placed in the ORTALIDAE,,and will be dealt with

in a forthcoming paper on this family.

17. HopLoLopHa,"Bezzi..
cristata, Bezzi.

18. TRIRHITHRUM, Bezzi:.
lycu, Coq.
nitidum, v. Rod.
albonigrum, End.
validum, Bezzi.

nigrum, Grah. (obscurum, End.)..

gagatinum, Bezzi.
occipitale, Bezzi.
bimaculatum, v. Rod.
nigerrimum, Bezzi.
leucopsis, Bezzi.

inscriptum, Grah. (coffeae, Bezzi)..

? albomaculatum, v. Rod.

19. XANTHORRHACHISTA, Hendel...

alata, Beck. (cephalia Hend.).

20. THEMARICTERA, Hendel,
flaveolata, F.
rufipennis, Hend.
laticeps, Loew.

21. THEmMARA, Walker.
fallacivena, End.

22. BARYGLOSSA, Bezzi.*-
histrio, Bezzi.

23. PTILONIOLA, Hendel..
neavel, Bezzi.
preussi, Hend.
tripunctulata, Karsch.
(Nov. GENUS.)
guttatolimbata, End.

24. RHACOCHLAENA, Loew...
fasciolata, Loew.
pulchella, Bezzi.
hammersteini, End.

25. PHORELLIA, R. Desv:.
brunithorax, R. D.
tristriata, Karsch.

26. TaomytA, Bezzi..
marshalli, Bezzi.
ocellata, Lamb.

27. NotomMa, Bezzi,,

bioculatum, Bezzi.
(Nov. GENUS.)
jucunda, Loew.

28. AcrpiA, R. Desv.
seychellensis, Lamb.
fossataeformis, Bezzi.
homogenea, Bezzi.

29. ** OcnEROS,”’ O. G. Costa..

mundus, Loew.
excellens, Loew.
sinuatus, Loew.
undatus, Bezzi.
bigemmatus, Bezzi.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 213

III. Subfamily CERATITINAE.—cont.

30. OCNERIOXA, Speiser.
pennata, Speis.
woodi, Bezzi.
disereta, Bezzi.
gracilis, Loew.

31. TErrparRitis, Latr.
indecora, Loew. ©
vernoniicola, Bezzi.

32. CarpomytA, A. Costa.
incompleta, Beck.

33. CRASPEDOXANTHA, Bezzi.
marginalis, Wied.
manengubae, Speis.

34. TeRELLIA, R. Desvoidy.
? hysia, Walk.

35. ALLOTRYPES, Bezzi.
brevicornis, Bezzi.

36. AcrIuRA, R. Desv.
perspicillaris, Bezzi.
angusta, Loew.
semiangusta, Bezzi.
caeca, Bezzi.
oborinia, Walk.
tetrachaeta, Bezzi.
capensis, Rond.
ternaria, Loew.
binaria, Loew.

37. SPHENISCOMYIA, Bezzi.
sexmaculata, Macq.
neavel, Bezzi.

38. TEPHRELLA, Bezzi.
bezziana, End.
nigricosta, Bezzi.
cyclopica, Bezzi (w-fuscum, End.).
tephronota, Bezzi.
rufiventris, Bezzi.
sexfissata, Beck.
hessii, Wied.

IV. Subfamily TRYPANEINAE.

39. PLATENSINA, Enderlein.
diaphasis, Big.

(Nov. GENUS).
lunifera, Loew.

40. EuTRETOSOMA, Hendel.
oculatum, Hend.
frauenfeldi, Schin.
bipunctatum, Loew.
millepunctatum, Bezzi.
polygramma, Walk.

41. ELAPHROMYIA, Bigot.

adatha, Walk. (melas, Big. ; wlula, Loew).

42. SpaTHULINA, Rondani.
semiatra, Loew.
bioculata, Bezzi.
parceguttata, Beck.
acrosticta, Bezzi.
aldabrensis, Lamb.
margaritifera, Bezzi.

43. PLIOMELAENA, Bezzi.
brevifrons, Bezzi.
strictifrons, Bezzi.

44. EUARESTA, Loew.
planifrons, Loew.
amplifrons, Bezzi.

45. Ensina, R. Desv.
sororcula, Wied.
gladiatrix, Bezzi.
myilopitoides, Bezzi.
anceps, Loew.
magnipalpis, Bezzi.
dubia, Walk.

ignobilis, Loew.
siphonina, Bezzi.

46. Eurtsia, Meigen.
perpallida, Bezzi.
discipulchra, Bezzi.
practexta, Loew.
caffra, Loew.
dissoluta, Loew.
tristrigata, Bezzi.
cyana, Walk.

47. CAMPIGLOSSA, Rondani,
perspicillata, Bezzi.

48. CAMAROMYIA, Hendel.
acrophthalma, Bezzi.

? helva, Loew.

49, ACANTHIOPHILUS, Becker,
helianthi, Rossi.
? ochraceus, Loew.

50. TRYPANEA, Schrank.
subcompleta, Bezzi.
hemimelaena, Bezzi.
peregrina, Adams (wrophora, Bezzi).
augur, F'rauenf.
tristicula, Hend. *
auguralis, Bezzi.
hexapoda, Bezzi.
confluens, Wied.
aucta, Bezzi.
repleta, Bezzi.
aira, Walk.
amoena, Frauenf.
decora, Loew.
diversa, Wied.

* This recently described species was erroneously included by Prof. Hendel in his great
work on the South American Trypaneids, 1914, p. 82; but it is from Mozambique, and
was accidently omitted by me in my previous paper ; it is very near augur, has about the
same wing pattern, and has likewise only two scutellar bristles.

(637)

»2

214 ' PROF. M. BEZZI.

V. Subfamily SCHISTOPTERINAE.

51. PERIRHITHRUM, Bezzi.

53. RHOCHMOPTERUM, Speiser.
marshalli, Bezzi.

neuropteripenne, Speis.
52. RHABDOCHAETA, de Meijere. 54. SCHISTOPTERUM, Becker.

spinosa, Lamb. moebiusi, Beck.
neavel, Bezzi.

* * * * * *

The addition of the very interesting new genus which follows makes necessary a |
modification of the beginning of the table of genera in my first paper, p. 216, nos.
1-5, which may be modified as follows :—

1(6). Thoracic chaetotaxy incomplete, hm., prst., dc. and st. being always wanting
(and also the prsc. in African forms).

a(d). Arista bare; abdomen short and broad, usually broader than the thorax ;
Wings with the first three longitudinal veins closely approximated, the
anterior cross-vein long and oblique, the second basal cell dilated and the
anal cell drawn out into a very long point oe Subfam. Dactnar.

b(c). Only two sa. present, the anterior one being entirely wanting

Dacus, ¥. (8. sir-):

c(b). Three sa., the anterior one being well developed .. .. Tridacus, Bezzi.

d(a). Arista plumose; abdomen long and linear, more narrow than the thorax ;
wings with the first three veins not approximated, the anterior cross-vein
short and placed perpendicularly, the second basal cell not dilated and
the anal cell with a short point .. .. Subfam. ADRAMINAE.

e(f). Femora not spinose beneath; thoracic suture interrupted in the middle ;
antennae shorter than the face ; point of the anal cell very short

Sosiopsila, gen. nov.

f(e). Femora spinose beneath; thoracic suture complete; antennae much longer

than the face; point of the anal cell longer .. Meracanthomyra, Hend.

Il. Subfam. ADRAMINAE.
The present subfamily 1s here accepted in a wider sense than that attributed
to it by Hendel in 1914, and corresponding to that adopted by me in 1916.* The
subfamily was believed to be almost exclusively Oriental; but the discovery of

the following new form shows that it is probably well represented in the Ethiopian
Region also.

Sosiopsila, gen. nov.

The present new genus seems to be allied to the Oriental Neosophira, Hendel
(Gen. Insect., 76, pl. i, figs. 57-59 and 64) and to Colobostroter, Enderlein (Zool.
Jahrb., xxxi, p. 445, fig. T.), but is distinct in having well developed outer vt., and
also in having only a pair of sct. (the basal pair) ; on the wings the third longitudinal
vein is not sinuous.

Head in front view more broad than high, a little broader than the thorax ; occiput
convex and prominent, with developed lower swellings; the head is attached to
the thorax about in the samé manner as described for the genus Xanthorrhachista.

* Bull. Ent. Res., vii, October 1916, p. 120.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 215

Eyes oval, less compressed, their vertical diameter being only a little longer than
the horizontal one. Frons flat, not at all prominent in profile, with parallel sides,
only a littie longer than broad ; lunula narrow, but much extended laterally ; face
flat in the middle, with shallow antennal grooves and with distinctly prominent
mouth-border. Cheeks lnear; jowls not broader than the breadth of the third
antennal joint. Antennae inserted at the middle of the eyes, only a little shorter
than the face ; third joint pubescent, rounded at tip; arista plumose, its feathering
being about as broad as the length of the third antennal joint. Proboscis short
and thick, with very broad flaps, which cover the palpi. Cephalic chaetotaxy
rather developed, only the oc., put. and genal bristles being wanting ; no ocp.; outer
vt. half as long as the strong inner pair; one strong s. or., and two pairs of long and
distant 7. or. Thorax elongate, with incomplete suture and reduced chaetotaxy ;
there are only two npl., one a., and two p. sa., one mpl., one pt.; the scp. are
rudimentary, only the middle pair being distinct. Scutellum triangular, flat above,
a little broader than long, distinctly carinate on the sides, with only the 0. sct., which
are very long, strong and diverging. Abdomen narrow and long, of cylindrical
shape, narrow at base but not properly pedunculate, in the male with 4, in the
female with 5 segments; the last two segments bear some strong bristles at the
sides behind. Male genitalia small and rounded; ovipositor short, depressed,
with the basal segment as long and nearly as broad as the terminal segment of the
abdomen, and with the apical segment much narrower and acute. Legs long and
simple, with non-spinose femora, even the front pair being without bristles below ;
middle femora elongate, as long as their tibiae, and a little thickened ; middle tibiae
with a single long spur at end; hind tibiae shorter than their femora, distinctly
curved, with scattered bristles on the outer side; claws very short. Wings short
and proportionally narrow, hyaline, with a broad black spot at apex. Costa
densely ciliate, with no distinct bristle; auxiliary vein closely approximate to the
first longitudinal vein, which ends at middle of the fore border, before the small
cross-vein, and is pilose above. Second longitudinal vein straight; third bristly
throughout its whole length and slightly curved but not sinuous ; last segment of the
fourth vein slightly diverging at end; small cross-vein a little beyond the middle ;
hind cross-vei straight and perpendicular, 24 times as long as the small one;
anal cell a little broader than the second basal cell, its inferior angle acute but not
produced and shorter than the upper one; sixth vein reaching the hind border.
Axillary lobe equal in the two sexes ; alula longer than broad.
Type: the following new species :—

Sosiopsila trisetosa, sp. nov. (Pl. xvu, fig. 1).

A narrow and elongate, entirely yellow species not unlike a Psila, with a broad
rounded black spot on the mesophragma, and with a broad black spot near the apex
of the wing.

¢ @. Length of body, 65-7 mm. ; of wing, 4°5-5 mm. ; of ovipositor, 0°7 mm.

Head and its appendages completely yellow. Occiput shining, unspotted ; frons
more yellow, dull and finely pubescent in the middle, but with shining ocellar and
vertical plates, the ocellar not being reddish brown, like the upper border of the
lunula. Face shining in the centre; cheeks and jowls with a faint whitish dust,

216 PROF. M. BEZZI.

like the lower occipital swellings. All the cephalic bristles are yellowish, only the
et. and the s. or. being a little darker. Thorax entirely yellow; darker, less shining
and punctulate on the dorsum, lighter, more shining and smooth on the pleurae ;
the very short and scattered pubescence of the dorsum is yellowish, like all the
macrochaetae, only the p. sa. being darkened, chiefly the interior one; the pleurae
have only sparse and very thin pale yellowish hairs on hind border of meso- and
on upper border of sternopleura. The scutellum is coloured like the mesonotum
and shining ; it bears some pale yellowish hairs, while its macrochaetae are black.
The mesophragma is shining yellow, but it has in the middle a broad, rounded, shining
black spot, which is prolonged above in the middle of the postscutellum. Halteres
and squamulae pale yellowish, the latter with a pale fringe. Abdomen shining
yellow like the mesonotum and similarly punctate, with longer and denser pale
yellowish pubescence; its terminal macrochaetae are black; venter yellow.
Male genitalia shining yellow and yellowish pubescent; the ovipositor likewise.
Legs entirely yellow, with pale pubescence and pale hairs, only the spurs of the
middle tibiae being black ; the middle femora have below in the middle three rigid
long hairs, directed downwards, stronger and more bristly in the male than in the
female. Wings quite hyaline and strongly iridescent; the veins are yellow, with
blackened ends, but the whole costa (with its ciliation) is black; the costal cell
is yellowish, and the short and narrow stigma dark yellow. The black apical spot
is attached to the costa by its base, which extends from the end of the second to
the end of the third vein ; it is continued inwardly, crossing the third vein and ending
roundly on the fourth vein ; the end of the first posterior cell is therefore completely
hyaline, while the end of the submarginal cell is broadly black. In the female the
apical spot extends only to the middle of the first posterior cell, without reaching
the fourth vein ; but even in the male the part of the spot placed beyond the cubital
fold is lighter.

Type ¢ and type 9, a single pair of specimens from Portuguese East Africa,
KH. of Mt. Mlanje, 21-23, xi., 1913 (S. A. Neave); an additional female specimen
from Nyasaland, Mt. Mlanje, 8.x.1913 (S. A. Neave).

IiI. Subfam. CERATITINAE.
In the new material are present some very important species, which belong to the
two genera Coelopacidia and Celidodacus, previously unknown to me; they may be
comprised in my table of the genera on p. 216, just after no. 8 (67), as follows :—
A(B). Prst., dc. and st. wanting; face very concave in the middle; arista bare ;
femora not spinose beneath ; third ne vein bristly throughout its
whole length .. ; ; he Coelopacidia, End.

B(A). The above-named eeieles wer or at elite not all wanting at the same
time ; face flat or even convex in the middle.

9(10). Femora spinose beneath; arista shortly plumose; third longitudinal vem
with only a few bristles near its extreme base.

a(b). Wings very broad, with the second and third longitudinal veins sinuous, the
last portion of the third being moreover bent downwards at the end.

Conradiina, End.

b(a). Wings narrower, with the above-named veins not sinuous and with the last

portion of the third straight hg. Bi a .. Celdodacus, Hend.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 217

Coelopacidia, Enderlein, 1911.

This genus was placed by its author near Acidia, while in Prof. Hendel’s table
it comes near Platyparea. I have before me two species, which agree in the form
of the body and head, and in the reduced wing pattern, with the description of Dr.
Enderlein ; but the author has not described the chaetotaxy, which is very reduced
‘in the species before me.

The systematic position of the genus is dubious. In the want of prst., dc., prscl.
and st. it agrees with the DactnaE and ADRAMINAE, but differs from them in the
presence of the hm. Owing to this last character, the presence of distinct ocp. and
to the presence of bristles on the underside of the front femora, I prefer to locate
the genus at the beginning of the CERATITINAE, as a connecting link between this
subfamily and the ADRAMINAE.

I have not seen the genotype, C. madagascariensis, End., but I will give here a
characterisation of the genus drawn from the two continental species at hand.

Head in front view about as broad as high, not broader than the thorax ; occiput
convex and prominent, with developed but not sharply differentiated lower swellings ;
neck inserted just below the middle of the occiput, which bears a tubercle above it.
Frons distinctly concave in the middle, parallel-sided, about twice as long as broad ;
lunula very narrow and not extended at sides. In profile the eyes are much rounded,
their vertical diameter being only a little longer than the horizontal one; they are
higher above than the middle of the hollowed frons, while in front the frons is promin-
ent as a distinct conical protuberance, which bears the antennae. Face deeply
hollowed in the middle, without median keel and with undifferentiated antennal
grooves, but distinctly prominent below at the mouth-border; cheeks linear ;
jowls as broad as the breadth of the third antennal jot. Antennae inserted at
middle of the eyes ; in consequence of the protuberance of the fore part of the frons
and of the excavation of the face, they are rather porrect; third jomt rounded
at tip, bare, not reaching the mouth; arista long, thin, with very short scattered
pubescence. Proboscis short and thick; palpi less clavate and shortly setulose.
Cephalic chaetotaxy reduced ; ocp. reduced to some thin acute yellowish bristles ;
put. very short, thin, almost indistinct ; no oc. ; four vt., the inner pair not much
longer than the outer one; one s. or., erect and directed backwards, placed near
the middle of the frons, at the front end of the long narrow curved ill-differentiated
vertical plates ; one 7. or. just above the lunula, directed inwards and decussate ;
genal bristle thin and short. Thorax elongate, with the suture narrowly interrupted
in the middle; scp. long and strong, those of the middle pair approximate ; two
npl.; one a. sa. and two p. sa.; one mpl. ; no distinct pt. and no st. Scutellum flat,
triangular, as long as broad, with lateral keels and with two pairs of strong bristles,
the apical ones parallel and not much weaker. Abdomen elongate, narrower than
the thorax, of linear shape, being as broad at end as at base; there are 4 segments
in the male and 5 in the female, but the sutures are not very distinct ; last segment
with some macrochaetae at hind border in both sexes ; male genitalia rounded and
concealed ; ovipositor with the basal segment depressed, triangular, obtuse, as long
as the last two abdominal segments together ; apical segment narrow, acute. Legs
slender ; front femora with only one bristle below near the end, middle femora
distinctly thickened, shorter than their tibiae, which bear two equally strong spurs

218 PROF. M. BEZZI.

at the end; hind tibiae straight, shorter than the femora, with some short bristles
on outer side; claws short. Wings long, hyaline, with a very small apical black
spot. Costa not ciliated, shortly pilose, without bristle; auxiliary vein closely
approximated to the first vein, which is pilose and ends before or opposite to
the small cross-vein; second and third vein straight and diverging at their ends,
the third bristly throughout ; small cross-vein beyond the middle of the discoidal
cell; hind cross-vein straight and perpendicular; last segments of the third and
fourth veins parallel; second basal cell not dilated, but very long; sixth vein
produced to the hind border ; anal cell not broader than the second basal cell, its
lower angle drawn out into a point which is rather long, but is always shorter than
the second basal cell; axillary lobe equal in the two sexes; alula longer than broad.
The known species may be distinguished as follows :—
1(2). Front femora with some bristles below near the end; wings without black
apical spot, but with a dark border to the hind cross-vein
madagascariensis, End.
2(1). Front femora with only one bristle below; wings with a black apical spot,
but the hind cross-vein not or less infuscated.
3(4). Thorax on the dorsum with a whitish pollinose, longitudinal, middle stripe ;

wings with yellow stigma and with a slightly infuscated hind cross-vein
strigata, sp. nov.

4(3). Thorax not striped; wings with black stigma and the hind cross-vein not
infuscated .. - ve wi sie ae melanostigma, sp. nov.

Coelopacidia strigaia, sp. nov.

An elongate, entirely yellow species, distinguished by the whitish thoracic stripe
and the yellow pterostigma.

3 2. Length of body 9-95 mm. ; of ovipositor 1 mm. ; of wing 8-8°5 mm.

Head entirely yellow, unspotted, only the very small ocellar dot being black.
Occiput very shining and bare, except for a few short hairs near the neck and on the
lower part ; frons wholly glistening, quite bare, with the front part slightly reddish
above the lunula, the upper border of whichis narrowly brown. Face very glistening ;
cheeks and jowls not pruinose. Antennae entirely yellow, even on the apical part
of the last joint ; second joint short and nearly bare ; third joint about twice as long
as broad; arista yellowish. Palpi pale yellowish; proboscis dirty reddish, with
pale hairs. Of the cephalic bristles the two pairs of vt. and the s. or. are black, the
thin pvt. and the ocp. are yellowish, like the 2. or. and the genal bristle ; the sparse
and short hairs of the lower part of the occiput are pale yellowish. Thorax entirely
yellow and shining; but the dorsum, being punctulate, is less shining than the
very glistening pleurae, which are moreover of a paler colour ; the whitish pollinose
middle stripe is broad, but only distinct when the thorax is viewed from in front ;
this stripe begins just beyond the the middle scp. and ends before reaching the
scutellum broadening gradually behind. The very short and sparse pubescence of
the dorsum is pale yellow, while the pleurae are quite bare, with only a few,
hardly visible, thin and whitish hairs on the mesopleura. Scutellum entirely
glistening yellow, bare, with some scattered pale hairs at the hind border.
All the macrochaetae of the thorax and scutellum are black. Postscutellum and

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 219

mesophragma shining yellow, the latter more or less reddish in middle or on upper
border. Squamulae very small, with sparse and short pale hairs at border ; halteres
pale yellowish, with more reddish knob. Abdomen entirely shining yellow; but
there is a narrow dark longitudinal stripe along the middle of the two basal
segments, rather faintly marked; the pubescence is more dense than that
of the thorax, but is always short and yellowish; the apical macrochaetae are
black. The venter is shining yellow in the middle, whitish dusted on the sides.
Male genitalia shining yellow, with pale pubescence; ovipositor shining yellow,
with very short and sparse pubescence. Legs entirely shining yellow, with more
reddish tarsi; their sparse pubescence is pale, but the bristles of the femora and
tibiae are black, like the apical spurs of the middle tibiae ; the middle tibiae have
a longitudinal row of three rather strong black bristles towards the middle of the
posterior side ; the hind tarsi are ciliated below, chiefly on the praetarsi ; the claws
are black, the pulvili whitish. Wings yellowish-hyaline, very shining and
iridescent ; costa and veins yellowish, but the thickened basal part of the first longi-
tudinal and the common stem of the second and third, are black ; the cross-veins
and the ends of the longitudinal veins are blackish ; the stigma is entirely yellow.
The apical dark spot is small, extending symmetrically on both sides of the end.
of the third longitudinal vein, along the wing border; it is broader in the female
than in the male. The infuscation of the hind cross-vein is very small and faintly
indicated, sometimes even indistinct. The first longitudinal vein ends in the costa
a little, but distinctly, before the small cross-vein.

Type ¢ and type 9, a couple of specimens from Nyasaland, Limbe, Chiromo,.
Ruo R., 22.1x.1916 (R. C. Wood). ee

Coelopacidia melanostigma, sp. nov.

Closely allied to the preceding species, but differing in some particulars of the
coloration and in the longer third antennal joint.

6. Length of body 85 mm. ; of wing 75mm. Head and its appendages exactly
as in C. strigata, but the put. more developed and blackish ; the third antennal joint
is distinctly longer and reaches the mouth-border, being about 2°5 times as long
as broad. Thorax, scutellum and halteres as in the preceding species, but there is
no trace of the whitish middle stripe; there is moreover a small but distinct pi.,
which is not to be seen in C. strigata. Abdomen the same, with the basal stripe
blacker and more sharply defined. The wings likewise, but the stigma is entirely
of a deep black colour ; the first longitudinal vein ends opposite to the small cross-
vein, not before it; the apical dark spot is similarly shaped, but there is no trace
of the infuscation of the hind cross-vein. The legs are identical.

Type J, a single specimen caught in the same locality and on the same day as the
specimens of C’. stregata, from which it is certainly distinct (R. C. Wood).

Conradtina acrodiauges, Speiser, 1913 (Pl. xvii, fig. 2.)
A couple of specimens of this characteristic fly from S.{Nigeria, Ibadan,
29.x!.1913—20.1.1914 (Dr. W. A. Lamborn).

220 PROF. M. BEZZI.

Celidodacus, Hendel, 1914.

Prof. Hendel has differentiated the present genus from Conradtina only on account
ot the wing pattern; the chaetotaxy of the head and thorax seems to be the same
in both genera ; but I think that the form of the wings and the shape of the second
and third longitudinal veins can be regarded as sufficient for generic separation.

I have before me what I believe to be the genotype, named apicalis, but not
described, by Prof. Hendel ; I think that it is the same as Acidia obnubila of Karsch,
even though the latter author does not mention the spines on the underside of the
femora. A close ally of this species seems to be Acidia coloniarum, Speiser, which
has a different wing pattern, having an isolated dark band across thesmall cross-vein.

I will give here a characterisation of the genus Celzdodacus, with some additional
notes on the type-species.

Head in front view distinctly broader than high. Occiput rather flat above and
less prominent below, the lower swellings being not developed. Eyes rounded,
with the vertical diameter longer than the horizontal one. Frons flat, parallel-
sided, 1°5 times longer than broad; in profile it is only a little prominent above
the root of the antennae; lunula very small. Face convex in the middle, with
broad antennal grooves ; mouth-border prominent ; cheeks linear; jowls narrow.
Antennae inserted a little above the middle of the eyes, with the third jomt acute
but rounded at tip, long, extending a little beyond the mouth-border; arista
plumose, the breadth of the feathering being equal to the breadth of the third antennal
joint towards the middle. Proboscis short and thick; palpi broad and bristly.
Cephalic bristles strongly developed ; ocp. acute and black; two vé., the mner pair
longer ; pvt. well developed ; no oc. ; two s. or. and three 7. or.; genal bristle well
developed ; there is also a bristle towards the middle of the lower part of the
occiput. Thorax broad and stout, with complete and strongly developed chaeto-
taxy; scp. long, those of the middle pair not approximated ; the dc. pair is placed
much behind; two mpl.; pt. strong. Scutellum triangular, as long as broad,
flat above, with lateral keels and four long strong bristles. Abdomen with 4
segments in the male and 5 in the female, narrowed at base, about as broad as the
thorax in the middle, with terminal bristles ; male genitalia small; ovipositor short
and broad, depressed. Legs short and stout, all the femora spinose beneath, but
those of the hind pair less so than the others; middle femora thickened ; middle
tibiae with a single spur at the end; hind tibiae with less developed posterior row
of bristles. Wings proportionately narrow, with well developed pattern. Costa
not ciliated and without bristle; auxiliary vein stout and well separated ; first
longitudinal vein ending before the middle of the costa and before the small cross-
vein ; second vein a little sinuous; third bristly throughout, straight on its last
portion and diverging from the fourth; small cross-vein much beyond the middle,
placed on the last third of the discoidal cell; hind cross-vein straight ; sixth vein
reaching the hind border; point of the anal cell acute, as long as the second basal
cell; axillary lobe broad; alula rounded.

Celidodacus obnubilus, Karsch, 1887 ; (apicalis, Hend., 1914). (P!. xvu, fig. 4).
A robust species, at once distinguished an account of the peculiar coloration
of the thorax and wings.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 221

6 9. Length of body 75-8 mm. ; of ovipositor 1 mm. ; of wing 65-7 mm.

Head entirely reddish yellow, only the small ocellar dot being black ; frons opaque
and with scattered hairs in the middle, shining at vertex and on vertical plates ;
antennae entirely yellowish, with the third joint three times as long as broad ; palpi
_and proboscis yellowish. All the cephalic bristles are black. The rather opaque
thorax has in the middle of the dorsum two broad longitudinal stripes of whitish
dust, which are visible only in certain lights; on their outer border there is on
each side a small black stripe, which is sometimes wanting ; the middle brown stripe,
_as described by Karsch, is formed only by the space between the two whitish
stripes. Very characteristic is the white, black-margined stripe on the notopleural
line. All the bristles are black, and the short pubescence of the dorsum is likewise
black. Scutellum like the mesonotum, with short and sparse pubescence on the
sides. Pleurae shining, with soft thin pubescence. Squamulae and_halteres
whitish. Mesophragma and postscutellum shining black, but sometimes reddish
or yellowish. Abdomen above either entirely shining black, or with the first two
segments more or less broadly yellowish ; the pubescence is blackish, the bristles
black. Male genitalia shining black, ovipositor shining black, sometimes reddish
brown towards the middle. Legs entirely yellow, with black bristles, spines and
spurs ; front femora with two rows of 5-6 spines each near the end, middle femora
with 7-8, hind femora with 4-5 much smaller ones. Wing pattern,as figured and
_ described by Karsch, the apical hyaline spot being whitish, as stated by Hendel.
The stigma is entirely black. Sometimes there is a small subhyaline dot in the
- dark base of the first posterior cell, near the fourth vein. The veins are yellowish
_ on the hyaline parts, and blackish on the dark parts.

Originally described from Pungo Andongo, and recorded subsequently from
S. Nigeria, the species seems to be widely distributed in Central Africa, from Hast
to West. There are some specimens of both sexes from S. Nigeria, Ibadan,
29.x1.1913 (Dr. W. A. Lamborn); from Nyasaland, Mt. Mlanje, 15.iii.13—vili.1913
(S. A. Neave) ; from Chintechi, Lake Nyasa, ii. 1913 (J. G. Morgan).

The other species, C. coloniarum, Speiser, has likewise a wide range, extending
from the Hast to the West Coast.

* *k * ES * *

As important additions have to be made to the forms grouped around the old
- genus “ Ceratitis,” a modification of my table on pp. 217-218 in my previous paper
is rendered necessary, beginning with no. 17 (28), as follows :—

17(28). Middle scp. well developed ; legs short and stout; wings of broader shape ;

head not balanced.

- 18(19). Antennae short, with the third joint distinctly pointed at end, and with
the second joint prominent and spinulose ; arista long plumose; usually
three 7. or. present ; genal bristle very strong.

A (B). Basal segment of the costa destitute of bristles before the costal one; basal
dark band of the wings perpendicular, beginning at stigma

Chelyophora, Round.

B (A). Basal segment of costa provided with two groups of strong bristles before

the costal one ; basal band oblique, beginning beyond the stigma
Bistrispmaria, Speis.

22(23).

93(22).

PROF. M. BEZZI.

. Third antennal joint rounded at end.

. Antennae very short, with the second joint prominent above and densely

spinulose; basal dark band of the wings oblique, beginning at fore
border beyond the stigma; costal bristle long and strong, as in the
preceding genera, like the genal bristle of Clinotaenra, gen. nov.

. Antennae less shortened, with the second jomt not prominent, never

spinulose above ; basal band perpendicular, beginning at stigma ; costal
bristle short and less developed; genal bristle much less developed.

. Head, in side view, less shortened, the eyes being therefore more rounded ;

thorax and scutellum prevalently yellowish or reddish, or densely
grey-dusted, with black spots (when the thorax is shining black, the
scutellum is yellow and not spotted) ; wings with yellowish bands, which
are only exceptionally broadly infuscated, and with blackish basal streaks ;
the band passing over the hind cross-vein is not, or only exceptionally,
united with the basal one.
Arista with short pubescence, which is more distinct on the upper side ;
frons of the male with conspicuous spatulate appendages
Ceratitis, MacLeay (s. str.).
Anzista with longer pubescence or even with short plumosity, Be
developed on both sides ; frons of the male without such appendages.

. Frons flat, normally shaped, with rather thin s. or., which are inserted on

Jess distinct and less converging plates.

. Middle legs of the male broadly feathered, at least on the tibiae; wings

with the dark marginal band separated from the basal one beyond the
stigma .. a Pterandrus, Bezzi.

. Middle legs of nals eptreh eho band typically united with the basal

one.

. Scutellum rounded, swollen, without distinct lateral keels, more or less

distinctly trilobate, black- spotted.

. A pair of well developed oc. present; first posterior cell not narrowed at

end ; fourth longitudinal vein ending much beyond the tip of the wing
Pardalaspis, Bezzi.

. Oc. quite wanting ; first posterior cell distinctly narrowed at end; fourth

vein ending at tip of wing - us Capparimyia, gen. nov.

. Scutellum flat and bluntly as ras with distinct lateral keels, entirely

yellow, not spotted .. wi : Re Perilampsis, gen. nov.

. Frons with very stout s. or., which are inserted on well developed and

converging plates, which are besides in the male very prominent and
cristate .. ae : f aS Hoplolopha, gen. nov.

B (A). Head distinctly en in ne view, the eyes being more narrow ; thorax

and scutellum shining black, sometimes with whitish markings, or even
the scutellum white with black spots; wings with blackish bands.
radiating from the base, which is typically destitute of streaks; arista
with long plumosity .. Me ai af .. Triuhithrum, Bezz.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 223

Leucotaeniella, Bezzi, 1918.
Leucotaeniella guttipennis, sp. nov. (Pl. xvu, fig. 3).

Very like trispila in the scutellar pattern, but distinct from it and from the other species
onaccount of its peculiar wing pattern, which approaches that of the reticulate type.

®. Length of body 6-6°5 mm. ; of ovipositor 1 mm. ; of wing 5-6°5 mm.

Head exactly as in trispila, but without the black ocellar dot, and the semi-
circular spot in the middle of the frontal stripe is paler and less distinct. Thorax,
scutellum and halteres as in érispila ; the black apical scutellar spot is less developed,
being much smaller than the space between the apical pair of bristles, and not
at all visible from above. Abdomen similar, but the ovipositor not darkened at
end; legs entirely similar. The wings in shape, coloration and venational
characters are exactly as in trispila, but they show conspicuous differences in the
pattern. This last in its general features is of the same type, but the middle and
cubital bands are broadly fused together and even with the marginal and anal
bands, forming thus a single broad patch filling up the apical two-thirds of the
wing, interrupted by some rounded hyaline spots, which form a kind of broad reticu-
lation. The basal band is broadly united with the anal one in the submarginal
and first basal cells, thus forming a rounded hyaline spot in the marginal cell just
below the stigma, and a hyaline indentation which ends above in the middle of
the first basal cell. The stigma is blackish on the basal half, and almost hyaline
in the apical one; the second costal cell is hyaline, with the two ends narrowly
infuscated. The anal band is broader, and infuscated on its lower part; the
marginal band has a broad hyaline spot before the end of the marginal cell (which
shows therefore two broad hyaline spots), and another at wing border before the
end of the submarginal cell, which has another hyaline spot in the middle just below
but much smaller than the apical one of the marginal cell; the first posterior cell
is entirely infuscated, yellowish on the basal part and fuscous on the apical, with
three rounded hyaline spots, two of which are in the middle (one behind the other),
and one at the end in the lower apical corner just above the end of the fourth
longitudinal vein and passing a little below it. The second posterior cell has two
hyaline spots at the border with a less defined light (but not hyaline) space above
them; the discoidal cell has a rounded hyaline spot before its apex, united with
the indentation of the third posterior cell. All these hyaline spots and indentations
are broadly margined with fuscous; but the hind cross-vein is entirely on the
yellowish part, being nearer to the fuscous border of the spots of the second posterior
cell than to that of the indentation of the discoidal cell.

Type 9, a single specimen from N. Nigeria, Zungeru,18.xi.1910 (Dr. J. W. Scott
Macfie). An additional female specimen from the same locality and collector differs
only in having three (instead of two) hyaline spots in the submarginal cell along
the costa, a third (half as small) hyaline spot being present between the two spots
of the type.

Bistrispinaria, Speiser, 1913.

Owing to the fact that the present very distinct genus was shortly characterised
by Dr. Speiser as a section of Ceratitis, I will give here a more complete description
of it. It is more closely allied to the genera Chelyophora and Acroceratitis, than
to those that have been separated by me from Ceratitis, s.].

294- PROF. M. BEZZI.

Head in front view a little broader than high, and distinctly broader than the-
thorax; eyes rather narrow; lower occipital swellings much developed; jowls
rather broad. rons broader than the eye, as broad as long, a little concave before:
the ocelli. Antennae broadly separated from each other at base, inserted near
the middle of the eyes, with the second joint very prominent and beset with strong
spinules, with the third joint pointed at the end and reaching to the middle of the
face ; arista long plumose on both sides. Cephalic bristles long and strong, black ;
the two s. or. are placed on converging plates and the second is very stout, curved
backwards and inserted over a small black tubercle ; three strong 7. or.; oc. very
long and strong; ocp. black and well developed ; genal bristle very long. Frontal
stripe with only a few hairs towards the middle. _ Thoracic chaetotaxy complete ;
dc. placed a little before the line of the a. sa.; two mpl.; pt. long and strong.
Scutellum rounded, convex, swollen, less distinctly trilobate, shining black, with
four bristies and some stout whitish hairs above. Abdomen ovate, as broad as
the thorax, acute at the end, bristly at hind border of the segments ;
male genitalia small. Legs stout, simple; front femora with a complete row
of bristles beneath; middle tibiae with two spurs at end ; hind tibiae without distinct:
posterior row. Wings rather long, with fuscous basal streaks and fuscous bands ;
basal band oblique, with its upper end beyond the stigma, united with the marginal
band ; middle band complete ; cubital band isolated. Basal segment of costa with
two groups of 2-3 erect spines each, the first placed just above the humeral cross-vein,
the second just before the costal bristle, which is well developed. First longitudinal
vein ending before the small cross-vein ; second vein nearly straight ; third vein
a little curved about the middle of its last section, and then parallel with the fourth
or slightly diverging, beset with scattered bristles to beyond the small cross-vein,
which is placed near the middle of the discoidal cell ; hind cross-vein long and rather
sinuous, inwardly oblique, the lower angle of the discoidal cell being acute ; discoidal
cell about three times as long as the hind cross-vein; anal cross-vein deeply sinuous;
the anal cell with a broad point, which is not longer than the second basal cell.

1. Bistrispinaria fortis, Speiser, 1913. (Pl. xvu, fig. 5).

A stout species, distinguished by its frons, antennae and legs.

¢ Length of body 5 mm. ; of wing 6 mm.

Head entirely pale yellowish, opaque, with a dark brown subocular spot; face
whitish, with a dark yellowish stripe at the mouth-border; frons with a darkish
rounded middle spot. Antennae entirely yellowish, with the third joint distinctly
darker; the spinules of the second joint are black and thick. Palpi whitish, with
black bristles; proboscis yellowish. All the bristles are black, lke the hairs on
middle of the frons. The lower occipital swellings are whitish and clothed with
white hairs. The thorax is blackish on the dorsum but densely grey-dusted and
clothed with pale yellowish hairs; on the disc there are three narrow longitudinal
darker stripes, the lateral ones being broadly interrupted at the suture; on the sides
there are two broad, subquadrate, shining black, ill-defined spots; the humeri are
yellowish; on the postalar calli there is a rounded, deep black, opaque spot.
The pleurae are reddish brown, rather shining, with a whitish yellow patch on
the upper border of the mesopleura, continued below the root of wings, and

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 225,

there united with the double whitish hypopleural spot; the hairs of the pleurae
are white on the whitish parts, and black on the dark parts. Postscutellum
shining black ; mesophragma reddish brown, but densely grey-dusted. Scutellum
shining black, with a narrow middle longitudinal yellow stripe, and with
a yellow stripe on each side; the bristles are black and the hairs pale yellowish.
Halteres pale yellowish. Abdomen with black bristles and yellowish hairs ; it is
blackish, but grey-dusted, with a narrow whitish hind border to the segments. Legs
pale yellowish, with the femora broadly infuscated and the four posterior tibiae
darker on the basal half; bristles black. Wings whitish hyaline, with blackish
basal streaks and with fuscous bands ; veins black, but the costa and the first longi-
tudinal vein whitish at the stigma. At the base there is a short blackish band,
beginning at the humeral cross-vein; there is a blackish streak in each cell,
broader and more sharply defined near the fore border; but the anal cell has no
middle streak. Stigma whitish, with the base narrowly darkened. The first band
is oblique, beginning at fore border just beyond the stigma, passing over the small
cross-vein, and ending evenly and symmetrically at the end of the sixth vein. The
costal band is united with the above-named band, extending below to the third.
vein and on the apical part passing even a little beyond it; it includes 6-7 darker
dots along the fore border, being moreover narrowly hyaline near the costa. About.
at middle of the last section of the third longitudinal vein there arises from the
costal band the middle band, which is narrow and oblique, irregularly interrupted.
at base, and crosses just beyond the middle the last section of the fourth vein, ending
at the border a little after the end of the fourth vein. The cubital band is broader.
than the middle one, extends from the hind border across the hind cross-vein, and.
ends roundly above it in the first posterior cell, being distinctly separated from.
the basal band. Axillary lobe hyaline, only with a aiffused greyish shade at the
apex.

Originally described in the male sex from Kamerun, Soppo, xu. 1912; there is:
a male from Uganda, Kampala, 10.xi.1915 (C. C. Gowdey).

Clinotaenia, gen. nov.

Notwithstanding the obtuse third antennal joint, the present new genus seems
to be more closely allied to the preceding one than to the following genera. Its.
wing pattern is very like that of Bastrispunaria, and is not very different from that
of Trypeta grata, Wied., a species which might therefore be better placed here than
in Leucotaenella. Loew (Berl. Ent. Zeits., v, 1861, p. 268) has pomted out that
the allies of 7. grata are to be found only in the neotropical fauna; the American
forms with a similar wing pattern are those of the genus Anastrepha, which are
however very different in their structural characters; but it is interesting to note
that the end of the fourth longitudinal vein in the species here described shows a
slight tendency to be curved upwards. The new genus Clinotaenia is closely allied
to the Oriental genus Gastrozona, but has only two z. or., and different body and wing
patterns. It must be recorded that Prof. De Meijere has recently described * from
Java a species of Anastrepha (A. extranea), which however cannot be placed in the
American genus, nor in the present one.

* Tijdschr. v. Entom., lvii, 1914, p. 193, pl. v, fig. F.

226 PROF. M. BEZZI.

Head in front view distinctly higher than broad, as broad as the thorax; frons
quite flat, proportionally narrow, as broad as the eye, longer than broad ; occipital
lower swellings distinct, but less developed than in Bistrispinaria; eyes narrow,
the head being very narrowed in profile; jowls rather broad. Antennae approxi-
mate at base, inserted distinctly above the middle of the eyes ; they are very short,
with the second joint rather prominent above and beset with short and thick spinules ;
third joint rounded at tip, not reaching the middle of face; arista long plumose.
Cephalic bristles well developed and black; the s. or. are inserted over less distinct
and not converging plates, and the anterior one is not specially dilated; only
two 2. or.; oc. well developed, but rather short ; ocp. black and acute ; genal bristle
of medium size. Thoracic chaetotaxy complete, with black and strong bristles ;
middle scp. long and approximated; dc. on a line with the a. sa. or only a little
beyond it; two mpl.; pt. as strong as the st. Scutellum swollen, rounded, not
trilobate, with four bristles, the apical ones decussate, and with sparse thin hairs ;
it is whitish yellow, with three black spots. Abdomen oval, as broad as the thorax,
bristly on sides and at end, with whitish transverse bands; ovipositor as long as
the last three abdominal segments together, triangular, rather swollen at base,
broadly obtuse at end. Legs short and stout; front femora with a complete row
of strong bristles beneath; middle tibiae with a single spur at end; hind tibiae
with a row of strong bristles in the middle of the outer side. Wings proportionally
long, with blackish basal streaks and with fuscous bands, the first of which is placed
obliquely (whence the generic name); costal bristle strong and long, but there are
no other strong bristles before it. First longitudinal vein ending before the small
cross-vein ; second vein about straight ; third vein beset with dense bristles through-
out its whole length, its last portion parallel with the last portion of the fourth ;
fourth vein slightly but distinctly curved upwards at end; small cross-veins long
and oblique, placed beyond the middle of the discoidal cell ; hind cross-vein straight,
inwardly oblique, the lower angle of the discoidal cell being acute; discoidal cell
three times as long as the hind cross-vein ; anal cross-vein deeply angulate in the
middle ; lower point of the anal cell of medium length, as long as the second basal cell.

Type: the following new species.

Clinotaenia anastrephina, sp. nov. (Pl. xvi, fig. 6).

A pretty species readily distinguished by its antennae and wing pattern.

Length of body 5:5 mm.; of ovipositor 1 mm. (when completely exserted 2:2
mm.); of wing 5°5 mm.

Head entirely pale yellowish, with a broad blackish patch on the occiput from the
neck to the upper corner of the eyes, with a small brownish subocular spot
on the jowls ; towards the middle of the frons there is a reddish yellow spot. The
frons 1s opaque, only the narrow plates bearing the s. or. being shining; ocellar
dot blackish. Antennae reddish yellow, the second joint appearing infuscated
owing to the numerous black spinules. Proboscis and palpi yellowish, the latter
with rather thick black bristles. Occipital swellings whitish and clothed with
whitish hairs, with a few short and thicker black ones at lower border. Frontal
stripe in the middle with a few short black hairs. The thorax is blackish on the

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. wealth

dorsum, but densely dark grey tomentose and with short golden-yellow pubescence ;
the sides above the notopleural line and above the root of the wings are shining
blackish brown, while on the postalar calli there is a rounded, velvety black spot;
on each side, behind the suture and above the blackish supra-alar patches, there
is a narrow and ill-defined, yellowish stripe. Pleurae blackish, partly reddish
brown, shining, with a broad oblique whitish patch across the meso- and pteropleura,
uniting in front with the whitish immaculate humeral callosities; there is also a
broad double whitish hypopleural spot; sternopleura shining black; mesopleura
hairs whitish. Postscutellum and mesophragma shining black. Scutellum shinin
whitish, with whitish hairs ; there is a narrow black stripe at base, and three broad
black spots, which are hardly visible from above, the apical one occupying the
whole space between the two apical bristles, and the lateral ones having the lateral
bristles inserted at their outer borders. Squamulae and halteres whitish. Abdomen
blackish, but clothed with dense grey and white dust; first segment shining black
on the basal half, white-dusted on the apical half, thus forming a complete and
broad cross band; second segment entirely brownish grey; third and fourth
entirely white, with a narrow blackish basal band. The abdominal hairs are white
on the white, black on the dark parts; all the bristles are black. Venter reddish
brown. Ovipositor entirely reddish, darker at base, with short black or dark
brownish hairs. Legs with dark reddish coxae and with blackish femora, which
are dark reddish at base; tibiae and tarsi pale yellowish, but the latter darkened
at end; bristles black ; hairs dark on femora and whitish on tibiae. Wings whitish
hyaline, with black veins, but the costa and the first longitudinal are whitish on
the hyaline parts, like the fourth and fifth veins. At the extreme base there is
a short fuscous band, extending from the humeral cross-vein to the base of the
second basal cell; after this band there are the blackish streaks, which are broad
and placed one in each cell, except the anal cell, and are present even in the base
of the submarginal cell and in the upper corner of the base of the discoidal cell.
The second costal cell is infuscated at the two ends; the stigma is hyaline, with
a blackish base. The bands are deep fuscous, the darker spots of the marginal band
along the costa being thus less distinct. Basal band oblique, beginning at fore
border beyond the stigma, passing across the small cross-vein and ending narrowly
at the tip of the sixth vein; it is almost straight on its outer border, with fuscous
teeth on the inner edge, two of which extend into the third posterior cell; at base
of the discoidal cell it is united with the fuscous basal streak of the same cell. The
marginal band extends below to the third vein, passing even below it near the apex
and ending at wing border near the middle of the first posterior cell. The oblique
middle band is rather broad ; it begins at middle of the last portion of the third
vein and ends at wing border just below the end of the fourth ven; the hyaline
indentation between the marginal and the middle band is narrower than the terminal
part of the marginal band itself. Cubital band less oblique, beginning at hind wing
border, passing across the hind cross-vein and ending rather narrowly above it in
the first posterior cell, quite separated from the basal band; its external border
is convex, while the internal one is concave, the band being thus distinctly arcuate.
Axillary lobe entirely hyaline, like the alula.

Type 9, a single specimen from Nyasaland, Mt. Mlanje, 3.xn.1913 (S. A. Neave).

(637) B

28: PROF. M. BEZZI.

Pterandrus, Bezzi, 1918.

Pterandrus rosa, Karsch, var. fasciventris, nov.

Differs from the type in having well developed blackish bands on the abdomen.

3. Length of body 5-5.5 mm. ; of ovipositor about 1 mm.; of wing 55-6 mm.

Head, its appendages and bristles exactly as in rosa. Thorax and scutellum like-
wise, but the dorsum of the mesonotum in fully coloured specimens seems to be darker
and more densely grey-dusted. The chaetotaxy is the same, the single mol.
being present. The abdomen in both sexes shows a distinct, complete, blackish.
rather shining cross band on the apical half of the second segment; the third and
fourth segments are whitish, but in the female they have a distinct dark cross
band at base. The ovipositor is shining reddish, with a distinct black base and with
a less intense apical black border, but these dark parts are not sharply separated.
The legs are as in rosa, the male having the middle femora quite simple, and the
middle tibiae feathered on the apical half alone; the front femora of the male are
not densely ciliated. The wing pattern is the same as in rosa, but the stigma is
blackish, with the apical third hyaline.

Type ¢, type 9, and an additional female specimen from Uganda, Entebbe,
17.vii.1911 (C. C. Gowdey); another male specimen (more lightly coloured owing
to immaturity) from Uganda, Kampala, 5.xu.1916 (C. C. Gowdey).

The species Pterandrus anonae, P. colae, P. rosa and Pardalaspis morstattv are
very closely allied and have a similar wing pattern; the feathering of the middle
legs in the male decreases from the two former (which have femora and tibiae
feathered) to the third, in which only the tibiae are feathered, and to the fourth,
in which they are not feathered at all. The two former species agree also in the
ciliation of the front femora, which is lacking in the two latter; morstattc in the
length of the ovipositor agrees with colae, while anonae and rosa have a short
ovipositor. Ina more natural classification P. morstatt: must be placed in the genus
Pterandrus, notwithstanding the lack of feathering on the middle legs in the male.

Pardalaspis, Bezzi, 1918.

I will give here another table for determining the rather numerous species of
the present genus, excluding znscripta, Graham (belonging to Trirhithrum), and
with the addition of two new species.

1(18). Wings with the band across the hind cross-vein quite isolated, and without
a complete middle band, or rarely with a small isolated streak on the
last portion of the fourth longitudinal vein.

2(15). The black spots of the scutellum very broad, quadrate or rectangular in
shape, closely approximated, and nearly covering the whole surface.

3 (4). Wings with the aE band isolated (as in Pterandrus); arista shortly
plumose .. ae morstaitt, Bezz.

4 (3). Wings usually with ie high eta, bevel united at stigma with the
basal one ; when they are separated, the arista is pubescent only.

5(14). Last portion of the fourth longitudinal vein destitute of an oblique dark
streak across its middle.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 229

6 (7). Two strong mpl. present; wings broad, with strikingly developed basal
streaks and dots, and with the bands much infuscated ; head of the male
with a peculiar, striking coloration .. a .. punctata, Wied.

7 (6). Only one mpl. present as a rule; wings with a less developed basal pattern
and with pale yellowish bands, which at most are infuscated only towards
the hind border ; head of the male not peculiarly coloured.

8 (9). Thorax destitute of black spots ; species of greater size (8 mm.)

brémer, Guér.

9 (8). Thorax with distinct black spots, at least on the sides beyond the suture ;
size smaller (not over 7 mm.).

10(11). Occiput above with broad, shining black spots; the shining black parts on
back of mesonotum more developed than the yellow ones; scutellum
black, with a narrow yellow base a e:: melanaspis, sp. nov.

11(10). Occiput yellow above or only a little infuscated; yellow parts of back
more developed than the black ones; scutellum yellow with black spots.

12(13). Humeri with a distinct black spot; sides of thorax with black spots before
the suture; scutellum with two basal black spots besides the three apical
ones 2: ; t .. cosyra, Walk. (giffards, Bezzi).

13(12). Humeri and sides of thorax eked the suture destitute of black spots ;
scutellum with no distinct basal spots.. de fe silvestrir, Bezzi.

14 (5). Last portion of the fourth vein crossed over the middle by a well developed,
oblique, dark streak ; wings short and broad .. .. antestictica, Bezzi.

15 (2). The black spots of the scutellum much smaller and broadly separated, the
outer ones being more rounded; the scutellum therefore mainly yellow.

16(17). Scutellum with three black spots ; wings with the Set and basal bands

wholly separated beyond the stigma .. flexuosa, Walk.
17(16). Scutellum with five spots; the two Lge tania Bands united and fused ©
with the stigma .. Ne : Bye quinaria, Bezzi.

18 (1). The band over the hind cross-vein not ceomiart or when its union with the
other band is less distinct, there is a complete middle band.

19(20). The marginal band united with the basal one and with the stigma; the
band over the hind cross-vein united with the basal one; no complete
middle band Ht oP fi stectuca, Bezzi.

20(19). The marginal band separated gi Te ale one beyond the stigma; the —
band on the hind cross-vein united with the marginal one; a “complete
middle band across the last section of the fourth vein. . aliena, Sp. Nov.

Pardalaspis melanaspis, sp. nov.

Allied to P. cosyra (giffardz), but smaller and at once distinguished by the broad
black occipital patches, and by the more extended shining black pattern of the
thorax and scutellum.

39. Length of body 3:2-3°5 mm.; of ovipositor 0°5 mm. (when completely
exserted 1°5 mm.); of wing 3°2-3°5 mm.

Head pale yellowish, with reddish-yellow frontal stripe, with a black, rather broad,
rounded ocellar spot, and a broad, shining black patch on upper part of the occiput,
extending from the neck to the eyes, but leaving a narrow yellow border near the

(637) BQ

230 PROF. M. BEZZI.

eyes. There is no subocular spot on the jowls. The antennae are entirely yellow,
about as long as the face; second joint not prominent, never spinulose; third
joint broad and long, attenuated but rounded at end ; arista dark, shortly pubescent
on both sides. Upper mouth-border rather prominent and shining. Palpi broad,
flat, yellow, with short black bristles ; proboscis yellowish. Cephalic bristles black,
but the pvt. and the very thin and indistinct genal bristles are yellowish, the former
sometimes infuscated ; the hairs of the jowls and of the lower part of the occiput
are yellowish ; frontal stripe with sparse indistinct darkish hairs. Thorax entirely
pale yellowish and rather shining, with broad black spots on the dorsum ; it is clothed
with short, pale yellowish hairs. Humeri with a broad rounded black spot; from
the humeri to the transverse suture extends a broad rectangular blackish-brown
shining patch, reaching inwardly to the dorsocentral lines, and including a deep black,
opaque spot above the notopleural line with the anterior npl. Between these two
black patches there is a rather broad, shining black, middle stripe, extending from
the black scapular region to the suture, and ending there truncately. Beyond the
suture the back is entirely shining black, with a concave front edge; along the middle
line there is a pale yellow, longitudinal stripe, which is a little narrower than the
presutural black stripe and ends with a transverse, pale yellowish streak in front
of the scutellum, thus forming a § shaped mark. Above and in front of the root
of the wings, and on the postalar calli there are rounded, deep black, opaque spots.
The pleurae are of a shining pale yellowish colour, whitish on the upper part of
mesopleura, and there with a narrow longitudinal black stripe, extending from
below the humeri to the mpl. bristle. The upper border of the sternopleura is pale ;
there is a broad double whitish hypopleural spot. All the bristles are black and strong,
but the well developed middle scp., and the thin pt. and st., are pale yellowish ;
one mpl. only ; dc. placed before the line of the a. sa. Scutellum rounded, convex,
slightly trilobate ; it is entirely shining black, with a narrow, pale yellowish, basal
border ; it has rather long and numerous whitish hairs on the disc, and four long
black bristles, the apical ones not decussate, or only shghtly so at the tips. Postscu-
tellum shining black; mesophragma yellow, clothed with dense whitish dust on
the upper half. Squamulae whitish ; halteres yellowish, Abdomen entirely yellow,
rather shining, with numerous blackish hairs, and with black bristles at end and
on sides. The narrow hind border of the first segment in both sexes and almost
the whole of the third segment in the female, are densely clothed with whitish dust,
forming thus two white cross bands. Ovipositor shining yellow, with yellowish
hairs, with the basal segment bluntly triangular, about as broad as long, swollen.
Legs entirely pale yellowish, with yellowish pubescence ; the long bristles of the
underside of the front femora, and those of the hind tibiae are likewise yellowish ;
the single spur of middle tibiae is black. Wings proportionally long, with short
and less developed costal bristle ; the veins are entirely pale yellowish, and typically
disposed ; the stigma is about twice as long as broad, quite yellow, or rarely with
the apical half subhyaline. The pattern is yellowish and shaped as in cosyra, but
the basal dark streaks are less numerous, the second basal cell having a single blackish
dot, and the anal cell having no dots at all.

Type dg, type 9, and an additional female specimen from Cape Colony, Grahams-
town, 9.0.1905 (C. W. Mally).

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 231

Pardalaspis aliena, sp. nov.

Allied to cosyra and silvestrw, but distinct from them as well from all the other
species in having the costal band of the wings separated from the basal one, and in
having a complete middle band across the last section of the fourth longitudinal vein.

9. Length of body 4°2 mm. ; of ovipositor 0°6 mm. ; of wing 4:2 mm. |

Head pale yellowish, without any dark spot, only with a small black ocellar dot.
Face whitish, not prominent at the mouth-border. Antennae wanting in
the type, only the shortened and pale yellowish basal joints being present. Palpi
whitish, not very broad, without bristles, with whitish hairs ; proboscis pale yellow.
Cephalic bristles black ; oc. well developed. Thorax shining yellow, with short
whitish pubescence. Humeri whitish, immaculate ; above them there is on each
side a shining black, longitudinal stripe, which ends before the suture, but is con-
tinued behind it along the dorsocentral lines by a black subquadrate spot on the dc.
bristles, and by a broad stripe in front of the scutellum, including a broad whitish
patch, in the shape of a rounded spot. Above the notopleural line and above the
root of the wings there are the usual velvety black, opaque spots, which are separated
by the whitish notopleural callus; on the postalar calli there is also a rounded,
deep black spot ; above the shining supra-alar patch there is on each side a short
whitish stripe. The pleurae are pale yellowish and shining, with a white meso-
pleural spot and a double white hypopleural spot. Postscutellum deep black;
mesophragma pale yellowish, with faint whitish dust. Scutellum shining whitish,
with three subquadrate, equally broad spots at hind border ; it is convex and less
distinctly trilobate. All the thoracic bristles are black, even the middle scp. and
the pt., only the st. being yellowish ; one mpl. Squamulae white ; halteres whitish.
Abdomen shining yellow, with whitish pubescence and with black bristles at sides
and at end; the narrow hind border of the first and third segments is covered with
whitish dust, forming thus a very narrow and ill-defined white cross band ; ovipositor
shining yellow; venter pale yellowish. Legs entirely pale yellowish and with
concolorous bristles and hairs, only the spur of middle tibiae being black. Wings
with entirely pale yellowish veins, and with a very short and black costal bristle ;
the bands are pale yellow, infuscated towards the hind border ; the stigma is yellow,
with the extreme upper corner subhyaline. At base there are some darkish dots
and streaks, three in the second costal cell, one streak in the other cells and none
in the anal cell. The basal band is perpendicular, beginning at stigma, including
the small cross-vein, which is situated just at its outer border, and ending with a
sheht curve at apex of the sixth longitudinal vein. After this band there is a complete
hyaline band, which extends to the fore border just beyond the stigma ; the marginal
band is therefore separated from the basal one. The marginal band extends below
to the third vein, passing beyond it with its infuscated end, and ending at apex
about in the middle of the first posterior cell ; 1t includes two darker marginal dots
in the marginal] cell, and a rounded hyaline spot at end of the second vein. A little
before the middle of the marginal band arises the narrow and complete middle band,
which, passing across and beyond the middle of the last section of the fourth longi-
tudinal vein, ends infuscated at hind border in the upper third of the second posterior
cell. The cubital band extends across the hind cross-vein and is only slightly oblique ;
above it becomes gradually paler, but is distinctly united with the costal band just

232 PROF. M. BEZZI.

near its base. The small cross-vein is placed noticeably before the middle of the
discoidal cell ; the last portions of the third and fourth longitudinal veins are slightly
divergent ; the hind cross-vein is straight and placed a little obliquely ; the produced
lower angle of the anal cell is rather broad, and a little shorter than the second basal
cell.

Type 9, a single damaged specimen from Cape Colony, Grahamstown, 30. xi.
1908 (C. W. Mally).

Capparimyia, gen. nov.

I will erect here this new genus (which, so far as at present known, is not Ethiopian
but only Mediterranean) with Ceratitis savastani, Martelli,* as type. This species
shows all the characters of Pardalaspis, as well as the same coloration of body
and the same wing pattern, but it is at once distinguished by the complete want
of the oc. bristles, a character which is present in none of the allied forms. The
ending of the fourth longitudinal vein at the tip of the wing is also unique among
the allies of Ceratitis, and it is rendered even more evident by the distinct narrowing
of the first posterior cell, due to the curving downwards of the third longitudinal
vein in its last portion.

The erection of a new genus for the present species 1s supported by the fact
that it is a gall-making insect ; the young buds of the flowers of Capparis spinosa,
L. (whence the generic name), in which the larvae live, become hypertrophied,
deformed and arrested in their development, the effects being similar to those
caused by the Itonid midge, Asphondylia capparis, Riibsaamen. The gall of
Capparimyia savastani is included in the great work of Prof. Houard}; and
it must be remembered that all the larvae of the allied forms live in fruits and are
not galligenous, and that the faculty of making galls is very rare in the subfamily
CERATITINAE, while it is not rare in the subfamily TRYPANEINAE.

The main characters of the new genus are as follows :—Pvt. yellow; ocp. black
and acute; two 2. or.; the s. or. are inserted on indistinct and not converging
plates; genal bristle thin and yellow; dc. placed only a little behind the line
of the a. sa.; one mpl.; pt. as developed as the st., and both yellowish in colour,
like the middle scp., while all the other bristles are black. Wings with a short
but strong costal bristle; first longitudinal vein ending noticeably beyond the
small cross-vein, which is placed before the middle of the discoidal cell; third
vein bristly to the middle of its last portion ; first posterior cell narrower at end
than in the middle, the last portion of the third vein being curved downwards
before the end; fourth vein ending very near the tip of the wing; lower angle
of the discoidal cell acute, the hind cross-vein being however only a little oblique ;
lower point of the anal cell acute and produced, but a little shorter than the second
basal cell.

The single species at present known has been found only in Sicily and in Southern
Italy.

* Boll. Arboric. Ital., vii, 1911, p. 19, figs. 1-4, and Mem. R. Accad. Zelanti, (3) vii,
1912, p. 50, figs. 1-4. In the Zoological Record, vol. L, p. 423, the species appears as
published in 1913.

t Les Zoocécid. des plantes d’Europe, etc., ili, 1913, p. 1355, n. 6730.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 233

Perilampsis, gen. nov.

This new genus is erected here with Carpophthoromyia pulchella, Austen, as
type, and includes Carpophthoromyia formosula, Austen, but not Trypeta grata,
Wied., which probably belongs to Clinotaena, as stated above. It comprises
very beautiful (whence the generic name) Ethiopian flies, which agree in coloration
of the body and in wing pattern, as well in the following characters :—

Head in front view broader than high, the face being a little shorter than the
frons and the jowls very narrow; eyes rounded; occipital lower swellings less
developed ; frons about as broad as long, flat, with the shining vertical plates
short, not prominent, slightly converging in front ; face distinctly concave in the
middle and prominent at mouth-border. Antennae inserted at middle of eyes,
rather long but always shorter than the face; second joint not prominent, never
spinulose above; third joint twice as long as the second and obtuse at end;
arista with long pubescence, or even shortly plumose, the feathering being much
narrower than the breadth of the third joint. Palpi broad and shortly bristled ;
proboscis short. Cephalic bristles black; ocp. numerous and long, black, acute ;
oc. strong and long; two 7. or.; genal bristle black and rather long. Thorax
shining black on the back of mesonotum and adorned with characteristic transverse
bands of minute, pale-coloured hairs; pleurae with whitish mesopleural stripe
and with double hypopleural spot ; scutellum quite yellow, destitute of black spots.
Thoracic chaetotaxy complete; middle scp. long and approximated; dc. placed
a little beyond the line of the a. sa.; one mpl.; pt. as strong as the st.; all the
bristles are black. Scutellum broadly triangular, not trilobate, rather flat above,
with distinct lateral keels, pubescent above and with four bristles. Abdomen
shortly oval, distinctly broader than the thorax at base, and with black bristles
at end and on sides. Legs stout, with complete rows of bristles on front femora
and on hind tibiae ; middle tibiae with a single spur. Wings broad and long, with
a blackish and characteristic pattern, the middle band of which is more or less
complete, and the basal band perpendicular and beginning at the stigma. Costal
bristle short but strong; first longitudinal vein ending opposite to the small
cross-vein ; second vein straight; third vein bristly to the middle of its last
section ; third and fourth veins slightly divergent towards the end, and thus the
first posterior cell not narrowed outwardly ; fourth vein ending much behind the
tip of wing; small cross-vein before the middle of the discoidal cell; hind cross-

vein oblique inwardly ; point of the anal cell a little shorter than or as long as the
second basal cell.

Perilampsis puichella, Austen, 1910.
Of the present species, known from Uganda and Kamerun, there is a specimen

from 8. Abyssinia (Rk. J. Stordy). The species seems therefore to have a very
wide distribution over the whole centre of the Ethiopian Region.

Hoplolopha, gen. nov.
This new genus is a very interesting one, because, although it is without doubt
closely allied to Ceratztis (s. 1.) and has a wing pattern very like that of the preceding
genus, it shows in the male sex on the frons the cristiform protuberances with

234 PROF. M. BEZZI.

_Incrassated bristles (whence the generic name), which are characteristic of other
Trypaneid genera, like Straussia, Vidalia, Stemonocera, ete.

Head in front view a little higher than broad. LHyes rather narrow; lower
occipital swellings well developed; jowls as broad as the breadth of the third
antennal jot. Frons longer than broad, narrower than the eye; in the female
it is flat, while in the male it is distinctly concave in the middle ; the shining orbital
plates are produced beyond the middle of the frons, and converge in front, and
moreover in the male they are prominent and cristiform. Face shorter than the
frons, rather flat, and with a less prominent mouth-border. Antennae inserted
at the middle of the eyes, proportionally long, but always shorter than the face ;
second joint not specially prominent, nor spinulose; third joint elongate linear,
rather narrow, about three times as long as the second, rounded at tip; arista
pubescent, the breadth of its feathering being much narrower than the breadth
of the third antennal jomt. Palpi broad; proboscis short. Cephalic bristles
black and strong; ocp. thin and pointed; oc. strong and long; the three s. or.
are in both sexes very long and incrassated, chiefly the two former, curved back-
wards, but in the male they are distinctly longer; two 7. or., rather short and
much shorter than the superior ones ; genal bristle rather short. Thorax yellowish
grey, pale pubescent, with ill-defined black pattern, and with less striking, whitish,
pleural markings ; scutellum swollen, rounded, yellowish, with three broad, shining
black spots. Middle scp. long, approximated, black lke all the other bristles ; de.
a little beyond the line of the a. sa.; one mpl.; pt. as strong as the st. ; four sct.,
the a. sct. diverging. Abdomen short and oval, with bristles on sides and at end ;
male genitalia developed ; ovipositor narrow and cylindro-conical, swollen, as long
as the last three abdominal segments together. Legs short and stout ; front femora
with a complete row of bristles below ; middle tibiae with one spur; middle and
hind tibiae with an external row. Wings long but proportionally not broad, with
a short but strong costal bristle, with basal dark streaks and dots, and yellowish
bands ; basal band beginning at stigma and perpendicular ; middle band complete.
First longitudinal vein produced beyond the small cross-vein ; second vein straight ;
third vein densely setigerous to the middle of its last portion; small cross-vein
oblique and placed considerably before the middle of the discoidal cell ; third and
fourth slightly divergent at end, the fourth ending much after the tip of wing
and the first posterior cell not narrowed outwardly ; hind cross-vein before the
middle of the first posterior cell, very oblique inwardly, the lower angle of the
discoidal cell therefore acute ; discoidal cell three times as long as the hind cross-
vein; anal cross-vein deeply sinuous, the lower angle of the anal cell very long.
longer than the second basal cell; axillary lobe broad and semicircular, in the
male separated by a deeper incision of the hind border at the end of the sixth
longitudinal vein.

Type: the following new species.

Hoplolopha cristata, sp. nov. (Pl. xvii, fig. 7).

A robust fly of comparatively large size, at once distinguished from all the allied
forms with a typical Ceratitis-like wing pattern on account of the cristate frons of the
male.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 255

$2. Length of body 6°5-7 mm.; of ovipositor 1 mm.; of wing 75-8 mm. ;
breadth of wing about 3 mm.

Head yellowish, but more whitish on the face, on the linear cheeks and on the lower
occipital swellings ; the frons is opaque and immaculate, but the ocellar and vertical
regions and orbital plates are shining and of a more reddish colour; below the
lower corner of the eyes there is an ill-defined brownish genal spot, more developed
an the female than in the male. Antennae entirely yellowish, the third joint infus-
cated along the upper border ; palpi pale yellowish, with some black bristles at the
end ; proboscis reddish. All the cephalic bristles are black, even the pvt. and the
genal ones ; the lower occipital hairs are white, but along the lateral borders of the
oral cavity there is a rather dense row of short bristly black hairs. Thorax on the
-dorsum reddish brown, but covered with a dense grey dust and with pale pubescence ;
‘it is opaque, with a more or less distinct trace of a middle longitudinal blackish
-stripe ; on the dorsocentral lines there is on each side a blackish stripe, which is
‘interrupted to form a rounded postsutural spot, and a broader and shining elongate
-spot in front of the scutellum ; above the wings there is also on each side a blackish
-stripe, and two broad blackish spots, the anterior shining, the posterior deep black.
Humeral calli reddish or pale yellowish, not spotted. Pleurae reddish yellow, rather
‘shining, with white dust and with whitish mesopleural stripe, more distinct in the
‘male, and with rather long, pale yellowish hairs, which are however dark on the
lower part of the mesopleura, while on the breast they are white. The double whitish
hypopleural spot is more distinct in the male. Postscutellum shining black, more
-or less reddish below and on sides ; mesophragma reddish with grey dust. Scutellum
with three very extended, shining black, rectangular spots, and between them
very narrow, yellowish or reddish stripes; in the narrowly yellow base there are
two dark spots, in contact with the hind border of mesonotum. MHalteres whitish,

-with reddish knob ; squamulae whitish, with narrow dark border. All the bristles
-are black, even the middle scp. Abdomen reddish yellow, with grey dust ; the
second, third and fourth segments are blackish at base, more broadly in the second,
which is almost entirely black, while the third is more densely clothed with whitish
dust, thus forming a transverse band, more distinct in the female. The hairs are
whitish on the pale, and blackish on the dark parts; the bristles are black. Male
genitalia black and black-haired ; ovipositor shining reddish, darkened at tip, with
-darkish hairs. Legs entirely pale yellowish, with blackish pubescence, black bristles
and black claws. Wings hyaline, iridescent, with yellowish veins, only those of the
lower part of the base being blackened. The extreme base, before the humeral
cross-vein, is infuscated as a short yellowish band. The basal blackish dots
-and streaks are numerous, and are more developed in the female than in the
male; they invade the anal cell also, and even the base of the discoidal and
third posterior cells. The stigma is yellowish, darker on the basal half. The bands
are yellowish, infuscated towards the apex and the hind border of wing. Basal
band perpendicular, going from the base of the stigma to the end of the sixth longi-
tudinal vein, but not extending into the axillary cell, or only slightly so; the small
cross-vein is placed just at the anterior border of this band. The costal band is
broadly connected with the basal one, its lower border going exactly along the third
“vein, surpassing it only near the tip of the wing, and there ending (infuscated)

236 PROF. M. BEZZI.

before the middle of the first posterior cell; it contains two rather large fuscous.
spots on the fore border in the marginal cell. From the costal band, about at middle
of the last portion of the third vein or a little beyond it, departs the fuscous narrow
middle band, which, crossing obliquely beyond the middle of the last portion of the:
fourth vein, ends at the hind border just below the tip of the fourth vein itself.
The fuscous cubital band is about twice as broad as the middle one; it begins at
the hind border of the wing symmetrically with the end of the fifth vei, and passing
over the hind cross-vein fills up broadly the end of the discoidal cell, and ends, more
pale-coloured, in the first posterior cell, but without reaching either the basal or
costal band; the base of the first posterior cell is thus completely hyaline. The
broad axillary lobe is hyaline, with an ill-defined fuscous middle spot; the alula is.
quite hyaline. The oblique folds in the discoidal and third posterior cells are about:
as dark as the axillary vein.

Type dj and type 9, a single couple of specimens from British East Africa, Kabete,.
28.vi.1914 (7. J. Anderson).

Trirhithrum, Bezzi, 1918.

The peculiar character of the wing pattern in the present genus is not (as wrongly
indicated on p. 233 of my previous paper) that the middle band is united to the:
marginal one, but that the cubital band (or the band on the hind cross-vein) is united
to the basal or anal band. Another error is to be found in p. 239 in the description
of the new species 7'r. occijntale, which has the thorax shining black, and not shining”
yellowish.

To the present genus is to be ascribed also Ceratitis inscripta, Graham, which
is the same as 7’. coffeae, Bezzi, as stated by Dr. G. A. K. Marshall ; and the following’
very distinct and robust new species, which seems to be allied to T. nitidum, v.
Roeder, in having a complete middle band on the wings.

Trirhithrum validum sp. nov.

A robust fly of comparatively large size, entirely shining black, with whitish
mesopleural stripe and whitish scutellum, and with a complete middle band on the:
wings, which have also rather distinct black streaks at the base.

©. Length of body 7°5 mm. ; of ovipositor 1°5 mm. ; of wing 8 mm.

Head with the occiput shining reddish in the middle and above, black on the:
sides; frons dull yellowish, face whitish on its lower part, jowls reddish like the
lower occipital swellings, which are shining black above. The frons is about as broad
as long, unspotted, only the ocellar and the narrow and short orbital plates being”
shining reddish. The cheeks are whitish, but their upper half, by the roots of the
antennae, is purplish. Antennae wanting in the case of the type. Upper half of
the face, just below the antennal root, blackish and grey-dusted. Jowls of a purple:
colour, with a dark brown spot below the eye. Palpi and proboscis reddish brown.
All the cephalic bristles are black, even the pvt. and the genal ones; two 2. or. ;
oc. very strong and long ; occipital hairs black, even below ; mouth-border on the.
sides with a row of black bristly hairs. Thorax shining black on the back, with
short dense black pubescence ; humeri black, narrowly reddish brown behind ;.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 237

pleurae shining brown, with a broad oblique whitish mesopleural stripe, on which
there are long and pale yellowish hairs, while the hairs are black on the rest; no
distinct lighter hypopleural spots. Postscutellum and mesophragma shining black.
Scutellum of a semicircular shape, flattened above, shining whitish, with three
broad shining black spots at hind border which are not visible from above ; it is
clothed above with rather long whitish hairs and bears four bristles, the middle
pair being not decussate. Thoracic bristles black, even the long and approximated
middle scp.; dc. placed on the line of the a. sa.; one mpl.; pt. stronger than the
st. Halteres blackish, with yellowish base ; squamulae dark yellowish, with blackish
border and with dark fringe. Abdomen with parallel sides, not broader than the
thorax and not narrowed at base; it is shining black, clothed with black hairs,
and with black bristles on sides and at end; the hind border of the first, and the
whole of the third segment are densely clothed with a whitish-grey dust, thus forming
two transverse bands, which are clothed with whitish hairs; the fourth segment
is reddish brown on the middle ; ovipositor elongate triangular, as long as the last
three abdominal segments together, depressed, shining black, with black hairs and
with the apical segment reddish ; venter shining black. Legs with reddish brown
coxae and femora, the latter more blackened outwardly, and with black pubescence ;
tibiae and tarsi pale yellowish and with whitish pubescence ; bristles of the femora
black, those of the tibiae tawny. Wings long and narrow, hyaline, with a blackish
pattern ; costal bristle strong but short; veins dark brown or blackish, but the
costa and the first longitudinal pale yellowish on the basal hyaline indentation.
The extreme base is blackened, being separated from the basal band by a triangular
hyaline indentation, the base of which is at costa in the two basal cells, and the
vertex is below the end of the second basal cell; in this hyaline part there are some
blackish dots, likewise in the basal blackish part there are some narrow hyaline
streaks, forming thus a Ceratatis-like pattern, though closer and less distinct. The
stigma is entirely black. The basal band begins at the stigma and extends to the hind
border below the end of the sixth vein, filling up almost the whole of the axillary
lobe, and being thus broadly united to the basal blackish part; the whole of the
first basal cell, and the base of the discoidal and third posterior cells are broadly
united with this band. The marginal band is broadly united at base with the basal
one, and does not extend beyond the third vein, passing only a little over it in its
terminal part, and ending at costa near the first third of the first posterior cell ;
it encloses two darker dots in the marginal cell, and two hyaline streaks at costa
before and behind the end of the second longitudinal vein. About at middle of
the last section of the third vein arises the narrow, but complete and even, middle
band, which, crossing beyond the middle of the last portion of the fourth vein, ends
at hind border just below the end of this same vein. The oblique cubital band is
broader than the middle one, and is broadly united with the basal one at the small
cross-vein ; it runs along the hind cross-vein, leaving hyaline the upper external
corner of the discoidal cell, and ends at hind border symmetrically with the end of
the fifth ven. Small cross-vein long and oblique, the first posterior cell being as
broad as the discoidal cell, which is very acute below, the hind cross-vein being
very oblique. Alula entirely infuscated.

Type 9, a single specimen from Uganda, Entebbe, 12—13.xii.1912 (C. C. Gowdey).

238 PROF. M. BEZZI.

Xanthorrhachista, Hendel, 1914.
Xanthorrhachista alata, Becker, 1910.
In my previous paper I overlooked the description of Acidia alata, Beck. (Ann.
Soc. Ent. France, 1910, p. 28), with which Xanthorrhachista cephalia, Hend. (1914),
is undoubtedly synonymous.

Themarictera, Hendel, 1914.
To the characters of the genus on p. 243 of the first paper must be added :

No oc. ; two pairs of vt.; two s. or. and two 2. or.; put. well developed ; genal
bristle thin ; ocp. long and acute, black. Scp. long, the middle ones approximated ;
no pt.; st. very short and thin, not properly wanting, as stated by Hendel; three
p. sa., the inner one thinner.

Themariciera laticeps, Loew, 1861.

Closely allied to T. flaveolata, F., but differing in the presence of two hyaline
spots, one in the submarginal and one in the first basal cell. The first of these spots
is placed on the limit between the brown and the yellow pattern, beginning at the
third longitudinal vein just beyond the small cross-vein, and extending above
nearly to the middle curve of the second vein; in this way is formed the black
indentation, which is exaggerated in Loew’s original figure. The second hyaline
spot is before the small cross-vein; it is ovate in shape, and extends from the third
to the fourth vein. These two spots are distinctly whitish, like the basal streaks,
as described by Loew. In the rest the unknown male is very like the female ;
the genitalia are rounded, not prominent, entirely shining yellow; the last
abdominal segment has numerous black bristles. The eyes are green, with two
purplish cross bands, converging anteriorly, on the upper half.

A male specimen from Natal, Durban, xi.1916 “reared from native fruit”
(C. P. van der Merwe, Agric. Dept. 8. Africa). This specimen is an aberrant one,
or is not fully coloured in the wing pattern, hyaline streaks being present in the
ends of the marginal, submarginal and discoidal cells and the whole middle of
the first posterior cell; the fuscous pattern is thus restricted to some broad
borders on the longitudinal and cross-veins.

Puparium. To the reared specimen is attached a puparium, which seems to be
rather small for such a fly ; it measures 6 mm. only in length, while the adult fly
is about 9 mm. long. It is of a dark reddish-brown colour and has a rather hard
and thick skin; the segmentation lines are not impressed ; the broadened areas
of the ventral side are not prominent ; the posterior spiracles are very approximated,
the surrounding parts being of a blackish colour. The surface is smooth, but quite
opaque.

Rhacochlaena, Loew, 1862.
The Ethiopian species of the present genus, with the addition of the new one,
may be distinguished as follows :
1(2). Wings destitute of a broad apical fuscous band, the first and second posterior
cells being almost completely hyaline wh .. » +. fascrolata, Lw.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 239

2(1). Wings with a broad fuscous apical band, the first and second posterior cells
, being for the greater part infuscated.

3(4). Apical hyaline spot of the wings broad, uae on both sides beyond the

ends of the 3rd and 4th veins He x pulchella, sp. n.
4(3). Apical hyaline spot small and not extending ee beyond the ends of
the above-named veins ae a i .. hammersten, End.

Rhacochlaena fasciolata, Loew, 1863.

The apex of the wing is occupied by a white rounded spot, which in certain lights
is very conspicuous. The ovipositor is 2 mm. long; it is cylindro-conical, shining
red, but broadly black at base above and more narrowly at tip ; it is clothed with
very short and dense, darkish pubescence. In the present species there is but a
single s. or.

A couple of specimens from Natal, Estcourt, ix. and x. 1896 (G. 4. K. Marshall).

Rhacochlaena puichella, sp. nov. (Pl. xvu, fig. 8).

Very near hammersteim, but at once distinguished by the much broader apical
hyaline spot of the wings.

g. Length of body 45 mm.; of wing 4 mm. Head reddish, face jowls and
lower part of occiput pale yellowish or whitish ; frons opaque, with shining vertical
plates and slightly whitish sides; the small ocellar dot is black. In the middle
of the face there is an ill-defined fuscous cross band, not reaching the sides.
Antennae entirely reddish, but the third joint a little infuscated at the end; the
third jomt is rather narrow, with the upper border distinctly concave ; it is 2}
times as long as broad, with rounded tip; arista shortly pubescent. Jowls very
narrow, narrower than the breadth of the third antennal jomt. Hyes greenish,
with two parallel cross bands of a purplish colour, one a little beneath the middle,
the other on the upper third. Palpi and proboscis yellowish. Cephalic bristles
black ; or. 1+ 3; no oc.; put. less developed; the longer bristles are rather
~nrtel at the sit. ilnoree: black, opaque, with greyish dust on the back; a
middle longitudinal stripe and the sides are reddish ; pleurae shining yellowish ;
along the notopleural line, from the humeri to the root of the wings, there is a
rather broad whitish stripe. Scutellum whitish in the middle and infuscated on
the sides. Mesophragma black and a little shining ; the hypopleurae are also black,
but with a narrow whitish stripe in continuation with the notopleural one.
Squamulae and halteres whitish. Chaetotaxy typical; all the bristles are black,
but the longer ones have yellowish ends; four scutellar; the short pubescence
is pale yellowish. Abdomen black, with the last segment reddish brown behind,
rather shining, with dark pubescence and black terminal bristles. Genitalia
reddish brown, shining, with short yellowish appendages beneath. Legs entirely
yellowish ; bristles of front femora and spurs of middle tibiae black. Wings
nearly the same as in Enderlein’s fig. R, p. 441, with the following differences :
(a) the first fuscous band is less developed and much shorter, ending near the upper
corner of the second basal cell; (6) the second fuscous band is longer, reaching
the hind border of the wing or nearly so ; (c) the third abbreviated band is narrower ;
(d) the apical hyaline spot is twice as broad, extending symmetrically above in

240 PROF. M. BEZZI.

the submarginal and below in the second posterior cell; it is distinctly whitish ;
(e) the hyaline indentation into the second posterior cell is narrower; (f) the hyaline
dot at the end of the second longitudinal vein and the hyaline streak in the middle
of the first posterior cell are faintly developed, or even indistinct.
Type 3, a single specimen from the Gold Coast, Accra, i. 1916 (Dr. J. W. S.
Macfie).
* x x * sk *k
The Ethiopian species referred by authors to Acidia, were provisionally placed
by me under the head of Philophylla; but after the removal of obnubila and
coloniarum to Celidodacus, there remain only ocellata (which belongs to a new genus)
and seychellarum, which on account of its eyes certainly cannot be ascribed to
Philophylla. This last genus seems therefore to be absent from the Ethiopian
fauna. The genera of this group must be disposed as follows, with a modification
of my table on pp. 218-219 of the first paper.
43(42). Prst. always present; two s. or.
44(45). Small cross-vein near or before the middle of the discoidal cell.
a(b). Wings of usual shape, with straight costa and long second posterior cell,
adorned with narrow vertical fuscous bands, which are often partly
broken into spots... : a Phorellia, R. D.
b(a). Wings towards the middle very Boas er the costa rounded outwardly
and with very short second posterior cell; they are adorned with a
broad rounded fuscous patch and with some curved dark bands
Taomyia, gen. nov. ©
45(44). Small cross-vein beyond the middle of the discoidal cell, and often very
near the hind cross-vein.
c(f). Eyes in profile narrow, much higher than broad.
d(e). Occipital lower swellings not developed; wings with yellow longitudinal
stripes, the lower one bifurcate at end; thorax with black spots
Notomma, gen. nov.
e(d). Occipital lower swellings rather developed ; wings with yellowish or fuscous,
partly confluent, cross bands; thorax not black-spotted
Philophylla, Rond.
f(c). Eyes in profile broad and rounded; if they are rather narrow, the
occipital lower swellings are not developed.
g(h). Wings of usual shape, with short stigma and distinct costal bristle ; oc. well
developed ; wings banded, at least on the apical half § Acrdra, R. D.
h(g). Wings rather narrow, with parallel sides, with elongate stigma and without
costal bristle ; oc. very thin or even wanting ; wings not banded, with
the fore half usually entirely infuscated .. Ocneros, O. G. Costa.

Taomyia, gen. nov.

The present new genus is nearly allied to Acedia and Philophylla, difiering from
both in the very different shape and venation of the wings, which recall that of the
Neotropical genus Parastenopa, Hendel, which however has no oc. and a differently
shaped pterostigma. Head in front view a little higher than broad, in side view

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 241

‘with rather narrow eyes and distinct lower occipital swellings. Frons as broad
as the eye, distinctly longer than broad, flat, with very short orbital plates ; lunula
broad. Face shorter than the frons, convex in the middle, with no prominent upper
mouth-border and with diverging antennal grooves; cheeks linear; jowls narrow.
Antennae inserted a little below the middle of the eyes, rather separated at base,
shorter than the face ; second joint less prominent above and not spinulose ; third
joint linear, twice as long as the basal ones, rounded at tip; arista quite bare.
Palpi not very broad, bristly ; proppscs short. Ocp. black and acute, less numerous ;
oc. very long and strong; one s. or. ; 2 or 37. or., only the middle pair being ise
the other very short; genal bretle well developed. Thorax elongated, about
twice as long as broad, with complete chaetotaxy, but the middle scp. are not
developed ; de. placed much behind the line of the a. sa.; one mpl. ; pt. rather
thin. Thoracic suture broadly interrupted in the middle. Scutellum bluntly
triangular, convex above, bare, with four bristles, those of the apical pair decussate
at the end. Abdomen oval, elongate, not broader than the thorax, with terminal
bristles; male genitalia prominent and appendiculate. Legs rather slender ;
front femora with 5 bristles below ; middle tibiae with one spur ; hind tibiae with
only 4-5 longer bristles on the middle of the outer side. Wings of usual shape
in the basal half, very broad and nearly semicircular in the apical half, broadly
infuscated, with hyaline stripes, surrounding a large rounded middle patch. Costa
very curved on its last portion ; costal bristle hardly distinguishable ; first vein
not specially curved upwards at the end, the stigma being longer than broad ; second
vein straight ; third vein bristly to the middle of its last portion, which is strongly
curved downwards towards the end, and thus very divergent from the second and
parallel with the last portion of the fourth ; submarginal cell very widened at end ;
discoidal cell very long and much broadened outwardly, the hind cross-vein being
perpendicular, arched outwardly, about three times as long as the basal cross-
vein of the discoidal cell, and about as long as the términal portion of the fourth
vein ; small cross-vein on or just beyond the middle of the discoidal cell ; second
posterior cell unusually short, higher than long ; lower angle of the anal cell drawn
out into an acute point, which is shorter than the second basal cell; axillary lobe
broad ; alula well developed and rounded.
—_ : the following new species.

It is very probable that Acidia (*%) ocellata, Lamb, 1914, from the Seychelles,
belongs to the present genus, having a similarly formed head and a very similar
wing pattern; but it has the small cross-vein placed beyond the middle of the
discoidal cell, and two s. or. on the frons.

Taomyia marshalli, sp. nov. (Pl. xvii, fig. 9).

A very distinct fly of comparatively large size, at once distinguishable on
account of its very characteristic wings.

¢. Length of body 7°5 mm. ; of wing 7 mm.; breadth of a wing 3°2 mm.

Head of a dark reddish-brown colour ; occiput shining, broadly blackened on the
sides, light yellowish near the eye-borders and below ; frons rather shining, with a
eas ocellar dot and an ill-defined blackish nes band in front above the
antennae ; lunula shining yellowish ; face shining, dark brown in the middle, more

242 PROF. M. BEZZI.

yellowish towards the antennal grooves and on the sides; cheeks whitish ; jowls
unspotted. Antennae entirely pale yellowish, the third joint a little infuscated
along the upper border ; arista pale yellowish at base. Palpi yellowish ; proboscis
dirty brownish. All the cephalic bristles are black and rather thin; the frontal
stripe seems to be bare on the middle. Thorax black on the back, but clothed
with greyish dust and with three narrow parallel longitudinal black stripes, those
on the dorsocentral lines extending about to the scutellum, while the middle one
is abbreviated behind the transverse suture. The short pubescence of the back
is black. The shoulders, the notopleural suture and the postalar calli are reddish
brown, like the pleurae, which are blackened on the middle of the meso- and sterno-
pleura, but are pale yellowish on the pteropleura, forming thus a broad perpendicular
stripe; hypopleura blackish and shining, like the postscutellum and the meso-
phragma. Scutellum entirely yellowish with a narrow black basal band. All
the thoracic bristles are black and rather thin. Squamulae dirty brownish ; halteres
pale yellowish. Abdomen black, shining, but with dark dust; pubescence and
bristles black ; genitalia yellow, black above and behind ; the unpaired and strong
appendage of the underside is similarly yellow. Legs pale yellowish, but the four
posterior femora above on the apical half, and all the tibiae except the tip, are more
or less blackened ; bristles black, those of the front femora very thin. Wings
whitish hyaline, with a very broad, dark brown pattern; an abbreviated stripe,
extending from the middle of the stigma to the third longitudinal vein is yellow.
Veins yellow, but darkened on the dark parts. Costal bristle very small. Extreme
base of the wing hyaline; an arcuate brown band begins along the costa, filling
up the whole of the first and the base of the second costal cell, and becoming gradually
broader passes over the middle of the second basal cell, fills up the whole of the
anal cell and is prolonged as an acute stripe along the sixth longitudinal vein, ending
with a point at the hind border of the wing. The axillary lobe and the alula are
quite hyaline. After this dark band there is a narrow, complete hyaline band,
going from the middle of the second costal cell, over the apex of the second basal
cell and over the base of the third posterior cell, to the hind border of the wing.
The extensive brown pattern of the broadened apical half of the wing is thus sharply
separated from the basal one ; it consists of a rounded, very broad patch, which
occupies the whole of the discoidal cell, and the greater part of the submarginal
and first basal cells, and the entire third posterior cell. At the fore border there
is the yellow stigmatic indentation, and after this a broader triangular oblique
hyaline indentation, which reaches with its obtuse poimt the fourth vein, entering
the end of the first basal cell ; from this indentation departs a rather narrow hyaline
arcuate band, which, running along the fore border, from which it is separated
by a narrower brown costal band, ends in the upper angle of the second posterior
«ell. The yellow stigma is blackish brown at base and on the lower spical half.

Type g, a single specimen from Natal, Verulam, vil. 1897, collected by Dr. G.
A. K. Marshall, in whose honour this fine insect is named.

Notomma, gen. nov.
The present genus is erected here for a species which has a wing pattern recalling
that of Xanthorrhachista, from which it differs in general appearance and in the

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 243

position of the small cross-vein, and Craspedozantha, from which it differs in the
bristly third longitudinal vein and in the position of the dc. bristles. In Prof.
Hendel’s table it runs down to Neanomoea, an undescribed Oriental genus from
Formosa. |

Trypeta jucunda, Loew, seems to be an allied form, but differs in the venation
and in the shape of head, belonging probably to some peculiar, as yet undescribed
genus. |

Head a little narrower than the thorax, in front view about as broad as high,
in profile rather flat, the eyes being narrow, with the vertical diameter about twice
as long as the horizontal. Occiput flat, even a little convex on the upper half,
destitute of distinct lower swellings. Frons gradually narrowed from the vertex
to the antennae, flat, narrower than one eye in front, about one and a half times
as long as broad, in profile a little prominent above the base of the antennae. Lunula
broad. Eyes greenish, with a purplish horizontal band just above the middle.
Face flat, with a distinct, broad and rounded middle keel and with shallow antennal
grooves ; mouth-border prominent; cheeks linear; jowls narrower than the
breadth of the third antennal jot. Antennae inserted a little above the middle
of the eyes ; second joint short, small, bristly above ; third joint broad, pubescent,
parallel-sided, rounded at tip, not reaching the mouth, less than twice as long as:
broad ; arista bare. Palpi broad, bristiy ; proboscis short and thick. Cephalic:
chaetotaxy complete and well developed ; ocp. acute, black ; put. thin and parallel ;:
two pairs of vt., the inner ones not much longer ; oc. robust but not long; two
s. or. and three 7. or. ; genal bristle strong, like the bristle on the lower part of the
occiput. Thorax robust, flat, subquadrate, with complete chaetotaxy ; the middle
scp. are approximated and as long as the outer ones; a. sa. strong; dc. placed
much behind the line of the a. sa; prsct. approximated ; two mpl. ; pt. about as
strong as the st. Scutellum convex above somewhat Ceratitis-like in shape but
not trilobate, with four strong bristles, the apical ones diverging. Abdomen in
the middle broader than the thorax, narrowed at base and at the end; it has five
segments, the last with bristles at the sides and behind; ovipositor elongate
triangular, rather swollen. Legs short and stout; front femora with a complete
row of strong bristles beneath ; middle tibiae with one spur; hind tibiae with a
complete row at the outer side; claws curved; middle tibiae without bristles
on the hind side. Wings elongated and proportionally narrow ; costa not ciliated:
and without costal bristle. They are adorned with two longitudinal yellow bands,
one along the fore border and the other along the middle; this last is bifurcate.
into a recurrent band across the hind cross-vein and along the hind border of the.
wing. Auxiliary vein strong; first longitudinal vein long, ending beyond the.
middle of the wing but always before the small cross-vein; stigma long, in the
middle about three times as long as broad; second and third veins rather:
approximated, slightly sinuous, the third bristly throughout its whole length ;
the last portion of the fourth vein is slightly sinuous and nearly parallel with the
last portion of the third, which ends nearer the tip of the wing, but the first posterior
cell is much broader at end than at base; small cross-vein placed exceedingly
near the end of the discoidal cell, its distance from the hind one being about equal
to its own length; hind cross-vein curved and oblique outwardly, its lower end:
(C637) Cc

244 PROF. M. BEZZI.

being very near the hind border of the wing; discoidal cell very broad at end ;

second posterior cell very short, above only a little longer than broad at base;

second basal cell rather broad and short; anal cell with the lower angle acute

and prolonged, but not longer than the second basal cell; sixth vein reaching

the hind border of the wing. Axillary lobe broad and short ; alula rounded.
Type: the following new species.

Notomma bioculatum, sp. nov. (Pl. xvu, fig. 10).

A reddish brown species, with two deep black spots on the middle of the back
of the mesonotum.

©. Length of body 6°5 mm. ; of wing 6°5 mm. ; of ovipositor 1°5 mm.

Head entirely yellowish, with only a small black dot on the ocelli. Occiput
shining, on the lower part with faint pale dust and with some rather long and
soft pale hairs. Frons opaque, with shining ocellar and vertical plates and some
scattered dark hairs on the middle. Face with whitish dust. Antennae entirely
yellowish, even the arista; palpi and proboscis yellowish, the latter with dirty
blackish flaps. All the cephalic bristles are black. Thorax reddish brown, shining
on the back; it is clothed with faint greyish dust and short black pubescence, disposed
in regular approximate longitudinal rows; the two very striking, deep black,
opaque spots are of oval shape and placed on the dorsocentral lines, just behind
the suture ; the pleurae are paler and more shining, and clothed with rather long,
scattered black hairs. Scutellum coloured and pubescent like the back of the
mesonotum ; at hind border it has three ill-defined, shining brown spots, a smaller
one between the apical pair of bristles, and a larger one on each side between the
basal and apical pairs of bristles. All the bristles are black. Squamulae reddish,
with a dark fringe; halteres reddish yellow. Postscutellum reddish, with two
rounded black spots ; mesophragma entirely of a shining reddish colour. Abdomen
shining reddish, with black pubescence and black bristles ; the hind border of the
segments is more lightly coloured ; the venter likewise ; ovipositor shining reddish.
Legs entirely reddish, with paler tibiae and tarsi ; bristles and spurs black. Wings
shining and iridescent with yellowish veins. The first yellowish band is complete
from the base to the tip of the wing, going throughout along the costa, but leaving
two hyaline streaks in the marginal and submarginal cells, just before and behind
the end of the second longitudinal vein ; m the middle this band does not extend
below the second vein, while at base it reaches the fourth vein just above the
second basal cell, and at the end it extends below the third vein to the middle of
the end of the second posterior cell; at the end there is a blackish infuscation,
on both sides of the apex of the third longitudinal vein; in the middle of the
marginal cell there is a black rounded spot, just beyond the entirely yellow stigma.
The second longitudinal band is of a paler yellowish colour above, but infuscated
below and at the end ; it begins on the anal cell and on the lower part of the second
basal cell, and goes across the base of the discoidal and the apex of the first basal
cell to the small cross-vein ; it there becomes narrower and is bent downwards,
going along the last portion of the fourth vein to the costa, which is reached just
below the tip of the fourth vein itself. From this band arises another, which
forms a border along the hind cross-vein and extends beyond it as a border along

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS, 245

the hind border of the wing to the end of the sixth vein; this recurrent band is
yellowish interiorly and fuscous exteriorly. The hyaline band between these two
yellow longitudinal bands is nearly equal in breadth throughout its whole length
and is bent just above the small cross-vein, entering there as a small streak into
the marginal cell, below and before the rounded black spot of the fore border.
The hyaline band between the central yellow band and the recurrent posterior
band ends with a rounded tip in the base of the first posterior cell just behind the
small cross-vein ; at base this band is united with the hyaline axillary lobe. The
second basal cell is almost entirely hyaline, while the discoidal and third posterior
cells are almost entirely infuscated except the broad hyaline band mentioned

above.
Type 9, a single specimen from the Gold Coast, Aburi, 1912-13 (W. H. Patterson).

Acidia, Rob.-Desv., 1830.

Three species are provisionally placed in the present genus, those here described
as new being very like the European A. caesio; they may be distinguished as
follows :—

1(4). Discoidal and third posterior cells with hyaline streaks or spots.

2(3). Thorax and femora black ; the apical dark bands of the wing broadly united
with the basal patch, the second hyaline indentation of the fore border
being quite wanting; discoidal and third posterior cells each with a
hyaline rounded spot *: P _ seychellensis, Lamb.

3(2). Thorax and femora entirely yellow ; iis apical dark bands quite separated
from the basal pattern by the second indentation, which forms a complete
hyaline cross-band ; discoidal and third Laos cells, each with a
hyaline streak .. = fossatacformis, sp. n.

4(1). The above-named cells at Aivaley fastens ana deseo of hyaline parts;
body and femora entirely black; wings with two hyaline indentations
at fore border .. as My. £4 af A homogenea, sp. n.

Acidia fossataeformis, sp. nov. (Pl. xvii, fig. 1).

An elegant species very like the Oriental A. fossata, F., but at once distinguished
by its entirely vellow thorax and legs.

g 2. Length of body 45-5 mm.; of wing 4-4'5 mm.; of ovipositor 0°8 mm.

Head entirely yellow and unspotted, except the very small black ocellar dot.
Occiput flat on the upper half, even a little concave above the median tubercle,
convex but not very prominent on the lower half, without distinct swellings.
Frons reddish yellow, opaque, with scattered dark hairs and moderately shining
vertical plates ; it is parallel-sided, narrower than the eye, 24 times as long as
broad ; lunula broad and short, semicircular, whitish, with brown upper border.
In profile the frons is only a little prominent above the antennae; the eyes are
rather rounded, their vertical diameter being only a little longer than the horizontal.
Face flat and long, not prominent at mouth-border, with a broad middle keel and
not very deep antennal grooves; cheeks linear, whitish ; jowls whitish, unspotted,
as broad as the breadth of the third antennal joint. Antennae entirely yellow,
inserted at the middle of the eyes, short, extending only to the middle of the face ;
second joint swollen and prominent, with a short bristle above ; third joint rather
(637) C2

246 PROF. M. BEZZI.

small, short, attenuated apically but with rounded tip; arista with rather long
pubescence, but the breadth of feathering narrower than the breadth of the third
antennal joimt. Palpi and proboscis short, yellowish, pale-haired. All the
cephalic bristles are black and rather thin; ocp. well developed and acute; pvt.
parallel ; outer vt. half as long as the inner ones; oc. short and thin; two s. or.
and three 2. or. ; genal bristle well developed. Thorax reddish yellow on the back,
very little shining, with faint whitish dust and short dark pubescence ; the pleurae
are paler and more shining, with scattered longer dark hairs. All the bristles.
are black; chaetotaxy complete ; scp. well developed, the middle ones approxi-
mated ; dc. placed much behind the line of the a. sa. and near to the prsct.; only
one strong mpl.; pt. as strong as the st. Scutellum triangular, as long as broad,
‘flat above, with lateral keels; it is coloured like the back of the mesonotum, but
is more bare and shining; it has four black bristles, the apical ones being shorter
and decussate. Squamulae and halteres pale yellowish ; the former with a short
thin white fringe. Postscutellum and mesophragma entirely reddish yellow and
shining. Abdomen of usual shape, as broad as the thorax towards the middle;
it is shining reddish yellow, but sometimes very infuscated in the male and even
brownish or blackish; the pubescence and terminal bristles are black; venter
more yellowish, with faint pale dust. Male genitalia shaped as in fossata, with a
pair of long and stout appendages, the basal one reddish, while the apical one is
black, and longer and stouter. Ovipositor with the basal segment broadly
triangular, shining black, the apical segment narrow and dark reddish; the contrast
between the reddish abdomen and the black ovipositor is very striking. Legs
rather slender, entirely yellowish ; front femora with 3-4 long black bristles beneath
at the end; middle tibiae without bristles, with one black spur at the end ; hind
tibiae with no distinct row. Wings shining and iridescent, with non-ciliated costa
and with a short but distinct costal bristle. The veins are yellowish at base and
on the hyaline parts, black on the rest; the second longitudinal vein is a little
sinuous ; the third vein is bristly throughout its whole length; the second, third
and fourth veins are equally divergent towards the apex, ending at equal distances ;
the last portion of the fourth vein is straight ; the small cross-vein is placed beyond
the middle of the discoidal cell; the hind cross-vein is straight, perpendicular,
a little longer than its distance from the small one; inferior angle of the anal cell
produced, long and narrow, but not longer than the second basal cell; sixth
longitudinal vein complete. The base of the wing is yellowish in the basal and
costal cells, even the brown stigma being yellow at base. The pattern is blackish
brown and about the same as in fossata (vide Bezzi, Mem. Ind. Mus., ui, 1913,
pl. ix, fig. 48), which however has the wings distinctly broader.

Type 3, type 2, and some additional specimens of both sexes from Nyasaland,
Chiromo, Ruo R., 2.1.1917 (R. C. Wood).

Acidia homogenea, sp. nov.

A black species, with a wing pattern very like that of the European A. caesio,
but at once distinguished from it and from the other species by the discoidal and
third posterior cells being completely infuscated, without hyaline spots or streaks.

Q. Length of body 4 mm. ; of wing 4 mm.; of ovipositor 0°6 mm.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 247

Entirely shining black, the head, antennae, tibiae and tarsi in part yellowish.
or whitish. Head as broad as the thorax, rounded in front view and about as high
as broad ; it is rather narrow in lateral view, the eyes being considerably higher
than broad. Occiput entirely black, shining above, grey-dusted below, without
lower swellings ; it is narrowly reddish on its lower portion. Frons narrow and
long, parallel-sided, narrower than the eye, twice as long as broad ; it is entirely
reddish, opaque, with shining white sides, and with a small blackish ocellar dot, the
narrow and short vertical plates are shining; lunula broad, semicircular, whitish.
Antennae inserted near the middle of the eyes ; the two basal joints are yellowish ;
third joint wanting in the type. Face only a little shorter than the frons, as broad
as it above, but twice as broad below; it is flat, with shallow antennal grooves,
and no prominent mouth-border; it is blackish, with a shining white dust,
yellowish only on the flat middle keel; cheeks linear, hardly distinguishable ;

jowls very narrow, reddish, unspotted. Palpi and proboscis dirty yellowish,
the former dilated at the end, the latter short and thick. Cephalic bristles missing
in the type, but judging by the insertion points there are two s. or. and three equally
distant 7. or. Thorax entirely shining black, not dusted on the back; scutellum
similar, triangular, flat above ; pubescence and bristles abraded in the type, but the
mpl. and st. are present on one side and are black; the scutellum has certainly
four bristles; the dc. are placed very near the suture. Squamulae brownish ;
halteres black. Abdomen entirely shining black, with short black pubescence
and black bristles at the end; ovipositor shining black, triangular, flat, pro-
portionally short; apical segment reddish; venter black. Legs with shining
black femora ; front tibiae entirely, four posterior ones on the apical half, whitish,
like the whole of the tarsi. Wings rather broad, with the membrane shining and
iridescent, chiefly on the hyaline parts; costa and veins black, but whitish on the
hyaline parts. First longitudinal vein short, the stigma being about as long as broad
at base; second vein straight; third vein bristly to the small cross-vein, bent
downwards beyond middle of its last portion ; last portion of the fourth vein curved
and slightly diverging from the third; small cross-vein beyond the middle of the
discoidal cell; hind cross-vein straight and perpendicular, a little longer than
its distance from the small one; lower angle of the anal cell narrow and shortly
produced. The blackish pattern is very like that of A. caesio, or of the Formosan
“ Trypeta” superflucta, End. (Zool. Jahrb. xxxi, 1911, p. 428, fig. J) ; but it differs in
having more than the basal half of the wing entirely blackish, without any hyaline
indentation of the fore border before the stigma, and without any hyaline spot or
streak in the discoidal or third posterior cells ; only the axillary lobe is hyaline.
The stigma is deeper black than the surrounding parts. The first hyaline indenta-
tion of the fore border beyond the stigma is as in caesio ; the second likewise, but
ending with a narrow point on the fourth longitudinal vein, without entering the
discoidal cell. The two hyaline apical indentations of the hind border are as in
caesio, the black streak between them being complete and not interrupted at

base (difierimg thus from that of fossata and fossataeformis, in which it is
isolated).

Type Q, a single specimen from Nyasaland, Port Herald, iv.-vi. 1913 (Dr. J. E.
S. Old).

248 PROF. M. BEZZI.

Oeneros, O. G. Costa, 1844.

In the typical species of the present genus the dc. are placed near the suture,
while in the known Ethiopian species they are placed much behind. It seems
thus probable that the latter species are not congeneric with the palaearctic ones ;
but I have not erected a special genus for them, because they may perhaps be
better placed in the following genus Ocnerioxa, which differs only in the non-
bristly third longitudinal vein.

The genus seems to be abundantly represented in the Ethiopian fauna; I have
before me two additional new species, which belong to the group of O. sinuatus.
This group is not only distinguished by its colour characters, but also
by the slender body, the longer antennae (which are nearly extended to the mouth-
border), the distinctly pubescent arista, the thin and prevalently yellow
macrochaetae of the head, thorax and scutellum, the longer pterostigma, and
by the last portion of fourth longitudinal vein being quite straight. To the
characters of the genus must be added that the 2. or. are accompanied by
additional bristly hairs.

The known species may be distinguished, by a modification of the table on p. 250
of my first paper, as follows :—

1(2). Scutellum entirely yellow ; discoidal cell wholly infuscated, without fe
spots ; stigma as long as or longer than the second costal cell; antennae
about as long as the face, with a shortly pubescent arista; macrochaetae |
mostly yellow.

a(d). Base of stigma narrowly hyaline, at least near the costa; there is a small
hyaline dot in the base of the first posterior cell, just above the hind cross-
vein.

b(c). Hyaline spot in the marginal cell beyond the stigma triangular and extended
to the second longitudinal vein; the fuscous portion of the wing has below
only a median broad projection, which extends about to the hind border
in the hyaline part of the wing ee . J. sinuatus, Loew.

c(b). Hyaline spot of the marginal cell in the shape of an elongate stripe near the
costa, which does not extend below beyond the middle of the cell; the
fuscous portion is wavy below, with 4-5 pro} ections, which are all far from
the hind border of wing of : He undatus, sp. n.

d(a). Base of stigma with a broad hyaline ale no palit dot in the base of the
first posterior cell; fuscous part of wing ending below with one acute
median projeeinon: which reaches the hind border of wing along the
fifth vein. 7 se .. bigemmatus, sp. n.

2(1). Scutellum Bevedie plaice in ate anianiloe discoidal cell with hyaline spots;
stigma shorter than the second costal cell; antennae much shorter than
the face, with bare arista ; macrochatae black mundus and excellens.

Ocneros undatus, sp. nov.

A slender and pale-coloured species of smaller size, very distinct on account of
the lower border of the fuscous part of the wing being wavy and remote from the
hind border of the wing.

Q. Length of body 4mm. ; of ovipositor 1 mm. ; of wing 4°5 mm.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 249

Head entirely yellow and opaque; occiput with a narrow black stripe on each
side, extending from the neck to the lower third of the eyes; frons narrower than
_the eye and longer than broad, unicolorous, with a very small black ocellar dot.
The face and unspotted jowls are yellowish, while the occipital lower swellings
are more lightly coloured and surrounded above with the black stripe; upper
mouth-border prominent. Antennae entirely pale yellowish, only a little shorter
than the face, with the third jomt rather broad, and with a shortly pubescent
arista. Palpi and proboscis pale yellowish. All the cephalic bristles are pale
yellowish, only the short outer vt., the acute and numerous ocp. and the 2. or. being
black ; there are no distinct ocellar bristles ; there are only two pairs of longer 7. or.,
but on the same line are some black, much shorter, bristly hairs. The lower occipital
hairs are whitish; the bristly hairs near the borders of the mouth are yellowish.
Thorax entirely yellowish, less shining, with pale short golden-yellowish pubescence ;
on each side there are two black longitudinal stripes, one extending from the
shoulders along the notopleural line to the root of the wing, and continuous with
the black postalar calli and the very glistening black mesophragma; the other
is parallel with the former but ends at the suture, being continued beyond it only
by a short ill-defined darkish stripe; the occipital black stripe is contmuous with
the notopleural one. The pleurae and the breast are entirely whitish yellow,
unspotted, the contrast between the black and the whitish parts being very striking,
when the thorax is viewed from the side. Hypopleural spot whitish. Scutellum
of a light yellowish colour, like the posterior part of the back of the mesonotum,
but a little more shining. Postscutellum shining black. All the macrochaetae
are yellowish, those of the pleurae being lighter; one mpl.; pt. much weaker
than the st. ; even the scp. are black. Squamulae blackish, with a long dark fringe ;
halteres pale yellowish. Abdomen shining yellow on the basal half, blackened
on the apical half, the last three segments having black lateral borders; venter
whitish ; hairs and bristles yellowish. Ovipositor flattened, broadly obtuse at
the end, shining reddish with pale pubescence. Legs entirely whitish yellow, those
of the front pair wanting in the type; apical spur of the middle tibiae black ;
hairs of the posterior row of the hind tibiae brownish. The wings are almost
wholly and evenly infuscated ; the lower border of the fuscous part ends about
on a line extending from the point of the anal cell to the lower anterior corner of
the discoidal and the upper third of the second posterior cell ; but this line is wavy,
having some rounded projections, three of which are in the second and three in the
third posterior cell. There is besides a very small subhyaline rounded dot in the
basal part of the first posterior cell, a little before the hind cross-vein and closer
to the fourth than to the third longitudinal vein. The stigma is as long as the
second costal cell and is black, with a narrow short subhyaline streak near the
base at costa, being as broad as a third only of the stigma itself. The hyaline streak
at fore border beyond the stigma is shorter than the stigma itself, and does not
extend below beyond the middle of the marginal cell. The small alula is entirely
brown, while the axillary lobe is entirely hyaline. The third longitudinal vein
has 2-3 bristles beyond the small cross-vein. The last portions of the third and
fourth longitudinal veins are quite straight and parallel, or only a little divergent.

Type 9, a single specimen from Nyasaland, Mt. Mlanje, 2.11.1913 (S. A. Neave).

250 PROF. M. BEZZI.

Ocneros bigemmatus, sp. nov. (Pl. xviu, fig. 2).

Very like the preceding species, but distinguished by its larger size, and the
different pattern of the thorax and wings. 7

©. Length of body 5°5 mm.; of ovipositor 1mm. ; of wings 6 mm.

Head, its appendages and bristles exactly as in the preceding species; the two
occipital black spots are broader, shaped as a triangular patch, the outer corner
of which does not reach the border of the eyes ; there are besides two approximated
black stripes above the neck, directed towards the vertex, wanting in O. undatus.
Thorax and scutellum, and their pubescence and bristles, exactly as in the preceding
species ; there is also an equally narrow, black, notopleural stripe in continuation
of the occipital one, and therefore there is the same striking contrast with the
whitish pleurae ; the interior stripe is however much broader and ends at the suture,
but is continued beyond it with a broad stripe on the dorsocentral lines, extending
nearly to the lateral edges of the scutellum. The mesophragma is shining reddish,
brownish fuscous on the sides, like the postscutellum. Halteres and squamulae
as in O. undatus. Abdomen more elongate, shining reddish; the three terminal
segments more broadly black on the sides and middle; bristles black; venter
darker; ovipositor shining black, with blackish pubescence. Legs as in the
preceding ; front femora with only three rather thin and yellowish bristles on the
apical half beneath. Wings as in O. undatus, but with the following differences :
(a) at base of the stigma there is a broad whitish-hyaline spot, extending below
to the first longitudinal vein; (b) the whitish hyaline spot beyond the stigma has
the shape of an oblique band, which goes from the end of the first vein across the
middle of the marginal cell to the second vein ; (c) there is no trace of the hyaline
basal dot of the first posterior cell ; (d) the lower border of the fuscous part is more
irregular, showing a blackish tooth in the middle of the upper border of the third
posterior cell, and another more elongate tooth on the outer inferior angle of the
discoidal cell, reaching the hind border of the wing along the fifth longitudinal
vein ; in the second posterior cell the border of the fuscous part forms a straight
line, which has no indentations or projections (or only a little so beyond the fourth
vein), the cell being thus mainly hyaline; along the middle of the anal cell there
is also a subhyaline longitudinal stripe, which is wanting in the preceding
species.

Type 9, a single specimen from British Kast Africa, Embu, 12.xu1.1913 (G. H. J.
Orde-Browne).

Ocnerioxa, Speiser, 1915.

The species of the present genus show a great resemblance to those of the first
group (sinuatus, undatus and bigemmatus) of Ocneros; they have the same
appearance and the same colour pattern of body and wings, being probably con-
generic. They can be distinguished only by the third longitudinal vein being
quite bare, or scarcely setigerous before the small cross-vein only ; and also by
the bare, non-pubescent arista and by the less developed orbital hairs between .
the fronto-orbital bristles.

In the new material there is the following undescribed form.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 251

Ocnerioxa discreta, sp. nov.

Very like O. wood: from Nyasaland, but more slender, smaller and distinguished

by some minor details in the coloration of the head, abdomen and wings.
~. Length of body 4 mm.; of wing 45 mm.

The body is narrower than in woodi, the thorax measuring about 1 mm. only
in breadth, while in woodi it is 15 mm. broad. Besides there are the following
differences : (a) the occiput has only the two lateral black stripes, the two middle
ones being quite absent; (6) the frons is entirely whitish, even above the roots
of antennae, the black band of woods being quite wanting, and thus the white
lunula is not conspicuous ; (c) the third antennal joint is not at all infuscated ;
(d) the black transverse band at the upper mouth-border is reduced to two small
spots, one on each side of the mouth-edges, and these spots are moreover dull ;
(e) the jowls are entirely whitish, quite destitute of the fuscous subocular spot ;
(f) the black notopleural stripe is narrower, and there is a faint indication of an
anterior median fuscous stripe, and the mesophragma is dark reddish in the middle ;
(g) the abdomen is not at all greenish, and has the fourth segment yellowish, with
narrow black sides and hind border; (4) the wing pattern is identical, but the
fuscous tooth below the outer lower angle of the discoidal cell along the fifth
longitudinal vein is much shortened, not being extended to the hind border of
the wing.

These small differences, together with the very different habitat, may be con-
sidered of specific value.

Type 3, a single specimen from N. Nigeria, Zungeru, 5.xi.1910 (Dr. J. W.
Scott Macfie).

Allotrypes, gen. nov.

This is a rather aberrant genus, which on account of its bare third longitudinal
vein, black and strong ocp. and its general appearance, comes in the vicinity of
Acura, being eee very distinct from it in the chaetotactic caer as well
as in the wing pattern.

Head a little broader than the thorax, but in front view about as broad as high.
Occiput distinctly concave above the neck, and less prominent below, with the
lateral lower swellings but little developed. Frons flat or even a little concave
before the lunula, with about parallel sides, as broad as one eye, only a little longer
than broad. In profile the frons is not at all prominent, not even above the
antennae; the eyes are rounded, with the vertical diameter only a little longer.
Lunula small. Face long, only a little narrower but distinctly a little longer than
the frons ; it is about parallel-sided, flat, with no prominent mouth-border, and with
shallow but long antennal grooves. Cheeks linear; jowls very narrow, narrower
than the third antennal joint. Antennae inserted distinctly above the middle
of the eyes; they are very short, not extending beyond the middle of the face ;
second joint not prominent, shortly setulose above, destitute of long bristles ;
third joint pubescent, attenuate at end but obtuse, 14 times as long as broad ;
arista shortly pubescent. Palpi and proboscis short, retracted into the oral cavity.
Cephalic bristles very strong and peculiarly curved; no distinct put.; ocp. numerous,

252, PROF. M. BEZZI.

long, unusually strong, black, acute at end ; inner vt. very long, more than twice
as long as the outer ones, strong, curved backwards beyond their middle; one
s. or. and one 2. or., equally strong and long and both curved backwards ; the
single 2. or. is placed much in front, near the lunula ; oc. strong, but not very long ;
on the sides of frons there are thus only two pairs of bristles, one near the base
and one near the end; genal bristle less strong. Thorax proportionally narrow,
less convex above, shining on the back, with the suture broadly interrupted. The:
bristles are strong and the chaetotaxy is complete, except the scp., which are not
distinct ; de. placed only a little behind the line of the a. sa.; prsct. only a little
closer together than the dc. ; one mpl. only ; pt. and st. very strong, curved upwards.
Scutellum triangular, as long as broad, flat above, with distinct lateral keels ;
it has four bristles, the basal ones very strong and long, the apical shorter, thinner
and decussate. Abdomen narrow and elongate, not narrowed at base, with about.
parallel sides throughout its whole length; it has five segments, the last one
with less developed bristles ; ovipositor flattened, triangular, about as long as the
last three abdominal segments together. Legs proportionally slender and bare ;
front femora with two long bristles below on the apical fourth ; middle tibiae with
a single spur; hind tibiae destitute of a distinct row of bristly hairs on the outer
side. Wings of usual shape, but rather narrow at base, the axillary lobe being
little developed and the alula very narrow. They have a characteristic and rather
aberrant pattern (for the group), consisting in a broad infuscation extending almost:
over the whole wing, with the fore border broadly yellowish-hyaline on a little
more than the basal half; there are no hyaline spots, and no hyaline indentations.
Costa not ciliated ; with a short but strong costal bristle. Auxiliary vein typical,
distinct throughout, with diffuse end; stigma short, about as long as_ broad ;
first vein hairy and rather short, ending much before the small cross-vein, its
distance from the end of the second vein being about twice as long as that of the
second from the third vein. Second vein in its last half strongly diverging from
the third, which is quite straight throughout and quite bare. Fourth vein straight
and parallel with the third ; small cross-vein placed beyond the middle of the long
discoidal cell; hind cross-vein straight, perpendicular and rather short, being
shorter than, or nearly as long as its distance from the small one; last portion
of the fifth vein proportionally long and strong; sixth vein complete, but colour-
less on more than its apical half. Second basal cell but little developed ; anal
cell broader but shorter than the second basal cell, with an acute but not or very
little produced lower angle.

Type: the following new species.

Allotrypes brevicornis, sp. nov. (Pl. xvii, fig. 3).

A narrow elongate species not unlike some forms of Aciwra, but very distinct
on account of its wholly testaceous thorax, and its pecular wing pattern.

Q. Length of body 4°5 mm.; of ovipositor 15 mm. ; of wing 4°5 mm.

Head reddish above, pale yellowish below; occiput with two broad shining
black spots of triangular shape, with the base on the eye-border and the narrow
apex reaching the insertion of the neck ; these two spots are separated by a yellowish
triangle which has the vertex at the neck and the base at the vertical edge.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 253:

Ocellar dot smal! and black. Frons reddish yellow, shining, with sparse short hairs
in the middle. Antennae entirely reddish yellow, with the third joint paler. Face
pale yellowish, shining in the middle below; the cheeks and jowls are clothed
with a faint whitish dust. Palpi and proboscis pale yellowish or whitish. All
the cephalic bristles are of a deep black colour. Thorax and scutellum entirely
shining testaceous ; the back has very short scattered black hairs and no distinct
dust; humeri and pleurae paler than the back; all the bristles are black.
Mesophragma shining, more or less infuscated in the middle, or even brownish.
Halteres and squamulae reddish yellow. Abdomen shining testaceous, distinctly
darker than the back of the mesonotum, even blackish brown in the middle of
the segments and behind; pubescence and bristles black; ovipositor shining
black on the basal half, dark reddish-brown on the apical half, but sometimes
entirely black. Legs entirely reddish yellow, with the front tibiae and the front
tarsi infuscated ; bristles and spurs black. Wings shining, but slightly iridescent ;
veins and costa yellowish on the light parts, and brownish on the dark ones.
Stigma yellowish. The yellowish-hyaline basal part of the pattern extends over
the costal, marginal and submarginal cells to the third vein; the marginal cell
however has its end brown, while the submarginal cell is brown on its apical half ;
this clear basomarginal band is yellowish in the middle, being hyaline at the two
ends only ; sometimes there is a distinct, but not sharply defined, rounded dark
spot in the middle of the hyaline part of the marginal cell near the costa (this
spot is wanting in the type). The rest of the wing is uniformly brown, only the
posterior border of the third posterior cell, of the axillary lobe and of the alula
being hyaline; the second posterior cell is somewhat more clear on the middle
of its hind border.

Type 9, a single specimen from Durban, Umbilo, 14.vi.1914 (ZL. Bevis); I
have seen an additional specimen from Zululand in the collection of the South
African Museum, Capetown.

Aciura, Rob. Desv., 1830.

Of the present “ genus ” there are in the new material before me three species,
one of which is not yet described and differs from all the others in having all the
occipital bristles black, like binaria, Loew, which is however a very different
species,

Aciura perpicillaris, sp. nov. (Pl. xviu, fig. 4).

A small elongate slender species, which is allied to angusta in having the axillary
cell rudimentary, but differs from it and from all the others of the same group in
having all the occipital bristles black, and the discoidal cell entirely black with a
single hyaline rounded spot near its extreme base.

§ 2. Length of body 3-3°2 mm.; of ovipositor 15 mm.; of wing 3‘6-4 mm.

Head entirely shining black and very glistening on the occiput, vertex and
orbital plates; the very narrow orbits and the linear cheeks are clothed with
greyish or whitish dust; the very narrow jowls are dark reddish brown, like the
anterior part of the frontal stripe. Lunula very small. Antennae inserted below
the middle of the eyes, entirely black ; third joint rounded at tip and grey-dusted ;

254 PROF. M. BEZZI.

arista long and with rather long pubescence, its feathering being about as broad
as the breadth of the third antennal joint. Palpi whitish; proboscis dirty brown.
All the cephalic bristles are black, even those of the occipital row; two s. or.,
the basal one much smaller; three 7. or., the apical ones much smaller; oc. as
developed as the basal pair of the s. or. ; occipital hairs sparse and black. Thorax
narrow and elongate, shining black, more glistening on the sides and on the pleurae ;
the short and very sparse pubescence of the back is black; all the bristles are
black; dc. placed before the line of the a. sa.; one mpl.; pt. weaker than the
st. Scutellum shining black, with only the b. scf., which are very long and divergent.
Postscutellum and mesophragma shining black like the hypopleural callosities ;
halteres black; squamulae dark and with a dark fringe. Abdomen narrow and
elongate, narrower than but about as long as the thorax ; it is shining black, with
sparse black pubescence and short black bristles; in the male it is obtuse at the
end, the genitalia being shining black and placed inferiorly at the side; in the
female the ovipositor is as long as the abdomen, glistening black and with black
pubescence, broad at base and acute at end, depressed ; the apical segment is dark
reddish-brown at base. Venter shining black, with a faint dark grey dust. Legs
rather long and slender; they have dark brown coxae and reddish trochanters ;
the femora are shining black, but narrowly brownish on the sides ; the four anterior
tibiae are yellow like their tarsi; the hind tibiae are broadly darkened at base,
more broadly in the female, and the tarsi are dark yellowish; front femora with
2-3 thin black bristles beneath on the apical half; hind tibiae destitute of a distinct
row of bristles on the outer side. Wings very narrow and elongate, with rudimentary
axillary lobe and alula, but appearing not so cuneate as in angusta, because they
are not dilated distally and have a much narrower third posterior cell. The veins
are black, but the costa and the first longitudinal are yellow in the two hyaline
identations of the fore border; costal bristle long and strong. The wing pattern
is like that of angusta, differing chiefly in the presence of two hyaline central spots,
and in the indentations of the hind border being more spot-like. The base is
entirely hyaline, without the obtuse black band of angusta, or only with a faint
indication of it below the humeral cross-vein. The stigma is entirely black and
about as long as the black apical part of the second costalcell. The two hyaline
indentations of the fore border are broader, more regularly triangular and of equal
size; the black band between them is broadly triangular and disposed perpendicu-
larly to the costa; the tips of the hyaline indentations are placed exactly on the
third vein and are obtuse, but at the upper end of the small cross-vein there is a
very narrow short hyaline streak, which is separated from the vertex of the second
hyaline indentation. A little before the middle of the first posterior cell there is
a rather broad, rounded or shortly oval, hyaline spot as in the following species.
The discoidal cell is entirely black, but near its base there is a rounded hyaline
dot, which is smaller than the above-named one. The three hyaline indentations
of the hind border are very different from those of angusta, being much shorter
and not produced into the discoidal cell ; the two indentations of the third posterior
cell have the shape of two rounded spots; the indentation of the second posterior
cell is twice as long as the other, but is broad and not stripe-like. The axillary
lobe and the alula are hyaline. The third posterior cell is much narrower than

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 255

in angusta, while the axillary cell is equally narrow; the hind cross-vein is short,
nearly straight and perpendicular, the lower outer angle of the discoidal cell being
thus not acute and not produced ; lower angle of the anal cell not acute.

Type 3g, a single specimen from British East Africa, Embu, 11.11.1913
(G. H. J. Orde-Browne); type 9, a single specimen from Gold Coast, Aburi,
27.x1.1914 (Dr. J. W. Scott Macfie).

Aciura oborinia, Walker (1849). (Pl. xvi, fig. 5).

A comparatively large species, like the following one, but at once distinguished
by the hyaline basal spot of the third posterior cell.

I assume the present species to be that of Walker, because it corresponds in the
size and answers very well to the description of the wing pattern; but Walker
says decidedly that the bristles at hind border of head are black, while in the present
specimens they are whitish yellow, like in the following and in the allied species.
‘In this character the species agrees better with the preceding one (persprcillaris),
which has also the hyaline spot of the first posterior cell; but I think that
persprcillaris cannot be Walker’s species, beg much smaller and having a very
different wing pattern.

I will give here a short redescription of what I believe to be Walker’s species.*

¢@. Length of body 45-5 mm. ; of ovipositor 2mm.; of wing 5°5-6 mm.

Head reddish brown, with black occiput, black ocellar plate and black vertical
plates ; antennae with reddish basal joints and infuscated third jomt; lunula not
specially developed. Thorax shining black, more glistening on the sides and pleurae ;
the short dorsal pubescence is yellowish; the bristles are black. The scutellum
is not well preserved, and seems to have no trace of a. sct. Halteres yellowish.
Abdomen narrow and elongate, shining black, with black pubescence and black
bristles ; ovipositor as long as the abdomen, shining black. Legs long and rather
slender, black, with the tibiae at end and the tarsi yellowish; front femora with
three black bristles beneath. Wings with a rather broad and quite hyaline axillary
lobe. The pattern is as described by Walker; the black basal stripe extends from
the first costal cell to the middle of the second, ending there obliquely and separated
from the black stigma. The second hyaline indentation of the fore border extends
with its acute point into the base of the first posterior cell, along the outer side
of the small cross-vein; the oval whitish spot at the base of the first posterior
cell is placed much before the middle of the cell itself, only a little beyond the upper
end of the posterior cross-vein. The three indentations of the hind border are narrow
and oblique, the basal one entering the base of the discoidal cell and reaching
the fourth vein, while the much shorter middle one enters only with a very short
point and the third stops just at the upper basal angle of the second posterior cell.
The last portions of the third and fourth longitudinal veins are slightly but distinctly
convergent towards the end; the two cross-veins are very close together, their
distance on the fourth vein being shorter than the length of the small cross-vein ;
the hind cross-vein is curved and oblique, the lower exterior angle of the discoidal
cell being acute and produced ; third posterior cell as broad as the axillary lobe ;
lower angle of the anal cell rather acute but not produced. Costal bristle long.

*[The specimens described by the author agree well with Walker’s type.—ED.] °*

256 PROF. M. BEZZI,

A couple of specimens from Uganda Protectorate, Entebbe (Forest), 3,800 feet,
5-11.vu.1911 (S. A. Neave).

Aciura tetrachaeta, Bezzi (1918).

Of this characteristic species, described from N. W. Rhodesia, there is a couple
of specimens from Natal, Estcourt, 1x.-x. 1896, and Malvern, 1897 (G. A. K. Marshall).

The as yet undescribed male is very like the female; its abdomen is obtuse at
the end, and the genitalia are black. The pubescence of the back of the thorax
is yellowish, and the bristles are likewise yellowish but more or less infuscated,
those of the pleurae being black. The front femora are yellow on the inner side
and bear four long bristles beneath. The greyish shade in the middle of the first
or basal indentation of the hind border is not distinct in the present specimens,

Spheniscomyia Bezzi, 1913.

I place in this genus a new species, which is rather aberrant in lacking the a. sct. ;
but in size, in the chaetotactic characters of the head, the black ocp., the form of
the scutellum and in the faint dust of the thorax, it agrees with the typical species.
On account of these characters, it is very different from the Ethiopian species which
form the very homogeneous group of Tephrella dealt with by me in the preceding

paper.

Spheniscomyia sexmaculata, Macquart (1843).
Of this common species there is a male specimen from Nyasaland, Mt. Mlanje,
4.vii.1913 (S. A. Neave), and another from Natal, Port Shepstone, v. 1897 (G. A. K.

Marshall).

Spheniscomyia neavei, sp. nov. (Pl. xviu, fig. 6).

A pretty species very distinct from any other in having no scé. and in having
two hyaline indentations at the fore border of the wing. It is possible that
A. capensis, Rond., which is described with a similar wing pattern, belongs to the
same natural group as the present species.

§ 2. Length of body 2°3-2°5mm.; of ovipositor about 1 mm.; of wing 3-3:1
mm. .

Occiput entirely black, but clothed with rather dense, dark grey dust; frons
reddish, with whitish orbits and black, but densely grey-dusted ocellar and orbital
plates ; lunula whitish, like the linear cheeks; face yellowish, with whitish dust ;
jowls very narrow, reddish, not spotted. Antennae entirely reddish yellow, with
a shortly pubescent arista. Palpi and proboscis pale yellowish. All the cephalic
bristles are black, even those of the occipital row ; oc. well developed ; two strong
2.o0r.; genal bristle well developed. Thorax and scutellum entirely black and rather
shining, the greyish dust of the back being very faint and even wanting on the
scutellum ; the bristles are black; dc. placed on the line of the a. sa.; one mpl. ;
pt. above as developed as the st.; no a. sct., the b. sct. being very long but not
diverging. The short pubescence of the back of the mesonotum is black, like the
rather sparse hairs of the scutellum. Mesophragma shining black; scutellum
broader than long and rather convex. Squamulae white; halteres pale yellowish.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 257

Abdomen shining black, with black pubescence and short black bristles; venter
black, but less shining; ovipositor glistening black and with black pubescence.
Legs black, with yellow tibiae and tarsi; the femora are narrowly yellowish at
tip, and the hind tibiae are infuscated at base; front femora with four long bristles
beneath. Wings with long and strong, but simple costal bristle; the veins are
black, but they are yellowish at base and on the hyaline indentations. Small
cross-vein placed on the apical third of the discoidal cell, which is much broadened
outwardly ; hind cross-vein convex outwardly ; last portions of the third and fourth
veins about parallel, the former being only a little curved downwards near the end ;
lower angle of the anal cell acute, but not produced. The base of the wing is
quite hyaline, the dividing line being straight and running obliquely from the
distal end of the second costal, second basal and anal cells to the hind border; the
stigma is entirely black. At the fore border there are two triangular, hyaline
indentations of equal size, just beyond the stigma and reaching with their obtuse
point the third longitudinal vein. The wholly black discoidal cell has near its base
a rounded hyaline spot, which is above in contact with the fourth vein, while
below it is distant from the fifth. At the hind border there are four hyaline identa-
tions, disposed in pairs; two shorter and broader into the third posterior cell,
reaching the fifth vein, the basal being twice as broad as the apical one; the other
two into the second posterior cell, the basal one extending along the hind cross-vein
(but without being in contact with it) to the fourth longitudinal vein; the apical
one is curved inwardly and ends broadly obtuse at the third vein, crossing the
first posterior cell in its last third. The hyaline parts are distinctly whitish ;
the dark parts are blackish, only the basal band below the sixth vein in the axillary
cell being more greyish.

Type ¢ and type 9, a couple of specimens from Nyasaland, Mt. Mlanje, 16.vi.
1913 ; an additional female specimen from the same locality, 31.v.1913, all collected
by S. A. Neave, in whose honour this interesting species is named. In the
Entomological Museum of the University of Kansas, at Lawrence, Kas., U.S.A.,
there is a female specimen collected by F. L. Snow at Salisbury, Rhodesia, 5050 ft.,
v. 1901; it differs in having the basal hyaline spot of the discoidal cell of greater
size, extending from the fourth to the fifth longitudinal vein.

Tephrella, Bezzi, (1913).
Tephrelia sexfissata, Becker (1910).
Acura sexfissata, Becker, Ann. Soc. Ent. France, 1910, p. 28.
In my previous paper (Bull. Ent. Res. ix, p. 22) I overlooked this species, which

is nearly allied to but distinct from 7. rufiventris, Bezzi ; it is recorded from British
Kast Airica, Nairobi (de Rothschild).

IV. Subfam. TRYPANEINAE.
Spathulina, Rondani, 1856.

This genus, as defined by me in the preceding paper, seems to have a near ally
in the South American genus Lamproxyna, Hendel (1914, p. 64), which shows also
a shining black, not dusted abdomen, but seems to be different in the much more
elongate and bicubitate proboscis, in the antennae being inserted below the middle

258 PROF. M. BEZZI,

of eyes, etc.; moreover the type-species, L. nitedula, Hendel, from Peru, has a
different wing pattern, with more numerous hyaline spots, which are disposed
in rows along the longitudinal veins. In the new material there is the following
new species, which is the first to be known from West Africa.

Spathulina bioculata, sp. nov.

Closely allied to sematra and running down to it in my table (1918, p. 28), but
at once distinguished by the pale yellowish macrochaetae of the head and thorax,
and by the presence of four (not three) hyaline spots in the second posterior cell,
and of two hyaline dots in the first posterior cell.*

©. Length of body 2°5 mm. ; of ovipositor about 1 mm.; of wing 3°5 mm.

Head not depressed, in front view about as high as broad, in side view narrower
than high; it is entirely pale yellowish, more whitish on the face, and blackish
but grey-dusted in the middle of the occiput ; middle frontal stripe more yellow
in the apical half, and there with some scattered yellowish hairs; all the bristles
are pale yellowish, those of the hind border quite whitish; there are three 2. or.,
but the apical one is only half as long as the others; oc. long and strong; ocp.
rather acute. Antennae pale yellowish, shorter than the face, with the pubescent
third joint as long as the first two joints together and a little pointed at its upper
terminal angle ; arista microscopically pubescent. The hairs of the lower occipital
border and of the chin are whitish. Palpi whitish ; proboscis yellow, rather thick,
as long as the lower border of the head, with the terminal portion thick, bent back-
wards and a little shorter than the basal one. Thorax black on the back, but
clothed with a dense opaque, grey dust ; a narrow stripe just above the notopleural
line is reddish yellow; the pleurae are black, with a darker grey dust, but they
are yellowish below the prothoracic stigma; the scutellum is grey on the middle,
broadly yellowish at the borders and below; the mesophragma is black, clothed
with a dark grey dust, like the pleurae. The short pubescence of the back is pale
yellowish like that of the pleurae, which is longer and more sparse; all the bristles
are yellowish ; the a. sct. are as long as the 0. sct. and are decussate ; the dc. are
placed before the line of the a. sa. Halteres whitish. Abdomen shining black,
with black hairs and black bristles ; the first two segments are dark reddish-brown
at the sides, and the first three segments have a narrow reddish hind border ;
ovipositor depressed, shining black. Legs, including coxae and tarsi, entirely
pale yellowish; their hairs are whitish, the spur of the middle tibiae reddish; hind
tibiae with a well developed posterior row of bristles. Wings proportionally long
and broad ; veins pale yellowish at base and on the hyaline dots, blackish on the
dark parts. Third and fourth veins distinctly diverging at the end; hind cross-
vein perpendicular, a little curved outwardly and shghtly longer than its distance
from the small one. The pattern is as in semiatra, with the following differences :
the subapical hyaline spot of the submarginal cell is much broader; there are
two small hyaline dots in the first posterior cell, the apical one being twice as broad
as that placed over the upper end of the hind cross-vein ; the second posterior cell

* These two dots are not shown in Loew’s figure, pl. ii, fig. 12 ; but in the description
they are recorded as being sometimes present, it seems therefore that it is a somewhat

variable character.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 259

is distinctly more elongate and shows a fourth hyaline spot before its upper apical
angle, a little smaller than the three others, which are disposed in a triangle; the
basal hyaline spot of the third posterior cell is broadly separated from that of the
axillary lobe, which shows two small hyaline dots in the dark space below the sixth
vein. Alula whitish hyaline and unspotted. The costal bristle is long and black.

Type 9, a single specimen from N. Nigeria, Zungeru, 15.1x.1910 (Dr. J. W.
Scott Macfie).

The character of the long and decussate a. sct. is common to the present species
and semiatra, Loew, and is in contrast with the other species in which these bristles
are rudimentary or even wanting ; in Lamproxyna there are likewise only two
sct., the apical ones being wanting. This character must also be taken with
caution in my table of genera ; even the character of the 2. or. can be misleading,
as in the present species there are three pairs of these bristles, while the other

species have only two pairs.

Pliomelaena, Bezzi, 1918.

Pliomelaena brevifrons, Bezzi (1918).
A female specimen from Natal, Port Shepstone, v. 1897 (G. A. K. Marshall).

Euaresta, Loew, 1873.

Euaresta amplifrons, sp. nov. (Pl. xvii, fig. 7).

A robust species of comparatively large size, very near planifrons, Loew, but
at once distinguishable by the presence of a pair only of scutellar bristles (agreeing
in this with the nearly allied megacephala, Loew), by the striking abdominal
coloration and by the somewhat different wing pattern.

g. Length of body 7 mm. ; breadth of abdomen 2 mm. ; length of wing 7 mm.

Head very broad, even a little broader than the thorax, in front view broader
than high, in side view about as broad as high ; it is entirely yellowish, but clothed
with a dense cinereous dust. Occiput concave, with a double blackish spot in the
middle, above and on sides of the neck ; the rather stout hairs of its lower portion
are whitish, like those of the chin. Frons very broad and quite bare, about three
times as broad as the eye, and even a little broader than long ; it is very flat, being
not produced above the eyes, but conspicuously prominent in front above the
antennae ; it is entirely of a reddish grey colour, the middle band being not distinct
from the orbits ; the small ocellar dot is blackish ; lunula very well developed, broad,
semicircular, rather prominent. Cheeks rather broad; face a little shorter than
the frons, concave in the middle, rather prominent at the mouth-border, clothed
with whitish dust; jowls twice as broad as the cheeks, a little broader than the

third antennal joint. Eyes rounded. Antennae inserted a little above the middle
of the eyes, very short and much shorter than the face; they are entirely

yellowish ; first jomt with pale yellowish hairs; second joint rather prominent
above; third joint shorter than the first two joints together, with the upper
terminal angle rather acute; arista yellowish, thickened at base, microscopically
pubescent. Proboscis thick, short, yellowish ; palpi as long as the proboscis,
broad, yellowish, with some bristly, yellowish hairs. All the cephalic bristles
(637) D

260 PROF. M. BEZZI.

are yellowish, those of the hind border of the head being paler ; two 2. or., which
are more developed than the s. or. ; oc. about as long as the vt. Thorax, scutellum,
postscutellum and mesophragma entirely black, but clothed with dense cinerous
tomentum, which on the back is more clear and shows a little bluish nuance ;
the short hairs are pale yellowish, like all the macrochaetae, but those of back
are darkened at the base and inserted on small blackish dots; dc. placed well in
front, only a little beneath the line of the prst.; pt. as strong as the mpl. ; st. even
stronger. Scutellum bare, with the b. sct. only, which are long, divergent,
inserted on rather large black spots and placed rather distant from the
borders ; there is no trace of the a. sct. Squamulae whitish, with a rather long
whitish fringe ; halteres pale yellowish, with a little infuscated knob. Abdomen
broad and very obtuse at the end ; it is reddish yellow above on the first three
segments, and quite black on the fourth, which is twice as long as the preceding
segment ; all the hairs are pale yellowish, like the lateral and apical bristles ;
venter entirely reddish, even on the last segment; genitalia rounded, blackish
brown. Legs rather short and stout; they are entirely reddish and rather bare,
with whitish dust ; front femora with a row of four stout yellowish bristles beneath
on the apical half; hind tibiae without a distinct row. Wings with a distinct
costal bristle and with yellowish veins; the distal portions of the 2nd, 3rd and
4th veins are straight and slightly divergent ; cross-veins rather close together,
their distance being shorter than the length of the straight posterior cross-vein,
and the anterior cross-vein being placed only a little before the last fourth of the
discoidal cell; lower angle of the anal cell rather acute, but less produced. The
pattern is very like that of planifrons, but the dark patch is more reduced, not
extending basally beyond the stigma, the basal portion of the submarginal cell
being thus hyaline, and the point of the bifurcation of the 2nd and 3rd veins being
in the shape of a conspicuous black callosity, which is very strikingly developed,
while in planifrons it is comprised in the dark patch. The hyaline spots at the
apex and hind border of the wing are of greater size, the radiating pattern being
thus more developed ; the discoidal cell is almost completely hyaline on its basal
half ; the second basal cell is quite hyaline ; the dark spot at the lower end of the
anal cell is less developed, and there is no dark spot at all below the end of the 6th
vein.

Type 3, a single, rather damaged specimen from Natal, Malvern, v. 1897
(G. A. K. Marshall).

The wing pattern of the present species is not unlike that of some species of
the genus T'rypanea, but is also very like that of Huaresta conjuncta, Loew. It
is very probable that the present species may be galligenous, like H. megacephala,
Loew, which in Sicily makes very conspicuous galls on the twigs of the Composite
plant, Inula crithmoides;* it is therefore possible that EH. megacephala,
E. planifrons and E. amplifrons will belong to a natural genus, which is biologically
characterized by the faculty of making so-called pleurocecidia on Composite plants,
a character rather uncommon in the present group of flower-head flies.

* T, De Stefani Perez, Marcellia, Avellino, iii, 1904, pp. 122-125; Trotter e Cecconi,
Cecidotheca italica, Fasc. xiv, no. 326 (1906); C. Houard, Les Zoocécidies des Plantes
ad’ Europe, etc. ti, 1909, p. 972, no. 5627.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 261

Ensina, Rob.-Desv., 1830.

Ensina sororcula, Wiedemann (1830).
Of this common species there is a male specimen from Natal, Ulundi, 5,000-6,500
ft., ix. 1896 (G. A. K. Marshall).

Ensina giadiatrix, sp. nov.

Closely allied to the preceding species, but distinguished by the infuscated
antennae, the black femora which have only a very narrow yellowish tip, and by
the exceedingly long ovipositor of the female.

©. Length of body 3 mm.; of the basal segment of the ovipositor 1°3 mm.,
and of the completely exserted ovipositor 3 mm. ; of the wing 3 mm.

Head as in sororcula, but less depressed and with the frons distinctly shorter ;
ocellar spot blackish ; third antennal joint distinctly infuscated. The thorax
seems to be darker and devoid of distinct pattern on the back; the pleurae are
more decidedly blackish ; there is no trace of a. sct. The abdomen seems likewise
to be darker above and devoid of distinct pattern ; the ovipositor is shining black,
depressed, with the basal segment a little longer than the whole abdomen (in
sororcula itis about a half as long as the abdomen), the second segment shining
black, the third segment reddish ; if completely exserted, as in the case of the
type, the ovipositor is longer than the entire body. In the legs all the femora
are entirely black, having only the extreme tip narrowly yellowish, while in sororcula
they have the apical half yellowish, and the four anterior ones are mainly or even
entirely yellow. The wings are proportionally broader and have the same
venation ; the pattern is very similar, but the stigma is quite black, even at the
base ; the fuscous reticulation is more developed on the posterior half of the wing,
chiefly in the discoidal and the second posterior cells ; in the first posterior cell the
hyaline spots are disposed in two distinct longitudinal rows, a condition which is
never to be observed in sororcula, in which they are not very distinct, never
ordinate, but are in one row only at least towards the middle of the cell.

Type 9, a single damaged specimen from Natal, Ulundi, 5,000-6,500 ft., ix.
1896 (G. A. K. Marshall).

Ensina magnipalpis, sp. nov.

Closely allied to dubia, Walker and Loew, but distinguished by the hyaline
spot of the stigma, and by the much less numerous hyaline spots of the first basal
and discoidal cells.

¢. Length of body 38-4 mm. ; of wing 4-4:2 mm.

Occiput black, grey-dusted above, yellowish below. Frons flat, not prominent
in profile, broad, parallel-sided, about 1°5 times as long as broad ; it is of a dark
orange-yellow colour, but a narrow longitudinal middle stripe and the orbits are
whitish ; lunula broad and rounded, yellowish ; cheeks narrow, with white shining
dust; jowls yellowish, unspotted, as broad as the third antennal joint. Face
narrower and shorter than the frons, with broad and very prominent mouth-border.
Eyes in profile rounded; antennae inserted a little below their middle; they
are entirely yellowish and a little shorter than the face; second joint with yellow

(637) p2

262 PROF. M. BEZZI.

hairs ; third joint broad, rounded at tip, about twice the second in length ; arista
microscopically pubescent. Palpi pale yellowish, with pale hairs; they are very
long and broad, flattened, as long as the basal part of the proboscis; the latter
is very long, about as long as the head and thorax together ; its basal part is pro-
duced much beyond the mouth-border, and the apical part is acute and much produced
behind the chin. Bristles abraded in the type, but they seem to be whitish on the
occipital border and black on the frons; pubescence pale yellowish. Thorax
clothed on the back with a dense opaque whitish-grey dust, but with a very broad
even longitudinal blackish stripe, which extends laterally to the dc.; a smaller
lateral blackish stripe on each side above the notopleural line; the back can
therefore be described as black with two lateral whitish stripes; the pleurae
are entirely greyish and unspotted. The bristles of the back are black; the
pubescence is pale yellowish ; on the pleurae the st. and the mpl. are black, while
the others are pale yellowish or whitish, like the hairs. Scutellum blackish, greyish
at the end and below, with four equally long and strong black bristles, the apical
ones diverging. Postscutellum and mesophragma coloured like the pleurae.
Squamulae and halteres dirty whitish, the former with a white fringe. Abdomen
broad and flat, grey, with two longitudinal rows of blackish spots, which are partly
confluent to form two longitudinal stripes; last segment with a broad yellowish
hind border ; pubescence pale yellowish ; apical bristles black ; genitalia yellowish ;
venter yellowish grey. Legs entirely yellowish, the femora more or less greyish
on the outer side; hairs and bristles pale yellowish or whitish, only the apical
spur of the middle tibiae being black. Wings with a well-developed black costal
bristle ; costa black on the dark, and yellowish on the hyaline parts ; veins black,
in part yellowish at base and on the hyaline spots; stigma longer than broad,
deep black, with a small circular, sharply defined, whitish hyaline spot before the
end. First vein ending before the small cross-vein; 2nd vein straight; 3rd vein
bare, straight throughout and parallel with the last portion of the 4th; smalt
cross-vein placed considerably beyond the middle of the discoidal cell; hind
cross-vein perpendicular, convex outwardly, distinctly longer than its distance
from the anterior one; discoidal cell long; anal cell broader than the second basal
cell, with an acute but not produced lower apical angle; 6th vein complete but
colourless at the end. The pattern is blackish, more intense on the fore half, paler
behind ; in general shape it is like that figured by Loew for dulia (pl. u, fig. 20)
with the following differences, besides the hyaline spot of the stigma; the hyaline
spots in the costal cell are broader ; in the base of the submarginal, 1st and 2nd basal
and discoidal cells there are only very few, much broader hyaline spots; the 3rd
posterior cell has the reticulation closer, being fuscous with about 8 hyaline spots.
The axillary lobe is likewise hyaline, with two fuscous rounded spots at the hind
border ; the anal cell is also hyaline.

Type 3; and an additional specimen of the same sex from Durban, Umbilo,
28.v1.1914 (L. Bevis).

These two species, dubia and magnipalyis, which agree in palpal and other
characters, and have a similar wing pattern with much extended black parts, require
the formation of a new genus, if they cannot be placed in the genera Spathulina
or Pliomelaena.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 263

Ensina siphonina, Bezzi (1918).
There is an additional specimen of this very distinct species, likewise from British
East Africa, Nakuru, i. 1913 (Dr. B. L. van Someren).

Trypanea, Schrank, 1795.

Of this genus, numerous species of which are tabulated in my previous paper,
there are in the new material before me two forms which are both very different
in wing pattern from all the other Hthiopian species before known, and are here
described as new.

Trypanea sudcompleta, sp. nov. (PI. xvii, fig. 8.).

Distinguishable by the four strong scutellar bristles and by the wings, which
have the typical star-shaped subapical pattern, but are marked to the base with a
faded but well developed reticulation, like that of the American species abstersa,
Loew, and reticulata, Hendel.

Q. Length of body 3°7 mm. ; of ovipositor 0°5 mm.; of wing 4mm.

Head of proportionally smaller size and rather narrow in profile, with the purplish
eyes higher than broad ; it is greyish, with a dark spot on the middle of the occiput ;
frons about as broad as long, not at all prominent in front, with whitish dust in the
middle; lunula rather broad; face short, not prominent below; cheeks linear
and jowls only a little broader. Antennae inserted at middle of the eyes, only
a little shorter than the face, entirely yellow, with the third joint rounded at end ;
arista bare. Palpi and proboscis yellowish, the latter short and retracted into the
oral cavity. Cephalic bristles dark yellowish, but those of the hind border whitish ;
oc. long and strong; three 7. or., but the anterior one less than half as long as the
others. Thorax blackish on the back, yellowish on the sides, but densely clothed
with grey dust, opaque and unicolorous; the short hairs are pale yellowish, the
bristles dark yellowish, those of the back being inserted over small black dots ;
pt. as strong as the st.; hairs of the hind border of mesopleura rather long.
Scutellum like the thorax, but broadly yellowish at the hind border; Db. sct. long,
diverging, inserted over rather broad black spots ; a. sct. a little shorter, decussate,
on smaller black spots. Postscutellum and mesophragma entirely grey. Squamulae
whitish, halteres pale yellowish. Abdomen like the thorax, with short pale yellowish
hairs and yellowish terminal bristles ; the first three segments have a rather broad
yellowish hind border; the 4th and 5th are blacker basally and more yellow
apically ; ovipositor depressed, shining black, short and obtuse, a little shorter
than the last two abdominal segments together. Venter pale yellowish, with
whitish hairs. Legs short and stout, entirely pale yellowish ; front femora above
with a complete row of short yellowish bristles, below with an apical row of four
very long yellowish bristles ; spur of the middle tibiae blackish ; hind tibiae with
a well developed row towards the middle. Wings long and narrow, with a distinct
costal bristle and pale yellowish veins, which are infuscated on the darker parts
of the pattern ; first vein ending much before the small cross-vein ; 2nd, 3rd and
4th veins straight in their terminal portions and equally and gradually divergent ;
cross-veins perpendicular and parallel, the hind one being longer than its distance
from the anterior one; lower angle of the anal cell acute but less produced ; 6th

264 PROF. M. BEZZI.

vein prolonged to the hind border. The star-shaped apical patch is blackish
brown; it encloses a triangular indentation into the marginal cell, two hyaline
spots near the border at end of the submarginal cell, one just below the end of the
2nd vein and the other broader and subquadrate before the end of the 3rd vein.
The first posterior cell has two hyaline spots at base, the first as long as the breadth
of the cell itself and in contact with the small cross-vein, the second rounded, half
as long, and in contact with the 4th longitudinal vein ; the cell has at the end the
three usual spots, which form the two usual radiating branches ; two other fuscous
parallel rays are in the second posterior cell, and another less distinct ray on the
hind cross-vein. The fuscous rays at the apex and hind border are therefore five
in number. The basal portion of the wing shows a pale but complete reticulation,
formed by rounded hyaline spots of rather large size; of these spots there is a row
only in the base of the submarginal and first basal cells; the discoidal cell has a
row in the basal, and two rows in the apical half; the third posterior cell has ten
broader spots disposed in two rows, and along the hind border three more spots ;
axillary cell with six spots in two rows. The base of the second posterior cell has
three spots, two larger and one small, between the ray of the hind cross-vein and
the middle ray. The short stigma is quite colourless, with only a black dot near
the costa at base of the costal bristle ; it is not united with the subapical patch ;
the oblique ray departing from it and extending to the small cross-vein is very
faint and belongs to the pale reticulation.

Type 9, a single specimen from British East Africa, Nairobi, 27.1v.1911 (7. J.
Anderson).

From 7. abstersa (which I have in my collection from California, New Mexico
and Colorado) the present species differs in the closer reticulation of the wings,
in the wholly colourless stigma, and in the want of the hyaline spot in the marginal
cell, just above the end of the second longitudinal vein (not to be confounded with
the hyaline spot in the submarginal cell just below the end of the same vein).
From 7’. reticulata it differs in the subapical brown patch being much more extended
and more typically shaped.

Trypanea peregrina, Adams (1905).

Having had occasion to see the type of this species in the Entomological Museum
of the University of Kansas, Lawrence, I must recognize that it is the same as my
T. urophora of the preceding paper, as already suspected. The figure 11 on pl. 1,
is too dark, the ray extending from the stigma to the small cross-vein being really
much paler and interrupted. The form of the ovipositor is the same, even though
it is depressed in the type of peregrina.

Trypanea hemimelaena, sp. nov. (PI. xviu, fig. 9).

A pretty species recalling 7. guimari, Becker, from the Canary Islands, and
hkewise with the apical half of the wing only occupied by a not typically star-
shaped pattern, which shows however more numerous hyaline spots.

S. Length of body 3°2 mm.; of wing 3°2 mm.

Head entirely pale yellowish, clothed with whitish dust, and only narrowly
‘blackened towards the middle of the occiput, above the neck ; it is rather large,

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 265

being in front view broader than high, and in side view higher than broad ; eyes
greenish, unicolorous, higher than broad. Frons flat, a little prominent in front
above the base of the antennae, as long as the face, with a broad bare yellow
middle stripe, and with a whitish-grey ocellar triangle and lateral orbits; lunula
broad, semicircular, whitish grey. Face flat, with shallow antennal grooves and
the upper mouth-border not prominent, cheeks linear ; jowls only about twice as
broad as the cheeks, unspotted. Antennae inserted above the middle of the eyes,
much shorter than the face ; they are pale yellowish ; third joint pubescent, with
the upper terminal angle a little infuscated and not acute ; arista microscopically
pubescent, yellowish, thickened at base. Palpi and proboscis yellowish, the latter
thick and short, retracted into the broad oral cavity. All the cephalic bristles
are pale yellowish, those of the hind border whitish; oc. long and strong; two
v. or.; genal bristle distinct, but short and whitish ; the rather short hairs of the
lower part of the occiput and of the chin are whitish. Thorax black in ground-
colour but clothed with a dense grey dust, which on the back is more clear and a
little yellowish, and darker on the pleurae; there is no distinct pattern on the
back, except an ill-defined dark stripe below the notopleural line; all the hairs
and bristles are whitish, the latter being inserted over very small but striking
black dots; dc. placed near the suture; pt., mpl. and st. well developed, but
. intermingled with rather long and stout hairs. Scutellum coloured like the back
of mesonotum and likewise pubescent ; the two basal bristles are inserted over
proportionally broad dark spots; the apical bristles are medium-sized and
decussate at the end. Postscutellum and mesophragma clothed with a dense
grey (not yellowish) dust. Squamulae and halteres pale yellowish. Abdomen
elongate and rather attenuate at the end; it is densely tomentose like the back ©
of the mesonotum (in guimarz it is less tomentose, the black ground-colour being
visible) and clothed with rather long pale yellowish hairs; hind border of the
first two segments distinctly yellowish; venter blackish, with short scattered
whitish hairs; genitalia black. Legs rather short and stout, entirely reddish,
the front coxae grey-dusted ; they are clothed with a short whitish pubescence ;
front femora with five stout yellowish bristles beneath; hind tibiae with no
distinct row. Wings proportionally short and broad, with a distinct costal bristle
and pale yellowish veins, which are blackened on the apical half; 3rd and 4th
veins straight and a little divergent at the end, while the 2nd is more strongly
divergent from the 3rd; cross-veins straight and parallel, the small one placed
near the last fourth of the discoidal cell, the posterior one longer than its distance
from the small one ; lower corner of the anal cell forming a wide angle; 6th vein
prolonged to the hind border, even if very faint. The stigma is very short and
pale yellowish, with a small black dot near the base of the costal bristle ; the first
vein ends much before the small cross-vein (in guemari it ends below or only a
little before it). The basal half of the wing is entirely whitish hyaline and quite
unspotted. The apical part is occupied by a dark brown patch, which is less dark
towards the hind border, and has a yellowish patch on the small cross-vein into.
the end of the first basal cell, prolonged above into the submarginal cell towards
the stigma. In the brown patch there are the following hyaline spots ; two spots
close together in the marginal cell just beyond the stigma, the first very broad

* 266 PROF. M. BEZZI.

and subquadrate, the second much narrower and subtriangular; in the sub-
marginal cell there are three rounded spots, one in the middle just above the upper
end of the small cross-vein and at the inner end of the upper part of the yellowish
patch, the other at the fore border just below the tip of the 2nd vein, the third
of greater size and subtriangular just before the lower apical angle. The first
posterior cell has a rounded spot near the base, half as broad as the distance
between the 3rd and 4th veins and in contact with this last vein before the upper
end of the hind cross-vein ; two other spots before the end, one over the other,
that in contact with the 3rd vein being smaller than that in contact with the
4th; and one apical, extending from the 3rd to the 4th vein, and thus forming
the normal apical dark fork of the Trypanea-pattern, though a little shorter than
usual. Discoidal cell with four about equally small, rounded spots, disposed as
a square. Second posterior cell with three larger rounded spots along the hind
border, of an indentation-like shape, and one of equal size near its upper inner
angle. Third posterior cell with a single rounded spot of greater size before its lower
apical angle and in contact with the hind border of the wing.

Type 3, a single specimen from the Gold Coast, Aburi, 1912-13 (W. H. Patterson).

Owing to the great affinity of the present species with 7. guimari, which shows
a very remarkable sexual dimorphism, it is possible that the unknown female
has a somewhat different wing pattern. These two species, like 7. peregrina,
belong evidently to the gnaphalu-mamulae group.

V. Subfam. ScHISTOPTERINAE.

This peculiar subfamily is represented in the new material before me by two
remarkable forms, one of which requires the formation of a new genus, while the
other, being closely allied to Rhabdochaeta spinosa, Lamb (1914), from the
Seychelles, may be provisionally placed in the Oriental genus Rhabdochaeta, de Mei}.

An important character of the subfamily is that the s. or. are placed on a line
converging towards the centre of the frons, being thus not in the same line with
the 7. or.

Perirhithrum, gen. nov.

This new genus is more closely allied to the Oriental Rhabdochaeta than to the
Ethiopian genera Schistopterum or Rhochmopterum, but differs from all the known
genera of the subfamily in having the first and third longitudinal veins bristly
above, in having a produced lower angle in the anal cell and in having a very
oblique hind cross-vein.

From Rhabdochaeta the new genus differs in the want of the middle pair of bristles
on the frontal stripe. These bristles are likewise wanting in Schistopterum and are
not mentioned for Rhochmopterum ; they are recorded for the type species of the
genus Khabdochaeta, as well as for the species spinosa, Lamb, and are present also
in my species baker: from the Philippines. This pair of bristles seems to correspond
to the crossed frontal pair of the ANTHoMyIIDAE: but in Rhabdochaeta they are not
crossed. It is interesting to note the presence of a similar pair of bristles in the
male of Euribia perpallida, Bezzi (as pointed out by me in vol. ix, p. 35, note),
the wing-pattern of which rather recalls that of the ScHISTOPTERINAE.

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 267

Body robust, of comparatively large size, with strong bristles of different
coloration, but not of different kind. Head in front view broader than high, in
side view rather narrow. Eyes bare, of large size, distinctly higher than broad.
Occiput rather flat and not swollen below. Frons about as long as broad, with
nearly parallel sides, not at all prominent in front; middle band devoid of
central bristles; lunula rather small and deeply excavated. Face shorter than
the frons, concave in the middle and with a very prominent border at the upper
mouth-edge and at the sides of the mouth; cheeks linear; jowls very narrow.
Antennae inserted at middle of the eyes and as long as the face; first joint very
short ; second joint prominent above and there clothed with short obtuse stout
bristles ; 3rd joint about as long as the first two joints together, of a rectangular
shape, with the upper terminal angle acute but not very prominent ; arista longer
than the antennae, with rather long pubescence. Mouth-opening broadened
anteriorly ; palpi long, flattened, bristly, prominent and of the usual shape for the
subfamily ; proboscis short, with short and broad terminal flaps. Occipital row
formed by stout obtuse whitish bristles; inner vf. very long; oc. long and strong,
widely distant from each other at the base; two s. or., not in the same line with
the three 2. or., but converging in front and thus the anterior pair about in front
of the oc.; middle frontal band bare; genal bristle strong; lower portion of the
occiput with scattered stout hairs. Thorax about as long as broad, moderately
convex above, with sparse scattered pubescence, even on the pleurae. Chaetotaxy
complete, but with no distinct scp.; dc. placed near the suture and a little before
the line of the a. sa.; one mpl. ; pt. a little weaker than the st. Scutellum broadly
triangular, gently convex above, clothed with sparse hairs ; 0. sct. long and diverging ;
a. sct. well developed, but broken in the type. Abdomen elongate, narrower
than the thorax, nearly bare above and on the sides, with some rather stout bristles
at the end; male genitalia globose and rather prominent downwards. Legs rather
slender, the femora not incrassated ; front femora with a complete row of strong
bristles below ; middle tibiae with a single spur; hind tibiae with no distinct row
on the outer side. Wings rounded and very broad ; costal bristle well developed,
with a second much smaller one before it; there is also a group of 4-5 short
bristly hairs at the costa near the humeral cross-vein and its incision. The incision
at the end of the auxiliary vein is present, but much shorter and less prominent
than typically. Auxiliary vein well developed, but ending at a straight angle
and there evanescent ; first vein ending about at middle of the fore border and
distant from the costa (the costal and subcostal cells being thus very broad), and
suddenly bent forwards at tip; it is clothed with bristles, which on its terminal
part are very long. Stigma distinct, about as broad as long. Second vein long,
straight, a little bent forwards at the apex, ending nearer to the end of the 3rd
than to that of the first vein. Third vein quite straight, clothed with short bristles
to the middle of its last portion; the submarginal cell is greatly widened at the
end, because the 2nd vein is strongly divergent from the 3rd, which is parallel with
the longitudinal axis of the wing. Fourth vein likewise straight and in its last
portion quite parallel with the last portion of the third. Small cross-vein normally
developed, straight, perpendicular, about as long as its distance from the upper
end of the hind cross-vein ; it is placed near the terminal third of the upper border

268 PROF. M. BEZZI.

of the discoidal cell, but near the middle of the lower border of the same cell.
Cubital fold well developed and forked outwardly, evenly chitinised. Hind cross-
vein broadly S-shaped, but placed very obliquely so that its upper end is near the
middle of the wing, while its lower end is very near the end of the fifth longitudinal
vein ; the lower external angle of the discoidal cell is therefore very acute and much
produced outwardly. Anal cell broad, with its lower outer angle acute and produced
longer than the 2nd basal cell ; sixth vein prolonged to the hind border of the wing.
Axillary lobe broad and rounded, with a distinct long axillary vein; alula well
developed, rounded. The wing pattern is a very characteristic one, consisting
of a broad, light brown patch, occupying the middle and extending over the whole
hind portion of the wing, except the apex; this patch shows black spots, shining
“bullae ’’ and subhyaline spots near the hind border; from this patch depart
numerous long rays, which extend to the costa along the entire fore border and the
entire apex.
Type: the following new species.

Perirhithrum marshalli, sp. nov. (PI. xvii, fig. 10).

A strange and beautiful insect of comparatively large size, strikingly distinguished
by the very peculiar wings. |

g. Length of body 5°5 mm. ; of wing 55 mm. ; breadth of wing 3 mm.

Head yellowish, darkened towards the middle of the occiput. Frons with a dark
ocellar spot and with a transverse dark band in front before the lunula. Antennae
reddish yellow, with the third joint blackened along its upper border; arista
whitish, cheeks white, with a small deep black, rounded spot above on each side,
near the roots of the antennae ; jowls yellowish, with a reddish spot on each side
in front; face yellowish, darkened towards the middle. Eyes reddish brown and
apparently unicolorous. Palpi whitish at base, reddish yellow on the apical half,
and clothed there with short black bristles; proboscis reddish yellow. The ocp.,
the outer vt. and the pvt. are whitish; the long inner vf. are black, while the oc.
are pale yellowish; the long and strong s. or. are black with a brownish end and
are inserted over distinct, black, slightly prominent tubercles, like the first (or
inner) of the 2. or., which is likewise black with a brownish tip; the two apical
pairs of the 7. or. are on the contrary-shorter and whitish. The genal bristle is
black, while the rather long, bristly hairs near it on the chin, and those on the
lower part of the occiput, are whitish. Thorax on the back black in ground-colour,
but covered with dark grey dust; the suture is broadly interrupted in the middle,
with whitish dust on the sides and has inwardly a brownish transverse band, which
is interrupted towards the middle; the points of insertion of the bristles are darker
and spot-like, and there are in the centre two dark brownish rounded spots, one
on each side, interiorly and posteriorly of the anterior dc.; the back is denuded
in the type, but in front of the scutellum it shows rather long and scattered whitish
hairs. The pleurae are yellowish near the sutures and towards the breast, but are
darkened towards the centre of the sclerites; they are clothed with a whitish-grey
dust. All the macrochaetae are pale yellowish, but those of the back are distinctly
black at base, though less so in the case of the mpl. and st., while the pt. is entirely
whitish. . Scutellum coloured like the back of the mesonotum, but less obviously —

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 269

yellowish on the sides and below; the b. sct. are very long, pale yellowish, with a
black base, and are inserted on a small black dot; the a. sct. are wanting in the
type. Postscutellum and mesophragma blackish, grey-dusted. MHalteres with
yellowish stalk and blackish knob. Abdomen reddish, with pale grey dust; each
segment bears on each side a very broad, brown spot, which is rectangular in shape
on the three basal segments, while it is subquadrate on the fourth; it could be
said that each segment is brown, with a narrow whitish hind border and a whitish
middle line; each segment bears moreover on the sides, in the whitish portion,
a smaller dark spot, and the 4th segment is broadly yellow at the hind border.
The entire abdomen seems to be bare, with only a few scattered short black hairs
on the sides of the segments, chiefly of the last one, which bears
moreover 8-10 not long but strong, black bristles at the hind border and on
the sides. Genitalia reddish yellow; venter entirely reddish, with whitish dust
and sparse whitish hairs. Legs entirely pale yellowish, even to the end of the
tarsi, but each femur is adorned below with two rounded dark spots, one near the
base, the other towards the middle; the hairs and bristles of the front femora
are pale yellowish, while the posterior femora have blackish pubescence and some
dark bristles at the end above; middle tibiae with a dark spur. The wing veins
are yellow but darkened on the darker parts of the pattern, and more strikingly
on the costa, which is alternatively black and yellow. The short costal bristles
near the humeral cross-vein are yellowish, while the two costal bristles at the
stigma are black, like the bristles on the Ist and 3rd longitudinal veins. The
pattern is brownish, more blackish at the end of the rays of the fore border and
on several of the central spots ; the membrane is iridescent, quite hyaline on the
fore border and at the apex, while the hyaline spots of the hind border are more
whitish. The radiating part shows the following dark rays, which become
progressively longer from the base to the apex of the wing; one at the extreme
base of the first costal cell; one like a narrow border of the humeral cross-vein ;
two in the second costal cell, one broader near the base, and one narrower but
' darker beyond the middle; one narrow and greyish along the diffuse terminal
part of the auxiliary vein; three blackish, close together and very marked on
the stigma, the first and broadest of all at the base, the second in the middle, the
third and narrowest along the terminal part of the first vein. In contact with
the last of these rays there is a broader triangular greyish ray departing from the
fuscous base of the marginal cell and showing some small, subhyaline, whitish
streaks on its base. This ray is followed by two other parallel rays at equal
distances in the marginal cell itself; these two rays are narrower and bent a little
outwardly, and have their central stripes whitish hyaline. There are subsequently
two dark rays in the submarginal cell, narrow and full, one ending at the end of
the second vein, the other ending at middle of the distance between the tips of the
2nd and 3rd veins; the Ist posterior cell has likewise two narrow rays which
follow exactly the two external branches of the cubital fork, ending at the tips
of the 3rd and 4th veins ; the 2nd posterior cell has the last three rays, which are
less sharply defined, the last one being even hardly distinguishable from
the dark patch; they are more irregular and sinuous and enclose two
small, hyaline spots and a hyaline streak at the base. On the disc of

270 PROF. M. BEZZI.

the wing there are: (1) the blackish patches, two of which more striking
and more defined are of subquadrate shape and placed one in the submarginal
cell before the small cross-vein, and the other near the base of the first posterior
cell, just above the upper end of the hind cross-vein ; (2) the shining “ bullae,”
the more distinct of them being placed around the small cross-vein at unequal
distances to form a kind of circle: one in the first basal, one in the submarginal
just above the small cross-vein, two in the first posterior above and in front of the
black spot, and one in the discoidal cell, at some distance below the above-named
black spot. There are moreover some other “ bullae,” one at the extreme base
of the submarginal cell, and one at the extreme base of the discoidal cell extending
into the end of the 2nd basal cell. The 3rd posterior cell and the axillary lobe
are entirely and evenly infuscated and show some hyaline spots forming a kind of
reticulation ; 9 of these spots are placed into the axillary cell, 4 above and 5 below
the axillary vein, and 7-8 are placed in the 3rd posterior cell, those near the apex
being more stripe-like. The hyaline parts of the stigma are colourless, like the
corresponding ones of the 2nd costal or marginal cells. The basal and anal cells
are infuscated to the base, only the extreme base of wing and the alula being greyish
hyaline.

Type J, a single partly damaged specimen from Natal, Port Shepstone, v. 1897 :
another of the interesting discoveries of Dr. G. A. K. Marshall, in whose honour
the species is named.

Rhabdochaeta neavei, sp. nov.

About the same as R. spinosa, Lamb, from the Seychelles, but seeming to differ
in the wing pattern, which is yellowish brown, not black, and shows therefore
a very remarkable deep black spot at the base of the first posterior cell, just beyond
the small cross-vein.

®. Length of body 2:2 mm. ; of ovipositor 0°8 mm. ; of wing 3 mm.

Head yellowish, with the occiput darkened towards the middle; frons yellow
on the apical half, with whitish orbits and a blackish ocellar dot ; cheeks whitish,
with a small, rounded, rather prominent, deep black dot above, between the root
of the antenna and the eye; face whitish, with a brown transverse band above
the mouth-border; jowls very narrow and unspotted. Antennae pale yellowish
at base, with the 3rd joint more yellow, very attenuated and very acute at the
end ; arista whitish, shortly pubescent. Palpi and proboscis dirty yellowish.
Middle frontal stripe bare, but with a pair of white bristles in the middle.
Cephalic bristles pale yellowish, the or. and the inner vt. darkish at base ; those
of the hind border are whitish ; two s. or. and three 7. or., counting that placed
far in front, just above the black dot of the cheeks ; the stout short hairs of the
lower part of the occiput are whitish. Thorax black in ground-colour, but
densely clothed with grey dust, which is darker on the pleurae; humeri and
notopleural region yellowish ; on the back there are distinct black spots on the
points of insertion of the macrochaetae, those of the prsct. being the most developed ;
the short hairs are pale yellowish ; the dorsal macrochaetae are distinctly blackened
at the extreme base, but are elsewhere yellowish; the pleural macrochaetae are

NOTES ON THE ETHIOPIAN FRUIT FLIES, OTHER THAN DACUS. 271

entirely pale. Scutellum like the back of the mesonotum, but broadly yellowish
at the sides and behind ; the 0. sct. are blackish at base and inserted over black
spots; the a. sct. are much smaller, decussate at the end, not darkened at the
base and not inserted on black spots. Postscutellum and mesophragma black
but clothed with grey dust, which is darker on the former. Squamulae and halteres
whitish. The abdomen seems to be differently coloured from that of spinosa,
being entirely reddish, with the base of the first two segments narrowly blackish ;
hairs and bristles pale yellowish; ovipositor broad, flattened, with the basal
segment shining reddish with black end. Legs pale yellowish, with dark pubescence
and whitish bristles ; front femora beneath before the end with a blackish spot,
forming an incomplete ring, and there with blackish bristles ; the four posterior
femora have this spot more developed, and have moreover a similar but more
elongate spot near the base ; all the tibiae have a narrow black ring near the base,
less developed on the front pair; hind tibiae with a short but distinct posterior
row of bristles. The wings are shaped as in Lamb’s figure 13, p.321; they have
the same venation, with a deeply sinuous fourth longitudinal vein and a nearly
straight hind cross-vein ; in the Oriental species pulchella and venusta the fourth
vein is a almost straight, but the hind cross-vein is likewise about straight ; in
bakert the fourth vein is straight, while the hind cross-vein is broadly S-shaped
and oblique. The veins are bare, as stated by Lamb; but on the third vein I
can perceive some scattered, almost microscopic, black bristles. The general
shape of the pattern is the same; it is however not black, but yellowish brown ;
the shining “ bullae’ are of the same number and have the same disposition ;
but the wing pattern being more clear, the black spot at the base of the first posterior
cell is very striking ; it is placed below the basal “ bulla” of the same cell, and
bears at base a small rounded whitish-hyaline dot, corresponding to that figured
by Lamb at the inner basal angle of the first posterior cell. Other differences are :
(1) the marginal cell has no hyaline spots in the fuscous part at the base of the
radiating streaks ; (2) the hyaline dots in the submarginal, first basal and first
posterior cells are less numerous and much smaller. It must be remarked that
in the present species, as well as in spunosa, the costal nick, even if distinct, is much
less developed than in pulchella and bakeri ; de Meiere’s figure of venusta shows
it however not developed.

Type 9, a single specimen from Nyasaland, 8.W. of Lake Chilwa, 13.1.1914,
collected by S. A. Neave, in whose honour the species is named.

Two species of Huribia, perpallida and discipulchra,* both likewise from
Nyasaland, show a very remarkable affinity with the above-described species

of Rhabdochaeta, and are perhaps allied; but they have different wing pattern,
unicolorous legs, ete.

~ Bull, Ent. Res. ix, 19138, pp. 35, 36.

ij ie Oe aa

ean "eee hes

i sobre Ne tors,

j Whe aol ey.

(a elieuea : ian vt

Or eave aed y Peek" Ra %,: srg wn bit
a

jal re tins ‘a

dws

a y

oa oh
} ae

biwrigy Ae
we Re |
gate ts naphi nN
f say aeties ou ae
i es ule ants mea
heey wae «ys oa

ee Tf! MY :
;

a Panam, | tah rte

far a So ae wis! “Dae:

aa ina sale diy “s = t

TAP IO VOW ”

uhs Ohh hee Bi Re aed Y So! pave \i \ ‘oh, wh

{ P
et TS ie 0 yi Ke raat Lat Par (ae ar, Fk, a
avon ste eens pera INE NPN ALY Yow ho} A

* 4

> c ast : ¥ = 7 mn bag bi ,
MOM 72 ISSOs oN rTM Dy AR

voy r
unt
Wa
.
rit

; yon ay
Cae eit

i Pt

aN LN

y h

ee ae

silk f al ;
7 Meret

RS re
a i oh q
= se Meas 8) eR

tMIS RONG eh
Sp» 48

: Ae had

. FID

+
i
rege
. a
fi
.
Se oe
< .
5 i
tore
‘
t
.
.
i
—s,
j
+
40)
‘y
i
Vy J
i
: \
| ower ay
ra ic
; -_*
—
’
.
¢
‘ n
‘ %,,
ie ‘
f S

Fig.

coc on DMD Oo FP OCF DD FE

EXPLANATION OF PLATE XVIL.

Wings of Trypetidae.

. Sosiopsila trisetosa, Bezzi, sp. nov.

. Conradtina acrodiauges, Speiser.

. Leucotaenella guttupennis, Bezzi, sp. nov.
. Celidodacus obnubilus, Karsch.

. Bistrispinaria fortis, Speiser.

. Clinotaenia anastrephina, Bezzi, sp. nov.
. Hoplolopha cristata, Bezz, sp. nov.

. Rhacochlaena pulchella, Bezzi, sp. nov.

. Taomyia marshalli, Bezzi,.sp. nov.

10.

Notomma bioculatum, Bezzi, sp. nov.

PLatTeE XVII.

WO, Ihe IPsares

BuLtit. ENT. RESEARCH.

ArPRICAN BRUIT FEIES.

al
i
re
t
}
‘
hy

X
hee |
i
inae
Ibe
y
-
‘
j
aul
‘
“
\
‘
»
v
ie
;
~
vt
she
er
%
£5)
2
‘Ff !
.
vi
a
x

4 a iy
if

i
«

ba A

wiv aT 4 Mort VA. OK .

7 apbseye? Jo: agri

You pe an pent: Rey ati Coa ae, oar

\ ie 9 a od + 7
insell eaatyosts gage ACOA) 8

Bel eiih is wis’ wt

aed 48 ys writ 1\ “eR 7 cit i WJ

“von qa Ai Eh ee ne a Vetere 8
wot We diosall: edi

vou hea geno SDN:

A. Jobin
seh ee SR aon
a e tas
Neat say ey eo iat tdiha ¢ Nod
je
7 -
t

«oi
Palle’
4 1
oo eet
i
Fi
i

EXPLANATION OF PLATE XVIII.

Wings of Trypetidae.

Fig. 1. Acidia fossataeformis, Bezzi, sp. nov.
. Ocneros bigemmatus, Bezzi, sp. nov.

. Allotrypes brevicornis, Bezzi, sp. nov.
. Aciura perpicillaris, Bezzi, sp. nov.

. Aciura oborinia, Walker.

. Spheniscomyia neavet, Bezzi, sp. nov.
. Euaresta amplifrons, Bezzi, sp. nov.

. Lrypanea subcompleta, Bezzi, sp. nov.

oO ao NT DO FS O&O DH &

. Trypanea hemimelaena, Bezzi, sp. nov.

10. Perirhithrum marshalli, Bezzi, sp. nov.

SUL, JBN, IRiIsSisnecs Wot, 24 JP Bi une DW

NERUCAIN IEIROIOE JRL Se

a I ‘ ‘ AL ak te i i

shen

273

SOME INJURIOUS SOUTH AFRICAN WEEVILS.
By Guy A. K. Marswattz, D.Sc.
(Pirate XIX.)

The following undescribed species of CURCULIONIDAE were all forwarded to the
Imperial Bureau of Entomology from the Division of Entomology, Pretoria, with
the information that they were attacking various cultivated plants.

Subfamily BRACHYDERINAE.

Prototsrophus* planatus, sp. nov. (Plate xix, fig. 4).

g. Unusually flattened dorso-ventrally. Colour black, with uniform and fairly
dense brownish-grey scaling above and grey scaling beneath.

Head separated from the rostrum by a slightly curved furrow which does not
nearly meet the lateral margins; the forehead flat, with a deep and complete
central furrow, and the sculpturing hidden by the scaling and the broad subre-
cumbent brown setae; the eyes strongly convex, moderately produced backwards,
deepest behind the middle, and the hind edge of the orbit not projecting. Rostrum
a little shorter than its basal width, strongly narrowed from base to apex, the sides
almost straight ; the upper surface with a rather deep median impression containing
a complete narrow central carina, and a very shallow longitudinal impression near
each lateral margin; the mandibles densely squamose. Antennae with joint 1 of
the funicle much longer and thicker than 2, which is nearly twice as long as 3,
and all the jomts much longer than broad. Prothorax twice as broad as long, the
sides strongly rounded, broadest well behind the middle, the basal margin arcuate,
the apex much narrower than the base and broadly sinuate; the upper surface
moderately convex transversely, but flat longitudinally, set with large shallow
separated setigerous punctures, which are normally hidden by the scaling, and
with a narrow smooth central line running from the base for three-fourths the length ;
the setae broad, flattened and recumbent. lytra oblong-ovate, obtusely pointed
behind, the basal margin jointly sinuate, with the angles not produced ; the striae
very shallow, the first three strongly curved outwards at the base, the very shallow
punctures normally quite hidden by the scaling; the intervals broad, flat and
minutely punctulate, set with irregular recumbent scale-like setae ; the scales very
short ovate or almost circular, smaller than those on the prothorax and closely
contiguous. Legs stout, densely clothed with grey scales and numerous oblique
scale-like setae ; the front tibiae strongly incurved at the apex, somewhat flattened
internally and there with a row of 4 or 5 spine-like bristles, the apical comb with
6 or 7 very short teeth; the hind tibiae slightly convex externally, internally
flattened and with 3 stout teeth near the apex, the corbels squamose; joint 2
of the front tarsi a little longer than its apical width. Sternwm with the mesepisterna
broadly separated from the base of the elytra.

* Ann. Mag. Nat. Hist. (9) i, 1919, p. 14.

274 GUY A. K. MARSHALL.

Length, 55-7 mm. ; breadth, 3-4 mm.

TRANSVAAL: East Rand, 30.1.1919.

Described from three specimens.

In its colourmg and general sculpture this species most closely resembles
P. (Strophosomus) gravis, Gyl., but is readily distinguished from it by its remarkably
flattened form and almost oblong elytra.

Recorded as injuring the foliage of young orange trees.

Protostrophus noxius, sp. nov. (Plate xix, fig. 1).

©. Colour black, with dense sandy brown scaling above, variegated with some
paler and darker markings, and uniform sandy grey beneath; the prothorax with
usually a broad lanceolate darker median stripe and a broader pale lateral stripe ;
the scaling on the elytra brown, with a broad whitish stripe on interval 6 and usually
with a macular stripe of alternating dark brown and pale spots along striae 1,
2 and 5.

Head separated from the rostrum by an abbreviated and deeply curved stria,
the forehead smooth, with a single deep furrow, and densely clothed with over-
lapping scales and raised scale-like setae ; the eyes moderately convex, deepest about
the middle, projecting strongly backwards, but the hind edge of the orbit not
produced. Rostrwm a little more than half as long as the basal width (5: 9), very
strongly narrowed from base to apex, with the sides slightly smuate; the upper
surface quite flat, with only a very shallow indistinct impression in the middle of
the apical half; the mandibles squamose, the mandibular appendages gently
curved, simple and tusk-like, flattened ventrally and with the surface partly striate
longitudinally. Antennae with joint 1 of the funicle about as long as the next
three, joint 2 about as long as broad, the remainder broader than long. Prothorar
twice as broad as long, a little broader than the base of the elytra, the basal margin
strongly arcuate, the apical margin somewhat narrower and trisinuate, the lateral
margins subdenticulate and gently rounded, broadest at the middle; the upper
surface strongly convex transversely and flat longitudinally, quite smooth and
evenly covered with dense scaling, without any trace of a furrow or carina. Slytra
broadly ovate, widest behind the middle, very broadly rounded behind, the basal
margin strongly sinuate, fitting closely to the prothorax, with the external angles
slightly projecting ; the extremely shallow striae bear rows of punctures that are
visible through the scaling, each containing a minute scale; the intervals broad,
almost flat, densely covered with almost circular overlapping scales equal to those
on the prothorax and each with a row of short recumbent scale-like setae. Legs
with the front tibiae moderately incurved at the apex, the inner surface with two
or three spine-like setae, the apical margin with a row of five or six stout teeth
and an additional one on the outer edge; the hind tibiae with two internal spines
near the apex, and the corbels densely scaly ; joint 2 of the front tarsi much broader
than long. Sternum with the mesepisterna separated from the elytra.

Length, 3°25-3°6 ; breadth, 2-2°2 mm.

ORANGE FREE State: Clocolan, 20.vu.1915.

Described from three specimens.

SOME INJURIOUS SOUTH AFRICAN WEEVILS. 275

Somewhat like a small P. funestus, Pér., but in the latter species the prothorax
is distinctly broader in relation to the elytra, its posterior angles are sharp right
angles and its anterior ones acute ; and the elytra fit less closely to the prothorax.

Found feeding on young wheat.

Protostrophus instabilis, sp. nov. (Plate xix, fig. 2).

§ 2. Colouring variable: piceous, with dense grey scaling; the prothorax
with a more brownish median lyre-shaped patch and two indistinct sinuous stripes
on each side of the same colour ; the elytra with interval 1 always fawn coloured,
elsewhere irregularly mottled with dark brown; usually pale spots stand out
rather conspicuously on intervals 3 and 5 behind the middle, and sometimes these
are linked up to form a continuous oblique pale band ; occasionally the elytra
are distinctly striped, intervals 3 and 5 being almost entirely dark brown and the
rest (except 1) light grey.

Head separated from the rostrum by a slightly sinuate stria that reaches the
sides, the forehead almost flat, with a single deep furrow and the sculpturing
quite hidden by the scaling, and set with obliquely raised flattened setae ;
the eyes not very prominent, deepest far behind the middle, the posterior
edge of the orbit not projecting. Rostrum distinctly broader than long, not
very strongly narrowed in front, the sides almost straight and with a slight
rounded projection above the apex of the scrobe; the upper surface nearly flat,
with a shallow median depression in the apical half, and without any distinct carina ;
the mandibles not squamose. Antennae with joint 1 of, the funicle much longer
and broader than 2, which is as long as 3 and 4 together, 3-7 transverse, 7 as broad
as 1. Prothorax about twice as broad as long, broadest at the middle, the sides
very strongly (subangulately) rounded, the base slightly arcuate, with the basal
angles rounded off, no broader than the apex, which is truncate ; the upper surface
flat longitudinally and smooth, the sculpturing completely hidden by scaling, except
for a shallow median stria, which is abbreviated at both ends. LHlytra oblong-ovate,
the sides obtusely angulated near the base and from there subparallel to beyond
the middle, broadly rounded behind, with the base very shallowly sinuate and
the basal angles not projecting; the shallow striae contain separated punctures
that are clearly visible through the scaling ; the intervals broad, almost flat, finely
coriaceous, and densely clothed with almost circular overlapping fluted scales
(of the same size as those on the prothorax) and each with a single row of raised
brownish scaie-like setae, which become longer and more erect behind, a few across
the top of the declivity beg white. Legs densely clothed with scales and sub-
recumbent scale-like setae; the front tibiae only slightly incurved at the apex,
with two spines on the inner face, and an apical row of three small and three stout
teeth and two stout ones on the outer edge; the hind tibiae with three spines
internally on the apical half and the corbels densely squamose. Sternum with
the mesepisterna narrowly separated from the elytra.

Length, 3-3°5 mm.; breadth, 1°75-2 mm.

TransvaaL: Nelspruit, 25.1x.1918.

Described from ten specimens.

(637) E

276 GUY A. K. MARSHALL.

This species differs from typical forms of the genus in the following points :—
the obliteration of the longitudinal impressions on the lower surface of the rostrum ;
the extension of the transverse cephalic stria to the lateral margin ; the narrowness
of the basal margin of the prothorax ; and the humeral angulation on the elytra.
in all these characters 1t agrees with Bradybamon, Mshl. (Ann. Mag. Nat. Hist.
(9) 1, 1919, p. 9), from which however it differs abundantly in the structure of the
epistome, mentum, eyes, venter, etc.

Recorded as injuring the foliage of young orange trees.

Eremnus horticola, sp. nov. (Plate xix, fig. 3).

g§ 9. Colour black, densely clothed with sandy grey scaling, irregularly mottled
above with brown markings.

Head with the rugose sculpturing entirely concealed by the scaling; the forehead
flattened, as broad as the base of the rostrum and with a central fovea. Rostrum
parallel-sided at the base, then gradually widening to the apex, very deeply com-
pressed behind the scrobes, leaving a raised dorsal area with perpendicular sides,
which rapidly narrows from the apex almost to a point at the base; the dorsal
area broadly impressed in the apical half and rugosely punctate, the punctures
hidden by almost circular large overlapping fluted scales. Antennae with the
scape gently curved just below the middle and slightly exceeding the hind margin
of the eye; the funicle with joint 2 half as long again as 1, joints 3-6 gradually
diminishing, and 7 as long as but broader than 4; the club as long as the three
preceding jomts. Prothorax half as broad again as the median length, the sides
rounded, broadest before the middle, deeply constricted and broadly sulcate
transversely near the apex; the basal margin slightly arcuate, the apex broadly
sinuate dorsally, the postocular lobes broad but rather short ; the upper surface
uneven, the central area granulato-punctate, with four foveae forming the corners
of a transverse parallelogram and with a shallow central furrow towards the base ;
on each side of the central area a very broad and deep furrow, bounded externally
by a sharp carina, beyond which the surface is rugosely granulate ; all the granules
covered with scaling. Hlytra rather broadly ovate in 9, narrower in g, broadest
about the middle and obtusely pointed behind ; the dorsal outline only slightly
curved, the posterior declivity very steep and curving inwards near the apex,
which is hidden from above (especially in the 9); the basal margin gently sinuate
and vertically truncate, almost marginate for a short distance on each side of the
suture ; the striae shallow and with separated shallow punctures, which are partly
hidden by the scaling ; the intervals moderately convex, the alternate ones higher
at the base, and the suture obtusely elevated on the declivity, less so in 3 than in
©; the scales much smaller than those on the prothorax, almost circular, fluted
and slightly overlapping, the short recumbent scale-like setae irregularly disposed.
Legs with dense brown scaling and pale recumbent scale-like setae ; the femora
with a pale band on the thickest portion and each with a small tooth.

Length, 7-8 mm.; breadth, 3:25-5 mm.

ORANGE FREE StaTE: Bloemfontein, 1.xu.1916.

Described from eight specimens.

Found feeding on dahlias and chrysanthemums.

BIE, laws, IyaSiascisi, Wom, ><, IDAinar 8). PLATE XIX.

INJURIOUS SOUTH AFRICAN CURCULIONIDE.

1. Protostrophus noxius, Mshl., sp.n. 2. Protostrophus instabilis, Mshl., sp.n.

38. Eremnus horticola, Mshl., sp.n. 4. Protostrophus planatus, Mshl., sp.n.

277

A CONTRIBUTION TO KNOWLEDGE OF THE 'TABANIDAE OF

PALESTINE.
By Masor E. E. Austen, D.S8.0O.

At the present moment our knowledge of the Dipterous fauna, including the
Tabanidae, of Palestine is little better than complete tgnorance, consisting as it
does solely of a few isolated records of species—chiefly mosquitos. The following
paper, which it is intended shall be succeeded by others dealing with other
families of Diptera, based upon material collected by the author in Palestine during
the recent campaign against the Turks, is a modest attempt to lift one small corner
of the veil of obscurity.

While, as in Syria according to previous writers,* the dominant Dipterous family
in Palestine is certainly the Bombyliudae,} the Tabanidae, of which twenty-one
species are recorded or described in the ensuing pages, are by no means poorly
represented. The extent to which the Dipterous fauna of Palestine is identical
with that of Syria has yet to be determined, and the present writer unfortunately
had no opportunity of pursuing his investigations north of Acre. So far, however,
as the Tabanidae—the only family yet worked out—are concerned, Loew’s
remarks (Verh. z.-b. Ges. Wien, Bd. vu, 1857, p. 79) with reference to the Diptera
of Syria would certainly apply to those of Palestine. “The Syrian Dipterous
fauna,’ wrote the author in question, “is on the one hand so closely connected
with that of Europe, and on the other exhibits so many peculiar species, that it
merits a high degree of interest.” |

With the exception of three or four examples previously contained in the
British Museum (Natural History), the material, amounting to one hundred and
forty-four specimens, upon which the following paper is based was collected by
the author, with occasional assistance from friends, during the Palestine Campaign
of 1917-18. So far as possible, special attention was devoted to the collection
ot Tabanidae in 1918, owing to the prevalence early in that year of a rumour that,
during 1917, the Turks in the Jordan Valley had lost a very large number of camels
from surra, a form of trypanosomiasis of which the causative agent, Trypanosoma
evansi, 18 at present believed to be capable of dissemination by more than one
species of Tabanus. In view of the imperative necessity of maintaining a large
mounted force in the Lower Jordan Valley until the date of the final general
advance, the report in question at first caused some apprehension at General Head-
quarters, Egyptian Expeditionary Force. Fortunately, however, any fears that
may have been entertained were not justified by the result, while the few cases

* Bezzi (Brotéria, Ser. Zool., Vol. viii, fasc. 2, 1909, p. 37), writing on Diptera from
Syria, remarks :—‘“‘ The peculiar character of the fauna is to be deduced from the
multitude of Bombyliidae: this family is preeminent among all others owing to the
beauty and wealth of its species, and constitutes 21:6 per cent of the entire collection.”
In a foot-note on the same page Bezzi points out that, among thirty-seven species of
Syrian Diptera recorded by Loew (‘‘ Nachricht tiber syrische Dipteren,” Verh.z.-b.
Ges. Wien, Bd. vii, 1857, pp. 79-80), the Bombyliidae amount to 56°7 per cent. :

+ The material belonging to this family collected by the author in Palestine during

1917-18 consists of 289 specimens—nearly twice as many as those obtained belonging
to any other family ; the precise number of Tabanidae brought back was 136.

(637) 2

278 ERNEST E. AUSTEN.

of surra diagnosed among animals belonging to the Camel Transport Corps failed
to spread, and were by no means certainly contracted in Palestine.

While, somewhat curiously, no species of Pangonia* was met with, and the large
numbers of specimens of Chrysops encountered all proved to belong to the same
species, the genus T'abanus, having regard to the relative smallness of the country
concerned, was found to be decidedly rich im species. Of this genus, examples
of no fewer than sixteen species were collected, and of these seven, or the high
proportion of 43°7 per cent., are apparently new. With one exception, the
remainder are more or less well known European forms, some of which have already
been shown to occur further afield, as in North Africa, Asia Minor, or Persian
Baluchistan. The applicability of Loew’s generalisation, already quoted, con-
cerning the Diptera of Syria, when transferred to those of Palestine, at least as
represented by Tabanidae of the genus Tabanus, is therefore obvious.

Attention is directed to the pursuit of a fast travelling motor car by specimens
of two species of Tabanus (cf. pp. 298, 300), one of which proved to be new. Some
three months later, a second instance of the same phenomenon was seen under
precisely similar circumstances, within a few miles of the same spot, the species
concerned in this case being probably Tabanus eggert, Schin., though, since no
specimen was caught on this occasion, the identification is uncertain. In
Nyasaland, tsetse-flies of the species known as Glossina morsitans, Westw., have
been observed to pursue and keep pace with a motor cycle travelling at high speed,
and the attraction exerted upon tsetse-flies by moving animals has been mentioned
by more than one observer. So far, however, as the writer is aware, the pursuit
of a motor car by Tabanids has not previously been recorded. It should be noted
that the chase was confined to short distances, so that the chance establishment
of a motor route passing through a local Tabanid area is unlikely to have much
effect in extending the distribution of species.

It is only necessary to add that the whole of the material referred to in the
following pages, including the types of all new species, is in the British Museum
(Natural History).

PANGONIINAE.
Genus Chrysops, Meigen.

In so far as a definite conclusion is warranted by the author’s necessarily limited
observations, the form mentioned below would appear to be the only representative
of its genus in Palestine. This paucity of species, however, is more than made
up for by the relative ubiquity and absolute aggressiveness of the exemplar con-
cerned.

Chrysops punctifera, Lw. (fig. 1).

Two 99, Wadi Ghuzze, nr. Tel el Fara, 26.vi.1917, biting horse; 1194, 6292,
same locality, 23.vii.1917, the $g numerous, resting on rocky bank of Wadi and
on plants above pool, many 929 ovipositing on reeds overhanging water; 3g,
499, same locality, 8.vili.1917 ; 14, Deir el Belah, 8 miles 8.-W. of Gaza, 31.vil.
1917; 1g, 19, Wadi Ghuzze, near El Gamli, 12.ix.1917; 299, Wadi Ghuzze,

* Bezzi (loc. cit., p. 41, Tab. ix, fig. 32) describes and figures, under the name
P. sulcata, a new species of Pangonia from Syria.

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 279

nr. Tel el Fara, 14.1x.1917, biting author ; 1g, Wadi Ghuzze, nr. Hl Shellal, 19.ix.
1917; 19, Wadi Ghuzze, nr. Tel el Fara, 28.1x.1917, biting author’s arm; 14,
Wadi Sukkereir, 10 miles 8. of Jaffa, 13.1v.1918, gd alone seen; 19, Jericho
Plain, nr. Kasr el Hajlah, 28.v.1918 ; 19, Latron, 18.vi.1918, biting author’s hand ;
19, left bank of R. Auja, nr. Sheikh Muannis, 5 miles N. of Jaffa, 29.vii.1918,
biting author’s arm; 19, Wadi Ishkar, R. Auja, nr. Ferekkiyeh, 8.viii.1918 ;
19, Wadi Barideh, nr. Sarona, 21.vii.1918, biting author’s hand; 19, same
locality, 2.ix.1918, biting author’s arm; 29° (one biting author’s arm), marshes
bordering R. Kishon, nr. Haifa, 2.x.1918. In addition to the foregoing, the
Museum already possesses a 2 of this species from the “ papyrus marsh above
Baheiret el Hileh (Waters of Merom), Upper Jordan Valley, 20.vii.1901 ” (the
late Dr. J. Cropper), with the collector’s field-note “ only amongst papyrus.”

In Chrysops punctifera the sexual colour-dimorphism, which is so striking a
feature in the case of the majority of the palaearctic representatives of its genus,
is especially pronounced, and at first sight it is difficult to believe that the
predominantly ochreous or olive-ochreous* female, with its brown-banded wings,
can possibly be conspecific with the mainly black and largely black-winged male.
This exceedingly aggressive and bloodthirsty fly would appear to be generally
distributed throughout the entire country, and to occur wherever the necessary
conditions for breeding exist. In July 1917, females were seen busily ovipositing

Fig. 1. Head of Chrysops punctifera, Lw., °9,
in profile, showing eye-markings. \ about 14.

in the Wadi Ghuzze. The elongate eggs, which are white when newly laid and
subsequently become opalescent brownish, are deposited in countless myriads,
closely packed in sheets, on leaves of reeds overhanging pools of slowly flowing
or stagnant water. The female of Ch. punctifera, wherever and whenever
encountered, seems almost always ready to bite both man and domestic animals,
and, so far as the author’s personal experience goes, the species is more aggressive
than any other Tabanid met with in Palestine.

* For names and illustrations of colours used for descriptive purposes in the present
paper, see Ridgway, “Color Standards and Color Nomenclature’ (Washington, D. C.
Published by the Author. 1912).

280 ERNEST FE. AUSTEN.

The range of the species extends at least into Syria. The type of Chrysops
punctifera, Lw., which was originally described (Neue Beitrdge zur Kenntnss der
Dipteren, iv, 1856, p. 24) from the male alone, was obtained at Beirut. Sub-
sequently (Verh. k.-k. zool.-bot. Ges. Wien, vii, 1858, p. 633) Loew, without
giving precise indication of habitat, stated that he possessed nine males and three
females of the species from “ Syria,’’ and added some notes on the female sex.
Quite recently, during the preparation of this paper, the British Museum (Natural
History) has received three males of Ch. punctifera, taken in Beirut Marshes,
11.iv.1919, by the donor, Lt.-Col. HE. P. Sewell, C.M.G., D.S.O., R.AMOC.

TABANINAE.
Genus Haematopota, Meigen.

The twenty-eight specimens of this genus obtained by the author in Palestine
appear to belong to four species, none of which can be identified with any of those
already known. They are accordingly described as new in the following pages.
and an attempt has been made to construct a diagnostic table for their distinction.
Since, however, all but three of the specimens are apparently conspecific, while
the remainder, representing as many species, are exclusively females, it has been
impossible to draw up a synoptic table for both sexes, and the subjoined key
should be regarded as purely tentative.

Specimens of Haematopota were met with for the first time on 22nd April 1918,
on horses near Jericho. None were encountered during the sprmg and summer
of 1917, and their apparent non-occurrence in the Wadi Ghuzze is somewhat remark-
able. A noteworthy instance of Syrphid-like hovering in the air by males of
Haematopota sewelli, sp. n., is recorded on page 285.

Key to Species described below.
1(2). First jomt of antenna, viewed from side, with a deep preapical groove* on
upper surface ; femora largely greyish fawn-coloured or greyish cinnamon
sewelli, sp. n. [1].
2(1). First jomt of antenna without such a groove; femora entirely or mainly

AG
3(4). First jomt of antenna, viewed from side, not at all swollen. Small or very
small species... Pe monuscularia, sp. Nn. [2].

4(3). First jomt of antenna, ed few als distinctly or strongly swollen.
5(6). First jomt of antenna, viewed from side, strongly swollen, its greatest vertical
diameter greater than or fully equal to half the length of the joint ;

dorsum of abdomen with a broad, grey, border on each side
minuscula, sp. n. [3].
6(5). First jomt of antenna, viewed from side, moderately swollen but elongate,
its greatest vertical diameter considerably less than half the length of
the jomt; dorsum of abdomen with a double row of grey spots, but
without grey lateral borders .. at i ennominata, sp. n. [4].

* Sometimes obscured by hair, in which case it may be difficult to distinguish. If,
however, the head be examined from the left side, with a hand-lens magnifying about
5 diameters, the groove, if present though indistinguishable in the left antenna, can
generally be seen distinctly on examining the inner side of the first joint of the right
antenna.

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 281

1. Haematopota sewelli, sp n. (figs. 2, 3).

3 9.—Length, $ (8 specimens) 11 to 12°2 mm., 9 (17 specimens) 9 to 12°5 mm. ;
width of head, 94 to 4:4 mm., 9 3 to 4 mm. ; width of front of 2 at vertex 1:2 to
15 mm. ; length of wing, 3 8°5 to 9:4 mm., 9 74 to 10 mm.

Blackish (3) or blackish brown, greyish olive, or deep greyish olive () species,
allied to and superficially closely resembling H. pluvialis, Linn.: distinguished in
the 3 by the har on the upper and under sides of the first antennal joint being shorter
and less fine, while the proximal half of the jount 1s enturely olive-grey pollinose, instead
of the pollinose covering being confined chiefly to the upper and inner side of the proximal
half or two-thirds ; by the femora of all three pairs of legs in the 3, or at least those
of the middle pair, having a complete or incomplete tawny or ochraceous-tawny band,
instead of all the femora beng unrcolorous dark neutral grey, vron grey, or dark olwve-
grey ; by the first antennal joint in the 2 beng somewhat longer, usually slightly less
swollen in proportion to ats length, duller (mouse grey, dark neutral grey, or connamon-
drab pollinose instead of partly shining black), and, if anything, more distinctly
constricted before the tup, but with rts upper surface not humped up before the con-
striction ; and by the femora in the 2 beng largely greyish fawn-coloured or greyish
cimnamon, instead of uniformly grey (deep olive-grey or neutral-grey).

Head: front in 9 smoke grey pollinose, clothed partly with short, erect, blackish
hair, partly with pale yellowish hair, vertex with a pair of admedian, mouse grey
or deep mouse grey blotches ; frontal triangle in $ greyish olive, both sexes between
and above bases of antennae with a large, dead black or blackish brown, hexagonal
or quadrate spot, which in 9 is bounded above by the frontal callus; face in 3
pallid neutral grey pollinose, in 2 whitish pollinose, in both sexes speckled with
blackish brown on each side, in 3 clothed on each side with fairly long blackish
brown hair, in @ with a pair of small but fairly conspicuous, admedian, blackish
brown spots below the antennae, and clothed with whitish hair; jowls whitish
pollinose, clothed with whitish hair, part of which in 3 sometimes has a yellowish
tinge ; occiput olive-grey, light olive-grey, or deep olive-grey, its upper margin
fringed in ¢ with a mixture of fairly long, erect, blackish and yellowish hair, and
in @ on each side of front with short yellowish hair; frontal spots in 9 blackish
brown or black, median spot usually small but distinct, lateral spots either roughly
triangular or more or less rounded, not in contact with eyes; frontal callus in 9
shining black, of moderate depth, its lateral extremities somewhat deeper and
narrowly connected below with each eye, its upper margin produced in middle
le into a small, upwardly directed angle; eyes in ¢ densely clothed with fine,
erect, drab-grey hair, hairy covering of eyes in 2 very short, but distinguishable
under a hand-lens magnifying about 5 diameters, larger facets in § contrasting
conspicuously with those forming remainder of surface, but not coarse ; proximal
segment of ¢ palpy mouse grey or deep mouse grey, clothed with long whitish
hair, terminal segment fairly large, ovate pyriform, pointed at distal extremity,
greyish pollinose pinkish buff, usually with an ill-defined mouse grey blotch on outer
side near base, clothed with whitish hair, which on upper margin beyond base is
mixed with fine, erect, black or blackish hair ; palpi in Q greyish pollinose pinkish

buff, proximal segment sometimes with a mouse grey spot on outer side near

282 ERNEST E. AUSTEN.

distal extremity, terminal segment moderately (in some specimens considerably)
thickened towards base, proximal segment and base of terminal segment below
clothed with fairly long whitish hair, terminal segment otherwise clothed on outer
side with short, appressed, glistening pale yellowish hair, usually mixed on upper
surface except at base, or at any rate towards distal extremity, with short,
appressed, black hairs; first joint of antennae in ¢ considerably swollen, when
viewed from above truncate ovate or elliptical ovate in outline, its distal half
shining black ; first joint of antennae in ¢ clothed below with yellowish white
hair and above with black or blackish hair, except proximal portion of pollinose
area at base, where hair is yellowish; second joint of antennae in ¢ blackish
mouse grey, clothed with short and rather coarse black hair; expanded portion
of third joint of antennae in ¢ of moderate depth as viewed from side, more or
less greyish at base, cinnamon in middle, and dark brownish towards distal
extremity, on inner side with a vertical band of minute, appressed, black hairs
between blunt angle on upper border and lower margin ; annulate portion of third
jomt of antennae in ¢ blackish brown, ending bluntly ; first joint of antennae
in @ thick, cylindrical, clothed with semi-appressed, short black hairs, lower surface,
except distal extremity, clothed with longer whitish hair; second and third joints
of antennae in 9 asin g. Thorax: dorsum blackish brown (¢) or dark greyish
olive ( ), in each case with pale smoke grey markings of the usual type; in 9 these
markings are but little conspicuous, and consist of a narrow median line, commencing
on fore border but disappearing before reaching transverse suture, and a pair of
admedian stripes, broader on anterior border but rapidly tapering off, interrupted
before reaching transverse suture and reappearing behind latter in shape of a pair
of more or less triangular spots resting on suture ; markings in Q similar to foregoing,
but in some specimens median line is continuous and reaches prescutellar groove,
usual crescentic marks in front of latter are distinguishable, there is a broader
pale smoke grey stripe on each side extending from inner extremity of depression
at end of transverse suture to basal angle of scutellum, and lateral border
on each side, from humeral callus to depression at end of transverse suture,
and from latter to basal angle of scutellum is pale smoke grey ; humeral calli in
©, and swelling occupying depression at each end of transverse suture in same sex
smoke grey ; scutellum in each sex agreeing in coloration with ground colour of
remainder of dorsum ; pleurae and pectus in both sexes pale smoke grey pollinose
and clothed with whitish hair (latter on mesopleurae in g sometimes appears to
have a yellowish tinge); dorsum, including scutellum, clothed in 3 with fine,
erect, yellowish hair, and in 9 with short, appressed, glistening yellowish hair,
which sometimes, especially towards lateral borders, appears to have a pinkish
tinge. Abdomen: dorsum in ¢ dull black, posterior angles of first (visible) tergite
and lateral extremities of two following tergites ochraceous-buff, ochraceous-tawny.
or cinnamon, posterior margins of second and following tergites (in case of second,
third, and fourth segments, and sometimes of fifth segment also, more or less
distinctly expanded in middle line into a wide flattened triangle) dark olive-buff :
second or third and following tergites to sixth inclusive in ¢ each with a pair of
more or less distinctly marked greyish olive or deep greyish olive, rounded
admedian spots (most distinct on fourth and two following segments), resting on

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 283

front margin in each case but usually not reaching or not quite reaching hind border,
a certain number of the tergites indicated, at least the fourth to the sixth, also
with a narrow median stripe of same colour as the spots; seventh tergite in d,
except posterior and lateral margins, deep olive or dark olive, without spots or
median stripe, lateral margins and posterior angles of third and following tergites
of same colour as posterior margins; dorsum in ¢ clothed mainly with blackish
hair, which in central region of first three and sometimes the following tergites
also may be largely mixed with glistening yellowish or cream-buff hair, posterior
angles of first and lateral extremities of two following tergites, posterior angles
of fourth to seventh tergites inclusive, hind margins of sixth and seventh tergites
and junction of median stripe with hind margin on the four preceding tergites

Fig. 2. Haematopota sewelli, Austen: a, head of g from above; a’,
antenna of ¢ in profile; b, head of Qin profile. x 14.

aoe

Fig. 3. Haematopota sewelli, Austen,
wine Of O27 XxX G.

clothed with bright, glistening, cream-buff hair, downwardly curved lateral
extremities of first tergite clothed with whitish hair; venter in 3 mouse grey or
deep mouse grey (seventh segment dark mouse grey), a larger or smaller area at
lateral extremities of second and third segments light greyish pollinose ochraceous-
salmon, hind margins of second and following segments light ochraceous buff ;
venter in ¢ clothed with fine, whitish or yellowish hair, seventh segment, except
lateral extremities, with usual coarse, erect, black hair, central area of sixth segment’
usually with finer black hair ; dorsum in @ blackish brown or dark blackish brown,
each segment with a pair of relatively large, greyish olive or deep greyish olive,
rounded, admedian spots, resting on front margin and not reaching hind border ;
Jatter, as well as posterior angles of each segment, a median, sub-triangular fleck

284 ERNEST E. AUSTEN.

on first (visible) tergite, and a median longitudinal stripe on each of the following
tergites, often indistinct or scarcely developed on seventh tergite but on the others
expanded laterally at its junction with hind border (on second tergite so much
expanded posteriorly that it has the shape of a median triangle), of same colour
as the paired spots; lateral extremities of first two or three tergites pale smoke
grey or deep gull grey, those of first (visible) tergite sometimes with a more or less
distinct pinkish cinnamon fleck on each side; ground colour of posterior angles
and hind margins of second or third and following tergites sometimes more or less
distinctly cinnamon-buff or cream-buff; dorsum clothed on darker areas, as also
on spots, with minute, appressed, black hairs, elsewhere clothed as a rule with
appressed, glistening, cream-bufi or cream-coloured hair; venter in Q light greyish
olive, smoke grey or pale smoke grey pollinose, second and following segments
each usually with a darker blotch in midde line, and with cream-buff or cream-
coloured hind borders; venter in 2 clothed with minute, appressed, glistening,
yellowish or cream coloured hairs, last two segments mainly with erect black hairs, a
few black hairs in centre of posterior half of antepenultimate segment also. Wangs
(fig. 3.) drab or light sepia-coloured ; usual three rosettes, pale mark in marginal cell at
distal extremity of stigma, and sinuous mark or transverse streak just before apex
of wing (starting from costa at anterior distal angle of first submarginal cell) well
developed and sharply defined; inner margin of proximal rosette as a rule
perfectly distinct from (not fused with, as is often the case in the wing of
H. pluvialis, Linn.) other pale marks crossing basal and anal cells ; distal marginal
angles of all posterior cells, or at least of second to fifth inclusive, often con-
spicuously pale; zig-zag pale mark running across proximal extremity of distal
third of anal and axillary cells, as also loop in base of latter, well developed;
discal cell with two pale transverse marks (portions of proximal and median rosettes)
at each extremity ; veins or portions of veins surrounded by centres of the three
rosettes more or less distinctly infuscated ; no conspicuous dark blotch beneath
stegma ; latter mummy brown or sepia-coloured, well developed, its proximal
extremity pale. Squamae light drab (3), with darker (drab-coloured) borders,
or pale smoke grey (2), with ivory yellow borders. Halteres cream-coloured,
knobs clove brown at base above and below. Legs: coxae in § neutral grey,
clothed with long whitish hair ; femora in § deep mouse grey, or dark mouse grey,
with a more or less conspicuous, complete or incomplete, tawny or ochraceous-
tawny band round middle, present at least on middle pair, but sometimes absent
in case of front femora and scarcely distinguishable on hind pair; femora in
clothed with fine whitish hair, which is long behind and below; distal half of
extensor surface of front femora in 3, and extreme tips of same surface in hind pair
clothed for-most part with minute, appressed, black hairs; front tibiae in g
blackish brown, with a conspicuous cream-buff band just beyond base, and with
distal two-thirds conspicuously or distinctly swollen; middle and hind tibiae
in ¢ dark brown, each with two cream-buff bands, hind tibiae thickened ; tibiae
in ¢ clothed with black or blackish hair, which on middle and hind pairs is fairly
long and fine, pale bands on all tibiae, as well as anterior surface of hind pair except
distal extremities clothed with glistening cream-coloured or Naples yellow hair ;
front tarsi in ¢ black, middle and hind tarsi in same sex dark brown, first joints

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 285

of middle and hind tarsi, except distal extremities, cream buff, hairy covering
of all tarsi black ; front coxae in Q greyish pollinose cream-coloured or light drab,
posterior surface and distal extremities usually neutral grey or light neutral grey,
middle and hind coxae in @ pale neutral grey, all coxae clothed with whitish hair,
longest in case of front pair, though front coxae sometimes show shorter, glistening,
appressed, cream-coloured hairs on outer surface; femora in Q greyish fawn-
coloured or greyish cinnamon, their extremities, at least distal extremities, more
or less mouse grey or deep mouse grey ; hairy covering of femora in Q as in J,
but shorter ; tibiae and tarsi in 9 as in 4, but hairy covering shorter.

Jafia district ; Lower Jordan Valley ; near Tul Keram; and Acre.

Type of ¢ and five para-types belonging to the same sex, near Jerisheh, 5 miles
N.-E. of Jaffa, 29.iv.1918 ; hovering at 9.0. a.m., poised in the air like Syrphids,
in front of a tent on the outside of which two 99 of the same species were resting.
Type of 9, same spot as foregoing, 5.v.1918. Additional para-types as follows :—
244, same spot as before, 26.1v. and 5.v.1918 respectively ; 999, Ain es-Sultan,
near Jericho, 22.iv.1918, on horses ; 499, same spot as that at which the typical
specimens were taken, 29.iv. and—.v.1918; one 9, W. bank of R. Jordan, near
Jericho, between El Ghoraniye and the Dead Sea; one Q, Kerkar (between Tul
Keram and Zimmarin), 28.1x.1918, biting author’s arm; 19, marsh near Acre,
close to mouth of Nahr Naaman (R. Belus), 5.x.1918—one of two specimens, one
of which bit author’s arm.

As a small acknowledgement of much kindness shown to the author, the species
just described has been named in honour of his friend Lieut.-Col. (temp. Col.) HE. P.
Sewell, C.M.G., D.S.0., R.A.M.C. (late D.D.M.S., XXIst Corps, HE. E. F.). So far as
the experience of 1918 warrants such a conclusion, Haematopota sewellt would
appear to be the commonest representative of its genus in Palestine. The remark-
able habit of the males, referred to above, of remaining poised in the air like Hover-
flies, although new to the present writer, would seem to be a generic character,
since Schiner (Fauna Austriaca. Die Fliegen (Diptera), I. Theil, 1862, p. 38),
at the end of his notes on the genus Haematopota, remarks :—“ The males too
hover in the air in the morning and towards evening.”

2. Haematopota minuscularia, sp. n. (fig. 4, a).

O.—Length (1 specimen) 6°6 mm.; width of head 2°6 mm.; width of front at
vertex 1 mm.; length of wing 5°8 mm.

Small, greyish black species, with somewhat elongate antennae, of which first joint
as cylindrical, not swollen, and, when viewed from side, without a noticeable, groove-lke
constriction before its distal extremity; wings light drab, with all three rosettes
well marked, adjacent margins of muddle and proximal rosettes fused together where they
cross discal cell, and with a pale border running along hind margin ; femora, except
extreme tups, which are narrowly warm buff, entirely grey, those of second paar of legs,
at proxumal ends of distal halves, each with a faint indication of a greyish fawn-
coloured band.

Head: front pale neutral grey, lateral margins, a narrow border encircling each
frontal spot, and a narrow median line from vertex to median frontal spot paler

284 ERNEST E. AUSTEN.

(pallid neutral grey); face and jowls pallid neutral grey pollinose, sides of face
punctate in usual manner, but pair of dark spots on central region of face, seen
in H. sewell1, Austen, 2, absent or but faintly indicated ; occiput pale neutral grey ;
with exception of a few minute black hairs in region of vertex, and on or near lateral
frontal spots, hair on head entirely white ; frontal spots dead black, median spot
small but distinct, lateral spots large, circular, not in contact with eyes ; frontal
callus shining black, of moderate and practically uniform depth, narrowly in contact
with eye on each side below, its upper margin, at least in case of type, not notice-
ably produced into an angle in middle line ; usual spot in middle line below callus
dull black, wider than deep, not conspicuous ; eyes clothed with short whitish hair ;
proximal joint of palpz neutral grey, clothed with long white hair, terminal joint
pale pinkish buff, not much swollen towards base, clothed with longer white hairs
on proximal portion of under surface, and on outer side with minute, appressed,
glistening creamy-white hairs, mixed towards distal extremity with minute black
hairs ; first jomt of antennae olive-grey pollinose (shining black on outer and inner
sides at distal extremity, which, though somewhat narrower than remainder of
jomt, is not marked off by a conspicuous constriction), clothed above and on

Fig. 4. a, Haematopota minuscularia, Austen; 6b, H. miéinuscula, Austen ;
c, H. innominata, Austen. Heads of 99 in profile. x 14.
outer side with very short, pale yellowish hairs, and below with longer white hair,
second joint of antennae greyish brown, clothed with minute black hairs, third
joint clove-brown, paler (cinnamon-brown) at extreme base, expanded portion
as viewed from side not very deep, roughly elliptical in outline, longer than
annulate portion. Thorax: dorsum blackish brown, clothed (including scutellum)
with short, fine, yellowish or yellowish white hair, and marked with six, more
or less distinct and continuous, pale olive-grey, longitudinal stripes, the two outer
stripes on each side fused together in front of transverse suture, and behind latter
much broader than remainder ; scutellum with extreme base of outer border on
each side, and a more or less indistinct median longitudinal stripe on upper surface
pale olive-grey ; pleurae and pectus pale olive-grey, clothed with long white hair.
Abdomen: dorsum blackish brown, seventh tergite mouse grey, lateral borders
and hind margins of first six tergites, as well as, on each segment from second to
sixth inclusive, a median, forwardly directed triangle, based on hind margin and
reaching or almost reaching front margin, and a pair of rounded, admedian spots,
resting on or close to anterior margin, pale olive-grey ; dorsum clothed with short,

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 287

appressed, whitish hair, except on blackish brown areas adjacent to median
triangles on second to fourth segments inclusive, where hairy covering is black or
blackish, and more erect ; venter pale neutral grey, clothed with minute, appressed,
creamy white hair, terminal segment with usual longer, coarse, erect, black hairs,
extreme hind margins of second to sixth segments inclusive paler (pallid neutral
grey or ivory yellow). Wangs: light markings in distal half, at least in case of
type, somewhat moniliform ; pale border running along hind margin fairly deep,
on distal side of each posterior marginal cell deeper than on proximal, commencing
in second submarginal and terminating at base of axillary cell, in proximal half
of which it becomes very narrow ; owing to fusion of adjacent margins of proximal
and middle rosettes, discal cell enclosing only three pale marks; imner margin of
proximal rosette almost or quite in contact with other pale marks crossing basal,
anal and axillary cells ; sinuous mark at tip of wing not reaching pale hind border ;
stigma light sepia-coloured, paler at proximal extremity ; veins or portions of
veins surrounded by the three rosettes more or less distinctly infuscated.
Squamae semi-hyaline, borders ivory vellow or cream-coloured. Halteres ivory
yellow, knobs clove brown above and below. Legs: coxae and femora light
neutral grey, clothed with whitish hair; front tibiae brownish black, with a fairly
broad, creamy white band just beyond base, distal two-thirds conspicuously swollen,
pale band clothed with glistening, appressed, cream-coloured hairs, a few of which
pass over on to the brownish black portion, which is clothed otherwise with black
hairs ; middle and hind tibiae clove brown, clothed, except at distal extremities,
chiefly with glistening, appressed, yellowish hairs, and each encircled with two
cream-buft bands, distal extremities clothed with minute black hairs, hind
tibiae somewhat thickened ; front tarsi brownish black, clothed with black hair ;
middle and hind tarsi clove brown, first joints, except distal extremities, cream-
buff.

Ain es-Sultan, near Jericho, 8.v1.1918, biting author’s arm.

Apart from its small size, this tiny species can be distinguished, at least in the
female sex, from all its congeners described or mentioned in the present paper
by the shape of the first jomt of the antenna.

3. Haematopota minuscula, sp. n. (fig. 4, b).

@.—Length (1 specimen) 7°6 mm.; width of head 3 mm.; width of front at
vertex 1:2 mm. ; length of wing 6:2 mm.

Dorsum of thorax olive, with lateral borders and three narrow, incomplete,
longitudinal strupes pale smoke grey; dorsum of abdomen with black, median,
longitudinal strvpe (tapering and becoming olivaceous black towards distal extremity),
and with lateral borders broadly olwve-grey ; first joint of antennae strongly swollen ;
wngs light drab, with usual pale markings well developed, and those in discal cell as
an H. pluvialis, Zann.

Head: front deep olive-grey, lateral margins and a ring encircling each lateral
frontal spot pale smoke grey, a median stripe extending from margin of occiput at
vertex to median frontal spot, and a small ring encircling latter, light olive-grey ;
face and jowls pale gull grey, sides of face punctate in usual manner, but pair of
dark spots on central region of face, seen in H. sewelli, Austen, 9, and H. plumalis,

288 ERNEST E. AUSTEN.

Linn., 2, absent; occiput light olive grey, upper border darker (greyish olive) ;
with exception of a few minute black or blackish hairs on front above median
frontal spot, and between upper extremities of inner sides of lateral frontal spots,
hair on head entirely pale, that on front and fringe of occiput above with a distinct
yellowish tinge, hair on face and jowls white; frontal spots dead black, median
spot distinct, lateral spots large, roughly oval, not in contact with eyes ; frontal
callus shining black, of moderate and practically uniform depth, narrowly
separated from eye on each side, and with upper margin produced in middle line
into an upwardly directed angle; spot in middle line below callus dull black,
triangular or wedge-shaped in outline (narrower below) ; eyes clothed with minute,
pale hairs ; palpi light buff, clothed with white hair, proximal joint light neutral
grey on outer side towards base, terminal joint slender, elongate and acuminate,
as viewed from side only slightly swollen towards base; first jomt of antennae
shining black, neutral grey pollinose on upper surface of proximal half, elliptical
oval in outline as seen from above, as viewed from side thickest a short distance
before distal extremity, but wrthout groove-like constriction before latter exhibited
by both H. pluvralis, Linn., 2 and H. sewells, Austen, 2 (vide swpra) ; first joint
of antennae, except distal extremity, clothed above and on upper part of outer
surface with glistening yellowish hair, and below with longer white hair, distal
extremity clothed with black hair; second joint of antennae pale (light drab on
outer side, inner surface greyish pollinose light ochraceous buff), distal angles
clothed above and below with black hairs; expanded portion of third joint of
antennae clove brown (base paler—ochraceous tawny), elongate and tapering,
moderately expanded a little beyond base, annulate portion brownish black, shorter
than expanded portion. Thorax: dorsum, including scutellum, clothed with
partly appressed, partly erect, glistening pale yellowish or whitish hair of medium
length, hair on front border whitish, somewhat longer and more erect ; fore border
of dorsum pallid neutral grey ; median pale smoke grey longitudinal stripe extend-
ing from fore border to a poimt two-thirds of distance between transverse suture
_and prescutellar groove ; admedian pale smoke grey longitudinal stripes not reach-
ing transverse suture, but each reappearing behind latter in usual shape of a
triangular spot resting on suture; pale smoke grey lateral border on each side
partially divided behind transverse suture into two longitudinal stripes ; swelling
occupying depression at each end of transverse suture clothed below with fairly
long and fine blackish hair; pleurae and pectus pale smoke grey, and clothed
with fine white hair. Abdomen: black median longitudinal stripe on dorsum
not sharpely defined on first (visible) tergite, attaining its greatest breadth (one-
third of transverse diameter of segment) on second tergite and thereafter becoming
progressively narrower, scarcely distinguishable on seventh tergite, ground-colour
of which is mouse grey ; hind margins of second and following tergites, as well as,
on second to sixth segments inclusive, a more or less distinct narrow median line
light greyish olive ; inner edge of olive-grey lateral border on each side of each
segment formed by a large rounded grey spot, in contact with anterior border of
each segment and on first three segments more or less completely fused with lateral
border, while on last four segments it may be partly or completely separated
therefrom, in which case the spot is partly surrounded with blackish brown ;

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 289

dorsum clothed on basal angles and on extreme lateral margins of first four
segments with fine white hair, clothed otherwise on grey lateral borders with minute,
appressed, glistening yellowish hairs (longer on hind margins of last three segments,
and on spots on fourth to sixth segments inclusive partly or entirely replaced by
minute black hairs), black or olivaceous black areas clothed for most part with
minute black hairs; venter smoke grey, hind margins of second and following
segments ivory yellow, third to sixth segments inclusive and posterior portion of
second segment clothed with minute, appressed, yellowish hairs, central portion
of seventh segment clothed as usual with longer, coarse, erect, black hair. Wangs:
outer borders of rosettes somewhat broken up, apical smuous mark (pale transverse
streak just before apex of wing) well developed and (at least in the type) reaching
hind margin ; distal extremity of each posterior marginal cell occupied by a large,
somewhat wedge-shaped pale blotch, posterior margin of axillary cell, from distal
extremity to fused ends of zig-zag transverse pale marks, also pale, so that hind
margin of wing has a deep pale border, interrupted at proximal marginal angles
of posterior marginal cells (broken up into a series of blotches); stigma light
sepia-coloured, not very noticeable, paler at proximal extremity. Squamae milk
white, borders ivory yellow. Halteres ivory yellow, knobs with a seal brown spot
at base above and below. Legs: coxae and femora neutral grey, clothed with
white hair, distal extremities of front and hind femora above and below and of
middle femora above clothed with black hairs, which also cover rather less than
distal half of antero-superior surface of front femora; extreme tips of all femora
cream-buff, middle femora with a more or less distinct, complete or incomplete,
light greyish cinnamon band, occupying approximately middle third ; front tibiae
brownish black, with a fairly broad cream-coloured band just beyond base, distal
two-thirds conspicuously swollen, pale band clothed with glistening, appressed,
yellowish hair, front tibiae otherwise clothed with black hair; middle and hind
tibiae clove brown, middle tibiae with extreme base and two bands cream-
coloured, hind tibiae with extreme base and two bands cream-buff, hair on middle
and hind tibiae mainly black on dark areas, glistening cream-coloured or yellowish
on pale bands, inner surface of hind tibiae clothed mainly with minute, appressed,
glistening, yellowish hairs, extensor surface of middle tibiae sparsely fringed with
long, pale hairs ; front tarsi black, clothed with black hair, third and fourth joints
somewhat expanded; middle and hind tarsi clove brown or blackish brown,
clothed with minute black hairs, first joints, except distal extremities, cream-buff
or cream-coloured, second, third, and fourth joints also narrowly pale at extreme
base.

Jericho Plain, near Makhadet Hajlah, R. Jordan, 9.vi.1918, biting mule.

The predominance of pale hair on the front in the 9, the pale second antennal
. joint, and the markings of the dorsum of the abdomen and of the discal cell in the
wing—the latter agreeing with those seen in the wing of Haematopota pluvialis,
Linn.—will serve to distinguish the species described above, at least in the female
sex, from H. crassicornis, Whlbg., with which it agrees in the absence of a groove-
like constriction before the distal extremity of the first joint of the antennae.

So far as it is possible to judge from the original description of Haematopota
pallens, Lw., the typical specimens of which were obtained in Turkestan,

290 ERNEST E. AUSTEN.

H. minuscula resembles that species, especially in the markings of the abdomen and
wings, but is distinguishable, inter alia, by its smaller size, by the less elongate
antennae, and by the more swollen first antennal joint.

4. Haematopota innominata, sp. n. (fig. 4, c).

°.—Length (1 specimen) 9 mm.; width of head 3°25 mm.; width of front at
vertex 1°2 mm.; length of wing 7°5 mm.

Dorsum of thorax dark olive, with lateral borders and three narrow, complete,
longitudinal stripes pale smoke grey or smoke grey ; dorsum of abdomen black, each
segment from second to seventh inclusive with hind border, a more or less distinct median
strvpe (practically obsolete on third segment in the type), and a pair of rounded
spots (relatively small as compared with corresponding spots in H. sewelli, Austen,
©—vide supra, p. 283) light olive-grey ; first joint of antennae as seen from side
distinctly incrassate, roughly elliptical, somewhat narrower in middle, narrowed at
each extrenuty, but without a sharply marked preayical groove.

Head: front deep mouse grey, posterior angles, border of occiput, a narrow
border surrounding median frontal spot, and a median stripe connecting latter
with occipital margin pale neutral grey, extreme lateral margins and a narrow
ring surrounding each lateral frontal spot pale smoke grey ; face and jowls pale
gull grey or whitish, sides of face punctate in usual manner, pair of dark spots on
central region of face, seen in H. sewelli, Austen, 9, and H. pluvialis, Linn., 9,
faintly marked ; occiput pale neutral grey ; hair on front mainly dusky (blackish ~
brown or blackish), that on margin of occiput pale yellowish and short, upper
margin of frontal callus partly concealed by a downwardly directed fringe of fairly
long, appressed, glistening whitish hair, face and jowls clothed with white hair ;
frontal spots dead black, median spot small but distinct, elongate (fusiform or
elliptical), lateral spots large, bluntly triangular, at least in type, not in contact
with eyes ; frontal callus shining black, of relatively considerable and practically
uniform depth, in contact with eye on each side below, but with its upper margin
scarcely produced into an angle in middle line ; spot in middle line below callus
dull black, truncate triangular, narrower below ; eyes sparsely clothed with minute,
pale hairs ; proximal joint of palpr neutral grey on outer surface and clothed with
whitish hair, terminal joint light buff, clothed below at base with fairly long and,
on outer side with short, appressed, glistening whitish hair, mixed on outer surface
beyond proximal fourth with minute black hairs, terminal joint acuminate,
proximal half, at least in type, considerably swollen; first joint of antennaz
neutral grey pollinose (narrower distal extremity shining black), clothed above and
on inner side with short black hair, on upper part of outer side with short whitish
hair, and below, especially on proximal half, with longer whitish hair; second
joint of antennae blackish brown, clothed with black or blackish hair (third joint
missing In case of type). Thorax: dorsum clothed with short, partially appressed,
glistening yellowish hair, dorsal surface of scutellum clothed with longer and more
erect hair of same kind; median longitudinal smoke grey stripe very narrow,
admedian stripes broader, but tapering to transverse suture and becoming partly
obsolete midway between latter and prescutellar groove, pale smoke grey lateral
border, behind depression at each end of transverse suture, indistinctly divided

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 291

into two fairly broad longitudinal stripes by means of an olive streak ; scutellum
with an ill-defined olive-grey median stripe, and base of each lateral border light
olive-grey ; swelling occupying depression at each end of transverse suture clothed
below with fine blackish hair; pleurae and pectus light neutral grey, clothed
with whitish hair. Abdomen: first (visible) tergite without either median stripe
or rounded spots, otherwise agreeing in coloration with following tergites, rounded
spots on fourth and subsequent tergites resting on base of segment, and in each
case partly concealed by hind margin of preceding tergite; lateral extremities
of first and second tergites pale neutral grey, those of first tergite clothed with fine
white hair, lateral extremities and posterior angles of third and following tergites
and hind borders of all tergites clothed with fine, glistening pale yellowish or
yellowish white hair, longer than elsewhere on posterior angles of fifth and sixth
tergites, and on hind borders of last two segments ; dorsum otherwise clothed
with minute, appressed, black hairs, second tergite with an area of minute,
appressed yellowish hairs extending from in front of each rounded spot towards
posterior angle on each side, spots on third tergite and area in front of them partly
clothed with similar hairs ; venter olive-grey, median third of last three scutes
darker (mouse grey), hind margins of second and following segments pale olive-
buff; ventral surface, except last segment, clothed with minute, appressed,
yellowish hairs, terminal segment clothed as usual with coarse, erect, black hairs,
a few black hairs also on preceding segment. Wings as in H. sewelli, Austen ;
distal margins of rosettes moniliform ; posterior margin of wing with a rather
narrow, interrupted, pale border, extending from distal extremity of first posterior
cell almost to middle of axillary cell, somewhat faint in distal extremity of first
posterior cell, and in each subsequent posterior cell broader in distal marginal
angle and tapering towards proximal marginal angle, where it becomes indistinct ;
stigma mummy brown, well marked, its proximal extremity pale. Squamae pale
smoke grey, borders ivory yellow. Halteres ivory yellow, knobs seal brown above
and below. Legs: coxae and femora light neutral grey or neutral grey, clothed
with glistening whitish or silvery white hair, extreme tips of femora as usual cream-
buff; front tibiae brownish black, with a fairly broad cream-coloured band just
beyond base, distal two thirds conspicuously swollen, pale band clothed with
glistening, appressed, silvery white hair, front tibiae otherwise clothed with black
hair ; middle and hind tibiae clove brown, each with two cream-buft or cream-
coloured bands and largely clothed with minute, glistening, yellowish hairs, fringes
on antero-extensor margins of hind tibiae consisting of rows of black and yellowish
hair alternately ; front tarsi black, clothed with black hair ; middle and hind tarsi
blackish brown, clothed with minute black hairs, proximal two-thirds of first joint
in each case cream-buff or cream-coloured.
Near Jerisheh, 5 miles N.-E. of Jaffa, first week of May 1918.

The species just described is allied to Haematopota plumalis, Linn., and H. sewelli,
Austen, but is distinguished from both in the female sex by the first joint of the
antenna being less swollen and more elongate, with the groove-like constriction
before the distal extremity less sharply marked, although the extremity itself is
narrowed. Haematopota innominata, Q, is further distinguished from H. sewellz,
2, by the dorsum of the abdomen being blacker, owing to the much less extensive
{637) F

292 ERNEST E. AUSTEN.

development of the grey markings; and by the femora, except the extreme tips,
which as usual are cream-buff, being entirely neutral grey, without any trace of
greyish fawn-coloured or greyish cinnamon markings.

From Haematopota minuscularia, Austen, 2°, H. nnominata, 9, is distinguishable
owing to its larger size, the hair on the front being mainly blackish or blackish
brown, the first joint of the antenna, as viewed from the side, being very distinctly
swollen and less conspicuously greyish pollinose, and by the grey markings on the
dorsum of the abdomen being much less extensive.

Lastly Haematopota innominata, 9, may be distinguished from H. minuscula,
Austen, 9, by the first joint of the antenna being less swollen and more elongate,
and by the absence of broad, grey, lateral borders to the dorsum of the abdomen.

Genus Tabanus, Linn.

The seventy specimens of this genus obtained by the author prove, as already
stated, to belong to sixteen species, of which seven, or 43°7 per cent—a fairly
high proportion—are apparently new. The list of species is as follows : OF

1. Tabanus decorus, Lw.

ay a alexandrinus, Wied.

De 44 unsecutor, Sp. N.

4 a gigas, Herbst.

5, At mendicus, Villen.

6. of lunatus, Fabr.

7 ‘4 nemoralis, Mg.

8 A eggert, Schin.

9, “3 autumnalis, Linn.
10. Ms regularis, Jaenn.
1% i rupinae, sp. N.
12. 3 arenwwvagus, Sp. 0.
Te a accensus, Sp. Ni.
14, ~ leleani, sp. n.

15. = ,, ~— pallidipes, sp. n.
16. c dalet, sp. n.

The representatives of Tabanus gigas and T. autumnalis show variations in
coloration already noticed more than sixty years ago by Loew (Verh. z.-b. Ges.
Wien, Bd. vin, 1858, pp. 582-583, 605), in the case of examples from Sicily, Cyprus,
or Asia Minor. In particular, the russet tint of the ground-colour of the dorsum
of the abdomen in females of 7. autumnalis, Linn., makes the recognition of such
individuals a matter of some difficulty for those who are only familiar with the
form of the species usually seen in Northern and Central Europe, in which the
abdomen in the female is black or blackish mouse grey.

While all of the previously described species of Tabanus met with in Palestine
are purely palaearctic, and the entire Tabanid fauna appears free from any
admixture of Oriental or Ethiopian forms, certain of the more or less grey or drab-
coloured species described below as new, such as 7. leleani, T. arenivagus, and
T. rupinae, appear to some extent to show the influence of a desert environment.

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 293

Key to the Sixteen Species of Tabanus recorded below.*

1 (1£). Eyes (under an ordinary hand lens magnifying about 5 diameters) distinctly
or conspicuously hairy.

2 (3). Body uniformly black, unrelieved by conspicuous lighter markings formed
by hair or ground-colour, or both; wings blackish brown, or with veins
strongly suffused with that colour .. oe alecandrinus, Wied. [2].

3 (2). Body not uniformly black, if mainly black or slate-black invariably relieved
by lighter markings formed by hair or ground-colour, or both ; wings not
blackish brown.

4 (5). Dorsum of abdomen with greyish olive median longitudinal stripe
lunatus, Fabr. [6].

5 (4). Dorsum of abdomen not so marked.

6 (7). Dorsum of abdomen with three longitudinal series of clearly defined, pale
neutral grey, white haired or whitish haired spots, those of median row
triangular Je A be nemoralis, Mg. [7].

7 (6). Dorsum of abdomen rath so faabltadh

8 (9). Large or very large species, at least 18 mm. in length; dorsum of thorax,
including scutellum, score clothed with bright ochreous or yellowish
harelvi¢. aie é ‘ gigas, Herbst (4).

9 (8). Small or medium- ee species, not Meocdine 14 mm. in length.

10(11). Second (visible) abdominal tergite in both sexes with a double, quadrate,
black or blackish mouse grey mark, occupying median third ‘and partly
divided by an indistinct greyish triangle, based on hind margin

decorus, Lw. [1].

11(10). Second (visible) abdominal tergite not so marked.

12(13). Legs in 9 (only sex at present known), except extreme tips of femora,
entirely black or slate-black ; no sharply marked contrast between first
two abdominal tergites and the two following ones . . insecutor, sp. n. [3].

13(12). Legs in 2 (only sex at present known) not entirely black, extreme base
of extensor surface of front tibiae, and middle and hind tibiae and first
jomt of middle and hind tarsi except tips, cream-buff; first two
abdominal tergites in 9 greyish olive pollinose, clothed with fairly long
ochreous hair, and, especially when viewed from the side, contrasting
sharply with next two tergites, which are shining black

mendicus, Villen. [5].

14 (1). Eyes (under an ordinary hand-lens magnifying about 5 diameters) bare or
apparently so.

*In this table, as in that for the genus Haematopota above, contrasted and mutually
exclusive statements and summaries of characters are indicated by the paired numbers
on the left, the contrasted conditions in each case being shown, either subsequently or
previously, under the second number of the pair (i.e., the one in round brackets).

The number in square brackets [ ] after the name of a species indicates the serial
position of the species in the ensuing pages.

It should be noted that the characters of a given species shown in the table are only
sufficient to distinguish it from the other species included in the present synopsis ;
they are not necessarily distinctive as regards known species not yet met with in
Palestine, but which may ultimately prove to form part of its fauna.

Tabanus insecutor, sp. n., is shown twice in the table, since it is included in both the
hairy-eyed and bare-eyed categories the very minute hairs on the eyes in the @
being liable to be overlooked on a first examination.

(637) F2

294

15(16).

16(15).

17(18).

18(17).
19(20).

20(19).
21(22).

22(21).

23(24).
24(23).

25(26).

96(25).

27(28).

ERNEST E. AUSTEN.

Blackish species. Dorsum of abdomen slate-black, transversely banded
with silvery-white hair; wings with extreme base and proximal two-
thirds of costal border ochraceous-tawny, contrasting sharply with
remainder of surface .. - a ss és dale, sp. n. [16].

Not blackish species. Dorsum of abdomen not slate-black, but with
markings arranged in longitudinal series ; no part of costal border pre-
senting a sharp colour contrast to remainder of wing.

Conspicuously narrow-bodied, elongate species.
a) Frontal triangle in g and subcallus in 2 with shining transverse
band, extending from eye to eye .. ». arenwagus, sp. n.[12].
b) Frontal triangle in g and subcallus in 9 without shining transverse
band, entirely dull, pollinose .. = accensus, sp.n. [13].
Species of ordinary shape, not conspicuously narrow-bodied or elongate.
Dorsum of abdomen with median longitudinal stripe.
a) Ground-colour of dorsum of abdomen ochraceous tawny
rupinae, sp. n. [11].
b) Ground-colour of dorsum of abdomen blackish brown ; much smaller
species than foregoing (length of 9 13-5 to 14 mm.)

| regularis, Jaenn. [10].
Dorsum of abdomen without median longitudinal stripe.

Legs in Q (only sex at present known), except extreme tips of femora,
entirely black or slate-black .. . msecutor, sp.n. [3].

Legs not entirely black or slate-black, at least proximal half of front
tibiae, and middle and hind tibiae except tips, pale (cream-buff or
ochraceous buff). an

Femora cinnamon-buff or pinkish cinnamon .. palladipes, sp. n. [15].

Femora, at least front pair, black or blackish, neutral grey or pale neutral
grey pollinose.

Mouse grey, dark mouse grey or blackish mouse grey species ; front in 2
broad (its length equal to rather less than four times its breadth at lower
margin of lower callus), with two large and completely separate frontal
calli, the lower callus large, quadrate, broader than high leleana, sp. n. [14].

Not mouse grey species ; front in Q narrow or of only moderate breadth
(its length equal to at least five times its breadth at lower extremity),
upper callus represented only, if at all, by a narrow median line running
upwards from lower callus, latter narrow, elongate, usually tapering to

its upper extremity, two and a half to three times as high as its breadth
below.

Sides of pale triangles in middle line of abdomen not concave; first
posterior cell in wing greatly narrowed (sometimes even closed) at or
before its distal extremity ; front in 9 at least seven and a half or eight
times as long as its breadth at lower extremity ; venter, except extreme
tip, and a broad, median, longitudinal, shining, slightly or not at all
infuscated stripe, uniformly light ochraceous-salmon or light ochraceous-
buff: large species, at least 19 or 20 mm. in length .. eggeri, Schin. [8].

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE, 295

28(27). Sides of pale triangles in middle line of abdomen usually more or less
concave ; first posterior cell in wing, though often narrowed at distal
extremity, seldom exceptionally so; front in Q five or five and a half
times as long as its breadth at lower extremity ; ground-colour of dorsum
of abdomen in Q either russet, black, or blackish grey—if russet, then
venter pale smoke grey pollinose orange-cinnamon or pinkish cinnamon,
without or with but faintly indicated, broad, dark, median, longitudinal
stripe: fairly large or large species, 15 to nearly 19 mm. in length

autumnalis, Linn. [9].

1. Tabanus decorus, Lw.

One ¢, three 99, near Jerisheh, 5 miles N.-H. of Jaffa, between 26.iv. and
8.v.1918; one 9, Valley of Jehoshaphat, Jerusalem, 20.1v.1918 (Captain (acting
Lt.-Col.) W. J. Dale, O.B.H., R.A.V.C.). |

Notwithstanding that the third joint of the antenna in three of the 99
mentioned above (it is wanting in the other two specimens) is somewhat broader
than it should be, according to Loew's own statement and the subsequent
description by Brauer (Denkschr. k. Akad. Wiss., xlu, 1880, pp. 152-153), there
can be little doubt that the determination here given is correct. The eye-
markings in this species appear to be in course of disappearance, and to be some-
what variable in consequence. Loew and Brauer describe the eyes as unbanded
in the g, but as having in the 9 a single, narrow, greenish-yellow transverse band,
or an indistinct trace of such a band. The ¢ mentioned above, and one of the
© 9 taken at Jerisheh, had in life green eyes, each marked with a single, faint, narrow,
median, golden horizontal band ; while another 9 captured at the same spot had
bronze-green eyes, each with three faint, golden, transverse bands, the middle one
being the more distinct.

It may be added that the type of Tabanus decorus was taken in Syria, while
Brauer (loc. cot.) records the species as occurring in the Lebanon.

2. Tabanus alexandrinus, Wied. (fig. 5.).

Five $d, three 99: near Jerisheh, 5 miles N.-E. of Jaffa, 26.iv.1918 (one J
beginning of May).

Although the brevity and lack of precision of the original description (Aussereurop.
zwerft. Ins., ii, 1830, p. 624) of this entirely black species would render identifi-
cation difficult, there can be no doubt, in spite of a slight discrepancy as regards
the coloration of the eyes, that the specimens referred to above really belong to
T. alecandrinus, as characterised and redescribed by Brauer (Denkschr. k. A kad.
Wiss., Bd. xlii,1880, p. 161, Taf. ii and v, figs. 23). This author records the species
as occurring in Syria and Italy, and states that Wiedemann’s collection contains
examples of it from Alexandria. A distinctive character of the typical form of
T. alexandrinus, especially in the Q, is the shape of the expanded portion of the third
joint of the antenna, the upper margin of which is convex, or at least shows no trace
of the usual excavation (fig. 5). The expanded portion of the third antennal
joint varies in width in different individuals, and is sometimes, especially in the
3, a good deal narrower than it appears in the figure. Macquart (Hist. Nat. Ins.,
Dipt., i, 1834, p. 199), who redescribed the species under the name Tabanus

296 ERNEST E. AUSTEN.

carbonatus, from a specimen from Sicily, notes the shape of the upper margin of
the third antennal joint.

According to Brauer (loc. cit.), the eyes in both sexes are dark green; in the
specimens collected near Jerisheh, however, the eyes of § and Q in life were
unicolorous black.

The Museum collection contains several examples (2 Jg, 3 99), from Algeria
and Tunis, of what must, at any rate provisionally, be regarded as a variety or
local race of the present species.* This western form is distinguished from the
typical one, as represented by Palestine specimens, by the absence of the greyish-
white hairs on the lateral margins of the dorsum of the thorax in the 9, on and
in front of the postalar calli, and by the distal extremity of the expanded portion
of the third joint of the antenna being distinctly emarginate above. This is at any
rate the case in the 9; in the two §¢ from Algeria and Tunis the third joints of
the antennae are unfortunately missing.

Brauer’s record of Tabanus alexandrinus from Syria, referred to above, gives
no more precise indication of locality, but Bezzi (Brotéria, Ser. Zool.,vii,1909, p. 42)
records the species as having been taken at Ghazir (about 15 miles N.-E. of Beirut),
on 14th June, 1904.

It is interesting to note that clinging to the hairclothing the front coxae of one
of the 3 specimens taken near Jérisheh on 26th April 1918, were six little, hexapod,
Campodea-like larvae, which Dr. Gahan states are the “triungulin” stage of an

oil-beetle (Meloe sp.).

3. Tabanus insecutor, sp. n. (fig. 6).

Q.—Length (1 specimen) 12°25 mm.; width of head 4 mm.; width of front
at vertex 0°6 mm. ; length of wing 10°25 mm.

Eyes unbanded, at first sight appearing bare, when more closely examined seen to be
sparsely clothed with very short, very minute hairs ; dorsum of thorax shining blackish
mouse-grey pollinose; dorsum of abdomen shining black, clothed partly with minute
appressed black hairs, and on hind borders of second and following segments with
appressed glistening whitish or pale ochreous hair, first and second tergites and hind
borders of two following segments greyish pollinose ; frontal callus of 9 large, elongate,
shining black ; palpi dark neutral grey, terminal segment in Q slender ; first two joints
of antennae dark neutral grey, third joint black, expanded portion broad, its upper
margin angulate; wings lightly tinged with brownish drab, extreme base and costal
border as far as stigma ochre-coloured, a small, ill-defined, oblique, faint cinnamon-
brown blotch extending from narrow, elongate, ochraceous-tawny stigma to discal cell ;
legs, except extreme tips of femora, which are cinnamon-buff, entirely black or slate-
black.

*The ¢ genitalia of this North African form have been examined microscopically
and compared with those of the typical race, as found near Jerisheh; a similar
comparison has been made in the case of the infra-anal plate of the 9. The differences
noted in each instance appear to be merély trivial, and, so far as the limited material
available for study permits a definite conclusion to be drawn (a pair of specimens of each
form is all that can at present be devoted to dissection), cannot be regarded as anything
more than individual.

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 297

Head light neutral grey pollinose, face and jowls somewhat thinly clothed with
whitish hair, longer in basi-occipital region, hind margin of occiput fringed above
and at sides with very short, pale yellowish hair ; front in 2 of medium and entirely
uniform width, about three and a half times as long as broad, clothed with minute,
appressed, glistening ochreous or pale yellowish hair, ground-colour slightly darker
in region of vertex but no trace of an ocellar tubercle; lower and upper frontal
calli fused together to form a single, elongate, shining black callus, extending from
level of inner angles of eyes to a point slightly more than one-third of length of
front from hind margin of vertex; callus exhibiting a trace of a fine, elongate,
median groove, and separated from eye on each side by an interspace equal to rather
less than half its width, lower margin of callus straight, upper extremity ogival,
the middle region somewhat constricted, so that outline of entire callus is suggestive
of that of an Ancient Greek, leaf-shaped sword ; eyes of 2 in life plain dark green,
unbanded ; proximal segment of palp: clothed with whitish hair, distal segment
in 2 acuminate, sparsely clothed on outer surface with minute, appressed, yellowish

‘hairs ; first two joints of antennae clothed above and below with short black hairs,
first jomt not swollen and not embracing second joint, upper distal angle of latter
moderately produced, annulate portion of third joint about one-fourth shorter
than expanded portion, depth of latter, from centre of lower margin to apex of
angle on upper border, about equal to its length, angle, which bears a few exceedingly
minute black hairs, situate in centre of upper border, which is slightly emarginate
in front of it. Thorax: dorsum (including scutellum) and pleurae clothed with
pale ochreous or pale yellowish hair, the hairy covering of the pleurae almost
whitish at the tips, lateral borders of dorsum and commencement of two
longitudinal stripes extending a short distance from front margin cinereous
pollinose. Abdomen: first and second (visible) tergites clothed with glistening
appressed, pale ochreous hair, the hair forming a tuft in the centre of the hind
border of the first tergite paler, and that on the hind border of the second tergite
glistening whitish ; glistening, appressed, pale hair on hind borders of fifth and
sixth segments extending forwards in middle line in each case in form of a broad
triangle, apex of which is almost or quite in contact with hind margin of preceding
seoment; sides of sixth tergite and entire surface of seventh clothed with
glistening, appressed, pale ochreous or pale yellowish hair ; ground colour of seventh
tergite cinereous pollinose ; venter blackish slate-coloured, neutral grey pollinose
but moderately shining, clothed with short, glistening, appressed, pale ochreous or
pale yellowish hair, extreme hind margins of second and following scutes cream-
buff. Wangs: proximal portions of first and fifth longitudinal veins ochraceous
tawny, bases of second and sixth veins ochreous, veins otherwise dark mummy-
brown or dark sepia-coloured, anterior branch of third longitudinal vein without
an appendix. Sguamae light buff, frmged with pale hair. Halteres: knobs dark
seal-brown above and below, stalks and tips of knobs cream-buff or cinnamon-
buff. Legs: coxae neutral grey pollinose, both they and femora clothed with fine,
yellowish hair ; tarsi and tips of tibiae, as also inner sides of front pair of latter,
clothed with minute, appressed, black hairs, tibiae elsewhere clothed with short,
glistening, appressed, yellowish-white, pale yellowish, or pale ochreous hair ;
front tibiae slightly thickened.

298 ERNEST E. AUSTEN.

Near Mulebbis (Jafia district), 14.v.1918 : caught by Lt.-Col. H. P. Sewell, C.M.G.,
D.S.0., R.A.M.C., from a light Ford motor-car (which the insect was pursuing)
travelling at 20 miles an hour across a grassy plain ; more than one specimen seen.

Apart from its smaller size, the species just described presents a superficial
resemblance to Tabanus aterrimus var. auripilus, Mg., but is distinguishable,
anter alia, by the elongate frontal callus of the 2, and by the conspicuously ochreous
coloration of the base of the wing.

Fig. 5. Tabanus alexandrinus,
Wied., head of @ (typical
form) in profile. X about 10.

‘ Vi

Fig. € Tabanus msecutor, Austen, 9: a, head from in front; b, head
in profile, showing shape of antenna. x 10.

Labanus insecutor is allied to T. umbrinus, Mg., with which it agrees in the
character of the frontal callus of the 9, the coloration of the eyes, and the shape
of the terminal segment of the @ palpi; the new species may however be
distinguished, inter alia, by the much greater development of the pale yellow hair

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 299

on the abdomen, particularly on the venter, and by the absence of the dark brown
costal border on the wing, extending from the base of the wing to the end of the
stigma, and of the broad dark brown blotch behind the latter.

4. Tabanus gigas, Herbst.

Two $3: one near Jerisheh, 5 miles N.-E. of Jaffa, 16.v.1918, hovering above
low hill, 9°15 a.m. (one other specimen seen) ; the other in marsh 1 mile S.-H. of
Tel Abu Zeitun (within 14 mile of spot at which foregoing specimen was taken),
18.v.1918, on flower (one other example seen).

This strikingly handsome species, in which, as shown by the first of the two
individuals recorded above, the eyes in life are purplish black, without bands,
is already represented in the British Museum collection by, among other specimens,
a ¢ from Galilee (B. T. Lowne, 1863-64). Statements by Brauer (Denkschr. k.
Akad. Wiss., Bd. xlii, 1880, p. 165) show that Tabanus gigas is found in France,
Austria, Greece, Sicily, Russia, Syria and Asia Minor, while the series in the
Museum includes specimens from Italy and Bulgaria.

In length, the two individuals taken by the writer measure respectively 19 and
18°5 mm.; their dimensions are therefore somewhat less than those given by
Brauer (loc. cit.), according to whom the length of the ¢ is 20 mm., while that of
the 2 may be as much as 23 mm. The variability of Tabanus gigas as regards
the coloration of the hairy covering of the head, body, and legs, according to the
country of origin, is dealt with at some length by Loew (Verh. z.-b. Ges. Wren, Bd.
viii, 1858, pp. 582-583), who refers to the species under its synonym T. albipes,
Fabr. Although the two specimens recorded above as having been caught
respectively on 16.v. and 18.v.1918 were captured within less than a mile and a
half of each other, they nevertheless show considerable colour-differences of the
kind indicated. While in the case of the ¢ taken in Tel Abu Zeitun marsh the pale
hair on the head, thorax, tibiae and first two segments of the abdomen is of a rich
ochreous tint, the corresponding hair in the case of the other specimen is much
lighter in hue, and on the tibiae and the hind border of the second abdominal
segment is even almost white. The individual with the paler hair also shows a
median triangular patch of whitish hair on the hind border of the third abdominal
tergite, and a smaller number of yellowish or ochreous hairs in the corresponding
position on the following three tergites. These patches of pale hairs are entirely
wanting in the other specimen, in which the paler fringe to the anal region is
represented by but a very small number of ochreous hairs.

5. Tabanus mendicus, Villen.

One ©: Ain Sinia, 10 miles N. of Jerusalem, 18.iv.1918, on bank of stream,
only specimen seen (Captain K. B. Williamson, R.A.M.C.).

In the living insect the eyes were bronze, with three horizontal, dark purple
bands, the lowest the broadest, the middle one narrow.

The type of this somewhat unusual-looking species (a 9, taken in the Oasis
of Damascus, in the second fortnight in April) is not inaptly compared by
Villeneuve (Bull. Soc. Amis Sc. Nat. Rouen (5) xlvii, 1912, pp. 41-42) to a
tiny example of Tabanus (Atylotus) gigas, Herbst. The specimen from Ain

300 ERNEST E. AUSTEN.

Sinia, recorded above, would seem to be only the second individual of Tabanus
mendicus met with up to the present time. It is perhaps worth mentioning
that, when viewed from the side, the insect appears to have a broad, shining,
black band across the middle of the abdomen, owing to the third and fourth
segments being clothed mainly with black hair, and thus contrasting with the
base and distal extremity, on which the hairy covering is honey-yellow or
cinnamon-buff, while the contrast is heightened by the ground-colour of the
first two segments being neutral grey or olive grey pollinose.

6. Tabanus lunatus, Fabr.

Jaffa district and Jerusalem: three 99, near Mulebbis, 14.v.1918, caught by
Li.-Col. FE. P. Sewell, C.M.G., D.S.0., R.A.M.C., from a light Ford motor-car
(which the insects were following), at the same time and under the same
circumstances as the type of Tabanus insecutor, Austen, see p. 298 (car proceeding ~
at 20 miles an hour across a grassy plain; several specimens seen); a fourth 9,
in the British Museum collection, bearing the label “Jerusalem. A. Béttcher,
Berlin,’ received in 1912 from the Zoologisches Museum, Berlin, and presented
by Miss G. Ricardo.

The four specimens enumerated above, which are the only representatives of
this small or medium-sized, greyish-olive-coloured, hairy-eyed species at present
contained in the National Collection, show that, in the female sex at any rate,
Tabanus lunatus, Fabr., is subject to considerable variation in size, the three
examples brought home by the writer measuring from 10°75 to 11°75 mm. in length,
as compared with 13:2 mm. in the case of the Q labelled “‘ Jerusalem.”

The eyes of the 9 of 7. lwnatus, which are thickly clothed with short, fine, whitish
hair, are bronze-green in life, with three narrow, purplish brown, transverse bands.

The description and figures of the head and palpus of the 2 given by Brauer
(Denkschr. k. Akad. Wiss., Bd. xlii, 1880, p. 172, Taf. ui, figs. 34)—who records
the species as occurring in Dalmatia, Italy, Sicily, Spain, Asia Minor and the
Caucasus—are somewhat misleading, and since in the preceding synoptic table
(J. cit., p. 134) the upper frontal callus is characterised as “ linear, connected with
the lower callus . . .”’ it would be difficult if not impossible to determine female
specimens of 7’. lunatus by means of the Austrian author’s well-known work on
“The European Species of the Genus Tabanus.” A careful examination and
comparison of the specimens brought back by the present writer shows that the
upper frontal callus in the Q of 7. lunatus, instead of being “ elongate, more or
less linear or fusiform ” (Brauer), is roughly oval or elliptical oval in outline, and
sometimes much larger than the lower callus, with which it is not connected ; also
that the terminal segment of the palpus in the same sex is acuminate, and only
moderately swollen at the base—much less so than would be imagined from
Brauer’s description and figure. The sharp colour-contrast in the third joint of
the antenna, owing to the expanded portion being cinnamon-rufous and the annulate
portion dark seal-brown or black, is very noticeable.

7. Tabanus nemoralis, Me.

Two 929, Mount of Olives, 20.iv.1918 (#. #. A.), and 4.v.1918 (Captain (acting
Li.-Cal.) W. J. Dale, O.B.E., R.A.V.C.), both specimens taken “ on ma.”

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 301

In life, the eyes of the Q of this species are purplish bronze, with three dark
purple horizontal bands, the middle one of which is the narrowest, across the centre.

Tabanus nemoralis is recorded by Brauer (Denkschr. k. Akad. Wiss., Wien., Bd.
xlu, pp. 175, 208-210 (1880)) as occurring in France, Switzerland, Italy, Sicily and
North Africa ; the British Museum collection contains specimens from Switzerland
(Upper Engadine), and Algeria (Rouiba, May-June, 1911, Baron J. R. M. Surcouf).

8. Tabanus eggeri, Schin.

One 9, Wilhelma (Jaffa district), 30.vu.1918, resting inside small, open shed,
at back of house in village ; one Q, near Jerisheh, 5 miles N.H. of Jaffa, 26. viii. 1918,
in tent, 7.45 a.m. (Zt.-Col. (temp. Col.) E. P. Sewell, C.M.G., D.S.O., R.A.M.C.) ;
one 9, near Yahudiyeh (Jaffa district), 1.1x.1918 (Captain (acting Major) W. S.
Corfield, R.A.M.C. (T.)).

In this species, as was shown by the specimen taken near Jerisheh by Colonel
Sewell, the eyes in life are purplish brown, with a bronze sheen, and are unbanded.

Tabanus eggert, which is already represented in the British Museum collection,
by, among others, specimens from Gibraltar, Italy and Algeria, is recorded by
Brauer (Denkschr. k. Akad. Wiss., Bd., xlii, 1880, p. 184—under the pre-occupied name
T. intermedius, Egg.) as occurring in Southern France, Spain, Switzerland, Sicily,
‘Corsica, Egypt and Asia Minor. As an aid to recognition it may be mentioned
that, in the species under consideration, the diameter of the first posterior cell
is greatly contracted on the hind margin of the wing. According to Brauer (é. cit.,
p. 183), the hind tibiae are fringed on the outer side with black hair ; the specimens
from Palestine, however, show a slight variation from the typical form, in that
the external fringe on the hind tibiae consists mainly of ochreous or ochraceous-
tawny hair.

9. Tabanus autumnalis, Linn.

One ¢, Baharet Katurieh, near El Jelil (about 10 miles N.E. of Jaffa), 28.1v.1918,
only specimen seen; two 99, Ain es-Sultan, near Jericho, 22.iv.1918—one taken
on a horse, the other, as also a third specimen, which was not captured, resting on
face of excavations.

This is a very widely distributed species, which, according to Brauer (Denkschr.
k. Akad. Wiss., Bd. xlu, 1880, p. 193), occurs throughout Central and Southern
Kurope, and is also found in Holland, Sweden, England, S. Russia, Poland, Corsica,
Corfu, Syria and Asia Minor. Loew (Verh. z.-b. Ges. Wien, Bd. viii, 1858, p. 605),
who draws attention to the variability of T. autwmnalis, like that of so many other
species, as regards both size and coloration, according to locality, records its range
as including “ the whole of Europe and a great part of the Near East.”” In addition
to specimens from Hungary, Bulgaria, England and elsewhere, the species is
represented in the British Museum (Natural History) by a Q from Algeria (Biskra,
18.v.1893, Rev. A. H. Eaton) and two 9° from Cyprus (one collected by P.
Gennadwus, the second specimen taken by Miss D. M. A. Bate, near Ktima, between
3.v. and 15.vu.1901). It may be added that the three last-mentioned examples
belong to the form of the species met with in Palestine, and also recorded by Loew
(loc. cit.) as occurring in Cyprus, in which the ground-colour of the dorsum of the

302 ERNEST E. AUSTEN.

abdomen in the @ is russet instead of black or blackish mouse grey, while the venter
in the same sex 1s pale smoke grey pollinose orange-cinnamon or pinkish cinnamon,
and the usual broad, dark, median longitudinal ventral stripe is absent or but
faintly indicated in both sexes. Two other Algerian 29Q of this species (taken
respectively at Algiers and Hammam R’Irha, in May 1908, by the Hon. W. (now
Lord) Rothschild), 11 the National Collection, belong to the typical form.

Loew (loc. cit.) remarks that South European examples of Tabanus autumnalis
are on the whole decidedly smaller than specimens from Germany and Northern
Europe. The respective lengths of the three individuals from Palestine, recorded
above, are 15°75, 15, and 19°5 mm.; thus it will be seen that, as compared for
instance with English representatives of the species under discussion, the two
former are markedly subnormal in size.

10. Tabanus reguiaris, Jaenn.

One 2, in Wadi el Kelt, Jericho Plain, 1.vi.1918. The eyes of this specimen in
life were dark purplish brown, without bands.

Tabanus regularis, which was described from a specimen from 8. France (Marseilles
district), and is included by Brauer among the Tabanidae of Italy and Greece, has
recently been recorded by Becker (Ann. Mus. Zool. Acad. Imp. St. Pétersbourg,
xvii, 1912, p. 596) as occurring in Persian Baluchistan.

11. Tabanus rupinae, sp. n. (figs. 7, 8).

¢ 9.—Length, 3 (2 specimens) 14°5 to 15-4 mm., 2 (3 specimens) 15 to 16-2 mm. ;
width of head, ¢ 5°4 to 5°6 mm., 9 5:2 to 5°6 mm.; width of front of 9 at vertex
0°75 to 0°83 mm. ; length of wing, § 11:25 to 12 mm., 2 12°6 to 13°25 mm.

Dorsum of thorax light greyish olive (deep greyish olive when rubbed) pollinose,
clothed with fine, silky, appressed as well as erect, ochreous har, mingled with fine,
erect, blackish hairs, the erect ochreous hair whitish at the tups; dorsum of abdomen
ochraceous-tawny, with a pale, median longitudinal stripe running from base of second.
to hind margin of penultimate tergite ; anterior branch of third longitudinal vein abruptly
angulate at the base, and provided with a long appendix; front femora dark mouse-
grey (smoke-grey pollinose on outer side), front tarsi and distal halves of front tibiae
black.

Head: face and jowls pale olive-grey in both sexes, clothed with whitish hair ;
occiput pale olive-grey in g, smoke-grey or pale smoke grey in 9, hind margin
of occiput in region of vertex clothed with very short whitish or pale yellowish
hair; frontal triangle in § smoke grey or pale smoke-grey pollinose; front in Q
olive-buff pollinose, moderately broad and of practically uniform width, its length
equal to about four and a half times its breadth between inner angles of eyes,
clothed with short, appressed, ochreous hairs, mingled in region of vertex with
minute blackish hairs; lower frontal callus in 2 blackish brown or sepia-coloured,
transversely oblong, its depth equal to about three-fourths of its breadth, narrowly
separated from eye on each side, its lower margin situate just above level of inner
angles of eyes; upper frontal callus in Q represented by a somewhat ill-defined,
dark brown, crescentic mark, with long horns extending upwards, but its precise

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 303

shape evidently dependent upon degree of rubbing to which the specimen has been
subjected ; ocellar tubercle in § very small and inconspicuous, situate at bottom
of cleft between eyes at vertex, no trace of ocellar tubercle in 2; subcallus in Q
pinkish buff pollinose, with a more or less distinct indication of a tawny-olive or
light mummy-brown transverse band extending from eye to eye at level of antennae ;
eyes under an ordinary hand-lens appearing bare in both sexes (closer scrutiny
sometimes reveals scattered minute hairs on lower portions), eyes in $ each with
a bluntly triangular area of greatly and abruptly enlarged facets, partly encircled
with a sharply contrasted zone of small facets occupying approximately lower
fourth and outer and upper border, latter being continued to vertex in practically
its full width ; eyes of 9 in life dark green, with purplish-brown sheen and without
bands; palpi of ¢ clothed with whitish hair, terminal segment cream-coloured,
greatly swollen, ovate pyriform and ending in a small curved point, proximal segment
of 9 palpi clothed with fairly long, whitish hair, terminal segment of 2 palpi long

SSSA

Se SY

Fig. 7. Tabanus rwpinae, Austen, 9. xX 4.

acuminate, considerably swollen at base, cream-coloured or ivory-yellow, clothed
on outer side with short, appressed, glistening hair of similar colour (occasionally
intermixed with a few minute black hairs towards distal extremity), and with
longer whitish hair at base below; first joint of antennae cinnamon-buff, cream-
buff pollinose, clothed with minute, appressed, glistening yellowish or whitish
hair, with longer pale hair below, and with minute black hairs on upper distal angle,
considerably swollen and expanded distally and partly embracing second joint,
latter agreeing with first joint in coloration and in its hairy covering and with its
upper distal angle considerably produced, third joint cinnamon, annulate portion
infuscated at tip, at least in dried specimens, expanded portion narrow and elongate
in g, of considerable depth in 9, with a prominent but blunt angle near base of
upper margin in both cases. . Thorax: anterior border of dorsum in certain
specimens showing more or less distinct vestiges of three pallid neutral grey

304 ERNEST E. AUSTEN,

longitudinal stripes ; swelling in depression at each end of transverse suture drab-
grey or light drab ; pleurae and pectus pale smoke-grey, clothed with whitish hair.
Abdomen: pale, median, longitudinal stripe on dorsum sometimes, especially
in §, formed by a linear series of anteriorly directed, truncated triangles ; median
stripe, at least on third (visible) and following tergites, bordered on each side with
a dusky, oblong blotch, extending for the full length of the segment except extreme
hind border, and clothed with minute, appressed, black hairs (like rest of hairy
covering of dorsum very easily removed by rubbing, in which case the blotches are
difficult to distinguish) ; on each side of third and two following tergites, outside
dusky blotch just described, a more or less distinct trace of another pale stripe or
oblique pale mark, sometimes bordered on outer side on fourth and fifth tergites,
at least in case of g, with another dusky, black-haired blotch, not reaching hind
margin ; sixth and seventh tergites, except lateral and hind borders of former and
its median stripe, dark brown and clothed with minute, appressed, black hairs
(posterior angles of sixth tergite in g, and hind border of seventh in both sexes

Fig. 8. Tabanus rupinae, Austen : a, head of gin profile ; b, head of 9 from in front,
x 10; b’, antenna of 9° from the side, greatly enlarged.

clothed with longer black hair); dorsum, except as already stated, clothed with
minute, appressed, glistening, ochreous hairs, longer and paler on hind borders
of segments, extreme lateral borders of tergites pale drab-grey pollinose and clothed
with longer whitish hair; ventral scute of first (visible) segment pale drab-grey
pollinose, remainder of venter except terminal segment pale cinnamon or pinkish
cinnamon, covered with cream-buff pollen and clothed with whitish or appressed,
glistening, pale yellowish or pale ochreous hair, hind borders of segments cream-
buff or pale drab-grey pollinose, clothed with longer pale hair; ventral surface
of terminal segment mouse-grey or deep mouse-grey, clothed as usual with coarse,
erect, black hair, in case of ¢ sometimes a few similar hairs on ventral surface of
preceding segment also. Wungs suffused with light drab; stigma inconspicuous,
elongate, pale tawny-olive; veins lighter or darker sepia-coloured. Squamae
cream-buff, borders cinnamon-buff. Halteres cream-bufi or cream-coloured, knobs

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 305

sometimes darker (seal-brown or light seal-brown) at base. Legs: coxae pale
drab-grey or smoke-grey pollinose, clothed with whitish hair; femora and tibiae
of middle and hind legs, and proximal halves or rather less than proximal halves
of front tibiae, cinnamon-buff or pinkish cinnamon-coloured, middle and hind
femora drab-grey pollinose ; middle and hind tarsi dark brown above, first joint
of middle tarsi sometimes paler towards base; third and fourth joints of middle
and hind tarsi distinctly expanded, rather broad; femora clothed with whitish
hair ; middle and hind tibiae and proximal halves of front tibiae clothed with short,
glistening, appressed yellowish-white hair, which on anterior margin of outer side
of hind tibiae is elongated so as to form a fringe ; upper surfaces of tarsi and distal
halves of front tibiae clothed with minute black hairs, extensor surfaces of middle
and hind tibiae, especially towards tips, also largely clothed with similar hairs.

Wadi el Kelt, Jordan Valley, near Jericho: type of ¢ and one para-type of the
same sex, 5.vi.1918, on cliff wall, 5.30 p.m.; type of 9 and one Q para-type L.vi.
1918; an additional 2 para-type 25.v.1918, resting on the author’s tent (pitched
above the Wadi, close to the point at which the old Jerusalem-Jericho road enters
the Valley), 6.30 a.m.

On several evenings at the beginning of June 1918, towards sunset, a number
of specimens of this species, which was not met with elsewhere, were found resting
on the precipitous earth cliffs forming the sides of the Wadi el Kelt, just where the
wild gorge in question leaves the hills.

Tabanus rupinae, which is not closely allied to any of its congeners known to the
writer, presents a very slight superficial resemblance to the Mesopotamian
T. polygonus, Walk., but is distinguishable at once, inter alia, by the somewhat wider
front in the 9; by the lower frontal callus in the same sex being transversely quad-
rate, instead of vertically ogival and prolonged above into a narrow line; _ by the
second abdominal tergite in both sexes being without any trace of a blackish median
spot at the base; by the anterior branch of the third longitudinal vein being
abruptly angulate at the base and provided with a long appendix; and by the very
different coloration of the front legs.

12. Tabanus arenivagus, sp. n. (figs. 9, 10).

¢ 9—Length, 3 (3 specimens) 13°5 to 14 mm.; © (10 specimens) 12 to just
over 14 mm.; width of head, ¢ 4°5 to 4°75 mm., 9 3°6 to 4:4 mm. ; width of front
of 9, at vertex 0°5 to 0°75 mm., between inner angles of eyes 0-4 to 0°5 mm. ; leneth
of wing, ¢ 105tollmm.; 910 to 11:2 mm.

Small, light drab-coloured, narrow-bodied species, with sides of abdomen straight,
tapering from base of fifth segment in 3, and from that of second segment in 2; eyes
under an ordinary hand-lens appearing bare (merely showing sparse microscopic
hairs below), in 3 with a sharply contrasted area of greatly enlarged facets, and two
purplish-brown transverse bands, in 2 with three sinularly coloured bands : frontal
treangle in 3 and subcallus in 2 with a shining transverse band, extending from eye to
eye; & with well-developed though sunken ocellar tubercle, of which there 1s no trace
m 2; latier with two frontal calla ; third joint of antennae ochraceous-buff or ochraceous-
orange, narrow, elongate and tapering, its upper margin not excavated but with a blunt

206 ; ERNEST FE, AUSTEN.

angle near base ; dorsum of abdomen with four longitudinal rows of oblique, elongate,
mummy-brown markings, more distinct in 2; wings hyaline, stigma brownish olive,
elongate and distinct, anterior branch of third vein without an appendia.

Head conspicuously enlarged in ¢; face and jowls in both sexes whitish
pollinose and clothed with whitish hair, a faint, not always distinguishable
infuscation between base of each antenna and corresponding eye; occiput pallid
neutral grey pollinose, its upper margin fringed behind with erect, pale yellowish
hair, very short in 9, longer and somewhat curved forward in 3, frontal triangle
in ¢ and front and sub-callus in 9 pale smoke grey or smoke grey pollinose, front
in 2 clothed with minute, appressed, pale ochreous hairs, and its ground-colour
somewhat darker (mouse grey) in vertical region; shining transverse band on
frontal triangle in ¢ nearly twice as deep as corresponding band on subcallus in
© and much darker (clove brown), its upper margin level with upper border of
area of small facets forming rather less than lower third of eyes, and its lower

Fig. 9. Tabanus arenivagus, Austen, 9. X 4.

margin separated from base of antennae by a space equal to about half the depth
of the band ; [corresponding band in @ brownish olive or isabella-coloured, and
occupying rather more than upper half of subcallus; front in 2 of medium width,
distinctly narrower below, neither callus in contact with eyes, lower callus quadrate,
shining, light sepia-coloured and situate well above inner angles of eyes, upper
callus much larger and darker (seal brown or dark brown) than lower, not shining,
quadrate, situate just below midway between vertex and lower callus, with which
it is not connected, and generally with a more or less distinct longitudinal groove
or depression in middle lme; area of enlarged facets in eye of 3 sharply defined,
bluntly triangular in outline, and occupying rather more than upper two-thirds,
with exception of a narrow hind border of uniform width; eyes of 3 in life
with area of enlarged facets pinkish isabella-coloured above, light grey on
lower border: area of small facets with hind border dark brown, with greenish
reflections, lower border greenish-bronze, with two narrow, purplish-brown,

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 307

transverse bands, extending to hind margin, and, below these, a very fine, and shorter
darkish, horizontal, transverse streak ; eyes of 9 in lafe greenish-bronze, with three
horizontal, purplish-brown bands, and a very narrow dark streak above and below ;
palm, pale cream-buff, those of g and proximal joint of those of 9 clothed with
long whitish hair, distal jomt of Q palpi clothed on outer surface with glistening,
appressed, pale yellowish hairs mixed with minute, appressed, black hairs, and
bearing longer pale yellowish or whitish hairs on under surface ; distal joint of ¢
palpi elongate ovate or ovate pyriform, corresponding joint in Q curved, tapering,
pointed at tip, not greatly swollen at base; first and second joints of antennae
cinnamon-buff, clothed on outer side with pale yellowish hair, and on distal margin
of second joint and upper distal angle of first joint with minute black hairs, first
joint somewhat greyish pollinose, moderately swollen in g, upper distal angle of
second joint not produced, annulate portion of third joint straight. Thorax:
dorsum with four dark mouse-grey or fuscous longitudinal stripes, separated by
narrower drab-grey or pale drab-grey interspaces, lateral borders pale drab-grey
or pale smoke-grey—swelling occupying depression at each end of transverse suture
pale brownish drab ; dorsum of $ clothed with fine, erect, pale yellowish or whitish
hair, mixed with short, glistening ochreous, appressed hairs, and sometimes also
interspersed with darker erect hairs, post-alar calli and swelling immediately above
base of each wing in ¢ clothed with whitish hair ; dorsum of @ clothed with short,
appressed, glistening pale yellowish or ochreous hair, mixed with a certain number
of short and more erect blackish hairs, hair on postalar calli and on swelling above
base of wing on each side as in 3; pleurae and pectus in both sexes pale smoke-
grey, clothed with whitish hair. Abdomen: pattern of markings on dorsum as
shown in fig. 9, mummy-brown markings on second (visible) and following segments
starting from base of segment but not reaching hind border, so that on each segment
‘ the drab-coloured markings, consisting of a median triangle and an oblique elliptical
ovate spot between latter and each lateral margin, are as a rule connected together
posteriorly ; on fifth and sixth segments, however, mummy-brown markings on
each side are often themselves connected together in each case, in such a way as to
enclose the drab-coloured ovate spot; lateral margins of dorsum pale smoke-grey ;
dorsum clothed with minute, appressed, shining pale yellowish hair (in case of 9,
with minute, appressed black hairs on the dark markings), lateral margins in ¢
clothed with longer whitish hair, seventh segment m © and sides of fifth, sixth and
seventh segments in ¢ with longer blackish hairs ; venter drab-grey (hind borders
of segments cream-buff), clothed with short, appressed, pale yellowish hair, seventh
segment also with the usual relatively coarse, erect, black hair. Wings: veins pale
sepia-coloured, here and there, especially anterior transverse vein and base of anterior
branch of third vein, distinctly darker. Squamae smoke-grey, borders paler.
Halteres : knobs seal-brown (cream-bufi at extreme tips), stalks cream-buff. Legs:
light buff, femora greyish pollinose, front pair sometimes streaked with mouse-
grey above or on outer side, tips of tibiae and upper surface of middle and hind
tarsi (at least distal extremities of joints of latter) sepia-coloured, front tarsi clove
brown (first jomt in 9 sometimes paler except at tip), slightly expanded in both
sexes, third and fourth joints of middle tarsi in g§ also somewhat expanded ;
femora clothed with whitish hair, long and fine in case of 4; tibiae clothed with

(637) val

308 ERNEST E. AUSTEN.

glistening pale yellowish hair, which on outer side of hind pair forms an out-
standing fringe (longer in J), which is usually mixed with black hairs; tips of
tibiae and upper surface of tarsi clothed with minute black hairs, similar hairs also
present to a varying extent on outer surface of tibiae, above tips.

S. Palestine : El Fukhari (8 miles W. of El Shellal, Wadi Ghuzze) and vicinity,
September-October, 1917. Type of g, Hl Fukhari, 21.ix.1917 (Li.-Col. P. 8S.
Lelean, C.B., C.M.G., R.A.M.C.); type of 2, same locality, 27.ix.1917 (entered
author's tent at dusk, attracted by lamp); eleven para-types as follows:—1 3, near
Wadi Ghuzze, 25.1x.1917 (Zt.-Col. E. P. Sewell, C.M.G., D.S.O., R.A.M Cc); Pee
Fukhari, 4.x.1917 (Col. J. A. Stewart, R.A.O.C.); 8 29, El Fukhari, 16.ix—7.x.
1917; 1 9, Weli Sheikh Nuran, 22.ix.1917 (Lt.-Col. Arnold, A.A.M.C.).

Tabanus arenwvagus, which was not met with outside the Fukhari area, was fairly
common during the latter half of September and early part of October, 1917, in
XXth Corps Headquarters Camp at E] Fukhari, where it was the only Tabanus
seen. The spot, though sparsely dotted with low scrub, was a perfectly dry, sandy

Fig. 10. Tabanus arenivagus, Austen: a, head of gin profile ; b, head of 9 from
in front, x 10; b’, antenna of 9 from the side, greatly enlarged.

one, entirely waterless, on the edge of a desert region, and there was no obvious
breeding-place whence the flies could have come, nearer than an occasional damp
hollow among the sand-dunes near the coast, several miles further west. The
author twice took QQ inside his tent at dusk, and at the same time on 22.1x.1917
was bitten on the hand by a third specimen in the open; on another occasion
(26.1x.1917) an officer was attacked by one of these flies, while sitting inside a mess-
shelter constructed of reed matting.

In size, general appearance, general coloration, and in the pattern of its
abdominal markings, Tabanus arenivagus somewhat resembles T. pulverifer,
Walk., the typical series of which was obtained at Baghdad. The species described
above, however, is distinguished from 7. pulverifer, inter alia, by the presence
of a shining band on the frontal triangle of the gj and subcallus of the 9; by the
lower frontal callus in the @ being paler, and usually smaller and Jess sharply
defined ; by the presence in the 9 of a relatively large, well-marked, subquadrate

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 309

upper (median) frontal callus, instead of the ill-defined transverse band, which
represents it in the case of 7’. pulverifer ; by the distal segment of the palpi in both
sexes being narrower (less sharply pointed in the 9) and deeper in tint, and in the
ease of the 2 bearing many black hairs on the outer surface; by the first joint
of the antennae in both sexes being smaller, and the third joint paler, narrower,
and much more elongate; by the wing-stigma being more sharply defined, and
by there being no trace of an appendix to the anterior branch of the third vein.

Tabanus arenivagus likewise resembles an undetermined species represented
in the British Museum collection by a solitary 9 from Arabia (Muscat), but is
distinguished by the front in the @ being more distinctly narrower below (the
inner margins of the eyes being convergent) ; by the upper portion of the sub-
callus in the @ being shining, and the lower frontal callus much smaller, and not
reaching the inner angles of the eyes; by the smaller size (less prominent upper
angle) of the first joint of the antennae ; by the dorsum of the abdomen having a
longitudinal series of pale triangles in the median line, instead of a broad, pale,
longitudinal stripe of more or less uniform width ; and by the stigma of the wing
being more distinctly marked.

13. Tabanus accensus, sp. n. (figs. 11, 12).

°.—Length (3 specimens) 12 to 13°4 mm.; width of head 3°8 to 4:25 mm. ;
width of front at vertex just under 0°6 to 0°6 mm. ; length of wing 9 to 9°75 mm.

Small, greyish species, with two frontal calli in Q; bare eyes ; dorsum of thorax
neutral grey or deep neutral grey, longitudinally strvped with pale olive-grey ; some-
what narrow straight-sided abdomen, smoke-grey or pale pinkish buff above, with four
longitudinal series of elongate blackish brown marks, lateral borders of segments pale
smoke-grey ‘pollinose, hind borders of second and following segments cream-buff ;
and parti-coloured legs.

Head: front in 9 smoke-grey pollinose, subcallus pale pinkish-buff pollinose,
front clothed with short, pale yellowish or bright ochreous appressed hair, in region
of vertex sometimes also with short blackish or dusky hair, no trace of an ocellar
tubercle but a tiny, dusky, mouse grey or deep mouse grey spot in middle line in
region of vertex; face and jowls greyish white pollinose, clothed with whitish
hair ; occiput pale smoke-grey pollinose, hind margin fringed above with very
short and inconspicuous whitish or yellowish white hair; front in 2 of moderate
breadth or relatively somewhat broad, slightly but distinctly narrower below, about
four times as long as its breadth at the lower end; lower frontal callus shining
blackish brown, large, quadrate and nearly square, extending from eye to eye or
separated from eye on each side by narrowest possible pollinose interval, lower
margin convex and descending slightly below level of inner angles of eyes, upper
edge with two small emarginations and consequently produced into three points,
upper frontal callus black or blackish brown, oval or elliptical oval, not connected
with lower; eyes of 2 wm life with upper angles and lower borders dark purplish
brown, and central portion metallic bronze or bronze-green, crossed by two narrow,
deep blue, horizontal bands; palpi ivory-yellow or cream-coloured, proximal
segment clothed with long whitish hair, distal segment in 2 acuminate, moderately
or considerably swollen at base, clothed on outer side with minute, appressed,

(637) G2

310 ERNEST E. AUSTEN.

glistening whitish hairs, mixed with minute black hairs; first joint of antennae
light drab or light pinkish cinnamon pollinose, moderately swollen towards distal
extremity, with upper distal angle considerably produced and partly embracing
second joint, clothed on outer side with glistening yellowish hairs and with minute
black hairs on upper border and towards upper distal angle, second joint of antennae
light drab pollinose, with distal margin fringed on outer side below with minute
glistening yellowish hairs and upper distal angle moderately or considerably
produced, third joint cinnamon- or orange-cinnamon-coloured (annulate portion
sometimes darker, dark brown or blackish brown), expanded portion of moderate
depth or rather narrow, with blunt but prominent angle on upper margin near
base. Thorax: dorsum clothed with minute, appressed, yellowish or whitish,
silky hairs, yellowish tinge in the hair being more apparent in interspace between
posterior ends of the two admedian pale olive-grey pollinose stripes (median pale
olive-grey stripe becoming obsolete at about level of transverse suture); lateral
borders of dorsum pallid neutral grey or pale smoke grey pollinose, swelling in

Fig. 11. Tabanus accensus, Austen, 9. 4.

depression at each end of transverse suture drab-grey, clothed below with long,
dusky or blackish hair, and above with shorter yellowish hair ; pleurae and pectus
pale smoke grey pollinose, clothed with whitish hair. Abdomen: middle line of
dorsum, from first (visible) to sixth segments inclusive, occupied by a longitudinal
series of anteriorly directed, truncated, smoke grey or pale pinkish buff triangles ;
blackish brown marks, except in some specimens admedian ones on first tergite,
not reaching hind borders of segments; in some cases, on each segment from
third or fourth to sixth inclusive, the ends of each pair of blackish brown marks
are fused together posteriorly in such a way as to enclose intervening smoke grey
or pale pinkish buff area, which then assumes appearance of an elliptical oval,
more or less oblique spot ; pale areas of dorsum, including hind borders of second
to sixth tergites inclusive, clothed with minute, appressed, glistening, pale yellowish
or pale ochreous hairs, blackish brown areas clothed with minute, appressed, black
or blackish hairs; seventh tergite, except lateral and hind borders, dark brown,
smoke grey pollinose at base, clothed with fine, fairly long, curving, black or blackish

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. a lp |

hair; venter pale smoke grey or pale drab-grey pollinose, clothed with minute,
appressed, glistening, ivory yellow hairs, hind borders of second to sixth scutes
inclusive ivory yellow, seventh segment deep greyish olive or mouse grey pollinose,
clothed with usual coarse, erect, blackish hair. Wangs: clear, hyaline, veins lighter
or darker olive-brown, anterior branch of third longitudinal vein without an
appendix, at least in typical form ; stigma pale isabella-coloured or almost colourless,
inconspicuous. Sguamae ivory yellow, borders cream-buff. Halteres: knobs
ivory yellow, stalks pale cream-buff. Legs: coxae smoke grey or pale smoke grey
pollinose, clothed with whitish hair; femora, except tips which are cream-bufi,
light greyish olive or smoke grey pollinose, clothed with whitish hair (extreme
tips, at least of front and middle femora, with minute black hairs above), inner and
under sides of front femora blackish brown or warm sepia-coloured, inner sides
more or less smoke grey pollinose ; tibiae cream-buff, tips (distal sixth in case of
middle and hind pairs, distal fourth or rather less in case of front pair) blackish
brown or dove brown, pale area of tibiae clothed with minute, appressed, glistening
ivory yellow hairs, mingled on extensor surfaces with minute black hairs; front

Fig. 12. Tabanus accensus, Austen, head of 9
from in front, x 10; a, antenna from the side,
greatly enlarged.

tarsi blackish brown, first joint more or less clothed above, except at distal extremity,
with minute appressed, glistening ivory yellow hairs; middle and hind tarsi clove
brown, proximal two-thirds or three-fourths of first joints cream-buff or cream-
coloured, clothed with minute, appressed, glistening ivory yellow hairs, mingled
with minute black hairs.

Lower Jordan Valley and Jerusalem: type from Wadi el Kelt, Jericho Plain
11.v1.1918, biting author’s arm; a para-type from Wadi Mellaha (about 63 miles
N.E. of Jericho, near R. Jordan), 26.v.1918, on horse; a third specimen labelled
* Jerusalem,” received in 1912 from the Zoologisches Museum, Berlin.

In size as well as in thoracic and abdominal markings Tabanus accensus, at least
in the female sex, presents a distinct resemblance to the Ethiopian 7. pertinens,
Austen; the female of the new species is, however, at once distinguishable,
unter alia, by the much narrower front, by the presence of two black or blackish
brown frontal calli instead of the single callus being either absent or scarcely notice-
able, and by the very different coloration and markings of the legs, especially
the tibiae and tarsi.

312 ERNEST E. AUSTEN.

14, Tabanus leleani, sp. n. (figs. 13, 14.)

$9.—Length, J (8 specimens) 12 to 14 mm., 9 (17 specimens) 11 to 14°6 mm. ;
width of head, j 4°4 to 5:4 mm., 9 4:25 to 5-4 mm.; width of front of 9, at vertex
08 to 1 mm., across lower edge of frontal callus 0°5 to 0°75 mm.; length of wing,
6 8°75 to 10°38 mm., 2 8°75 to 11°6 mm.

In general appearance looking like a greyish form of Tabanus cordiger, Mg., with
which it closely agrees as regards pattern of abdominal markings, and dimensions and
other details of 9 front. Eyes under an ordinary hand-lens appearing bare in both
sexes, in § with a sharply contrasted area of greatly enlarged facets, and in both sexes
with a single; horizontal, transverse band ; dorsum of thorax (in undamaged specimens)
mouse grey, clothed with fine erect hawr, interspersed with a coating of fine, appressed,
short, silky hair, of an ochreous or somewhat paler colour; dorsum of abdomen dark
mouse grey or blackish mouse grey (when viewed at a low angle from behind suffused,
like thorax, with pale neutral grey blcom), with, in middle line, a longitudinal series
of pallid neutral grey triangles, and, on each side, midway between middle line and
lateral margin, a series of oblique and roughly elliptical spots of same colour, sides of
two or more of the proximal segments, from posterior angles of first (visible) to fourth
inclusive, often more or less cinnamon ; wings hyaline, stugma scarcely distinguishable,
anterior branch of third vein typically without appendia.

Fig. 13. Tabanus leleani, Austen,2. 4.

Head: face and jowls whitish pollinose and clothed with white hair, occiput
in both sexes, frontal triangle in g, and front and subcallus in 9 pale olive-grey
pollinose ; frontal triangle in J, immediately below level of enlarged facets of eyes,
crossed by a somewhat deep, dusky, shimmering pollinose, horizontal band ; in
both sexes a dark clove-brown horizontal band between base of antenna and eye
on each side; hind margin of upper border of occiput fringed with short, erect or
forward-curving, whitish or pale yellowish hair, among which are occasionally
intermixed on each side a few short black hairs, though, at least in the case of the
g, the fringe is much shorter than the corresponding one in T. cordiger, Mg. ; front

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 313

in 2 broad, slightly narrower below, its length equal to rather less than four times
its breadth at lower margin of lower callus, clothed above with minute black hairs
and below with appressed yellowish hairs, vertex darker (dark olive-grey), but
without trace of ocellar tubercle ; two frontal calli in 9, lower frontal callus shining
black or blackish-brown, large, broad, quadrate, about four-fifths as high as broad,
only separated from eye on each side by an exceedingly narrow, pollinose interval ;
upper frontal callus in 9 represented by a large black mark, rather less than half-
way between upper margin of lower callus and vertex, separated from eye on each
side, bare of pollen but in undamaged specimens clothed with minute, appressed,
yellowish hairs, and usually either roughly V-shaped, or transversely oval or
elliptical, with a deep, narrow indentation in middle line above; eyes in 3
with sharply defined hind and upper border of small facets, not tapering off but
extending in full depth to vertex, while greatest depth of lower border, which also
consists of minute facets, is about 24 times that of upper border; eyes of 3 in
life with area of large facets greyish brown or a beautiful steely grey, lower
border consisting of small facets purplish bronze or reddish brown, a dark
horizontal band, coloured like the band in 9, close to upper margin of lower border,
hind and upper border dark purplish brown ; eyes of 9 in life bronze-green with
a purplish sheen, or purplish bronze, with a single, horizontal, dark purple, dark
purplish brown, or dark brown band, starting from inner margin of each eye
adjacent to lower frontal callus, extending outwards and backwards, and ending bluntly
at about one-fifth of the diameter from hind margin; palpo light buff, or ivory
white (proximal segment mouse grey at base), clothed with whitish hair, in case
of g usually mingled towards distal extremity of greatly swollen, pyriform or ovate
pyriform, terminal segment with a few minute black hairs; terminal segment
of 2 palpi acuminate (proximal half greatly swollen), its outer surface clothed with
minute, appressed, glistening white or yellowish white hairs, frequently mixed
with minute, scattered, appressed, black hairs ; first and second joints of antennae
blackish (first joint somewhat greyish pollinose above), clothed below with
glistening whitish hair and above with minute black hairs, first joint greatly swollen,
its bluntly produced upper distal angle overlapping the sharp corresponding angle
of the small second joint, third joint warm sepia-coloured to fuscous-black, its
expanded portion moderately broad, with a blunt but fairly prominent angle close
to base of upper margin. Thorax: dorsum with usual light grey longitudinal
stripes more or less distinctly marked, erect hair partly blackish, partly pale (light
buff), except on anterior margin, where it is entirely whitish ; swelling in depression
at each end of transverse suture greyish pollinose vinaceous-fawn, clothed mainly
with black hair; humeral and postalar calli, and lateral border of dorsum above
base of wing on each side, clothed with whitish hair; pleurae and pectus pallid
neutral grey pollinose, clothed with fairly long, fine hair of same tint. Abdomen:
lateral margins of tergites, from second or third (visible) to distal extremity, and
posterior borders of dorsal and ventral scutes of all segments cream-buff; basal
angles of first (visible) tergite more or less greyish pollinose ; pallid neutral grey
markings of dorsum, described in diagnosis above, clothed with minute, appressed,
cream-coloured hairs, and usually most distinct (sharply defined in 9) on second and
three following segments, but sometimes distinguishable on first and sixth segments

314 ERNEST E. AUSTEN.

also; in the case of the second to fifth segments inclusive the median triangles,
which are based on the hind margins and taper sharply forwards, reach the anterior
margins, while in the 9 the oblique spots sometimes similarly extend over the
full length of the segments ; dark areas of dorsum clothed with minute black hairs—
towards lateral margins of fifth and following segments in g, and on sixth and
seventh segments in Q, the hair is longer; lateral margins of all segments from
second to sixth inclusive clothed with fairly long whitish hair ; venter pallid neutral
grey pollinose, clothed, except in case of last (visible) segment, which is bedecked
with the usual coarse, outstanding black hairs, with minute, appressed, glistening,
creamy-white hairs, a certain number of black hairs on penultimate segment also,
on median area in front of hind margin. Wangs: veins dark clove-brown, except
basal portions of second and fifth longitudinal veins, which are paler (sepia-
coloured) ; stagma when discernible, pale cinnamon, narrow. Squamae cream-
buff, borders darker (isabella-coloured). Halteres ivory yellow, stalks somewhat
infuscated, and knobs usually seal-brown at base above and below. Legs: coxae
and femora (except extreme tips of latter, which are cinnamon-buff, and inner
and under sides of front femora, which are black—sometimes greyish pollinose
in case of inner side) pale neutral grey pollinose, clothed with whitish hair ; front
tibiae black, their proximal halves, or rather more or rather less, cinnamon-buff
or pale pinkish buff, and clothed with minute, appressed, whitish hairs, which may
extend on to the black area, outer surface of front tibiae also with longer hairs,
some whitish others black, or whitish on proximal half and black towards distal
extremity ; middle and hind tibiae, except extreme tips, which are tinged with
seal brown, cinnamon-buff, clothed with whitish hair, which is long (relatively
dense also in the case of the 3) on outer side of hind pair ; it is also long on outer
side of middle tibiae, where however it is sparse and intermixed with scattered
black hairs (likewise sometimes present on outer side of hind tibiae) ; front tarsi
black, not expanded, hind tarsi and upper surface of middle tarsi seal brown or dark
brown, under surface of middle tarsi, as also upper surface of fifth joint and of
proximal two-thirds of first joint, cinnamon.

Type of ¢ in Wadi el Kelt, Jordan Valley, near Jericho, 1.vi.1918; type of 9,
and three para-types of the same sex in Wadi el Aujah (64 miles north of Jericho),
Jordan Valley, 19.1v.1918; additional para-types as follows :—3 gd, (1 resting
on rocky bank above pool) 3 29, in Wadi Ghuzze, near Tel el Fara, 5.vi., 4-23.vi.,
and 8.vii.l1917; 3 99, Ain es-Sultan, near Jericho, 22.1v.1918; 3 39 (resting
on cliff wall, 5.30 p.m.), 1 & (tried to bite author, 10.30 a.m.), in Wadi el Kelt, near
Jericho, 3-5.vi.1918; 1 9, Wadi ez Zerka, 5.viii.1918 (Captain (acting Lt.-Col.)
W.J. Dale, O.B.E., R.A.V.C.), on horse.

As throwing light on the range of this species, it may be noted that in addition
to the foregoing, the Museum also possesses the following specimens of 7’. leleant :—
1 g (ex coll. Saunders), 1 9 (1909, Dr. G.A. Williamson), from Cyprus ; 1 9, Sbeitla,
Tunis, 6.v.1913 (G.C.Champron) ; 2 29, Biskra, Algeria, 14.1v.1894, 31.v.1893
(Rev. A. EH. Baton), and 1 2 from same locality, 1908 (Hon. W—now Lord—
Rothschild); 19, Nasiryah, R. Euphrates, Mesopotamia, April-June 1916
(Major W.S. Patton, I.M.S.); and 1 2 from Kangra Valley, Punjab, 4,500 ft.,
July 1899 (G. C. Dudgeon).

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 315

The author has much pleasure in naming this species in honour of his friend
Lt.-Col. P.S. Lelean, C.B.,C.M.G., R.A.M.C. (his cheery tent-companion during many
months of the Sinai-Palestine campaign), whose ingenuity, resource, and untiring
energy in the cause of field sanitation contributed in no small degree to the healthy
efficiency of the H.E.F.

Tabanus leleani, which was the characteristic, and indeed the only representative
of its genus met with in the Wadi Ghuzze in 1917, was again common a year later
im wadis running down to the Lower Jordan, near Jericho. In the Wadi el Aujah,
19.iv.1918, both sexes were seen resting on stones near the water’s edge, while at
the beginning of the following June, towards sunset, males were noticed in some
numbers sluggishly resting on the precipitous, cliff-like walls of the Wadi el Kelt,
a short distance above the spot at which the latter leaves the hills and enters the
plain ; under such conditions the insects could generally be boxed or captured
in tubes with little difficulty. The solitary occasion on which T. leleani was
observed to attack man has already been recorded.

Fig. 14. Tabanus leleani, Austen: a, head of gin profile; b, head of 2 from in
front, x 10; b, antenna of 9° from the side, greatly enlarged.

Apart from its much paler (greyer) appearance, T'abanus leleanc is distinguished
ifrom J’. cordiger, Mg., by the presence of a band on the eyes, by the row of erect
chair on the upper margin of the occiput in the ¢ being much shorter, finer and
less conspicuous, by the erect hair on the dorsum of the thorax being shorter and
finer, and the covering of pale yellowish, silky hair more appressed.

In general appearance and the majority of the external characters, except as
regards size, which is usually distinctly larger, the species described above agrees
with Tabanus albifacies, Lw., found in Egypt and Persia. 7. albsfacies, however,
is distinguishable at once by the presence on the eyes (at least in the Q) of three
bands instead of only a single band. As described by Loew (Neue Bewtrdge, iv,
p. 28 (1856)), the wing of 7’. albsfacies shows a well-marked appendix to the anterior
branch of the third longitudinal vein. In the specimens of 7. leleant examined
‘such an appendix is absent, except in the case of two 9 9 from Nasiryah, Mesopotamia
(Major W. S. Patton, I.M.S.).

So far as can be judged from the original description of Tabanus wnicinctus, Lw.,
the type of which, like that of the foregoing, was obtained in Egypt, 7’. leleani also

316 ERNEST E. AUSTEN.

resembles that species in general appearance, width of front in 9, and size.
T. unicinctus is however distinguishable, inter alia, by the veins in the centre of the
wing being suffused with brown, by the presence of a dark brown stigma, and by
the ground colour of the legs, with exception of a pale ring at the base of the front.
tibiae, being entirely black.

Finally it may be mentioned that Tabanus leleanz is nearly allied to T. unifasciatus
Lw., but may be distinguished by the sharply marked contrast in size between the
es and small facets in the eyes of the g, and by the greater = and blunter
termination of the purple eye band in the 9.

15. Tabanus pallidipes, sp. n. (figs. 15, 16.)

°.—Length (2 specimens) 11 to 13°6 mm. ; width of head 3°8 to 4°75 mm. ; width
of front at vertex 0°4 to 0°5 mm. ; length of wing 8°6 to 10°2 mm.

Small, greyish species, with bare eyes, narrow front in 9, two frontal callt on same
sex, the upper one narrow and elongate or almost linear in shape ; dorsum of abdomen
greyish cunnamon-coloured, darker at distal extrennty and with lighter and darker
markings ; and pale femora and tibiae.

Head: front and subcallus in 9 dark olive-buff pollinose, front lathe for most
part with minute, appressed, ochreous hairs, and with short dusky hairs on vertex,
no trace of an ocellar tubercle, but a tiny, dark mouse grey spot in ocellar region ;
face and jowls pallid neutral grey pollinose, clothed with whitish hair; occiput
light olive-grey pollinose, upper hind margin with a scarcely noticeable fringe of
short, whitish hair ; front in 9 very narrow, tapering somewhat to lower extremity,
about six-and-a-half times as long as its breadth at lower end ; lower frontal callus
shining dark brown (mummy brown) or cinnamon-brown, oval or quadrate with
its angles somewhat produced upwards and downwards, narrowly separate from
eye on each side, sometimes with a fine, median, longitudinal groove, upper frontal
callus dark brown, linear or narrowly elliptical, indistinctly connected with lower
callus by a fine median line; eyes of type an lufe dark brown, unbanded ; palm
creamy white or cream-buff, proximal half of distal segment moderately or
considerably swollen, distal half acuminate, proximal segment clothed below with
whitish hair, distal segment clothed on outer side with minute, appressed, glistening,
cream-coloured or yellowish hairs, mixed with minute, appressed, black hairs, which.
sometimes appear to be in greater number than the pale hairs ; antennae cinnamon-
coloured, first joint (sometimes second joint also) drab-grey pollinose on outer
side, moderately swollen distally, with its upper distal angle considerably produced.
and partly embracing second joint, first joint clothed on outer side above and
distally with minute black hairs, mixed with a certain number of glistening ochreous.
hairs, and below with longer yellowish hairs, second joint clothed on distal margin.
with minute black hairs and with its upper distal angle sometimes considerably
produced, third joint with expanded portion of moderate depth and having a blunt
angle on upper margin a little before middle, expanded portion slightly shorter
than annulate portion, which may be infuscated. Thorax: dorsum deep neutral,
grey, anterior border pale neutral grey, lateral borders and five more or less distinct
longitudinal stripes pale smoke grey or smoke grey, outer stripes visible only behind

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. SLT

transverse suture above bases of wings, all stripes sometimes difficult to distinguish ;
dorsum clothed with short, appressed, glistening, pale ochreous or greyish hair,
on darker areas mixed with short, erect, blackish hair, ground colour
of rather less than distal half of scutellum when denuded sometimes greyish
cinnamon-coloured ; swelling occupying depression at each end of transverse suture:
clothed with fairly long, fine black hair, mixed with yellowish hair, which
predominates above; pleurae and pectus pale smoke-grey pollinose, clothed with
whitish hair. Abdomen: dorsum from second to sixth segments inclusive with
four longitudinal rows of elongate, more or less oblique, blackish brown spots, four:
spots on each segment, namely an admedian pair (extending from anterior margin
outwards, nearly or quite reaching hind margin and enclosing an anteriorly directed,
drab-grey, truncated, median triangle), and a lateral oblique spot on each side:
extending backwards and outwards from anterior margin, partly cutting off basal
angle but not reaching hind margin, area between lateral and admedian dark.
brown spots on each side of each segment thus taking shape of a more or less oval
or elliptical oval, backwardly and outwardly directed, drab-grey spot; seventh
tergite drab, without darker markings; median drab-grey triangles resting on
hind margins and extending to front margins of segments, with their posterior
angles somewhat produced outwards; drab-grey or greyish cinnamon areas of
dorsum clothed with minute, appressed, glistening, yellowish hairs (hair on extreme.
_ lateral extremities of anterior tergites more whitish), blackish brown areas of
dorsum clothed with minute, appressed, black hairs, longer black or blackish
hairs also present on lateral extremities and posterior angles of last three tergites ;
venter light pinkish cinnamon (hind margins of second to sixth segments inclusive:
paler—light buff), suffused with pale smoke-grey or pale drab-grey pollen, and clothed.
with minute, appressed, glistening, cream-coloured or pale yellowish hair; ventral
scute of seventh segment bearing the usual coarse, erect, black hairs. Wangs
hyaline, clear or tinged with light drab, veins lighter or darker olive-brown,
auxiliary and first longitudinal veins, and proximal portion of second and fifth:
veins, paler (tawny-olive or cinnamon); stigma faintly tinged with yellowish or
almost colourless, inconspicuous. Squamae whitish or ivory yellow, borders cream-
coloured. Halteres ivory yellow or cream-coloured, distal extremity of stalks
slightly darker, at least in dried specimens. Legs: coxae drab-grey or smoke-
grey pollinose (ground-colour of front pair tawny-olive), clothed with whitish hair ;
femora cinnamon-buff or pinkish cinnamon, pale smoke-grey pollinose on outer
side, clothed with whitish or pale yellowish hair, front femora also with black
hairs above ; tibiae warm buff or cinnamon-buff (distal fourth or rather more than
distal fourth of front pair brownish), clothed with minute, appressed, glistening,
cream-coloured or pale yellowish hair, infuscated distal extremities of front pair
and extensor surfaces of all tibiae also largely clothed with minute black hairs,
hind tibiae on outer side with a fringe of longer hair, in which in some specimens.
cream-coloured in others black hairs predominate ; front tarsi dark brown or dark
sepia coloured, third and fourth joints somewhat expanded, middle and hind tarsi
mummy brown (first joints paler at base), all tarsi clothed above with minute black
hairs, front tarsi sometimes also with yellowish hairs on proximal two-thirds of
anterior surface.

318 ERNEST E, AUSTEN,

Lower Jordan Valley and Jerusalem: type from Wadi el Kelt, Jericho Plain,
11.vi.1918 ; a second specimen labelled “‘ Jerusalem,” received in 1912 from the
Zoologisches Museum, Berlin.

The species just described, individuals of which, as will have been seen, sometimes
‘show considerable variation in size, appears to be allied to the Austrian Tabanus
miki, Brauer, from which however it is distinguishable, inter alia, by the much
paler legs ; in T. mikii the femora, front tarsi and tips of the front tibiae are black.

™ ———s
"2 REDE LAAN Dai

Fig. 15. Tabanus pallidipes, Austen, 2. x 4.

Fig. 16. Tabanus pallidipes, Austen, head
of ° from in front, x 10; a, antenna from
the side, greatly enlarged.

Tabanus pallidipes is possibly identical with the Arabian T. arabicus, Macq.,
in which the femora were stated by Macquart (Mém. Soc. Roy. Inlle, 1838, p. 299;
Dipt. Exot., i, 2, 1838, p. 183) to be tawny or fawn-coloured (fauves). The original
-description of this species, however, which makes no mention of the dimensions of
the front im the © and leaves the abdominal markings to the imagination of the
reader, is so brief and faulty that, in the absence of the typical specimens, precise
identification is impossible.

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 319:

The species characterised above is also allied to 7’. laetetunctus, Becker, the typical
series of which was obtained in Persian Baluchistan ; so far as can be judged from.
Becker’s description, however, 7. pallidipes is distinguished, inter alia, by the much.
narrower front in the 9, and by the infuscated tarsi.

16. Tabanus dalei, sp. n. (figs. 17, 18.)

g.—Length (1 specimen) 125 mm.; width of head 475 mm.; length of wing
1] mm.

Dorsum of thorax dusky neutral grey, shining; dorsum of abdomen slate-black
second and four following segments with postervor transverse bands of appressed,
glistening, silvery-white ha; eyes bare; wings hyaline, extreme base, costal and
subcostal cells, and proximal half (narrow portion) of marginal cell ochraceous tawny,
contrasting sharply with remainder of surface ; legs black, extensor surfaces of all tibiae,
except at distal extremities, clothed with glistening whitish or yellowish harr.

Fig. 17. Tabanus dale,
Austen, head of ¢ in pro-
file. > about 10. .

Head of ¢ small, frontal triangle, face and jowls smoke-grey pollinose, face and
jowls clothed with whitish hair; occiput neutral grey pollinose, posterior margin:
fringed at sides and above with erect, yellowish hair, longer in region of vertex,
and mingled on each side of vertical region with a certain number of black or blackish
hairs ; ocellar tubercle inconspicuous, sunk in vertical triangle; eyes in 3 each
with an area of moderately enlarged facets abutting on inner margin, but such
facets progressively diminishing in size towards exterior and merging gradually
into remainder, to which they nowhere present a sharp contrast ; palpi in ¢ dark
mouse grey pollinose, clothed with whitish hair, terminal segment small, bluntly
elliptical ovate, not pointed at distal extremity ; first joint of antennae in ¢ blackish
mouse grey, scarcely swollen distally, sparsely clothed with short yellowish hairs,
second joint blackish brown, with a few yellowish hairs on inner side and minute
black hairs on upper distal angle, which is considerably produced, third joint dark
clove brown or blackishjbrown (slightly paler at extreme base), expanded portion
of moderate depth or fairly deep, at least in g, with a blunt but prominent

320 ERNEST E. AUSTEN.

angle about middle of upper margin, which is moderately excavated in front of the
angle. Thorax: dorsum clothed with fine, erect, silky, pale yellowish hair,
scattered among which is a certain number of fine black or blackish hairs, which
-on scutellum are much more numerous and even appear to predominate, anterior
border of dorsum with commencements of two light greyish olive, pollinose,
longitudinal stripes; swelling occupying depression at each end of transverse
suture agreeing in coloration with remainder of dorsum and similarly clothed with
hair, but without black hair; pleurae and pectus clothed with yellowish white or
whitish hair, the tuft of long hair on hind margin of mesopleura conspicuously
ochreous at base; pleurae neutral grey pollinose above, pectus and lower part of
pleurae deep neutral grey. Abdomen: dorsum of first (visible) segment neutral grey
pollinose (greyish olive pollinose on each side), clothed in centre and on each side with
glistening yellowish white hair ; posterior transverse bands of appressed, glistening,
silvery-white hair, on second to sixth segments inclusive, expanded in middle
line so as to form anteriorly directed triangles ; silvery-white hair also extending

Saeco
Zz tii

Fig. 18. Tabanus dalei, Austen, wing of g. x 10.

from front to hind margins on lateral borders of segments referred to; on third,
fourth and fifth segments commencement of expansion of hind border on each
side includes a certain amount of appressed, glistening, ochreous hair; ground-
colour of hind margins of first six tergites cimnamon-drab; dorsum, except as
already stated, clothed with short, appressed, black hair; venter deep neutral
orey pollinose (hind margins of segments as on dorsum), clothed with silvery white
or appressed silvery white hair, which is more conspicuous on hind borders of
segments, and in central area of anterior part of second, third and fourth segments
is largely if not entirely replaced by minute black hairs. Wungs: veins dark brown
or mummy brown, proximal portion of fifth longitudinal vein, forming lower
boundary of second basal cell, ochraceous-tawny and, like corresponding portion
of fourth longitudinal vein, narrowly suffused with similar colour ; anterior branch
of third longitudinal vein without appendix; stigma ochraceous tawny, indis-
tinguishable from coloured costal border ; proximal portion of marginal cell with
a narrow hyaline streak on its anterior border. Sqguamae ivory yellow, borders
‘cream-buff or cream-coloured. Halteres: knobs clove brown, stalks mummy
brown. Legs: coxae neutral grey pollinose, clothed with yellowish white hair ;
middle and hind femora and posterior surface of front femora clothed with

CONTRIBUTION TO KNOWLEDGE OF TABANIDAE OF PALESTINE. 321

glistening, yellowish white or ochreous hair; pale hair on hind tibiae in ¢ fairly
long and bushy; distal extremities of tibiae and upper surfaces of tarsi clothed
with minute black hairs.

Near Jerusalem, May, 1918 (Captain (acting Lt.-Col.) W. J. Dale, O.B.E., R.A.V.C.).

The species characterised above, with which the author is glad to have the oppor-
tunity of associating the name of its discoverer (A.D.V.S., XXth Corps, 1917-19),
is not closely related to any of its congeners known to the writer, and, owing to the
ochraceous-tawny coloration of the base and proximal two-thirds of. the costal

border of the wing, is readily distinguishable from all other species described or
mentioned in the present paper.

) OR) SES aaa NORTON ERROR
¥ Ss SPO Sites is SO,
5)
‘ :
yok ney
AL

‘
ji
1%
1 f
}
r .
‘ i Ay
on
ie
ie

eh hid».

AL:

few,
’ ee

323

NOTES ON SOME CULICIDAE COLLECTED IN LOWER MESOPOTAMIA.
By Lieut. P. J. Barraup, R.A.M.C.

From October 1918 to February 1919 inclusive I was stationed at Basrah,
Lower Mesopotamia, and during that time I was able to devote some time to the
collecting of CuLicrpDaE and their larvae.

My investigations were necessarily confined to a somewhat restricted area,
possibly five miles from north to south, and one mile wide, embracing the districts
known as Ashar, Makina, and Magil, along the right bank of the Shatt-el-Arab
river. Some collecting was also done on the left bank from Tonooma, opposite
Ashar, northwards. Practically the whole of this ground is within the cultivated
belt lying between the river banks and the desert, and is intersected by numerous
creeks, and irrigation canals. There were also in some places many pools and
swampy areas forming good collecting ground.

The wet season commenced towards the end of November and continued to the
end of my stay, but the rainfall was intermittent and not considerable in amount.
In December there were several violent storms with heavy rain for a time.

From official temperature records I made the following notes :—

October 1918.—Maximum shade temperature between 102°F. (5th and 7th) and
88°F. (23rd and 30th). Minimum shade between 75 (5th) and 57 (24th).

November 1918.—Maximum shade between 89 (3rd, 6th and 7th) and 69 (9th).
Minimum shade between 63 (7th) and 49 (several days in the last week).

December (up to 22nd, beyond which date I have no records).—Maximum shade
between 76 (Ist) and 60 (10th, 11th and 12th). Minimum shade between 60 (17th)
and 38 (13th).

There was no frost or snow throughout the winter.

The “ main river ” referred to in the following notes is the Shatt-el-Arab, formed
by the junction of one of the mouths of the Euphrates and the Tigris, about five
miles north of Ashar.

I am much indebted to Dr. Guy A. K. Marshall, Imperial Bureau of Entomology,
and Mr. F. W. Edwards for kind assistance in the working out of my material ;
to Major 8. R Christophers, I.M.S., for giving me facilities and help in my work
at Basrah ; and to Capt. H. E. Shortt, I.M.S., for his valued companionship on many
a delightful excursion. |

Anopheles stephensi, Liston.

This appears to be the chief malaria carrier of the district and bred continuously
during the time I was at Basrah. Numbers of adults and larvae were found in
October, November and December.

The larvae occur chiefly in the clearer pools in the cultivated area. On 14th
January 1919 [ found about one hundred larvae in a small pool near Ashar, many
of them full-grown. Adults were bred out in the laboratory during the following
ten days. In February only a few larvae were found.

(637) H

324 P. J. BARRAUD.

Anopheles pulcherrimus, Theo.

Anti-malarial work in the cultivated belt seems to have ia more effect on this
species than upon the preceding one, and the numbers have been greatly reduced.

Large numbers of larvae were, however, found in pools near the edge of the desert
in October. Sometimes the larvae of this species and those of A. stephensi occur
together in the same pool, but those of A. pulcherrimus usually prefer weed-grown
stagnant and more brackish water. Camps near the desert are at times infested with
large numbers of A. pulcherrimus adults.

Anopheles sinensis, Wied.

Major Christophers has described a new form of this species occurring in
Mesopotamia, viz. var. mesopotamiae (Indian Jl. Med. Res. i, p. 196). The
species is not very common in the Basrah area. On 18th December I saw it
for the first time and caught three females and some small larvae possibly of this
species, in the marshes on the left bank of the main river. Subsequently a few more
were taken in the same place. On 18th January one male emerged in the
laboratory from a nymph collected in Makina area. Later on more were found
on the banks of a creek between Magil and the Euphrates.

Culex fatigans, Wied.

Abundant throughout the cultivated area during my stay. The larvae were
found in extraordinary numbers in some pools, appearing as a dark mass beneath
the surface.

Culex pipiens, L.

Far less common in my experience than C. fatigans. I did not identify any larvae
of this-species until January. In mounting numbers of larval skins I did not find
much difficulty in separating the two species.

Culex modestus, Fc.

Some larvae of this species were collected in October and November, but they were
not very numerous. In one case they were found in a pool containing numbers of
A. stephensi and A. pulcherrimus.

Gulex tritaeniorhynchus, Giles.
Larvae were found in October and November in Ashar and Makina areas in small
numbers.

Culex tipuliformis, Theo.
Larvae of this species were not found until January, first in isolated pools on the
left bank of the main river, and later in larger numbers in Ashar district.

Stegomyia fasciata, I.

Very few specimens met with in Basrah area during my stay, and no larvae found.
On 23rd November, when visiting Mohommerah, lower down the main river on the
Persian side, adults were found in small numbers in a building on the river bank.

NOTES ON SOME CULICIDAE COLLECTED IN LOWER MESOPOTAMIA. a2)

Ochierotatus dorsalis, Me.

This mosquito was found quite commonly over the whole area, both adults and
larvae, and is the only species of those mentioned which occurred in about the
same numbers in the cultivated belt and in the outlying places visited on the edge
of the desert. In one case larvae were found in a collection of rain-water in the
bottom of a boat in Ashar town, and some were found in pools in the desert at
some distance from cultivation. Adults and larvae occurred from December to
February.

Theobaldia longiareolata, Macq.

The first specimen emerged on 27th December in the laboratory, the larva having
been found in Ashar area. Towards the end of January and in February the larvae
became more numerous, being found chiefly in the deeper pools. The pupa may
often be seen about four inches below the surface of the water and appears to
retain that position by gentle movement, remaining at that depth for quite a long
time.

Theobaldia annulata, Schranck.

Larvae were not found until 24th February and were then not uncommon in Ashar
district.

The adults have the thorax and abdomen almost uniformly light reddish brown,
with very little trace of the normally conspicuous abdominal banding. This inter-
esting colour modification, which is unaccompanied by any structural change, is
no doubt adaptive to desert conditions. A similar variation has been noticed in
several other mosquitos.

Port Said, June 1919.

(637 a

; la | AAW RS oH: reg, a eh Rio scan | |

| te ae pach iia ;
dea oe pat ag a abe
| : . he hia Spee, ne ' i a ie i ae 7 aces

74 onl eR AD DG

nf ; se boieh cis a, Hest nese Saas BP
a f ing adie i FS Osh too, ea aed
6 , en ba hee eee
; em Bh phy sth Vea)
a, Dag Diya: nieici ae ;
‘ ¥ ie j is "
i idle . 4 thay: pie ib add ah iene . “bawet! Deri si
¥ 4 # gt Vial ihe. thw tsa bis ‘salad fee his OR, os |
. ibaa fag hate LALLA GON ce ues "ees Anne or
AN 4 ie ha & = ’ “is be Pe eae NRE iv aye & 1 se: ’ - y
3 4 ie a bane oe ¥ i trek. yaa? A i vpeeailn ; es sia ‘siete pe Bic ” oii i
83 ie siearenas) dod ‘his wali “fbn ne tig beat, fone. nies

es is > etter: Seah “ ; :
Bete i's isl dalbha: 7 ian, ingle wane
ae aint ean ji Spo |
ae Ie ata Spek foeicaaungeth
re "et 1H) wk anid ical e AAlinde he

| “meh 9

ae es ; \ ‘
ee ee fee
AK ; Se ae 9 ay 1S eat ae ENG ad Chee ee Fi H seid Au wis dbs
’ .
| ae hc yt es ae lc ke TCLS. wuigtboias i werete Ney

F pi iy eo r é \ V2 Te gk a t vii this

i = 4 f

* s } pe

all . iy

—
»~

rm so Bes! a he atl t

eet) eeu sid th gine vont Hse iat ta vie beh 2 amb hl eh
i. F “ 7 be dj a : oot)

: io rt Ty | pr

’ | a re ; 1 : id 5 o = ai 4 J
i" si by ay a
+4 4 \ ’ i]
ee ae. s sane xipuctn, | i
a af ‘ ibe hon pi ero ‘ ; roe Shad
Ney, ey An pas 5 Shes Porat: et 3 ea Pao PY ‘, 2% ie Beret nurs:
>
j
fi
hy
yeey Nas
j 1%.
a
, a es
1 Pi oe cre

oat

ON A NEW MOUTILLID PARASITE OF GLOSSINA MORSITANS.
By Row ann EH. TuRNER.

Mutilla auxiliaris, sp. nov.

Q. Nigra; thorace brevissimo, fusco-ferrugineo ; tergitis tribus basalibus fascia
apicali angusta pallide aureo-sericea ; area pygidiali nulla.

§. Niger; prothorace, mesonoto, scutelloque ferrugineis; tergitis quatuor
basalibus albido-fimbriatis ; alis fuscis, basi dilutioribus, posticis basi hyalinis,
venis nigris, calcaribus intermediis posticisque pallidis.

Long. 9,5 mm.; ¢, 7 mm.

Q. Mandibles simple ; antennae rather stout ; second joint of the flagellum short,
scarcely longer than the third; the apical joint rather slender, longer than the
penultimate. Head rugosely punctured ; eyes oval, situated nearer to the base of

Fig. 1. Mutilla auxiliaris, Turner, sp.n., 2d.

the mandibles than to the hind margin of the head. The whole insect rather sparsely
clothed with long blackish hairs. Thorax as broad as the head, punctured-rugose,
almost as broad in the middle as long ; the sides distinctly, but not strongly, convex ;
the anterior margin not quite straight, widely and very shallowly emarginate, the
posterior slope rather abrupt, no scutellar tubercle, the sides of the apical slope
with several minute teeth. Abdomen strongly punctured, the punctures more or
or less confluent longitudinally ; first tergite broad, transverse, almost as broad
as the second, abruptly truncate anteriorly ; second tergite as broad in the middle
as long, the sides distinctly, but not very strongly convex ; no pygidial area. Hind
tibiae with a row of three spines ; tarsal ungues simple.

§. Clypeus strongly concave, shallowly emarginate at the apex ; bordered laterally
by carinae which converge towards the base, the apical margin with several pale
setae. Mandibles bidentate on the right side, tridentate on the left ; labrum trans-
verse. Head closely punctured, the clypeus smooth and shining; antennal
tubercles strongly developed, the front with an obscure longitudinal sulcus which does

a) a ROWLAND E. TURNER.

not extend to the anterior ocellus. Second joint of the flagellum distinctly shorter
than the third, the fourth slightly longer then the third. Pronotum and mesonotum
evenly and rather strongly punctured, the mesonotum short and broad ; scutellum
more coarsely punctured ; median segment short and broad, very coarsely reticulate.
Abdominal tergites rather less strongly punctured than the thorax; first tergite short,
transverse, narrower than the second which is very broad, being nearly three times
as broad at the apex as long, with the sides strongly convex ; seventh tergite broadly
subtruncate at the apex. Second abscissa of the radius longer than the third, first
transverse cubital nervure curved inwards, second curved outwards, first recurrent
nervure received close to the middle of the second cubital cell, second beyond
two-thirds from the base of the third cubital cell.

PortuGuEsE East Arrica: Kanezi, Mossurise district, vu. 1918 (C. F. M.
Swynnerton).

Bred from puparia of Glossina morsitans, 5 99, 1 ¢.

Allied to M. glossinae, Turn., but easily distinguished by the shorter thorax and
second tergite of the female ; and in the male by the much more distinct transverse
carina at the base of the dorsal surface of the first tergite and the distinctly shorter
second tergite. The two species are very closely related, the neuration of the males
being very similar, both differmg in this and many other respects from
M. benefactriz, Turn., the only other species which has yet been bred from Glossina.

O29

NOTES ON THE IDENTIFICATION OF ANOPHELINAE AND THEIR
LARVAE IN THE ZANZIBAR PROTECTORATE.
By W. MANSFIELD-ADERS,
Economic Biologist, Zanzibar Government.

Medical Officers and others interested in tropical sanitation often find great
difficulty in rapidly identifying adult mosquitos. Four Anophelines have been
found in the Zanzibar Protectorate, two of these being well-known malaria carriers.
The identification of the two latter with the aid of a pocket lens is quite simple.

Adult Anophelines.
Anopheles costalis, Lw.

This is the commonest Anopheline in the Protectorate and the species most often
brought in by the mosquito brigade for identification. It is a large insect and can
at once be recognised by the golden yellow stippling on the first pair of legs, which
is specially marked on the tarsal joints.

Anopheles funestus, Guiles.

Adults of this species are very uncommon in the town, but are constantly captured
in the outlying districts. This Anopheline is easily recognised by its small size
and generally black colour. There are no signs of stippling on the legs, therefore
it cannot be confused with A. costalis.

Anopheles mauritianus, Grp.

Adults of this form are very rarely captured. It is one of the easiest species to
identify, its chief characteristics being its large size, the very distinct long black

spots on the costal vein, and especially the pure white tarsal joints of the hind pair
of legs.

Anopheles squamosus, Theo.

This mosquito somewhat resembles Stegomyza fasciata in having the legs spotted
with black and white markings, but the outstanding characteristic peculiar to it
is the presence of lateral abdominal hair tufts.

Anopheline Larvae.

The identification of the Anopheline larvae is a more difficult problem, though

three of them, A. costalis, A. funestus and A. mauritianus, have marked
characteristics.

330 W. MANSFIELD-ADERS.

Anopheles costalis, Lw.

This is the commonest Anopheline larva. It is generally found in small collections
of rain-water, such as road-side puddles, borrow-pits, water-holes (dug by the natives
for storing water), shallow swamps surrounded by grass, occasionally in cement
tanks, oftener in those with algid growth attached to their sides, at times in rain-
water collected in dug-out canoes and boats. The larvae of this species vary greatly
in colour, some being quite light, some dark, others almost green, the variations
being caused by the type of food ingested.

The larva has a small head, the hairs on the first three segments of the abdomen
are not markedly feathered and the palmate hairs from the second to the seventh
segment are not very conspicuous.

Anopheles funestus, Giles.

These larvae are generally found in backwaters of streams and at the edge of the
sluggish rivers where the water is overhung by grass or other vegetation.
Occasionally they are found in association with A. costals in the Zanzibar

Fig 1. Thorax of larvae of Anopheles funestus, Giles (4th instar); a, one of the
plumose hairs, greatly enlarged.

Protectorate. The larvae are small and black in colour and much more heavily
plumed than those of A. costalis. With a pocket lens a pair of thick feathered hairs

IDENTIFICATION OF ANOPHELINAE AND THEIR LARVAE. ool

springing from a chitinous pocket near the middle of the thorax are easily recog-
nisable (fig. 1, a). This characteristic alone suffices to distinguish the larvae of
A. funestus from those of the other island species. It must be remembered that the
larvae of A. costalis carry the same hairs, but these are not nearly so conspicuous
and are difficult to detect even with a lens. The palmate hairs are better defined
than those of A. costalis.

Anopheles mauritianus, Grp.

The larvae are easily identified ; the antennae carry a conspicuous hair-tuft in
the median region and there are two very characteristic fan-shaped tufts of short
black hairs above the mouth-brushes. These two characters are sufficient for the
identification of these larvae. The abdomen in both old and young larvae is
marked with whitish bands. The living larvae have the peculiar habit of twisting
themselves into an S-shape, a habit which I have not noticed in any other Anopheline
larva.

Anopheles squamosus, Theo.
The larvae of this Anopheline are unknown to me.

Technique and Methods of Control.

A simple method of examining mosquito larvae is to remove one from the glass
collecting jar by means of a pipette, place it in a drop of clean water on a microscope
slide, and anaesthetise it by placing near it a pledget of cotton-wool soaked in chloro-
form covering all with an inverted petri dish for a few minutes. The larva can then
be moved about and turned by means of a fine dissecting needle with a minimum of
injury to its finer hairy structures.

The Public Health Officer will be greatly aided in his work by a mosquito brigade
of native boys trained to recognise the larvae of the three genera Anopheles, Culex
and Stegomyia. We supply our mosquito brigade with glass jars for collecting
larvae and with long-handled white enamelled iron ladles, which make most efficient
‘dippers and in which the most minute larvae are easily seen against the white back-
ground.

A further valuable means of indicating the presence of mosquitos and of obtaining
specimens of larvae is the use of mosquito traps. The traps used are ordinary
wooden tubs about 24 feet in diameter cut down to. a height of 6 inches. These
are sunk till their rims are flush with the ground, soil to the depth of some 2-inches
is spread in them, water is poured on to the requisite depth and finally some aquatic
weeds are floated on the surface. These traps, under the control of the mosquito
brigade, are placed at intervals round the periphery of the town.

The traps are examined every six days for the presence of larvae, all species being
recorded. The method of examining the traps is as follows. Each trap is visited
by an inspector on the sixth day, and its whole contents after agitation, are emptied
into a bucket bearing the same number as the trap. The buckets are brought to
the Health Office for critical examination. The contents of the buckets are
‘strained through fine muslin, the residue and debris left adhering to the muslin
‘being washed down into a large white enamelled dish, in which the larvae can easily

332 W. MANSFIELD-ADERS.

be identified. The muslin is soaked and washed in clean water after each
examination. Generally mixed infections of larvae are found, and these have
proved of great use in training the staff of the mosquito brigade.

All members of the mosquito brigade are trained to capture and identify adults,
a most valuable adjunct to the larval index of a town, and the senior members are
taught to recognise males and females of Anopheles, Culex and Stegomyia.

For obtaining adults they are supplied with short glass tubes about six inches
long and one inch in diameter; a plug of cotton-wool soaked with chloroform is
placed in the bottom of each and the tube is corked. With this simple contrivance:
resting adults can be caught with ease. The adults so collected are pinned out by
their respective captors and on their return to the Health Office are submitted for
examination. By these means we secure an ample supply of specimens for exchange:
or for teaching purposes.

The training of natives requires a great amount of patience, but if one is able
to interest them in their work, the trouble spent on explaining the life-history
and habits of the insects they are expected to control is well repaid.

330

THE MOSQUITOS OF FAR EASTERN PORTS WITH SPECIAL REFERENCE
TO THE PREVALENCE OF STEGOMYIA FASCIATA, F.

By A. T. STANTON,
Bacteriologist, Institute for Medical Research, Federated Malay States.

I. INTRODUCTION.

The danger of the introduction of yellow fever into Oriental countries has been.
the subject of much discussion since Sir Patrick Manson first drew attention to it
in 1903. In 1911 Major (now Lt.-Col.) 8. P. James, I.M.S., was deputed by the
Government of India to study the conditions in the endemic areas of Central America
and in the principal sea-ports between that country and India. His report, published
in 1913, was a most valuable contribution to our knowledge of the problems
surrounding this subject. Among other recommendations, Major James suggested
that further inquiry should be made with reference to the occurrence of Stegomyra
fasciata in Far Eastern ports.

Any measures that may be devised for the protection of India and the Far Eastern
colonies from yellow fever must take account of the distribution and prevalence
of the known carrier of the disease, perhaps also of its near allies, and of the sanitary
and other circumstances affecting the prevalence of mosquitos generally. It was
considered desirable to supplement the data hitherto obtained, and in September
1915 at the suggestion of the Imperial Bureau of Entomology I was instructed by
the Government of the Federated Malay States to pursue these inquiries.

The following ports were visited between October 1915 and March 1916 ; Bangkok
(Siam), Saigon (Cochin-China), Haiphong (Tonkin), Canton (South China), Batavia,
Samarang and Soerabaia (Java), Makasser (Celebes) and Tjilatjap (Java). Through
the courtesy of several correspondents I have been able also to examine specimens.
from other places in this region. It was not found possible to complete the observa-
tions by a survey of the ports of China and Japan as had been planned.

Reports of the observations (with specimens of mosquitos from the different
ports) were sent to the Imperial Bureau of Entomology as the work progressed and
these reports are now reproduced in a form convenient for reference. There is added
a summary of the observations and of the conclusions that can be drawn from a
study of them. -

I am under obligation to many officers of the Governments of Siam, French
Indo-China, and the Netherlands East Indies for generous assistance in connection
with the enquiry, especially to Drs. M. Carthew, M.O.H., and Malcolm Smith
(Bangkok), Dr. A. Denier (Saigon) and Drs. W. T. de Vogel, N.F. Lim, H. Werkman,
L. 8S. von Romer and M. L. van Breemen (Java). To Dr. G. A. K. Marshall, Director
of the Imperial Bureau of Entomology, and Mr. F. W. Edwards, of the British
Museum, I am indebted for confirmation or correction of my identifications of
mosquitos.

334 of A. T. STANTON.

If. Tur Ports VisitEp.

Bangkok.

Bangkok, the modern capital of Siam, is situated on both banks of the River
Menam, in latitude 13° 45’ North and longitude 100° 30’ East, about twenty-five
miles up-stream from the bar at the river’s mouth. The city proper, the government
offices, banks and business houses are on the left bank of the river ; on the right bank
are rice mills and other industries and the dwellings of a large part of the native
population. The city is intersected in all directions by a network of canals.
The estimated population is 625,000.

T'rade.—The trade of the port is mainly with Singapore, Hong Kong, Swatow and
Hoihow. There is also direct steamer connection with Saigon and Batavia.

Temperature and Rainfall—The monthly mean temperature of Bangkok is above
80°F. for all months of the year except November, December and January. The
annual mean rainfall is about 57 inches. The months May to October are wet
months ; December, January and February are dry.

Water Supply.—A public water supply for Bangkok was installed in November
1914. The water is taken from the Menam about twenty-five miles north of the city,
carried by a canal to the filtering plant at Bangkok and thence distributed in pipes
over the greater part of the city on the left bank of the river.

In addition to this public supply, which is as yet available for only a small part
of the population, there are artesian wells, shallow wells and the river and canals,
all of which are drawn upon. The storage of water in jars for household purposes
is the rule in Bangkok among the poorer classes.

Samitation and Quarantine.—The sanitary administration of the city is controlled
by the Ministry of Local Government. Provision is made for three European medical
officers, a Bacteriologist, and a number of Siamese assistants. No special measures
are in operation for the reduction of mosquitos.

This department also deals with quarantine matters. The quarantine station,
is situated at Koh Phra, an island about 60 miles from Bangkok. It is proposed
to construct a new station on the river about 15 miles from the city.

Mosquitos—An investigation was made in late October and early November
of the distribution of mosquitos in all parts of the city by examining specimens
taken in houses and by the identification of larvae taken in artificial and natural
collections of water in the neighbourhood of houses.

The following list of species is in order of comparative frequency of their
occurrence. |

(A). Adult mosquitos taken in houses :—Culex fatigans, Wied. ; Stegomyia fas-
ciata, F.; Stegomyia albopicta, Skuse (scutellaris Theo.) ; Mansomoides uniforms,
Theo. ; M. annulipes, Walk. ; Culex gelidus, Theo. ; C. concolor, R. D.; C. sitrens,
Wied. ; Mansonioides annuliferus, Theo. ; Anopheles sinensis, Wied.; A. senensis
var. peditaeniatus, Leic. ; A. rossi var. indefimtus, Ludl.; A. barbirostris, Wulp ;
A. fuliginosus, Giles; A. fuliginosus var. nivipes, Theo. ; Culex halifaxi, Theo. ;
C. fuscocephalus, Theo. ; Uranotaenia cancer, Leic. ; Anopheles tessellatus,.Theo.

MOSQUITOS OF FAR EASTERN PORTS. 335

(B). Larvae from artificial collections of water: jars, tubs, barrels, tins, etc. :—
Stegomyia fasciata ; Armigeres obturbans, Walk. ; Stegomyia albopicta (scutellaris) ;
Cyathomyra brevipalpis, Giles ; Toxorhynchites immasericors, Walk. ; Culex halifax: ;
C. fategans ; Anopheles rossi var. indefinitus.

(C). Larvae from natural collections of water: pools, canals, ditches, etc. :—
Culex fatigans, C. geludus, Armigeres obturbans, Culex concolor, Anopheles rossi var.
ndefinitus, A. sinensis.

The cannibal larvae of Toxorhynchites immisericors, Walk., were here frequently
encountered in water-jars, and their astonishing voracity for Stegomyia larvae,
as many as twenty-five of the latter being devoured in a single night, made one think
of the possibility of their utilisation in Stegomyia reduction. In Bangkok through-
out the city mosquitos are numerous and Stegomyra fasciaia is abundant. The
habits of the people in the storage of water and the restricted distribution of the
public water supply combine to make the task of its reduction a very difficult one.

Saigon.

Saigon, the capital of Cochin-China, is situated on the banks of the River Saigon,
a triobutary of the Donnai, in latitude 10° 46’ North and longitude 106° 30’ East.
The town lies about 40 miles up-stream from Cape St. James at the mouth of the
river Donnai and is accessible to the largest steamers. The population of Saigon
is about 70,000. The city of Cholon, which for practical purposes is a part of the
port, is situated about four miles further up-stream and is connected with Saigon
by a steam tramway. Cholon is the centre of much commercial activity in con-
nexion with the milling and export of rice, the staple product of Cochin-China.
The population of Cholon is 165,000.

Trade.—The trade of the port is with Hong Kong, Singapore, Manila, Japan
and the Netherlands Hast Indies by direct services, and with EHurope by way of
intermediate ports.

Temperature and Rainfall—The monthly mean temperature is above 80°F.
for all months of the year except December, when it falls shghtly. The annual
mean rainfall is about 65 inches. The raimy season begins about the middle of
May and ends about the middle of November, January to March being the driest
months.

Water Supply.—The public water supply of Saigon is collected from a number
of surface wells within the city and from a small catchment area about two miles
outside it. The supply is now intermittent, necessitating storage in tanks or jars.
A plan for a continuous high pressure supply is under consideration.

Samtation and Quarantine.—The sanitary administration is controlled by the
Civic Corporation through their medical officer of health, who is assisted by sanitary
inspectors. Though no special measures are directed against mosquitos, the
efficient civil sanitary service ensures that breeding places are not neglected.

The quarantine service is directed by the Chief of the Health Service of Cochin
China, an officer of the Sanitary and Medical Services of Indo-China. The
quarantine station is situated at Nhabeé, fifteen miles down-stream from the city ;
the buildings, accommodation and equipment are excellent.

336 A. T. STANTON.

These services throughout Indo-China are well maintained. All the principal
ports and many of the smaller ones are provided with rooms for disinfection and
mobile apparatus for the disinfection of ships.

Mosquitos.—The following species were met with, given in order of frequency.

(A). Adult mosquitos taken in houses :—Culex fatigans, Stegomyia fasciata,
S. albopicta (scutellaris), Armigeres obturbans, Mansoniordes uniformis, Culex
gelidus, C. sitiens, C. concolor, Anopheles rossi var. indefinitus, A. sinensis.

(B). Larvae taken in artificial collections of water :—Stegomyia fasciata,
S. albopicta (scutellaris), Ochlerotatus gubernatorius, Giles, Cyathomyia brevipalpis,
Culex concolor.

(C). Larvae taken in natural collections of water :—Culex fatigans, C. gelidus,
C. concolor, A. rossi var. wndefimtus.

At the time of my visit, in November towards the end of the rainy season,
mosquitos were not numerous in Saigon, but Stegomyra fascrata was not uncommonly
met with. In the native city of Cholon larvae of this species were common.

In the annual report for 1914 of the “ Ambulance du Cap St. Jacques ” Major
J. H. Rencurel describes an outbreak of fever of obscure causation among officers
and their families. A further small outbreak at Cape St. James occurred in 1915
and similar cases were observed at Saigon. Major Rencurel identified this fever
with the “six day fever of ports’ described by Leonard Rogers and attributed its
spread to a biting insect, mosquito or fly. Certain officers who had experience of
the fever and of the conditions at Cape St. James entertained the suspicion that
the disease might be related in some way to yellow fever.

Haiphong.

Haiphong is the shipping port for Hanoi and other commercial centres in Tonkin
and for the Chinese province of Yunnan. It is situated in latitude 20° 50’ North
and longitude 106° 42’ East, on the banks of the rivers Cua Cam and Song Tam Bac.
These rivers are joined by several channels with the river Song Koi, a great water-
way connecting Yunnan with the Gulf of Tonkin. The banks of the rivers are low
and consist of alluvial mud, from which the present site of the town of Haiphong
has been reclaimed. Scattered throughout the town are extensive pools and areas
of marsh land. The population of Haiphong is about 45,000.

Trade.—Haiphong is accessible to large ocean-going steamers. The sea-borne
trade of the port is mainly with Hong Kong, Saigon, Pakhoi in the Chinese
province of Kwang-si, and Hoihow on the island of Hainan.

Temperature.and Ranfall.—The monthly mean temperature rises above 80°F.
only during the months June to September. From December to February it is
about 62°F. The annual mean rainfall is about 69 inches, May to October being
the rainy months.

Water Supply.—The public water supply is a continuous high pressure supply
drawn from storage reservoirs in the hills and distributed over the whole area of
the city.

Samtation and Quarantine-—These services are controlled by the director of
Health Services of Tonkin, whose headquarters are at Hanoi and who is represented
at Haiphong by an official of the department.

MOSQUITOS OF FAR EASTERN PORTS. 337

The quarantine station is at Binh-Dong, about ten miles down-stream. There
is a complete equipment of buildings, materials, and staff.

Mosquitos.

(A). Adult mosquitos taken in houses :—Culex fatigans, Stegomyia fasciata,
S. albopicta (scutellaris), Armigeres obturbans, Mansomordes annulrferus, M.annulupes,
Culex sitiens, Anopheles rossi var. indefimtus, A. sinensis, A. minimus, Theo.

(B). Larvae taken in artificial collections of water :—Stegomya fasciaia,
S. albopicta (scutellaris), Armageres obturbans.

(C). Larvae taken in natural collections of water :—Culex fairgans, C. satiens.

At the time of my visit to Haiphong in December the temperature was low, 64°—
72°F., and mosquitos were few in number. A few specimens of Stegomyia
fasciata were taken.

Dr. P. L. Simond, Inspector General of the Sanitary and Medical Services of
Indo-China, reports (Bull. Soc. Med. Chir. Indochine, vii. 2,1916) that he found
a specimen of Stegomyra fascrata in Hanoi, Tonkin, in November 1915, but considers
that the species is rare in this region. Dr. Simond has also made observations
in Cochin-China, where he found specimens of Stegomyia fasciata, but concludes
that it is a rare species. This conclusion was not borne out by my examination
of breeding places in Saigon.

Canton.

Canton, the capital of the Chinese Province Kwang Tung, is situated on the
Chu Kiang or Pearl River in latitude 23° 7’ North and longitude 113° 14’ Hast.
The city proper extends to a breadth of about two miles and is about six miles
in circumference, being enclosed by a wall. The suburbs extend along the river
for about five miles. The population has been estimated by the Customs
authorities at 2,500,000.

Trade-——Ample means of communication exist between Canton and Hong Kong,
a distance of 95 miles by water and 112 miles by railway. Foreign steamers and a
large number of native craft ply daily. There is daily steamer communication
with Macao, and regular communication with Wuchow and other West River towns
and with Shanghai and other Chinese ports. Ocean-going vessels of considerable
draught can proceed to Canton.

Temperature and Ramfall—The monthly mean temperature exceeds 80°F.
only during the four months June to September. During the months December
to February it falls to about 60°F. The annual mean rainfall is about 65 inches,
April to August being the rainy months.

Water Supply.—tThere is a public water supply of limited extent brought from
the hills eight miles from the city, but both within and without the city walls
there are large numbers of wells. The practice of storing water in jars is very
general.

Sanitation and Quarantine.—In the Chinese city little is attempted in the way
of sanitation as understood in Europe. The Municipality of Shameen (British
Concession) employs a medical officer of health.

338 A. T. STANTON.

There is a quarantine anchorage down river from Canton, but there are no
buildings or staff to deal with infected ships. The Medical Officer of Health,
Shameen, has assigned to him by the Chinese Customs Department some duties
in connexion with quarantine matters, but these duties appear to be ill-defined.

Mosquitos.

(A). Adult mosquitos taken in houses :—Culex fatigans, C. sitiens, Armigeres
obturbans, Stegomyia albopicta (scutellaris), Anopheles sinensis.

(B). Larvae from artificial collections of water :—Culex fatigans, Stegomyia
albopicta (scutellaris), Armageres obturbans.

(C). Larvae from natural collections of water :—C. fatogans.

The weather conditions were unfavourable for the development of mosquitos
during the time of my visit. No Stegomyia fasciata, either larvae or adults, were
found. S. albopicta larvae were present in fair numbers in jars both within and
without the city, but their development was slow. Larvae of this species taken
at Canton on 13th December 1915, which were apparently then half-grown,
transformed to pupae on 6th January 1916 in Kuala Lumpur and a few adults
hatched out 8th-10th January 1916.

. Since my visit I have had sent to me collections of mosquitos taken in Canton
during the month of March. Among these were no specimens of Stegomyia.

Batavia.

Batavia, the principal city in West Java, is situated on the Bay of Batavia in
latitude 6° 7’ South and longitude 106° 48’ East. It is a long narrow city extending
along the banks of the river Tjiliwong and comprises the port (Tandjong Priok),
the old town (Batavia), and the new town (Weltevreden). The population is
about 138,000.

Trade——Batavia has extensive trade relations with all the neighbouring ports
and is the point of departure for lines of steamers to Australia, EKurope, America
(by way of the Philippine Islands) and Japan.

Temperature and Rainfall—The annual mean temperature is about 79°F. The
temperature is nearly the same throughout the year, the greatest difference
between the highest and lowest monthly mean being only about 2°. The annual
mean rainfall is 71 inches, October to April being the wet months.

Water Supply—The public water supply is drawn from artesian wells and is
distributed throughout the city by means of pipes to hydrants and private houses.
From the hydrants water is carried in tins to the houses in the native quarters
and 13 there stored in earthenware jars. There are also large numbers of open
wells, :

A scheme for a constant high pressure supply of water from the hills has been
under consideration for some years.

Sanitation and Quarantine.—Special officers devote their whole time to sanitary
work and important progress has been made during recent years in improving
the condition of the native quarters of the city.

MOSQUITOS OF FAR EASTERN PORTS. 009

There is a special staff which devotes attention to anti-mosquito measures.
Native assistants are taught to recognise the different kinds of larvae and to deal
with breeding places. By means of posters in the vernacular languages, cinema-
tograph films and lectures, the people are taught the importance of mosquito-
borne diseases and the means of preventing them.

The quarantine station is situated on two islands in the Bay of Batavia about
6 miles from the town. It is supplied with ample accommodation to meet any
emergency. There is a complete equipment of apparatus for disinfection.

Mosquitos.

(A). Adult mosquitos taken in houses :—Culex gelidus, C. tritaenorhynchus,
Giles, Stegomyia fasciata, Mansonioides annulipes, M. annuliferus, S. fasciata var.
luciensis, Theo., Mansonioides umformis, Culex fatigans, Armigeres obturbans,
Stegomyra albopicta (scutellaris), Culex whitmorer, Giles, C. vishnui, Theo., Aédomyta
catasticta, Knab, Anopheles rossi var. indefimtus, A. sinensis, A. ludlowi, Theo.,
A. fuliginosus, Banksinella lineatopennis, Ludl., Uranotaenia campestris, Leic.,
Culex bitaenorhynchus, Theo., Rachionotomyia aranoides, Theo., Lophoceratomyva
barkert, Theo., Culex fuscocephalus, Cyathomyia brevipalpis.

(B). Larvae from artificial collections of water :—Stegomyia fasciata, S. fascrata
var. luciensis, Culex fatigans, Armigeres obturbans, S. albopicta (scutellaris).

(C). Larvae from natural collections of water :—Culex fatigans, C. vishnu,
C. tritaenorhynchus, Anopheles tessellatus, A. koch, A. rossi var. indefinitus,
A. sinensis, A. ludlowi, A. barbirostris.

At the time of my visit to Batavia mosquitos were very numerous, as may be
inferred from the large number of species recorded. Stegomyia fasciata was pre-
valent in all parts of the city. Special measures are in operation to reduce the
numbers of mosquitos; the task, which is under highly competent direction, is a’
formidable one.

Samarang.

Samarang, the principal port of Mid-Java, is situated on the north coast about
midway between Batavia and Soerabaia, in latitude 7° 15’ South and longitude
110° 38’ East. The population is about 96,000.

Trade-—Samarang is the commercial centre of a very large agricultural district.
There are rail connections to all points in Java and many steamships touch at this
port on their way to Singapore, Australia, China, India, Hurope and America.

Temperature and Rainfall—tThe climate of Samarang resembles that of Batavia.
The annual mean rainfall is 86 inches.

Water Supply—The public water supply is drawn from the hills. It is a con-
stant high pressure supply and is distributed over the whole area of the city.

Samtation and Quarantine.—These are supervised by officers of the Civil Medical
Service. Samarang is a second-class harbour and the quarantine station is not
provided with its own apparatus for fumigation. A new quarantine station is
projected.

There are no special measures in force directed against mosquitos.

340 A. T. STANTON,

Mosquitos.

(A). Adults taken in houses :—Culex fatigans, Stegomyia fasciata, S. fasciata
var. luciensis, Uranotaenia cancer, Anopheles ludlowi, C. tritaeniorhynchus.

(B). Larvae from artificial collections of water :—Stegomyia fasciata, S. fasciata
var. luciensis, Culex fatigans.

(C). Larvae from natural collections of water :—Culex fatigans, C. bitaennorhynchus,
Anopheles ludlowi, A. barbirostris.

At the time of my visit mosquitos were not numerous at Samarang and Stegomyia
fasciata was rarely encountered in houses. Larvae of this species were found in
jars used for storing water in houses. The admirable public water supply is, I think,
an important factor in keeping down the numbers of mosquitos, which were less
numerous than might have been expected.

Dr. W. T. De Vogel published in 1909 (Geneesk. Tijd. v. Ned. Ind. xlix, no. 5)
an account of malaria infection experiments with Anopheline mosquitos bred
from larvae taken in salt-water pools near the sea-coast at Samarang. These
mosquitos were then identified as Anopheles (Myzomyia) rossi, Giles, and Dr. De
Vogel’s results gave rise to some controversy, as it had been thought that A. rossa
was not a malaria-carrying species. During my visit to Samarang I re-examined
these pools, but found in them only Anopheles ludlowi, Theo., and A. barbirostris,
Wulp. At Soerabaia, however, in similar situations I took large numbers of larvae
of Anopheles rossi, and I think it quite probable that Dr. De Vogel experimented
with this latter species.

Soerabaia.

Soerabaia is situated on the Kali Mas and its tributary the Kali Pegirian in
latitude 7° 14’ South and longitude 112° 44’ East. It is a long narrow city stretch-
ing along the banks of these rivers. Population 160,000.

Trade.—Soerabaia is the commercial metropolis of Java, and the greater part
of the foreign trade is executed from this port. Regular steamship services are
maintained with Makasser and other ports in Netherlands East Indies and with
Australia, India and Hurope.

Temperature and Rainfall—The climate of Soerabaia resembles that of Batavia
and other coast towns. The annual mean rainfall is 68 inches.

Water Supply.—The public water supply of Soerabaia comes from the hills. It
is a constant high pressure supply and is distributed over the whole city.

Dr. J. T. Terburgh, Chief of the Public Health Service of Kast Java, has during
several years studied the important question of supplying the native population
with pure water at small cost. This problem, of vital importance from the point
of view of water-borne diseases such as cholera, is also intimately connected with
the question of Stegomyia reduction in sea-port towns. Dr. Terburgh was good
enough to demonstrate to me the methods which he advocates and to show me
their practical working in a native kampong of Soerabaia.

Water is brought to each house in pipes and by means of a permanent water-
meter, devised by Dr. Terburgh, a quantity of water, estimated to be sufficient for
drinking and domestic purposes, is delivered at a uniform rate, The water is stored

MOSQUITOS OF FAR EASTERN PORTS. 341

in cement reservoirs of appropriate capacity, whence it can be taken as required.
Dr. Terburgh claims for his method that an adequate supply of water is delivered,
that wastage is prevented, and that after making provision for interest on capital
expenditure and for repayment, the cost to the consumer may be less than he now
pays for the transport of water from hydrants.

Owing to the more or less constant disturbance of the water surface in these
reservoirs there is less tendency for the female Stegomyra fasciata to lay eggs in
them than in the common water-jar. Nevertheless larvae of this species were not
uncommonly found in the reservoirs. It was observed also that despite the
existence of an adequate supply of water at all times in the reservoir, the natives
removed water from them and stored it in jars in accordance with a custom that
obtains everywhere throughout Eastern countries.

In his experiments Dr. Terburgh had not in mind the question of the reduction
of breeding places of Stegomyra, but I understand that he proposes to carry out
further experiments in an endeavour to meet this indication also. I was much
impressed by the possibilities revealed by Dr. Terburgh’s proposals for the solution
of the problem of mosquito control in the densely populated towns of the Hast.

Samtary Service and Quarantine.—Special officers are in charge of these services
and they are provided with subordinate stafis who are now being instructed in the
details of anti-mosquito work.

A new quarantine station has been approved and is to be built on the island
of Madura opposite Soerabaia.

Mosquitos.

(A). Adults taken in houses :—Culex fatigans, Stegomyia fasciata, S. fasciata var.
luciensis, C. sittens, C. vishnur, C. minimus, Anopheles rossi, A. rossi var. indefinitus,
Culex bitaeniorhynchus, Anopheles barbirostris.

(B). Larvae from artificial collections of water :—S. fascoata, C. fategans.

(C.) Larvae from the water collected in cut bamboos :—S. albopicta (scutellaris) ;
Rachonotomyia aranoides.

(D). Larvae from natural collections of water :—C. fatigans, C. vishnus, C. sitrens,
C. bitaemorhynchus, Anopheles rossi var. indefimtus, A. rosse.

Mosquitos were numerous in Soerabaia and S. fasciata was a common species.
The problem of their reduction is receiving special attention.

Makasser.

Makasser, the capital of Celebes Island and its dependencies, is situated on the
west coast of Southern Celebes in latitude 5° 20’ South and longitude 119° 40’ Hast.
Population 26,000.

Trade.—Makasser is a trade centre for Eastern Netherlands East Indies and a
port of call for steamships proceeding to Australia, Manila and Hong Kong.

Temperature and Rainfall—The mean temperature is about 80°F. The climate
is less equable than that of Java and the daily variation may be as much as 18°F.
The annual mean rainfall is 115 inches.

342 A. T. STANTON.

Water Supply.—The water supply is drawn from surface wells in compounds.
The question of a piped water supply has been mooted, but the expense involved is
great.

Sanitation and Quarantine —These services are supervised by the Military Medical
Authorities. There are no special measures in force directed against mosquito
breeding.

Mosquitos.

(A). Adults taken in houses :—Culex fatigans, Stegomyia fasciata, C. tritaeniorhyn-
chus, C. vishnu, C. fuscocephalus, Anopheles rossi var. indefinitus, A. sinensis,
A. barbirostris.

(B). Larvae from artificial collections of water :—Stegomyia fasciata.

(C). Larvae from natural collections of water :—C. fatigans, C. vishnui, A. rossi
var. undefimtus. .

-(D). Larvae in cut bamboos:—S. albopicta (scutellaris), Rachionotomyra
aranotdes.

Mosquitos are exceedingly prevalent at Makasser and Stegomyia fasciata very
numerous, but no special measures were in operation for their destruction. There
were large numbers of mosquito larvae in the cement reservoirs used for storing
water in bath-rooms.

III. OrHER RECORDS OF STEGOMYIA.

During the course of this inquiry I have received a number of collections of mos-
quitos from different parts of the Malay Pensinsula and neighbouring countries.
Among these the new records of the occurrence of species of Stegomyia are as
follows :—

1. Stegomyia fasciata, F.

Malay Peninsula : Kelantan (Tumpat) ; Kedah (Alor Star) ; Pahang (Kuantan).
Borneo : British North Borneo (Jesselton) ; Pulau Laut (Stagen).

Java: Tegal; Djokjakarta ; Tjilatjap ; Garoet ; Billiton I.

Sumatra : Djambi; Padang ; Indrapura ; Pulau Weh.

Celebes : Boetan I.; Soembawa I.

New Guinea: Port Moresby.

2. Stegomyia albopicta, Skuse.

Malay Peninsula: Kelantan ; Kedah ; Ginting Simpah, Selangor (2,000 {t.).
Siam : Krabin ; Koh Chang I.
Borneo : British North Borneo.
Java: Djokjakarta ; Tjilatjap ; Biliton I.; Lombok I.
Celebes: Boetan I.
Sumatra : Djambi ; Indrapura ; Padang ; Palambang.
3. Stegomyra variegata, Dol.
New Guinea : Port Moresby.

4. Stegomyia fusca, Leic.

Siam : Krabin ; Koh Chang I.; Koh Mohsi I.; Koh Kra I.; Koh Klun I.; Koh
Rang I.; Koh Kut I.
Malay Peninsula : Kuala Lumpur.

MOSQUITOS OF FAR EASTERN PORTS. 343

5. Stegomyia W-alba, Theo.

Malay Peninsula : Kuala Lumpur.

Siam : Bangkok.

Sumatra : Pulau Weh.
6. Stegomyia trilineata, Leic.

Malay Peninsula: Ulu Gombak, Selangor.
7. Stegomyia desmotes, Giles.

Malay Peninsula : Ulu Gombak, Selangor.
8. Stegomyra albolmeata, Theo.

Malay Peninsula: Kuala Lumpur.

TV. GENERAL REMARKS.

Mosquitos on Shvps.—It was pertinent to the subject of my enquiry to endeavour
to learn to what extent mosquitos were in fact conveyed on ship-board from port
to port. In the course of this enquiry, I travelled on a number of ships, some of
them of recent construction, some of them old. Careful examination was made in
port and at sea. Once only was Stegomyia fasciata observed, and on that occasion
the ship lay in a river on both banks of which this species was breeding in very
large numbers.

Occurrence of S. fasciata inland.—It is often stated that in the East S. fascrata
is found on land only near the sea coast. It is therefore of some importance to note
its occurrence at inland points in Java and the Malay Peninsula. This species
was taken at Garoet, in West Java, 40 miles from the sea and 2,300 feet above sea-
level. Within the past ten years S. fasciata has replaced S. albopicta (scutellaris)
as the common Stegomyia species in Kuala Lumpur, an inland town of the Malay
Peninsula 27 miles from the sea. -

Trade Routes —The old trade routes between America and the Orient have been
little affected by the opening of the Panama Canal to traffic. These routes still
pass northward by way of Honolulu, Japan and China. There is little direct traffic
within the tropical zone. :

V. SUMMARY AND CONCLUSIONS.

1. In Far Eastern ports of the equatorial region the yellow fever carrier, Stegomyia
jasciata, is prevalent at all seasons of the year. Even where efforts have been
made to reduce its numbers it is common; where no such efforts have been made,
it is a veritable plague.

In this region the conditions are highly favourable for the spread of yellow fever
if it should once be introduced.

2. The old trade routes between America and the Orient have not so far been
altered with the opening of the Panama Canal to traffic, these routes still pass
northward by way of Honolulu, China and Japan. It was found in the course
of this enquiry that in ports situated towards the northerly limit of the Eastern
tropical belt S. fasciata occurred in relatively small numbers in the colder months
or was not met with at all.

344 A. T. STANTON.

It is thought probable that a survey of the ports of China and apan would
show that the conditions there are unfavourable for the propagation of S. fasciata
even in the warmer months.

If this supposition should prove correct it would be shown that the path for the
conveyance of yellow fever infection to the Orient is cut at this point. It is therefore
important that a survey of these ports should be undertaken in completion of the
observations here recorded.

3. Systematic observations on a number of ships within the tropical belt failed
to disclose the presence of S. fascoata on board ships at sea, even when coming from
ports heavily infested with this mosquito. It issuggested that too much importance
has hitherto been attached to the possibility of the conveyance of infected mosquitos
over long distances in ships, and that with modern steel ships the danger is minimal.

4. With the prospective increase in sea-borne traflic it is more than ever imperative
that measures for Stegomyia reduction in Far Eastern ports should be vigorously
pursued. Examples of effective action are not wanting,and the work of James
and his colleagues in Colombo has shown what may be accomplished in this direction.

5. Trained entomologists should be engaged to give their whole time to the study
of the highly specialised problems connected with mosquito reduction and to the
supervision of measures. These officers should work in close co-operation with the
executive public health authorities.

345

COLLECTIONS RECEIVED.

The following collections were received by the Imperial Bureau of Entomology
between Ist July and 30th September, 1919, and the thanks of the Managing
Committee are tendered to the contributors for their kind assistance :—

Dr. W. Mansfield-Aders, Government Economic Biologist :— 6 Culicidae and
53 other Diptera ; from Zanzibar.

Capt. P. J. Barraud :—150 Culicidae, 4 Phlebotomus, 1 Hippoboscid, 20 other
Diptera, 19 Hymenoptera, 32 Coleoptera, 7 Planipennia, 15 Lepidoptera, 11
Rhynchota, 7 Orthoptera, 1 Mayfly, and 17 Odonata ; from Palestine.

Mr. G. E. Bodkin, Government Economic Biologist :—7 Diptera, 1 Flea, 26
Termites, 160 Ants, 8 other Hymenoptera, 39 Coleoptera 4 Lepidoptera, 22 Rhyn-
chota, 1 Grasshopper, 23 Ticks, 4 Millipedes, 1 Scorpion, and 3 Worms ; from British
Guiana.

Dr. C. K. Brain :—A number of Chalcids ; from Surrey.

Capt. P. A. Buxton :—15 Culicidae, 1 Tabanus, 1 Hippoboscid, 12 Lepidoptera,
and 10 Rhynchota ; from N. W. Persia.

Division of Entomology, Pretoria :—4 Hippoboscidae, 38 other Diptera, 12
Hymenoptera, 162 Coleoptera, 71 Rhynchota, and 20 Orthoptera; from South
Africa.

Mr. E. Melville Du Porte :—5 Coleoptera ; from Canada.

Durban Museum :—41 Coleoptera ; from Natal.

Dr. Eustace W. Ferguson :—17 Tabanidae : from New South Wales.

Dr. Lewis H. Gough, Government Entomologist :—3 Trypetid Diptera ; from Egypt.

The Government Entomologist, Madras :—113 Diptera, 12 Hymenoptera, 14
Coleoptera, 1 species of Coccidae, and 6 other Rhynchota ; from Southern India.
Mr. C. C. Gowdey, Government Entomologist :—56 Diptera, 1 tube of Braconid
parasites, 110 other Hymenoptera, 70 Coleoptera, 11 Lepidoptera, 20 Rhynchota,
5 Orthoptera, and 2 Odonata ; from Uganda.

Mr. KE. Hargreaves :—52 Culicidae, 5 Tabanidae, 1 Hippoboscid, 1 tube of Chalcids,
5 Coleoptera, and 4 Rhynchota ; from Italy.

Mr. William Harris, Government Botanist :—A species of Coccidae, and 41 Cole-
optera associated with it ; from Jamaica.

Dr. C. Gordon Hewitt, Dominion Entomologist :—20 Psyllidae ; from Nova Scotia.

Mr. C. W. Hobley :—5 Haematopota, 9 Tabanus, 1 Glossina, and 1 Tachinid ;
from British Hast Africa.

The Imperial Departmant of Agriculture, West Indies :—24 Melolonthid beetles,
11 larvae, and 11 pupae ; from Antigua, etc.

Mr. H. H. King, Government Entomologist :— 3 Culicidae, 1 Chrysops, 3 Tabanus,
and 30 other Diptera ; from the Anglo-Egyptian Sudan,

346 COLLECTIONS RECEIVED.

Mr. A Loveridge :— 30 Glossina, 449 other Diptera, 56 Dipterous larvae, 49
Dipterous pupae, 93 Fleas, 109 Thysanoptera, 1,471 Ants, 71 other Hymenoptera,
2,252 Coleoptera, 332 Coleopterous larvae, 93 Coleopterous pupae, 40 Lepidoptera,
179 Lepidopterous larvae, 12 Lepidopterous pupae, 2 Caddis-flies, 282 Termites,
9 species of Coccidae, 95 Cimicidae, 500 Aphididae, 852 other Rhynchota, 20
Odonate nymphs, 5 May-flies, 21 Mallophaga, 4 Anoplura, 29 Ticks, 400 Mites,
99 Spiders, 17 Scorpions, 3 Centipedes, 41 Millipedes, and 27 Worms; from
Portuguese and “ German ”’ Hast Africa.

Dr. R. E. McConnell :— 1 Tabanus, 30 other Diptera, and 4 Rhynchota; from
Uganda.

Dr. J. W. Scott Macfie :— 438 Culicidae, 104 Phlebotomus, 5 other Psychodidae,
17 Ceratopogoninae, 8 Cecidomyiidae, 2 Tabanus, a number of other Diptera, 7
Hymenoptera, 5 Coleoptera, 21 Lepidoptera, 20 Termites, 3 Rhynchota, 3 Orthoptera,
and samples of grain attacked by insects ; from the Gold Coast.

Capt. Malcolm E. MacGregor :—-54 Anopheles maculipennis, with a number of
eggs ; from Kent.

Mr. Arthur W. J. Pomeroy, Government Entomologist, Nigeria :—17 Tabanidae,
and 354 other Diptera; from the United States of America: 24 Hippoboscidae,
35 other Diptera, 16 Chalcids, 57 other Hymenoptera, 287 Coleoptera, 72 Lepidoptera,
75 Rhynchota, 11 Orthoptera, and 15 Odonata ; from Southern Ni geria.

Dr. J. O. Shircore :—5 Diptera and 1 Bug ; from British East Africa.

Mrs. W. Smith :—19 Culicidae, 2 Glossina, 49 other Diptera, 11 Hymenoptera,
12 Coleoptera, 16 Rhynchota, and 9 Orthoptera ; from Kumasi, Gold Coast.

Dr. F. H. Storey :—181 Culicidae, 10 Tabanidae, 163 Glossina, and 30 other
Diptera ; from Koforidua, Gold Coast.

Mr. C. F. M. Swynnerton :— 19 Glossina and their pupa cases, 3 other Diptera,
and 6 Hymenoptera ; from Portuguese Hast Africa.

Mr. F. V. Theobald :— 1 Orthopteron ; from Mesopotamia.

Mr. R. Lowe Thompson :—113 Coleoptera ; from Southern*Rhodesia.

Mr. F. W. Urich, Government Entomologist :—33 Culicidae, and 24 Mutillid
Hymenoptera; from Trinidad.

Mr. F. C. Willcocks, Entomologist to the Sultanic Agricultural Society :—24
slides of Mallophaga, Anoplura, and Acarina ; from Kgypt.

Mr. C. B. Williams :—54 Chrysops and 45 Tabanus ; from Trinidad.

IMPERIAL BUREAU. OF ‘ENTOMOLOGY,
“HON NORARY OOMMITTEE Or MANAGEMENT

THE VISCOUNT Bue a ne
_ Limvr.-Coronen A. ALCOCK, C.LE., ERS. peepee
Maton E.R AUBTEN, DS.0/% 0 yo se cc aa
Dr. A. G. BAGSHAWE, C.M.G. : : | | Re
Mayor-Genzrat Siz JOHN R. BRADFORD, K.0.M.G., ERS. val Ee
Masor-Generat Sir DAVID BRUCH, K. C. B., BRR es
Met, CROP M RR) oy ea Ree
$m §. BF. HARMER, KBE. PRS. ee | fie
| Prov. H. MAXWELL LEFROY.
Hon: E. LUCAS. |
De. R. STEWART MACDOUGALE.
Sm JOHN McFADYEHAN. i
‘Sm PATRICK MANSON, OMG. ERS. )
Sm DANIEL MORRIS, K.C.M.G. -
Pror, R. NEWSTEAD, F.RS.
Pror. G. H. F. NUTTALL, F.RS.
Prov. E. B. POULTON, F.R.S. | at ns
Lizut.-Cotone, Sir DAVID PRAIN, OMG., OLE, BRS his aie a
Sm H. J. READ, K.0.M.G., CB. Pee
fue Hon. N. C, ROTHSCHILD. ;
Dr. HUGH,SCOTT. a He
Sin A. E. PLEY, G.B.E., PRS. he
Mn R. A. C. SPERLING, CMG.
'.Sm STEWART STOCKMAN.
Me. F. V, THEOBALD. | | | ? We
Mr. C.\ WARBURTON. | | fees enna

a ee
sit 3 yet

a Director. 4
Dr, GUY A. K. MARSHALL ah

Assistant . Director.
har ar. 8): A. NEAVE.

‘Secretary. oy :
Mr. ae Gi: Cc PAREINSON.

347

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS.
By W F. Fiske.

INTRODUCTION AND GENERAL DISCUSSION.

These studies were made in Uganda on the islands and shores of Victoria Nyanza,
In territory which had been depopulated some years before on account of tsetse-fly
and sleeping sickness. They were begun in October 1913, and extended over two
full years, of which eighteen months were spent in the fly belt on Victoria Nyanza,
and three months on a tour into Bunyoro, where for the time being (the spring
of 1914) Glossina morsitans was the centre of greater attraction as a possible vector
of human trypanosomiasis. Seven tours were made from headquarters at Kntebbe—
the shortest, of three days only, being interrupted by the War, and the longest,
of eight full months, being much prolonged by the War.

The working basis unquestioningly accepted in the beginning was that the
measures adopted in 1906-07 for the suppression of sleeping sickness in Uganda
were both wise and necessary. These measures called for complete severance of
contact between the fly and the native populations so long as infection was present.
They had contemplated originally the elimination of infection through complete
severance of contact between fly and population, but in this particular had failed
of their object, for the trypanosome persisted both within and without the depopulated
zone; as a parasite of game and fly within it, and as a human parasite in the
tiparian populations of adjoming sectors of the lake region without it. They
therefore required the extermination of fly in populated districts, or as a prerequisite
to reclamation and repopulation of territory from which the inhabitants had been
removed. They were exactly such measures as are absolutely required for suppression
and prevention of the cattle trypanosomiasis transmitted by Glossina morsitans
in other parts of Uganda; and going on the perfectly natural assumption that a
close parallel existed between human and bovine trypanosomiasis, as transmitted
by Glossina palpalis and Glossina morsiians, respectively, the wisdom and necessity
of the measures was not questioned.

The economic or practical objective of the studies was the sanitation and
reclamation of the fly-infested territory ; the more immediate and technical objective,
therefore (proceeding on the above assumption), was the extermination of fly as
a sanitary prerequisite to reclamation. To exterminate a species is to restrict
its range or distribution. Therefore, the most immediate objective was made
that of identifymg and studying those factors in the bionomics or “ control’’ of
the species which operate in nature, or which might be operated by artifice, to delimit
-and determine or to “control” its range.

(659) Wt.P4/140. 1,000 5.20 B.&F.,Ltd. Gp. 11. A

348 W. F. FISKE.

It was an integral and most important feature of the original plan that, after
the factors which operate in the “ natural control” of the insect should have been
identified and studied, an experiment should be conducted on a “ practical” scale
to demonstrate at the same time the accuracy and the economic applicability of
the knowledge acquired. By the spring of 1915 sufficiently definite conclusions.
had been reached on the particular points at issue to justify this experiment.
Until then the investigations had been conducted on the islands. The experiment
could not be made on the islands, but must be made on the mainland. Consultation.
with the Principal Medical Officer as to the best site for it led to the selection of the
Buddu district, which borders the lake for some ninety miles just north of the old
German frontier. Therefore a tour was made with the object of surveying this.
reach of shore carefully, to ascertain the degree of infestation by the tsetse; the
probable cost of exterminating it; the precise location of old village sites, canoe
landings, fishing grounds, etc., with respect to fly; and such other points as required
consideration preliminary to any experiment of the character proposed.

The results of this survey were unexpected, and led to a complete revolution
of ideas previously entertained. The reach of shore is naturally divided into a
considerable number of semi-isolated districts, having from two and a half to.
seventeen miles of frontage on the lake. Some of these districts were found to.
be infested by tsetse to a degree of density never before encountered. Others were:
very lightly infested. In one district a few days labour with a good gang of men
would have sufficed to exterminate all the fly on a reach of some five miles. In
this case the natives would very willingly have given the labour in return for the
lands and the fishing grounds off shore, for the lands were good and extensive and
the fishing grounds productive. In several other districts it was thought probable:
that the value of the land and water rights was sufficient to induce the natives.
to undertake all the labour requisite to exterminate fly without other expense to.
the Government than that of supervision and imspection. Any of these would.
have been an excellent site for the proposed experiment.

There was one district, however, Bukakata, which had a frontage of some six
and a half miles on the lake and included the steamer landing for Masaka station,.
that presented a strange problem. The lands adjoining the lake in the depopulated:
or forbidden zone were valueless for agriculture. The fishing rights were valuable,.
and would have been a considerable inducement to the natives to clear the shore
of tsetse, except for the fact that the natives were already occupying them, openly,
for the entire reach of six miles. Infestation by tsetse was much heavier than in
certain other districts, but still moderate, exceeding the average for the lake shore
and islands as a whole at only a few points, and averaging for the district about
half the average for the region generally. Careful inquiry failed to elicit a particle
of evidence that any of the native fishermen had suffered in the slightest degree.
from long-continued exposure to tsetse under these conditions.

At certain other points along the shore surveyed poaching was free and flagrant ;
at one point in particular, where density of fly was more than seven times greater’
than the average for the region generally, and where, in addition to this, food.
(of fly) was very scarce and the flies literally ravenous, twenty-three fish traps,

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 349

some new and some old and discarded, were found along a two-mile reach of shore,
indicating regular and long-continued contact between trespassers and fly at its
very worst. Less aggravated conditions were frequently encountered along many
reaches of mainland shore.

On returning to Entebbe a special survey was made for fly within the inhabited
precincts and for trespassing by natives beyond them. It was found that a
considerable portion of the township was very lightly, but constantly or regularly,
infested by fly (to an easily measurable degree), and that at certain points in the
environs natives went freely into contact with fly at a moderate—but nowhere
excessive— degree of density. Similar conditions prevailed at Jinja and at Kampala,
the two other lake ports.

With these conditions in mind, inquiry was made concerning the number of
cases of sleeping sickness recorded in the official death returns or coming to the
attention of Medical Officers; for it was thought certain that some must occur.
To my very great surprise I learned that, so far as known, not a single case had
been contracted within the Province of Buganda—for which the returns are most
reliable—since 1912, and that with the exception of two cases (one of them not surely
trypanosomiasis, and the other possibly contracted in Busoga) among the men who
accompanied Dr. G. D. H. Carpenter on his tour to the islands in 1911-12, no cases
were known or suspected to have been contracted since the islands were depopulated.
in 1909.

A curious situation was thus created, which led to the abandonment of the
proposed experiment and to complete readjustment of ideas and preconceptions.
In theory—the theory upon which the suppressive and preventive measures had.
been based—complete severance of contact between fly and population was necessary
in order to control sleeping sickness, but in actual practice complete severance.
of contact was found to be unnecessary. In theory it was necessary either to exter-
minate fly from populated districts or to make removal of inhabitants ‘ from the
vicinity of tsetse complete and without exception” ; but in practice it was proved.
sufficient to reduce the density of fly to within moderate limits in populated districts:
or to reduce density of population to within moderate limits in fly-infested territory.

If this is really sufficient, knowledge concerning factors which control range
of the insect is more or less superfluous, whereas knowledge of factors which operate
to control breadth of contact between fly and population—equivalent to frequency
of contact between hungry flies and men—is specifically required.

Two very different lines of study are thus outlined. Which of them ought to be
followed ?

As a matter of fact, both were followed: the first up to 1915, and the second
during the last months of field work. One of them was impractical, and the specific
results of it are practically valueless, except as disjointed fragments of information
gained can be salvaged for use in other connexions. Which of these two lines
of investigation ought to be reported upon ?

In the accompanying manuscript neither has been reported upon, because this. |
question has uot been answered, nor any specific object for the preparation of any

(659)

300 W. F. FISKE.

report defined. It contains a mere mass of unapplied, and for the present inapplicable,
information of no practical value until a decision has been reached on the following
points :—

(1) Shall the original plan for these investigations be followed, and an experiment
be planned and conducted to test the accuracy and applicability of the
conclusions reached ?

(2) If so, shall the object be—

(a) to exterminate fly and prevent all contact between fly and population,
in accordance with the theories held in 1906-07 ; or

(b) to reduce excessive density of fly and to prevent excessive breadth of
contact between fly and population, in accordance with present practice
—as in the Bukakata District in 1915-16 ?

Unfortunately the decision cannot rest on the results of these studies in tsetse
bionomics, which have only the most indirect bearing upon the point at issue.
Every bit of information which seemed, even remotely, to bear upon them, and
which could be gleaned from any available source, was included in a report presented
in 1916, upon which no action has been taken at this date of writing. The one phase
of tsetse bionomics having any bearing on these points is the fly’s choice of hosts
and host preferences. This is enlarged upon in the following pages.

The methods used in conducting these studies consisted in the main of a “ fly
survey ” of the lake shore and islands. This survey was designed to measure
as accurately as possible all peculiarities and variations in the range of the
insect, and all variations in its density as they occurred from time to time in
the same localities, or from one locality to another at the same time. Coincidently
observations were made and notes kept concerning every factor known or suspected
to operate in the “natural control” of range and density, with the object of
identifying those of major importance, 7.e., such as are responsible for easily
measurable peculiarities in range or variations in density. No attempt was made
to study specific factors—such as natural enemies—unless they were indicated to
be of sufficient importance to account for measurable variations in density. By
this procedure much useless work was avoided.

These methods were especially applicable to studies having as their ultimate
or practical objective the extermination of the insect. But after the readjustment
of old ideas and preconceptions of the economic problem, made necessary by the
discovery that it is practically unnecessary, and even, for economic reasons, unde-
sirable to undertake extermination of fly, entirely different methods of study were
demanded. ‘The questions to be answered involved less the range and density of
tsetse and the factors controlling them than the injuriousness of the tsetse and
the factors in its control. The injuriousness of the insect, or its injurious status
in relation to a population occupying the same or adjoining territory, is, im part,
but only in relatively small part, determined by density of infestation. Factors
of equal or even greater importance are those which control the frequency of contact
between flies and persons, and these include the relative abundance of host animals
. (such as crocodiles), the principal occupations of the population, the precise location
of points of occupation or concourse of the population with respect to colony centres

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS, 351

of fly, and various others to which the original fly surveys had accorded only in-
cidental attention or had left out of consideration entirely. Reduction in density
of fly is one object ; reduction in the injuriousness or injurious status of fly is, in
actuality, a totally different object, and methods of study must vary accordingly.

Finally, it was disclosed that the question of paramount importance was one
that required yet different methods of study. It is to define the extent to which
frequency of contact between flies and persons inhabiting a given region must be
reduced in order to bring human trypanosomiasis under effective and satisfactory
control. It is a question which can only be answered by measuring the frequency
of contact and the coincidence of trypanosome infection in the population. It
is, | regret to say, impossible for me to carry on investigations along this most
necessary line unaided. It is for the entomologist to measure density of fly and
frequency of contact between flies and men, and to identify and study the factors
in control, but unless he is specifically trained to diagnose and detect trypanosome
infection, his studies are incomplete and of relatively little value. —

Every effort was made in 1915 to secure the co-operation of a Medical Officer
for a tour along the coast and islands of Nyanza Province (British East Africa),
where the riparian populations had not been removed; but, on account of the
War, nothing could be done. Some little information on this pomt was gleaned
from old (manuscript) reports of Medical Officers employed on “ Sleeping Sickness
Extended Investigations ” m Uganda, which, as far as it went, strongly confirmed
the presumption that the disease is incapable of spreading unless there is excessive
breadth of contact between fly and population—in excess, for example, of that
encountered in the survey of the Buddu shore in the Bukakata District—and a few
other scraps of information have been found in literature—for example, in Todd
and Wolbach’s survey for human trypanosomiasis in the Gambia, which contaims
a few brief references to prevailing density of fly and frequency of contact—but
it is a lme of investigation which has yet to be systematically followed.

I. DispaRiry BETWEEN THE SEXES OF GLOSSINA PALPALIS.

When any considerable number of flies of this species of Glossina are caught, it
is unusual to find the sexes evenly represented. Although they are produced in
equal or approximately equal numbers, they are caught in unequal numbers.*

This disparity between the sexes of Glossina palpalis is an extremely variable
quantity, ranging in different localities (in catches of 100 flies or more) from 1°9 per
cent. to 85:0 per cent. of females or from 15-0 per cent. to 98:1 per cent. of males.
Females never predominate to quite the extent of males, and most frequently the
males are in excess.

Various hypothetical explanations for this phenomenon have been proposed by
Medical Officers and Entomologists who have observed it, and prior to the inception
of these investigations all these explanations presupposed that “ caught flies are
hungry flies ”’ (attracted to their captor through desire to feed on his person), therefore

*<< Thus of 1,400 flies bred from pupae obtained on Damba Island the proportions were
6:9:: 48:52. Whereas of 5,000 flies caughs during the period in which the pupae were

co lected the proportions were ¢: 9:: 78°6: 21:4.” Carpenter: Repts. of the 8.8.Comm.
of the Royal Soc., xii, p. 105.

W. F. FISKE.

352

‘SpuvIST Ossog oT} JO dnory equIvg-oMoy oy} jo dvur-qoyoyg “[ “O1T

The

enjwiny

AO =
nuny ak
ebuemg =
Vv
oO
| ebues
OF
ebvema as
oD
ambuesemq ose =) (ah

nae.) Pa

uss SP/el PROS

WY S27/S dwey

oesny

0 IMHO]

: S v € @ | O
NWoOlq

that the ratio between the sexes in the catch would be an index to the ratio actually

existing between all flies in the locality or district where the catch was made.

deduction is logical from every point of view

“SAIL gO 2]€9S

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 300

' But what becomes of the surplus males in localities where females predominate,
or of surplus females, when, as usually is the case, males predominate, were questions
that had not been answered when these investigations were begun. Whatever
the explanation might be, it was certain to be intimately involved with phases in
the bionomics of tsetse the nature of which could only be guessed, but the importance
of which could not be doubted.

A minor problem was thus defined which served as a starting point for further
investigations into the broad and complicated subject of the bionomics of Glossina
palpalis, and the first field-work done was the series of experiments an account of
which follows,

I(a). Experiment to determine if Variations in Sex Ratio of Glossina palpalis are
due to differences in the purely local, or in the climatic or seasonal Conditions
of Life.

The Islands of Bulago and Tavu (fig. 1) are separated by only a few thousand
yards. Tavu is much the smaller and differs also in topography and vegetation.
But the two islands are so near together that any difference in sex ratio, if constant
and if occurring at the same seasons of the year, would necessarily be due to purely
local conditions.

Large collections of fly were made on the two islands as shown in Table I. There
was considerable variation in sex ratio observed on Bulago on different days
(when collections were made at different points on the island), but in the total
catch of 4,405 flies there were proportionately twice as many females as in the
total catch of 3,126 flies from Tavu. This difference could only be attributed to
purely local differences in conditions of life between these two islands.

TABLE J.

Catch of Fly from Bulago and Tavu Islands, showing relatively constant difference
in Sex Ratio.

Daily Catch. Total Catch.
1913. Island.

Total. Females. Total. Females.

Ist Nov. ..| Bulago hss 170 20% 170 35 %
Onde) >.". e e 18 V7 £9, 188 33%
Sed is Le af te 60 24 % 248 32 %
AU iige Oe o s BS oe 630 25: % 878 28. %
Sth >, fe yi 564 27 OF 1442 Dy eR
S15 as oe ae As LP 656 1S) 95 2098 25 G
10th?) ~,, --.. * A; 651 207 EG 2749 24 %
MG 7 2. 35 Ay 1057 2h ie 3803 24°5 %
on Ki 167 ey 3970 24 9
Pst . y 435 29 % 4405 25 %
Bon (Wa ys] Taye 43 136 AS of 136 Airy
bhi yy ge oe. ¥ Lf 1179 8-3 % 1315 7°83 %,
TA LOORY Sa a aa e 945 | 13 2260 yA
Bile iiieg bc sli 2 $6601), ty auld 3126 12-4 %

354

W. F. FISKE.

Much wider variations in sex ratio between different islands were subsequently
noted, and the extremes are given in Table IJ. Im all such cases—or at least in
most of them—explanations for variations in sex ratio must be found in the purely
local conditions of life, and not in climatic or seasonal differences.

The absolute extremes encountered in the course of the investigations, together
with the records for all colonies of fly or infected districts in which female percentage
exceeded 45°0 are presented in Tables III and IV.

TABLE II.

Extreme Range nm Variation of Sex Ratio (observed) between different Islands

an Victoria Nyanza.

Island.

Date.

.| February 1915

TABLE III.

Catch of Fly.

Total. Females.
7065 63°O) a

557 44°3 %

1413 39° 9%

fle buat 39°0 %
i yee 79%
704 fie as

73 4°1 %

101 3-0 %

Absolute Extremes in Sex Ratio as encountered in course of these Investigations
on Islands and Mainland.

Largest percentage of females in catch

‘6 | Largest percentage of females in catch

Catch in those Districts on Mainland where Female Percentage averaged 45:0
or over for entire District.

District.

Date.

Mujuzi Creek Colony, Buddu
Bukakata South Colony, Buddu
Kaziru District, Buddu bes
Kitebo District Mawakotu ..
Gwamba District, South Buddu

..| June

- | July
..| duly

..| July

MONS i
1915..
1915..
..| August 1915..
HONS

Catch.

Percentage of

Females.
1611 67°6
298 51:0
991 50°6
1199 47°5
3422 45-0

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 355.

I(b). Relative Inactivity of Females of Glossina palpalis as the Explanation of
Low Female Percentage in Catch.

The figures of the catches of fly from Bulago and Tavu Islands presented in
Table I show a discrepancy in that the percentage of females in the total catch
from Bulago Island tended to fall as the total grew, whereas on Tavu it rose sharply.
This might be accidental, but the circumstances under which the experiment
was conducted indicated otherwise. Tavu Island is much smaller than Bulago,
and the catching of some 1,300 flies on the 5th and 6th of November seemed
to have reduced the local density of the species appreciably, as measured by the
average number of flies which could be caught per boy per hour. A corps of 9
or 10 boys was employed. During these first days of the investigations the exact
time spent in catching was not recorded. It was approximately known, however,
and when the catch of males and females per boy per hour (approximately) is
tabulated it stands as in Table V.

TABLE V.
Reduction in Density of Glossma palpalis on Tavu Island due to catching
of Fly.
Catch of Fly. «“ Boy Density of Fly Indicated.
1913.
| ye) | »
Males. Females. | Hours. | Males. | Females. | Total
|
6th November .. 1081 98 36 30°0 OI 3207
7th a mn 734. 169 45 16°3 3°7 20°0
8th he a. 698 111 45 15°1 2°5 17°6
TABLE VI.
Catch of Glossina palpalis on Lula Islet, showing the relative Inactivity of
Females.
; Catch for Period. Catch to end of Period.
Period
(November 1913).
Fe- Female
Males. maloe) oa Total. | Females.
18th Afternoon .. a are ..| 280 18 60% 298 60%
19th Mor: ing .. me .-| 308 ee 19:0 % 678 1333 %
19th Afternoon .. Bb Bs yo ess 68 42:2 % 839 18°8 %
20th Morning .. oe ve see So. 4G 64°3 % 1068 28°5 %
20th Afternoon .. a we Clu 89 G2, oF 1193 3279 9
21st All Day... at ee aly oOo 224 10-2, oF 1512 40°38 %
22nd... sa on aig peilia aged 79 60°38 % 1642 42°4 %
PATH... ss - Da 2O% ee Teme mgles ©,” | 74d 43°5 9%,
25th 55 ste aie au au Ore 93 TORE 1865 46°2 %
26th sk a8 ath aes eee 65 69°1% 1959 473 %
27th Morning \ BA <u tog naan 75°17 % | 2029 48°3 %

*These figures, apparently, represent about the daily emergence of males from pu ae;
all the old flies of this sex appear to have been caught off 4 re

356 F. W. FISKE.

These data indicate that whereas the density of male flies had been reduced from
30°0 caught per boy per hour on the 6th to 15:1 on the 8th, that of female flies had
not been appreciably reduced—exactly as though the females were actually present
on the island but for some reason not so easily caught as either (a) the males on the
same island, or (b) the females on Bulago Island (Table 1).

This hypothetical explanation for a low percentage of females in the catch suggested
the experiment of catching all the flies from an islet. Tavu was rather larger than
convenient and the yet smaller islet of Lula (Fig. I) with an area of hardly more
than 10,000 square yards was selected.

The results of this experiment are presented in Table VI. It was not carried
to a final conclusion because of the time which would be required to catch all the
flies emerging from pupae on the island (at least three weeks longer than the period
devoted to it), but it was carried far enough to demonstrate conclusively that a very
low percentage of females may be due to the relative inactivity, and not, as had
always been assumed, to the absence of the sex.

I (c). The Percentage of Females among caught Flies as an Index to Food Supply.

An analysis of the data presented (in part) in Table VI and secured during the
course of the Lula experiment above described disclosed the curious fact that although
the catch of males fell sharply from 14°0 per boy per hour on the the first day to
2°6 and 1°7 on the morning and afternoon of the third day (see Table VII); that of
females rose strikingly from ‘9 per boy per hour on the first to 45 and 4°4 on the
morning and afternoon of the third day. Otherwise stated, the density of active
male flies was quickly and permanently reduced to one-eighth or less of the original,
but that of active female flies actually increased by no less than five times during
the same period that that of the males was being reduced by seven-eighths.

TaBLE VII.
Analysis of Catch of Fly made on Lula Islet showing Variation in Activity of
Females.
Density* of Active Flies.
Period
(November 1913).

Males. Females. Total.
18th Afternoon . 14:0 “9 14°9
19th Morning .. 9°7 2°2 | 11°9
19th Afternoon . 3°5 2°8 | 6°3
20th Morning .. 2°6 4:5 if
20th Afternoon . 17 4:4 61
21st All Day .. 1°8 41 5°9
Donan ee, 1-0 1-6 2-6
24th G a od 2°4
25th ne “) 2°9 3°8
26th ies Me rai | 1°4 2°1
27th Morning .. 6 1°6 2°2

* By “density ’’ is meant the number ef flies which can be caught per boy per hour.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 357

This strange increase in the activity of the female flies on this islet was obviously
associated with the experiment being conducted there, and that which seemed the
most plausible explanation was that the male flies are normally active and easily
caught at all times during good weather, whether they are hungry or not, but that
the females are normally inactive and not to be caught except when hungry and
seeking food. This hypothesis was borne out by the observation (made during the
course of the experiment) that the mere presence of the corps of fly boys (9 or 10 of
them) on the islet had temporarily banished from it the several crocodiles and Varanus
which had formerly frequented it and which were the only visible sources of food
for the flies. Assuming that only hungry females are active and easily caught,
the effect of this would be to increase the number of them and therefore the number
caught per boy per hour, through decreasing their food supply.

On this hypothesis such variations in the percentage of females amongst caught
flies as are presented in Tables I and II (7.e., such as occur between different islands)
would indicate corresponding variations in abundance of food, or density of host
animals. It was resolved to test this hypothesis by banishing the host animals
from a small islet without catching off any of the flies, and observing the effect on
sex ratio.

This experiment was made on the small island of Lugazi, in the following manner
and with the following results.

19th and 20th December 1913. A total of 197 flies were caught on the islet, the
sex was determined and they were then liberated (in order that catching off of males
should not affect the sex ratio). The ratio was, males: females: : 166: 31=15°5
per cent. females.

22nd to 27th December 1913. All host animals known to be fed upon by Glossina,
consisting of several Varanus and crocodiles, were systematically hunted from the
islet,

26th and 27th December 1913. A total of 208 flies were caught, showing sex
ratio, males : females : ; 89: 119 =57-2 per cent. females.

This experiment was carefully conducted, and careful notes were kept upon the
behaviour of the flies on the islet towards man and also towards certain domestic
animals which were tethered there (see Sect. II (b)). They were so strikingly affected
by the banishment of their reptilian hosts as to leave no doubt that they had been
principally dependent on them, and that the increase in percentage of females from
15°5 to 57:2 was the direct result of food shortage. Data on this phase of the
experiment appear in Sect. II following.

I (d). The Real Ratio between the Sexes of Glossina palpalis.

The foregoing observations and experiments demonstrate conclusively enough
that the ratio between the sexes of caught flies is no criterion of the real sex ratio
in the locality where the catch is made. It would be impossible to determine the
real ratio accurately, otherwise than by the continuation of such an experiment
as was made on Lula until all flies had been caught, and this would require more
time than the knowledge is likely to be worth.

358 W. F. FISKE.

It is believed, however, that the ratio in the catch on Lugazi after the hosts had
been banished from the island (57°2 per cent. of females) is not far from representing
the real ratio between the sexes on that particular islet. The more active males.
pretty certainly run greater risks, and do not on the average live so long as the less.
active females, and the real sex ratio would be determined by respective longevity
of the sexes—probably a variable quantity.

An explanation is thus provided for any ratio between the sexes of caught flies
from, say, 60 per cent. to 70 per cent. of females down, but doubtfully for female-
percentages of 70, 80 or even 90, such as are occasionally encountered. The explana-
tion for these excessively high percentages was a mystery until long after the
conclusion of the Lugazi experiment, when investigations into the movements of
flies along the shore of a lake or stream provided the explanation (see Sect. III}
and at the same time additional confirmation of the conclusions, tentatively reached
through the Lugazi experiment, that the percentage of females in caught flies may
serve as an index to the abundance of food (or density of host-animals) in the
region where the catch is made. This conclusion is undoubtedly correct, provided
that the catch is truly representative of the district or region; it is not correct if
the catch is made under specific conditions (with respect to shelter, etc.) described
in Sect. III.

IJ. THe VARIABLE BEHAVIOUR OF GLOSSINA PALPALIS TOWARDS MAN AND-
Domestic ANIMALS AS CORRELATED WITH VARIATIONS IN SEx Ratio.

The conclusions tentatively reached through the foregoing experiments are of
interest in a vital connection. They suggest that if the female percentage is low,
food must be plentiful, and man would be less hable to attack. Moreover, when
a great majority of the flies are feeding regularly on reptilian hosts—which do not
carry the virus of the human disease—many flies, or a dense infestation by fly,.
would be less injurious to man and less hable to transmit disease from man to
man than if there were only a few flies, or a light infestation, feeding principally
or exclusively on man, or on animals which may carry the virus of human disease.

Therefore, if the conclusions are correct, the injuriousness of Glossina palpalis to a
population living in constant contact with it would be subject to variations indepen-
dently of any variations in density of the fly, or of any other factor than the
abundance of host animals incapable of harbouring the virus of human disease, and
if it were attempted to reduce the density of fly, with the object of minimising its -
injuriousness, through measures directed against its host animals, the results might .
be the opposite of those desired. Through a campaign of extermination directed
against its reptilian host, its density might be reduced, but its injuriousness increased.

Several experiments were conducted and many observations were made which
have a bearing on the point, and which were also designed to confirm the conclusions
to the Lula and Lugazi experiments cited above ; for if these conclusions are correct, .
there must be a conspicuous correlation between variations in female percentage
and in the persistency with which tsetse-flies press their attacks upon man.

In part, these experiments and observations are cited in this immediate connection -
(as confirmatory of the preceding conclusion); but they are further cited in subse-
quent pages in other connections.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 359

II (a). Methods used for Estimating the Density and Economic Status of Glossina
palpalis.

The most convenient method for measuring variations in the density of Glossina
-palpalis is that of employing expert “fly boys” and of counting the number of
flies which can be caught per boy per hour under standardised conditions. This
method had already been used by Dr. G. D. H. Carpenter and others, and when
care was taken to eliminate sources of error very reliable figures were secured.*

On account of the very variable activity of the females it 1s obvious that estimates
-of local density must be based on the catch of males. That of females is of little
significance by itself in indicating density, but taken in connection with the catch
of males it possesses a large significance, and is indicative of variations in the
economic status of the fly. This, hike density, varies extremely (a) from time to
time in the same region or locality, and (6) from one district or locality to another
at the same time. It is in part determined by the density of fly, and variations
in density correspond to variations m economic status (2.e., by reducing density
we assume that we depreciate the injurious status of the insect, and its status must
naturally vary with natural fluctuations in density); but in larger part it is
determined by the behaviour of the fly towards man, and this in turn by abundance
-of wild hosts. The female percentage is an index to abundance of hosts, and therefore
to the liability of man to. attack.

To illustrate the above, two extremes cited in Table III may be used: the
infestations as measured in Bugaba and Bale beaches respectively. In each case
the actual catches of fly were made at three pomts a few hundred yards apart on
‘these beaches, and the complete data concerning degree of infestation are presented
‘in Table VIII.

At first glance, in comparing the density of fly at these two pots on the basis
of the total catch—both sexes—per boy per hour, it would appear that there were
twice as many flies along the Bale as along the Bugaba beach, for the total catch
is more than twice as great. But the difference is very largely due to the enormously
greater activity of the females at Bale, and if allowance is made for as many,
proportionately to the males, at Bugaba as at Bale, it appears that there are almost
three times as many fly at Bugaba, instead of less than half as many.

* Proper attention to the following suggestions will serve to eliminate various possibilities
-of error :-—

(1) Boys should be trained at least one month; all new boys in a corps without one or
-or two experts as teachers would require longer training.
(2) Nets must have a standard-sized ring and a standard length of handle; an 8-inch
-ring and 18-inch handle were used.
(3) Catches should not be made before 8.30 or 9 a.m. nor later than 2.30 or 3 p.m.
(4) Boys will make more even catches if provided each with a dark-coloured umbrella,
‘upon which the flies will settle, and from which they are easily caught.
(5) Unless density is very low—less than 1:0—it is better not to spend more than
2 to 3 “boy hours”’ at the same point on the same day; density is easily reduced,
‘temporarily, by catching.
_(6) Boys should be stationed along the routes most likely to be followed by moving
flies, and always, if there is shadow, at the edge of it.
(7) Estimates of density must be based on catch of males, on account of the variability
“of female activity.
(8) Good results cannot be secured on cool, cloudy or windy days.

360 W. F. FISKE.

TABLE VIII.

Extremes of Variations in Economic Status of Glossina palpalis as indicated by
Female Percentage.

Catch of Fly. Indicated Density.
Loeality and ke No. of | Female
Point. hae |. Boy. | | Ratio.
Males.) ales.| Lotal. route Males. | Females! Total. |
Bugaba—Pt. 15..) 33 1 34 3 11-0 3 11°3 30 %
TOR AS 1 40 14 26°0 6 26°6 2°5 OF
Ak by ae ate 4 0 32 14 23 ‘0 21:3 Om
Bale—Pt. 49 . ath a0 84 104 2 10°0 42-0 52°0 808%
», 90 ed me 0) 80 90 2 5:0 40-0 45-0 88°8 %
fea) be i 40 47 2 3°5 20°0 23'5 85°11 %
Totals—Bugaba..| 104 2 106 6 17°3 3 WRT Loe
Bale) ssl a0 | 2045) 244 6 6°2 34-0 40°2 84°6 %
|

It is also necessary to take account of the reason for the excessive activity of
female fliesat Bale. If this is caused by scarcity of food, it appears that all the females,
and obviously the males as well, must be hunting for food. Therefore a man fishing
on the beach would be likely to encounter no less than 40 hungry flies per hour.

But at Bugaba there would be only one hungry female each three hours, and if the
males are proportionately hungry, not more than two flies that were actively seeking
food would be encountered each three hours by the fisherman. This would indicate
that the fisherman on Bale beach would be exposed to 60 food-hunting flies to 1 for-
the fisherman at Bugaba; 2z.e., the fly at Bale, though less numerous by half, is
indicated to be 60-fold more ees ae to a population living in contact gue it.
than the fly at Bugaba.

This point seemed so important that confirmation of conclusions tentatively
deduced by the above method was sought in various ways, and the final conclusions.
are that the fly really varies in economic status over a much wider range than as.
above indicated. There are several reasons for making this the final conclusion
which need not be gone into; the poit upon which evidence is submitted in the
pages immediately following is that of the much greater freedom with which fly
attacks man when the female percentage (for a district or region) is high than whew
it is low.

II (6). Correlation between High Percentage of Females, artificially produced,
and Persistence of Attack upon Man.

In the course of the Lugazi experiment to which reference has already been made
(p. 357), two adult goats and one—subsequently another—small pig (of European
stock) were tethered, equally exposed to attack by Glossina, along the shores of the:
island. From the 15th to the 20th of December great care was taken not to disturb:
the wild hosts of the fly on this islet. The female percentage taken on the 19th and.
20th was found to be 15:5, which was probably maintained throughout this period..

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 361

From the 22nd December the wild hosts (crocodile and Varanus) were systema-
tically hunted from the island. The effect on the behaviour of fly toward the
tethered host animals was not noticeable on the first day, but appeared on the second
in the increased number engorging upon them. The third day was overcast and
flies were inactive, but on the fourth day, as may be seen from Table IX, a very
notable increase in the number of flies engorging was brought about.

The percentage of females was determined anew on the 26th and 27th, and was
found to have increased from 15°5 to 57:°2._ This was certainly due to decrease in
number of wild hosts ; the increase in number of flies engorging upon domestic animals
was certainly due to the same cause, and is thus directly correlated with variations.
im sex ratio.

TaBLe IX,

Showing Effect of hunting Wild Hosts from Lugaz Islet upon Behaviour of
Glossina palpalis toward Goat and Pig.

Pig. Goat.
ny Date | .

Conditions on Islet. (1913). Hours} Flies | Bites | Hours; Flies | Bites
ex- lengorg-| per ex- lengorg-| per
posed.| ing Hour. | posed.| ing. | Hour.

15th December}; — — ~-- 9f | — —

16th oe) S25 4) ae 2 meeS See

Wild hosts undisturbed. | |17th oe 4 os — 8 = —

Female ratio 15°5 %.||18th - 32 a —— 74 — —

19th BS 8 1 12 16 — —

20th es 8 — — 16) — —

Wild hosts hunted. ae 2 ic br cg 28 ~ 3 “17

Female ratio increasing) |o4¢h Toye tat hehe: eae nes 19

to 57°2 %, llpsthe 18 oh gra Gis hy 272 18 Guess

The behaviour of fly toward man underwent a change quite as striking as in its
behaviour toward pig. At first it was attempted to keep an exact record of the
number of flies attacking the observer and the fly boys. From the 15th to the 20th
inclusive only one fly bit and none engorged upon the observer. The fly boys
were equally immune. During this period no one paid any attention to the fly nor
was any effort required to ward it off.

On the 22nd December no difference was noted and no flies bit the observer.

On the 23rd ‘“‘ Flies were more active than I have seen them.” Four flies bit the
observer.

The 24th was a dull day. “Flies very inactive but bothersome. Four flies
bit; others would have bitten if permitted to do so.” The fly boys were equally
annoyed, and considerable effort was required to keep flies at a distance.

On the 25th it was necessary to be constantly on guard against the flies. “ The
flies are hungry—very hungry. They attack man viciously, and it is impossible
to estimate the number of times I have been bitten or would have been bitten but

362 W. F. FISKE.

for constant vigilance. All the fly boys were on the qua vive, and continually fighting
off flies, in sharp contrast to complete indifference toward them on the 22nd and
previously.”

Undoubtedly more flies would have bitten man on the 25th than actually engorged
upon the pigs that day, and their changed behaviour toward man was equally as
notable as shown by the table (IX) toward pig.

This experiment of hunting the wild hosts of tsetse was accidentally repeated
on the peninsula of Neozi on the island of Bugalla (Sesse) where camp was pitched
in November 1914. There was a not heavy or noticeable infestation by tsetse,
and the principle host of it was situtunga. But no sooner was the camp occupied
than these animals evacuated the peninsula and were seen crossing the isthmus
connecting it with the mainland one to two miles distant only a few hours after the
men began work on the tent and huts. On the following day (Sunday) the behaviour
of the flies was not notably changed, but on the third day they became so unbearably
persistent in their attack that (in view of the possibility of human infection from
their bites) the camp was abandoned.

Such conditions are likely to be produced in almost any fly-infested locality by the
advent of a temporary population sufficiently large to produce a measurable effect
upon host animals. They are particularly likely to be created when, as was formerly
a custom of the natives, temporary fishing camps are located on small islets which
are otherwise uninhabited. If a person infected by the human parasite is a member
of such a fishing party conditions will be extremely favourable to the transmission
of the organism to other members of the expedition.

II (c). Correlation between High Percentage of Females naturally prevailing and
Persistence of Attack upon Man.

The islands listed in Table X were visited in January and February 1914 in company
with Dr. Carpenter. Large collections of fly were made and the average of male
density and of female percentage recorded for the several observation points on
each island is also given in the table. Dr. Carpenter and I were agreed, at the close
of the tour, that the relative persistency of attack by fly upon man was fairly estimated
as in the table.

The difference between behaviour of fly towards man on Kimmi and Damba,
respectively, was truly extraordimary—equally as extraordinary as its changed
behaviour on Lugazi islet following the banishment of its wild hosts. And these
differences are strikingly correlated with variations in sex ratio, and not at all, or
only as by accident, with variations in density of infestation (as measured by density
of active males).

These observations have been many times confirmed, notably by the behaviour
of flies on Bale fly beach (Table VIII), where the female percentage was 84°6, and
where, despite constant vigilance, the observer was actually bitten about 15 times
per hour. This is more times in one hour than he was bitten in more than six weeks
spent in field work on the islands of Bulago, Kome, Damba, Tavu, Lula and others,
from 1st November to 15th December 1913. A record was kept of the number of
bites inflicted during this period and it totalled only 9. When the fly is no more

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 363

troublesome than this, the ordinary person makes little effort to ward off attack,
and nearly every fly that cares to bite does so, but when the fly is as troublesome as at
Bale, where “ nearly every fly tried to bite ” the ordinary person is continually on
the alert to ward them off, and not nearly so many flies bite as would otherwise do
so. The difference in behaviour of fly towards man on such islands as Tavu,
Bulago, etc., and on the fly beach at Bale is even greater than the figures given
above would indicate.

TABLE X,

Correlation between Female Percentage and Behaviour of Glossina palpalis
towards Man.

Ai i | |
|

Yempaita .} 1382 42°6 37°9 % |About as bad as Wema would average.

Bulago .. 580 2130 29:0 94 |Not as bad as on preceding island, but bad
enough.

Nsadzi ..| 1080 18°4 24:6 % |Not as bad as Buiago, but worse than on
Tavu at times.

Kizima .. 513 Lori 18-9 % |Not at all troublesome.

Tavu ie 674 46°6 17°71 % |Not as bad as on Bulago.

Damba .. 975 30°7 12:2 % |Less troublesome than on any other badly

infested island.

* Owing to the varying degree of activity of females the density of fiy in any district
or island must be based, for comparative purposes, on the catch of males alone.

II (d). Experiment to test Variability in Behaviour of Fly towards Man on Kome
or Damba Islands, as correlated with Variations in Abundance of Host Animals
and Female Percentage.

In the autumn of 1915 a definite and detailed scheme was worked out for the
reclamation of and re-establishment of the populations upon the SesseIslands. One
detail of this scheme was the placing of a colony (on the island of Kome (Fig. I),
which in itself is one of the safest in the lake and could easily be made much safer,
- but which suffered severely from sleeping sickness owing to the nearness of the
island of Damba, which, when populated, was probably the most insanitary island in the
Sesse group ; it must have been continuously much as Lugazi islet was after being
occupied for three days by hunters of the wild nosts of tsetse. It would be entirely
impractical to make Damba a safe place for a native population to live unless there
were much more need for agricultural land than at present. But wild hosts were
so numerous and female ratio so low that theoretically Damba, if unpopulated,
and set aside as a bit of game reserve, would not be a source of danger to a population
on Kome.

It was desired to test this theory by pitching camp on Kome, and spending some
days hunting over and surveying both it and Damba, keeping accurate count of the
number of times that flies actually bit—without engorging—during the days spent in

(659) B

964 W. F. FISKE.

the experiment. This wasdone; camp was pitched squarely in the fly belt on Kome ;
three days were spent in exploration and survey of bush and old plantations along
its shore; four hunting trips were made to Damba, and afterwards two days were
spent in a camp located squarely in the fly belt on that island.

The results of the experiment are summarised in Table XI. Host animals were
more than five times as numerous on Damba as on Kome, with the result that female
percentage was very much lower. Density was not less than six times greater on
Damba, but owing to the relative scarcity of wild hosts, fly was much more prone
to bite on Kome, when one-sixth the number of flies mflicted nie times as many
bites. This would indicate the fly to be nine times more likely to transmit disease _
on Kome than on Damba, although the density is only one-sixth as great.

TABLE XI.

Comparison of the Behaviour of Glossina palpalis towards Man on Kome and
Damba Islands.

Basis of Comparison. Kome. | Damba.
Number of hours spent in bush along shore at times when fly was |
active .. 14 [a2
Number of hours spent. in camp in fy belt, at times when fly was
active .. a oe 56 Beales! 8) 8
Average density ‘of fly i in bush, about. ae Mes a Ding eae 0) 30°0*
Average density of fly at camp, about th oe a ae 5:0 15:0
Female percentage in bush _ ... a A a an a eed 10-2
Wild hosts of fly seen or flushed :—
Varanus .. 0 12
Situtunga . 7 50
PORE OL AOA P IS NS Eee eee es
Wild hosts seen or flushed perhour .. ae ef fs aol bene 2°8
Number of times observer was bitten in bush. ds ae a 3 0
oS oa Ss camp ae Ik x 4 1
(VA areas me AR’ Mennaaedi/ acre
Number of times bitten per hourin bush ... me a ein O2ke4 0-0
>» » >» camp - - vs eh es Oe Cem Or 12

Number of times bitten per hour spent on island .. af a 0-28 | 0-03

*The catches made during this trip indicate a somewhat heavier infestation than this,
but it is known that they were too high; they were made at points where flies were
concentrated, and allowance is made for lower average density.

The really striking comparison, however, is between the figures as given for Damba,
and the note made during field work on the beach at Bale (see p. 362) to the effect
that “ nearly every fly tried to bite,” and that 15 actually bit in less than one hour.

Density, as indicated by the male catch at Damba, is three or six times greater
than at Bale, but at Damba 22 hours were spent in the bush without being bitten
once, as compared with 15 bites in one hour at Bale, a proportion of less than 1 to
more than 15 x 22 = 330.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 365

Making allowances for error at every point, it is clear that density of infestation
being equal, the fly is several hundred times more likely to feed upon man where
‘wild hosts are very few and female percentage very high than when they are very
many and female percentage low.

II (e). Female Percentage as an Index to the Chances favouring Transmission of
Human Trypanosomiasis.

Perhaps the most pertinent point in this connection is that the chances favouring
transmission of the virus of sleeping sickness from man to man are vastly less
proportionately when few flies feed on man than when many do so. The same fly
‘must feed on or bite the human host twice in order to transmit disease from an
infected to a healthy man. If only one fly in 500 or 1,000 actually bites man, the
chances that that same fly will attack man a second time are absurdly small; if
every second or third fly feeds upon or bites man the chances that the same fly will
attack man a second time aré stupendous in comparison.

The female percentage may thus be a very valuable index to the chances favouring
transmission of human disease.

Ill. Tot Lonc-sHoRE MOVEMENTS OF GLOSSINA PALPALIS.

The explanation of the variable ratio between the sexes of Glossina palpalis provided
by the early experiments proved inapplicable in many cases, and numerous vagaries
‘in the sex ratio were observed which were for long inexplicable.

Eventually investigations into the long-shore movements of flies and the routes
and courses followed by them im their food-hunting and otherwise stimulated flights,
led back into the old problem of sex disparity and provided logical explanation
for many variations in it which had been observed but which could not be correlated
with abundance or scarcity of food. The same study also required consideration
of “shelter” (arborescent vegetation sought or required by the flies for their protection)
and of the relative attractiveness of different types of it.

These three, quite different topics—sex disparity, shelter, and movements of flies
from place to place—had to be considered coincidently in the field and cannot
be entirely separated in reporting upon field work.

In the following sub-sections they are discussed as they were studied, inter-
-dependently.

IIT (a). Experiment to test the Movements of Flies along the Lake Shore.

The larger catches of fly made on Bulago Island and recorded in Table I were
‘from near the extremity of a long, narrow spit of sandy land designated in the
notes as “Crocodile Point” (see Fig. I). The area of this spit was considerably
less than of the islet of Lula, from which the flies were caught as recorded in Tables
VI and VII. But though these catching experiments on Lula quickly brought
about reduction in density of active flies, 1t was observed that no such effect was
produced on the density of the fly at Crocodile Pomt. On the contrary, although
more than 2,000 flies were caught there during the period 10th to 13th November,
neither density nor female percentage underwent notable change (see Table XII).

(659) B2

366

W. F. FISKE,

TABLE XII,

Catch of Fly at Crocodile Point, Bulago Island, showing Existence of Rapid
Movement along Shore.

| Catch of Fly.

Density of Active Fiy.*

Date. re
Total. | Females. Males. Females. Total.
5th November 1913.. 564 72 Tha 14°2 5°8 20°0
10th 4 patie: 630 20 2G 16-0 4-0 20-0
lith ” aca 944 28. 6 12°5 5°0 17°5
12th 53 cree 167 1S: oe 10°3 2°4 12°57
13th > pone 435 29° '% 14°7 6:0 20°7
Totals and averages Ist experiment) 2740 24°4 % 13°5 4-6 18°1
17th January 1914 912 40-3 % 24-4 17-1 42°5
20th 09 5. fee se 1102 48°2 % 27 26°4 53°9
21st ” anh pete a oe 303 44°2 % LB 9°0 20°3T
22nd’ ;, 55%, eee 2a: se 606 48:0 % 26°5 28°9 55°4
Totals and averages 2nd experiment 2923 45°2 % 22°7 20°3 42°9

* “ Density ” is the number of flies caught per fly-boy per hour.
+ These days were somewhat dull and overcast, rendering flies inactive.

The fact that density of both sexes was perfectly maintained on the 13th November
after four consecutive days of catching, when viewed in the light of the results of less
extensive catchings on Tavu (‘Table V) and Lula (Tables VI and VII), can only be
explained by the movements of fly along shore, and these movements must be quite
free and rapid in order to account for it.

Notwithstanding the semi-isolated position of Crocodile Point, as many flies
penetrated its area each day as were caught on it, leaving no other conclusion possible
than that, if no flies had been caught, as many would have moved away from it.
Otherwise stated, the fly population of this region was so far from permanently
fixed there that hardly any individuals sojourned there for more than a single day.

The experiment of the 5th to 13th November 1913 was repeated on the 17th to
22nd January 1914 (Table XII) and completely confirmed it. The conclusions
were subsequently confirmed in various other ways, and there is no doubt that the
flies of this species move freely about from place to place, forming continuous streams
of fly traffic along the shores of lakes, banks of streams, and, it was subsequently
ascertained, along the borders of woodland, game trails or human pathways, ete.

ITI (6). Movement of the Sexes along the Lake Shore.

Although it was impossible to reduce the local density of fly under such conditions.
as exist at Crocodile Point if an interval of several hours is permitted to elapse between
periods of catching, it was easily possible to reduce the density of male flies—not
of active females—by even a single hour’s catching, provided no interval elapsed

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 367

before the experiment was repeated. In other words, if catching is continued over
several consecutive hours, the density of male flies, but not of active females, will
be reduced during the first hour, and will not rise again to normal until several hours
have elapsed.

TABLE XIII.

Catches of Fly for consecutive hours at points on Lake Shore, showing Reduction
im Density of Males but not of Active Females, and demonstrating more rapid
movement of Females.

Catch of Fly. Density of Active Fly.

Locality. Date. Hour.
Total. | Females. | Males. |Females.| Total.
Crocodile Point.) (17th Jan. | Ist 343 30°9 do 35°0 14-1 49°1
HulapoIsland|| gia ..j| 22¢ | 319 | 414% | 267 | 185 | 45:2
: 3rd 250 5652 9 14°5 18°7 33°2
Ist 419 IO Oe 623 15 69'°8
| ond | 250 | 10:4% | 37-4 4:3 41-7
ed isle | Sra) isa) 213% | 24-0 6-5 | 305
| He athe 16s eh 25-294 o088 6:8 27-1
sth 167 26-3 % | 203 75 27°8

| : |

———

These experiments were many times repeated with results similar to those presented
in Table XIII. It was found that the local density of male flies could be very quickly
and easily reduced, whereas that of active females was not affected in this manner.
It was conclusively proved by experiments cited on p. 365, and illustrated by
Table XII, that reduction in density of males would be made good by incoming
flies if some hours were allowed to elapse before a second catch was made, but that
no period of waiting was required to make good any reduction in density of active
females caused by catching.

The only conclusion that can be drawn is that the active females habitually move
along shore much more rapidly than the males, or than many of the males, This
_is entirely in accord with the conclusions reached through experiments cited in Sect. I,
that active females are hungry, and actively seeking food, and that degree of activity
is correlated with abundance or scarcity of food. Their movements along shore
are stimulated by hunger. The movements of the males are in part stimulated by
hunger, but also in part (see Sect. IV) by sex instinct. The object of the females—
and of such males as require it—is to seek food; the object of many of the males
(forming a majority when females are inactive and the female percentage low) is
merely to seek the females, and this is accomplished by loitering along the routes
most freely followed by the food-hunting flies.

This conclusion found ample confirmation as the investigations progressed ; as,
for example, by the fact that when food is so scanty on an island (as on Kimmi
Island in January 1914) that the female percentage is very high, it is impossible
to reduce the local density of either males or females by catching experiments at

368 W. F. FISKE.

points where they are passing. Under such conditions nearly all flies are seeking
food, and all are equally active, as clearly shown by a comparison between the figures
presented in Table XIV and those presented in Table XIII.

TABLE XIV.

Catch of Fly at a point on Shore of an Island where Food was deficient, and
where nearly all Flies were seeking it, showing equal degree of Actwity on the
part of both Sexes.

Catch of Fly. | Density of Fly.

Locality. Date. Hour. | Sg
Total. | Females. | Males. | Females.| Total.
; Ist | 205 | 66:3°%!| 93-0. | 45:2) || (ieee
ond| 249 | 66:3 ° 28 0 55:0 83-0
Fee aces al pond) 4ae). ost, 1) oi) | AO eay mae
L “Hath! A684) Goa, | | S7e0 50:0 77°0
\| sth | 289 | 640% | 17:3 30°8 48-1*

* Falling off in density during the 5th hour was due to approach of evening. The sexes.
remained proportionately as active as before, as shown by the female percentage for this.
period.

lil (c). Effect of Shelter, or of Type of Vegetation, upon Long-shore Move-
ments of Fly and upon Percentage of Females in the Catch.

On 4th September 1914, a catch of fly was made under unusual circumstances
on the island of Bukassa. The shore at this point was lined with a very dense
fringe of reeds only two or three yards in width. Inside the fringe was open grass
land, cropped very short by hippo, with scattered clumps of thick bush and trees
o° a sort affording attractive shelter to fly. At the observation point itself was
a landing place of hippo with a trail forming a tunnel through the reeds. The
prow of the canoe was thrust into this tunnel, with the stern projecting beyond
the reeds into the open lake, and collections were made simultaneously by one
boy stationed in the stern of the canoe, outside the thick fringe of reeds, and by
two boys at and near the point where the hippo trail entered the opening, imside
the fringe of reeds, These reeds, it should be noted, were at least 10 feet in height,.
or higher than tsetse is at all likely to rise from the ground.

The three boys worked for two hours with the following results:

Inside Reeds. Outside Reeds.

Catch per boy hour—
Males a) ie Be fig 510 3D
Females .. ans as es 45) ES
Female Percentage .. M a 143% 68:29,

The extraordinary feature of this catch was the low percentage of females (14°3):
wmside and the high percentage (68°2) outside the barrier formed by the reeds; the
two points being separated by not more than 10 yards.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 369

This same phenomenon was observed a second time under different conditions
in a catch made on the 23rd September 1914 at a point on the island of Bussi,
where two boys worked ankle-deep in water outside a thick mass of reedy vegetation
that prohibited landing, and two other boys on a point of shelving rock, backed
by an open space and bushy forest of a type much favoured by fly as shelter. The
poorly sheltered, reed-fringed point on the shore was only about 100 yards distant
from the attractively sheltered point.

The catch was as follows, for the four boys for half an hour:

Well sheltered Poorly sheltered

Point. Point.

Catch per boy hour—
Males sf ot AG a 49-0 8°0
Females .. ie * oe 3°0 13°0
Female percentage .. i we 58 60°8

As before, there is an extraordinary discrepancy in sex ratio between points
separated, in this case, by only about 100 yards. And in both cases the high
percentage of females is associated with a type of vegetation known to be especially
repugnant as shelter and the low percentage with a type of vegetation known
to be attractive.

A hypothetical explanation for this phenomenon, which has withstood all tests
applied to it, is as follows :—

(a) The body or mass of active flies is contmually in movement, and
streams of flies are continually passing points along shore (see Sect. III (a) ).

(b) These streams of moving flies are made up of (1) food-hunting flies
of both sexes, which compose a variable proportion dependent upon abundance
of food and which move rapidly ; and (2) male flies which are not seeking
food, but which frequent the routes followed by food-hunting flies, and which
move much more slowly (see Sect. III (0)).

(c) The relatively idle and lingering males tend to prolong their sojourns
at points where sheltering vegetation is of the most attractive type, and to.
pass quickly, or not at all, by points where the vegetation is of an unattractive
type. Itfollows that (1) density of active males will be greatest where shelter
is most attractive, and least where it is least attractive, and (2) that the
percentage of females amongst caught flies will be greatest where shelter is
least, and least where shelter is most attractive to the lingering males
(because food-hunting flies must, for several reasons,* consist principally
of females),

When confirmation of this hypothesis was sought it was found on every hand,
not only in new catching experiments and fly survey work, but in old records, made

*'These reasons include the following :—

Ist. Becaus> the females in any district or region appear to outnumber the males
(see Sect. I (d) ).

2nd. Becaus, the females must nourish their young as well as themselves, and most
probably require food somewhat more frequently, in nature, than males.

3rd. Because the males are normally active at all times during good weather, and

are more apt than the females to encounter host animals without specifically seeking
for them.

370 W. F. FISKE.

ong before. The records made on the island of Kimmi in January 1914 are an
example, As presented in tabular form (Table XV) they are not particularly

TABLE XV.

Catch of Fly on Island of Kimmi showing effect of variable character of Shelter
on Density, Sex Ratio and Movements of Glossina palpalis.

Catch of Fly.

Date. Locality on Island.
(January

1914.) (Observation Point.) Male | Female | Total in

Density.| Ratio. | Catch.

27th ..|South-easternmost point. ee a Bee ee 3 J) 53°2 506
29th ..|North-eastern point ne sah “if “fell cael coal 66:9 14z
29th ..|North-western point .. ay ne an) whee 522 272
24th ..|South-western point .. a Ka Aa -Baeow 629 384
27th ../Southern shore .. We nm fs soe luted 53°0 276

* Shelier was appreciably less attractive at these two points.

striking, but the correlation between shelter, local density of males, and female
percentage is quite strikingly shown in figure 2. In this case the “ shore line ”
in the graph represents the entire circumference of the island, a distance of between
2 and 3 miles, with five observation points located as stated. The difference in
character of shelter was not particularly notable, but it was enough to produce
the effect shown.

Observation | South- —— —-North-.—North- South- : |
Point eastern eastern ——_—s western western Southern

| Percentage
lot Females

Density of
Malés

Shore Line |
Shelter Superior Inferior Superior Inferior Superior

——

~~ Fig, ae Graphic representation of dala given in Table XV.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. Bel

Records made in the course of the fly survey of Bugaba Island as presented in
‘fig. 3 afford a second excellent example of the correlation between character of
sheltering vegetation, density of male flies and percentage of females in the catch.
In this particular case there was a short reach of sandy shore at observation point
A3 which served as a breeding ground for the flies, but which was poorly (too
slightly) sheltered. At either end of this bit of sandy shore was forest, with marshy,
reed-srown fore-shore of a distinctly unattractive type. Only where the forest
‘growth came down to the more open sandy shore at points 42 and 44 was shelter
really attractive

[Observation
Points a4

Pees [pas ub — as

Female
Percentage 3050 0 80 a O 2530
Female | | er a
Percentage = ;
\
\
\
| \
\ tf
| \ vals
\ “a \
| \ 7 \
{ \ oe \
| ee
| yy, NUE \
' vg \
ee asi oe \
Density |
so vace-sseerbee = Dom LUE 2. Al out 3000 yards - - Agi ae hea

Poor Poor Good Poor Good Poor Poor |

Shelter

; Breeding
_Ground

None None None Poor None None NOs

Vig. 3. Correlation between character of shelter, density of male flies ana
percentage of females.

Figure 3 shows an excerpt from the data secured during the “ fly survey ” of the
island of Bugaba (vide Sect. VI. (g) ) and illustrates conditions of infestation by fly
along a reach of shore about 12 milesin length. In all, more than 150 miles of mainland
and island shore were systematically surveyed in this manner, and the data secured
have served in scores of cases to confirm the conclusions herein stated concerning
the correlation between character of shelter, density of males and female percentage.
A single other example may be given in Sect. III (d) which follows.

III (d). Notes on Survey of Lutoboka Bay, Bugalla (Sesse) Island.

Figure 4 presents the results of a survey of some 5 miles of shore, principally
included in the large circular bay that lies within the peninsula of Lutoboka on
Bugalla (Sesse) Island. This survey was undertaken with the object of collecting
-data relative to the effect of environment upon range and density of Glossina palpalis.

SnD W. F. FISKE.

It is impracticable to include in the graph data concerning environmental features.
responsible for the striking variations in male density and sex ratio shown by the
curves. In general they are as follows :

Points 1 to 4 inclusive are typical of shore infested to slightly above the average
extent for this general region. This reach of 14 miles lies on the western shore of the:
bay, and is densely wooded, with bits of open shore alternating with reed-grown and
jungly reaches. There were some sand and gravel deposits affording good breeding
grounds.

Points 5 to 8 inclusive :—The forest continues, but there is no more open shore,
nor any breeding places. Instead there is a fringe of floating sudd (papyrus and
saw-grass) lying off-shore and no open space between it and the massive shelter of
the forest behind. Catches were made from a canoe outside the sudd, which, with
the exception of crass, 1s the most repulsive type of vegetation.

Only the hungry food-hunting flies (principally females) pass beyond the limits
of the open shore at point 4. The relatively idle males turn back.

The minimum density of males and maximum percentage of females is recorded
from point 6. At pomt 7 males are more numerous and female percentage has
fallen, indicating that a new fly colony is being approached. The catch at point 8
confirms this oa makes it certain that males are coming into the repulsively sheltered
reach from the other direction.

At point 9 there is a break in the sudd, and an open grassy bit of fore-shore,
scattered with bushes and backed by massive forest. Shelter is attractive, but as
yet no breeding grounds occur. The excellent shelter makes for a sharp increase.
in male density and a corresponding decrease in female percentage.

At point 10, the forest, which has continued unbroken until now, ceases and open
grassland comes down to near the water’s edge. The foreshore is open, with scattered
bushes, and deposits of beach sand afford excellent breeding grounds. For a short
space between points 9 and 10 this series of open sandy belts is backed by the massive
shelter of the forest. This combination affords complete protection to the flies,
both as pupae and adults, and forms the centre of the colony, or centre of infestation
for the shore on either side.

Beyond point 10 the open fore-shore, with sand deposits and the slight shelter
of scattered bush, continues (as at point 10), but is no longer backed by massive
shelter. The effect on male density and female percentage is precisely as caused
by the fringe of sudd. The idle males turn back when massive shelter ends, and only
the food-hunting flies, in part, continue. 7

At point 14 is a relatively thick mass of bush, which tempts a few males to linger,
but not for long, and density does not rise again until, some distance further on,
forest shelter 1s again encountered.

There is not the slightest doubt that in these observation points high percentage
of females coupled with low density of males is due to insufficient or unattractive
sheltering vegetation, leading to a partial segregation of the food-hunting flies of
both sexes from the relatively idle and satiated males.

Fema
Percen? 0

1g effect of environment

Bale & Danielsson, LtY

Observation
Point

Density

Female
Percentage

Density

Female
Percentage

Shore Line
Scale of Miles

Fig.4. Fly survey of a reach of shore in the district of Lutoboka, Island of Sesse (Bugalla), illustrating effect of environment
on range, density and movements of Glossina palpalis.

Bale & Danielsson, 67

<
‘
- a :
ssanealad Tita iebeeenesl hh mrecdemeettied ieee Ae eee : 7 ,
. ‘ aye 8am tu, Pers ; ._- ae ¢
' 4%, ale NOTEVISEOU F
. el ( : $ Mio. mee © 1
ht eed nie nae mse rye pip nwa en ne ts tail 5
¢ " . he ‘, } * i? Ai. iPS
a ; 1 , “iene fe
Ma oT ry
=
F +
fm.
ss hy
' 7"

SE —— 8 gy enn

a
'
a
” o
™ 7%
“-
o ~
ae
-
¥, r
\
- | Mb we _ —
o") :.
ry
;
censunhtiiemidiamecddiecca te eect ea ae
es pa ieee
: at

Sry. Ma eT te, aay:
1 SH A SUAS. mG

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. o1e.
III (e). Maximum Percentage of Females brought about by Segregation
through Movements.

The maximum percentage of females amongst active flies encountered at any
point on island or mainland was in the district of Buddu between the old landing
places known as Kalkosa (or Sekwe) and Bale. The figures of the catches made
during a fly survey of this coast are presented in Table XVI. The distance from

TABLE XVI.

Catch of Fly wn the Mujuzi Colony.

Catch of Fly.
Date. Observa- | Density | Percentage
(June 1915.) tion Point. of Males jof Females.
Males. | Females.) Total.
24th by: me aM 40 26 39 65 | 26:0 60°0 %
24th “oH 4s os 4] 12 90 102) 40 88°3 %
22nd a a We 42 33 65 Gar 165 66°0 %
22nd aye die Be 43 23 68 91 115 OEE Ss
22nd i a a 44, 58 Wipe ils 19°3 Tar 0 Of
24th e ne és 45 196 162 358 98°0 AD 5
22nd is “ a 46 106 103 209) 45320 49°2 %
26th 2 t oe 47 4] 73 114 |. 205 64:0 %
26th SS, ie ores 48 45 120 165 19°3 TT
26th fi nae ae 49 20 84 104 10°0 80:8 %
26th Bec ahs oe! 50 10 80 90 5:0 88°8 %
Totals and averages A. 570 104] 1611 26°0 67°6°%%
Average by points se a oe 26°6 69°5 %
l

point 40 to point 50 is about 34 miles. Point 45 is at the mouth of Mujuzi Creek,
which affords excellent harbour for crocodiles, as well as breeding places for the
fly on its banks. It is the colony centre, and centre of infestation for a considerable
reach of shore. Points 47 to 50 are-in the Bale fly beach mentioned in Sect. II (a),
Table VIII.

The average female percentage for this district is either 64°6 or 69°5 according
to the method of computation, and compared with the maximum encountered on
the islands (57°5 or 64°6 for the colony on Kimmi Island; vide pp. 363 and 370)
is about 6°0 points higher. This is undoubtedly due to scarcity of food, which only
occurred sufficiently or at all at the very mouth of Mujuzi Creek (point 45). The
behaviour of flies bore this out completely. Nowhere else on island or mainland
were they so troublesome; it was impossible to avoid being bitten, even with both
hands employed wielding fly switches.

The extraordinarily high proportion of females at points 41, 49 and 50 break all
records made elsewhere (for equally large catches), and are certainly explained by

374 W. F. FISKE.

the unattractive character of the environment* at these points retarding the movement
of the relatively idle and inactive males along shore. These males were concentrated
along the short reach of shore including points 45 and 46 ;_ and elsewhere it was almost
entirely the actively moving, food-hunting flies that were caught.

This particular colony of Mujuzi is in several other respects one of the most interest-
ing and deserving of study of any encountered on either island or mainland. It
is mentioned elsewhere, and a graph of it is presented facing p. 388. I: was non-
existent in 1906 but came into being subsequently as a result of environmental
changes described on p. 458.

III (f). The Full Significance of the Phenomenon of Sex Disparity in
Glossina palpalis with Relation to Food Supply, Shelter and Movements
of Fly.

The phenomenon of a variable ratio between the sexes of active flies of Glossina

palpalis proved on investigation, as I have attempted to make clear, to be replete
with interest and significance,

Taken over any considerable district or region it is a fair index to the abundance
of preferred hosts, and thereby of the relative immunity of man to attack (see
Sects. I and II).

Taken locally, pomt by pomt along a reach of shore, and coupled with the local
density of active males at these same points, it becomes a subtle index to the
attractiveness or repulsiveness of the local environment to fly. This fact having
been conclusively proved, thereafter data in such local variations become of really
great value in reaching conclusions concerning the relative attraction of various
types of vegetation, etc., to the flies.

Finally, through study of this sex disparity and its causes, a very clear conception
is obtained of the ordinary movements of the flies along the shores of the lake, banks
of stream, borders of woodland, game trails, foot-paths, or other favourite courses.
A great many little experiments and informal observations were made to ascertain
the routes most freely followed by the active food-hunting flies—too many to mention
in detail. The lake shore is the most favoured of them all, but, in general, they
follow quite closely the line separating sunlight from shadow.

The flies are averse to penetrating shadow, unless sunlight is perceptible beyond,
and even more averse to crossing sunlit spaces unless shadow or massive shelter

is perceptible beyond.

*It was not necessarily unattractive shelter that prevented the males from
congregating at these particular points. They were distant from all sources of food,
and however attractive shelter may have been (it was in fact very attractive at 49 and 50,
but not at 41) no permanent concentration of males could occur.

The males tended to congregate at point 46, which was well sheltered; but not at
point 44, just the other side of the creek, where shelter was not attractive. But by
the time the relatively idle males had dispersed much beyond point 46, towards points
49 and 50, they began to require food, and there was no food unless they passed back
to point 45, or passed far beyond point 50.

In other words, a reach of foodless shore, although supplied with both shelter and
breeding ground, produces the same falling off in male density and elevation of female
‘percentage as the shelterless shore described in the pages preceding this.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. od

It was found by experiment, details of which need not be given, that flies will cross
open channels up to at least 300 yards in width, but that they do not cross them
very freely. Itis certain that the width of the channel likely to be crossed is dependent
on the character of the farther shore. Certainly they do not cross much wider channels
than 300 to 500 yards at all freely, for from the beginning evidence of their doing
so was carefully sought, and no shred of it secured.

It is the movements of fly along shore, or along the border line between sunlight
and shadow, with which we are principally concerned, and these studies have shown
how the flies tend to congregate in attractively sheltered localities, where, in conse-
quence, density of males is high and percentage of females low, and how, pressed
by hunge~, both sexes, but more especially the females, tend to disperse from these
centres of greater density and range along the lake shore or other favoured route
in search of the sluggish amphibian animals on which they principally feed, thus
accounting for the extraordinary preponderance of females amongst caught flies
in repulsively sheltered localities.

IV. Tue Assempiine Hapits oF GLOSSINA PALPALIS.

The stimuli which keep in motion the streams of tsetse described in the preceding
section are known to include, and believed in all ordinary circumstances to be
limited, to the following :—

(A) Hunger. Hvery second or third day the flies of both sexes undertake
food-hunting flights. These are relatively rapid and easy to follow.

(B) Self-protection. At the close of each period of activity the flies of both
sexes are moved to seek massive shelter, where they remain in seclusion
during periods of inactivity. They appear very unwilling to move far
away from such shelter; hence flights in search of it are usually of short
duration.

(C) Maternal Instanct. At intervals of eight days, or longer, the females are
stimulated to seek breeding grounds wherein to deposit their young. The
great concentration of puparia in very attractive breeding places is proof
that flights in search of them were of considerable duration. The location.
of deposits of puparia near the well-marked courses followed by food-
hunting flies is proof that these are followed. But the flights themselves.
are very hard to follow.

(D) Masculine Instinct. Hvery day, if the weather is good, the males appear
all to leave shelter to undertake assembling flights. These are readily
followed, and as described in the preceding section, are relatively slow.
The insects tend to congregate or to loiter at points or along courses where
the females are most likely to congregate or to pass. It is these special
activities which are now to be considered.

Several points in connection with the assembling habits of Glossina palpalis in
addition to those touched upon in the preceding sections are discussed in the sub-
sections following. The topic is inter-dependent with several others, including,
especially, host preferences of the flies,

LO We P. PISKE:

IV (a). The “Following Swarm’’ of Glossina palpalis.

Many observations, but few detailed experiments were made upon the curious
habit of Glossina palpahs of approaching and scrutinising strange objects. It was
noted that there is ordinarily little discernible difference in this behaviour toward
inanimate and animate objects other than their favoured hosts. When the percentage
of females amongst active flies is low, man is almost immune to attack, but his person
is none the less, and equally with any other similarly conspicuous object, strangely
attractive both to the food-hunting flies of both sexes, and to the relatively inactive
males. It is on this account that the males are so easily caught, whether or not
they are seeking food.

Of all objects, apart from their favoured hosts, the most curiously attractive
to flies is, perhaps, an open dark-coloured umbrella, and in the catching experiments
and fly survey work the fly boys were each provided with one. When displayed
at any point in or near the course followed by food-hunting flies, or the relatively
inactive males, it becomes so conspicuous an object as to attract to it virtually
all flies which perceive it. They crawl over it, especially its underside, inspecting
it curiously until curiosity is satiated—or for, perhaps, one to five minutes—when they
pass on. And this is equally true of both idle males and food-hunting individuals
of both sexes. Despite the extraordinary attraction of the object there is no tendency
on the part of any flies to linger more than the few minutes required for its inspection,
so that a swarm will never collect. If the density of the stream of passing flies is
20 or 30 per hour, there will be from one to four or five flies resting on the attractive
object at all times, but the same flies do not remain for long.

As above stated, the behaviour of the flies toward man 1s, ordinarily—z.e., when
the man is alert, food is plentiful and the female percentage low—aindistinguishable
from their behaviour toward his coat thrown over a bush or his umbrella openly
displayed. It is not necessary to think about the flies or to ward them off;
occasionally one will bite, but the number doing so is so small that it is not worth
thinking about. Man is attractive to flies, but does not hold them after their curiosity
is satisfied. The same is true of goats, sheep, and many other animals.

But with certain animals under certain conditions the behaviour of the flies is
conspicuously different. When a crocodile or Varanus is basking in half open
sunlight, the flies attracted to them are apt not to pass on, after cursory examination,
but to linger indefinitely. The result is the gradual accumulation of an accompanying
or “ following ” swarm, which may grow to number several times more flies than a
boy stationed at the same spot would be able to catch in an hour’s time, and which
may certainly represent (sometimes) the accumulation of the majority of flies passing
the spot during several hours’ time. Similar swarms may collect upon or near a
foraging Varanus, as it wanders slowly through the woodland, or a grazing situtunga,
and may then follow the animal out of woodland into open grassy spaces, where the
flies would not go of themselves,

Mention of these conspicuous swarms of flies in connection with Varanus, crocodile,
tortoise, hippopotamus and situtunga is made in Sect. VI (e). A few additional
notes excerpted from my field books follow.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPATIS. Old

On Varanus. Kitobo Island, 3rd December 1913.

Varanus observed excavating burrow in sandy soil some little distance from
shelter. On approaching it made off rapidly, and on reaching the spot I was assailed
by a great number of tsetse, which swept back and forth and around me like angry
bees, “‘ buzzing ” in their flight in a manner never before noted. After a few minutes
they all dispersed, without any of them alighting upon me.

‘On Varanus. Nsadzi Island, 3rd February 1914.

A Varanus was suddenly flushed in grass land (grass about knee-high) not much
less than a hundred yards from any shelter. It made off rapidly and a swarm of
flies, numbering several dozen at least, rose and filled the air with their angry, bee-like
“buzzing.” As on Kitobo they swept back and forth and around about me, but
none alighted or offered to attack, and in a minute or two all had dispersed.

On Situtunga. Bugalla Island, 23rd November 1914.

A large male situtunga was approached as it was feeding with its head concealed
in a dense thicket. With glasses (Zeiss prismatic x12) it was possible to make out
that a peculiar dark colour of fore leg, lower shoulder and thorax, which were plainly
seen through an opening in the bushes, was due to an unprecedented number of
tsetse, which literally blackened its coat. It seemed entirely unmoved and
phlegmatic under attack.

On being shot, the animal plunged directly through the thicket ; ran a few yards
at great speed and fell. On proceeding to the spot where it was feeding, I found
a “following swarm ”’ of fly of unprecedented size (probably not less than 200 flies)
buzzing like a great swarm of angry bees. They surrounded me, but hardly any
alighted on me or followed me to where the antelope lay.

On Situtunga. Damba Island, 13th September 1915.

A large male situtunga was shot in an opening in the forest in the dusk of evening.
It ran into the thickly shaded forest and fell. On reaching it I was amazed to find a
considerable swarm of flies, partly outside, but judging by the noise they made,
more inside than outside the forest (it was so dark inside they could not be seen). Is
it possible that a swarm will follow an animal into the night, and perhaps remain
on its body all night ?

On Varanus. Manene Island, 16th March 1915.

While passing through forest a movement in the vegetation (the ground was
completely covered with broad-leaved herbage rather less than knee-high) indicated
the approach of some small animal. It proved to be a Varanus, which came very
slowly, evidently hunting insects and molluscs, to very near me. At no time could
I see its body, but keeping pace with it and evidently following the movements of
the herbage was a swarm of perhaps two dozen tsetse. The flies were not anxious
to feed or to reach the body of the animal, but merely followed its movements,
alighting on the herbage or hovering about. A movement on my part sent the
animal scuttling rapidly away, whereupon the disappointed flies rose and swept
back and forth, buzzing angrily as usual in such cases,

378 W. F. FISKE.

On Situtunga. Bugalla Island, December 1914.

(Original note lost.) On entering an open space in the jungle where formerly
were plantations, a small herd of two female and two half-grown male situtunga was.
seen, with other animals feeding in the edge of the jungle out of sight. Those in
sight did not immediately see me, who stood motionless watching them, nor upon
seeing me did they betray alarm or more than mild curiosity. The whole herd
moved in my direction and one female approached within three yards. Hach animal
was followed by a small swarm of tsetse—perhaps 15 or 20 flies—few of them on the
animal itself, but principally on the vegetation close at hand, or hovering about.
Not one of the flies was seen to feed, nor did the animals show annoyance at their
presence. On becoming alarmed the antelopes made off without undue haste,
the flies following.

The above may serve in some manner to make clear the difference between the
behaviour of fly toward a favoured and complacent host, and such another as man,
who is neither favoured nor complacent. Anyone passing through infested territory
and aware of two or three, or perhaps half a dozen, tsetse constantly hovering about
him is certain to receive the impression that this number of flies is persistently
following him. In the case of Glossina morsitans the impression would, probably,.
be correct, for man will collect a following swarm of this species; but with Glossina
palpalis 1t is incorrect. The same individual flies will not follow a man for more:
than a few minutes unless they are very hungry, but will quickly drop behind and
be replaced, perhaps by others. This is easily proved by catching them just as
they appear; one will quickly catch many times more than appear to be following
at any one time.

The explanation of the “following swarm” with Glossina palpalis is, without
doubt, identical with that put forward by Lloyd for the “following swarm” of
Glossina morsitans. It is the assembling habit of the males; first, to linger along
the routes followed by the food-hunting females; secondly, to examine any strange
object coming in range of vision; and finally, on encountering a complacent host
of a favoured species, to remain by it ; for in these ways they are most likely to come
into contact with the seclusive females.

Nothing else in the habits of the fly affords more convincing evidence of their
unwillin: ness to feed upon man, unless they are forcedby hunger. Far from attracting
a following swarm, man cannot even hold one that has already been attracted to some
other host ; if by his approach he startles the favoured host into flight, there is not.
the slightest tendency on the part of the following flies to turn their attention to him,
but invariably, as in the cases noted above, the swarm has dispersed without the
flies paying as much attention to the intruder as is usual when other hosts are about.

IV (6). Conditions under which a Following Swarm will Collect.

More than 25 separate experiments were made with various wild and dumestic
animals tethered at poimts where they were exposed to columns of fvoa-nunting
flies, mainly to ascertain the host preferences and feeding habits of the insect, and
finally in a specific effort to induce a “‘ following swarm ”’ to collect under observation.
In the very first experiment in the series such a swarm did collect upon a tethered
Varanus, but all other experiments failed in this respect.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 379

Under entirely natural conditions swarms had been seen on Varanus, situtunga,
crocodile, hippopotamus, pig (of domestic stock, but run wild) and tortoise, and
most freely of all on Varanus. The fact that repeated attempts to induce one
to form on Varanus which were held in constraint all failed, excepting only the first
(see Sect. VI (c), experiment 4), convinced one that something more than the mere
presence of a host animal of a favoured species is necessary, and failure is believed
to be due to the impatience of these animals under restraint. In other words, a
host must be complacent under attack as well as of a favoured species or it will not
be favoured by tsetse to the extent of attracting a “ following swarm.”

The final experiments with Varanus, sheep and oxen are described in Sect. VI (c),
experiment 3. They led to the conclusion that a host to be “ favoured,” must
always be complacent, and suggested that perhaps almost any of the larger animals,
including man, might be favoured, even to the extent of collecting a “ following
swarm,” if it were in them to suffer the attacks of Glossina without sign of protest.

‘\V. Tae Errect of OPpEN WATER AND Humip CoNDITIONS UPON THE RANGE
OF GLOSSINA PALPALIS.

No factor in the bionomics of Glossina palpalis is more obviously correlated with
the range or distribution of this species than open water. It is a riparian species,
and its occurrence far from shores of lakes or banks of streams is a subject for remark
whenever observed.

Exceptional dispersion inland, however, is not at all rare, and many cases of it
were encountered and studied. In all cases the explanation was the same, and
involved, as virtually every topic in Glossina bionomics seems bound to involve,
the subject of hosts and host preferences. The two topics are inseparable in field
work and cannot well be separated in discussing and recounting field work.

These studies were more convincing to the observer than it is feared they may
prove to the reader, that the correlation in question is purely coincidental; that water
or humid conditions are not of direct benefit to the insect, nor required by it; but
that a proper combination of food, shelter and breeding places which is requisite
to its existence occurs so infrequently away from the shores of lakes or banks of
larger streams that it is perforce riparian in habit.

V (a). Range of Fly inland from Lake Shore.

Ordinarily the range of fly inland from the lake shore is extremely limited; at
50 yards there will usually be a sharp reduction in density ; beyond 100 to 200 yards
flies will appear only as stragglers; and at 300 to 500 yards they will disappear,
or, at least, density will drop below any figure that can be conveniently measured.
If range inland is more extensive than this, some special cause for it must exist.

Attempts to measure the ordinary inland range of fly were limited to a single
experiment on the island of Kitobo in December 1913, which resulted as follows—
the catch being made at the base of a slope that rose rather steeply to a plateau with
abandoned plantations.

(659) Cc

389 W. F. FISKE.

Catch at base of slope, 54 “boy hours,”

showed density

(both sexes) of KA se ae BS 15:50
Catch ha.f-way up slope, 16‘ ‘boy hours,” showed density of 1:05
Catch at crest of slope, 28 ‘“‘ boy hours,” showed density of ‘26

(This point was well under 300 yards from the water’s edge.)

Catch in old plantations beyond the crest of slope, 14 ‘ en
hours,” was Nil.

This is probably a sharper deloe, in denear ee cee bee not ea eaustale SO.

V (0). Effect of Marshes and Rivulets on Range of the Fly.

The effect of lakes or streams on the range of Glossina pa palis is so conspicuous
that whenever. any unusual extension of its range inland from lake shore or bank
of large stream has been observed it has almost always been accounted for as being
due to presence of small streams, marshes or humid conditions of life generally.
Therefore if any island were to show inland range of fly too much in'excess of that
noted as “ normal” in the preceding sub-section, the question of interior conditions
with respect to water and elevation (dryness) is the first to be considered.

A survey of the islands was begun in January 1914 to ascertain and measure
variations in degree of local infestations, as they might be correlated with differences
in local environment. Mostly small islands, of less than 1,000 yards short diameter,
were visited ; but by the Ist September out of a total of some 40, 8 of larger size
had been included in the survey, and conditions with respect to infestation by fly,
humidity and topography of the interior of each of these islands were found to be
as stated in the accompanying table (XVII).

TaBLE XVII.

Fly Survey of Islands in Victoria Nyanza showwmq Lack of Correlation between
Humidity of Interior and Infestation by Glossina palpalis.

Conditions in Interior. Infestation by Glossina.
Island.
Topography. Humidity. Shore Line. Interior.
Wema Hilly Diya e, aca Dense None.
Damba .. Flat Sinem and Dense Medium.
marshes
Kome .: Hilly Streams. . Light None.
Kitobo .. Plateau Dry* Medium. . None.
Bukassa . Hilly Dry* Light Light.
Bugaba.. Hilly Streams. . Light None.
Bubembe Hilly Dryt Light None.
Bugovu. ..| Hilly Dryt Light None.

* Careful survey showed no streams near points of interior infestation.
+ Streams may exist on paris of islands that were not explored.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 381

It will be seen that of the eight islands listed, two differed from the rest in being
infested in the interior. One of these, Damba, is flat. with streams and marshes
in its interior, and with a dense infestation of fly along shore. Two others—Bugaba
and Kome—also had streams flowing down from the hills, but they were hilly and
the infestation along shore was light. Hither elevation or lightness of infestation
along shore might be a possible explanation for absence of fly mland, and its
absence is merely negative evidence that it does not always occur in humid localities
near lake shore colonies.

But in the case of Bukassa Island we have a dry, hilly interior, with only a light
infestation along shore, and for all this a general infestation of the interior. This
is not negative, but is absolutely positive and very strong evidence that, whatever
it may be which restricts the range of this tsetse to a narrow belt along the shore
of the islands generally, it is not the presence of water or soil humidity, for the interior
of Bukassa is, perhaps, the driest and hilliest of any of those listed except Wema.
And whereas it is perfectly possible that the presence of open water might account
for infestation by tsetse of a belt several miles in width along the borders of it. it is
impossible ta consider water alone as the determining factor when, on some islands,
the inland range of fly is virtually limited to 500 yards or less, whereas on others it
is extended to 1,500 yards and more, and with no surface water in sight.

Slight though this evidence is, it is of such a positive character that it 1s practically
conclusive. Some other factor than open water must be held accountable for
limitation to inland range of Glossina palpalis from shores of lakes or banks of streams.

V (c). Effect of Food Supply on Range of the Fly.
A comparison between the two islands, Bukassa and Damba, in search of points

of similarity, which should at the same time be points of difference between these
and the other islands named in Table XVII, led to the conclusion that the most
probable explanation for their infestation interiorly by tsetse was the presence of
unusually large numbers of game animals: of situtunga on Damba, and of domestic
pigs run wild on Bukassa. This conclusion was reached reluctantly, because at that
time (September 1918) it was believed that the only favoured hosts of this species
of tsetse were reptilian.

The course of the fly survey of the islands was thereupon set to include careful
study of conditions on all such as were known or suspected to be more than usually
well stocked with game. These included specifically the group of semi detached
islands of Bukone, Serina and Lulamba, and the island of Buvu, upon which
pigs were known to occur, and the (relatively) very large island of Sesse (Bugalla-
Buninga), which had been reported to be overrun with situtunga.

Examination of the pig islands disclosed the fact that though great numbers
of the animals had been present (enough to cover them with networks of trails
and to uproot the soil nearly everywhere in the woodland) some great catastrophe
had reduced their numbers to a few stragglers. More skeletons were found than
traces of living animals. But one result of the visit was to discover a “ following
swarm” of fly about a pig that was shot, and to prove conclusively enough that
this animal may be a favoured host. None of the islands was found to be infested
by fly except very narrowly along the lake shore.
(659)

382 W. F. FISKE.

On Sesse the outcome was much more decisive. Situtunga were found everywhere
in numbers which are extraordinary for an antelope accounted “ rare’ and a denizen
only of marshes. They were even more numerous in the overgrown plantations
and village sites on the plateaus, and elsewhere on the higher levels, than in the
woodland and marshes along shore. The island is nearly forty miles in length, and
eventually was explored from end to end. Practically every bit of woodland or
jungle upon it was found to be infested by Glossina, including points in the interior
of the western peninsula known as Buninga that are at least 24 miles from the lake.
The only exceptions were certain areas of woodland in the central portion of the
eastern half of the island (Bugalla) which, though hardly a thousand yards from
the lake, were completely surrounded by open grassland. Elsewhere a great diversity
of inland environment was found _ high hills and low; marshes and rocky summits ;
original forest and the frequently impenetrable jungles which have sprung up in
the abandoned plantations; but everywhere that shelter at all attractive to the
flies was found, there would be both situtunga and fly.

Except at the very beginning formal records of “‘ catches ”’ in the interior were not
kept, because the work of exploration was done almost entirely on dull days or at
hours (after 3 p.m.) when the flies are not as active as earlier. The records made
at the beginning are presented in Table XVIII, in comparison with those from
Bukassa, where inland extension of range is due to great numbers of pigs.

TaBLE XVIII.

Catches of Fly on Islands of Bukassa and Sesse (Bugalla) showing Inland
Extension of Range due to Presence of Game, made in September 1914.

| Bukassa Island. _ Bugalla Island.
Collection Points.

_ Female Male Female

Rite. ratio. | density. ratio.

Lake shore... w ne v adh; O78 Ay) LOMES,. etga-p 22°5 %

Sandy plain, 400 yards inland me Ce lene — 25°0 lob

Hills, 1,000 to 2,000 yards inland .. a RE Eh EOF 10°0 Tore

57. OUD 59) BLOOD res ‘aila, Se elder rch) hawt 47 94 %

» 90,000 ,, 4,000 oe af % ee ee — 1°3 =~|\ None

S400) OF more. |”. iy ey ..| — ‘4 |J Caught,

Of significance is the low percentage of females in the interior points; indicating
attractive shelter or ample food supply or both.

The only other host animal occurring on the islands, that is found at all commonly
more than a few yards from the water front during hours when flies are active, is
the monitor lizard (Varanus).* On certain islands—notably Manene and Dziru,

*Execept on the sudd-surrounded islands of Baujako and Binga, where bush-buck,
bush-pig and buffalo occur, and which are faunistically a part of the mainland.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 383

both of very small size—it is extremely common, and on the island of Mbugwe
it is far more numerous inland than is usual. On these islands, also, unusual extension
of inland range of fly was noted, as described more in detail in Sect. V (d) following.

At the end of the island survey the conclusion was definitely reached that
the range of fly inland up to a distance of approximately 24 miles is controlled in the
islands by the abundance of food. A summary of observations on this point in tsetse
bionomics is presented in Table XIX.

TABLE XIX,

Fly Survey of Islands in Victoria Nyanza showing Correlation between
Abundance of Host Animals and Infestation by Glossina palpalis in the Interior.

Infestation by Fly. Host Animals in Interior.
Island.
Shore, Interior Pig. Situtunga.| Varanus.

Wema ..\Heavy ..\None ..\None None .. Pew.
Kome Oy eden” ally a’ ee ..|Very few. .|Very few.
Bugabu . .|Laght a as | i Alpe » --iKew,
Bubembe . |Light Oe | Rs Saintes . .|Rew Eat
Bugovu . Light NW i a Wer le Welles:
Buvumira .. (Light i SALA Ess a iets pan ial ne
Fumbe . (Light a 4 ane .. (Few Seats
Bunyama . (Light sis sialy ess ..|Very few. ./Very few.
Lulamba ..|Light NTA ..|Very few. .|None .|Few.
Buvu. ..(Medium ..| ,, re S ae ‘¢
Bukone . {Light 5 Bia si - a
Serinya ..|(Medium ..| ,, BS ; " is
Kitobo ..|Medium ..1 ,, None i =
Bukana ..(Light ..(/Light Many 55 5
Damba .|\Heavy ..|Medium ..jNone Many r
Zinga a sag .|Very heavy|Light * i a
Sesse (Bugalla) :—

Western Portion . .|Heavy sHhaghiby ls) .aliag Very many ,,

Northern ,,. .. . .|Heavy Medium . a % ae

North Central Portion ..|Medium . .|Light ~ s am i

South ,, vA ..|Medium . .\Light fA. cH ys 3

;

*Manene and Dziru are small islands less than 1,000 yards in short diameter, but

are remarkable in that infestation of the interior is equally as heavy in the case of
Dziru and heavier in the case of Manene than along the shore.

V (d). Unland Range of Fly as affected by Varanus on the ‘Islands.
Manene Island.

Manene Island is, perhaps, 1,000 yards in length by 500 in breadth. Its northern
half is elevated and rocky, its southern half low and fertile. Mostly it is covered
with thick bush or forest, except for the rockiest places, and also excepting the
site of old plantations in the very middle of the southern half, which are kept open

and closely grazed by hippopotamus. These spaces are entirely surrounded by
thick bush or forest.

384. AW, aE SU UBISIKGR,

Varanus was more common than on any island of similar size visited. Perhaps
as many as six per hour were flushed during the survey of it. They were everywhere,
but were especially common in the clear space above-mentioned, where many had
their burrows.

There were several bits of sand or gravel beach which offered good breeding grounds
and excellent shelter for fly, but the shelter in these was less attractive than that
in the close-cropped clearings, surrounded by forest, bush or jungle and with bits
of massed vegetation breaking the openings everywhere

The catch of fly was as follows :

Male Female
Density. Percentage.
Shore points .. a bt mo ¥ 12°3 ie 217
Inland clearing ay sis si8 21:0 a 18

This is quite an exceptional case of concentration of fly away from the water
(150 to 200 yards), due to more attractive shelter and constant abundance of food.
No breeding places could be found except on the water front.

Dzru Island.

Conditions generally similar to those on Manene, except that shelter in the interior
was not so attractive and concentration of fly less notable.
Mbugwe Island.

Mbuegwe Island is about two miles in length by one in greatest breadth. It is
densely forested, and appears never to have been cleared for cultivation except
at a few points, now overgrown with jungle. The one open space discovered on
the island is on the crest of a steep hill that occupies the centre of the broadest
portion. The very summit of the hill is flat rock, partly covered with thin soil,
rank grass and scattered bush. The shore is generally rocky with here and there
bits of beach.

Varanus is common. Ordinarily this reptile occupies burrows in rocky places
or excavated in light soil in grassy openings, near which are located its habitual
basking spots. On Mbugwe there are no such openings as are frequented by i1t—
at least none were seen—except on the crest of the hill above-mentioned, the slopes
of which to the very edge of the water were covered with dense, old forest. Hence,
although this opening was fully 800 to 1.200 yards from the water, it was much
frequented by the animals, which had fairly covered the soil in places with their
excreta (containing shells of crabs and molluscs from the lake).

The catches of fly made on this island (4th and 5th March 1915) were as follows :—

Male Female

Catch. Density. Percentage.
Points on shore at Py Se ks oe 16°6 ie 33°5
Crest of hill .. eS $e 83 Be 13:7, vi 50°6

Breeding grounds were sought for and found in and near the basking spots of
Varanus, in dry vegetable debris that partly covered the rocks. Six boys, searching

for one hour found :
Total. Per Boy Hour.
Empty shells :% Ae beh yi ene o she 9°2
Healthy puparia .. ey “4 4 agg" ie 3°5

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 385

The shells had all hatched, and none had been eaten or destroyed by small
predators.

This particular colony is the one and only inland infestation discovered at any
pomt which appeared to be quite independent of colonies on shore, and this is the
only occasion on which breeding grounds adequate to sustain an independent colony
have been found except in old or recent deposits of beach sand or gravel.

The conditions are unique in many respects. The very dense forest which clothed
the slopes of the hill was a strong deterrent to dispersions of fly. The local environ-
ment was far from being attractive, and it is exceedingly doubtful if the flies would
have remained in it if the thick forest had not been so relatively less attractive.
But they were walled in as on an island, or as in a great breeding cage—it chanced
that there was a regular supply of food—it chanced that there was protection for
the pupae—and in consequence the flies remained and bred.

There is no reason why such colonies should not be found at any distance from
the water front, except that such conditions are exceedingly rare in this region.

V (e). Dispersion of Fly inland from the Mainland Shore.

At numerous points on the mainland it was ascertained that inland dispersion
of fly was no more extensive than is usual on theislands. The two notable exceptions
follow. |

Dumo Point, Buddu, June 1915.

Dumo Point is a peninsula. with a hill and old plantations, separated from the
mainland by an open plain. It is in large part densejungle. Near the very centre,
about 500 yards from the lake shore on two or three sides was a semi-open space,
kept grazed by hippo, in which grew great quantities of ga Bush-buck was
rather common. A herd of bush-pig had a retreat in the jungle and clearly marked
trails showed where it made regular rounds of the guava thickets for the fallen fruit.
In the immediate vicinity of ae thicket, but not elsewhere except in the usual
narrow belt along shore, Glossina was common. The catches stand as follows:

Male Female
Density. Percentage.
Points along shore of peninsula .. #': SONGS ARO ez
At guava thicket, 500 yardsinland ia Sorc awa

The shelter at this point was quite attractive, but no more so than at many other
points ; the inland dispersion of fly was distinctly unusual and the presence of both
bush-buck and bush-pig, the latter particularly, led to the conclusion that the presence
of fly was thus explained.

Bujaju Pernnsula, July-August 1915.

Bujaju Peninsula is an extensive tract of land, almost an island, but with the
deep bays behind it choked permanently with sudd. _Its shore is very marshy, and
principally fringed with papyrus; but inside the papyrus are open spaces, with
occasional bits of sandy soil, or traces of old beach line, and these are sparsely infested
by tsetse. The interior is much of it “ impenetrable ” jungle, cut with great numbers
of hippo trails, and with clearings, closely cropped by hippo, where once were villages

386 W. F. FISKE.

and plantations. Not far from the northern extremity are natural clearings, with
very short grass and flat outcropping ledges of rock, and where the soil is
sufficient, exceedingly dense clumps and thickets of bushes and vines.

Everywhere on the peninsula the inland range of fly appeared to be normal, except
in these natural clearings, which extended a mile or more and were distant 800 to
1,200 yards from the water. Catches were attempted in the clearings kept open
by hippo, but without result, or only straggling flies were caught ; but in the natural
clearings the greatest density of fly was found of any point on the peninsula. The
records are

Male Female
Density. Percentage.

Points along shore of peninsula opposite clearmgs 59 .. 42°3
Point nearest clearing on foreshore... My. OO ase Eile
In clear space kept open by hippo 100 yards

inland opposite natural clearings .. oe Gey A sD
In hippo trail leading to clearing 400 yardsinland 100 .. 12:0
In clearing, 800 to 1,000 yards inland .. ac RUM vee slid

In open space 300 to 1,500 yards beyond clearing nil Wo ks

Breeding grounds were sought but not found ; but this is no proof that they may
not have existed. The shelter was of a most attractive type, but no more so than
in other clearings, kept open by hippo, where the soil was much deeper and the
vegetation more luxuriant. The only explanation for the presence of fly in such
unusual numbers—for, in fact, the major concentration of fly on the peninsula so
far as the survey of 1t extended—was the excellent shelter (probably coupled with
breeding grounds) and rather unusual numbers of bush-buck, which appeared to be
virtually the only source of food. It was less the numbers, however, than the easy
accessibility of these animals which was conceived to be mainly responsible. They
found refuge in the dense thickets and clumps of bush, which were so small in extent
that the flies could find them at all times without penetrating far into the shadow.

Kitebo Peninsula, August 1915.

Kitebo Peninsula in the district of Mawakota, is bounded by permanent sudd fields,
except for a reach of about 3 miles along its south-eastern and southern shore.
Here it is narrowly-fringed with sudd of more recent growth and less permanently
lodged, which in 1915 was broken at three points. Hach break was the centre of
a colony, or centre of infestation, for the space between the breaks and on either
side of the semi-open reach.

To the northward, behind the permanent sudd fields that fill the channel between
the peninsula and the island of Bujako, infestation falls away abruptly. To the
westward, behind the permanent sudd fields that bind the western shore and south-
western extremity of the peninsula a curious situation was discovered.

The south-western extremity of the peninsula is a marshy plain with belts and
patches of drier land. The wetter parts of it are overgrown with a terrible tangle
of tall grass, shrubbery and briery vines that defies penetration. The drier parts
are in places open game pasture, and in part covered with patches of dense jungle.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 387

The whole area forms a triangle, with its base against the higher ground (densely
forested) and its apex lost in the tangled mass of vegetation that stretches outwards
into the sudd.

Crossing this area, some distance from and very roughly parallel to the forested
base of it, is a sandy ridge, densely overgrown with shrubbery and jungle. The
tangled marsh is beyond this ridge, and inside it are the game pastures, very closely
cropped by hippo and bush-buck. (Hippo crop almost as closely as sheep, using
their horny-edged lips.) The ridge is an ancient beach line and affords fair breeding
grounds and adequate shelter, which is supplemented by the patches of jungle
scattered over the game pasture.

In addition to bush-buck, which were unusually common, were plenty of more or
less recent tracks of water-buck, bush-pig and buffalo, which came out from the
forest, and from other game pastures 2,000 to 3.000 yards through the forest.

The infestation along the ridge was interesting :—

Male Female
Catch. Density. Ratio.

At break in sudd, and hippo landing 195 44:5 a eee

200 yards from open water, on ridge 114 29-0 49°1%,

700 ., : . 64 20:0 BY abs
1,100 ip e : BD 20:0 27-29/,
1,500 f 4 17 6-5 93°59,

There could be little doubt, in this case, that the bush-buck were mainly responsible
for inland dispersion, for they greatly outnumbered the other game animals ;
of Varanus there was none, of crocodile but few, and the hippo herd slept far away
from its nightly pasture.

There was also infestation at open points in the forest at distances believed to be
2,500 yards from the lake, but it was slight and not accurately measured. At the
game pastures from 3,000 to 5,000 yards away from the open water—not far from the
western sudd field—no fly could be found; water-buck was the principal species
grazing in them.

V (f). Dispersion of Fly inland from Lake Shore along small Rivers.

At various points on the western shore of the Lake, in the district of Buddu, larger
streams or creeks than exist on the islands flow into it. One of these, Mujuzi, enters
the lake about two miles north of the old canoe landing at Kalkosa, and three miles
south of that at Bale. South of the creek all the way to Kalkosa the shore is marshy,
with an old beach line overgrown with dense vegetation. At the mouth of the
creek an old sand beach begins, at a level of some two to three feet above the present
beach line and continues nearly to Bale. The foreshore also is sandy, but except
for a reach of 600 to 800 yards just north of the creek, it is overgrown with dense
vegetation of an impenetrable character to the water’s edge.

At the mouth of the creek the breeding grounds and shelter are of the best,
and food is provided by a large number of crocodiles which find harbour there, and
which also rest in the sand. To the north there is some breeding ground, but

388 Wi Re RIGIKOR:

shelter is repulsive and no indication of any sort of food was seen. To the south
there are situtunga and perhaps other host animals, but shelter is repulsive and
no breeding grounds were seen. Conditions of life at the very mouth of the creek,
over a radius of hardly more than 50 yards, are highly favourable for tsetse,
but favourable conditions are extremely localised, making for a sharply defined
colony that fuses with a small colony at Kalkosa landing (point 40 in Fig. 5),
to the southward, but tapers away to a minimum of ‘5 at Bale (about 1,200 yards
beyond point 53), where a colony yet further north fuses narrowly with it.

Mujuzi Creek is almost a river, being at points some 10 yards in width and of
considerable depth. Its shores are thickly wooded, and cannot well be followed
because of the tangled vegetation. There are occasional pools with low sand bars,
as at points 45C and 45D. The stream was quite open as far as point 45H, where
it was choked with floating vegetation which extended for a considerable distance.
At points beyond 45F it was open again, and at one point ran through a tract of
woodland which seemed the headquarters of a herd of buffalo, but beyond point
45F, which was approximately 2 miles from its mouth, no fly was seen. At no
point from the mouth inland was shelter especially attractive; no breeding grounds
were seen, and host animals were certainly not abundant.

If allowance is made for the general attractiveness of the shelter from point 45
to beyond point 46 along the lake shore, the falling off in density of fly appears
to be approximately the same on the lake shore northward to Bale, and along the
creek inland. It is more pronounced along the creek, but not very much more.
The conditions of life are unfavourable to tsetse in all three directions from the
colony centre, and infestations of the creek banks and lake shore alike for a distance
of from two to three miles is plainly due to the existence of exceptionally favourable
conditions in the extremely limited space around the mouth of the creek.

Hixtension of range inland along other creeks wide enough to create something
of an alley-way through the bush or forest for the moving flies to follow has been
found to be much as in this instance.* The flies follow the waterways much as they
follow the lake shcre. But as yet no colony centre, or semblance of such comparable
to the hill colony on Mbugwe Island (Sect. V (d) ) or the inland colony in Bujaju
(Sect. V (e) ) has been discovered on the banks of any stream.

Every bit of evidence that has been gathered supports this final conclusion con-
cerning the range of fly inland from the lake shore; it is primarily controlled by
the distribution of host animals, and there is nothing to indicate that this tsetse
requires open water or humid conditions, or that water is even attractive to tt.
Its favoured hosts are amphibious in habit; the most attractive types of shelter
occur more commonly near the shore than inland ; and it rarely finds good breeding
places elsewhere. :

It may be added that every instance in which fly was encountered other than
as mere stragglers at a distance greater than 500 yards from the lake shore has
been covered in this section.

*If the stream is so small that no break is caused in the continuity of the forest, fly
has not been observed to follow it.

Observa
Point

Observa e
Meee lhe
{a ’

Density
Male
Flies

Solid li
densi
along sh
Dotted
dens!

on Mujuzi Creek.

1e
er

fa Be
iy
a ;
ce
1

;

4
i re
- ee)
; Ate
y
F
> {i ‘ 4
BE
j
i
,
:
— i
a ss Bit ut

we
rer, oh
Ls if «

Observation
Point

Density

Observation
Point

Density

Density of
Male
Flies

Sekwe or
Solid line Kalkosa Landing
density

along shore.
Dotted line

density

\
Mujuzi Créek|
\
along creek

\
\
\
\
\
1
i)

Shore Line

Scale of Miles

Fig. S.

Fly survey of mainland shore from Kalkosa to Bale, Buddu, showing inland spread of fly on Mujuzi Creek.

we
¢
;
)
,
t
‘
,
re
‘
‘
‘
pain cl ety a
é 4
i A

ae) 1%, Cen

Hanis ‘

(

| cee wineemteter te ep LETTE

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 389

VY (g). The Sex Ratio at Points where Range is extended inland.

In the records of catches at inland points which have been given in the preceding
paragraphs a very significant feature is the falling off of female percentage coinci-
‘dently with the falling off of male density, as distance from the water point increases.

This is notable in the catches on Bugalla Island (Table XVIII), at Dumo Point,
on Bujaju Peninsula, and on Kitebo Peninsula (Sect. V (e) ). On Bujaju and also on
Manene Island the male density was greater inland, but the female percentage
‘strikingly lower. On Bukassa Island male density fell and female percentage
remained the same, with no serious exception. The Mbugwe hill-top colony is the
one striking exception, and this is also unique in being, apparently, quite independent
‘of colonies along the shore and entirely different in this respect from the infestations
at Dumo, Bujaju, Kitebo and on Bugalla Island, which are extensions inland of shore
colonies.

This same phenomenon has been noted repeatedly at other points on Bugalla
Island, where collections were made at distances of 500 to 2,000 yards inland, and the
whole series of observations falling into this category stand in the sharpest contrast
to the rule that applies along shore (so well illustrated by the curves in figure 4, p.
372) that female percentage rises as the distance from the centre of infestation
‘imcreases and male density decreases. And it stands, perhaps as the best bit of
other than purely empirical evidence, in proof of the effect of food supply or
‘distribution of host animals upon inland range of fly.

The shelter on Bugalla Island, for example, or at Kitebo, is in every respect
as good and attractive to the flies as on the foreshore. Food is better, and more
abundant. It is only the protection afforded by the breeding grounds that is
lacking to make the inland conditions of life entirely favourable.

TABLE XX.
Mujuzi Colony. Kitebo Colony.
Dispersion forced, by hunger, into an Dispersion induced (inland) by superior
intolerable environment. attraction of environment,
Distance of observa- Female | Distance of observa- Female
A Male : “| Male
tion point from colony density per- | tion point from colony) ee per-
centre. "| centage. eentre. | CenST'Y- | centage.
0 yards 98°0 45°3 0 yards 44°5 54:3
700." ;; 53°0 49°2 200K Fe 29-0 49:1
300): 555 20°5 64°0 700 ,, 20 0 SSeS)
E00 ~., Ons 2h TG ae 20°0 Of
ZIaOO . 55 10°0 80°8 T500t 6°5 Pe mS)
2,900 ,, a0) Ses

The fly cannot breed inland, and hold its own numerically, on account of lack
of good breeding places; but there is continual inland dispersion of flies from the
colony centres at the breeding places on the foreshore, induced by relatively better
or more food and equally attractive shelter. The flies are drawn away from the

390 W. F. FISKE.

colony centre, or centres of infestation, by more attractive conditions elsewhere.
In the colony at Lutoboka Bay (fig. 4) or at Mujuzi Creek (Sect. III (e) ) dispersion
is, on the other hand, forced, and the flies, instead of being drawn outward by a superior
attraction, are forced outwards into a relatively repulsive or intolerable environment
by the stimulus of hunger. In this latter case female percentage tends to soar
as male density falls; im the former male density and female percentage fall
together, as shown in Table XX. And it is certainly superior food in combination
with equally attractive shelter which explains inland dispersion in all the cases
that have come under observation, except inland from the lake along the banks of the

Mujuzi and a few other creeks, which is comparable to dispersion along the shore of
the lake.

VI. Hosts anp Host PREFERENCES OF GLOSSINA PALPALIS.

VI (a). Methods of Studying Hosts and Host Preferences.

These methods were used for the study of hosts and host preferences of Glossina
palpalis :—

(1) Experiment. The animals were tethered along fly-infested shore (see
Sect. VI (c)).

(2) Approach and observation. Wild and domestic animals were approached
and the behaviour of fly toward them was observed under entirely natural
conditions (see Sect. VI (d) ).

(3) Shooting. When possible, wild animals were shot in the fly belt at hours
when flies were active. If the animal falls without running the flies which

have been following may linger by it for a time, and some of them will
usually feed (see Sect. VI (e) ).

In addition to the above the study of sex ratio, as affected by the density of favoured
hosts, affords an opportunity for securing supplementary or confirmatory data
(see Sect. V (g) ).

A list of the principal animals that occur within the fly belt, together with methods.
by which they were studied as hosts of fly, and a summary of conclusions reached
is presented in Table XXI. The list excludes many small mammals, such as rats,
bats, etc., which though common are certainly of no importance as hosts of fiy, and
it also excludes others such as Hyraz, the “ edible rat,’”’ and others which are rare or
nocturnal and of no importance, in this region at least.

All animals which are fed upon by Glossina, and thus technically included amongst
its hosts, are not equally favoured or preferred by it, and a broad line of distinction
may be drawn between the two following categories :—

(A). Hosts which are positively attractive and which are favoured or preferred
to such an extent as to cause concentration of fly in localities that they
frequent.

(B). Hosts which are rather repulsive than attractive and which are avoided
to such an extent that the flies tend to scatter and disperse from localities.
in which no more favoured hosts occur.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. aol:

In numerous cases there is no doubt into which category an animal falls, and such
are designated by an exclamation point (!) in the table. In other cases there is con-
siderable doubt. These are indicated by an interrogation point (?). In one or two,
notably that of the domestic ox (see Sect. VI (c), Experiment 3), some individual
animals appear to be attractive and others repulsive.

There are two tests for distinguishing between attractive and repulsive hosts :—

(1) An attractive host is apt to collect a “followmg swarm” of fly. This
curious phenomenon is associated with the assembling of the sexes and is
described in Sect. IV.

(2) The range of fly—especially its range inland from the shore—is likely to
be notably extended through the presence of favoured hosts. This
phenomenon is discussed in Sect. V.

TABLE XXI.

Summary of Observations on Hosts and Host-Preference of Glossina palpalis
on Victoria Nyanza in Uganda.

Methods of Study. Conclusions.
Animals

observed in fly Approach Is animal |Is range of

belts. Experi- and Sout Is animal | a favoured |fly affected

ment. observa- cotms: || fed upon ? host? |by animal ?
tion.
| | |

Crocodile ..| Yes wilh Hees ..| Yes We eet aia Westic ic. i) Mes
Varanus ..| Yes wl MUER sale Les blll NAC Lat ctl MOR cons la Mega
Tortoise eo) ee eo ees ot NO ..l| Yes aa es Hole NO
Serpents :.| Ne oo Mes ..| No | No? .| No | No
Lizards . ae. No Sale Gs ..| No | No? .| No! No
Situtunga | NG ae fo wal nes sill ee EN Gwe ce alee Ses:
Hippo .. ..| No eo ae es Pl) Yes 2. Mes ..) Les?
Bush-buck ..| No ..| No .«| Yes LI NGEh Ala New ..| Yes??
Bush-pig =|) DO .2| Noe 2c ees pli INO? \ chcNo? Ves ae
Water-buck ..| No eee NO ..| Yes alt NO 2 NO? ..| No?
Buffalo . ..| Ne ..| No ..| No ..|| O Bur ae : ! No?
Duiker.. ..| No s shaNe ae. lea © spsille <e)-0 /.|,Nio.
Zebra .. De |, INO Syiic. ..| No hee ily O) la NOe
Wart-hog ..| No ..| No NG sO al @ .:| No.
Reed-buck ..| No style INO ..| No wcll £0 2a ING
Monkeys ..| No ..| Yes sos GOS ..|| No? ..| No! ..| No.
Otter .. ..| No spallenies ..| Yes ..|| No? eo BINGE
Mongoose J. WNo ..| Yes aa NO ..|| No? ..| No! .| No.
Hyaena ..| No ..| No Vala tlle) sthsO PB Co
Leopard ..| No edie ..| No «lf O ae) ub INIO:
Birds .. ..| No ..| Yes ..| No ..| No? ..| No! ..| No!
Man .| Yes Sal oes No oil -¥eslo (sci Ne! we Oe
Re 4. 2 «| Ses | wes ..| No .. Yes vay Ves ob No?
1 | Wes ..| No ..| Yes Tees seh ees eo] ese
Goat .. oo) Mes san es ..| No ..|| Yes ».| No! Tape:
Sheep .. ‘| Yes 2 NO ..| No “| Yes .-| No! ..| No.
Dog poe Ser ‘ES )es o No Yes ic| NO! it NGS

* Human presence affects range of fly negatively.
O. No evidence either way: animals rare or uncommonly seen by day.

392 W. F. FISKE.

Aenl

It is important to know into which category any given host species falls, if squarely
in either. It is certain that a repulsive host will not attract a “ following swarm,”
and notes on observations of such swarms in association with any particular host
are valuable in doubtful cases.

There are two host species, crocodile and Varanus, which form, apparently by them-
selves, a category of most favoured hosts. They are not only most attractive to fly, but
the most favoured and best protected breeding grounds of the fly are frequently identical
with either (a) spots selected by crocodiles as its own breeding grounds, or (b) spots.
selected by Varanus as a basking ground. (The female crocodiles habitually frequent
their breeding grounds, and bask over the spot where their eggs are buried.)

On this account Glossina palpalis becomes something like a specific parasite of
these reptiles, suggesting not distantly those domiciliary parasites that do not live
on the body but in the nests or domiciles of their hosts, and which are specifically

adapted to such a mode of lite.

VI (6). The Habits and Habitats of the Hosts of Glossina palpalis in the
Region of Victoria Nyanza.

The following very general notes and observations are drawn up with the idea of
indicating the practicability of controlling the fly through extermination of its hosts.
No proposition of this character should be considered, however, without also
considering the data presented in Sect. II (d).

Crocodile.

Crocodile is undoubtedly the most important host of Glossina palpalis in this.
region. It is common nearly everywhere, on islands and mainland alike, and on
or off rocky, marshy, clay-banked, sandy and gravelly shores. It is excessively
numerous at times along certain reaches of marshy or papyrus-fringed shore (notably
on certain floating islets north of the island of Bunjako where in August 1915 dozens.
and scores were seen basking together), but it cannot breed in such places. These:
are its feeding grounds, and individuals frequenting them must go, in some cases.
considerable distances, in search of the sandy or gravelly places in which alone they
deposit their eggs. Much frequented breeding grounds were found—notably on the
mainland in that district of Buddu known as the Swamba, and on certain islets,
notably that known as Dwavannu, which lies just off the southern tip of Luambu.
Island, where nearly a score of occupied “ nests’ were found in July 1915.

The females brood daily for long periods over the precise spot where their eggs.
lie buried, and at such times are especially exposed to attack by Glossina, and.
provide for it a regular supply of food.

Neither breeding nor hasking spots are very far from the shore; perhaps never
more than 100 and rarely more than 50 yards. In consequence inland range of the
fly is not affected by this host, as it may be by Varanus.

If it were deemed desirable to reduce the numbers of these reptiles, it could probably
be done by placing a small bounty on their eggs or by locating all the favoured breeding
places and making regular rounds of egg destruction. InGermanHast Africa, I was in-
formed, abounty had beenplaced on eggsasa measure for reducing the density of Glossina.
But such measures, to be really effective, ought to include the lake as a whole, for the
animals undoubtedly move about considerably in search of feeding and breeding grounds.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 293

They are not ordinarily hunted by the natives, who make no use of either flesh
or skin, and who, though they detest the animals, will not take the trouble to
destroy their eggs when found. Depopulation, the natives assert, has had no effect
on the numbers of crocodile infesting the islands. They are probably almost
unmitigated vermin. |
Varanus.

Varanus, or the monitor lizard, is second only to the crocodile in importance
as a host for Glossina palpalis. It is distributed on and near the rocky shores
of islands and mainland. It is much more at home on land than the crocodile,
and may wander so freely and to such considerable distances inland as to affect
the range of the fly notably (see Sect. V (d)).

Few animals enjoy a greater diversity ot purely animal food. Its excreta,
sometimes abundant in much-frequented basking grounds, show it to feed largely
on molluscs, both terrestrial and aquatic, and very freely on the fresh-water crabs
that often abound under stones in shallow water. An abundance of crabs is sure
to attract great numbers of Varanus, and nearly always, except on small and poorly
sheltered islets, an abundance of Varanus is correlated with abundance of tsetse.
But the reptile also feeds freely on insects, both aquatic and terrestrial, including
among the latter such unsavoury types as cockroaches and scavenger beetles ;
fish and other small vertebrates, including certainly snakes and frogs; the eggs
of birds, and also (so it is said) eggs of crocodile.

It domiciles itself in burrows which it excavates in sandy soil, in crevices of rocks
and, quite commonly, in termite nests.. Usually there is a well marked basking
spot near at hand, and there occur quite frequently the breeding places of tsetse.

Its own breeding places are variously described, and declared by some European
observers, positively, to be in old termite nests. But it appears more probable
that they are in large mounds of earth and debris thrown up by the animals, but
not watched over by them. These curious mounds are common, but of those
excavated none were found occupied. The tsetse is only by chance associated with
its host at these (assumed) breeding places, and there is no coincidence between
the location of them and the haunts or breeding places of fly.

On certain islands and islets, notably Manene and Dziru, Varanus was found
so commonly that an average of four or five would be flushed per hour of hunting.
Such abundance, the natives assert, was unheard of previous to the depopulation
of the islands, and undoubtedly the species has increased notably, and the tsetse
with it. They are so eagerly hunted by the natives, partly for their flesh, but
especially for their skins, that their numbers would certainly be reduced if the
islands were repopulated. The skins are so highly prized—for making musical
instruments—that it is doubtful if any reward or bounty would lead to the
destruction of much larger numbers.

Tortoise.

No species of tortoise (or of turtle) is at all commonly encountered in this region.
Therefore, though tortoise is freely fed upon and apparently a favoured host it is
of insignificant account in this particular region. Elsewhere it might be of prime
importance as a factor in Glossina bionomics.

394 W. F. FISKE.

Serpents.

Serpents are not commonly encountered, except, perhaps, on a few islets (notably
Tavu, where a cobra was unusually numerous) and on certain sandy plains much
frequented by a small earth-coloured snake. Of those observed, a python and a
pufi-adder seemed most likely to serve as hosts of tsetse, but both were so infre-
quently seen that no python and only one adder was observed in a fly-infested
locality at hours when the flies were active.

In view of the notable proclivity of this tsetse to feed {on reptiles, it pretty
certainly attacks serpents, but in this region they cannot be of any consequence

as hosts.

Small Inzards.

Small lizards of several species are very numerous in fly-infested localities, but
all evidence that they are ever fed upon is negative. Certainly they are of no
importance as hosts.

Situtunga.

Although usually accounted a rare species, this antelope 1s easily third in importance
among the hosts of Glossina palpalis in the region under consideration. On the
mainland its range is restricted by the leopard or human hunters to marshy areas,
either forested or overgrown with reeds, grass or papyrus, and to a zone of varying
width (maximum observed about 2 miles) surrounding these protected areas.

It is possibly unique (together with other “ species” of its subgenus) amongst
antelopes in that it swims well and takes voluntarily to the water. This permits
it to occupy islands where no other antelope occurs, and it is the only species occurring
on any of the islands visited, except those of the Mawakota group (Bunjako, Binga
and perhaps Bussi and Zinga), which lie near the western shore of the lake, and
are connected with the mainland by fields of floating sudd.

Except on these sudd-bound islands there are no leopards on any that have
been visited. Previous to the removal of the island population, in 1909, the antelope
was restricted in insular range to the uninhabited islands of Nkose and Mwanse, and
to the larger mhabited islands upon which large areas of marshy land afiorded
the protection required against leopard and human hunter alike. But following
depopulation it was free from destructive enemies and not only multiplied with
remarkable rapidity on the islands it formerly occupied (7.e., those with large
marshy areas), but extended its range to other islets and islands hitherto free

from it.*

*The following notes were made on distributions in 1915 :—

Damba Island. Originally infested, now densely infested.

Kome Island. Originally infested, now lightly infested. Spread (since 1909) from
Kome to Namba and Nsadzi Islands.

Bugalla-Buninga (Sesse) Island. Originally infested, now very densely infested.
Spread (since 1909) from Sesse to Fumve, Bubembe, Bunyama, Manene, Bufumira,
Bugaba and Mbugwe Islands.

Nkose Island. Originally and still densely infested.

Luambu Island. Newly infested from mainland.

Bunjako, Binga, Zinga, Bussi and other sudd-bound islands all originally infested.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS, 395

On these islands it has forsaken the unnecessary protection of the marshes, and
assumed habits much more like those of the bush-buck on the mainland. It is
most common in the jungles which have sprung up in the abandoned plantations
and village sites, and is generally more common on the highlands than in its original
habitat. At points it has browsed on shrubbery growing on precipitous rocky slopes
which would seem to afford it the scantiest foothold. Its elongate hoofs spread so
easily as not to impede rapid flight, if alarmed, even on rocky soil, and they are also
worn away nearly to the length that is normal for animals that frequent rocky places.

It has developed into a most destructive enemy of many species of plants. Some
of its most favoured food-plants, formerly very common, have been virtually
exterminated over wide areas. There is one plant, allied to the cultivated canna,
which was perhaps the most common and conspicuous of any forest herb. It formerly
covered the forest floor over extensive areas on Sesse Island, growing to a height
of 2 to 7 feet, and effectually concealing and protecting (in a crouching position)
the antelope that fed upon it. Hundreds of acres were covered with this plant
to this degree of density in the autumn of 1914, but a year later the plants had been
browsed to the roots nearly everywhere. Of one species of woodland shrub, formerly
very common, not a living shoot could be found in 1915. Of many species of trees,
vines and large shrubs not a living seedling, leaf or young shoot within reach of the
animals could be found. Areas formerly thickly massed with fern had been browsed
to the bare earth and roots. And when trees fell in the forest it happened repeatedly
that the animals prevented other trees from springing up, and the spaces thus became
open glades.

The animals on Sesse Island were rather badly afflicted as they grew older with a
mange or “ scab ” parasite, which seemed to be causing more trouble in 1915 than in
1914. On Damba Island this parasite did not occur. It appeared possible that it
might prevent much further increase on Sesse. If not, and if nothing else disastrous
supervenes, the animal must speedily adapt itself to the grazing habit, for it had
reached and passed far beyond the maximum density which the island would sustain
if it retained its woodland habit of browsing.

This “ outbreak” of situtunga presents so many points of interest that it was
carefully studied as a legitimate phase of these investigations, and much more was
learned of it than can well be included here.

The effect on the range of Glossina of the increasing numbers of situtunga has been
already described in Sect. V (c). The further effects are bound to be profound if
increase continues. There are certain islands (notably Bunyama) where density
of fly was well below the average in 1914-15, but where conditions are very favourable
to rapid increase of the very small herds of situtunga then domiciled upon them,
and also such as to make spectacular increase in density of fly almost a certainty
if the antelope increases.

The animal is assiduously hunted by the natives for flesh, skin and horns. As
already noted, they had either exterminated or prevented establishment of the species
on islands devoid of the specific protection afforded it by marshy tracts, and, in
time, they would pretty certainly exterminate it again if the islands were repopulated.
‘To exterminate it from the great fields of sudd which are permanently anchored

(659) D

396 W. F. FISKE.

in the more protected bays and channels of islands and mainland would doubtfully
be possible; nor would it be necessary, for these tracts are beyond the range of tsetse.

It can exist despite beasts of prey and European or native hunters in such natural
protection, and requires no additional legal protection to save it from extinction ;
and under such conditions it is harmless. But on the islands where leopards are absent,
it is absolutely to be classed as vermin wherever a human population exists or may
exist.

Hippopotamus.

Hippopotamus is generally distributed and, for an animal of its size, common
nearly everywhere on the island and mainland shores. It stands fourth in importance
among the hosts of Glossina, but is of much less importance than its numbers might
imply, because of its habits. There are isolated individuals and small herds (herds
of more than 15 or 20 are rarely encountered) which habitually sleep on land or with
their bodies partly exposed along shore during the hours of activity of the fly, but
these comprise only a small minority of the animals in this region. Those composing
the large majority habitually sleep in deep water, and are very rarely seen on land
or near enough to it to be reached by the fly during hours of its greater activity.

In a very few cases, where herds or individuals habitually repaired to some small
islet or isolated peninsula to sleep by day, the density of tsetse appeared to be greater
than would otherwise have been expected. But it is improbable that the effect
of the extermination of this animal on range and density of fly could be easily
measured or appreciated. At all events it is only those particular individuals or
herds that sleep on or close to shore which are of any account in this connection.

The animals have not increased in numbers according to some natives, and have
done so according to others, as a result of depopulation. They were formerly hunted,
not very successfully, for flesh and skin, and also as pests, for they may be terribly
destructive to plantations. On many islands it was necessary to protect plantations
from their ravages by deep ditches on the water side, and sometimes on all sides.

Probably the most pronounced effect produced by the presence and activities
of these animals on Glossina is that they keep open and closely cropped considerable
spaces on shore, especially in the old plantations. These open spaces or grazing
grounds are sometimes two miles or more inland, with hard beaten trails leading
to landing places on shore. They are also favoured haunts of situtunga on the
islands and of bush-buck on the mainland, and since they are usually in localities
where the soil is fertile, they are generally banked about with thickly massed
vegetation which affords ideal shelter for tsetse. Many points along shore, or inland
on the islands overrun with pig, situtunga or Varanus, are thus made attractive
to tsetse, which would otherwise be covered with vegetation of an unattractive or
positively repulsive type.

Bush-buck.

Bush-buck is common nearly everywhere along the mainland shore and on the
sudd-bound islands of Bunjako and Binga. It is by its habits and habitat much
less suitable than the situtunga as a host for tsetse, for its best natural protection
is found in the thick tangles of shrubbery and vines that grow in dry and rocky
places where sylvan or jungly vegetation will not flourish.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 397

Hividence concerning its status as a host of tsetse is conflicting ; possibly it is to
be ranked fifth in importance of the wild hosts in the general region.

It is hunted by the natives, but less assiduously than situtunga, and is the most
persistent of the antelopes, except situtunga, found in this region. It is said to have
increased considerably in the depopulated belt; certainly it has found favourable
conditions of life in many of the deserted plantations and village sites along shore.
Bush-pig.

The bush-pig ranges with bush-buck and is about equally common. Its preferred .
habitat is the jungles that grows in marshes which are not too wet, or on hills which
are not too dry, and which are its natural protection against leopard. It is by its
habitat no very suitable host for tsetse, but it may tie with bush-buck for the position
of fifth in relative importance.

It is assiduously hunted by the natives, both for its flesh, and as a major pest
of plantations. It is properly classed as vermin from every point of view, but as
long as its natural protection is provided, it will doubtless defy native and Huropean
hunters, and beasts of prey alike successfully.

The natives aver that its numbers have increased notably as a result of depopulation.

Water-buck. |
There is no water-buck on any island, but at certain points on the mainland it has

become very common as a result of depopulation. No evidence associating it with
tsetse was secured, but its numbers appear to be increasing rapidly in the depopulated
zone and it might easily become a host of some importance.

The natives positively assert that it was formerly unknown in certain localities—
the peninsula of Gova especially—where it is now common and increasing.

Zebra.

A few zebra are running with herds of water-buck on Buganga and Kiteba penin-
sulas. The natives assert (positively) that it was formerly unknown ; in fact, they
showed complete scepticism concerning its present existence on the shore, until
visually convinced.

Wart-hog.

Wart-hog has come into Buganga peninsula along with zebra and water-buck.
The natives assert that it was unknown before depopulation, but it is as yet very
rare, |

(The return of these animals to territory from which they had been exterminated
by the natives illustrates well the effect of native population on game, and indirectly,
perhaps on tsetse.)

Reed-buck.

Reed-buck is a rare species in or near the fly belt and finds its best natural
protection in vegetation that is positively repulsive to tsetse. It is of no consequence
whatever as a host.

Dunker.

A small duiker, of unknown species, was seen in some numbers on the peninsula
of Bunjako. Its habits keep it well away from tsetse, so far as observed.

(659)

398 W. F. FISKE.

; Buffalo.

Numerous herds of buffalo, said to be increasing, range over much of the main-
land shore, and others exist on the sudd-bound islands of Binga and Bunjako.
The animals come to the shore itself, but the herds keep well together and each
ranges over a considerable territory, so that despite the number of animals, they
do not afford an at all constant or regular supply of food for tsetse. A herd may
remain for several weeks without coming near the shore, and then for a week or
two range nowhere but along it.

It is believed that buffalo would be a favoured host of fly if it were less inclined
to wander, but no evidence whatever could be secured.

Leopard.

Leopard is of very great importance in the bionomics of tsetse as a deprivative
enemy, for it is a major factor in controlling the range and density of bush-buck,
situtunga, bush-pig and perhaps other hosts of the fly. As such it is discussed else-
where.

It is very doubtfully a host itself, for though common enough in the fly belt, it is
extremely seclusive by day.

It has probably increased with the game, and its range has been extended since
depopulation to include the islands of Bussi and Zinga. Formerly it occurred only
on the mainland and the sudd-bound islands of Bunjako and Binga. Bussi and Zinga
Islands were not completely sudd-bound, but were separated from the mainland by a
narrow open channel, used in canoe traffic. With cessation of this traffic these
waterways were allowed to become choked—permanently unless the Government
reopens them—and the leopards promptly crossed to the islands. The channel
between Bussi Island and the mainland was the first to close, and the leopards crossed
and increased to greater numbers (as evidenced by tracks and excreta) than in
any other district visited. The channel between Zinga and the mainland was not
known to be closed until in August 1915, when excreta of leopard were found on the
island—once only seen in some five days spent in fly survey—and on investigations
the channel was found choked.

Monkey.

The common Cercopithecus (‘ green’? monkey) occurs everywhere in the fly
belt on the mainland, and on most of the larger islands. On the islands, in the
absence of leopard and man, it is usually much more numerous than on the main-
land. It was hunted as a pest by the natives and is said by them to have increased.
An impression, not confirmed by actual study, was to the effect that the animals
on different islands were developing into fairly well marked local varieties.

The animal appears to be consistently avoided by tsetse, and despite the facts
that the flies are known to be pathogenic towards monkeys (of this species) on ©
several of the islands, and that the monkeys forage freely along beaches densely
infested by fly, they take no visible harm. Their numbers appear to be limited
only by the amount of food available.

Other monkeys occur in the fly belt on the mainland, but they are more strictly
arboreal. A lemur occurs on Bunyama and perhaps other islands.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 399

Small Mammals.

Otter is very common on the islands; mongoose on certain islands; “ edible
rats” on Sesse (Bugalla) and perhaps others; Hyrax on Sesse; civet cat on Sesse ;
rats (in excessive numbers, sometimes) on many islands; and one or two other
small mammals were observed. None are of the slightest account as hosts of

tsetse, so far as any evidence even remotely indicates.

Fruit bats, including a large “ flying fox,” are characteristic of the island fauna,

on account of the great quantities of wild figs that abound along rocky shores.
The smaller species sleep suspended from low bushes in such a manner as to expose
themselves to tsetse, but there is no shred of evidence that they are ever fed upon.
Birds.

A very conspicuous feature of the island fauna is the excessive abundance of shore
birds : cormorants of two or three species, and darters. They breed in low bush along
the shore and roost in great numbers on trees and (frequently) on low rocks, which
bring them within easy reach of tsetse. The flies must occasionally feed upon
these birds, but every bit of evidence which could be secured was to the effect that
they are of no importance whatever as hosts. They outnumber other large birds
occurring along the shore. i

These others include egret, heron, ibis, crested crane, open-billed stork, saddle-
billed stork, whale-headed stork (or shoe-bill—not uncommon in certain environ-
ments), Nile goose, spur-wing goose, and a few others which are potential hosts of
the fly. But none, apparently, are fed upon except, possibly, in emergency.

Domestic Animals.

Oxen and goats are occasionally to be seen herded on the borders of the fly belt,
as at Entebbe. Formerly oxen, goats, sheep and dogs were domesticated by the
islanders and, despite constant contact with tsetse, appear to have done very
well. There appear to be no cattle ticks on the islands at the present time, and
before depopulation not all the disease-bearing ticks were present, according to
available evidence. This and the absence of beasts of prey make the islands highly
favourable for breeding of live-stock. |

Domestic pigs were introduced to the islands by Europeans. <A few escaped into
the bush and were left when the islands were depopulated. They increased to
excessive numbers, which were latterly reduced through unknown causes—possibly
superabundance and famine. They have adapted themselves well to wild life, and
have habits very suggestive of the bush-pigs (not in the least like wart-hog) on
the mainland. But they would fare badly in open competition with the wild pigs,
for their efforts to uproot the earth are relatively puny and bush-pigs will secure
food where domestic pigs might starve.

On the one island where they occur abundantly (1914-15) they are favoured
hosts of the fly, and responsible for its notably wide dispersion, if not greater density.
They could doubtfully exist in the wild state in competition with bush-pig and might
not be able to escape leopard, but on ‘the islands they thrive so exceedingly well
in the bush that they are to be classed as vermin, and ought certainly not to be
introduced into any island unless strictly confined.

400 W. F. FISKE,

Sheep and goats are not at all favoured hosts of tsetse; the status of the domestic
ox is somewhat doubtful, but probably differs greatly from time to time and place
to place.

Man.

Previous to the sleeping sickness epidemic all the larger islands were permanently
inhabited, and fishing camps or temporary villages were located in many of the
smaller ones. The mainland was also well populated, and temporary fishing
villages or camps occupied by fishermen were located on reaches unfit for
permanent habitation. |

About 1900, the population of Buvuma Island and adjacent parts of the main-
land was decimated by famine, accompanied by the outbreaking epidemic of
sleeping sickness, which latterly spread into Buganda and the Sesse Islands, to
the westward. In 1907 a belt along the mainland shore was evacuated by the
natives at the advice of the Government, and two years later the islands
were similarly depopulated. Thereafter, except at the lake ports for steamers,
which were protected by local clearings, the whole fly belt was a proscribed zone,
into which few natives or Europeans were authorised to enter. This zone includes
all the islands (except the western shore of Bussi) and a two mile belt along the
mainland shore.

Notwithstanding this proscription, man continued to come into measurably broad
contact with tsetse-fly. Although the clearing at the open ports reduced density
of fly to a perfectly innocuous minimum, they did not exterminate it. With enough
patience I could always find tsetse in the Botanic Gardens and at certain other spots
in Entebbe. Density was very low—not more than 0-2 or 0°3 by the standard adopted
herein for comparative measurement—but considerable portions of Entebbe township
lay within the range of the species, and a large population lived in very narrow
contact with the fly.

At other open ports, notably at Bukakata, larger portions were included within
the fly belt ; density of fly was greater, and the local population came into broader
contact with it. At Bukakata the population was very small, but it lived very broadly
in contact with a not inconsiderable infestation.

Poaching within the proscribed area was, at points on the mainland, rather flagrant,
and some of these points were very badly infested. The reach of shore covered by
the northern spread of the Mujuzi colony (p. 387), which is the very worst infestation
discovered, was, strangely, a favourite haunt of poachers. No less than 22 discarded
or hidden fish-traps were found along some two miles of shore. Man was by no-
means a rare or uncommon host of tsetse at many points covered by the fly
survey.

The islands were much less freely visited by poachers, especially after my fly survey
was begun, but those nearer the mainland showed many traces of illicit occupation,
most of them however old.

Taking the mainland only, man was about as frequent a visitor in the fly belt as
water-buck or buffalo. ,

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. AOL

VI (c). Experiments with Animals under Constraint to determine Host
Preferences of Glossina palpalis.

1. Comparison between Goat, Pig, Varanus and Man.

This experiment was made on Lugazi Island in December 1913. Mention has
already been made of it (Sect. I (b)). The animals were exposed as equally as possible
along a bit of beach. The goats were native stock and full-grown; the pigs of
Huropean stock and only about one-quarter grown ; the Varanus were newly caught,
one being full-grown and the other about half. i

The bites inflicted upon man were counted for comparative purposes, but the
comparison is not exact, because only the flies actually engorging on the animals
were counted, and some bit without engorging. The goats appeared to have been
bitten several times, and each time prevented the fly from feeding. The complete
immunity of goats and man, in so far as engorgement is concerned, is due to the same
cause—an instinctive movement in self-protection which causes the fly to desist
from attack. Neither were immune from being bitten, and bites on goats were
probably about as numerous as on man.

TABLE XXII.

The Relative Attractiweness of Varanus, Goat, Pig and Man to Glossina palpalis.

Host. No. of hours exposed| No. of bites inflicted. |No. of bites per hour.
Varanus Pe q 24 60 2°50
2 lhc Bf us 32k 1 03
uGOab. ... eh a 76 0 “00 on
Man, African .. ies 38 I °03
» Kuropean bs 202 5 "02

Infestation by fly at the point where the experiment was made was :—Male density
6:0; female percentage 15:5.

The strong preference displayed by fly for Varanus over the other hosts was
confirmed in subsequent experiments. The results with pig proved untrustworthy.

It was later discovered (p. 381) that this animal is sometimes, at least, a favoured
host.

2. Comparison between Goat, Varanus and Crocodile.

Two goats, one Varanus and one crocodile were used in a short experiment on
Kimmi Island on the 27th January 1914. The crocodile was young and active,
and between 3 and 4 feet in length. The Varanus was slightly larger. All animals
were exposed as equally as possible. The crocodile first, and soon after the Varanus,
became intractable, apparently because of the swarms of flies which assailed them,
and the experiment had to be discontinued.

The infestation of this island at this time was:—Male density 24:3; female
percentage 57:5.

402 W. F. FISKE.

Taste XXIII.
The Relative Attractiveness of Varanus, Crocodile and Goat to Glossina palpalis.
Host. No. of hours exposed.| No. of bites inflicted.| No. of bites per hour,
Varanus 2 98 49-0
Crocodile 12 40 22°8
Goat .. 4 3 oy

The comparison between Varanus and goat is believed to be a fair one. That
between Varanus and Crocodile is not so fair, because of the small size and greater
activity of the crocodile. Both are favoured hosts, and about equally so.

3. Comparison between Ox, Sheep and Varanus.

This experiment was conducted in April 1915 on the Peninsula of Nubiru, on
the mainland. Two native sheep, both full-grown, were used, and three young
bulls about half-crown. One of these was an apparently pure-blooded Ankole
(Bos aegypticus) ; the second was of a strongly marked zebu or Indian type, with
fully developed hump and other characteristics ; the third had the outward appear-
ance of some nondescript European breed, and was probably a half-breed. Special
care was used to expose the various animals as equally as possible. The results
are summarised in Table XXIV.

TasLe XXIV.
The Relative Attractweness of Varanus, Sheep and Ox to Glossina palpalis.
Host. No. of hours exposed | No. of bites inflicted.| No. of bites per hour,

Varanus ss a4 4 | 29 ed
Ankole Bull.. iy. 1] 32 2°9
Indian?" 5,° 2 oy 11 if “1
Half-breed,,.. Se 11 20 ey
Sheep No. 1.. FS 11 1 1

Se OND ae a 11 0 0

Infestation by fly in this locality was, at this time :—Male density 5°3; female
percentage 41:2.

There were two extraordinary features of this experiment, one of which is brought
out in the table above: the immunity of the Indian bull to attack as compared with
the Ankole and half-breed. This was entirely due to his very excitable tempera-
ment. He was intractable, and perhaps a bit dangerous, and became almost as
excited under attack by Glossina as the sheep, which were even more intolerant of
fly than the goats used in previous experiments. The Ankole, on the other extreme,
was absolutely tractable and docile, and refused to become annoyed at the attack
of either Glossina or Stomoxys, and the half-breed was not much different. Both
these animals were covered with ticks, but the Indian would not permit even these
pests, and by remarkable contortions succeeded in freeing himself of them on nearly
every part of his body.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 403

The second notable feature of the experiment was the behaviour of fly toward
the bulls at different hours of the day. The experiment extended over three days.
Hach day the morning was dull and rainy and each day the animals were taken
to the shore as soon as the clouds broke and while the vegetation was still wet.
As will be seen by Table XXV the bulls were fed upon much more frequently during
the first hours following the clearing, but the Varanus was somewhat more freely
attacked later on.

TABLE XX V.

Behaviour of Glossina towards Domestic Ox as affected by Presence of Tabanus.

Bites per hour inflicted
during first hours following| Bites per hour inflicted
Animals, clearing of weather; during later hours ;
Tabanus not active Tabanus active.
Ankole Bull te 11
Indian BS oo x0)
Half breed ,, 5:0 17
Varanus .. 50 74

The fact that the number of flies feeding on Varanus increased after the firs!
hour, instead of falling off sharply, proved conclusively that the flies were no less
active or less willing to feed, and the explanation for the sharp decline in number
of bites inflicted on the bulls is unquestionably due to the fact. that an hour or so
after the clouds broke and when the vegetation was quite dry a large species of
Tabanus appeared and caused the animals much annoyance. Not even the phleg-
matic Ankole would permit these great flies to approach it; the Glossina were
incidently included in the interdiction, and were effectually prevented from attacking
if there was any Tabanus about. But no Tabanus or any other flies than Glossina
attacked the Varanus.

These experiments were inconclusive in so far as proving the relative status of
the domestic ox as a host of Glossina palpalis ; but they conclusively showed
that where such animals are concerned freedom of attack by Glossina is to a
very large extent determined by the temperament of the individual, and also by
extraneous circumstances and conditions of time and place.

The experiments were continued over several days, more in an effort to find
some locality where there was tsetse but no Tabanus, and also in an effort to induce
a “following swarm” to collect on either an ox or Varanus, but without result.
The cattle found the Tabanus so annoying that it was necessary to constrain them to |
remain near the Lake shore at hours when these flies were active, and they were
generally distributed at all points along this reach of shore. The Varanus, also
proved intractable, and would not remain quiet in restraint. On this account,
it is believed, the “following swarm” (vide Sect. IV) refused to collect.

404. W. F. FISKE.

4. Comparison between Varanus and Man.

On 31st October 1913 a large Varanus was tethered at a point on the shore of
Bulago Island infested by Glossina. The morning was cool, and few flies were
active until about 11 o'clock, when the weather cleared. It was then noted that
the observer appeared rather more attractive to them than the Varanus.

From noon until. 1.30 the animal was left alone, and on returning to it, more than
30 (perhaps more than 50) flies were resting on or near it, and many more were
close at hand. Many were feeding, and about 100 were observed to engorge in
the course of the afternoon.

No fly bit the observer during this period, and it was estimated that 30 flies came
to the reptile to 1 that would come to him—in sharp contrast to the conditions
in the morning.

This was a true “following swarm,” and the only one that could ever be induced to
gather on an animal in captivity or under constraint. The failure of all other experi-
ments in this respect is believed, certainly, to be due to the uneasiness of the animal,
and it was long afterwards recalled that the animal in this particular experiment
had been quite severely wounded, and was disinclined to activity.

VI (d). Approach and Observation of Animals in a Natural Environment to
determine Hosts and Host Preferences of Glossina palpalis.*

Varanus.

Varanus were approached on a fair number of occasions, and ample confirmation
was secured of their attractiveness to tsetse as a host.

Crocodile.

Crocodile was never actually approached, for there are few animals more quick
to take alarm at the sound of an intruder, and at the first alarm a crocodile—
unlike many other animals—is sure to make off. But on several occasions when
they have been seen to slip quietly into the water, the vegetation near the spot
where they were resting has been found covered with engorged flies; on one
occasion in such numbers that the small bushes nearest at hand seemed thick
with red berries.

Si-utunga.

Attempts to approach situtunga were singularly successful. Many of the animals
on the islands had never seen man, and though the scent of him would stampede
them instantly, they were not, as a rule, at all alarmed at the sight of him if h>
remained motionless. Perhaps a dozen in all were successfully approached.

Notes on behaviour of fly toward this host have already been presented. The
behaviour of the host toward fly was rarely positive. They are phlegmatic, and very
bovine in their actions, only occasionally showing signs of annoyance by a move-
ment of the head, if flies are very persistent. An exception was a half-grown male,
ac-ompanying several adults, which displayed almost continual annoyance, with
Sfomoxys no less than with Glossina.

* See also notes on ‘ following swarm,” Sect. IV.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 405

No large Tabanids or other large biting flies, apart from Glossina and Siomoxys,
were ever seen on situtunga on the islands.

Tortoise.

Tortoise have twice been encountered on land in infested territory, and on each
occasion were accompanied by so many flies as to constitute a “ following swarm.”
The flies were seen to feed on both occasions, and in one instance the animal was
very plainly annoyed, frequently brushing the flies from its head with its foot.
The flies fed principally on the orbits, and the extent to which the subject had
suffered from their attack was witnessed by a whitish ring surrounding each eye
composed of the dried serum which is exuded in droplets by the flies while feeding.

Monkeys.

Troops of the Cercopithecus monkey common to the region have twice been
approached closely enough to observe the movements of flies in their vicinity.
The most successful attempt was on Bugalla (Sesse) Island in January 1915. A
considerable troop was deployed on the beach at a point where infestation was
heavy. FElies could be seen resting on rocks around which the monkeys were
foraging (for crabs), but none was seen to approach the animals, nor did the latter
show signs of being attacked.

When goats or sheep are driven into infested spots like this, enough flies will
approach to keep them on the gui vive. There is no doubt that this species of monkey
is avoided by fly yet more completely than sheep or goats.

Otter.

The mammalian life on the islands is so lacking in diversity (as compared with
the mainland) that otter, next to situtunga, hippopotamus and pig, is the most
logical warm-blooded host for tsetse. On three occasions otters were observed
on land freely exposed to attack by tsetse and under conditions which would certainly
have led to attack upon Varanus, but the flies paid no attention to them.

Mongoose.

A large mongoose is common on certain islands—notably Kitobo, where more
were seen in a week’s time than on all other islands together. It has a habit of
basking in the sun, selecting a spot where its ground-colour (much like dead leaves)
harmonises completely with its surroundings. By its habits it is freely exposed
to attacks of fly, but on several occasions when animals basking in fly-infested spots
have been flushed, no fly has been observed.

Domestic Anmals.

Cattle were more or less regularly, and goats occasionally, herded on the borders
of the fly belt near Hntebbe, and fly was occasionally seen on the former. But
it was especially noted (after the experiment with cattle at Mbiru described in
the preceding sub-section) that the presence of other flies—and others were usually
present in very much larger numbers than tsetse—reacted upon the cattle, and
their behaviour upon that of Glossina, to protect them measurably from attack.

406 W. F. FISKE.

It is one of the most important traits of Glossina to be, perhaps, the quickest
of all biting flies to detect and evade retaliatory movements on the part of its hosts.
There is no large biting fly known to me which does not possess this self-protecting
trait; otherwise stated, there is no large biting fly the behaviour of which I have
observed that will habitually press its attack upon an aroused and unwilling host
regardless of consequences to itself. Recently I have been studying the behaviour
of Tabanids of various species toward animals of many kinds in the Washington
Zoological Park, and while they are much more persistent than Glossina, they
clearly “know when they have had enough” and desist from pressing an attack
on an unwilling host. It is this same trait that is developed to a conspicuous degree
in Glossina palpalis which explains the great relative immunity of so many animals,
including probably man and monkey. It is doubtiul whether it is nearly so
strongly developed in Glossina morsitans.

Man.

Life in the fly belt affords a perpetual opportunity to observe and study behaviour
of fly toward the human host, and the important reactions of the host toward
the parasite.

The idea that the European is less freely attacked than the native is apparently
baseless, in so far as the original movement of the insect is concerned. The Huropean
is more quickly stimulated to retaliatory action—of a purely instinctive and almost
uncontrollable sort—than some natives, but there is almost or quite as much difference
between men as between the Indian and Ankole bulls used in the experiment at
Mbiru. Some of my porters or boys would react to the first approach of a tsetse,
and these the flies left severely alone, unless there was very great scarcity of other
food. Others of my canoe-men were, in contrast, extremely apathetic and phleg-
matic, and would not react to anything short of a severe bite,* and such as these
were very frequently bitten. Moreover men of phlegmatic temperament would
calmly compose themselves to sleep squarely in line with the stream of passing
flies along shore, rather than miss recall to the canoe for foraging expeditions, though
they might have escaped the fly by going a hundred or two yards inland; and
on one occasion something suggestive of the nucleus of a “following swarm”
was observed collected upon the sleeping forms of the lazier members of the canoe’s
crew who had thus exposed themselves to attack.

Next to sleeping men, men engaged in an absorbing occupation are most liable
to attack. I estimate to have been bitten 1,500 times in 18 months, and if from
this total is subtracted such bites as were inflicted in localities when both male
density and female percentage were high, and when the writing of field notes
would hardly be undertaken, it is probable that a majority of these bites were
inflicted while I was note writing. This was my only absorbing occupation in the
fly belt, but the men had many—repairing canoes; manufacture and repair of
fishing tackle; cleaning of fibre (for lines and cordage), etc. In such cases the
busiest—not, as above, the laziest—men came in for freest attack.

*The bite of G. palpalis may be absolutely painless, or it may be almost stinglike in
sharpness, or anywhere between.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA .PALPALIS, 407

But most frequently of all, men were bitten while paddling the canoe. If the
landing were infested, a number of flies—sometimes from badly infested landings
a dozen to twenty—would follow the canoe from shore, and once fairly away from
shore would be very loath to leave it. And these flies, seemingly because there
was nothing else to do, were remarkably prone to feed *—not at first, but some
time after the canoe had left the land. The work of paddling is purely mechanical
and rhythmic, and not in the least absorbing except for the steersman, but it is
of a sort that cannot well be interrupted to ward off a hovering fly. Consequently
the flies would be left undisturbed until one actually bit, when the paddler, if he
felt it, would interrupt his work to drive it away. The fly would then pass to
another man, and the process might be repeated many times. I have actually seen
five men bitten in quick succession by a single fly, and it is not at all easy to follow
the movements of a fly in a crowded canoe.

The engorged flies curiously pass forward in the canoe—against the wind of
passage—and will cluster on the prow of it. On one occasion, while rounding a
densely infested point of land on the Buddu shore and passing too closely in doing
so, a swarm of flies assailed the canoe, and in due course a cluster of eight, fully
engorged, had formed on the prow.

This rounding of fly-infested points too closely is even more conducive to attack
upon man than the departure from fly-infested landings, for if the canoe skirts
the shore for any considerable distance it may gather flies in large numbers, and
carry them out to sea.

I have not seen flies come to a canoe much, if any, more than a hundred yards
out, and not many will come to one over fifty yards out ; but at forty or fifty yards
a perfect swarm of flies may make a whirlwind descent.

All things considered, the hunter or forager is probably most immune to attack
by tsetse of any individual, for his senses are keyed up, and he is wide-awake and
conscious of little things without being physically absorbed in any occupation that
deters him from warding off a menacing fly. Certainly I have been bitten least
frequently while engaged in hunting or in patrolling the shore in fly survey work,
always intent on whatever might be taking place in the vicinity.

As noted of cattle, the relative susceptibility or immunity of man and of individual
men to attack is very largely a question of temperament, and of conditions and
circumstances of time and place—not forgetting to include among these the relative
abundance of other food than human blood.

VI (e). Hosts and Host Preferences of Glossina as determined by Flies
found on Animals shot in Fly Belts.

Crocodile—Island of Tavu, 13th September 1915.

A crocodile between 9 and 10 feet in length was shot in such manner as com-
pletely to paralyse the body but not to stop circulation of blood (these animals
will live and move the body vigorously for an hour or more after a brain shot; this

*It is a curious trait of Glossina morsitans, and probably also of palpalis, to feed in
confinement when they would not if unrestrained. Freshly captured, caged flies will

usually feed promptly on the body of an animal pressed against the wire screen, when
they would most certainly not have fed if uncaught,

408 W. F. FISKE.

one was shot through the neck vertebrae, and did not bleed, as when shot through +
the brain). “ Following swarm ’”’ present, estimated at over 100 flies. In course
of one hour 18 flies fed. No flies bit or annoyed the observer. Infestation at this
point :—Male density 41:0; female percentage 12:2.

Situtunga.—Island. of Sesse (Bugalla), 4th October 1914.

A nearly full-grown male was shot in an old plantation on a hill top about 1,000
yards from water at 5.30 p.m. Small “following swarm”’ of 22 flies present, of
which 6 fed. None annoyed observer. Infestations at this point :—Male density
12:0; female percentage 8-0.

Situtunga.—Island of Sesse (Bugalla), 10th October 1914.

Large male shot in old plantations about 500 yards from shore at 11.0 a.m.
Day dry, but overcast and conditions unsuitable for fly to be active. Following
swarm of from 30 to 40 flies, of which 11 fed. None annoyed observer. Average
number following observer through this region (while hunting) between 2 and 3.
Infestation at this point (taken after the weather had cleared and flies became
more active) :—Male density 15:3; female percentage 23°3.

Situtunga.—Dumo Point, Buddu, 13th June 1915.

Adult female shot about 2.30 p.m. in thick swamp of raphia palms. There were
3 flies following her. Distance 400 to 500 yards from shore, which is marshy and
thinly infested. Density at this pomt (4 boy-hours spent there) nil.

Bush Pig—Dumo Point, Buddu, 13th June 1915.
Adult male shot about 9 a.m., within 100 yards of spot where situtunga noted
above was shot in p.m. No fly.

Bush Pig—Dumo Point, Buddu, 17th June 1915. |

Adult female shot about 5.30 p.m. in guava thicket (mentioned on p. 385)
500 yards from Lake shore. No fly. Infestation at this poimt:—Male density
8:8; female percentage 21-4.

Altogether 9 bush-pigs were shot in or so near to fly belt that presence of fly
on them would have caused no surprise. But none were shot under more favourable
conditions than the two noted above. If a situtunga had been shot under the
conditions cited in the second case, absence of fly would have caused surprise. In
view of the infestation of this particular point, and of no visible cause for it except
the frequency and regularity with which it was visited by bush-pigs in search of
the falling guavas, considerable surprise was felt that no flies were following this
animal. Tue evidence concerning preference displayed by Glossina for this host is
conflicting, as in the case of bush-buck.

Bush-buck.—Island of Bunjako, 17th August 1915.

Three nearly full-grown males shot about 12.30 in old plantations about 200
yards from densely infested shore. No fly on animals. Infestation on shore :—
Male density 26°0; female percentage 29:1. Infestation at point where animals
were shot :—Density 45; females 39-2 per cent.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 409

Altogether about 10 or 12 bush-buck were shot under conditions which would
have made presence of fly explicable. In only one case was fly found,* but in no
case did its absence cause real surprise except in the one cited above. Had these
animals been situtunga, absence of fly upon them would have been extraordinarily
exceptional. The evidence concerning preference of Glossina palpalis for bush-
buck is completely conflicting. As in the case of bush-pig, dispersion of fly mland
from water indicates bush-buck to be a favoured host, but absence of fly on shot
animals indicates the contrary.

Water-buck.—Bugunga Peninsula, 15th August 1915.

Very large male shot about noon, grazing on edge of thicket about 200 yards from
shore. Good weather, and boys were catching fly near at hand at the time. No
fly on animal. Infestation at this point :—Male density 4:5; female percentage
25:0.

The evidence is negative, in view of the above notes on bush-buck and bush-pig.

Domestic Pig.—Island of Serinya, 15th September 1914.

Large fat female, one of two, shot about 10.30 a.m., 150 yards from shore, behind
dense fringe of reeds, in open forest. Day overcast, but clearing. Large following
swarm, considering conditions—perhaps 40 flies, of which many fed. Density at
this point very low, but tracks of pig followed backwards led to a densely infested
reach of shore 500 to 800 yards distant.

A number of pigs were shot on the Island of Bukassa under conditions almost
equally favourable, but no following swarm was observed on them. In this case
on Serinya the evidence secured in this manner confirms precisely the evidence
presented in Sect. V (c) leaving no doubt as to the status of domestic pig as a
favoured host—at least under certain conditions.

Hippopotamus.—Island of Buninga (Sesse), 26th July 1915.

Adult female shot about 9 a.m. It was apparently sleeping on the shore, and,
aroused by approach of canoe, was cut off from the water. Body fell on land, less
than 100 yards from the point where shot. Large following swarm (estimated at
125 flies), of which many (estimated at 30 flies) fed. They fed on eyelids, nose,
feet and belly, and engorged freely despite the thickness of the skin—evidently
on blood in the skin, if this is possible. (It occurred to me afterwards that though
gorged or partly gorged flies were seen apparently feeding on the legs and belly,
no flies were actually observed engorging except on the eyelids and nose.) Infesta-
tion at this point :—Male density 26:0; female percentage 22:0.

This is the one instance in which a hippo was shot (a) on land, (b) without plunging
into water, (c) in a fly-infested locality, and (d) at an hour when fly was active.
The conditions are not easy to meet, but this single case fixes the status of hippo
as a favoured host.

*In this case there was but one fly and it is not certain that it was attracted to the
antelope.

410 W. F. FISKE.

VI (f). Long-shore Range and Density of Fly as determined by Food Supply on

Victoria Nyanza.

Glossina must, of course, have food, and no other food than the blood of vertebrates
is required by flies in confinement. If any other sort is ever taken, evidence or
indication that it is required by the flies is entirely lacking.

Evidence is abundant and conclusive that the four most common of the large
shore-loving or amphibious animals found in this region are all favoured hosts
(crocodile, Varanus, situtunga and hippo). Evidence with respect to common
game animals of the mainland (bush-buck, bush-pig, water-buck and buffalo) is
negative or conflicting, but it would probably be a conservative estimate that the
four favoured hosts first mentioned provide 95 per cent. of the food taken by Glossina
palpalis on the mainland and 98 or 99 per cent. of its food on the islands (excepting
the islands of Bunjako and Binga, where bush-buck, etc., occur, and Bukassa, where
pig is a factor).

If this is correct, it might be assumed that complete extermination of the four
principle hosts would—in the absence of man and domestic animals—cause reduc-
tion in density of infestation by 95 to 99 per cent. I think this would undoubtedly
follow, if the extermination of the hosts was complete, for there are several reaches
of mainland shore in the district of Buddu where absence of all host animals of favoured
species from points within range or reach of tsetse from good shelter and good
breeding places is specifically correlated with absence of fly. The best case of this
sort is illustrated by figure 6 and the explanatory note accompanying it. Another
striking case is the Mujuzi colony, which is remarkable in so many ways, and which
is illustrated by fig. 5 (p. 388). This colony centres conspicuously at the point
(45 in graph) where alone food is plentiful. North of the colony centre (points
46 and 47) there are good breeding grounds and good shelter, but the falling off
in density, coupled with the uniquely high female percentage (p. 373), left no
doubt that absence of food explained scarcity of fly away from the colony
centre.

It is absolutely necessary, however, that adequate protection in the form of
both shelter and breeding grounds shall be provided for the insect within reach
of food ; otherwise the most abundant supply of the most preferred food is valueless
to the species. These conditions occur in many localities, and are described in some
detail in discussions on the effect of shelter and breeding grounds on range and
density of fly. Small islets, like that south of Dziru or that east of Bubembe
(p. 429), may fairly swarm with crocodile and Varanus, and, in addition, provide
breeding places of the very best, but if devoid of adequate massive shelter they are
fly free, or infested only by stragglers from elsewhere.

It is wholly impracticable to attempt a presentation of general data concerning
environmental conditions at the many points on the mainland and on the more
than 70 islands that were wholly or in part surveyed. (It would require not less
that 150 pages of descriptive matter apart from tables and graphs.) The selected
cases cited in various connections must suffice, together with the conclusions drawn
from a thorough study of original data in the field (November-December 1914),
followed by extensive fly surveys (Sesse Islands and the Buddu and Mawakota shores)

i 3

a ee

eae shinies

ye
‘i
i.

bh)

Observation
Point

Density of

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. All

designed in part to confirm conclusions drawn from co-ordination of data collected
in the course of the Island survey, viz. :—

I. If protection (shelter and breeding ground) is more than adequate, range
and density of Glossina palpalis are controlled by quantity of food of
preferred sorts occurring within range or reach of flies from both kinds
of protection required (see fig. 6 and note following).

II. Food, even of the most favoured sorts, is valueless to the species unless
protection (of both sorts, shelter and breeding grounds) occurs within
range or reach of individual flies from it.

III. The distance separating food from protection (as it may be measured
by yards) is a factor of the very first importance in determining the
prevailing degree of infestation by the species.

VI (g). Effect of Crocodile on Range and Density of Fiy along Shore.

The curve in the accompanying graph (fig. 6) illustrates the variations in local
density of Glossina palpalis along a reach of four miles of shore which was surveyed
in May and June 1915. |

The foreshore at these points was an open sandy beach, continuous except at
observation points 5 and 9, where it was cut by small creeks. Immediately behind
the beach was a belt of open grass-land with a few scattered bushes—not enough
to afford shelter for fly. The soil was beach sand, and this open belt or natural
clearing was from 40 to 100 yards in width. Behind it was thick bush, jungle
and forest, affording excellent shelter for fly; good breeding ground was almost
continuous along the border of the woodland from point 1 to point 9.

Food only was lacking, There were tracks of hippo, bush-buck, bush-pig, water-
buck and buffalo in the open belt, but the animals were not seen; their spoor was
not fresh, and they were obviously only occasional visitors. Crocodiles and the
fresh spoor of crocodiles were seen along the water’s edge, but they were effectually
eut off from shelter by the open belt, and beyond reach of flies.

Only at points 5 and 9, where the creeks cut through the open belt and sand
beach, did the crocodiles come into range or reach of flies from shelter. At these
points they found harbour in the mouths of the creeks, and basked on their banks
within easy reach of the flies from shelter, and here, as the graph clearly shows,
two small semi-distinct colonies of fly existed. There could be no doubt that the
existence of these colonies was due to the presence of a regular supply of food,
represented by the crocodiles, within range or reach of the flies from shelter,

VII. Taz BREEDING GROUNDS OF GLOSSINA PALPALIS.

It is a strongly marked characteristic of Glossina palpalis to seek specifically pro-
tected spots wherein to deposit its larvae, and extraordinary accumulations of puparia
and the empty shells of them (from which adults have issued) may be found in
especially attractive situations.

To what extent the species is dependent on breeding grounds of sorts easily
found, and recognisable as such, was unknown, and it was one of the principle objects
of these investigations to secure positive information on this point, and in the hope
that it would prove practicable to control the density of the insect by seeking and
destroying its specific breeding places.

(659) E

Observation
ont

Density of
Fly

Shore Line

10 WA 2 Ye 3
Scale of Miles Creek with Crocodiles Creek with Crocodiles

Fig.6. Fly survey of a reach of shore in the district of Bwendi, Saza of Buddu, showing effect of Crocodile
on the range and density of Glossina palpalis.

412 W. F. FISKE.

VII (a). Breeding Grounds on Sand and Gravel Beaches.

The larva of Glossina palpalis is nourished by the female until full-fed and ready
to pupate, and is then carried by her and deposited in protected situations known
as breeding places. The most attractive breeding grounds are the deposits of clean
dry beach sand or gravel that occur more or less frequently along the shore. They
must be in or near to the course followed by the streams of moving flies, 7.e., very
few puparia can be found in situations much off the course of fly traffic—and they
must be shaded.

Shade may be afforded by almost anything, provided only that it is not more
than two or three feet above the surface of the soil. If much higher the spot will
be avoided. Vegetation, both shrubby and herbaceous, rocks, logs, stumps, etc.,
will provide attractive shelter from the sun. The need of it is imperative, for the
larvae never penetrate more than two or three inches below the surface, and the sun
striking full on the surface will kill the puparia quickly. Even in the shaded, sand-
strewn caves which occur along the shores of certain islands, into which the sun
never penetrates, and which are sometimes attractive as breeding places, the flies
will only deposit their larvae around the edges, close under the overhanging rock,
as though fearful lest the rock farther above the surface of the soil might not
provide the continuous shelter required.

In especially attractive spots—if there are not too many of them—large deposits
of puparia and puparium shells may frequently be found. The record “ catch”
which was accurately counted was made in September 1914 on the island of Zinga, |
beneath an old stump thrown upon the beach by some storm, and overgrown with
“morning glory ” vines. It required one boy 35 minutes to sift the soil beneath this
bit of shelter. In a space hardly more than a yard square he found:

Empty shells ee .. 478 Rate per boy hour 814
Healthy puparia .. eo hy AE: i be 247
Total 622 e " 1061

Relatively few accurate counts of puparia and pupal shells were attempted,
for unless inordinate pains are taken, the figures mean little or nothing. There
may be much or little attractive breeding ground, and if much of it, puparia are
difficult to find—if little of it, they are more easy to find. The number of puparia
found at different seasons of the year means nothing, for the upspringing or withering
of vegetation affects the extent and attractiveness of breeding ground profoundly ;
and the same is true of the more or less regular fluctuations in level of the lake.
The accompanying table (Table X XVI) will indicate the numbers likely to be found
by expert fly boys under favourable conditions—as along the “ fly beaches ”’ (each
locality except that on Tavu Island was a “ fly beach”’).

For a time it was thought that data on the proportionate numbers of pupae in
different stages of their development might be worth collecting, but after a time
the idea was discarded. The healthy puparia were opened, and those showing
the grey colour of the adult were classed as “late stages,’ while those showing
no colour except on the eyes were classed as “early stages.” The data of this

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 413

character collected are, in part, presented in Table XXVII. They are mainly
interesting on account of the much lower percentage of late stages on the island
of Kimmi in March 1914 than during the previous January. In this specific case

TABLE XX VI.

Showing Number of Puparva of Glossina palpalis that may be found per Boy

Island.

Tavu
Wema..
Zinga ..
Damba.
Yempaita
Nsadzi

Date.

January 1914
February ,,
September ,,
HeDruUaTY: | jos

99 29

TABLE XXVII.

Hour in Attractive Breeding Grounds on Islands in Victoria Nyanza.

Find of Healthy Puparia.

Per boy per hour.
Gross.

Maxim. | Average,
086 268 65°1
775 196 34°2
144 247 =
og ral 39°3
445 211 49°3
133 46 16°8

Showing Variable Proportions of Pupae of Glossina in Late Stages of Develop-
ment, possibly indicative of a correspondingly Variable Rate of Reproduction.

Island.

Date. .

Wema..
Yempaita
Damba .
Nsadzi..
Tavu
Karambide
Mugogoya
Zinga ..
Namba .

392 2
Kimmi .

9°

Total and Average of total
Average by Islands

February 1914

33 29

93 29

él September 1915 uy

95 1914 ee

99 39

99 29
January m
March of
January ~
March bs

Finds of Puparia.

Per cent

Early Late :
Y of later
stages. | stages. stages,
1928 | 447 vee ee
267 157 oT Oy,
349 it 1679096
73 43 SOR
320 79 20°0 %
6 25 80°6 %

12 5 29°4 %
92 52 36°71 Y%
18 6 25°70 %
30 1 28°6 %
101 87 46°3 %
130 Oy a) ert

2626 989

* This is unduly high on account of the find on Karambide, where the high percentage
of late stages was due to the breeding grounds being recently abandoned by flies.

(695

414. We FISKE

especial care was taken to make the count and test a fair one, in connection with
the unusual conditions prevailing on the island in January. It was believed at
that time that on account of shortage of food the insects were unable to breed at
all freely, and that a high proportion of late stages of pupae would confirm this
fact. Surely enough, the proportion of late stages was abnormally high—and
surely enough, it was abnormally low in March when food had become more plentiful ;
but whether the data sustain the hypothesis is somewhat doubtful. A more complete
account of the experiment is given on page 451.

The proportion of living puparia to empty shells varies extremely, and variations
may be due to many causes. The normal proportions would appear to be about
1 to 5 (if there is a “ normal”’), but there are so many things which might explain
variations from it that data are of slight value, except perhaps in special cases.
The first citation in the accompanying table is such a special case. The find was made
at a spot where Varanus had habitually basked, but there were no recent signs
of the presence of the animal, and, with its abandonment of its favourite basking
spot, the flies ceased to find the spot attractive breeding ground.

Of somewhat greater interest is the percentage of empty shells which have been
eaten or destroyed by some predatory creature—ants, or beetle larvae—which extract
the contents without completely destroying the shell. The small percentage of eaten
shells and the large percentage from which the flies had issued, shown in Table XXIX,
is a clear indication of the remarkable degree of protection afforded by good breeding
places to the pupae of Glossina. The subject is discussed somewhat more fully

elsewhere (p. 435).
TaBLE XXVIII.

Showing Proportion of Inving Puparia of Glossina palpalis to Empty Shells
found on Islands in Victoria Nyanza.

Finds of puparia and empty shells.

Island. Date.
Puparia.| Empty | Total. |Per cent. of

shells. puparia.

Limnaiba e. ..| September 1914... 3 283 286 rie
Kiuwa.. ks =| Lebruary s.,) sit 6 456 462 13 %
Kimmi, Pt. 1*. &/2 \; i ist V7 1077 1094. 1:6 %
Damba, Pt. 1.. Sis So Be 39 1163 1202 a 206
Kimmi, Pt. 2*. et i 59 1412 1471 4:0 %
Dampa, Pi, 2.) iu: . ..| 797 | 4493°"| 5299 (|) aieiees
Nsadzi. . ae i as = ous 133 979 1412 1Lo ee
Mugogoya .. ..| September ,, 4S 17 85 102 16°7 %
Zinga .. ss Mt a = 2 144 478 622 el ee
Karambide .. wf nf 5 ve 31 719 110 28°2%
Tavu .. we a ay TOUS: ae 207 344 601 34°4 %
Wema.. ae ..| February 1914 ..| 1755 3044 4799 36°6 %
Total and Average ne ol) 3208." |, 13943) V7 151 18:7 5

* Point 1 on Kimmi Island was in a cave, where the empty shells might be expected
to persist indefinitely. Point 2 is the total for several finds elsewhere on the island.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 415

TABLE X XIX.

Showing Degree of Protection against Predatory Destructors provided for
pupae of Glossina palpalis by good Breeding Places.

Find of Empty Shells.
Locality.
Wema Island,

February 1914. Fly Destroyed by Total Percentage
emerged. predators. eae destroyed,

First Beach Le ut 46 0 46 070%

Second ,, my ae 46 5 51 OFS U6

Third ys iG ae 292 7 299 Zp Oe

Fourth ,, ty: Bs 318 9 327 28 Of,

Fifth 25 sis Si 213 3 216 14%

Sixth 35 ba Me 521 il 532 Pall MOB

Kighth ,, Sty “ie Lily 5 122 aed Gy

Total and average .. 1553 4.0 , 1593 27D

The need for the secure protection afforded by good breeding places in the case of
an insect like Glossina is fairly obvious. The adult has taken it upon itself to feed
and protect the larva, but the puparia are exposed to all the innumerable parasitic
and predatory destructors of Dipterous puparia in general, which are usually common
throughout the world. Owing to the fostering care lavished upon the larva,
very few young are produced—less than four per month per female—and it is
doubtful if the average longevity exceeds four or five months. About the
maximum possible rate of increase, therefore, would be 10-fold per generation,
as compared with anywhere from several hundred to several thousand fold potential
increase for Diptera generally. The species cannot withstand heavy mortality
im its pupal stages under such conditions. Secure protection for the puparia is
necessary, and the presence of the female in seeking localities for deposition of her
larvae where parasites and predatory destructors rarely penetrate is extremely
useful to the species. It is doubtful if any better protected places for breeding
exist (at all commonly) than those actually selected most freely, in dry, clean deposits
of beach sand or gravel. Except for the ants which wander over the surface, such
spots are positively avoided by nearly all insects except Glossina, and therefore by
the predatory and parasitic destructors of insects in general ;* and no parasitic or

*It is possibly significant that the most favoured breeding places of the tsetse and
of its principal host, the crocodile, are virtually identical. Much the largest deposits
of puparia have been found within a few yards of crocodile nests, in the same type of
soil, and under the same type of vegetation that serves partially to shade the female
crocodile as she “‘ broods”’ above her egg deposit.

An almost equally striking correlation between breeding places of tsetse in vegetable
debris and basking spots of Varanus, coupled with the finding of large deposits of
puparia in the sunning spots of situtunga, suggests that the specific or characteristic
preferences displayed by the flies for these hosts originated in the circumstance that
flies feeding on them were most likely to propagate (instead of the converse: that flies
propagating on sand beaches or dry vegetable debris would therefore be most likely
to feed on animals inhabiting the same localities.)

It is, indeed, most probable that the preference for certain hosts and for certain types

of breeding grounds, now developed into specific characteristics, originated coincidentally,
and together served to segregate Glossina palpalis from its congeners.

416 W. F. FISKE.

predatory entomophagid has attached itself to Glossina palpalis, as a “ specific
enemy,” to follow and prey upon it in the protection of its breeding grounds.

VII (6). Breeding Grounds in Vegetable Debris.

Quite early in these investigations a second and very distinct type of breeding
ground was discovered in fine dry vegetable debris—so dry and undecomposed
as to be as unattractive to insects in general as the deposits of clean-washed beach
sand or gravel which are most favoured breeding places.

These deposits of vegetable debris have been found rather sparingly both undor
and over rocks and over coarse gravel or pebbles, which preclude soil moisture irom
rising to dampen them and aid the growth of fungi and other saprophagous organisms.
If under rock, they must be in crevices so deep and well protected as not frequently
to be wetted by storms. If over rock or pebbles, they are most frequently shaded
by fern and composed of fragments of fern fronds and stems—material which does
not readily decompose or attract mycetophagous or saprophagous insects or the
destructors of insects.

TABLE XXX.

Finds of Shells and Puparia of Glossina palpalis on Breeding Grounds in
Vegetable Debris.

Finds of puparia. (4) Finds of empty shells.

| Per boy per | Per boy per
Island. Date. hour. hour,
Gross. | Gross.

_ Maxi- | Aver- | Maxi-| Aver-

| mum. | age. | mum, age.
Kiuwa (1)) Feb. 19, 1914.. 3 2 05 317 94 57°7
Namba (2)) Mars T, 1914... 2 4 —- 76 155 —=
Namba (3)| Mar. 17, 1914...) 12 6 16 173 50 24°7
Nsadzi (2)) Feb. 4, 1914..| 6 12 = 86 172 ae
Karambide . (3) Aug. 24, 1914..| 34 136 — 79 316 —
Mbugwe (3)| Sept. 1, 1914. . 8 24 — 88 358 =
Mugogoya .. (2) Sept. 4, 1914..| 24 48 — 85 170 —
Limnaiba (2)) eps 55, L014)... 3 6 — 283 576 =

(1) In very fine, dry debris under rock shelter.

(2) In basking spot of Varanus under fern or other low overhanging shelter.

(3) Under fern or other shelter, over pebble or cobble ; in each case a basking ground
of Varanus was near at hand. ;

(4) Compare maximum finds with finds from sunny and gravel breeding grounds
(Table XXVI, p. 413). The average finds per boy hour in most cases would be not
much greater than in the two cases given, for the breeding places are so small, as a rule,
that one boy will exhaust them in 15 minutes to half an hour in cases where the
breeding ground is specifically associated with a Varanus basking ground.

It is precisely in such spots that Varanus likes best to sun itself: over a rock or
beach of pebbles with its body partly shaded, as by over hanging fern, and partly

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. A417

exposed to the sun.* The same animals return daily to thesame places, and n moving
about break up and pack down dead stems, twigs, foliage, etc., to form a finely broken,
firm-surfaced mass of debris, right at the edge of low overhanging shelter. The result
is excellent breeding ground for fly, definitely associated with perhaps its most
favoured host, and very many puparia have frequently been found in such spot..
Some of the finds made under these conditions are cited in the accompanying Table.

These breeding grounds in vegetable debris are so few and restricted in extent
as compared with those in beach sand or gravel that it was for a time believed that
they play no part of measurable importance in the bionomics of Glossina. Later
on, as recounted in Sect. VII (c) following, 1t was concluded that they may play a
part of considerable importance locally, or under rather unusual conditions—unusual,
that is, for this particular region.

VII (c). Observations on Breeding Grounds found at a Distance from the
Lake Shore.

Many small, but very few large finds of puparia have been made at points above
the old beach line of 1906. The following are the more conspicuous cases.

Bugalla Island, February 1915.

On the bay side of Lutoboka peninsula, Bugalla (Sesse) Island, is a reach of
some 2,000 yards of very old beach line, about 100 to 200 yards back from the
beach line of 1906. The intervening space is open grass land, the soil being very
light and sandy. The forest rises abruptly behind the very old beach line, and
continues unbroken to the shore on the lake side of the peninsula. The infestation
along the present shore is heavy on the lake side and low on the bay side. The
infestation in the forest along the old beach line is heavy, owing to the great numbers
of situtunga.

At certain points along this old beach line in the border of the forest, where
the vegetation is dense, the antelope have made sunning places for themselves
by preventing new growth from springing up in openings left by falling trees. Two
such points were found, near together, about 200 yards from the foreshore on the
bay side and 400 yards from the shore on the lake side. The infestation was :—

Male Density. Female Percentage.

At shore on lake side . . ty 24-6 15°8
M » bayside .. ie Tw ie 58°8
At sunning spots of antelopes 25-0 16°6

In the sunning spots the sandy soil was laid bare by the hoofs of the antelope,
and was fine and dry. There were a few bits of it shaded by tangles of dead vines,
by fallen logs, or by tufted vegetation of sorts repugnant to the animals, and puparia
were found in large numbers. Exact counts were not made, but it was noted that
the breeding places were as attractive and as safe as any along the shore, and that
the deposits of puparia were larger than could be found at any point along the shore.

*It is perhaps worth noting that both crocodile and Varanus like best to bask, not
in the fuli sun, but with a part of the body shaded. The reason was learned through
use of tethered Varanus in feeding experiments. It was found that an hour or two of
exposure to the full effects of the sun would kill the cold-blooded reptile—more
quickly by far than similar exposure would kill a warm-blooded animal. Varanus
at least (and probably crocodile) is more susceptible to ‘‘ sun-stroke ’’ than man.

418 W. F. FISKE.

The conclusion was reached that under such conditions, with good food in abun-
dance, good shelter and good breeding ground, the fly would exist and thrive at
any distatice from the water. But it is a very rare combination at points beyond
the old beach line of 1906. Usually, when the soil is light and sandy back from the
water, it will not sustain vegetation sufficiently dense to serve for shelter, or for
protection of any great number of antelope. Had leopards been present on this
island the antelope could not have found protection from them in an environment
of this sort.

It was furthermore proved by experiment that except at points where the blanket
of vegetable mould that covered the sand on the old beach line had been ground
under the feet of the antelope and thoroughly dried by the sun, predatory destructors
of the puparia were present in exterminative numbers (see p. 443). The breeding
places were made by the antelope, which thus came to provide not only food but
protection to its parasites.

Bugalla Island, December 1914 and February 1915.

In December 1914 a remarkably heavy infestation was found on the very summit
of a high hill near the centre of the southernmost peninsula on Bugalla Island.
The catch of fly made at this point indicated a male density of 23:0 and female
percentage of 2‘5—but the record was discarded because the boys flushed a herd
of situtunga and caught the “ following swarm.” Probably a density of 12-0 would
be nearer the facts.

The presence of inland breeding grounds was suggested by the circumstance, and
by the further fact that infestation was heavier on the summit of the hill than on
its slopes. They were sought for at the time, and a special trip was made to the
locality in the February following to seek for them again, but without result.

This was the only inland point found on Bugalla Island, except the one above
noted, where density of infestation away from the water suggested the presence of
breeding places of sufficient extent to permit the fly to exist independently of breeding
grounds along shore.

Bujaju Peninsula, July 1915.
The presence of inland breeding grounds was strongly suggested by dispersion
of fly, as noted on page 385.

Kitebo Pennsula, August 1915.

Breeding grounds of fair quality were found on a ridge of sand representing an
ancient beach line at distances up to 1,100 yards from the water. The conditions
(see p. 386) suggested that the infestation at these points was mainly an extension
of a riparian colony of fly inland.

Damba Island, October 1915.

Breeding grounds were found on a very old beach line at probably 300 yards
from water, in a sunning spot of situtunga, much as noted on Bugalla Island. The
infestation appeared to be an extension inland of infestation along shore. Six
days were spent in exploration of Damba in search of inland colonies. Infestation
inland is general, and at points quite heavy, but at no point could evidence be
found indicative of any real centre of infestation with the possible exception of this one.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 419

Mbugwe Island, March 1915.

The one case in which what appeared certainly to be an independent inland
colony, and the only case of a colony independent of breeding grounds in sand or
gravel except those on tiny islets, is described in Sect. V (d). This is a unique and
important observation.

VII (d). The ‘Fly Beaches.’’

The very notable attractiveness of shaded deposits of beach sand or gravel to
the females of Glossina palpalis, coupled with the need of the species for protection
during the pupal stages and the high degree of protection provided by the most
attractive breeding places, suggested, naturally enough, that these deposits of sand
or gravel are requisite to the life of the species, and that a careful survey of the
region would reveal a close correlation between extent of breeding ground of this
particular character and prevailing local density of the fly.

Attention was first directed toward the “ fly beaches,” of which there are many
on island and mainland shores. They are merely open beaches of anything from
fine white sand to coarse brown gravel or small pebbles mingled with more or less
sand ; sometimes 600 to 1,000 yards or more in length, but usually less, and backed
up, first by low, bushy, semi-open shelter and secondly, by more massive shelter
such as may be provided by vine-clad cliffs, but which is nearly always provided by
forest.

When, as is usually the case, food is plentiful, this combination undoubtedly
provides the most favourable conditions of life for tsetse that occur in this region.
The combination of light backed by massive shelter is the most attractive of any
to the relatively idle male flies, and the combination of light shelter and masses of
beach sand or gravel affords the type of breeding ground most attractive to female
flies.

In consequence the maximum degree of density (measured always by density of male
flies—since activity of females is so variable a quantity) anywhere encountered in the
course of these investigations, was along such a beach (Zinga Island—Table XX X]).

But though the fly beaches afford the very best condition of life for Glossina
palpalis, other combinations are nearly as good. This is brought out by the table
accompanying (Table XX XI) in which are compared :—

I. Maximum degree of infestation (male density) encountered during fly survey
of islands in January and February 1914 along fly beaches, with maximum
infestation encountered elsewhere during same period.

II. Maximum degree of infestation encountered at any time during course
of investigations along fly beaches, with maximum infestation encountered
at any other points.

These data indicate the absolute maximum of density encountered in the course
of the investigations. The average density for the entire region was calculated on
the basis of data from the first 300 observation points to be very close to 12:5.

The density at fly beaches is seen to average something over 20-0 points higher
than at the most densely infested points other than fly beaches. At the same time
infestation elsewhere is often very heavy.

490 W. F. FISKE.

TABLE XX XI.

Comparison between Maximum Infestations by Glossina palpalis at Points on
Fly Beaches and ai other Points on Victoria Nyanza.

Infestation by Glossina.
Locality. Date.
Catch. Male Female
Density |Percentage.
“ \\17 {| Zinga Island .. ..| 28nd Sept. 1914 ..|. 206 | 1580 23°3 %
g | Ni Pt. Nov80;, Buddy ),..2) 3rd July 19 lisse 288 89°5 oto OG
cS) ee ie pei
S (| Wema Island .. 4) 95th Heb., 1914: eo eas 67-0 340 %
PA I { Yempaita Island . ||) 24th Peb., 1914. 377 59°6 3252794
tee Damba Island .. .;| d2th, Web, 1914 ae oge 55°8 67 %
| Nsadzi Island .. .:| 3rd Feb... 1914... .)_. 201 36°4. 16°71 %
eee ed ge Ge AE tts Aa 2a =| are

& \| yy {| Mujuzi Colony .. a) 24th Jume;400s 5 .| 25958 98°0 45°3 %
s \| Lutoboka Bay .. o-( ith Septelol4e iG." 0 kon 60°7 150 %
je} — — = = a a
a | Tavulsland .. a) Lothodans Hobo: be eA) Goes 104 %
Oa | Bulago Island .. -.| 22nd Jan, lola 271 42°4 19 :3%G
= {| .1| Kimmi Island .. aa | 20h Jan.., 1on4 yo) 366 38:0 53°38 %
© }| (| Kiuwa Island .. —..| 18th Feb., 1914 ..| 245 | 22-2 184 %

Total and averages for fly beaches 7 1734 Tie 25° 1.0%

Total and averages for other points 1820 54:1 271 oe

VII (e). Breeding Grounds on Old Beach Line.

In every case cited in Table XXXI of heavy infestation elsewhere than along
fly beaches the breeding grounds were found to be in sand or gravel deposits two or
three feet above the level of the lake at that time and anywhere from 10 to 100 yards.
back from the existing shore line. It was quickly obvious that the lake level had
fallen recently, and that the character of the shore had frequently undergone radical
change in consequence.

What had been a narrow sandy beach open to the lake and overhung by large trees
(a fly beach, in other words) had been left high and dry. and was separated from a
reed-grown, marshy foreshore by a more or less marshy and open belt, usually grass-
grown and closely cropped by grazing hippo. From the water there would be no.
indication of any open space, or of any sand beach, nor of anything but a bed of
reeds and forest to the back of it.

At cther points where low flat rocks overgrown with tangled vegetation marked
the existing shore line, considerable deposits of beach sand or gravel would be found
a little way inland, washed up by the waves upon the shelving rocks, left behind
by the receding waters, and hidden by bushy growth.

At yet other points these conditions were reversed, and what had been a rock-
bound shore had become a sand or gravel beach backed by a line of rocks hidden.
by vegetation.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 421

In one case what had been two small islands lying off the shore of a larger one had
been united by a low isthmus, and were connected with the larger island by an almost
continuous line of “ merinde”’ or ambatch trees growing in the shallowed water.

Many other radical changes in the character of the shore line were noted, and it
soon became evident that although the fly was to be found in maximum density along
the open fly beaches, the great mass or body of fly bred in the deposits of sand and
gravel marking the old beach line, usually hidden from sight.

Moreover in not a few cases it appeared certain that with the recession of the lake
conditions of life formerly intolerable for tsetse had become favourable to it in many
localities, and that in other localities the effect had been the opposite.

The question of lake level, and of fluctuations of it, was thus injected into the
general subject of tsetse bionomics as one of its more important phases. Data were
sought, and it was discovered that the old beach line marked the high lake-level
of 1906; that the existing low level was likely to be only temporary ; and that at
any time the waters might rise to the old level.

In consequence great changes and fluctuations in density of tsetse might be expected
to occur without other warning than rise or fall of the water. Eventually, in comparing
conditions on a certain reach of shore as they existed in 1915, with conditions as they
had been specifically described in old reports of the officers employed on Sleeping
Sickness Extended Investigations in 1906, it was found that changes in the degree
of infestation had been of the most extreme character. The Mujuzi colony, for
the most striking example, already mentioned several times, and last cited in Table
XX XI just preceding, was non-existent in 1906.

A more detailed account of the changes wrought in specific cases is presented in
Sect. XI.

In certain respects no factor in the bionomics of Glossina palpalis in this region
is more important than this of the fluctuations in lake level, as the breeding places
of the insect are affected. The rising waters uproot vegetation, clear the shore,
and wash clean the beaches. Falling waters leave ideal conditions of life for tsetse
where formerly they were hard or intolerable. Then the rapidly growing, tropical
vegetation closes in on the breeding places, gradually shutting them off from the
flies, until rising waters’ shall again open them to occupation.

VII (f). Correlation between Extent and Character of Breeding Grounds and
Degree of Infestation of small Islets by Glossina palpals.

A fly survey of small islets of less than one square mile in Victoria Nyanza was
begun in December 1913 and continued through parts of January, February, August
and September following. The object was to ascertain the extent to which the degree
of infestation is correlated with amount and character of breeding grounds as
represented by deposits of beach sand or gravel along either the old or new beach
line. First and last, some 50 islets were visited. Conditions with respect to extent
and character of breeding grounds were extremely diverse. On some islets no deposits
of beach sand or gravel: could be found—on others they were numerous and extensive.
The islets may be roughly assigned to five categories, accordingly as breeding grounds
of this type are “ very good,” “good,” “ poor,” “ very poor” or “ none,” and the
accompanying table (Table XXXII) has been framed to include a fairly represen-
tative number from each group.

499 W. F. FISKE.

TABLE XXXII.

Fly Survey of Small Islands in Victoria Nyanza showing Lack of Correlation
between Extent of Breeding Grounds and Degree of Infestation by Glossina

palpalis.
Infestation.
Breeding pty
Island. Date. Size (1).
x Grounds, Male Female
Density.) Ratio.
7, |
Pardes is ..(6)| Janz 1914 ..| B or C | Very Good..| 46:6 17 18%
Kimmi . As 2 ACopil irda, VOM A be Shae ae 5D°S (9G
Lukalu West.. ..(2)| Feb. 1914 B - 5 9°7 28°5 %
Islet S. of Dziru ..(2)| Mar. 1915 D > ‘9 a8 26°6 %
Lukalu East .. ..(2)| Feb. 1914 D o 33 0°0 =
Kukassu be ..(2)| Aug. 1914 D i - 00 a
Dwavannu .. ~(6)| July 1915 C Good 23°0 98%
Manene. te .-(6)| Mar, 1915 B Re: LET 12°74 %
Dyavadermi .. ..(2)| Mar. 1914 B or C 5 80 214,08
Kawari West . ..(2)| Dec. 1913 Cor D) |. .3 7°8 12°75 %
Ziro .. *) ..(3)| Feb. 1914 A 3 54 23°7 Of
Kirengi. in ..(3)| Feb, 1914 A te 3°8 24:7 Of
Kizima . an ..(2)| Jan, 1914 B or C 55 Le 2°5 LisOneF
Karambidi_ .. ..(4)| Aug, 1914 B Poor. 16°7 Li 7oSa
Namba. .. ..(6)| Jan. 1914 A - 9:4 79 %
Kakide . id . .(6)| Sept. 1914 D Bhd 8°0 —
Kibibi, . sit ..(6)| Sept. 1914 B aus 78 162 %
Darn 2. Ys ..(6)| Sept. 1914 B Adee es 13°74, 9%
Kawari East .. ..(6)| Dec. 1913 C or D re 4°5 47 %
Marida . se ..(3)| Feb. 1914 B or C a At LOS,
Mugogoya .. ..(4)| Sept. 1914 B Very Poor ae 44%
Lugazi . ial (47) Dees 19i3 C or D be * 6°0 la7 Se
Kisigalla a ..(6)| Feb. 1914 D 3 i, 17 —
Limnaiba .. . .(6)| Sept. 1914 B a 59 1°6 =
Sanga.. by ..(6)| Feb. 1914 D os 0:2 —:
Tenaga: *) ie ..(4)| Nov. 1913 D None 14°0 6°0ne,
DAT, on. sad ..(4)| Aug. 1915 C r 9°6 974 %
Dwanga ’Mto . ..(3)| Feb. 1914 B Re Ol —
Semanya ie ..(6)| Dec, 1913 C or D 5 Trace* —
Dwanga ’Mkur ..(3)} Feb. 1914 A i Trace* ==
Dwasengwe .. ..(3)| Feb. 1914 C . 0:0 nF
Sentette ne ..(6)| Dee. 1913 Cor D 9 0°0 al
Kadenga a ..(6)| Aug. 1914 C08 Dilabes 0:0 7e
Kaziru . oe ..(6)| Aug. 1914 Cor Dyin. 0-0 ay

* One fly caught possibly followed canoe to island.
(1). Size of islets: A 125 to 625 acres, roughly estimated.

B 25 29 125 99 9 99
C 5 39 25 39 99 99
D 1 5

(2). See Sect. VIII. The factor of shelter is paramount.

(3). See Sect. IX, notes on spiders. The factor of natural enemies enters in.

(4). See Sect. VII (h), following. The factor of insularity and dispersion of fly as
arbitrarily restricted on small islands enters in.

(5). See Seet. VII (7), following. Ths factor of food supply, coupled with that of
insularity and restriction of dispersion, enters.

(6). These islands are: (a) well sheltered, (b) without undue abundance of natural
enemies, (c) well provided with food, and (d) the factor of insularity appears not to. count
heavily or at all in its effect on infestation. They appear fairly to represent the effect
of breeding grounds in sand and gravel on density of tsetse.

ak
‘ ue }

‘tie Ares ae ee ee ca z° ce aes ‘ .
ath ire eae ah op doen a deren vag Taal erate sienteiiieae: eet er tl ep i RO Tip onl eR GN ae joni ainbdia

> PANE een ew seit
i :

‘ ? , : :
Me ‘ Sothee ee
sarc A arg: ae ey ar w ? — ” y i b~ s
4 } t, 5 3 ; i
4 ' 5
ie é
4 = .
Bi'st X 2 -
fa ui =a i o
rf } : F
; is Pa ae d
Praainn? reemeoe ath ny orien ae ee ne, eee ancl A “ a ee
gi 5 . :
; * f
.
f
¢ 4
eae ey “ie
Ais seth tien
” *
; i:
he 1
ie ‘ ae
FSD a Ls Nene
my Wy:
Fe 1
Pi
ah
.
J :
4,
t ;

x
=
Sat

3 = x Saee
Feb ipa annie aU Neely apie nes
can
‘ 1
.
rs
~
\
= a ‘
$ FJ
1
i

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 493

A certain correlation between extent or character of breeding grounds and degree
of infestation is at once apparent, but the lack of it is rather more conspicuous
than the presence of it. The scarcity of fly on the island of Lukalu West and the
islet south of Dziru, and its complete absence on the very small islets of Lukalu
Hast and Kukassu, notwithstanding the presence of good and extensive breeding
places, as well as the lightness of infestation on many of the islets falling in the second
group (Kizima, Kirengi, Ziro, etc.), is proof that something else that is requisite to
the life of the species is either deficient or lacking altogether.

Yet more significant is the relatively heavy infestation of islands like Karambidi,
Mugogoya, or most strikingly, Lula, on which breeding grounds of the sand or gravel
type are poor, very poor or lacking altogether.

The food supply was adequate or much more than adequate on virtually all these
islets, with the single exception of Kimmi. Both crocodile and Varanus were
common on nearly all the smallest islets, and particularly so on certain of them
that were least densely or not at all infested.

Many other factors were taken into account, but none seemed adequate to explain
the facts, and in August 1914 it was resolved to extend the fly survey to include
long reaches of shore on the larger islands and mainland on the theory that mere
insularity (v.e., excessive insularity of tiny islets of only a few acres in extent)
might prove the principle source of confusion. A fairly solid foundation for this
theory was subsequently found.

The first extensive reach of shore systematically surveyed was the circumference
of the island of Bugaba (fig. 7), a reach of some 164 miles. The results were
most encouraging. A summary follows.

The characters of the foreshore of the lake, and of the soil at points along the shore.
is indicated in the chart by letters, as follows.

Soils of types which will serve as breeding grounds :—
A. Brown beach gravel and sand.
B. White sand, mixed with coarse gravel or small pebbles.
C. Fine sandy soil, or loose sandy loam a little back from the beach line.
D. Pebbles mixed with a little sand and gravel.
E. Flat rock, with sand or gravel lodged in depressions.
F. Wet sandy beaches, overgrown with grass to water’s edge.
Soils and shores of types which do not serve as breeding grounds :—
G. Bold rocky shore, sometimes precipitous.
H. Rocky shore, neither very bold nor very flat.
I. Flat rocky shore (without sand or gravel).
J. Rock and marsh; or marsh with rocks protruding.
K. Clay banks, with or without marshy foreshore.
L. Marsh, with some floating vegetation (sudd).
Shelter immediately back from the fore-shore :—
M. Forest, or very thick bush.
N. Scattered bush, with open spaces. | Sooient shelter for fly.
O. eal fringe, with grass land behind.
P. Thin bush fringe, with grass land behind. .
Q. Open grass land to water's edge. [Insufficient SaieE ES,

It is to be noted that concentration of fly is apt to be greatest at a short distance
from the breeding ground.

At point 43 shelter at the breeding ground is insufficient and there is a double-
peaked colony, greatest concentration occurring on both sides, about equally.

Scale of Miles °

os
1 2 1 3 4 9 10 1 12 13 14 15 16 7
i V4. 25 2| 27 2 25 30 31 33/34 | 35 36 3 43 4/4 45 |46 47 48 !
Obsenverion was 1 2 3 5 | eel 3 Jo i 2 13 a B [38 | 39] 40/41 S| ESEU
Male F Fi . ~ 4 63/54 B5100I5 9 3 “0 io i 4 75 713 3:0 10-7 10 2:0 | 217 53/63 7 7 13 45 4 “1 53 12-7 7-6 140 40 2-5 15 “0 0
Density 0 ze pe Hs be & q sit He a fee i
aS | =I | 2 a 4 —
20 = ARE: ar
; |
el
10 i. “hs | +
}
5 ——— ifs a
mete | wi me _
Shore tine T 1 | q i U v Jad wdisil 6 Ic
Character of cl Gc ¢j¢ |x k K (Pee RS a RS a ER H H H H mb Hi] & i iL H HOF AIP Kk ik KHOR J] sq alk ka BS Se a K| Cc 4|
Character oF any iM m | ™ M N |N Q| M M M M m|™m M M M M M M M M M M Ce) o |e a ol] m M MNjo p MiP my Mm M
elter tt tt
Fig.7. Fly survey of Bugaba Island.
For explanations see text.

ADA W. F. FISKE.

At point 15, where there is a fine “ fly beach” with full southern exposure, the
trees grow so luxuriantly at the back of it as completely to shade it during the
summer months, when the sun is in the north. At points 16 and 17 the breeding
grounds are no better, but the sun strikes the beach and the flies tend to concentrate
at the border line between light and shade.

Similar explanations apply at other points.

VII (g). Gorrelation between Density of Infestation by Glossina pilpalis and
Location and Extent of Breeding Grounds in Beach Sand in Regions
where Food and Shelter are both more than adequate.

Bugaba was the first of the larger islands selected for exhaustive survey. It
lies near the centre of the Bugalla-Bukassa group, and in its general aspects
resembles those surrounding it.

Its area was roughly estimated at 4,500 acres, made up of (a) forest, rather more
than one-third ; (6) open grassland (formerly pasture), about one-third ; (c) jungle,
occupying sites of abandoned villages and plantations, rather less than one-third.

The forest fills the ravines, covers rocky and unarable slopes and ridges, and forms
an almost continuous belt along the shore. Thegrass land lies at the back of the forest
belt along the shore, and extends up the slopes to the borders of the old plantations.
The villages and plantations were located on a plateau that occupied the whole
central portion of the island. None were seen on the shore. There were three canoe
landings, with broad foot-paths leading to the villages and plantations.

The shore line was estimated at between 16 and 17 miles. It was sheltered by
forest for almost the entire distance. The southern and eastern shores, except
in the large southern bay, were generally bold and rocky—at points precipitous.
Elsewhere the shore was for the most part marshy, or with clay banks, but with
rock at many points. In the aggregate there was between 2 and 24 miles of sandy
or gravelly shore which might be accounted as potential breeding ground. This
was broken into numerous short reaches of from 20 to 100 yards in length, distributed
irregularly along the shore line.

Infestation by Glossina was limited to a narrow belt along shore. At no point
could flies be detected more than 400*yards inland, and at only one point could
any be found more than 200 yards inland.

The longshore infestation was light. The maximum male density at any of the
48 observations points was 26-0. The average for all pomts was 6-7—far below
the average for the lake region as a whole of approximately 12:5.

Food was plentiful. Crocodiles, Varanus and hippo were all common ; situtunga
had newly immigrated from Sesse (Bugalla) Island, and a few tracks were seen
at only one point. The female percentage was very low: 7:9 for the total catch
on the island.

The coincidence of infestation with the location and character of the sandy
or gravelly reaches that offered potential breeding ground was gratifyingly precise.
At or adjacent to every point on the shore where sandy or gravelly soil was exposed
to the surface (7.e., not concealed beneath grass or leaf-mould) and properly
shaded, the local density of fly exceeded the average. At no point not at or adjacent

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 425

to potential breeding ground was the average exceeded. At points closely adjacent
to breeding ground the shore was invariably infested, whatsoever its character—
whether flat or bold, rocky or marshy or clay bank. At no point exceeding 1}
miles distance from breeding ground did the density exceed 1:0. At various points
no flies were caught in from 2 to 5 boy-hours passed there. Absence or scarcity
of fly away from breeding ground was, like presence of it at points adjacent, in
no manner affected by the character of the shore, as long as it was devoid of sand
or gravel deposits.

In brief, the evidence, as far as it went, was entirely confirmatory of the theory
that on large islands, at least, the fly requires in addition to food and shelter the
protection provided for its pupae by sand or gravel deposits and by nothing else.

Finally, it was estimated that the clearing of bush or forest from about 45 acres,
or roughly 1 per cent. of the total or 4 per cent. of the arable area of the island,
would suffice virtually to exterminate the fly.

The data collected during this bit of survey work are presented in fig. 6 (p. 411).

VII (A). The Factor of insularity.

The fly survey of Bugaba Island disclosed the existence of semi-detached colonies
of flies, centring at or very near to the relatively few and short reaches of shore
that would serve as breeding grounds.

From these colony centres flies dispersed and infested reaches of shore on either
side along which it was plain that conditions of life were unfavourable to the existence
of the species, except as migrants from the better protected colony centres, or centres
of infestation.

It was very plain that the flies actually living in the zone of dispersion and beyond
the protected precincts could not perpetuate themselves unless the females returned
to the breeding grounds with their young. The chances that they may do this are
best nearest the breeding grounds, and decrease as distance from the breeding grounds
increases. Unless gifted with a “ homing” instinct they are as likely to seek breeding
grounds in the wrong as in the right direction. There is nothing whatever to indicate
or suggest that they have any such extraordinary prescience. :

On the contrary, the dispersion of fly from centres of infestation in protected precincts
is shown by the graph of the Bugaba colonies and others, to be not incomparable
to the dispersion of water welling from a spring in an arid region. At the colony
centre—t.e., within the protected precincts—the fly is perpetually increasing in
density, and as perpetually flowing or dispersing outward into unprotected territory,
where its numbers are perpetually decreasing ; much as water from a desert spring
is continually dispersing outwards into the arid territory surrounding, and as con-
tinually being lost by evaporation. There is a perfect natural balance between
outflow of water and loss through evaporation ; an increased outflow causes more
humid conditions near the source and extension of the irrigated area, until increased
evaporation compensates, and wice versa.

It is much the same with fly. An increased amount or degree of protection at
the colony centre, or anything else which is conducive to increased rate of

426 W. F. FISKE.

reproduction, leads to increased density of infestation at and near the source of it,
and to wider dispersion into the unprotected territory surrounding, until increasing
losses through dispersion compensate for increased rate of reproduction.

But if the dispersion of water at the desert spring were to be completely restricted
to a very much smaller area than it naturally irrigates, it would tend to accumulate,
and the “ balance” would be destroyed. And so with fly ; if dispersion of it from
protected precincts is arbitrarily restricted—as actually occurs on small islets—the
“balance” is destroyed, and almost anything may happen. And, indeed, almost
everything likely to happen in such circumstances actually occurs on the small
islets in Victoria Nyanza. The balance in the case of this species of insect is
indubitably stabilised by dispersion of it from protected precincts, where conditions
of life are favourable to numerical increase from generation to generation, into a
surrounding unprotected “zone of dispersion,” wherein conditions of life are
unfavourable to numerical increase.

The mechanism of the balance rarely fails to function smoothly, except when
dispersion is In some manner arbitrarily restricted. The islands afford the
commonest exceptions.

The islets devoid of sand or gravel breeding places, but yet infested by Glossina,
afford excellent proof of this conception of the “ balance” with Glossina palpalis,
and this conception provides the explanation for the occurrence of fly under these
conditions of extreme insularity.

The very best example is Lula islet, which is almost, or quite, the smallest in the
lake that is infested. It contains very good, but very small, breeding grounds in
vegetable debris. They are so small that they would certainly not serve as colony
centres if located on the shore of a large island, with no restriction upon dispersion
of fly from them. But on Lula dispersion is arbitrarily restricted to a space of less
than 5 acres, and to an extreme distance of, probably, less than 200 yards. The
flies cannot get so far from them that they cannot readily return to them, and in
whichsoever direction the flies seek breeding places for their young they are quickly
led to this particular spot.

If Lula were a promontory on Kome Island, where conditions along shore are much
the same, the infestation would be, as on Kome, extremely light or nil—the flies
would naturally disperse from so restricted a bit of protected area if they could.
But they are confined, as in a great breeding-cage, in a corner of which their young,
if deposited there, will find protection; and they live as flies in a breeding-cage
would live, not because the environment is pleasing or attractive, but because it
happens to be favourable to existence and because they cannot well escape it.

The same conditions prevail on Karambidi, Mugogoya, Lugazi, Sari and Limnaiba
Islands. All these, except Limnaiba, were infested to a higher degree than the
extent and character of the breeding grounds seemed to warrant, until this question
of insularity and restriction of dispersion from breeding grounds was considered.

On Limnaiba Island there was a very good, but very small, bit of breeding ground

found, and a very much less dense infestation than experience on Lula, Mugogoya, etc.,
would lead one to expect. Search in this breeding ground disclosed such numbers

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 427

proportionately, of hatched shells to whole puparia (see Table XXVIII, p. 414) as
to prove that a more dense infestation had recently existed ; and reduction in density
was explicable because of the fact that this particular spot also bore evidence of
having been recently abandoned as a basking ground by Varanus. A similar
instance was encountered on Karambidi, where in what plainly had been until very
recently the basking spot of Varanus had also been a favoured breeding spot of tsetse,
which had been very recently abandoned, as proved in this case by the most excep-
tional proportion of puparia in a late stage of development.

Of most particular interest in this connection are the circumstances surrounding
the occurrence of an island colony of tsetse, breeding exclusively in vegetable debris
on the very crest of a heavily forested hill. In this case, cited more at length in
Sect. V (d), isolation was produced by the forest, which is, like open water, an
obstruction to dispersion without being an insuperable obstacle to it.

It can safely be predicted that if any long reach of shore which provides good
shelter and plenty of food is found with breeding places in vegetable debris at
sufficiently frequent intervals, it will be found infested by fly ; for then flies dispersing
from one would find another. But such conditions had not been found at the close
of these investigations.

VIII. Tot SHELTER REQUIRED BY GLOSSINA PALPALIS.

Two points in the bionomics of Glossina palpalis are so conspicuous as to be noted
by every observer: its riparian habit, and its dependence on the shelter provided
by arborescent or rankly-growing herbaceous vegetation. It is rarely found far from
open water and never far from sheltering vegetation.

The effect of water upon range and density of the species appears to be indirect
and coincidental, as concluded in Sect. IV ; that of shelter is certainly not so. The
flies absolutely require the kind of protection it affords them, and the species could
not exist in the absence of it.

It was not expected that these studies would leadto any new discoveries concerning
the specific types of sheltering vegetation that provide necessary protection, but
it was taken too much for granted that where the flies were found in largest numbers
the shelter would be, at least, adequate. This proved not so; and it was discovered
that the shelter most attractive to the active flies, or that serves best to shade breeding
grounds and as protection for the puparia, is frequently inadequate as protection
for the inactive flies. Two kinds of shelter are requisite, (a) light, such as serves
at breeding grounds and for the active flies; and (6) massive, or forest-like, which
is required by the inactive flies.

In no other manner was it possible to explain the distribution of fly on the islands,
as disclosed by the island survey. The idea was made a working hypothesis, and
subsequent survey work sufficiently confirmed it, as recounted in the following pages.

VIII (a). Correlation between Character of Shelter and Infestation of small
Islands by Fly.
Reference to Table XX XII will show that though on certain small islands, notably
Lukalu East and Kukassu, infestation was nil despite the presence of “ very good”
(659) FP

428 W. F. FISKE.

breeding grounds. Food, also, was especially common on these islets, and the
only reason which could be assigned for absence of fly was the total lack of what
has been called “‘ massive shelter’ of the sort provided by large trees in masses,
or very heavy masses of shrubbery ; in short, by either true forest, or by a type
of vegetation approximating true forest.

If this were the true explanation it would serve also to account for lightness of
infestation on various other islands—notably Lukalu West and Kizima, which
possess relatively to such islands as Tavu, Kimmi, and nearly all of those of larger
size, but little massive shelter.

TaBLE XXXITI.

Showing Correlation between Character of Shelter and Infestation of Small
Islands by Glossina palpalis.

| |
‘Character of Shelter, | Infestation.

Sharacter of | |

Island. breeding 5
ound. | : : _ Male | Female
| Light. Massive. density. |percentage.

Opposed to the theory that massive shelter is necessary was the fact that the
active flies showed a strong preference for such light shelter as is provided by low
massed shrubbery with open spaces between thickets or clumps of it, masses or
clumps of rank-growing herbs, vine-covered bushes, or stumps, etc. This is the
kind of shelter that provides shade for the most attractive breeding places, and it
is indubitably most attractive to the active flies of both sexes.

Opposed to the theory also was the fact that such long, narrow sand-pits as
Crocodile Point on Bulago Island, which are devoid of massive shelter and semi-
detached from the main body of the island (or mainland), are apt to be densely
infested by fly. There is more and better massive shelter, and much better breeding
grounds, on the island of Kizima, for example, than on Crocodile Point, Bulago ;
yet the infestation}of Kizima, in January 1914, was only 2°5 as compared with 22°7
at Crocodile Point.

This last objection is met, however, by the facts disclosed in the catching experi-
ment on Crocodile Point cited in Sect. III (a), which proved conclusively enough

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 429

that the fly population was not permanently resident, but was continually moving,
and that virtually all flies infesting it at a given time were likely to move away
from it in the course of a day or two, to be replaced by others. On Kizima the
flies have not this privilege. They must remain on the islet and put up with whatever
shelter it affords, or leave it permanently. If Kizima were a peninsula of Bulago,
like Crocodile Point, its light shelter and breeding grounds, and the great quantity
of food it provides, would certainly attract to it even more flies than infest Crocodile
Poimt ; but it must be concluded that its heavy infestations under such hypothetical
conditions, ike the heavy infestation actually existing on Crocodile Point, would
be, as the other certainly is, due to the existence of good massive shelter within
range or reach of the flies infesting it.

These same facts concerning the movements of flies along shore, or along other
favourite courses, answer also the first objections made above. The active flies
are active with some positive object ; the females are seeking either food or breeding
places; the males are seeking either food or association with the females. Light
shelter provides the best shade for breeding places, and food (Varanus and crocodile
especially) is most frequently encountered associated with it; therefore it is sought
by the females and equally by the males during their hours of activity. At other
times (during the night, storms and dull weather generally), both sexes remain
in seclusion, and it is then that massive shelter is presumptively necessary for their
protection.

A large number of data collected during the course of the survey of small islands
sustained this as the true explanation for lightness or absence of infestation on a
fair number of them. The data are of a sort not easily presented, and Table XX XIII
is suggestive without being conclusive. Reliance in drawing conclusions was
principally placed in such comparative studies as between infestations at points
such as are illustrated in the sketches accompanying (fig. 8).

WY) ln

“Crocodile

Dziru
a

‘(00 yds. Island

ne Islet

Bulago
Island

Light Shelter & eos
Massive Shelter LMA breeding grownd _ Cas

Fig. 8. Illustrating the correlation between character of shelter and infestation of smail
islands by Glossina.

In each of the three cases we have an area of light shelter with good to very
good breeding grounds located in certain relations to massive shelter. These three
localities are roughly equal in gross area and attractiveness to fly; the two islets
being rather superior to the peninsula than otherwise. |

(659) F2

430 W. F. FISKE.

The first, at Crocodile Point, lies from 0 to 400 yards from massive shelter on
Bulago Island. Infestation was :—
Crocodile Pomt (Jan. 1914): male density 22:7; female percentage 42:9.
Near massive shelter on south shore of island: male density 20:4; female
percentage 40°6.

The second area of light shelter and breeding ground was on an islet south of
Dziru, and separated from massive shelter by an open channel of about 100 yards
in width. This served as a serious obstruction in free movement of fly. Infestation
in March 1915 was :—

Islet: male density 3:8; female percentage 26:6. Island: male density 5:0;
female percentage 44:4.

In September 1914 infestation at these same points had been :—

Islet: male density 1:5; female percentage 0. Island: male density 6:5;
female percentage 13°3.

The third area is the islet of Kukassu, which is more than 600 yards off the shore
of Bubambe Island, which is the nearest massive shelter. This width of the channel
is a complete obstruction to movement of fly. In consequence, infestation, in
August 1914, was found to be :—

Kukassu Islet: nul.
Bubambe Island (at point opposite): male density 13°7 ; female percentage 7°8.

These and other data of a similar character provided a weight of evidence that
was in the end conclusive; lightly sheltered areas may provide the best and most
attractive hunting, breeding and assembling grounds for Gilossina palpalis, but
unless they lie within easy range or reach of the flies from massive shelter they
will not be infested. If there is no obstacle to free movement of flies between such
areas and massively sheltered areas, they are apt to be more densely infested than
the other. But as distance or obstacles to free movement increase, infestation
diminishes, until finally beyond certain limits they cease to be infested.

VIII (6). Correlation between Type of Shelter and Density of Fly as disclosed
by Survey of large Islands and Mainland.

In the beginning it was expected that the survey of small islands and comparisons
between them with respect to environmental conditions and density of infestation
would be productive of the most valuable data. But it was discovered that insularity,
as it affected the normal dispersion of fly (p. 425), was so confusing a factor that the
data secured through the survey of long reaches of shore on the mainland and large
islands were much more satisfactory.

Especially is this true where the survey reveals, as on Bugaba Island (fig. 7, p. 423),
in the Bwendi District (fig. 6, p.411), at Dumo Point and Mujuzi Creek (fig. 5,
p. 388), etc., the existence of well defined, isolated or semi-isolated colonies of fly,
each with a sharply defined nucleus, from which flies are continually dispersing to
infest a zone surrounding.

The most interesting of all the colonies encountezed in the fly survey was certainly
that at Mujuzi Creek, ot which so much has been written, and which centres at a
point where food is provided in a region that is deficient in food. The second in

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS, 431

interest :s probably that on Lutoboka Bay, illustrated by figure 9, which centres
at a point where shelter is adequate in a region that is devoid of good shelter. Study
of it in September 1914, and at various subsequent visits, provided evidence deemed
absolutely conclusive concerning the need of massive shelter within easy range or
reach of Glossina palpalis. The colony (fig. 9) is seen to centre at the point where
the heavy shelter of a forest belt ceases, and light shelter of a type especially attractive
to active flies begins. There are breeding grounds at intervals all the way from
point 10 to point 17, but density falls away almost as abruptly in this direction as
in the opposite, where light shelter is lacking, and where a positively repulsive
belt of sudd (papyrus and saw-grass) fringes the shore.

Observation Fornt
6 i ae We I \4 paige 17

ake, ae eae ae

Li Grasisland Grassfand
Yi Lo ll be na nn xi ix x x x oo i ae

aides i silat eh cay er and breeding fncsitid Wltaba

600 yds. 750 ibid st a a 300 | 350 | 600 yds. bene
Points yds.| yds.

Fig. 9. Diagram showing effect of shelter on range and density of G. palpalis in
Lutoboka Bay.

Another interesting colony is that shown at points 40 to 46 in figure 7 (p. 423),
and also in another connection in figure 3 (p. 371). In this case a short reach of light
shelter is flanked on either side by massive shelter, and a double-peaked colony
results: 7.e., it is either two colonies fused broadly together, or one colony with two
nuclei.

432 W. F. FISKE.

An exceedingly interesting reach of shore is one of about five miles along the
southern margin of Buganga District in Buddu. Its western extremity is plunged
into a great sudd field. Beginning at a point where the sudd gives way and permits
landing, there is a reach of 200 yards (estimated) where an almost continuous line
of thick shrubbery and other low vegetation grows along a sand embankment which
represents the beach line of 1906. Inside this is a short grass, sandy plaim, with
large herds of water-buck and some zebra. Just outside it are many semi-open
spaces of a sort very attractive to fly, with most excellently shaded breeding grounds.
Crocodile and hippo are unusually numerous. (Points 1 to 6, inclusive, Table
XXXIV.) For the next 2,700 yards the shore becomes very marshy and no attractive
shelter could be detected from the water, until at point 8 a narrow ridge of sand
outside a belt of light shelter again appeared, continuing to point 9, where for the
first time massive shelter was encountered. Beyond this point the shore was diverse
and fairly heavily infested.

TABLE XXXIV.

Catches of Fly along Southern Shore of Buganga District, Buddu, showing Effect
of Massive Shelter on Density.

* Heavier density represents dispersion from point 9, which was a colony centre.

The point of exceptional interest, however, is No. 4, which was selected as that
one where the fringe of light shelter seemed to be the heaviest along this first section
of lightly sheltered shore. To all appearances such shelter as was present had sprung
up since the high lake-level of 1906, for it was principally growing on an embankment
that was the beach line at that time. Moreover, it appeared to be growing heavier
and denser in 1915; promising to become, in time, a forest fringe separating the
short grass plain from the lake. It had already become so heavy or massive that
it would have caused no surprise to have found the shore infested to an average
degree. The extreme lightness of the infestation was certainly due to lack of more
massive shelter, however, and could not be attributed to any other cause.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 433

But shelter was growing—or seemed to be growing—more massive with each
year that passed. It seemed altogether probable that in another year it would become
sufficient to protect Glossina adequately, and at point 4 the embryo of a future
colony appeared actually to exist. Despite the small size of it, it seemed to represent
a real independent colony, that would grow, as its protection grew, to become in a
few years just such a colony as occurs in the Bukakata district. The conditions of life
were almost identically the same in the two localities, except that along the reach of
shore in the Bukakata district the vegetation was a few degrees more massive,
and afforded that much better protection. The difference cannot be described,
and would be difficult to measure, but it represents the line of distinction between "
shelter that is adequate and shelter that is not adequate to protect the fly.

The main point is that light, low shelter, of the type that is most attractive to the
active flies, and that is most certainly to be associated in the mind of the casual
observer with the fly, is entirely insufficient to afford the species the protection
it requires. Massive or forest-like shelter must be had. It may not exist within
several thousand yards of a point where fly occurs, seemingly well satisfied with its
environment, but in such cases it-will certainly be possible to trace the source of
infestation to centres located close to the forest growth.

It may be added that a cliff overgrown with vines and shrubbery has been observed
to serve as massive shelter.
Reference back to figure 3 (p. 371) will show again the effect of absence of massive
shelter on density.
At points 40, 41, 45 and 46 massive shelter was good but light shelter absent
or lacking.
At point 43 light shelter was good but massive shelter lacking.
At points 42 and 44 the light and massive shelter were in juxtaposition, and
here were the centres of greatest concentration of fly.

VIII (c). Occurrence of Glossina palpalis behind Papyrus or Sudd.

At many points on islands and mainland narrow belts of floating vegetation
(usually papyrus) have formed along sheltered reaches of shore which appear to
have been open in 1906—or if not then, very recently. The old beach line of 1906,
not infrequently with good breeding grounds, lies behind the sudd belt, which is
traversed by passage ways kept open by hippo and crocodile. The crocodiles land
and sometimes breed in these places, which retain much of their original character.
Fly is commonly found, and in considerable density, in such situations. It would
be strange if it were not.

Much of the shore of the lake, especially in the large sheltered bays and channels
in the north-western limb of it, is permanently bound by much older fields and banks
of sudd. Inside of these the shore line has changed from its original character to
resemble the border of a marsh. Sand and gravel beaches, where they existed, have
usually been buried beneath humus or are densely overgrown with vegetation.

In July and August 1915 special effort was made to discover if fly were bred under
these conditions. The following surveys were made.

434 W. F. FISKE.

Bujaju Pemnsula, 30th July and 2nd August 1915.

A great bay (not shown on Whitehouse’s Chart) lying west of Bujaju Peninsula
is entirely sudd-bound, but with a central channel open. At several points hippo
trails traverse the sudd belt. At one point, where the soil is sandy, crocodiles come
ashore to breed. The width of the sudd belt is unknown. The distance from the
crocodile breeding place to any break in the sudd which would make a canoe landing
practicable is not less than 5 miles. The distance across the arm of the peninsula
to any point on the eastern shore, or any point where fly has been found inland, is
not less than three miles.

What appeared to be a small colony of fly was found. The catch (in 16 boy-hours)
was males 5, females 1, giving a density of 13 and female percentage of 16°6.

Bunjako Island, 10th August 1915.

The western shore, separated by sudd-bound channels from the mainland was
surveyed for about 5 miles. Situtunga were fairly common at the pomt, bush-pig
common, a few bush-buck, and a very small herd of buffalo; no landing places of
amphibious animals, and no trace of fly.

Kitebo Peninsula, 13th August 1915.

Surveyed eastern shore of peninsula, being the western shore of sudd-bound
channel separating mainland from Bunjako Island, for distance of 34 miles north
from last break in sudd belt. At the point sandy soil and ancient beach line, semi-
open, offered fair breeding places and good shelter. Game not common, no amphi-
bious animals, and no trace of fly beyond the first few hundred yards.

Zinga Island, 20th-22nd August, 1915.

Surveyed north-western peninsula (not shown on Whitehouse’s Chart) on northern
and western shores. It is anciently sudd-bound. Many situtunga were seen and
one animal that appeared to be bush-buck ; excreta of leopard at one point; no
amphibious animals.

Dense infestation at south-western point of the peninsula (where there is no sudd)
spread inland for at least 1,000 yards, owing to situtunga. No trace of fly at much

more than 1,000 yards from open shore on this portion of island. No sand or gravel
found at the back of the old sudd-fields.

Binga Island, 23rd August 1915.

Binga is entirely sudd-bound. The narrowest portion of sudd-belt is about 100
yards in width; except at this one point it is much wider. Bush-buck, bush-pig
and buffalo all common ; situtunga not common ; hippo land frequently where sudd-
belt is narrowest.

Fly was found near hippo landings; in 13 boy-hours 11 were caught, making
density of ‘5, and the female percentage was 45-4 There is a chance that these flies
followed the outer border of the sudd-field from the point where it breaks on the
stem of Bunjako Island, 3 to 4 miles distant, and that they turned up the hippo
trails to Binga. There seemed less probability of this being an independent colony—
a.€., a self-perpetuating colony—than in the case of fly caught on the western shore
of Bujaju.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 435

Conditions at the back of ancient, permanently anchored fields and belts of sudd
are somewhat more likely to be favourable to the existence of Glossina palpalis
than conditions at inland points generally, because there is apt to be sandy soil
sufficiently well watered to sustain sheltering vegetation. But it is improbable
that the necessary combination of food, shelter and breeding places would exist
at all commonly unless the sudd-belt is so narrow as to permit landing of amphibious
animals.

It is a safe enough presumption that a sudd-belt not traversed by hippo or crocodile
trails is free of fly on its landward side, but narrow belts crossed by hippo or crocodile
trails should not be regarded as sufficient protection against fly.

TX. Tae Inimicat Factors In THE BIONOMICS OF GLOSSINA PALPALIS,

The natural enemies and other inimical factors in the bionomics of Glossina
are numerous and diverse. Tor this very reason they are peculiarly difficult to identify
and study. It is much more simple and practical to study the more specific and
easily defined protective factors than the more numerous, less specific and less
easily defined destructive and otherwise inimical factors against which they afford
protection.

For example, shade at breeding grounds is protective against the destructive
effects of sun, and the clean-washed deposits of sand or gravel which are most favoured
as breeding places are protection against many different species of parasitic and
predatory destructors of insects, that are generally distributed and common in most
other types of soil than these. If shade at the breeding grounds is destroyed, the
puparia in these are quickly killed by the sun; and if there are no other breeding
grounds left to the fly in that district, it will soon be exterminated, either through
lack of protection or through the operation of destructive factors, according to the
point of view. But though shade is specifically protective against a specific destructive
factor, it is not that factor which operates exterminatively when protection against
it is destroyed ; it is the parasitic and predatory enemies which become exterminative,
because the flies will most certainly not deposit their young in unshaded situations
to be destroyed by sun. They will seek shaded places, wherein ants, perhaps, or
ground beetles or beetle larvae, or yet other predatory creatures will be the destructive
agents, according to the particular spot that they select.

It is impossible in such cases to specify the destructive agents actually responsible
for extermination of the species. The breeding grounds are a protection against
not one or two but against a whole coalition of enemies, any one of which may be
exterminative if the flies were to deposit all their larvae in situations exposed to that
particular one, and all of which are impotent to destroy if the one specific form of
protection is provided and utilised by the flies.

Moreover it is difficult to see how knowledge of the inimical factors can be made
of practical application. If we have a knowledge of the protective factors, they
can be utilised, for by depriving the insect of its protection very practical utilisation
can be made of its destructors. But knowledge of the destructors is valueless unless
it includes knowledge of the protection which serves the species against them. It

436 W. F. FISKE.

is necessary to deprive the fly of protection in any event (before its destructors can
be utilised) and knowledge of protective factors in its bionomics must be acquired
sooner or later.

The deprivative factors in the bionomics of Glossina appear, on the whole, to be
more efficacious in restricting its range and density than the destructive. They
include all the destructive and otherwise inimical factors in the bionomics of its
host species, and, in addition, whatever else may operate to render its food supply
less easily accessible or available. But the many deprivative factors are as difficult
to identify and define as the destructive, and it is simpler and more practical to
study the more specific food factors than to attempt to identify and study the less
specific destructors or deprivators of food.

In short, except as a matter of curiosity or technical interest, or as it may assist,
sometimes, in confirming conclusions or supplementing evidence bearing on doubtful
points, it appears neither necessary nor desirable to devote time and energy to
investigation into the identity and nature of the factors inimical to Glossina palpalis.
If the object for investigation into its bionomics was to cultivate it, as a beneficial
species, there would be the same object for studying its enemies that there is for
studying it as an enemy of domestic animals or human populations, namely, to find
means of protection where protection is lacking. But inasmuch as the object is to
destroy and not to cultivate it, and because, obviously, it is protected against all its
destructors and other inimical factors in the districts which it infests, it appears
much the more logical as well as the simpler and easier course to limit all specialised
or intensive studies to the factors responsible for its presence rather than to attempt
inclusion of factors manifestly unable to prevent its presence in injurious density.
On the face of it, destructive factors are locally impotent and inoperative, and the
reason for this must be found before they can be made operative and useful ; if it
can be learned how the pest is protected against them it may be possible to deprive
it of such protection and thus to make practical utilisation of them, but not otherwise.

On these accounts relatively little attention was devoted to the imimical factors
in the bionomics of Glossina palpalis. A few experiments were made, and these,
together with a summary of miscellaneous observations covering a very wide diversity
of enemy species, are cited in the following pages.

IX (a). The Hosts of the Fly as Destructive Factors in its Bionomics.

The bite of the tsetse is frequently painful enough to invite vigorous retaliation
on the part of any but the thicker skinned and least sensitive host animals, and an
active, sensitive host would thus become a destructive enemy of the first importance
if the flies were forced to feed upon it, or were foolish enough to press their attack
upon it regardless of consequences.* It is not improbable, however, that the tsetse is

* The very active and very sensitive host is as much an enemy of the flies as any
parasite or predatory destructor known. There are many parasites and predators which
would destroy the flies if they did not employ self-protective tactics against them ;
there are several species of host animals which would do the same if the flies were
devoid of self-protective instincts. The flies are actually exterminated by neither because
of specialised instincts of self-preservation.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 437

quicker to detect danger from this source and more successful in protecting itself
against it than any other biting fly. Certainly it is quicker to “ dodge” a blow and
more difficult to capture bare-handed than any other biting fly known to me, and it
is also the quickest to desist from attack upon a dangerously alert and sensitive
host.

The most dangerous of its potential hosts are probably monkeys, and it is
exceedingly doubtful if it could exist if forced to feed off them alone. Monkeys
are not only alert, sensitive and very quick of movement, but are expert catchers
of insects.

Sheep and goats are almost equally intolerant, but less dangerous, by far. They
strive to evade being bitten rather than to destroy the aggressor.

It was noted of the young bull used in the experiment cited on p. 402 that he
strove to destroy his tormentors, striking vigorously with head or foot, in pale
sharp contrast to the behaviour of the sheep used in the same experiment.

Man, however, is probably the most potentially destructive host next to monkey.
The ordinarily sensitive individual would destroy nearly every fly which he detected
in the act of feeding (occasionally the bite is absolutely painless and the fly engorges
before being detected), were it not for the extreme quickness of the insect. The
chances of a fly’s being able to engorge undetected are not at all good, and if deprived
of all other than human hosts many would become very hungry before they succeeded
in engorging. The very hungry flies take desperate chances, and press their attack
so viciously and persistently that they are frequently destroyed. This was
characteristic of the flies in the Mujuzi Creek colony, where they pressed their
attack so persistently that so many were destroyed as to make it difficult to imagine
that the species could exist as a parasite of man alone.*

IX (6). Predatory Destruction of Adult Flies.
(1). Sprders.

One of the conspicuous biotic phenomena along the shores of Victoria Nyanza
is the incredible number of spiders that occur locally at points along the mainland
shore, or on small islets. There are numerous species, the most conspicuous of
which is a gigantic web-spinner identified (by Mr. C. C. Gowdey, Government
Entomologist of Uganda) as Nephele pilipes. The full-grown individuals of this
species habitually stretch their webs from one tree or bush to another across openings
in bush or forest, or between the trees and bushes growing not too distantly separated
along the lake shore.

There are several other large species which appear like the Nephele to be quite
independent, and many small species which are not independent, but which live—
in proportionately incredible numbers—as guests or inquilines of the larger species—
sometimes by robbing their webs of the small insects entangled in them, and some-
times by spinning their finer webs in the coarse meshes of the webs spun by the
larger species.

*So alert, quick and sensitive a host as man must be approached by insect parasites
with considerable circumspection ; the successful insect parasites of man appear either
to approach him while asleep, as in the case of the bed-bug, the floor maggot, most
mosquitos, etc., or to be extraordinarily insidious, like the chigger, or to be exceptionally
resistent to retaliatory activities like fleas, and some others.

438 W. F. FISKE.

At times the great webs of the Nephele are so numerous that in passing through
semi-open spaces along shore one must strike them down at every step, and literally
hundreds of the gigantic spiders can be seen through openings between the trees,
each suspended in the centre of its web and silhouetted against the sky ; and not
the least remarkable feature of the phenomenon is that one small islet may be thus
festooned, and another, a few hundreds or thousands of yards away, may be entirely
free. Several of the islets and small islands lying off the shores of Kome (fig. 1)
have been found densely infested, but Kome itself has always been entirely free.

These webs are frequently stretched squarely across the courses which would
naturally be followed by streams of food-hunting flies, and are inevitably a source
of considerable danger. The flies are well aware of it, and remarkably successful
in avoiding capture; still many of them become entangled. Of these a fair
proportion escape by the strength and vigour of their actions, but a proportion
are destroyed.

On the island of Wema, in February 1914, a light infestation by spider and
dense infestation by fly afforded opportunity for study, and counts were made of
insects as large as or larger than tsetse that were found entangled in the webs, with
results as shown in Table XX XV. _ It should be noted that of the seven tsetse recorded
five were found entangled on approaching the webs, and two became entangled
while the webs were under examination.

TaBLE XXXYV.
Insects found entangled in Webs of Nephele pilipe:.

In webs along|In webs in old
Insects. fly beach, | plantations, Total.
27th Feb. Ist March.

Odonata . 1 2 3
Ephemeridae .. 4 1 5
Orthoptera (Acridiidae) 9 13
Homoptera eon 0 14 14
Hemiptera ae 4 4 8
Lepidoptera 9 14 23
Coleoptera 3 6 14
Hymenoptera .. 5 8 11 14
Miscellaneous Diptera : 5 2 ra
Glossina . : 5 2 7

Total .. ue ae : dint 43 65 108

Percentage Glossina a0 mE ay 11-6 a6 65 %
|

The unexpectedly high percentage of Glossina amongst the insects of equal or
larger size captured by the Nephele rendered it extremely probable that where
infestation by spider is excessively heavy it must have some effect in reducing
local density of tsetse, and it is believed that its effect is fairly measured in the

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 439

accompanying table (Table XXXVI). The two islands of Ziro and Kerenge were
very densely infested by spiders, but otherwise they appeared to be in all respects
as favourable to tsetse as the other islands named in the table. Both were accounted
“better than Namba” and “as good as Bulago” in the final summing up of
impressions gained during the several days spent in the survey of each.

TaBLE XXXVI.
Effect of Spider on Density of Glossina palpalis.
Character of Knvironment. Infestation.
Island. Date.
Shelter. | Breeding Food. Spider. Male | Female
grounds. density.| ratio.
Bulago .| Jan. 1914; Good ..| Good . | Good ..| None 27°5 29°0 %
Ngsadzi..| ,, x ee Se x LG ou, a 5 18°4 24°6 9
Kitebo..| Dec. 1913) ,, %. = Bhi oy. bi “ 14:5 3011 9
Namba .| Jan. 1914) Very good) Poor 4 Very good|_,, bik 9°4 LO%
Ziro ..| Feb. 1914; Good ..| Good . 1 Good \ Exceeding-|f 5°4 rae) LO
Kerenge * “4 £ Laine (ets OY Mee ly many |\ 3°8 24-7
Average infestation of islands with no spider .. 3% ak Hee TD DB as
rs * i i » Many spiders .. a af 4°6 2422, Of,
Reduced infestation attributed to spider vs sis sie TEE PAG Se oTleran

Numerous other islands were found to be heavily or excessively infested by
spiders, but in the case of no others was comparison with spider-free islands so
generally fair.

‘It is possible that if conditions of life otherwise were less favourable than on
Ziro and Kerenge, the spiders in comparable density might prove an exterminative
destructor. The islands of Dwanga Mkuru, Dwanga Mto and Dwasengwe noted
in Table XXXII (p. 422) as very lightly or not at all infested may possibly owe
freedom from fly to infestations by spiders, which was fairly heavy in each case,
as well as to absence of breeding grounds.

(2). Dragonflies.

A dragonfly of a rather small and undetermined species occurs commonly, but
irregularly distributed, throughout the islands and riparian belt. There are other
species, but this one is conspicuous for its numbers and for a well marked habit of
following moving animals and man and feeding off the flies which are attracted to
them. Large numbers—in exceptional cases perhaps as many as 40 or 50—of these
dragonflies may follow a man along the open shore, and they have been observed to
capture tsetse many times. _

The species varies greatly in local density from one island or district to another.
It was observed most abundantly along the western shore of Bukone Island in
September 1914. A fairly complete survey of the island was made and the infesta-
tion indicated (male density 8-6; female ratio 23-4 per cent.) was considerably

440 W. F. FISKE.

lighter than conditions, apart from the very exceptional number of dragonflies,
would have led one to expect. It seemed quite probable that the destructor was
responsible for a considerably less heavy infestation than would otherwise have
occurred.

There is this difference between spider and dragonfly—that sheltering vegetation
is of little or no protective value against spiders, but of much protective value against
dragonflies. They will not follow a moving animal into cover, and the flies are not
likely to be captured unless they hunt in the open.

Apart from spiders and dragonflies no destructive enemies of tsetse have been
observed that appeared to be specifically responsible for appreciable reduction in
density of Glossina palpals.

(3). Bembex.

A wasp of the genus Bembex works somewhat like the dragonfly in following moving
animals and picking off flies attracted to them. Its habits have been described by
Dr. Carpenter.. Though very variable in density it is nowhere excessively common,
and at most points was not observed to occur at all. It would stand third in
importance of the destructors of adult flies observed at work, but at no point did
it appear to produce an appreciable effect on density of fly.

(4). Muscellaneous Destructors.

Small lizards are among the most numerous of entomophagous creatures which
hunt in the haunts of tsetse, and it has seemed probable that they destroy many
flies. They are nearly everywhere common, and vary relatively little in local density.

A fair number of other entomophagids have been observed working in the haunts
of tsetse, but none among them has appeared to be sufficiently numerous to be of
any consequence as an enemy.

It is to be noted, however, that it is virtually impossible to study the actual or
potential destructors of the inactive flies. It is almost impossible to discover the
inactive flies without putting them to flight, and nothing is known of the specific
dangers that beset them.

IX. (c). Parasitic and Predatory Destruction of the Larvae and Puparia.

(LpAnis:

Almost the only insects common at breeding grounds in clean-washed deposits
of beach sand and gravel are the ants which run about over the surface. These would
destroy the weak larvae of tsetse if the latter did-not bury themselves quickly below
the surface. The prescience of the female flies guides them to spots where their
young are not unduly exposed to this danger.

There are numerous species of ants which will destroy the puparia if encountered,
but none of these are known or suspected to burrow beneath the surface of the soil in
breeding grounds in search of buried puparia. The “red driver” and some other
of the larger ants will carry the puparia away bodily.

In woodland, or where there is a thick blanket of decaying leaves on the surface
of the soil, there are species of ants which work under the loosely packed material
and which will destroy any Dipterous puparia encountered init. If flies ever selected

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 441

such localities as breeding places it is probable that their young would not penetrate
deeply enough to escape these wandering ants, but the flies appear never to deposit
their young in such places.

It is with ants as with so many other potential destructors, including monkeys
and spiders. If flies fed or attempted to feed on monkeys they would be destroyed ;
if they flew headlong into the webs of spiders their species would probably be
exterminated on the spider islands ; if they failed to display characteristic prescience
in the selection of breeding places, their young would be destroyed by ants, parasites,
etc.: but in each case their instincts save them from destruction.

(2). Hymenopterous Parasites.

The parasitism of puparia of Glossina palpalis in good breeding grounds by Hymen-
optera was found to be absolutely negligible. The only instance encountered was
on Wema Island, where in a catch of 203 puparia, 3 were found attacked by a small
gregarious Chalcidid. A percentage of parasitism at this poimt of 1:5 was thus
indicated. At other points near at hand enough more unbroken puparia were collected
and examined to make a total of 1,775, and not another parasitic specimen was found.
From other points, first and last, more than 3,000 living puparia were collected and
examined, making some 5,000 in all, and no other case of parasitism was encountered.

Occasionally empty shells are found with small round holes such as are left by
Hymenopterous parasites in emerging, but except at the one point noted above,
examination of such shells has never disclosed the characteristic (and unmistakable)
exuviae of such parasites. The selected breeding grounds of the species are practically
absolute protection against this class of destructors.

The parasite discovered on Wema bred with the greatest freedom on puparia of
Glossina in confinement. About one month was required for the generation. It
could be an enemy if it would, but its instincts lead it elsewhere than in the breeding
places of Glossina in search of prey.

(3). Miscellaneous Predatory Destructors.

Ground beetles and their larvae of several species, Carabids and Elaterids, are
known to be destructive to puparia, and believed to be very destructive to them
except in the dry deposits of sand, gravel and fine vegetable debris normally chosen
by the female flies. These, or other destructors unknown, frequently destroy
appreciable proportions of puparia, varying in different localities. It was thought
probable that counts of “ eaten ”’ and “ hatched ”’ shells found at breeding grounds
in sand and gravel would show a much lower proportion of “ eaten ” than at breeding
grounds in vegetable debris. The actual counts, however, showed no great difference
and indicated that the degree of protection provided by vegetable debris is almost
equal to that provided by beach sand and gravel (Table XX XVII). There is a much
sreater difference between good and poor breeding places in sand and gravel or
in vegetable debris than between sand or gravel as compared with vegetable debris,

The two instances on Mbugwe and Nsadzi Islands where destruction by predators
amounted to 27:2 per cent. and 31:2 per cent., respectively, are worthy of special
mention. In each case the prescience of the female was deceived by superficial
conditions. Both cases were in basking spots of Varanus. That on Nsadzi Island

442 W. F. FISKE.

was in the very centre of a great tuft of grass that had been killed by overhanging
(encroaching) shrubbery. The fine dead leaves of grass had been pressed down by
the body of the animal to present a firm smooth surface, quite dry. But underneath,
the grass leaves were moist and decaying, and the larvae of tsetse, penetrating
the surface and forming puparia in the damp interior were largely destroyed. The
case on Mbugwe Island was very different. Here there was a layer of vegetable
debris under fern and over pebbles that had been packed so firmly by the body
of the Varanus that the larvae could only penetrate with difficulty. Many were
forced to form puparia only half concealed and of these all were destroyed.

TABLE XXXVII.

Showing Relatwe Degree of Protection to Puparia of Glossina palpalis provided
by Breeding Grounds in Beach Sands and Gravels and in Vegetable Debris.

; Finds of puparia shells. | Proportion
D Character of breeding destroyed
Locality and Date. ground. ue | by weedeat
ed Eaten. Total. tors.
Damba .. Feb. 1914] Coarse sand .. .-| 1074 89 1163 T6%
Kerenge .. Feb. 1914| Sand and gravel ..| 152 3 155 20%
Wema .. Feb. 1914] Brown sands and
gravels .. ..| 2974 70 3044 273%
Bugalla .. Sept. 1914) White sand .. as 81 3 84 So oA
Zinga .. Sept. 1914 3 is ioe? ate 457 ZY. 478 4°4 %
Damba .. Feb. 1914 2 B90 tte --| 3939 554 4493 12°93 %
Nsadzi .. Feb. 1914] Very fine white sand.| 417 86 503 1771-87,
Namba .. Mar. 1914} Vegetable debris... 76 0 76 —
Mbugwe .. Mar. 1915 i a ar 55 0 55 —
Kiuwa .. Feb. 1914 a 4 ce 310 7 317 222%
Karambidi Sept. 1914 » % bla 76 3 79 Ba yA
Limnaiba . Sept. 1914 ee - ae 268 15 283 58%
Mugogoya Sept. 1914 a be as Tid 8 85 94%
Mbugwe .. Sept. 1914 % < os 62 24 86 272%
Nsadzi .. Feb. 1914 a x ig 55 25 80 312%
Total and average in sand A ..| 9044 826 9920 1 0.0
998 99 », vegetable debris .. 979 82 1061 9°9 %*

* Average by localities.

A yet more striking case of this same sort was observed in the interior of Bugalla
Island, where a surface of hard clay was very thinly strewn with gravel. Shelter
and surface conditions were very attractive to the flies, and it is probable that
many larvae were deposited there, but though an occasional perfectly fresh puparium
—formed the same day—was found, no shells or old puparia could be found on
any of the several occasions the point was visited. It appeared to be a perfect
death-trap.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 443,

First and last, a considerable number of experiments were made in the planting
of puparia in localities where none could be found, but where it was conceived they
might be deposited by females which could find no safe and attractive breeding
places. These experiments proved that certain places—notably tufts of a certain
species of grass—would afford adequate protection to the puparia if the flies had the
prescience to select them. They also proved that in many places which the flies
might conceivably select in lieu of better, the puparia would almost certainly be
destroyed.

It was concluded rather definitely that the fly would be exterminated by the
predatory destructors of puparia if the females were to deposit their young in the
massive shelter—.e., in forest or forest-like shelter—which they require as protection
for themselves. It is most especially in such places, where the surface of the soil
is covered with dead leaves and leaf-mould, that the puparia are found and destroyed

by predators.

IX (d). Inanimate Destructive Factors.

(1). Sun.

There are very few insects which can withstand for long full exposure to sun,
and Glossina is no exception. Flies have been inadvertently killed through less than
half an hour’s exposure of the cage in which they were confined. The cage rested
on sun-baked earth, however, and if unconfined the flies could probably have lived
much longer. But shade, and of a substantial sort, is requisite during the heat
of the day, and flies cannot be tempted far from it when the sun is at maximum
power.

Sun striking on soil in which puparia are buried no more deeply than the larvae
naturally penetrate—rarely more than 14 or 2 inches—will destroy them quickly.

It not infrequently happens that a week or a fortnight of dry weather will
destroy—or at least cause to droop and wither—herbaceous vegetation that
previously afforded attractive and adequate shade for breeding grounds, and great
numbers of puparia, in the aggregate, are destroyed by drought and sun in this
manner. The most curious case was encountered on Tavu Island in September 1915.
Very attractive breeding grounds, adequately protected by shrubs of a sort known
to the natives as “ Kinsembwe,” had been exposed by the activities of a defoliating
caterpillar (a species of Acraea) and quite extraordinary numbers of puparia were
destroyed. Game breaking down vegetation at breeding grounds on Bugalla
Island produced the same effect on a smaller scale.

All things considered the sun must be regarded as the most destructive—potentially
—of any “enemy” of Glossina palpalis and the one against which protection is

most urgently required.

(2). Storms.

Light shelter is sufficient protection against sun, for both pupae and adults.
Massive shelter is nevertheless certainly necessary to the species, and the one
destructive agency against which it is obviously of much greater protection than
light shelter is storms.

(659) G

444 W. F. FISKE.

It is very difficult to study the effect of storms on adult flies—mainly because
they are almost impossible to find except when active. One can only be sure that
their hiding places have been discovered by searching at hours when none are
active—for they are quick to take flight if disturbed and quickly confused with
active flies if there are any about.

It is a strongly marked characteristic of the active flies to seek shelter when a
storm is brewing, especially the females, as shown by the accompanying graph
(fig. 10), which was secured in the course of an experiment with the streams of
moving flies. The catches were made at a point when all the “ resident”’ flies
had been caught off, and only those in the passing streams were being caught.
About 11 o’clock a storm gathered on the horizon. The temperature fell a few
degrees, and a light breeze sprung up. This was all that happened. The sun
was not obscured, and the storm did not come within miles of the island (Manene,
17th March 1915), but almost half the females sought shelter, and remained secluded
until the cool breeze dropped. Then they quickly became even more active than
before.

Certainly the insects would not flee so quickly at first signs of a storm unless
they were fearful of it, and next to sun, storms are probably the most dangerous
“enemy ”’ of the species.

(3). Flood.

The rising waters of lake or stream may destroy many puparia in breeding grounds
in sand or gravel deposits, and would destroy many more than is in fact the case
were it not that the females avoid depositing their larvae in spots too near the
water. Even on the “ fly beaches ”’ it is in the old beach line beyond reach of even
the heaviest waves that nearly all the large deposits of pupae have been found.

The effect of such a flood tide as that of 1906 upon the puparia must be very
destructive, for in May and June of that year there could have been very few good
breeding places—relatively—beyond reach of the waves (see p. 457).

In the bed of the Victoria Nile, some miles from its source, excellent and much
frequented breeding grounds were found (April 1914) at points where they would
almost certainly be destroyed in May or June.

(4). Fire.

Bush or grass fires are potent factors in the bionomics of Glossina morsitans, but
at no point visited on the islands or riparian zone of Victoria Nyanza was fire likely
to affect Glossina palpalis to an appreciable extent.

IX (e). Secondary Environmental Factors detrimental to Glossina.

Whatever benefit may be derived by humanity from the activities of the two
natural enemies of Glossina palpalis which produced an appreciable effect on its
density—spider and dragonfly—is inseparably associated with whatever it is which
permits these entomophagids to multiply to such excessive numbers as is required
to produce that effect.

445

INVESTIGATIONS INTO THE BIONOMICS} OF GLOSSINA PALPALIS.

“OOULISIP B 4V Sas WI0}S & JO syyvdjod vursso}H JO SElJIATJO’ WO Yoo ede WvIseIg “OT ‘Shy

al | | ] fe | CD2EF| GAEL | 0}
| | | Pode Nee, ‘ a a J
| | | Pa =A r |
| ae | eA ‘S oe eee es
a4 = a SS aa A n= -
Sees ;
en f X
y. 4 \
ca g :

Vik) ISO
IZIIING IZIIIG JOO S)

O22 002 Ori ~o2l OT Gr2) 0fz2m O02 Ovim Oz) 001 cr Oly 02:01 ODOM ore.

(659)

446 W. F. FISKE.

Nothing but “ insularity ” could be attributed as the determining factor in the
case of spider. Its great abundance is partly due to the incredible numbers of
“lake flies’ (Chironomus, mainly) which breed in the lake and frequently rise
from it in such swarms as to appear like the smoke of a large steamer on the horizon.
Swarms of them that must number thousands of millions gather and sweep on shore,
where, if there chances to be a generation of newly hatched spiders coming on, they
provide an abundance of food at a critical time. Possibly there is some correlation
between the incidence of lake fly and of spider, which works out to the disadvantage
of Glossina in the end.

There is less mystery in the case of the dragonfly. The particular species before
mentioned feeds as nymphs in shallow, rush-grown water, more particularly where
there is sand and pebbly bottom. The existence of such shallows off shore is detri-
mental to tsetse on shore.

The Bembex breeds in fine, dry sand, of a sort that it can burrow into without
its galleries collapsing. It is very irregularly distributed, because this type of soil
is not at all common. The type is mildly detrimental to tsetse where it occurs.

Absence of rock, both on shore and in the shallows off shore, is a more potent
factor in its inimical effect on tsetse, for Varanus is a rock-dweller on land, and its
most favoured food is the crabs and molluscs—especially the crabs—which are only
found in rocky shallows. There are long reaches of sandy shore south of Dumo
village, in Buddu, which are virtually free from fly because of absence of food ; only
crocodile occurs, and only at a few points where buck comes down to the water are
flies and crocodiles likely to meet.

Notrace of Varanus was seen along these reaches. Had it occurred (7.e., if there
had been food for it, and a rock-bound bottom), its inland wanderings across the
natural clearing that bordered the shore would have brought it into contact with
shelter and breeding places, and conditions of life would have been excellent for tsetse.

IX (f). Leopard as a Deprivative Enemy of Glossina palpalis.

Through a combination of partly fortuitous circumstances it is possible to measure
or estimate, roughly, the effect of the presence and activities of leopards in restricting
the range and density of Glossina palpahs. It is plainly a more efficacious deprivative
enemy than spider is a destructive enemy of the insect. The data follow, item
by item. |

(1). On the Sesse Islands there are no leopards or other large beasts of prey.
Following elimination of the human factor situtunga increased at a rapid rate on
every island where it occurred, and spread from the marshes to the hills and
abandoned plantations. As a result the range of fly inland increased from a low
average of perhaps 300 yards—not counting stragglers—to one that is at least 5 times
ereater wherever shelter is good and continuous for this distance inland.

(2). On the two sudd-bound islands of Bunjako and Binga leopard occurred,
and also situtunga. There was no increase of the latter, nor could evidence be found
on Bunjako, which is densely fly-infested along its shore, of more than ordinary
extension of infestation inland. ;

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS.¢ 447

(3). On the island of Bussi, largely sudd-bound, but previous to 1909 separated
from the mainland by a canoe track through the sudd, no leopard formerly occurred.
The canoe track was choked by sudd, however, and leopards crossed it soon after the
eastern shore of the island was depopulated. In October 1914 the island was surveyed
for game and inland infestation by fly. Nowhere else were the excreta of leopards
discovered in such extraordinary quantity, and examination of them—exact figures
were not kept—disclosed unmistakable remains of either situtunga or monkey in
nearly every case. No other traces of situtunga were found on the island except,
as is usual on the mainland, along the borders of the sudd-fields. There was no
inland extension to the infestation by fly, which was fairly dense along the shore.

(4). Zinga Island was sudd-bound like Bussi, and like it, separated from the main-
land by a canoe track. It was said by the natives to be entirely free from leopards.
Situtunga increased to very large numbers, as on the Sesse Islands, and extended
itsrange inland. The range of fly was also extended inland up to at least 1,000 yards.
In exploring the island for inland infestation in August 1915 fresh excreta of leopard
were found, and examination of the old canoe track showed that it had become
broadly choked with sudd. If the history of the incursion of leopards in Bussi is
repeated, the animals should become very numerous by about 1920; situtunga
should be driven back into the protection of the sudd-fields, and monkeys, which
were excessively numerous in 1915, should be greatly reduced in density.

(5). At no point on the mainland has situtunga increased to much, ifat all, beyond
its former density, nor has it extended its range beyond easy reach of the animals
from the protection of sudd-fields and marshes.

(6). At about three points on the mainland where the local environment affords
exceptionally good protection to bush-buck, inland extension of infestation by
fly corresponds to that on islands where leopard is absent and situtunga has increased
to exceptional density.

There can be no doubt that the range and density of situtunga are controlled by
the leopard, which is unable to harm it in the protection afforded by sudd-fields
and marshes, but which hunts it assiduously along the border of these protected
precincts. There can be little doubt that the range and density of bush-buck are
controlled in the same manner; protection in its case being provided by densely
tangled thickets, through which it plunges with strange facility, and within which
no large animal can approach it without betraying itself.

Wherever leopard is absent and situtunga present in sufficient density, range of
fly is extended inland, and situtunga has increased or is increasing (if present at all)
everywhere that leopard is absent. There is every reason to assume that bush-buck |
would do the same, in the absence of leopard and of human hunters.

The correlation is perfect at every point, and it is an entirely reasonable conclusion
that in the absence of leopard inland range of fly would on the average be as extensive
along well protected portions of the riparian belt as on the islands where situtunga
has increased or as at the few points along the mainland shore where the bush-buck
has found exceptionally secure protection.

There would be inland extension of infestation to approximately 5 times its present
depth at many points, and along extensive reaches of mainland shore, if it were not

448 W. F. FISKE. .
for the deprivative effect of the activities of leopards. They are certainly a more
efficacious natural enemy of the tsetse than spiders, or than any other destructive

enemy that has been identified.

IX (g). Relations between Glossina palpalis and other Biting Flies.

The data secured during the experiments with cattle described in Sect. VI (c) (Table
XXV) indicated that the presence of other biting flies than Glossina might have an
effect on host animals that would effectually prevent Glossina from feeding upon
them. This has since been confirmed in various ways; partly by a series of
observations on the behaviour of domestic animals, or of animals in Zoological
Gardens, under attack by flies of various species. The Zoological Park at Washington
affords particularly good opportunities for such observations, being situated well
out from the city in a naturally wooded ravine where Stomoxys and Tabanus are
numerous. Not less than 25 species of ungulates occupy paddocks that are badly
infested by these flies, and various others, including several of the familiar game
animals of Africa, have runways that are partly exposed.

Variations in degree of tolerance or intolerance to attack were very remarkable,
ranging from the absolute intolerance of Barbary sheep, which would seek the seclusion
of their hut at the approach of a single Stomoxys, to the phlegmatic indifference of
red deer, and wapiti, which would permit Tabanus to engorge without serious protest.
But there was no animal that was not roused to protest and retaliation if flies became
too numerous and persistent, and any animal that is thoroughly aroused and excited
becomes of little value as a host to any biting fly.

Eland, for example, was passive under attack by Stomoxys in moderate numbers,
but repeatedly on approach of Tabanus the animals would betray annoyance, and
move to protect themselves not only against the greater but the lesser pest. A single
Tabanus would cause all the elands to seek shelter, and effectually deprive several
dozen Stomoxys of this particular source of food. Bison was more tolerant than
eland, but a few Tabanus would sometimes set a whole herd in motion. Zebra
and wild horse were less tolerant than eland, but would suffer Stomoxys to feed in
small numbers until Tabanus appeared.

None of the larger biting flies observed on these or any other occasions have
pressed their attack on an unwilling host regardless of risk to themselves, but none
has shown more regard for its own safety or been quicker to desist from attack upon
an intolerant host than Glossina palpalis. A Tabanus has been observed to return
to the attack more than 30 times before admitting defeat. A Glossina, unless
positively famishing, doubtfully returns more than 4 or 5 times at the most—perhaps
not more than once or twice. With man they usually desist after the very first
repulse. The more persistent pest may, therefore (as shown by the experiment—
Table XXV) become a very efficacious deprivative enemy of the less persistent
and the effect is felt in two distinctly different ways :—(a) the host animal may
be induced to leave a locality or district where it is liable to be annoyed, or (b) the
less persistent and more easily discouraged flies may be induced to leave the locality
where a more persistent and annoying species is active.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 449

The dependence of Glossina palpalis upon specific forms of protection—breeding
grounds and shelter—has been shown in Sects. VI. and VIII. Except when provided
with both kinds within easy reach (of the flies) from each other the species cannot
exist. Where both kinds occur in sufficient proximity the fly can exist if also
provided with food, but food, also, must be within easy reach of flies from both
kinds of protection.

It is this factor of the actual location of food with respect to protection that
counts most heavily in determining the range and density of fly. It is much less
a question how much food—how many host animals—exists in a given region or
district, or on a given island, than it is one of the quantity of food or number of
host animals existing in the adequately protected precincts. It is not necessary
that the host animals shall be destroyed in order to injure the fly; but is quite
sufficient that they should move a little outside the range or reach of flies from
either shelter or breeding places.

Therefore if the animal is induced by Tabanus, or any other biting fly, to move
a little farther away from the specifically protected precincts, Glossina is forced
to follow, and in following is exposed to all the risks, and enemies of a destructive
nature, that make specific protection a requisite for its existence. The effect of the
rival is precisely equivalent to either of the following :—(a) reduction in quantity
of available food, (6) reduction in quantity or degree of protection, or (c) increase
in number or destructiveness of enemies.

This is if the host is induced to leave the locality. It is the same if retaliatory
activities on the part of the host induce the flies to desist from attack and to move
on in search of another, more submissive anima!. ‘To do this the flies must leave
protected precincts and undergo greater risk of being destroyed themselves, or of
being unable to find protection for their young.

The activities of Tabanus have the direct effect of reducing the quantity of available
food in protected precincts, which is the equivalent of either reducing the amount
of degree of protection, or increasing the number or destructiveness of enemies.

IX (hf). Super-density of Glossina palpalis.

There appears to be nothing to prevent the multiplication of Glossina palpalis
to any degree of density which the food supply permits in any locality where protec-
tion is provided for it, except the continued dispersion of flies outwards from the
protected precincts to infest a surrounding zone of unprotected territory. Within
the protected precincts the natural increase of the species is greater than the
mortality, but gain through natural increase is equal to loss through mortality
plus loss through dispersion. Hence density does not increase beyond certain limits,
because when the flies become sufficiently numerous to arouse their host animals
to the point of retaliation, the immediate effect is to accelerate dispersion, thus
reducing density in the protective precincts. Outside these precincts mortality
is greater than reproduction, but loss through mortality is equal to gain through
natural increase plus gain through immigration.

450 W. F. FISKE.

A delicate “ natural balance’ is thus created, which remains in a state of great
stability as long as nothing interferes with the free dispersion of flies. Super-density
(or ‘“ outbreak ”’ or “ epidemic’) of fly is impossible. It cannot increase so as to
cause direct injury to its host species through excessive blood-letting ; it cannot
even increase so as to cause its host animals excessive annoyance.

But if dispersion is for any reason obstructed or interfered with, the mechanism
of the balance between the fly and it host species is thrown out of gear, and its
density continues to increase until in their own protection its hosts are forced to
abandon their haunts. The fly thus becomes an “ enemy ”’ of its hosts in actuality.
It has “broken out” or become “ epidemic,” in the terminology of economic
entomology. It has increased to a state of super-density when the flies have become
the real deprivative enemies of each other, just as flies of the genus Tabanus may
become the deprivative enemies of tsetse-flies. One tsetse can be the deprivative
or competitive enemy of another as easily as a Tabanus can be the enemy of a tsetse.

On a very few occasions only were such conditions found to prevail in the region
covered by these investigations. The first instance was on the island of Kimmi,
which alone of all the islands visited—some 70 in all—appeared to constitute almost
entirely a protected area. There was no opportunity for dispersion from it unless
the flies struck out over open water, which they were plainly loath to do, and there
was hardly a spot on the island where the flies could fail to find both shelter and
breeding grounds within easy reach. There was no apparent reason why density
should not increase under such circumstances to the point of super-density, which
would either injure the host animals or force them to abandon the district.

Conditions encountered on this island in January 1914 indicated that something
of this sort was actually taking place. The infestation was heavy (37:2 according
to methods finally accepted for estimating male density) and the female percentage
was very high (63:5 in total catch and 55°3 at the lowest estimate). The flies were
as ravenous as the sex ratio indicates, and host animals seen during 34 days spent
on the island by Dr. Carpenter and me totalled a single Varanus seen by him and
a single hippo heard crashing about in a reed thicket by me.

No crocodiles were seen, but there was abundant evidence of their very recent
presence in the form of nesting places with hatched egg-shells, and runways through
the reed fringe along the fore-shore. There was also plenty of evidence of the recent
presence of Varanus, the mounds thrown up by which were more numerous on
this than on any other island visited, except Manene. It seemed not at all improbable
that the fly had actually increased to such super-density as to force even the crocodiles
to abandon the island, temporarily, until the density of the pest should be reduced.

This idea was sustained by the failure of the experiment cited in Sect. VI (c),
Table XXIII (p. 402), in which tethered crocodile and Varanus (on this same island
at this same time) became so quickly intractable under attack that experiments had
to be discontinued.

The idea also suggested the following experiment :— .

If, as appeared probable, super-density of fly had caused the host animals to
abandon the island temporarily, the sudden deprivation of food would result in a

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 451

quick fall in the number of young produced.* If such decrease in rate of reproduction
had taken place within the period of 3 to 4 weeks required for the pupal stages, the
fact would be disclosed by examination of the puparia, and the proportionate
numbers of them found in early and late stages of development. At a certain late
stage in development the pupa is found within the puparium shell showing traces
of adult coloration. Such pupae must have been from larvae deposited at least
two weeks before. In freshly formed puparia the pupa is unrecognisable, and it
requires at least two weeks for it to form and to acquire colour, except in the eyes,
which are coloured almost as soon as it takes recognisable form. A count of late
and early stages of puparia collected at various points on Kimmi Island was therefore
made, and as a check upon it some 3,000 puparia collected m various other localities
were examined and counted to ascertain the normal proportions.

The normal, as indicated by the check, was 27°3 per cent. late stages, but on
Kimmi Island the late stages composed 46°3 per cent. of the total found and examined.
There should have been—if the normal is correctly calculated—very close to 16
early stage to 6 late stage puparia ; actually there were only 7 early to 6 late, indicating
that less than half as many young were being produced at the time the collection
was made as were being produced two to three weeks before.

In March the island was revisited, the infestations measured, and a second search
made for puparia. Conditions were changed notably. Density was somewhat
lower than in January, but food had become more plentiful, female percentage
had dropped to a figure not far above the ordinary, and the flies were no longer
ravenous or especially troublesome. The examination of puparia disclosed only

TaBLE XXXVIII.

Showing Variation in Rate of Reproduction (Number of Young produced) of
Glossina palpalis as reflected by Proportion of Late and Early Stages of Puparia.

. . Total and - }
Kimmi Kimmi
Felandepesl ees aoe Island.
Analysis of finds of puparia. aes
January fe ane (1) March
1914, vege 1914,
normal),
No. of puparia found .. ha ae ap 188 3615 155
No. of early stages... as ie uy 101 2626 130
No. of late stages us a pa ae 87 989 25
Percentage of late stages eS ma oh 46°3 % 27°73 % (2) G1 %
Ratio of late to early .. = oF Se Le EG bi: 266 (2) 1: 5°20
Excess or deficit, early stages . oe oo ee oe 0 15-99%

(1) See Sect. VII (a), Table X XVII.
(2) This ratio would be a constant if the rate of reproduction of the species (number
of young deposited) did not vary.

* The females nourish the young larva until it is full fed and ready to transform into
the pupa, equal in size to herself. Normally she must feed (engorge) at least 3, probably
4 and possibly 5 times in order to provide sufficient food for a single larva.

452 W. F. FISKE.

16-1 per cent. of late stages. There should have been 16 early to 6 late stages,
whereas there were found approximately 32 early to 6 late, or almost exactly
twice more than normal, indicating that the flies were depositing twice as many
young as were being deposited two or three weeks before.

The experiments were conducted with sufficient care to eliminate most chances
of error, and little doubt is felt that there was actually a falling off in the number
of young produced followed by a proportionately heavy increase in numbers. The
only question is whether this falling off was due, as assumed, to the temporary
abandonment of the island by host animals, and whether this in its turn was due,
as assumed, to increase in density of fly. The experiment turned out exactly as it.
should, in accordance with the working hypothesis. It sustains the hypothesis,
though it does not prove it. The point in question is, in fact, one that could only
be proved by investigations conducted over a considerable period on an island
like Kimmi, with exact observations on the movements of host animals and
variations in degree of infestation.

TABLE XX XIX.

Showing Variation in Infestation of Kimnu Island between January and March

S.W. Point oP Orde 2a3 62°93 &% 908. 1o"t 26°38 %
N.W. Point Y lect 43°3 O22 OF) Me BOR eek O 277397
S.E. Point oo Wels «06 39°5 532,95 4) SSL hae 2774 %
Southern Shore .. ele han G 42°7 53°0 % 292 | wy Lee 137 3,
Totals and Averages... 1438 | 37:2 3°3 0/, 1353 | 32°0 23°7 0,

The only other localities visited in the course of these investigations in which the
female percentage was in excess of 50 (for the entire island or district) are :—
(1) Bunyama Island (data mislaid*)

(2) Bale Beach and Mujuzi Creek colony Me mi 64:6%
(3) Bukakata South colony 4: ae Hy ce 51:0°,
(4) Kaziru District (Buddu) ae as Pee ire 50°6%

* Since recovered, but not included here.

In none of these localities did it appear that scarcity of food was due to density
of fly. The fly was suffering from want of food, certainly, but merely because there
was so little food and this so hard to find. Conditions were as they would become
on Kimmi if the host animals abandoned the island permanently, all save two or
three Varanus or a single crocodile. The point where food chanced to be, at any
time, would then be a “ colony centre ’”’—exactly like the colony centres in the

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 453

Bwendi district (fig. 6, p. 411) or at Mujuzi Creek, and there would be dispersion from
them as from the points in the Bwendi district or from the point at the mouth of
Mujuzi Creek where crocodiles harbour (p. 388), or in Bugaba Island where breeding
ground marks the colony centres (p. 423). There can be no super-density of fly
except under exactly such conditions as prevailed at Kimmi when there is food
generally distributed throughout the island, shelter and breeding ground within
easy reach of food, and no opportunity for dispersion. There may be starvation
and famine, and a greatly reduced reproductivity, but it cannot be induced by the
fly’s own activities, unless (1) conditions of life are locally favourable to mncrease,
and (2) dispersion from the locality is arbitrarily restricted.

X. GENERAL CONCLUSIONS.

There are three specific requisites to life of Glossina palpalis, one sustentative and
two protective :—good shelter and breeding ground. One kind of food, the blood
of vertebrates, suffices for the insect in all its stages. The two kinds of protection
serve adults and pupae respectively.

All these appear equally requisite. There is some slight evidence that other food
than blood may be taken, but none concerning its nature, and no evidence that any
other is required. There is much evidence that blood is absolutely required—as
supplied by the experiment cited in Sect. IT (6), and the effect upon behaviour of flies
of depriving them of their normal food supply ; by the apparent reduction in rate
of reproduction and in density of fly on Kimmi Island associated with obvious
shortage of food (Sect. IX (h)); by the long-shore distribution of fly in the Bwendi
district as described in Sect. VI (g); by the distribution of fly in the Mujuzi Creek
district (note to Sect. III (e)), and in various other cases. No one of these cases,
of itself, would establish the fact, but the cumulative evidence is all to one effect,
and the mass of it convincing.

The absolute necessity for the kind of protection that is provided by breeding
grounds is particularly well shown by the fly survey of Bugaba Island (Sect. VI (g) and
fig. 7). This was confirmed by similar surveys of Bubembe and Bunyama Islands,
- which have not been included herein for the sake of brevity, and by observations
many times repeated. It is not requisite that the breeding grounds should be in sand
or gravel—fine, dry vegetable debris will serve as well; but it happens that the sand
and gravel breeding places are most extensive in this particular region, and more attrac-
tive than any other, so that very much the larger number of flies breed in them.

A surprise was the discovery that “ massive shelter ” is requisite to the life of this
fly. This discovery was entirely unexpected, and for a long time it seemed incredible
that absence of massive shelter should be the explanation for absence or relative
scarcity of fly on the islands cited in Table XX XIII (p. 428). Cumulative evidence
finally dispelled all doubts—notably that supplied by the study of the Lutoboka
Bay colony (fig. 9, p. 431); by conditions along the southern shore of Buganga
Peninsula (Table XXXIV, p. 432); and various other less striking cases not cited
in these pages. There could be no doubt, in the end, that the light shelter that is
most attractive to the active flies, and which also serves in the majority of cases
to shade breeding grounds, is insufficient to protect the inactive flies from destructors,
animate or inanimate, as the case may be.

454 W. F. FISKE.

Apart from these things the insect appears to require nothing except that which
is inseparable from an equable, tropical climate. It is believed to be pure coincidence
that the insect is never found far from water, and it is confidently believed that it
would exist, and that it will be found eventually to exist, in any inland localities
where host animals of favoured species occur in well sheltered areas provided
with suitable breeding places. In fact the interior infestation on Mbugwe Island,
described on page 384 and mentioned in several other connections, is believed to be
a true inland colony, such as might occur equally well at any distance from lake
or river under a comparably favourable (and extremely unusual) combination of
circumstances.

The three requisites bemg equally indispensable, it follows that in the absence
of any one the species cannot exist. All three must be present in proper combina-
tion, and local density is likely to be governed by the quantity or amount of
whichever one of the three is least abundantly provided :—

(a) by amount of food, if protection is more than sufficient, relatively ; (b) by
amount of shelter, if food and breeding ground are both more than sufficient,
relatively ; and (c) by amount of breeding ground, if food and shelter are both
more than sufficient, relatively.

This is indeed the case. In the Bwendi district (Sect. VI (g) ) shelter and breeding
grounds were both very good and extensive, but food was very scanty and density of
fly low; along the southern shore of Buganga Peninsula, (p. 432, Table XXXIV)
food and breeding grounds were both good and plentiful, but shelter was poor or
lacking, and density of fly was therefore low; on Bugaba Island (p. 424) both
food and shelter were good and plentiful, but breeding grounds were poor, or few,
and density of fly was low. These are merely specific instances selected to illustrate
the general rule, which was upheld everywhere.

A very important question is: What constitutes a proper combination of the
requisites to life? It is obvious that they must occur within the radius of movement
and perception of individual insects, from a central point, or from each other; but
this does not answer the question. What is the radius of movement and perception—
or it may be called the radius of range or reach—of individual insects? This |
question can only be answered empirically, and unfortunately it did not occur
to me until nearly the close of the investigations to attempt to answer it. Careful
review of field notes, supplemented by memory of conditions not accurately described,
leads to the tentative conclusion, that unless all these requisites to life occur within
a radius of less than 100 yards from a central point, the conditions of life are so
unfavourable that the species cannot exist.

This does not mean that territory deficient in one or more of the requisites to life
is never infested. On the contrary, in the major portion of the infested territory
conditions of life are unfavourable to the existence of the species, and these portions
owe their infestations to their contiguity to territory in which the life requisites
occur in proper combination, and which thus becomes a centre of infestation for
a considerable zone surrounding. Within these centres natural increase of fly
from generation to generation exceeds mortality, but in addition to loss through
mortality there is loss through emigration or dispersion of flies into the surrounding

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 455

zone where conditions of life are positively unfavourable. In this surrounding
zone loss through mortality exceeds gain through natural increase, but in addition
there is gain through immigration from the centres of infestation. A perfect
“natural balance” is thus established and perpetually maintained, so long as flies
are free to disperse from the infested centres into the surrounding zone. If dispersion
is interfered with, the mechanism of this balance is thrown out of gear; but this
very rarely happens in nature so far as I am aware it failed to function in this
region only on a few small islands, such as Kimmi, where the conditions of life
appear to be so favourable that super-density results at frequent intervals, causing
the host animals to abandon the locality temporarily until the density of fly is reduced.
Here we have a natural balance constructed on an entirely different mechanical
principle, which operates as effectually, but not nearly so smoothly, as in the
majority of cases, in which it is the flies and not the host animals that are moved
to disperse from over-densely infested localities.

The réle played by natural enemies and destructive factors generally is to destroy
the insects, and to prevent the existence of the species everywhere beyond the limits
of dispersion from protected localities. The protection of shelter and breeding
grounds suffices against all destructive factors, animate and inanimate, except possibly
and very rarely, against extraordinary numbers of spiders or dragonflies. The
protective factors in the bionomics of the species are far more specific in their
nature, and easily studied, than the destructive. It is, therefore inadvisible to
study the destructive factors specifically. It is possible to utilise them for controlling
the density of the pest easily enough by merely depriving it of its specific protection
against them, but altogether impossible to expect them to penetrate protected
precincts, or to utilise them otherwise than in the manner indicated by the facts.

It is possible to argue successfully that the species is normally controlled in range
and density by the amount of available food, provided that the word “ available ”’
is broadly enough interpreted. Food, to be available, must be within reach of the
insects from protection (of both sorts). It is also necessary that the host animals
should be complacent under attack, for there is no host of the insect which cannot
protect itself against attack if it will. Monkeys are “ unavailable” as hosts, not
because their blood is less suitable as food, nor because the flies are unable to draw
it, but entirely because the animals are so active in protecting themselves. Sheep
and goats are more complacent and less active in protecting themselves than
monkeys, but much less complacent and more actively self-protective than oxen
or antelopes. Men vary greatly in temperament, but the average man stands
between goats and sheep on the one hand and oxen and antelopes on the other.
The large reptiles are the most complacent of all, and most available as hosts,
despite the fact that the flies experience measurably greater difficulty in drawing
blood from them. Availability .of food thus varies greatly with the species, and
with the temperament of the individual host, and also with circumstances and
conditions of time and place. If large Tabanid flies are annoying oxen, they become
less available as sources of food ; if Glossina itself increases in local density, its host
animals become less available. This is strikingly exemplified by human behaviour.
If there are very few flies a man pays little or no attention to them, and they have .

456 W. F. FISKE.

relatively little difficulty in feeding; but if there are very many flies, the man
provides himself with one or two fly switches, and they have proportionately greater
difficulty in feeding.

Availability of food is also measured by the precise location of host animals with
respect to protection. If separated from shelter by as much as 20 yards of unsheltered
terrain, food may be entirely unavailable to the flies. This was proved by conditions
along certain reaches of shore in Buddu, where there was abundance of crocodiles
along a sandy beach separated by from 20 to 100 yards of grass land from shelter and
breeding grounds. The flies could not perceive the hosts at this distance, and they
do not freely range so far from shelter in search of food.

The factors which operate naturally and control the range and density of Clasgae
palpalis in this region are, mainly, as described above; the factors which can be
operated by human effort to reduce obnoxious density to within innocuous limits
are identical. It is useless to contemplate destructive control measures. We can
see from the experiment and observations made on the “ spider islands” that it
requires an immense amount of destruction by specific agency to produce a
measurable effect on the density of the species. The spiders on Wema Island actually
destroyed enormous numbers of flies, but for the most part these flies would have
been as promptly destroyed by any one of various other agencies in the absence
of the spiders. If the spiders are about, the flies increase to slightly greater density,
and then, through wider dispersion, run greater risks and the natural balance is
struck at a slightly higher level. Or, stating the same truth in another way, the main
effect of artificial or unduly heavy natural destruction of flies by any specific agency
isto make conditions of life much more favourable to the rapid multiplication of the
survivors. It is wholly impractical to consider any control measures involving
artificial destruction of flies, and wholly necessary to rely upon measures designed to
deprive the insects of either food or protection or to render food less available to them.

For reasons in part set forth in Sect. II preceding it is regarded as inadvisable
and even dangerous to contemplate control of the pest through depriving it of food.
It is probable that if completely deprived of all favoured hosts it would be unable to
exist on hosts favoured to a no greater degree than sheep, goats and man, but it is
probable that it would continue to exist if cattle or pigs were provided. It is also
certain that where favoured hosts are plentiful, man is almost immune to attack, but
that when they are few man is freely attacked, and our object must always be to
protect man rather than to destroy flies.

There is no such objection to the proposition of controlling fly through depriving
it. of protection, and it is on measures designed to this end that we must chiefly
rely. They are the clearing measures already in use, and they have been proved
efficacious on many occasions. The maximum of economy and efficiency is to be
gained through clearing at precisely the right points—i.c., at the centres of infestation
wherein natural increase of fly ismost rapid. By clearing these the dispersion of flies
into the surrounding zone is prevented, and the effect is general. By clearing in
the zone infested by immigrating flies the effect is local at best. A small amount of
clearing at the centres of infestation may be much more effective than a large
‘amount of clearing away from them.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. ADT

It is useless, however, to discuss ways and means for reducing obnoxious density
of this pest until some decision has been reached concerning what constitutes obnoxious
density. It has been shown by three investigations that the density of infestations
by this insect, as measured by the males to be caught “ per boy per hour,” ranges
all the way from less than 0-1 to more than 150-0. A locality isnolesstruly infested
if density is 0-1 than if it is 150°0—the difference is wholly one of degree. We know
that a density of 0-1 to 1-0 is innocuous, because density in populated districts along
shore appears never to have been reduced below these figures, despite which human
trypanosomiasis formerly prevalent—when density was much higher—seems
completely to have disappeared. ‘There is good reason to believe that a density of
6-0 is ordinarily safe and sanitary, but that one of 15:0 or 20-0 would ordinarily be
dangerous. Admittedly it would be desirable for purely sanitary reasons to reduce
density to 0 everywhere, but the expense would be so enormous as to render it
completely impracticable. Practical measures must combine efficiency with economy,
and the most economical measures are such as do not carry reduction in density
of the pest beyond what is necessary for sanitary reasons.

XI. Errecr oF Fiuctruatine Laxe LEVEL oN GLOSSINA PALPALIS.

In Sect. VII (e) mention was made of the effect of fluctuating lake-level on range
and density of Glossina palpalis. Following are data bearing on this point, which
could not well be included in that section, and from which certain conclusions may
be drawn, as presented.

XI (a). Fluctuations in Level of Victoria Nyanza.

Victoria Nyanza is usually highest in May and lowest in November, with a tendency
to rise in December and to fall in February. The average yearly fluctuation is about
11 feet. But the movements are very irregular, and in 1906, instead of falling in
February it rose steadily from November 1905 until the following May, and reached
an unprecedented height. Again in 1910 it failed to rise in December, but continued
to fall steadily until it reached an unprecedentedly low level in February 1911. The
maximum in May 1906 was about 44 feet above the minimum in February 1911,
and was nearly 3 feet higher than the maximum from 1911 to 1913 inclusive.

The fluctuations during the ten year period ending in 1913 are shown on a
graph recently published by Dr. H. Lyndhurst Duke (Bull. Ent. Res. EX, p. 270).

XI (6). Effect of Fluctuations on Shelter and Breeding Grounds of Tsetse.

The high level of 1906 uprooted much vegetation along the shores and washed
clean the sand and gravel beaches that in 1913-14 lay some 3 feet higher than and
anywhere up to 200 yards inland from the foreshore.- Following recession of the
water, new vegetation sprung up to shade the newly washed sands and gravels, and
ideal breeding grounds were created where none, perhaps, existed before. But
already in 1913 and 1914 these new breeding grounds were beginning to deteriorate.
Leaf-mould was accumulating and covering the sand, vegetation was massing above it
and keeping it cold and damp; and already some of these ibreeding places had
become second-rate or quite unattractive. They were fast becoming like the deposits
of sand and gravel that lay beyond the flood mark of 1906, which are nearly always
hidden from sight beneath blankets of mould and quite valueless as protection to tsetse.

458 W. F. FISKE.

These were the principal changes that were taking place in the conditions of life
for tsetse along the lake shore as a result of the flood tide of 1906 and subsequent
At some places the changes were of a very different character, and more

low water.
In general,

kinds of changes were progressing than can well be described in detail.
however, it can be stated that wherever the shore rises very gradually, and most
particularly at the many points where the gradient is less than 1 per cent., extra-
ordinary changes in the conditions of life for tsetse are likely to follow any such
unusual flood as that of 1906. But if the land rises sharply from the water the changes
may be insignificant, and usually are of no consequence.

In a general way the island survey had shown that the range and density of fly
must have been profoundly affected by the movement of the water, but the real
extent of the changes wrought was not appreciated until, in 1915, a fly survey was
made of the Buddu shore. This region had been surveyed in August and September
1906 by Dr. Van Someren, whose reports were in considerable detail, and permitted
the quite accurate comparison which follows.

XI (c). Increase in Density of Glossina palpalis due to falling of Lake Level.
The lake shore of the district of Buddu was surveyed for fly in 1906 by Dr. R.
Van Someren, working under the direction of Dr. A. D. P. Hodges in the Sleeping
Sickness Extended Investigations. This same shore was surveyed by me in 1915,
and a most extraordinary change in conditions of infestation was found to have
occurred. Following is an extract from the MS. report of Dr. Van Someren, kindly

supplied me by Dr. Hodges, paralleled by comments on the present conditions.

Conditions im 1906.
(Report of Dr. R. Van Someren.)

‘When I came to Bale, a little south
of Namirembe I found the distribution
again interesting.

‘““Northwards from Bale I found the
whole coast some 3 miles up to Namirembe
devoid of fly, and*from Namirembe for
some further 2 to 24 miles, roughly, into
the Gwamba also with no fly.

“After this point one meets them at
first very scantily indeed; then in in-
creasing numbers right up to our limit in
the Gwamba some 8 to 10 miles from
Namirembe.

“For some 3 miles southward from
Bale no fly are met with till a small point
near the landing place Sekwe (Calcosa)
close to Dumo, where I came across them
in small numbers and limited practically
to this point.

Oonditions in 1915.

(Fly Survey.)

Northwards from Bale the whole coast
to Namirembe was found infested to the
average density of 6°3. At Namirembe it
was very low, but about half a mile beyond
the landing it began to increase rapidly,
until at points from 2 to 24 miles beyond
it had reached the very high average of
63°5. I have never but twice seen fly
more numerous or more annoying (see
fig. 11).
After this point the density rises to yet
higher figures (76°5, 73°5, 89°5, etc.) and
then falls away to moderate figures (5:3,
7°5, 5°5, etc.) but only to rise again later.
The average for this 6 to 8 miles of shore
is the very high one of 37°6.

For some 5 (not 3) miles southward from
Bale to Sekwe the shore is continuously
infested, and at 3 miles (2 miles north of
Kalkosa or Sekwe) is the centre of the
Mujuzi Creek colony described on p. 387
and illustrated by figure 5, with the
extreme density of 98:0.

NVorth

TE TH
ae es

ro Er
aE First Fly 1906

line in {915.

on |
7
y
a i] py
€
j F;
a
me afl syed
7
= ? 1
)
rc
}
A

7)
ar
heck
M: st hs
ao

Scale of Miles _| 2 3 4 5 6 7 8 9 10 I 12

Namirembe

Fig.1l. The dotted line shows the occurrence and density of Glossina palpalis in 1906, and the solid line in !91I5.

Pe ae

oe | ay i felt ie

oe es i > . * ¥ dade es tae ee eee eee ene
r 5 . v
t :
: wy | i
j f ‘
} ;
{ f
ve i“ ¥ " a a eer —_ Te

4 5 ‘ ‘
ti } j Peer
) ' ;
ine a
; }
be ; '
rf ' i \
ge % - mae . « ~—_ i ~ tn -
‘ '

} rt - , %
Ps t ~ o ~ Wy wre pte
' : —
f ;
4 i
} ; :
o; * « ° ~ r - _ ——
| 4 a y
j : . d
; ( bt mo ti

ryote sare of roe f a a a eetilic
« . y
+
. i‘ ; }
} d \
Ai j >
{ Py v f
rs meaner s eetaemie sg ng Aeron, mtg ets, Seal
i | : |
~ eee lion : wi
i aad
bb “ [ha Le NF os
te a eS A a
, i anem,
hom ne § 4 d w pe ii
iy “ngs “4 ye | ea “wren ign “nso At gered
’ °

: i _ a iy , ae

“ay aah unm . PiVAS ‘Tewih ain ae cee” oma. i

o
- et I, fel ae wae 00 olmak

AS), SPREE aw (Grow Le. a qe thee Sy wi 1 abel ee

,
la q
‘
3 =
bane a =
+
i
ag r
f te ’
P| i
eR
‘ fae eI 2
‘ : es a
4 ; hi Read
i) i
t Ny wile
, r/o
“a ie R m
4
‘
*
Z
¢ ny Cs
% 4 ;
Aa
4 Vv ey
= +, / A |
ete 4 ‘ 7 4

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS.

Conditions in 1906.
(Report of Dr. R. Van Someren.)

“While on one side of the point we
caught 8 on the two occasions we waited
for them, on the rocks on the other side
some 10 yards [rods?] across, where
several men and boys were sitting fishing,
we did not observe a single fly about their
persons, and amongst a group of some 20
drawing a seine a little further along we
caught only one the second day.

“The physical conditions in these sec-
tions, starting from our limit in the
Gwamba southwards, are as follows :—

“A sandy beach extends practically the
whole way, except close to Sekwe, where
the trees are practically at the water’s edge.

“The land behind is low and covered
in parts with scrub; in others with
larger trees with here and there wide
open areas intervening again. At some
portions the tree-covered areas are swampy
some distance back from the beach.

“At the fly limit north of Namirembe
this swampy part comes quite close té
the beach and remains so more or less
continuously till close to Bale, and con-
tinues thus practically to Sekwe, except
for portions here and there where the
Swamp line again comes close to the
beach. In many parts the beach is not
so high that during high winds the waves
wash right over into the swamp.

(659)

459

Conditions m 1915.
(Ely Survey.)

The density at these points (Sekwe)
averaged 9°2 in 1915—probably not far
different from that observed by Dr. Van
Someren in 1906, except that it applies
to the shore in both directions from the
rocks. (These are the only rocks for
miles in either direction and identify the
locality absolutely.)

The physical conditions in these sections
have undergone radical changes owing to
the falling of the level of the lake.

A sandy beach is practically continuous
the whole way except for the reach of two
miles north of Sekwe. Here the beach
of 1906 is separated from the present shore
line by a belt of marsh, very soft and
muddy, into which one sinks knee to thigh
deep. Outside the marsh are the beginnings
of a new beach, very narrow and wet.
Outside of this is a hedge of ambatch, or
merinde, that only permits passage of
canoes inward to the land at a few points.
The old beach of 1906, thus completely
cut off from the present waterline, is over-
grown with an excessively dense tangle
of vegetation. This continues to a point
“close to Sekwe’’ where the trees are still
‘‘practically at the water’s edge,’ and
(the shore being rocky or falling away
steeply) the conditions are virtually un-
changed, either physically or, as nearly
as can be judged, with respect to infestation
by fly. .

These conditions are unchanged; there
is more of swampy land back from beach
than Dr. Van Someren’s notes might imply.

The falling water wrought most radical
changes in the conditions here described.
The beach of 1906 is now high and dry,
and there may be 20, 30, 50 or in some
places 100 yards of dry sandy embank-
ment between the present. beach line and
the tree or reed grown swamps and marshes
back of it. This almost continuous em-
bankment is partly open in places, and in
other places overgrown with dense jungle.
Instead of a narrow wet beach with high
waves breaking over it, we have a broad
sandy embankment with swampy areas
“some distance back from the beach,”
and providing ideal shelter and breeding
ground for the fly. It should be added
that in 1915 somewhat higher water than
for the past few years was working still
further changes ; building new beaches and
cutting off new bits of newly made marsh
from the open lake at certain points, and

460

Conditions in 1906.
(Report of Dr. R. Van Someren.)

‘““A more or less continuous line of
merinde trees occurs some 5 to 15 yards
off the shore.

‘““I formed the opinion therefore that
the absence of fly from the point north of
Namirembe until Bale could be accounted
for by absence of suitable breeding places
owing to the conformation mentioned.

‘“‘ Southwards of Bale however the beach
is narrower and fringed with small bushes
and apparently suitable for fly.

“I was surprised, therefore, to find

none, and cannot account for it.

“This part of the coast is being gradually
encroached upon owing to the action of
the waves killing the trees and bushes by
washing away the soil, which is chiefly sand.

“I repeatedly examined these fly-free
areas under favourable weather conditions,
but the results were uniformly negative.

Note on Sex disparity.

“As regards the relative number of
male and female flies I have evidence
which I think goes to support Dr. Bag-
shawe’s observations (No. 42, §.S.E.I.) as
‘I have had most favourable opportunities
for the purpose.

“On three occasions with the aid of
the fly boys we examined simultaneously
that part of the Gwamba where numerous
fishermen were daily working the sambas,
and another part some miles beyond the
last net where practically only passing
canoes may chance to camp the night.
The results were :—

Frequented Area.
Ist day 52¢ to 589
2nd , 40¢ to 379
3rd_ ,, 24¢ to 299
Unfrequented Area.
Ist day 75 to 409
2nd. ,, 113¢ to 689
3rd_ , 956 to 379

W. 8S. FISKE.

Conditions im 1915.
(Fly Survey.)

cutting away the embankment at others,
so that the young trees and jungles growing
upon it were being undermined and thrown
into the lake. |

The new shore line cannot be less than
20 or 30 yards, average distance, outside
the shore line of 1906. The old line of
merinde trees has mostly disappeared,
with only here and there a few straggling
survivors or decaying trunks. A new
line, continuous or nearly so for some 3
miles north of Sekwe, has sprung up, and
at other points a new line is forming.

Certainly the conditions described could
not possibly provide good breeding places ;
neither would the shelter be suitable.
Either cause would suffice in explanation.

Southward from Bale the beach is now
from 20 to 100 yards from bush or forest
with open grass land intervening. Then
for rather more than a mile thick jungle
grows to the very edge of the water.
Beyond this it is open or semi-open to
Mujuzi Creek.

In 1915 this reach of shore appeared
to be foodless, and the flies dispersing
into it from the great colony at Mujuzi
creek were plainly famishing (88°8 per
cent. of females at one point, see p. 373).

This part of the coast is now (1915)
being added to, and the conditions described
by Dr. Van Someren prevail at points
north of Bale, as noted above.

—

The catch of fly made over approximately
this same territory in 1915 consisted of
768 males to 591 females = female per-

‘centage of 43°9.

This compares with the following figures
of percentages from Dr. Van Someren’s
report. :

Frequented area :—Males 116 to females
124 = female percentage of 51°7. |

Unfrequented area :—Males 283 to females
135 = female percentage of 32:1.

Total catch both areas :—Males 399 to
females 259 = percentage of females of
39°2,

Dr.Van Someren does not state the number
of fly boys employed or time spent in
catching, but it is interesting to note that
the flies are about as hungry in the general
region now as they were in 1906, there
being a difference of but 4°7 points between
the figure for the total catches at the two
widely separated periods.

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS.

461

As will be seen by recerence to the graph (fig. 11), the increase of fly in the region

south of Bale and just north of Namirembe has been phenomenal.

Below Sekwe,

however, the reverse has taken place, as shown by comparison between conditions
in 1915, and as they were described by Dr. Van Someren in 1906 in his report for

that region.
Conditions in 1906.
(Report of Dr. R. Van Someren.)

“The only fly met with (in the region
south of Sekwe) occurs along the shore
from a point near the village of Dumo
northward for about one and one-half
miles to the Kirala River, and gradually
tailing off from this river toward Sekwe.

From Dumo toward the river the fly
is very abundant.

“The range inland (between Dumo and
the Kirala river) is practically nil, as the
extent. of dry land at the lake-side is not
more than five yards in width at the
broadest part. Inside this dyke is exten-
sive dense forest of kibo palm (Raphia)
down to the lake-side, standing in very
swampy ground.

“Opposite Dumo, where the ground
rises fairly sharply the fly crosses a belt
of scrub and long grass intervening between
lake and village and is found in the planta-
tions immediately bordering the belt. I
frequently saw them on women digging
in this part.

Conditions in 1915.
(Fly Survey.)

The fly still occurs from the same point
near the old site of the village of Dumo
northward, but its density is notably
changed at several points.

The fly is hardly what would be called
“very abundant” at any point, but it
is not at all abundant from Dumo toward
the river. On the contrary, it falls off
to a low minimum before the river is
reached, and then begins to rise slowly,
Owing to migration from the colonies at
Sekwe and Mujuzi Creek. The infestation
at the Kirala River now consists mainly
of flies coming in from Sekwe, instead of
being at the edge of a centre of infestation
with fly tailing off toward Sekwe.

The extensive dense forest still exists,
and is still exceedingly swampy, the water
at the back of the dyke standing at a higher
level than the lake. But instead of
coming down to the lake-side there is a
belt of exceedingly soft marsh, with grass
and a little papyrus lying outside the old
beach, from which the open lake is all but
completely hidden. The dyke, or beach
of 1906, is now a terrible tangle of shrubbery
weighted down with rank vines, and with
marsh grasses growing rankly along its
outer edge. Outside the marshy belt,
10 to 30 yards from the old beach line of
1906, a new beach was forming in 1915.
Range inland at this point is noted on
page 408. Flies were found on a situtunga
shot 400 yards from the lake shore.

At the very end of the fly beach described
by Dr.Van Someren, where the ground begins
to rise fairly sharply, is found the remains
of the large colony he describes as existing
in 1906. A few hundred yards south the
belt of scrub and tall grass—elephant grass—
still separates the lake from what were the
old plantations. There is now exceedingly
dense jungle with occasional openings kept
clear by the hippo and permeated with
a net work of hippo trails. The fly still
penetrates the belt of tall grass, following
the hippo trails, which have replaced the
foot-paths leading to watering places, and
it is now found inland, in the very centre
of the old village site (see p. 408 note on
inland range at Dumo),

462
Conditions in 1906.
(Report of Dr. R. Van Someren.)

“Just beyond Dumo there is again a
forest area for some half mile. Where
the forest ceases until the German boun-
dary there is no fly to be found.

“Of the flies caught in this area 187
were males and 72 females (= 28 per cent.
females).

‘** At Dumo by one of the watering places
I caught 20 females to 14 males (=59
per cent. females).

W. F. FISKE.

Conditions in 1915.
(Fly Survey.)

The fly stops at a point only 100 yards
beyond that indicated by Dr. Van Someren.
But there are several colonies south of the
point.

The total catch in this area was 215
males and 184 females, making 46 per cent.
females.

At the identical watering place (the
only one in the fly area) two catches were
made on different days as below :—

**On another occasion along the forest

up to the Kirala River my boys caught Female
106 males to 2 females (=1:9 per cent. Males. Females percentage.
females). This is of interest in connection 15 4] 73%
with the physical features. Since fibre 19 om 15%
collecting ceased there practically no one — —

visits it. 34 98 74%

Along the forest up to the Kirala River
the catch of males was identical, 106, but
that of females was 86, making 45 per cent 9.

Again it is interesting to note that sex ratio shows that the watering place,
which is (or was) just outside the “ belt of scrub and long grass,” is now, as it was
in 1906, unattractive to the idle males, and is the point south of the colony centre
where female percentage is highest. But northward, along the beach where
* practically no one” went in 1906, the conditions are absolutely changed.

The foregoing brings out clearly the need for eternal vigilance if any systematised —
effort is ever made to repopulate the riparian zone and islands, and to protect the
population against recurrence of sleeping sickness. There is no more dangerous
reach of shore in Africa, perhaps, than that between the points Sekwe and Bale,
yet it was perfectly safe apparently in 1906, and the changes have been wrought
by the falling of the lake level.

Wherever the shore falls gradually into the lake, similar changes are likely to
occur at any time, especially if there is much sand in the district. The danger
points can be found by a “ fly survey” of the shore, and must be watched to
preclude the chance of conditions eminently favourable to spread of the human
parasite coming about unknown to sanitary or administrative officials.

XI (d). Gauses for Fluctuation in Lake Level.

Causes for the irregular fluctuation in level of Victoria Nyanza are, in part,
its great size and narrow outlet. From so great an area loss through evaporation
must be enormous, and relatively constant as compared with the overflow. The
latter represents the surplus of rainfall over evaporation, and if evaporation is a
relatively constant quantity, any variations in rainfall are likely to be greatly
exaggerated. Thesurplus not being so very large, it does not require much to double it,

INVESTIGATIONS INTO THE BIONOMICS OF GLOSSINA PALPALIS. 463

The narrow outlet to the enormous volume of the lake acts like a choked safety
valve. Rise in lake level may be very rapid, but fall is more gradual. It seems
to have required more than a year for the unusual surplus that gathered so
quickly in 1906 to find its way through the walls of rock at the source of the
Victoria Nile. If the outlet were not choked in this manner the level of Victoria
Nyanza would be somewhat more constant, and the flow of the Victoria Nile
somewhat more variable. If it were yet more effectually choked the flow of the
river would be more constant and the level of the lake subject to yet wider
fluctuations.

XI. (e). The probable Effect of a Dam at the Outiet upon the Tsetse along the
Shores of Victoria Nyanza.

The outlet to the lake into the Victoria Nile is a superb source of water-power,
and one of Uganda’s great assets. The topography permits its relatively cheap
development. There is a natural dam, and excellent opportunity for a canal along
the eastern shore of the river, which does not immediately fall away. But a dam
would permit a much more complete utilisation of the power that it is as certain
to be built as that the economic development of the country will continue.

It is neither a high nor a low level of the lake that creates favourable conditions |
_ of life for tsetse, but a fluctuating level. The rising waters clear away massed
vegetation and wash clean the breeding grounds ; the falling waters permit the growth
of new shelter and the breeding of fly in the renovated terrain. Continued growth
chokes the breeding grounds, and hides them beneath massed vegetation or the
accumulations of mould.

If a dam at the outlet of the lake could be so constructed as to equalise the level
of it by permitting more rapid egress of rising waters, it might have a beneficial
effect in reducing the density of tsetse along its borders. But the object in building
a dam would be the very opposite; it would be to equalise the flow of water in the
power canal and the Victoria Nile, and thus to accentuate the fluctuations of the
lake.

This might, conceivably, be prejudicial to the fly below the dam, but it would
be beneficial to fly above it.

It would be delightful if construction of a dam, and utilisation of the water-
power at the source of the Nile could be urged as a sanitary measure, but unfortunately
its construction will have the opposite to the desired effect.

ant
‘

NOD Me

; ‘ Mes.

‘465

COLLECTIONS RECEIVED.

The following collections were received by the Imperial Bureau of Entomology
between 1st October and 31st December, 1919, and the thanks of the ee
Committee are tendered to the contributors for their kind assistance :—

Mr. E. Ballard, Government Entomologist, Madras :—56 Tabanidae, 8 ou
Diptera, and 13 Lepidoptera ; from South India.

Capt. P. J. Barraud, Entomologist to the Beyptian Expeditionary Force :—d2
Culicidae, 6 Tabanidae, 49 other Diptera, 13 Hymenoptera, 102 Coleoptera, 3
Planipennia, | species of Coccidae, 3 other Rhynchota, 4 Orthoptera, and 9 Odonata ;
from Palestine.

Mr. C. Beeson, Forest Zoologist :—93 Braconidae and 157 Curculionidae ; Aon
Dehra Dun, India.

Mr. G. E. Bodkin, Government Economic Biologist :—8 Tabanidae, 72 other
Diptera, 28 Hymenoptera, 200 Coleoptera, 4 Planipennia, 1 species of Coccidae, a
number of Aphids, 16 other Rhynchota, 12 Orthoptera, 22 Ticks, 133 Spiders, 4
Scorpions, 3 Centipedes, and 6 Intestinal worms ; from British Guiana.

Mr. John R. Bovell, Superintendent of eee —144 Diptera, 37 Parasitic
Hymenoptera, and 2 Lepidoptera ; from Barbados.

Dr. 8. L. Brohier, W.A.M.S. :—11 Lepidopterous pupae: from the Gold Coast. 3

Mr. J. H. Burkill, Director of the Botanic Gardens, Svraits Settlements :—13
Agromyzid Diptera and 8 Coleoptera ; from Singapore.

Mr. P. A. Buxton :—11 Culicidae, 31 Tabanidae, 16 Coleoptera, 380 Aphididae,
and 3 species of Coccidae ; from Persia and Mesopotamia.

Dr. G. D. H. Carpenter :—2 Chrysops, 19 other Diptera, 7 Lepidoptera, 170
Coleoptera, 3 Planipennia, 70 Hymenoptera, 2 male Coccidae, 75 other Rhynchota,
19 Orthoptera, and 2 Odonata; from Uganda.

The Divisron of Entomology, Pretoria :—22 Hymenoptera, 303 Coleoptera, 26
tubes and 107 slides of Aphididae, and 72 other Rhynchota ; from South Africa.

_ Dr. Lewis H. Gough, Government Entomologist :—24 Diptera, 52 Hymenoptera,
57 Coleoptera, 12 Planipennia, 6 Rhynchota, and 19 Orthoptera ; from Egypt.

The Government Entomologist, Punjab :—33 Diptera ; from Lyallpur.

Mr. K. Hargreaves :—17 Culicid larvae, 11 other Diptera, 40 Parasitised dipterous
puparia, 250 Thrips, 150 Parasitic Hymenoptera, 2 Coleoptera, 51 Coleopterous
larvae and pupae, 250 Lepidopterous larvae, 5 species of Coccidae, and 54 Ticks ;
from North America, etc. ; 61 early stages of Pieris rapae ; from England.

Hawaiian Sugar Planters’ Association :—87 Weevils ; from Honolulu.

Dr. J. F. [lingworth :—90 Coleoptera and 2 Rhynchota ; from Queensland.

Imperial Department of Agriculture for the West Indies :—3 Weevils, and 1
pupa; from Curacao.

Imperial Institute :—3 Melipona bees; from Colombia; and 6 Rhynchota;
from Dakar.

Dr. Jawaera, Jewish Health Bureau :—12 Culicidae; from J erusalem.

466 COLLECTIONS RECEIVED.

Mr. 8. Leefmans, Instituut voor Plantenziekten, Buitenzorg :—4 Weevils ; from
Java. : :

Dr. J. W. Scott Macfie :-—304 Culicidae, 10 Pericoma, 11 Stomoxys, 3 tubes of
small Diptera, 3 Lepidoptera, and 20 small Leeches ; from the Gold Coast.

Dr. J. M. O’Brien :—1 Queen termite ; from the Gold Coast.

Rev. Father J. A. O’Niel, S.J. :—7 Moths ; from Salisbury, Rhodesia.

Mr. Arthur W. J. Pomeroy :—4 Tabanus, 20 other Diptera, 93 Hymenoptera,
69 Coleoptera, 3 Planipennia, 21 Lepidoptera, 10 Mallophaga, 9 Cimicidae, 8 other
Rhynchota, 38 Orthoptera, 121 Ticks, 2 Chelifers, 7 Spiders, 10 Scorpions, and 12
Entozoa; from British E. Africa and the Tanganyika Territory; 8 Tabanus, 6
other Diptera, 3 Coleoptera, 2 Lepidoptera, 2 Rhynchota, and 2 Orthoptera ;
from Southern Nigeria; 12 Nycteribiidae, and 35 Mallophaga; from Kamerun ;
and 10 Tabanidae, 338 other Diptera, 2 Chalcids, 138 other Hymenptera, 62
Lepidoptera, 3 Trichoptera, 70 Coleoptera, 57 Rhynchota; from U.S.A.

Lieut.-Col. W. Rainey, Government Veterinary Officer :—1 Bot-fly ; from Fiji.

Dr. W. Roepke :—50 Lepidoptera ; from Java.

Dr. A. T. Stanton :—79 Culicidae, 1 Phlebotomus, 2 Haematopota, 4 Tabanus, 12
Stomoxys, 6 other Diptera, 2 Hymenoptera, and 3 Rhynchota; from the Federated
Malay States.

Mr. Robert Veitch :—84 Diptera, 119 Lepidoptera, 117 Coleoptera, 67 Hymenop-
tera, 1 slide of Thrips, 25 Rhynchota, and 19 Orthoptera ; from Fiji. 7

467

GENERAL INDEX.

Numerals in heavy type indicate that the species is illustrated on that page.

Aberia caffra, Coccids on, 121.
abstersa, T'rypanea.

Abyssinia, Perilampsis pulchella in,
233.

Acacia, 80, 112.

» caffra, new Coccid on, 97.

» horrida, Coccids on, 116.

» karroo, Coceids on, 125, 126.
Acanthiophilus helianthi, 213.

EY ? ochraceus, 213.

accensus, Tabanus.
Acidia, 240.

» caesio, 245-247.

» fossata, 245-247.

» jfossataeformis, sp. n., 212,
245-247.

» homogenea, sp.n., 212, 246—
247.

» seychellensis, 212, 245.
Acura, 213, 251.

» angusta, 213, 253-255.

» vbinaria, 213, 253.

sis caeca, 213.

» capensis, 213, 256.

» oborinia, 213, 255-256.

» perspicillaris, 213, 253-255.

» semiangusta, 213.

» sexfissata (see Tephrella).

» ternaria, 213.

» tetrachaeta, 213, 256.
Aclerda berleset, 183.
Acroceratitis, 223.

Acrocercops bifasciata, a cotton pest
in Zanzibar, 148.

Acromyrmex sp., enclosing Coccids in
paper nests, 179, 190, 191.

acrodiauges, Conradtina.
acroleuca, Conradtina.
acrophthalma, Camaromyia.
acrosticta, Spathulina.

actinella, Tachardia.

aculeata, Montrichardia.

adatha, Hlaphromyia.

adersi, Lecanium ; Rhynchocyon.

Adoretus versutus, attacking sugar-cane
in Fiji, 29-30.

Aédes curriet (see Ochlerotatus).

Aédomyia catasticta, in Far Eastern
ports, 339.

affluens, Tachardia.
africana, Lecaniodiaspis.

Africa, South, Coccidae of, 95-128,
fruit-flies from, 230, 232, 238, 253,
262; injurious weevils from, 273-276.

Africa, West, early stages of mosquitos
in, 56-69, 134-135.

africanus, Dacus; Mansonioides ; Phle-
botomus minutus var.

ager var. ethiopicus, Culex.

Agromyza, on pigeon-pea in Zanzibar,
151.

atra, Trypanea.

alaskaensis, Theobaldia.

alata, Xanthorrhachista.
albida, Tachardia (Tachardina).
albidipennis, Oxycarenus.
albifacies, Tabanus.

albipes, Tabanus (see T’. gigas).
albocephalus, Ochlerotatus.
albolineata, Stegomyia.
albomaculatum, Trirhithrum.
albonigrum, Trirhithrum.
albopicta, Stegomyia.
aldabrensis, Spathulina.
alexandrinus, Tabanus.

Aleurodes comata, a minor pest of
Sugar-cane in Fiji, 38.

Algeria, mosquitos from, 134.
aliena, Pardalaspis.

Allotrypes brevicornis, gen. et;sp.n.,
213, 251-253.

468 GENERAL

Alternanthera sessilis, Coccids on, 103.

amabilis, Mausoleopis.

Amaryllis, insects injurious to, in
Zanzibar, 154.

ambigua, Coryna.

ambiguus, Chionaspis.

amoena, Trypaneda.

Amorphococcus, 108.

Ampera intrusa, gen. et sp. n., in stored
rice in Java, 41.

amplectens, Mylabris.

amplicollis, Rhina.

amplifrons, Huaresta.

Anagrus, egg-parasite of Perkinsiella
vitiensis, 37.

Anastrepha, 225.

ss extranea, 225.

anastrephina, Clinotaenia.

anceps, Hnsina.

angusta, Acvwra.

annulata, Theobaldia.

annulatus, Dacus.

annuliferus, Mansonioides.

annulioris, Culex.

annulipes, Mansoniordes.

Anomala orientalis, controlled by Scolia
manilae, 30.

Anona, Coccids on, 105.

anonae, Pterandrus.

Anopheles barbirostris, in Far Eastern
ports, 334, 339-342.

costalis, in N. Gold Coast,
49-52, 54-57; early
stages, 60-62; in Zan-
zibar, 329, 330.

fuliginosus, in Far Eastern
ports, 334, 339.

fuliginosus var. nivipes, in
Far Eastern port, 334.

funestus, in Gold Coast,
52-57; early _ stages,
59-61; in Zanzibar,
329-331.

hyrcanus, synonomy of, 129.

kochi, in Far Eastern port,
| Sou:
ludlowi, in Far Eastern
ports, 339, 340.
marshalli, early stages
60-61.

INDEX.

Anopheles mauritianus, in Gold Coast,
54-57 ; early stages, 62;
in Zanzibar, 329, 331.

as minimus, in Far Eastern
port, 337.

a pharoensis, early *stages of,
60-61.

E pictus (see A. hyrcanus). -

Bs plumbeus, 91.

igh pretoriensis, in N. Gold

Coast, 49, 54; early
stages of, 59-61.
pseedopiclus (see A. hyr-

caunus).

53 pulcherrimus, in Mesopo-
tamia, 324.

- rossi, 340, 341; (see A. sub-
pictus).

oe vosst var. indefinitus, in
Far Eastern ports, 334-
DO 1s ON, otlwete.

rufipes, in N. Gold Coast,
54 ; early stages of, 61-62.

ms sinensis, 129, 324; in Far
Iuastern ports, 334-339,
342; (see also A. hyrca-

nus).

- smmensis var. mesopotamiae,
a)
Aya

sinensis var. peditaeniatus,
in Far Eastern port, 334.

= squamosus, in Zanzibar, 329,
Sol.

a stephensi, in Mesopotamia,
323.

subpictus, synonomy of, 129;
(see also A. ross?).

tessellatus, in I‘ar Eastern
ports, 334, 339.

antalus, Deudoria.

antennata, Meracanthomyia.

antipodeus, Ochlerotatus.

antistictica, Pardalaspis.

Antonina australis, 185.

socialis, 185.

waterstoni, sp. N., from
Macedonia, 182-185.

Ants, attacking pupae of Glossina, 440—
441; associated with Coccids, 179,
190, 191.

aonidum, Aspidiotus.

Apanteles sp., bred from Outirphis
unipuncta, 34.

GENERAL INDEX. 469

Aphis tavaresi, on Citrus in Zanzibar,
he.

apicalis, Celidodacus (see C. obnubilus) ;
Culex.

apicatus, Aspidiotus (Chrysomphalus).

Apple, Coccids on, in 8. Africa, 111.

arabica, Mansonia (see Ochlerotatus
caspius).

arabicus, Culex (see
caspius); Tabanus.

aranoides, Rachionotomyia.

arctica, Theobaldia.

Ochlerotatus

areniwagus, Tabanus.
argentea, Stegomyia (see S. fasciata).

Argyroploce leucotreta, infesting man-
darins in Zanzibar, 152.

armatus, Tridacus.

Armigeres obturbans, in Far Eastern
ports, 335, 337-339.

articulatus, Aspidiotus (Selenaspidus).

Arundo phragmites, new Coccid on, in
Macedonia, 184.

Ashanti, domestic mosquitos in, 52, 58.
Asphondylia capparis, galls of, 232.

Aspidiotus aonidum, on rose in
Zanzibar, 154.
AP (Chrysomphalus) apicatus,
sp.n., on Avicennia in
Brit. Guiana, 195-196.

a (Selenaspidus) articulatus,
on cottee in Uganda, 197—
198, 200.

. (Selenaspidus) articulatus
var. magnospinus, 188,
197-198.

| oe citricola (see Lepidosaphes
pinnaeformis).

5 cyanophylli, on coconut in
Zanzibar, 147.

ee destructor, on mango and
coconut in Zanzibar, 147,
152.

a (Chrysomphalus) dictyos-
permi, on mango in
Zanzibar, 152.

2a fiorineides, sp.n., on coffee
in Uganda, 199-200.

= (Selenaspidus) kamerunicus,
198.

a lataniae, on coconut in

Zanzibar, 147.
a3 longispina, 194-195.

Aspidiotus maskelli (see A. longispina).

on (Chrysomphalus) paulistus,
197.

3 perseae, 197.

3 pinnaeformis (see Lepido-
saphes).

. (Odonas pis) rhizophilus,
sp.n., on Chloris in
E. Africa, 198-199.
ee (Pseudaonidia) trilobiti-
formis, on Citrus, 152;
on Ficus elastica, 153.
5 (Chrysomphalus) wmboni-
ferus, sp.n., on Lecythis
in Brit. Guiana, 196-197.
Aspidistra, Coecids on, 104.
aspidistrae, Chionaspis (Hemichion-
aspts).
Aspidomorpha puncticosta, on sweet
potato in Zanzibar, 151.
Aspidoproctus gowdeyi, sp.n., on
plumbago and rose in Uganda, 177.
Asterodiaspis pustulans (see Astero-
lecanium).
s variolosus (see Astero-
lecanium variolosum).

Asterolecanium, 108-115, 119.
- bambusae, 112-1138.

5 borboniae, sp.n., on
Borbonia in 5%.
Airica, lik

e brevispinum, sp. n., in
S. Africa, 109.

Sy coffeae, 109.

x conspicuum, Sp. D., on
Acacia in §. Africa,
112.

3 euryopsis, 113-114.

2 pustulans, 111.

ne quereicola (see A.
variolosum).

3 stentae, sp.n., in S.
Africa, 114-115.

- _thesvi, 115.

et variolosum, 110.

Asura saginaria, on cinnamon trees in
Zanzibar, 153.

ataeniata, Culex (see C. univitiatus).
aterrimus, Tabanus.

atomosa, Marasmarcha.

atrox, Brachycerus.

470 GENERAL INDEX.

Atylotus (see Tabanus).
aucta, Trypanea.

Aucuba, Coccids on, 108.
auguralis, Trypanea.
aurantapex, Culex.

auratilis, Chionaspis.
auripilus, Tabanus aterrimus.
Australia, Rioxa termitoxena from, 1—5.
australis, Antonina.
autumnalis, Tabanus.
aucxtiliaris, Mutilla.

Avicennia nitida, new Coccids on, 186,
196.

Avocado Pear, Dinoderus minutus
attacking bark of, 153; termites
attacking roots of, 153.

azadirachtae, Lecaniodiaspis.

Baccacoccus elytropappi, gen. et sp. n.,
127-128.

badius, Phyllodinus.

Baikea eminti, new Coccid on, 178.
bakeri, Perkinsiella ; Rhabdochaeta.
ballardi, Phenacoccus.

Bamboo, Coccids on, 97, 113; Dino-
derus minutus a pest of dried, 153.

bambusae, Asterolecanium.

bancrofti, Filaria.

Banksinella lineatopennis, in Far
Eastern ports, 339.

barbirostris, Anopheles.

barkeri, Lophoceratomyia.

Baryglossa histrio, 212.

beckit, Lepidosaphes (Mytilaspis).

bellicosus, Termes.

Bembex, destroying Glossina, 440, 446.

benefactrix, Mutilla.

berlesei, Aclerda.

bezziana, Tephrella.

bicolor, Rioxa ; Trypeta ; Urophora.

bifasciata, Acrocercops.

bigemmatus, Ocneros.

bimaculata, Sturmia.

bimaculatum, Trirhithrum.

binaria, Aciura.

binotalis, Crocidolomia.

binotatus, Dacus.

bioculata, Spathulina.

bioculatum, Notomma.

bipartitus, Tridacus.

bipunctatum, Eutretosoma.

Birchippia (see Lecaniodiaspis).

Birds, and Glossina, 391, 399.

bistrigatus, Dacus.

bistrigulatus, Dacus.

Bistrispinaria, 212, 221, 223-226.
me fortis, 212, 224-225.

bitaeniorhynchus, Culex.

bivittatus, Tridacus.

blepharogaster, Dacus.

Blighia sapida, 53.

boas, Oryctes.

Borbonia cordata, new Coccid on, in §,
Atriea, 111.

borboniae, Asterolecanium.
braber, Lecaniodiaspis.

Brabeium stellatifolium, new Coccid on;
in 8. Africa, 118.

Brachycerus atrox, on leaves of Ama-
ryllis in Zanzibar, 154.

Brachyplatys pacificus, life-history of,
in Fiji, 37-38.

Brachystegia, 80.

Bradybamon, 276.

brasiliensis, Chionaspis (see COC. aspi-
distrae).

Brazil, new Coccid from, 176.

Breeding cage, useful insect, 43-44.

bréemei, Pardalaspis.

brevicornis, Allotrypes ; Pulvinaria.

brevifrons, Pliomelaena.

brevipalpis. Cyathomyia ; Glossina.

brevis, Dacus.

brevispinosa, Cremastogaster.

brevispinum, Asterolecanium.

brevistylus, Dacus.

breviusculus, Lyprops.

bridwelli, Delphacodes ; Dicranotropis.

Brithys pancratii, on lilies in Zanzibar,
154.

British Guiana, Coccids from, 179, 185—
187, 189-191, 194-197.

broadwayi var. echinopsidis, Pulvinaria.
bromeliae, Pseudococcus.

broquetti, Grabhamia (See Ochlerotatus
cwrriet).

GENERAL INDEX. AT]

Bruchus chinensis, in stored beans, 155.

Ha ornatus, in stored beans, 155.
brunnipennis, Tropidocephalus.
brunithorax, Phorellia.

Bucculatrix loxoptila, on cotton in
Zanzibar, 148.

Buffalo, and Glossina, 387-391, 398,
400, 434.

Bupalus piniarius, on pines in Sweden,
157-160.

Bush-buck, and Glossina, 382, note,
385-387, 391, 396-397, 408-409,
434, 447.

Bush-pig, and Glossina, 382, 383, 385,
387, 391, 397, 408, 434.

Busseola fusca, on millet in Zanzibar,
149.

caeca, Aciura.
caesio, Acidia.
caffra, Chionaspis ; Euribia.

Cajanus indicus, Pseudococcus obtusus
on, in Zanzibar, 150; Bruchus
chinensis in stored, 155.

caladiui, Pentalonia.

Caladium, Hntyposis impressa on, in
Zanzibar, 154.

calamistis, Sesamia.

Calandra oryzae, common in stored
products in Zanzibar, 155.

caliginosus, Ochlerotatus.
Calodendron capense, Coccids on, 121.
Calotropis procera, Coccids on, 105,
Camaromyia acrophthalma, 213.

35 2 helva, 213.
campestris, Uranotaenia.
Campiglossa perspicillata, 213.
Campodea, 296. i
cancer, Uranotaenia.

capensis, Aciura; Chionaspis (see C.
lounsburyi) ; Duomitus.

capitata, Ceratitis.
Capparimyia, gen. nov., 232.
Capparimyia savastani, 222, 232.

Capparis spinosa, Capparimyia savas-
tant making galls on, 232.

capparis, Asphondylia.

Capsicum, Acanthocoris fasciculatus on,
150.

Caralluma caudata, new Coccid on, 115.
(693)

carbonatus, Tabanus (see T. alexan-
drinus).

Carissa grandiflora, Coccids on, 103.
carissae, Chionaspis (Poliaspis).
Carpomyia incompleta, 213.

Carpophilus humeralis, in maize cobs
in Zanzibar, 155.

Carpophthoromyia formosula (see Peri-

lampsis).
pseudotritea, 212.
pulchella (see Peri-

lampsis).
scutellata, 212.
superba, 212.
tritea, 212.

e vittata, 212.

Carteria decorella (see Tachardia).
caspius, Ochlerotatus (Culex).

Cassava, little attacked by insect pests
in Zanzibar, 151.

cassiae, Chionaspis.
castaneum, Tribolium.

Castor Plant, insects injurious to, in
Zanzibar, 154.

Casuarina equisetifolia, Cirina forda
on, 158.

catasticta, Aédomyia.
catoirt, Ceratitis.
cautella, E'phestia.
Celidodacus apicalis (see C. obnubilus).
.: coloniarum, 212, 221.
: obnubilus, 212, 220-221.
cephalia, Xanthorrachista (see X. alata).

Cerataphis lataniae, on palms, 147; a
pest of Citrus, 152.

Ceratitis capitata, 212; in Zanzibar, 152.
catoiri, 212.
inscripta (see Trirhithrum).

rosa, in soursop in Zanzibar,
152.

3 savastant (see Capparimyia).
Cercopithecus, and Glossina, 398.
Ceriops candolliana, Termes bellicosus

attacking timber of, 153.
Cerococcus, 108.
ornatus, 120-121.
ovoides, 119-120.
45 passerinae, Sp. n., 121, 122.
royenae, sp. n., 121-122.
B

472 GENERAL

Ceromasia sphenophori, 25.

Ceroplastes rubens, on Citrus in Zanzi-
bar, 151.

cervinus, Gyponychus.

Chaetachme aristata, Coccids on, 103,
108.

Chelyophora, 212, 221, 223.
. magniceps, 212.

Chilo suppressalis, on maize and millet
in Zanzibar, 149.

chinensis, Bruchus.
chionaspiformis, Diaspis.
Chionaspis ambiguus, sp. n., 97-98.
5s aspidistrae, 104.
a auratilis, sp.n., 204-205.

A brasiliensis (see C. aspidis-
trae).

55 caffra, sp. n., 97.

55 capensis (see C. lounsburyt).

ne carissae, 102-103.

ss cassiae, 96.

ah chionaspitiformis, 96.

ae coronifera, 205.

x cyanogena, 103.

Ne diosmae, sp. n., 101.

ss distincta, 102.

ne distorta (see C. distincta).

Ay dura, sp. u., 205-206.

a herbae, 202.

s imbricata, sp. n., 100.

a laniger, sp. 0., 206—207.

a latus (see C. aspidistrae).

a leucadendri, sp. n., 98.

a lounsburyt, 99, 101.

« 4 var. ekebergiae,
ne. 90%

+ madiunensis, 200-202.

5 natalensis, 100.

iis praelonga, sp. n., 203-204.

he proxvma, 104-105.

ue scutiae, sp. n., 95.

a simplex, 96-97.

re spneaeie E sp.n., 202-

chionaspitiformis, Chionaspis (Hemi-
chionaspis).

Chloris incompleta, new Coccid on, 199.

chlorogramma, Trachycentra.

INDEX.

Chrotogonus hemipterus, on castor in
Zanzibar, 154.

Chrysomphalus (see Aspidiotus).

Chrysops punctifera, common in
Palestine, 278-280.

ciliatus, Dacus.
cinnamomeus, Simodactylus.

Cinnamon, Asura saginaria on, in
Zanzibar, 153.

Cirina forda, on Casuarina equisetifolia.
in Zanzibar, 153.

Cirphis loreyi, on maize and millet in
Zanzibar, 149; on sugar-cane
in Fiji, 35.
» wunipuncta, on sugar-cane in
Fiji, 34.
citrt, Pseudococcus.
citricola, Mytilaspis (see Lepidosaphes
pinnaeformis).
Citrus, Coceids on, 107, 108; insects.
on, in Zanzibar, 151-152.

Civet Cat, and Glossina, 399.
Ciaius, 177.
Cladoderris, 212, note.

Clinotaenia anastrephina, g. et sp. n.,
212, 225-227.

39 grata, 212, Dos
Cloves, insects on, in Zanzibar, 145.
Clypeococcus hempeli, 175-177.

Coccidae, monograph of 8. African,
95-128; observations on, 175-207.
Coccus viridis, on Citrus in Zanzibar,

ay IE
Coconut, Oryctes on, in Zanzibar, 145—
147; Termes bellicosus destroying
nuts of, 147.
Coelaenomenodera costulata, 171.
ee elaeidis, Sp.0., on
oil palm in Gold
Coast, 171-174.

nS signifera, 171.
>» speciosa, 171.

5 thomson, 171.
OCoelopacidia madagascariensis, 212,
216-218.
sis melanostiqma, sp. u., 212,

218-219.
uy strigata, sp. u., 212, 218-
219.

Coffea robusta, new Coccids on, in.
Uganda, 188, 200.

GENERAL INDEX. Ailes

coffeae, Asterolecanium; Trirhithrum
(see 7’. inscriptum).

Coffee, Coccids on, 180, 188, 200.
colae, Pterandrus.

coloniarum, Celidodacus.

comata, Aleurodes.

comstocki, Pseudococcus.

concolor, Culex.

confluens, Trypanea.

conjuncta, Huaresta.

Conradtina, 212, 216, 220.

Le acrodiauges, 212, 219.
rs acroleuca, 212.

oe longicormis, 212.

af suspensa, 212.

consimilis, Culex.

conspicuum, Asterolecanium.
cordiger, Tabanus.

cornuta, Tachardia (Tachardiella).
coronifera, Chionaspis.

Corvus scapulatus, feeding on Oryctes
larvae, 147.

» splendens, feeding on Oryctes
larvae, 147.

Coryna ambigua, on pigeon-pea in
Zanzibar, 150.

Cosmopteryx, sp. n., on sugar-cane in
Hii; 36.

Cossonus suturalis, in sweet potatoes
in Zanzibar, 155.

costalis, Anopheles.

costulata, Coelaenomenodera.

cosyra, Pardalaspis.

Cotton, insects attacking, in Zanzibar,
148-149.

Craspedoxantha, 213, 243.
55 manengubae, 213.
he marginalis, 213.
crassicornis, Haematopota.

Cremastogaster brevispinosa var.
tumulifera, new Coccid in nest of,
182.

cristata, Hoplolopha.

Crocidolomia binotalis, on cabbage in
Zanzibar, 149.

Crocodile, and Glossina palpalis, 373>
376, 379, 387, 391-393, 401-402,
404, 407-408, 410, 411, 415, note,
417, note, 424, 429, 432, 434, 435,
452-456.

(693)

Croton, Coccids on, 106.
Cryptaspidiotus, 200.
Crypticerya hempeli (see Olypeococcus).

Cryptorrhynchus mangiferae (see Ster-
nochetus). )

cucumarius, Tridacus.

Cucurbitaceae, pests of, in Zanzibar,
150.

Culex ager var. ethiopicus, in Gold
Coast, 49, 54-57 ; early stages
of, 65-68.
» annulioris, larva of, 49.
» apicalis, 134.
» arabicus (see Ochlerotatus caspius).
» ataeniata (see C. univittatus).
» aurantapex, 134-1385.
» bitaeniorhynchus, 135; in Far
Eastern ports, 339-341.
», bitaeniorhynchus var. ethiopicus,
HB }5s
» casprius (see Ochlerotatus).
» concolor, in Far Eastern ports,
334-336.
» consimilis, larva of, 49.
» currier (see Ochlerotatus).
. decens, in Gold Coast, 49, 50, 53—
Dil.
», diversus (see Ochlerotatus rus-
ticus).
» dorsalis (see Ochlerotatus caspius
and O. curriet).
» duttoni, in N. Gold Coast, 49-58 ;
notes on, 136.
» fatigans, in Accra, 50; -not
recorded from N. Gold Coast,
51; in Mesopotamia, 324; in
Far Eastern ports, 334-342.
» fuscocephalus, in Far Eastern
ports, 334, 339, 342.
» gelidus, in Far Eastern ports,
334-336, 339.
» halifaxi, in Far Eastern ports,
Se pas
» hortensis (see C. apicalis).
invidiosus, in Gold Coast, 49-57.
» maculiventris (see Ochlerotatus
caspiws).
minimus, in Far Eastern ports,

341.
modestus, in Mesopotamia, 324.

» onondagensis (see Ochlerotatus
currier).
B2

474. GENERAL INDEX.

Culex penicillaris (see Ochlerotatus cas-
pus).
>, prpiens, in Mesopotamia, 324.
>, pulchripalpis (see Ochlerotatus

caspiws).

>, punctatus (see Ochlerotatus cur-
rev).

> pungens (see Ochlerotatus rus-
ticus).

5s pyrenaicus (see C. apicalis).
,, gquadratimaculatus (see Ochlero-
tatus rusticus).

5, guasigelidus, in Gold Coast, 54-
56; early stages of, 66—68.

» Tusticus (see Ochlerotatus).
», saxatilis (see C. apicalis).

, sitiens, in Far Eastern ports,
334, 336-338, 341.

» territans (see C. apicalis).
,, testaceus (see C. apicalis).

,, tigripes var. fuscus, in Gold
Coast, 49-58.

, tipuliformis, im Mesopotamia,
324.

» trifilatus, 135.

, tritaeniorhynchus, in Gold Coast,
54; in Mesopotamia, 324; in
Far Eastern ports, 339, 340,
342.

» univittatus, in Gold Coast, 55;
early stages of, 68-69;
C. ataeniata asynonym of, 136.

» vishnui, in Far Eastern ports,
339, 341, 342.

» whitmoret, in Far Eastern ports,

339.
watti, sp. u., from Gold Coast,
135-136.
Culicoides, in Gold Coast, 47, 51, 53,
55-58.
os grahami, rare in Gold Coast,
Bl. b2: im Ashante: 52,
58.

Culiciomyia nebulosa, in Gold Coast,
49-58; in Ashanti, 52, 58.

curriei, Ochlerotatus (Aédes).

cyana, Ewribia.

cyanogena, Chionas pis (Hemichionaspis).

cyanophyll, Aspidiotus.

Cyathomyia brevipalpis, in Far Eastern
ports, 335, 336, 339.

cycadis, Poliaspis.

cyclopica, Tephrella.

Cylas formicarius, in sweet potato in
Zanzibar, 151.

Cyrtacanthacris guttulosa, in Fiji, 39.

Dactylopius (see Pseudococcus).
Dacus, list of African spp. of, 211.
» brevistylus, in Zanzibar, 150.
» punetatifrons, in Zanzibar, 150.
» vertebratus, in Zanzibar, 150.
dalei, Tabanus.
danica, Locusta.
debeaufortii, Rioxa.
decemplex, Lecanium (Hucalymnatus).
decens, Culex.
decora, Trypanea.
decorella, Tachardia (Carteria).
decorus, Tabanus.
deformosum, Lecanium (Hulecanium).
Delphacidae, new African, 139-144.
Delphacodes bridwelli, sp. n., 140-141.

a disonymos, 139-140.

se erectus, 139.

35 BS nigrupennis, 139.

3 matanitu (see D. dis-
onymos).

¥ nigeriensis, sp. n., 140.

Vemmerezi, Tridacus.
demodocus, Papilio.

Dermestes vulpinus, abundant in dried
fish, 155.

derogata, Sylepta.
desmotes, Stegomyia.
destructor, Aspidiotus.

Deudorix antalus, on pigeon pea in
Zanzibar, 150.

diaphasis, Platensina.
Diarrhegma, 1.
Diaspis chionaspiformis (see Chio-
NAS Pts).
dicincta, Mylabris.
Dicranotropis bridwelli, sp. un., 141.
see ibadanensis, sp. n., 141—

dictyospermi, Aspidiotus
phalus).

digitifrons, Walkeriana.
digressa, Thalassoides.
Dinaspis (see Chionaspis).

(Chrysom-

GENERAL

Dinoderus minutus, a timber pest in
Zanzibar, 153.

Diocalandra frumenti, on coconut husks,
147.

Diosma crenata, new Coccid on, in S.

Africa, 101.

diosmae, Chionaspis.
Diospyros mashuna, 80.
discipulchra, Huribia.

Discoha ovalauensis,
Rhopaea vestita, 28.

discreta, Ocnerioxa.

disjunctus, Tridacus,

disonymos, Delphacodes.

dissoluta, Huribia.

distuencta, Chionaspis.

distorta, Chionaspis (see C. distincta).
diversa, Trypanea.

diversus, Culex (see
rusticus).

Dog, and Glossina, 391,
Dombeya rotundifolia, Coccids on, 119.
domestica, Musca.

domesticus, Taeniorhynchus (see Culex
aurantapex).

domimea, Khizopertha.
dorsalis, Ochlerotatus (see O. caspwus):
Dorset, Ochlerotatus curriet in, 131.

Dracaena australis, Coeccids on, in S.
Africa, 108.

Dragonflies, destruction of Glossina
by, 439-440, 444, 446.

dubia, Ensina.
Duiker, and Glossina, 391, 397.

Duomitus capensis, on castor plant,
154.

dura, Chionaspis.
duttoni, Culex.

Dysdercus fasciatus, on cotton in Zanzi—
bar, 148, 153.

superstitiosus, in Zanzibar,
148, 153.

EKarias wnsulana, on Hibiscus esculentus,
148. ;

East Africa, mosquitos from, 134, 136 ;
Coccids from, 180, 199; fruit-flies
from, 216, 236, 250, 255, 263, 264;
new Mutillid from, 327-328.

eburneus, Tridacus.
echinopsidis, Pulvinaria broadway.
Echinopsis latiflora, new Coceid on, 187.

parasite of

Ochlerotatus

INDEX. 475

echinus, Ochlerotatus (Finlayia).

Egg Plant, Acanthocoris fasciculatus on>
in Zanzibar, 150.

eggert, Tabanus.

Egypt, mosquitos from, 130.

Ekebergia, new Coccid on, 99.

ekebergiae, Chionaspis lounsburyt.

elaeidis, Coelaenomenodera.

Elaeis guineensis, Rhynchophorus phoe -
mcis on, 147; new Hispid injurious
to, 171-174.

Hlaphromyia adatha, 213.

oh melas (see EH. adatha).
“ ulula (see H. adatha).

Elenchus tenuicornis, parasite of Perkin-
siella vitiensis in Fiji, 37.

elytropappt, Baccacoccus.

Elytropappus rhinocerotis, new Coccids
on, 124, 128.

Ensina anceps, 213.
» dubia, 213, 261-262.
» gladiatrix, sp. n., 213, 261.
» tgnobilis, 213.
» magnipalpis, sp. u., 213, 261-
262.

» myropitoides, 213.
» stiphonina, 213.
» sororcula, 213, 261.

Entyposis impressa, on castor and
caladium in Zanzibar, 154.

Ephestia cautella, in Zanzibar, 155.

Epping Forest, Stegomyia fasciata in,
91.

erectus, Delphacodes.

Eremnus horticola, sp. n., on, chry-
santhemums in 8. Africa, 276.

Eriodendron anfractuosum, and Culex
larvae, 53; Dysdercus on, in Zanzi-
bar, 153 ; food-plant of Pectinophora
gossyprella, 148.

erythraeus, Dacus.

ethiopicus, Culex bitaeniorhynchus.

Euaresta amplifrons, sp. n., 213, 259-
260.

ee conjuncta, 260.
megacephala, 259-260.
P planifrons, 213, 259-260.

Euchromia formosa, on Ipomoea, in
Zanzibar, 154.

476 -_ GENERAL INDEX.

Huclea sp., new Coceid on, 126.
» natalensis, new Coccid on, 100.
Hugenia caryophyllata (see Clove).

Euproctis producta,not attacking cloves,
145; on cotton in Zanzibar, 148.

Euribia caffra, 213.

ua cyana, 213.

ae discipulchra, 213, 271.

Bs dissoluta, 213.

» perpallida, 213, 266, 271.

aS praetexta, 213.

Be tristrigata, 213.
Euryops tenuissimus, Coccids on, 114.
euryopsis, Asterolecanium.
Eutretosoma bipunctatum, 213.

5 frauenfeldi, 213.

Bs millepunctatum, 213.
a oculatum, 213.
» polygramma, 213.

evansi, Trypanosoma.
excellens, Ocneros.
eximius, Tropidocephalus.
extranea, Anastrepha.

fallacivena, Themara.

Far Eastern Ports, mosquitos of, 333—
344.

fasciata, Stegomyia.

fasciatus, Dysdercus.

fasciolata, Rhacochlaena.
fasciventris, Pterandrus rosa.
fatigans, Culex.

Ferns, Coceids on, in 8. Africa, 104.
ficicola, Dacus.

Ficus, harbouring mosquito larvae, 50,
55; and tsetse-flies, 80.

» elastica, Aspidiotus trilobitifor-
mis on, 153.

Fig, harbouring mosquito larvae, 50,
55; Coecids on, 120; and tsetse-
flies, 80.

Fiji, sugar-cane pests in, 21—39.
filamentosus, Pseudococcus.
Filaria bancrofti, 51.

#5 nocturna, 51.
filiformis, Ischnaspis (see I.longirostris).
jimbriata, Planchonia.
Fiorinia, 200.

fiorineides, Aspidiotus.

Fires, bush or grass, and Glossina, 444.

flava, Mytilaspis (see Lepidosaphes
gloveri) ; Rioxa. -

flaveolata, Themarictera.

flavescens, Mytilaspis (see Lepidosaphes
pinnaeformis).

flavicrus, Tridacus.

flavolineatus, Megamelus.

flexuosa, Pardalaspis.

Flood, Glossina destroyed by, 444,
457—458.

Forcipomyia, in Gold Coast, 52, 53.

> incomptifeminibus, in
Gold Coast, 57.

_ ingrami, in Gold Coast,
Bile

forda, Cirina,

Forest Insects, use of experimental
plots when studying, 157-160.

formicarius, Cylas.

formosa, Euchromia.

formosipennis, Rioxa.

formosula, Perilampsis (Carpophthoro-
mya).

fortis, Bistrispinaria.

fossata, Acidia.

fossataeformis, Acidia.

frauenfeldi, Eutretosoma.

frumentt, Diocalandra.

fuliginosus, Anopheles.

Fullawayella kirkaldyi, 45, 46.

as nigra, 46.
as tulipaella, 46.
. violae, 45, 46.

fulva, Mytilaspis (see Lepidosaphes
pinnaeformis).

funestus, Anopheles ; Protostrophus.

furcifer, Megamelus.

fusca, Busseola ; Stegomyia.

fuscatus, Tridacus.

fuscocephalus, Culex.

fuscovittatus, Tridacus.

fuscus, Culex tigripes.

gagatinum, Trirhithrum.

Game, relation of Glossina to, 7-20,
394-409.

Gardenia, Glyphodes sericea on, 154.

GENERAL INDEX. ATT

Gastrozona, 225.

gelidus, Culex.

geniculatus, Ochlerotatus.

giffardi, Pardalaspis (see P. cosyra).

gigas, Tabanus (Atylotus).

gladiatrix, Hnsina.

Glossina brevipalpis, in S. Rhodesia,
78.

5 morsitans, and ungulate game,
7-20; in Gold Coast, 58;
in §. Rhodesia, 71-90;
attracted bymoving objects,
278; new Mutillid parasite
of, 327-328 ; habits of, 347,
378, 406, 407, note; fire a
factor in bionomics of, 444.

ee pallidipes, and ungulate game,
8-18; in Gold Coast, 58;
in S. Rhodesia, 78.

os palpalis, and ungulate game,
8-18; in Gold Coast, 58;
bionomies of, 87, 347-463.

<6 submorsitans, 58.
glossinae, Mutilla.
glovert, Lepidosaphes (Mytilaspis).
Glyphodes_ sericea, on Gardenia in
Zanzibar, 154.
gnaphalii, Trypanea.
Goat, and Glossina, 360-361, 391, 399,
400, 401, 402, 405, 437, 455, 456.

Gold Coast, domestic mosquitos of
northern, 47—58, 61-69, 129, 136;
new Hispid injurious to oil palm in,
171-174; fruit-flies from, 240, 245,
266.

gossypiella, Pectinophora (Gelechia).

gowdeyt, Aspidoproctus.

Grabhamia broquetti (see Ochlerotatus

curriet).

. dorsalis (see O. caspius).

a longisquamosa (see O.
caspwus).

» subtilis (see O. caspius).

9 willcocksi (see O. caspius).

gracilis, Ocnerioxa.

grahami, Culicoides.

grata, Clinotaenia (Trypeta)..

gravis, Protostrophus (Strophosomus).
gubernatorius, Ochlerotatus.

guimari, Trypanea.

guttipennis, Lewcotaeniella.

guttulosa, Cyrtacanthacris. ae

Gymnospora buxifolia, Coccids on, 101.

Gyponychus cervinus, on Citrus in
Zanzibar, 152.

Haematopota crassicornis, 289.

55 innominata, sp.n., in
Palestine, 280, 290-
292. .

3 minuscula, sp. n., in
Palestine, 280, 287—
290, 292.

minuscularia, sp.n., in
Palestine, 280, 285—
287, 292.

és pallens, 289.

55 pluvialis, 287-291.

am sewelli, sp.n., in Pales-

tine, 280-285, 287,
288, 290, 291.

halifaxi, Culex.

Halimococcus lampas, 126-127.
hamatus, Dacus.

hammersteint, Rhacochlaena.
hargreavesi, Ochlerotatus caspius.
heliantht, Acanthiophilus.

helva, Camaromyia.
Hemichionaspis (see Chionaspis).
hemimelaena, Trypanea.
hemupterus, Chrotogonus.

hempeli, Clypeococcus (Crypticerya,
Icerya).

herbae, Chionaspis.

hessu, Tephrella.

Hexacinia stellata, 1.

hexapoda, Trypanea.

Hibiscus, Coccids on, 117, 119.

- esculentus, pests of, in
Zanzibar, 148, 151.
ae sabdariffa, Nisotra weiset on,

in Zanzibar, 151.

Hippocentrum trimaculatum, in Gold
Coast, 52.

Hippopotamus, and Glossina palpalis,
376, 379, 383, 385, 387, 391, 396, 405,
409, 410, 424, 432, 434, 445, 450.

hirsutus, Ochlerotatus.
histrio, Baryglossa.
homogenea, Acidia.

Hoplolopha cristata, g. et sp.n., 212,
222, 233-236.

478 GENERAL INDEX.

hortensis, Culex (see C. apicalis).
horticola, Hremnus.

Huernia transvaalensis, new Coccid on,
115.

humeralis, Carpophilus ; Tridacus.
Hyaena, and Glossina, 391.
Hyphaene crinita, Coccids on, 127.
Hyrax, and Glossina, 390, 399.
hyrcanus, Anopheles.

ibadanensis, Dicranotropis.

Icerya (Crypticerya) hempeli (see
Clypeococcus).

» purchasi, on Citrus in Zanzibar,
Vol.

» seychellarum, on lemons in
Zanzibar, 151.

Idiopterus nephrolepidis, 45, 46.
ignobilis, Hnsina.

imbricata, Chionaspis.
immaculatus, Dacus.
immisericors, Toxorhynchites.
impressa, Eintyposis.

incompleta, Carpomyia.
incomptifeminibus, Forcipomyia.
indecora, Tephritis.

indefinitus, Anopheles rossi.
India, new Theobaldia from, 137.
indica, Theobaldia.

ingrami, Forcipomyia.
imnominata, Haematopota.
inornatus, Dacus.

inquilina, Pseudophilippia.
inquilinum, Lecanium.
inquilinus, Pseudococcus.
inseriptum, Trirhithrum.
ansecutor, Tabanus.

insignis, Rioxa.

instabilis, Protostrophus.
insulana, Harias.

intermedius, Tabanus (see T. eggeri).
wntrusa, Ampera.

Inula crithmoides, Huaresta megacephala
making galls on, 260.

invidiosus, Culex.

Ipomoea batatas, Cylas formicarius and
Aspidomorpha puncticosta on, in
Zanzibar, 151; Huchromia formosa
on, 154.

Isaria sp., attacking Brachyplatys
pacificus, 38.

Ischnaspis filiformis (see ip longirostris).
ae longirostris, 107-108.

Italy, mosquitos from, 130; _ fruit-
flies from, 232.

Jamaica, Coccids from, 182.

Java, a new Lygaeid on stored rice in,
41.

jucunda, Trypeta.

Kamerun, fruit-flies from, 233.
kamerunicus, Aspidiotus (Selenaspidus).
karroo, Tachardia.

Kigilia pinnata, 80.

kingt, Dacus.

kirkaldyi, Fullawayella (Macrosiphum).
kochi, Anopheles.

kolophon, Megamelus.

lacca, Tachardia.

Lacon stricticollis, on sugar-cane in
Fiji, 31-33.

Laemophloeus pusillus, in stored maize,
155.

laetetinctus, Tabanus.
lampas, Halimococcus.
Lamproxyna, 257-259.

a: nitidula, 258.
laniger, Chionaspis.
lataniae, Aspidiotus ; Cerataphis.

lateralis, Ochlerotatus (see O. echinus) ;
Rioxa.

laticeps, Themarictera.

Latheticus oryzae, in stored maize, 55.
latus, Chionaspis (see C. aspidistrae).
Lecaniodiaspis africana, 118.

e azadirachtae, 118.
ee brabei, sp. n., 117-118.
55 magna, sp. n., 117.
os mimosae, 116.
<e natalensis, sp. n., 116—
LE
a tarsalis, 118-119.
Lecanium adersi, on mango in Zanzibar,.
152%
“5 (Hucalymnatus) decemplex,
sp. n., 188-189.
bd (Hulecanium) deformosum, sp-

n., 179, 190-191.

GENERAL INDEX. 479

Lecanium inquilinum, sp. n., 179, 189—

191.

nigrum, 191; on mango in
Zanzibar, 152.

a nigrum var. nitidum, n., 191-

192.

5p punctuliferum, on mango in
Zanzibar, 152.

5 quercus (see Asterolecanium
variolosum).

A subacutum, sp. n., 187-188,
200.

65 tarsale (see Lecaniodiaspis).

Lecythis sp., new Coccids on, in British
Guiana, 189, 197.

leleant, Tabanus.

Leopard, and Glossina palpalis, 391,
398, 399, 434, 446-448.

lepidonotus, Ochlerotatus.
Lepidosaphes beckii (see L. pinnaefor-

mis); a pest of Citrus
in Zanzibar, 151.

>» glovert, 106-107
oe pinnaeformis, 106.
leucadendri, Chionaspis.

Leucadendron argenteum, new Coccid
on, 98.

leucopsis, Trirhithrum.
Leucotaeniella guttipennis, sp. n., 212,

223:
oe pentaspila, 212.
NS trispila, 212, 223.

leucotreta, Argyroploce.

Lilac, new Coccid on, 98.

Lily, Brithys pancratii on, in Zanzibar,
154.

lineatopennis, Banksinella.

litura, Prodenia.

Lizards, and Glossina, 391, 394, 440.

Locusta danica, in Fiji, 39.

longiareolata, Theobaldia.

longicornis, Conradtina.

longivostris, Ischnaspis (Mytilaspis).

longisetosa, Tachardia.

longispina, Aspidiotus.

longispinus, Pseudococcus.

longisquamosa, Grabhamia (see Ochler-
otatus caspius).

longistylus, Dacus.

Lophoceratomyia barkeri, in Far
Eastern ports, 339. °

Loranthus entebbiensis, new Coccid on,
206.

loreyi, Cirphis.

lounsburyi, Chionaspis ; Tridacus.
loxoptila, Bucculatria.

luciensis, Stegomyia fasciata.

ludlowt, Anopheles.

Luffa acutangula, Dacus vertebratus on,

lugubris, Tephritis.

lunatus, Tabanus.

lunifera, Platensina.
Luzibarziba, new Coccid on, 192.
lyew, Trirhithrum.

Lyprops breviusculus, on pigeon-pea in
Zanzibar, 150.

macaensis, Polistes.

Macedonia, mosquitos from, 130, 133,
134; Coccids from, 184.

macer, Dacus.

Macrosiphum kirkaldyt (see Fullaway-
ella).

Macrotoma palmata, a timber pest in
Zanzibar, 153.

maculatissima, Porphyronota.

maculipennis, Plutella.

maculiventris, Culex (see Ochlerotatus
caspius).

madagascariensis, Coelopacidia ; Pseu-
dococcus virgatus.

madiunensis, Chionaspis.

magna, Lecaniodiaspis.

magniceps, Chelyphora.
magnipalpis, Hnsina.

magnospinus, Aspidiotus articuiatus.
Mahogany, African, Coccids on, 96.
maidis, Peregrinus.

Maize, pests of, in Zanzibar, 149, 155.
mamulae, Trypanea.

Man, and Glossina palpalis, 347-463.
manengubae, Craspedoxantha.

mangiferae, Sternochetus (Cryptorrhyn-
chus).

Mango, harbouring mosquito larvae,
50; Coccids on, in Natal, 100; insects
on, in Zanzibar, 152.

480 GENERAL INDEX.

Mangrove, Macrotoma palmata in, in
Zanzibar, 153.

Manihot utilissima, little attacked by
pests in Zanzibar, 151.

manilae, Scolia,

Mansomia arabica (see Ochlerotatus
caspwws).
Mansonioides africanus, 56, in Ashanti,

58.

= annuliferus, in Far
Eastern ports, 334,
337, 339.

oe annulipes, in Far
Eastern ports, 334,
Dok Bowe

ws uniformis, in Far

Eastern ports, 334,
336, 339; in Gold
Coast, 51, 56, 57.

Marasmarcha atomosa, on pigeon-pea
in Zanzibar, 150.

margaritifera, Spathulina.
marginalis, Craspedoxantha.
marginatus, Psewdokermes.

marshal, Anopheles; Perirhithrum ;

Taomyia.

maskellt (Aspidiotus, Morganella) (see
A. longispina).

Mastotermes, Rioxa termitoxena bred
from galleries of, 4.

matanitu, Delphacodes (see D. dis-
onymos).

mathias, Parnara.
mauritanicus, Tenebroides.
mauritianus, Anopheles.

Mausoleopis amabilis, on rose in
Zanzibar, 154.

media, Poliaspis.
megacephala, Huaresta ; Pheidole.
Megamelus flavolineatus, sp.n., 143-

144.
3 fureifer, 142.
x a var. nigeriensis,

nov., 142-143.
ie kolophon, 142-143.
melanaspis, Pardalaspis.

melanosticta, Miresa.
melanostigma, Coelopacidia.

melas, Hlaphromyia.

Meloe sp., larvae of, on Tabanus alex-
andrinus, in Palestine, 296.

mendicus, Tabanus.
M eracanthomyia antennata, 2 11, 214.
mesomelas, Dacus.

Mesopotamia, MOSqUILOS from, 130,
323-325

mesopotamiae, Anopheles sinensis.
metallica, Stegomyia.

Micromyzus nigrum (see Fullawayella).
Microparsus variabilis, 46.

miki, Tabanus.

millepunctatum, Hutretosoma.

Millet, insects on, in Zanzibar, 149.

Mimomyia mimomyiaformis, in Gold
Coast, 54, 56, 57.

e splendens, in Gold Coast, 54.
mimomyiaformis, Mimomyia.
mimosae,. Lecaniodiaspis

phora).

minimus, Anopheles ; Culex.
minor, Tachardia.
minuscula, Haematopota.
minuscularia, Haematopota.

minutus, Dinoderus; Ochlerotatus ;
Phlebotomus.

Miresa melanosticta, a minor pest of
African almond, 153.

(Prosopo-

mochit, Dacus.

modestus, Culex ; Dacus.
moebiust, Schistopterum.
momordicae, Tridacus.

Mongoose, and Glossina, 391, 405.

Monkeys, and Glossina palpalis, 391,
398, 405, 406, 437, 455.

monoceros, Oryctes.

Monocrepidius wpallipes, destroying
grubs of Rhopaea vestita in Fiji, 28,
31, 33-34.

Montrichardia aculeata, new Coccid on,
194.

montrichardiae, Platysaissetia.

Morganella maskel (see Aspidiotus
longispina).

Morocco, Mosquitos from, 134.

morsitans, Glossina.

morstath, Pardalaspis.

mundus, Ocneros.

Murraya, Coccids on, 106.

musae, Rioxa (Trypeta).

Musca domestica, 52.

GENERAL INDEX. 48]

Mutilla auailiaris, sp. n., parasite
of Glossina morsitans in
Portuguese E. Africa, 327—
328.

= benefactrix, 328.
» glossinae, 328.
mytopitoides, Ensina.

Mylabris amplectens, on pigeon-pea in
Zanzibar, 150.

. dicincta, on pigeon-pea in
Zanzibar, 150.

Mytilaspis beckii (see Lepidosaphes

pinnaeformis).
i, citricola (see L. pinnae-
forms).

» flava (see L. gloveri).

x flavescens (see L. pinnae-
forms).

at fulva (see L. pinnaeformis).

ae glovert (see Lepidosaphes).

At longirostris (see Ischnaspis).

oA pallida (see L. glovert).

a pinnaeformis (see Lepi-
dosaphes).

Naseberry, a food-plant of Cuvrina
forda in Zanzibar, 153.

Nasturtium, Crocidolomia binotalis on,
149.

Natal, fruit-flies from, 242, 260, 261,
270,

natalensis, Chionaspis; Huclea; Lecan-
todiaspis.
Neanomoea, 243.

neavei, Ptiloniola ; Rhabdochaeta ;

Spheniscomyia.
nebulosa, Culiciomyia.
nebulosus, Tridacus.
Necrobia rufipes, in dried copra, 155.
Nectandra, new Coccid on, 185.
nemoralis, Tabanus.
Neotoxoptera violae, 45-46.

Nephele pilipes, destruction of Glossina
by, 437-439, 444, 446, 456.

nephrolepidis, Idiopterus.
neuropteripenne, Rhochmopterum.
New Guinea, Rioxa from, 5.

New Zealand, mosquitos from, 132.

Nigeria, Delphacidae from, 139-144 ;
fruit-flies from, 219, 221, 223, 251,
259.

nigeriensis, Delphacodes; -Megamelus
furcifer ; Ochlerotatus.

nigerromum, Trirhithrum.

nigra, Fullawayella; Rioxa.
nigricosta, Tephrella.

nigripennis, Delphacodes erectus ; Rioxa.
nigronervosa, Pentalonia.

nigrum, Lecanium ; Micromyzus
(see Fullawayella) ; Trirhithrum.

Nisotra weisei, on Hibiscus in Zanzibar,
ote

nitens, Pseudokermes.
nitidula, Lamproxyna.

nitidum, Lecanium nigrum; Trirhi-
thrum.

nivipes, Anopheles fuliginosus.
nocturna, Filaria.

Notomma bioculatum, g. et sp. n., 212,
242-245.

noxius, Protostrophus.

Nyasaland, fruit-flies from, 219, 221,
227, IAG 247) 2405 256, Zot, 20.

nymphaeae, Rhopalosiphum.

obnubilus, Celidodacus.

oborinia, Aciura.

obscura, Rhabdocnemis.

obscurum, Trirhithrum (see T. nigrum).
obturbans, Armigeres.

obtusus, Pseudococcus.

occipitale, Trirhithrum.

ocellata, Taomyia.

Ochlerotatus albocephalus, 64.

ae antipodeus, sp. n., from
New Zealand, 132.

‘a caliginosus, 64.

Fy! caspius, 130.

of ie var. hargreavesi,
n., from Italy, 130.

a curriet, 130-131.

4 dorsalis, 130-131; in
Mesopotamia, 325.

sf (Finlaya) echinus, sp. 0.,
from Macedonia, 133—
134.

3 geniculatus, 91, 133-134.

v gubernatorius, in Saigon,
336,

- hirsutus, pupa of, 62-64.

a lateralis (see O. echinus).

482 GENERAL INDEX.

Ochlerotatus lepidonotus, sp. n., from
Macedonia, 132-133.

<5 minutus, 64.

Be nigeriensis, in N. Gold
Coast, 56; pupa, 63-64.

is ornatus (see O. echinus).

55 punctothoracis, 64.

35 rusticus, 131-133.

rs salinus, 130.

ochraceus, Acanthiophilus.
Ocnerioxa, 213, 248, 250-251.

a discreta, sp. n., 213, 251.

5 gracilis, 213.

a pennata, 213.

ue woodi, 213, 250.
Ocneros, 240.

Bs bigemmatus, sp. n., 212, 248,
250.

Ep excellens, 212, 248.

a mundus, 212,248.

59 sinuatus, 212, 248, 250.

ee undatus, 212, 248-250.
oculatum, Hutretosoma.
Odonaspis (see Aspidiotus).
oleae, Dacus.
Oleander, Coccids on, 111.

onondagensis, Culex (see Ochlerotatus
curriet).

Ooencyrtus pacificus, controlling
Brachyplatys pacificus in Fiji, 38.
Ootetrastichus, egg-parasite of Perkin-

stella vitiensis in Fiji, 37.
Orange, Coccids on, in 8. Africa, 106,
124; pests of, in Zanzibar, 151-152.

Orgyia vetusta, on cotton in Zanzibar,
148.

orientalis, Anomala.
ornata, Aspidiotus longispina.

ornatus, Bruchus; Cerococcus ; Och-
lerotatus (see O. echinus).

Ortaloptera, 3, note.
Orthezia, 116.

Orthopodomyia albionensis, in Epping
Forest, 91.

Oryctes boas, in coconut in yee ee
145-147.

pe monoceros, in coconut, 145-147.
oryzae, Calandra ; Latheticus.
Otter, and Glossina, 391, 405.

ovalauensis, Discolia.

ovoides, Cerococcus ; Pollinia.

Oxycarenus albidipennis, on cotton im
Zanzibar, 148.

Pachyura murina, 147.

pacificus, Brachyplatys; Ooencyrtus.

Palestine, mosquitos from, 130; Taba-
nidae of, 277—321.

pallens, Haematopota.

pallida, Mytilaspis (see Lepidosaphes:
glovert).

pallidipes, Glossina; Tabanus.

pallipes, Monocrepidius.

Palms, Coccids on, in 8. Africa, 19),
108, 127.

palmata, Macrotoma.

palpalis, Glossina.

Pamera, 41.

pancratit, Brithys.

Pangonia sulcata, 278, note.

Me: Coccids on, in British Guiana,

Papilio demodocus, on Citrus in Zanzi-
bar. "i52.

Paranagrus, egg-parasite of Perkin-
siella vitiensis in Bij, 37:

Parastenopa, 240.
parceguttata, Spathulina.
Pardalaspis, 212, 222, 228-232.

ay aliena, sp. nu., 212, 22%
231-232.

is antistictica, 212, 229.

As brémet, 212, 229.

“ cosyra, 212, 229-231.

by flexuosa, 212, 229.

55 giffardi (see P. cosyra).

Ss melanaspis, sp. 0., 212,
229-230.

bs morstatti, 212, 228.

5 punctata, 212, 229.

Bs quinaria, 212, 229.

6 senegalensis, 212.

ah silvestrii, 212, 229.

ee stictica, 212, 229.

Parinarium mobola, 80.

Parkia biglobosa, harbouring insane
larvae, 50, 51, 53.

Parnara mathias, on maize in Zanzibar,
149.

GENERAL INDEX. 483

Passerina ericoides, new Coccid on, 121.
passerinae, Cerococcus.
passerinit, Pycnocerus.
paulistus, Aspidiotus (Chrysomphalus).
pecorum, Trypanosoma.

Pectinophora (Gelechia) gossypiella, on
cotton in Zanzibar, 148 ; on Hibiscus
esculentus, 151.

pectoralis, Tridacus (see T. bivittatus).
penicillaris, Culex (see Ochlerotatus cas-
pius).
penicillatus, Pterandrus.
peditaeniatus, Anopheles sinensis.
Pentalonia caladu, 46.
7 Sh nigronervosa, 45, 46.
pennata, Ocnerioxa.
pennipes, Pterandrus (see P. anonae).
pentaspila, Leucotaeniella.
peregrina, Trypanea.

Peregrinus maidis, on imported maize
in Zanzibar, 149.

Perilampsis, gen. nov., 212, 222, 233.
formosula, 212, 233.
- pulchella, 212, 233.

Perirhithrum marshall, g. et sp. n.,
214, 266-270.

Perkinsiella sp., from Nigeria, 139.
= bakeri, 139.

vitiensis, controlled by egg-
parasites in Fiji, 37.

perniciosus, Pseudococcus.
perpallida, Huribia.

Persea gratissima, Dinoderus minutus
in bark of, 153.

perseae, Aspidiotus.
perspicillaris, Acwura.
perspicillata, Campiglossa.
pertinens, Tabanus.
Petrodomus sultan, 147.
pharoensis, Anopheles.

Phaseolus mungo, Bruchus chinensis in
stored, 155.

Pheidole megacephala, 23, 35.
Phenacoccus ballardi, 181.
Philophylla, 240..

Phlebotomus, in Gold Coast, 47, 51,
53-58; in Ashanti, 58.

minutus var. africanus,
in Gold Coast, 52.

phoenicis, Rhynchophorus.
Phorellia, 212, 240.
"7 brunithorax, 212.
be tristriata, 212.
Phyllodinus badius, sp. n., 144.
Phytometra signata, a minor pest in
Zanzibar, 149.
pictus, Anopheles.
Pig, Domestic, and Glossina, 361, 379,
381, 385, 391, 399, 401, 409.
Pigeon Pea, pests of, in Zanzibar, 150
pilipes, Nephele.
pimarius, Bupalus.
pinnaeformis, Aspidiotus; Lepidosa-
phes (Mytilaspis).
Pinnaspis (see Chionaspis).
pupiens, Culex.
Pistia stratiotes, relation of mosquito
larvae to, 54, 56, 57.
planatus, Protostrophus.
Planchonia (see Asterolecanium).
planifrons, Huaresta.
Platensina diaphasis, 213.
a lunifera, 213.
Platysaissetia montrichardiae, sp. n.,
192-194.
Pliomelaena brevifrons, 213, 259.
= strictifrons, 213.
Plumbago, new Coccid on, 177.
plumbeus, Anopheles.

Plutella maculipennis, on cabbage in
Zanzibar, 149.

pluvialis, Haematopota.

Poinciana regia, harbouring mosquito
larvae, 50.

Poliaspis carissae (see Chionaspis).
a cycadis, 102, 103.
ue media, 103.

Polistes macaensis, controlling Oirphis
unipuncta in Fiji, 34; life-history of,
36-37.

pollim, Pollinia.

Pollinia ovoides (see Cerococcus).

AS pollini, 120.

polygonus, Tabanus.

polygramma, Hutretosoma.

Pomegranate, Coccids on, 124.

pornia, Trypeta.

484 GENERAL INDEX.

Porphyronota maculatissima, on Citrus
in Zanzibar, 152.

praelonga, Chionaspis.

praetexta, Hurtbia.

pretoriensis, Anopheles.

preusst, Ptiloniola.

Prodenia litura, in Fiji, 35 ; on cotton
in Zanzibar, 148, 151.

producta, Hwuproctis.

prosopidis, Prosopophora.

Prosopophora prosopidis var. mimosae
(see Lecanodiaspis mimosae).

Protea hirta, Cocecids on, 102.

Protostrophus funestus, 275.

5 (Strophosomus) gravis,
274.

a instabilis, sp.n., on
Citrus in §. Africa,
275-276.

a noxius, Sp.n., on wheat

in §. Africa, 274-275.

3 planatus, sp.n., on
Citrus in 8S. Africa,
273-274.

proxima, Chionaspis (Hemichionas pis).
Pseudaonidia (see Aspidiotus).
Pseudococcus bromeliae, in Fiji, 39.

ss citri, on cotton in Zanzi-
bar, 149 ; a minor pest
of Citrus in Zanzibar,

151.
. comstocki, 179.
5 jilamentosus, 180.
% inquilinus, sp. n., in

British Guiana, 178—
179, 190, 191.

as longispinus, 178.

ie obtusus, on cotton in Zan-
zibar, 149; on pigeon-
pea, 150; on orange,
151; on Casuarina,153 ;
on mango, 152.

A perniciosus var., 179-180.

As virgatus, on cotton in
Zanzibar, 149; on
creeper, 154.

a virgatus var. madagas-
carensis, On cassava,
Haile

Pseudokermes marginatus, sp.u., in
British Guiana, 185.

‘o nitens, 185.

Pseudophilippia inquilina, Sp. n., ip
Jamaica, 181-182.

pseudopictus, Anopheles.

pseudotritea, Carpophthoromyia.

Psila, 215.

Psychidae, on African almond in
Zanzibar, 153.

Pterandrus, 212, 222, 228.

a anonae, 212, 228.

colae, 212, 228.

> penicillatus, 212.

- pennipes (see P. anonae)..

5 TOSQ, 2025220.

> » var. fasciventris, n.,.
212; 228:

9 rubivorus, 212.

it volucris, 212.

Ptilona, 3, note.
Ptiloniola neavei, 212.
bes preussi, 212.
«is tripunctulata, 212.

pulchella, Perilampsis (Carpophthoro-
myia); Rhabdochaeta; Rhacochlaena.
pulcherrimus, Anopheles.
pulchripalpis, Culex (see Ochlerotatus:
caspius).
pulverifer, Tabanus.
Pulvinaria brevicornis,
Avicenmia in
Guiana, 186.

“4 broadwayi var. echinopsidis,.
n., in British Guiana, 187..

punctata, Pardalaspis.
punctatifrons, Dacus (Tridacus).

punctatus, Culex (see Ochlerotatus:
currier).

Sp: 0.5 Om
British

puncticosta, Aspidomorpha.
punctifera, Chrysops.
punctothoracis, Ochlerotatus.
punctuliferum, Lecanium.

pungens, Culex (see
rusticus).

purchasi, Icerya.
pusillus, Laemophloeus.

pustulans, Asterolecanium (Asterodias--
pis, Planchonia).

Pycnocerus passerinii, on dead bark of.
clove in Zanzibar, 145.

pyrenaicus, Culex (see C. apicalis).

Ochlerotatus:

GENERAL INDEX. 485

Pyroderces simplex, in stored cotton-
seed in Zanzibar, 148.

quadratimaculatus, Culex (see Ochlero-
tatus rusticus).

quadrifera, Themaroides (hioxa).
quasigelidus, Culex.

quercicola, Asterolecanium (Planchoma).
Quercus, Coccids on, 110.

quercus, Lecanium (see Asterolecanium
variolosum).

quinaria, Pardalaspis.

Rachionotomyia aranoides, in Far
Eastern ports, 339, 341, 342.

Radish, Crocidolomia binotalis on, in
Zanzibar, 149.

Rats, and Glossina, 399.

Reed-buck, and Glossina, 391, 397.
regularis, Tabanus.

repleta, Trypanea.

reticulata, Trypanea.

Rhabdochaeta, 214, 266, 270-271.

¥ bakeri, 266.

ae neavet, Sp. n., 214, 270—
2A Me

- pulchella, 271.

sa spinosa, 214, 266, 270,
PATS

a venusta, 271.

Rhabdocnemis obscura, life-history of
in Fiji, 22-25.
Rhacochlaena fasciolata, 212, 238-239.

>» hammersteint, 212, 239.
3 pulchella, sp.n., 212, 239-
240.

Rhina amplicollis, infesting coconut in
Zanzibar, 147.

Rhizopertha dominica, in stored maize
in Zanzibar, 155.

rhizophilus, Aspidiotus (Odonaspis).
Ehochmopterum, 214, 266.

oe neuroptertpenne, 214.
Rhodesia, tsetse in, 71—90.

Khopaea subnitida, life-history of, on
sugar-cane in Fiji, 26-29.

Be vestita, life-history of, 26-29;
parasites of, 30, 31, 33, 34.

Rhopalosiphum nymphaeae, 45.
ue violae (see Fullawayella).

Rhus thunbergi, Coccids on, 99.
Ehynchocyon adersi, 147.

Khynchophorus phoenicis, on coconut
in Zanzibar, 147.

Rice, Ampera intrusa in stored, 41 ;
insects attacking, in Zanzibar, 149,
155.

Rinderpest, relation of Glossina morsi-
tans to, 7-20.

Rioxa bicolor, 4-5.
» debeaufortii, 5.
» jrava, 5.
» formosipennis, 5.
» wsignis, 5.
» lateralis, 5.
» musae, 1-2.
» nigra, 5.
» mnigripennis, 5.
» quadrifera (see Themaroides).
» serrata, 5.
» sexguttata, 5.

» termitoxena, sp. n., from Aus-
tralia, 1-5.

Rioxoptilona, 1.
rosa, Pterandrus (Ceratitis).

Rose, insects injurious to, in Zanzibar,
154; new Coccid on, 177.

rossi, Anopheles.

Royena pallens, new Coccid on, 122.
royeniae, Cerococcus.

rubens, Ceroplastes.

rubworus, Pterandrus.

rufipennis, Themarictera.

rufipes, Anopheles ; Necrobia.
rufiventris, Tephrella.

rufus, Dacus.

rupinae, Tabanus.

Russia, North, new Theobaldia from,
136.

rusticus, Ochlerotatus.

saginaria, Asura.

Saissetia (see Lecaniwm).

salinus, Ochlerotatus.

Salonica, mosquitos in, 134.
savastam, Capparimytia (Ceratitis).
saxatilis, Culex (see C. apicalis).
scaber, Tridacus.

486 GENERAL INDEX.

scapulatus, Corvus.
Schistopterum, 214, 266.
Be moebiusi, 214.

Scola manilae, parasite of Lamelli-
corn beetles in Hawaii, 30.

scutellaris, Stegomyia.

scutellata, Carpophthoromyia.
Scutia indica, new Chionaspis on, 95.
scutiae, Chionaspis.
Selenaspidus (see Aspidiotus).
semiangusta, Aciura.

semiatra, Spathulina.
semisphaerus, Dacus.
senegalensis, Pardalaspis.
seriata, Rioxa.

sericea, Glyphodes.

Serpents, and Glossina, 391, 394.

Sesamia calamistis, on maize and
millet in Zanzibar, 149.

sewelli, Haematopota.

sexfissata, Aciura; Tephrella.

sexguitata, Rioxa.

sexmaculata, Spheniscomyia.

sexmaculatus, Dacus.

seychellarum, Icerya.

seychellensis, Acidia.

Sheep, and Glossina, 391, 400, 402, 405,

437, 455, 456.

Sigmoides, Dacus.

signata, Phytometra.

signifera, Coelaenomenodera.

Silk-cotton Trees, harbouring mosquito
larvae, 50.

silvestrui, Pardalaspis.

Silvanus surinamensis, in stored maize,
155.

Simodactylus cinnamomeus, on sugar-
cane in Fiji, 31-33.

simplex, Chionaspis ; Pyroderces.

sinensis, Anopheles.

sinuatus, Ocneros.

siphonina, Ensina.

sitiens, Culex.

Situtunga, and Glossina, 362, 376-379,
381-383, 388, 391, 394-396, 404-405,
408, 418, 424, 434, 446, 447, 455.

socialis, Antonina.

Solanum melongena, Acanthocoris fasci-
culatus on, in Zanzibar, 150.

Sorghum vulgare, 149 ; Calandra oryzae
in stored, 155.

sororcula, Ensina.

Sostopsila trisetosa, g. et sp.n., 211,
214-216.

Spathulina, 213, 257-259.
Re acrosticta, 213.
Us aldabrensis, 213.

a bioculata, sp. n., 213, 258—
259.

50 margaritifera, 213
parceguttata, 213.
a semiatra, 213, 258-259.
Spheniscomyia oe Sp. n., 213, 256-
57.

ue sexmaculata, 213, 256.

Spirogyra, Culex larvae protected by,
49.

speciosa, Coelaenomenodera.
sphaeristicus, Tridacus.
sphenophori, Ceromasia.
spinosa, Rhabdochaeta.
splendens, Corvus ; Mimomyia.
squamosus, Anopheles.
Stapelia, new Coccid on, 115.
Stegomyia eos (scutellaris), 334,

albolineata, in Malay Penin-
sula, 343.

is argentea (see S. fasciata).

desmotes, in Malay Peninsula,
343

fasciata, in Gold Coast, 47—
58; in Ashanti, 58; in
tree-hole in Epping Forest,
91; chaetotaxy of pupa
of, 161-169; in Mesopo-
tamia, 324; prevalence
of, in Far Eastern ports,
333-344.

P fasciata var. luciensis, in

Far Eastern

ports, 339-341.

fusca, in Siam, 342.
metallica, in Gold Coast, 55.
sugens (see S. vittata).

trilineata, in Malay Penin-
sula, 343.

unilineata, in Gold Coast, 53.

variegata, in New Guinea,
342.

GENERAL INDEX. — A87

Stegomyia vittata (sugens), in N. Gold
Coast, 49-57, 61; not
reported from Accra, 51.

Pe w-alba, in Malay Peninsula,
‘Siam and Sumatra, 343.

stellata, Hexacinia.
stentae, Asterolecanium.
stephensi, Anopheles.

Sternochetus (Cryptorrhynchus) mangt-
ferae, on mango in Zanzibar, 152.

stictica, Pardalaspis.
Stizolobium aterrimum, 22.
Stomoxys, and Glossina, 448.

Storms, destructive to Glossina, 443-
444.,

strangulatus, Trochorrhopalus.
stricticollis, Lacon.

strictifrons, Plhomelaena.

strigata, Coclopacidia.
Strophosomus (see Protostrophus).

Sturmia bimaculata, bred from Cirphis
unipuncta, in Fiji, 34; from C.
loreyt, 35.

stylifer, Tridacus.

subacutum, Lecaniwm.

subangustus, Tabanus.

subcompleta, Trypanea.

submorsiians, Glossina.

subnitida, Rhopaea.

subpicius, Anopheles.

subtilis, Grabhamia (see Ochlerotatus
caspius).

Sugar-cane, pests of, in Fiji, 21-39;
Coccid on, in Uganda, 201.
sugens, Stegomyia (see S. vittata).

suleata, Pangonia.

sultan, Petrodomus.
superba, Carpophthoromyia.
superflucta, Trypeta.
superstitiosus, Dysdercus.
suppressalis, Chilo.
surinamensis, Silvanus.
suspensa, Oonradtina.
suturalis, Cossonus.

Sweden, method of studying forest
insects in, 157-160.

ee Potato, pests of, in Zanzibar,

(693)

Sylepta derogata, on cotton in Zanzibar,
148, 151.

Tabanidae, of Palestine, 277-321.

Tabanus, 277-278, 292-321, 403, 448—
450.

ie accensus, Sp. n., 292, 294,
309-311.

ae albifacies, 315.
se albipes (see T. gigas).
a alexandrinus, 292-296, 298.

se arabicus, 318.

A areniwagus, Sp. n., 292, 294,
305-309.

BL aterrimus var. auripilus, 298.

ee autumnalis, 292, 294, 301—
302.

3 cardonatus (see T. alexan-
drimus).

a cordiger, 312, 325.

- dalet, sp. n., 292, 294, 319-
oad.

Me decorus, 292, 293, 295.

- eggeri, 278, 292, 294, 301.

» (Atylotus) gigas, 292, 293,
299.

$5 ansecutor, sp. n., 292-294,

296-300.
roy intermedius (see T. eggert).
ass laetetinctus, 319.
- leleani, sp. n., 292, 294, 312-—
316.

Ze lunatus, 292, 293, 300.

a4 mendicus, 292, 293, 299-300.
a miki, 318.

Ds nemoralis, 292, 293, 300-301.

95 pallidipes, sp. n., 292, 294,
316-319.

. pertinens, 311.

a polygonus, 305.

sy pulverifer, 308.

wd regularis, 292, 294, 302.

ss rupinae, Sp. n., 292, 294, 302—

305.
» subangustus, in N. Gold Coast,
53.
» wmbrinus, 298.
pi wunicinctus, 315, 316.
43 unifasciatus, 316,

488 GENERAL INDEX.

Tachardia, 119, 122-126.
ss actinella, 123-124.
9 affluens, sp.n., 125-126.
55 (Tachardina) albida, 122,

126.
Ha (Tachardiella) cornuta, 122.
ae decorella, 125.
‘i karroo, sp.n., 124.
bs lacea, 122, 126.
oe minor, sp. n., 124.
A longisetosa, 126.

Tachardiella, new subgenus of Tach-
ardia, 122.

Tachardina, new subgenus of Tach-
ardia, 122.

Taeniorhynchus domesticus (see Culex
aurantapez).

Tamarinds, harbouring mosquito
larvae, 50.
Taomyia marshalli, sp.n., 212, 241-
ot a

As ocellata, 212, 241.
tarsalis, Lecaniodiaspis (Lecanium).
tavarest, Aphis.

Teak, Dinoderus minutus in, in Zan-
wibar, ldo.

Tenebroides mauritanicus, in stored
maize in Zanzibar, 155.

tenuicornis, Hlenchus.
tenuidisculus, Chionaspis.
Tephrella bezziana, 215.
cyclopica, 213.
ey hessti, 213.
a nigricosta, 213.
ae rufiventris, 213, 257.
% sexfissata, 213, 257.
a tephronota, 213.

: w-fuscum (see T. cyclopica).
Tephritis indecora, 213.

b lugubris, 5.

a vernonticola, 213.

tephronota, Tephrella.

Terellia % hysia, 213.

Termes bellicosus, killing young clove
trees in Zanzibar, 145; destroying
coconuts, 147 ; a timber pest, 153.

Terminalia catappa, attacked by
Psychidae in Zanzibar, 153,

termitoxena, Rioxa (Trypeta).

ternaria, Aciura.
territans, Culex (see O. apicalis).
tessellatus, Anopheles.

testaceus, Culex (see OC. apicalis) ;
Dacus.

tetrachaeta, Aciwra.

Tetranychus, on cassava in Zanzibar,
151.

Thalassoides digressa, on castor plant
in Zanzibar, 154.

Themara fallacivena, 212.
Themarictera flaveolata, 212, 238.
— laticeps, 212, 238.
35 rufipennis, 212.
Themaroides quadrifera, 5.
Theobaldia alaskaensis, 136.

As annulata, 137; in Meso-
potamia, 325.
arctica, sp. n., from N.
Russia, 136-137.
iy indica, sp. n., from Punjab,
137s
5 longiareolata, in Meso-

potamia, 325.
thesit, Asterolecanium.
thomsoni, Coelaenomenodera.
ligripes, Culex.
tipuliformis, Culex.

Tortoise, and Glossina, 376, 379, 391,
393, 405.

Toxorhynchites wmmisericors, in Far
Eastern ports, 335.

Trachycentra chlorogramma, larva of,
attacking diseased sugar-cane, 35—
36.

Tragocephala variegata, in pigeon-pea
in Zanzibar, 150; in COvttrus in
Zanzibar, 152.

Tribolium castaneum, in stored grain,
155.

Tridacus, list of African spp. of, 211.

trifilatus, Culex.

trigonus, Dacus.

trilineata, Stegomyia.

trilobitiformis, Aspidiotus (Pseudaoni-
dia).

trimaculatum, Hippocentrum.

tripunctulata, Ptiloniola.

Trirhithrum, 212, 222, 236—237.

- 2 albomaculatum, 212,

GENERAL INDEX. 489

Trirhithrum albonigrum, 212.
bimaculatum, 212.

coffeae (see T. mseriptum).
gagatinum, 212.

992

inscriptum, 212, 236.
leucopsis, 212.

a lycti, 212.

a nigerrimum, 212.

+3 nigrum, 212.

a nitidum, 212, 236.

5 obscurum (see T. nigrum).

ae occeipttale, 212, 236.

- validum, sp. n., 212, 236—
23s

trisetosa, Sosiopsila.
trisptla, Leucotaeniella.
tristicula, Trypanea.
tristriata, Phorellia.
tristrigata, Huribia.
tritaentorhynchus, Culex.
tritea, Carpophthoromyia.

Trochorrhopalus _ strangulatus, _ life-
history of, 25-26.

Tropidocephalus brunnipennis, from
Nigeria, 139.

bi eximius, 139.
Trypanea, 213, 260, 263-266.
neh abstersa, 263, 264.

cee avira, 213.

am amoena, 213.
$3 aucta, 213.

a augur, 213.

ss auguralis, 213.
ia confluens, 213.
$s decora, 213.

8 diversa, 213.

» gnaphalii, 266.
on guimari, 264—266.

fs hemimelaena, sp.n., 213.
se hexapoda, 213.
ra mamulae, 266.

si peregrina, 213, 264.

repleta, 213.

reticulata, 263, 264.
subcompleta, sp. n., 213,

263-264.
Re tristicula, 213.
urophora (see T'. peregrina).

(693)

Trypaneidae, new Australian, 1-5;
new African, 211-271.

Trypanosoma evanst, transmitted by
more than one sp. of
Tabanus, 277.

i pecorum, transmitted in
absence of tsetse, 81.

Trypeta bicolor, 4-5.

ay grata, placed in Clinotaenia,
229. 20a:

5 jucunda, 212, 243.

iD musae (see Rioxa).

2 pornia, probably a Rioxa, 5
cm superfiucta, 247.

tulipaella, Fullawayella.

tumulifera, Cremastogaster brevispinosa.
Turnip, Crocidolomia binotalis on, 149.

Uganda, relations of Glossina morsitans
and game in, 7-20; Coccids from,
177, 178, 188, 192, 198, 200-207 ;
feuit-thies, irom (225,228,233. 237,
256; bionomics of Glossina palpalis
in, 347-463.

ulula, Hlaphromyia.
umboniferus, Aspidiotus (Chrysompha-
lus).
umbrinus, Tabanus.
undatus, Ocneros.
unicinctus, Tabanus.
unifasciatus, Tabanus.
uniformis, Mansoniordes.
unilineata, Stegomyia.
unipuncta, Cirphis.
univittatus, Culex.

Uranotaenia campestris, in Far Eastern
ports, 339.

a cancer, in Far Eastern
ports, 334, 340.

Urophora bicolor (see Rioxa).
urophora, Trypanea (see T’. peregrina).

validum, Trirhithrum.

Varanus, and Glossina, 357, 376-379,
382-385, 387, 391-393, 401-405,
410, 415, note, 416-418, 424, 427, 429,
441, 442, 446, 450, 452.

variabilis, Microparsus.

variegata, Stegomyia ; Tragocephala.

variolosum, Asterolecanium (Astero-
diaspis).

490 GENERAL INDEX.

wvenusta, Rhabdochaeta.
vernoniicola, Tephritis.
versutus, Adoretus
vertebratus, Dacus.
vestita, Rhopaea.
vetusta, Orgyia.

violae, Fullawayella (Neotoxoptera,
Rhopalosiphum).

virgatus, Pseudococcus.

viridis, Coccus.

vishnui, Culex.

vitiensis, Perkinsiella.

vittata, Carpophthoromyia ; Stegomyia.
Volucris, Pterandrus.

vulpinus, Dermestes.

w-alba, Stegomyia.

w-fuscum, Tephrella (see T. cyclopica).
Walkeriana digitifrons, sp. n., 177-178.
Wart-hog, and Glossina, 391, 397.

Water-buck, and Glossina, 387, 391,
397,.400, 409, 432.

Watercress, Crocidolomia binotalis on,
149.

waterstonit, Antonina.

watti, Culex.

Weevils, injurious species of, in 8.
Africa, 273-276,

weisei, Nisotra.

whitmorei, Culex.

willeocksi, Grabhamia (see Ochlerotatus
caspius).

woodi, Dacus ; Ocnerioxa.

Xanthogramma Sp. preying on Alew-
rodes comata in Fiji, 38.

xanthopterus, Tridacus.
Xanthorrhachista, 242.

5s alata, 212, 238.

fe cephalia(see X.alata).

Zanzibar, injurious insects in, 145-155 ;
Anophelines from, 329-332.

Zebra, and Glossina, 391, 397, 432.

491

INDEX TO NAMES OF PERSONS.

Abraham, A. A., 179, 185, 190, 191.

Aders, Dr. W. M., 93, 145-155, 209,
329--332, 345.

Anderson, T. J., 93, 147, 180, 209, 236,
264. .

Arnold, Lt.-Col., 308.
Arrow, G. J.. 26.

Austen, Major E. E., 20, 87, 277.
Bain, 14.

Baker, A. C., 45—46.-

pallard, E., 465. |
Bandyopadhyay, B. N., $3, 209.
Barnett, 9. |

Barraud, Capt. P. SAO: 209, 323-225,
345, 465. .

ibanrow, Col. EL. J. W., 137.
Bate, Miss D. M. A., 301.
Becker, 302. _

Beeson, C. F. C., 465.
Bevan, ul. E. WL 76:
Bevis, L., 253, 262.

Bezzt, Prof. M., 1—5,. 211-271, 277,
note, 278, note.

Bodkin, G. E., 44, 93, 179, 185-187,
189-191, 194, 196, 197, 209, 345, 465.

Bottcher, A., 300.

Bovell, J. R., 465. .

Brain, C. K., 95-128, 180, note, 345.
Brauer, 295, 296, 299-302.
Breemen, Dr. M. L. van, 333.
Bridwell, J. C., 139-144.

Brohier, Dr. 8. L., 465.

Brookes, A. E., 132.:

Bruce, Sir David, 8.

Burkill, J. H., 209, 465.

Bunton. He bios [32%

Buxton, Capt. P. A., 93,130, 345, 465.

Carment, Capt. A. G., 136.

Carpenter, Dr. G. D. H., 93, 349, 359,
362, 440, 450, 465.

Carthew, Dr. M., 333.

Champion, G. €., 314.

Chapman, J., 72, note.
Christophers, Major 8. R., 323, 325.
Cleazes i, Ds dur. 43-445

Corfield, Capt. W.S., 301.
Coryndon, R. T., 74.
Crawley, W. C., 93.
Cropper, Dr. J., 279.

Dale, Capt. W.:J., 295, 300, 314, 321.
Dash, J: 'S:,-93.

Davis, 45.

Deakin, R. H., 199..

del Guercio.-G., 45.

de Meijere, Prof. J. C. H., 225.
d’Emmerez de Charmoy, D., 93.
Denier, Dr. A., 333.

de Vogel, Dr. W. T., 333, 340.
Distant, W. J.., 41.

Dreyer) 0. E., Vik

Dry, F. W., 180, 199.

Dudgeon, G. C., 314.

Duke, Dr. Tf. Lyndhurst, 7-20, 457.
Duponi, P. B., 93.

Duporte, E. M., 345.

Dutton, Dr., 51.

Dyar, H. G., 230, 134, 136, note.

faton, Rev. A. E., 301, 314.

Edwards, F. W., 59, 61, 63, 65, 66, 68,
129-137, 223, 333.

Essig, EB. A., 46, 46. |

492 INDEX TO NAMES OF PERSONS.

Faure, J Cale:

Ferguson, Dr. E. W., 345,

Fiske, W. F., 7, 10, 138-15, 18--20, 347—
463.

Fowler, Major C. E. P., 134.

Froggatt, J. L., 5, 185.

Fullaway, D. 'T’., 45, 46.

Fuller, €., 96; 99, 10@, 103; 112, 120.
127.

Gahan, Dr. C. J., 296.

Gennadius, P., 301.

Gibbons, Major A. St. H., 12, 20.

Gielgud, Val, 8, 72, note.

Goetz, Rev. E., 84.

Goot, P. v. d., 46.

Gough, Dr. L. H., 345, 465.

Gowdey, C. C., 93, 177, 178, 188, 192,
198, 200, 201, 203--207, 209, 225, 228,
251g Da, ole

Green, E. E., 176, 202.

Guercio, G. del, 45.

Hall, Dr: C23. 5. van, 41,
Hardenburg, C. B., 111.
Hargreaves, E., 93, 130, 209, 345, 465.
Harris, W., 345.

Harrison, H., 196.

Hendel, Prof. F., 1, 5, 214.
Hewitt, Dr. C. G., 345.

Hill, Gebel, a

Hobley, C. W., 345.

Hodges, Dr. A. D. P., 458.
Howard, Dr. L. C., 9, 136, note.
Hutchins, E., 7, 8, 15.

iatsom, J. a5 9a.

lingworth, Dr. J. F., 25, 39, 465.
Ingram, Dr, A., 47-58, 59-69.

Jack, R. W., 12, 20, 71-90.
Jackson, Sir F. J., 8-10, 82.
James, Lt.-Col. 8. P., 333.
Janson, O. E., 176.
Jardine, N. K., 93.

Jepson, F. P., 39.

Kelly, A., 97, 100, 113, 116, 119-121.
King, H. H., 345.

Kinghorn, Dr. A., 87.

Knab, Dr. F., 136, note.

Kotinsky, 195.

Lamborn, Dr. W. A., 79, 80, 82, 88,
ZAC 29.

Leefmans, §., 209, 456.

Leicester, 134-135.

Lelean, Lt.-Col. P. S., 308, 315.
Leonardi, Prof., 104.

Lim; Dr, Non sso:

Lindesay, Capt., 137.

Lloyd, Ll., 11, 14, 79, 82, 83, 84, note.

Loew, H., 277--278, 280, 295, 299, 301.
302.

Lounsbury. C. P., 95, 97-99, 107-109,
111, 213; 116 Ee e228

Loveridge, A., 346.
Lowne, B. 'T., 299.

McAdams, J., 75.
McConnell, Dr. R. E., 346.

Macfie, Dr. J. W. S., 59-69, 129, 161-
169, 223, 240, 251, 255, 259, 346, 466.

MacGregor, Capt. M. E., 91, 94, 346.
Mackenzie, Dr. Alex., 76.
Mally, C. W., 117; 198:'230; 232:

Mansfield-Aders, Dr. W., 93, 145-155
209, 329-332, 345.

Manson, Sir P., 333.

Marshall, Dr. G.. A. K., 1, :20,) 41, 52.
129, note, A71, 211,236, 239) 242
256, 259-261, 270, 2738-276; 323, 330.

Maskell, W. M., 195.

Masters, W. ae os

Maulik, S., 171-174.

Meijere, Prof. J. C. H. de, 225.

Montgomery, R. E., 8, 87.
Morgan, E. M., 195.

Morgan, J. G., 221.

Morrison, H., 187.

Muir, F., 25, 39, 139-144, 209.

Neave, Dr. S. Aly Zit, 224. 227; "249;
256, 2515. 200 Me
Neumann, O., 9-11, 20.

INDEX TO NAMES OF PERSONS. 493

Newstead, Prof. R., 102, 122, 175-207,
181, note, 182.

North, D.S., 25.
Nuttall, Prof. G. H. F., 209.

O’Brien, Dr. J. M., 466.

Olds Drs J... 8., 247.

O’Neil, Rev. Falher J. A., 466.
Orde-Browne, G. H. St. J., 250, 255.

Patterson, W. H., 171, 209, 245, 266.
Patton, Major W. S., 314, 315,
Pearson, Prof. HW. H. W., 98.

Pomeroy, A. W. J., 135, 136, 346, 466.

Rainey, Lt.-Col. W., 466.
Rencurel, Major J. E., 336.
Ricardo, Miss G., 500.
Richardson, V. F., 17.

Romer, Dr. L. 8S. von, 333.
Roepke, Dr. W., 466.

Ross, Sir R., 129.
Ross-Tewnsend, 96.
Rothschild, Lord, 257, 302, 314.

Sander, Dr., 20.
Schneider, 134.
Scott, 74.

Scott, H., 131.
Sergent, Dr., 134.

Sewell, Lt.-Col. E. P., 286, 285, 298,
300, 301, 308.

Sharpe, Sir A., 8.

Shireore, Dr. J. O., 11, 13, 20, 346,
Shortt, Capt. H. E., 323.

Shroff, K. D., 94.

Silvestri, Prof. F., 105.

Simond, Dr. P. lu., 337.

Simpson, Dr. J. J., 47.

Smith, Dr. Malcolm, 333.

Smith, Mrs. W., 346.

Snow, F. L., 257.

Someren, Dr. B.I:., van, 263, 458-462.

Stanton, A. T., 333-344, 466.

Stent, Miss S., 115.
Stevenson-Hamilton, Major J., 8-10,

12, 192-20.
Stewart, Col. J. A., 308.
Stordy, R. J., 233.
Storey, Dr. F. H., 346.
Surcouf, Baron J. RK. M., 301.
Swainson-Hall, R., 209.
Swynnerton, C. FE. M., 328, 346,

Terburgh, Dr. J. T., 340-341.
Theobald, F. V., 45, 230, 346.
Thompson, Rh. L., 101i, 346.
Thornton, Capt., 74.
Tragardh, I., 157-160.
Turner, R. E., 327-228.

Urich, F. W., 346.

van Breemen, Dr. M. L., 333.
van der Goot, P., 46.

van der Merwe, C. F., 101, 117, 121,
124, 238

van Hall, Dr. C. J. J., 41.

van Someren, Dr. B. L., 263, 458-462.
Veitch, R., 21-39, 94, 209, 466.
Villeneuve, Dr. J., 132, 134, 299.
Vogel, Dr. W. T. de, 333, 340.

von Romer, Dr. lL. S8., 3338.

Waterston, Capt. J., 94, 133, 134, 184.
Wait, Dr. W. G., 94, 136, 210.
Werkman, Dr. H., 333.

Wesché, W., 66, 67.

Willeocks, F. C., 346.

Williams, C. B., 210, 346.

Williamson, Dr. G. A., 314.
Williamson, Capt. K. B., 299.

Wood, R. C., 210, 219, 246.
Woosnam, R. P., 11.

Yorke, Dr. W., 11.

PRINTED BY BARCLAY AND FRY, LTD., THE GROVE, SOUTHWARK, LONDON, S.E.

pee oes dy mene aie Be ue
; Rte isdn my

Mie | ate Py ta
AGH UO REE tad i sis i
ite i aepops ype et '

L sO Crea Wed . ET oe ae ig nat et ; PR |
oe i a vitae
we recy Ei ehdinieina’s Sh Ay oe cae Jovy

Se hihi : fod i fone WE of

GR ey Eo
: pee A }

"Vere ure
oat reds ¥

* Sa ROE a

et eae + ay

Deh ie Pat peed Gist ae feats
oD, COG Gah ‘Abe. tee aay ploy
ere ee fm aL te as sere eth. AK:

BES yo 8 ae
ab a ac sip 14 th
eth al AEM pvgh oar

9 by ie: ae

eal aad AD Ke.
cfTR wy

AG airy el
OE iow ce
A a way inf v

2 af MT Se"

2 Pee igh AY . ; ‘
Te) ttscwe, ; ae ay shee aks Ay ma
A i a & ~The kes antl

Ra -
“a te Lb Th wee F Rese yriea ibli era
her weary hy mee cee bot,

ia fel - «tea eine ie

a o ee § aft Maceo om

Sh h Sere wee = Sheil
eae ame hrs oe

e , 4

ADVERTISEMENTS.

Illustrated Scientific Work a Speciality.

WE INVITE YOU TO SEND

AN INQUIRY FOR YOUB

Book and Magazine Printing

TO OUR . .

Printing Department, 83-91, Great Titchfield Street.

WATKINS &

‘Sa
BS
eA
SS
S

DONCASTER,

Naturalists and Manufacturers of every kind of
Apparatus and Cabinets for Collectors of Insects, Birds’ Eggs,
Plants, Minerals, &c.

Plain Ring Nets, Wire or Cane, ls. 9d.

Folding Nets,

Pocket Boxes, 8d.; corked both sides, ls. 1d.,
Is. 6d., 2s.

Zinc Relaxing Boxes, Is.,1s. 3d., Is. 10d., 2s. 6d.

Japanned Collecting Box, with strap, 6s. 6d.

Entomological Pins, mixed, ls. and ls. 3d. to 3s.
per oz.

Mite Destroyer, 14d. per oz., Is. 6d. per lb.

owas Tin, with brush, ls. 6d., Is. 9d., 2s. and
2s. 3d.

Store Boxes, with camphor cells, 2s. to 7s. 6d.
Ditto, Book Pattern, 12s., 13s. 3d., 14s. 6d.
Setting Boards, flat or oval, lin., 7d. ; 14 in., 9d. ;
T#in., 10d.; 2in., 11d.; 24in., 1s. 1d. ; 3in.,
Is. 2d. ; 34in., Is. 4d. ; 4in., 1s. 4d. ; 44in.,

Is. 8d.; 5in. 1s. Lld. Complete set of 14

boards, 13s.

Setting Houses, 11s. 6d., 13s., 16s.

Zinc Larva Boxes, ls., Is. 3d.

Breeding Cage, 5s., 6s. 6d., 7s. 6d.

Coleopterist’s Collecting Bottle, 1s. 6d., 1s. 8d.

Botanical Cases, japanned, double tin, Is. 9d.
2s. 9d., 3s. 6d., 4s. 6d., 5s. 6d.

Steel forceps for removing Insects, 3s. per pair

Cabinet Cork, 7 by 34, Is. 5d. per doz,

Pupa Diggers, Insect Lens, 1s. to 12s. 6d.

Egg Drills, 2d., 3d., and Is.

Metal Blow-pipes, 6d. and 8d.

Taxidermist’s Companion, %.¢., a pocket
leather case containing tools for skinning,
12s. 6d.

Egg Collector’s Outfit, containing Drills, Blow-
pipes, and all requisites, 48, 3d. post
free.

A Large Stock of Insects and Birds’ Eggs. (Singles and in Clutches.)
New and second-hand Books on Natural History.

Birds and Mammals, ¢-c., Preserved and Mounted by jirst-class workmen.

3G, STRAND, LONDON, W.C. 2.

Five doors east from Charing Cross Railway Station (S.E. & C.R.)

(C572)

i ADVERTISEMENTS.
DULAU & Co., Ltd.,

37, SOHO SQUARE, LONDON, W.1,

Have always on hand a very extensive stock of works on the Natural

and Physical Sciences—both Secondhand and New.

CATALOGUES SENT FREE ON APPLICATION.

Libraries or small Collections of Books Purchased. Distance no object.

_COLLINGE (W. E.). The Food of Wild Birds- 4/-
COLLINGE (W. E.). A el ed of injurious
Insects S - - - - - 12/6

WORSDELL (W. C.). The Principies of Plant-
Teratology. 78 Puates. 2 Vols. Ray

Society - - - . - - £2 10/-
Journal of Economic Biology. Edited by

W. E. COLLINGE. ILiustraTEep. 10 Vols.

1905-13. 8vo. Cloth - - - 2 - £8 8/-

JOHN WHELDON & Co. have the largest stock in the
country of Books in all departments of Science and Natural
History, also Transactions and Journals of Learned Societies,
etc., in sets, runs, and single volumes or numbers.
A LARGE SELECTION OF BOOKS ON ENTOMOLOGY. ANY WORK
QUOTED FOR.
LIBRARIES OR SMALL PARCELS PURCHASED.

SPECIAL CATALOGUES: Entomological, Botanical, War Economy (Agriculture,
Gardening, Economic Botany, etc.), Zoological, Chemical, Geological, etc.
2d. each post free.

38, GREAT Nite gar STREET, KINGSWAY, LONDON, W.SC. 2.

hele phone? GERRARD 1412.

| =
The Editor will be pleased to receive ‘tor ‘publication papers: ‘or |
notes dealing with any insects which are of economic importance.

a Such communications to sin addressed to

THE DIRECTOR, : |
Imperial Bureau of Entomology,
British Museum ogra History),
London, SW. 7.

‘The Rublisation and distribution of the “Bulletin of Entomological

we Hemearch ” has now been taken over by

THE ASSISTANT DIRECTOR,
imperial Bureau of Entomology,
‘88, Queens Gate, London, S.W. oe

and orders ad subscriptions should. be sent direct to him or through.

any bookseller.

"The annual subscription is Ten Shillings, post free»

(C572)

CONTENTS.

ORIGINAL ARTICLES.
egy Dr A 0. Neotovoplera violae, Theo. and i its Allis ¢
(PLATE bY eee tee

Brzzi, Prof. M. A new Runeclnn species of cia with a BOA
remarkable peyety (Dipt.) (illustrated) — =

) CLEARE, LAvRENcE D., Ju unt. A useful Breeding Cage (ustrated).

Disranr, W. L. A new Lygecia Pag. pes puns Stored Be. Bria!
in Java (dlustraied) — - ; :

DUuUKE, Dr. H. LynpuuRst. An Rnquicy into the Badu Oboes
- Glossina morsttans and cee pik ee ee reference. = as ce
to Rinderpest mg fh os

Incram, Dr. A. The Domestic Breeding Mosquitos of the a iets
Northern Territories of the Gold Coast -~— ata ae id

= INGRAM, Dr. A., and Macrm, Dr. J. W. 8. The Barly Sages of
~ West African Mosquitos—LV ree) =

J ACK, RUPERT W. Tsetse tie in Southern bears gt ‘118
_ {Puates I-IV, Maps I-II -

MAogREGOR, Capt. M. E. On the Dain of Seyomgi Lapwae ys *
» in a Hole in a Beech Tree in Epping Forest -- =e )

_VEITOH, Rozear. Notes on the more important Insects in wee va
_Pagarroane Plantations i in ‘By (dlustratedy) - =

- MISCELLANEOUS. Pe
Obllectionsresivad Se ee eee eo Se ei: i

Pe)

The Editor will be pleased to receive tor publication papers or
notes dealing with any insects which are of economic dasa sna

~ Such communications to be addressed to

THE DIRECTOR,

: Imperial Bureau of Entomology, :
Bt we . British Museum (Natural History)
| London, S.W. 7

ee The publication and distribution of the “Bulletin of Entomological

- Research” has now been taken over by

+
ed

_ THE ASSISTANT DIRECTOR,
Imperial Bureau of Entomology,
88, Queen's Gate, London, S.W. ue

and orders and subscriptions should be sent direct to him or through

any bookseller. :
i Th
The annual ibcaston to Vol. X is Ten Shillings, post free ; that

for subsequent Volumes will be Fifteen ees

(637) |

_ The Editor will be pleased to. receive for publication papers or

“notes: dealing with any insects which are’ of. ais bal Hilde alain
"Such communications’ to be addressed to. ae
ee ee ‘THE DIRECTOR,

ae | | ane Bureau of Entomology,

. British Museum (Natural iistoey,
| London, S.W. 7.

The publication and distribution of. the vs “Bulletin of mac ea

“Research ” has now been taken over r by

THE ASSISTANT DIRECTOR,
| imperial Bureau of Entomology,

88, Queens Sette London, SW. ke
%

te and orders and subscriptions should be dent dikect to Bi or thradek

G,
i any bookseller.

The annual subscription to Vol. Xi is Fifteen’ Shiites, post free.

“oats

(659)

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Bulletin of entomological research"

See other formats