wl ~ ar Preeti beSSSbEel 06Gkabeteteded hdd. lk ee Pubshoveledabetubcia an oa eee iptatetatubiinletkbettdetetasaticataee ow ere Fer met ati ee ae) letutalelss ahaa ee OPE Br eee dr gg wen, PP me ed oer we. abs ent eee rt ey RENEE NS Or Hh pet sew4, ON EE ee TE ee I ee mht... tasetehe ong sind Pe PP”. + © SOS whehntes, rs : O80 ebaleteietetyte: —~ cart wie: Wer eret sv hevgtewmie:» ; > o Danae a e . 7 ayers wt i ee eee Mote tere ia rte Als iin Raia whos ndas Pee ain we oe wrth ade ale alerwwes Aes Werner 'p Wigtd verte riety tfeibdh-deae le Tn ot te . i ie ~~ ee OP Pele sents athe , He ee : . a fee eee Leietelwigre ebele : Fea a altel Saas rdicontite. : ae =: : é le hetbieiarsre “ ; Wheat hse lip wah ens ~euteze - os tots wer wth wrotaletvie’s 09 vate} ‘ * oe! 5 sat oF ‘ edie I es ee ees —e a = v ‘ wie letehe ae , s tet Oe awe wesw wwwx state arene Me eee ee ' euhere ee ¢ eke ‘ * r —— D OC wwe Ww aie elon whelg She are tete ele eine tates a ate ee Fete el ebe Se Saw 5 2: P f 2 6 siete ‘ ‘a co Peewee oe ahebete fw 4 Be Wale eebime ota OP ® tree es op! u'si oles alps oe eee pe ea ele ate ate e's otcfelstetdi " 7 9 e'> wtete + o> ad Class Book 2 SMITHSONIAN DEPOSIT. ' / COR ie CS MS aCe. «ean ee ik PER * SRP RES 6 ‘y | sf . im? , ; A LAT bee on aes ‘ | F a AVA 2 IP; ‘ ~ . we + q i p nd «ih oo ° 0 Re Bae MER 0 Oe ‘ eAal. F orf ', : , y i ih . 8 6 : : aa | eee ) J By ‘ Nae a8 2 y ; : sm wean i? " oa a” ¢ wi AS > Gas 4 vcs * % wel , rast, ah H " % ff “? ¥ ¥ A y wee rip) oe 43 { bath 4 . ¥ ‘earn = yi ; h 7 ee 2 a Ss ’ oT fi)? é : (rep he rot i 2 ' a, z AL, 4 La | - 7 RA a : 4 ie i di TAL, ; - sh re ’ ‘ + 4 ip i \ ae dees P i. 163 s ” . how ow Wt Agia | ‘tm 25% bY ; J 4) , * Fe i fi nee . ia wi ; en ro), BOL Rs | y ira) Ab , ; ee i nn ‘ fig y Beer ere y's "5 +f pa. | ax i oF ee \ belie : ic OS eee Oe) oes ee Lindley ; MR Ay slowed Lime, deh ah irafnge ha? 0 a ‘ ii te We ripe ’ ; , z . ' . My ha by Ta hi in) ‘ ie 4 i / ‘ ‘ wa P, 7 a ne =) - rae 4 a a a ay as ae y va ie +: oh: aa ‘om L Py kn tt We j . Ayn. te by ° é P ri ‘ 4 os GAD ni SE oe OI ane a rake ie LCA as ee abe j ye a el : i vod if i & : 7 ‘ ya eh — | ¢ »* Z ‘ ; ‘ ia. bes me ee . ee, Mae te At Mee a Tem Cie Sa : wee : a ra my. anes PS ly er ‘ vies : ae Ai So. Cee By) { a ’ ; ! ae er, . \ y fc A \ 4 =. ae 0 h eee / ee vy. On. ~ ee ox LD > BULLETIN OF ENTOMOLOGICAL RESEARCH ISSUED BY THE IMPERIAL BUREAU OF ENTOMOLOGY. EDITOR: THE DIRECTOR. VOL XI, LONDON: Tue Imppria, Bureau or Entomo.noey, 41, QubEN’s Gate, 8.W.7. 1920-1921. 4 ~~ est th HA} NET OLE: 6 rtp OHO PRE *Lt} Chae : wok % Po aT Asia Ba earn ok & Vigtert era eee x) ve thf 7M gna hietha thet FIN SNe eas i* Fa fa tema” ic ae if : ree Fa}? - | Fi et i 7 Pay ‘s > se dj : ie als a Sita¢ fs. P A *) > * my. . 7 7a) Co 4 a el -™ i hee! J 4 a “ne i. | 7 an ase ay ' Sa "hi We i eee ak 7. Bore ee tar ie A Te 7-75 ise fi thé i} 7 ‘ ‘ ai 44 Lae hud REL. AUG. Vl keg J j | >? ur ig ' i ; | ee ati Ried) Pe Mart q f + if JAF MO < °T ar ’ ‘ie 7 , > { | j i R At! 4 f LZ j . ‘ ey a a, ? 17 hares}: ‘ Tal i} ¢ ‘ ) ° ’ eh etl ACE Lif . , a r¢ ihe} 45 J F nen cee! z i i? j ( r we d 7 ‘yy va j a Y saat 4 AYi Tl ewe 4 ~ | ae 7 OPAL! 12h HUA ( os Ioagtoliedy leet uimal iH igh eer cra ; a 7 Mote 4.08 © Sa ( gaat! rept? wet a) Fi id wi) Ags Ae in ton? “ ae « wt § ? mi ettgion . 4° Pt - av ION ? i wel! MOY strana aM rate. SS ium A} uvard..A & sf + AGIOMERD hMiaheY) compe ae Q | | WA VER ober sitet e'aoangl (i Byaae) e hte oe ‘7 U - a i em Jv gy J i. a) ONO TAWHYS: TA oe | f J 7h ue wad . » J) 454 f Lateran fi) . ‘ ‘\) eit ie s eit tien sf ' i 4 Pa dj viola 2% Ey COME, US. ORIGINAL ARTICLES. Arrow, GILBERT J. PAGE. Some New West Indian Species of the Melolonthid Genus Lachnosterna 189 AustTEN, Masor E. E. The Percy Sladen Trust rape to the Indian Ocean in 1905, and 1907-1900, © Diptera > Tabanidae?..... es eae 43 Notes on the Nomenclature of certain African Tabanidae (Sub-family Pangoniinae) with Descriptions of a new Genus and new Species.... 139 BARRAUD, Capt. P. J. Mosquitos collected in Palestine and adjacent Territories............ 387 Bezzi, Pror. M. Further Notes on the Lonchaeidae (Dipt.) with Descriptions of New Rarer ret APTI SA BIS css. cick ace e oe ving oye pee a cs 199 Brain, Dr. Cuas. K. ‘ ine Woceidac of South Atri¢a Vi. 2.55 dee eee ae then bie 1 Buxton, P. A. A Liparid Moth (Ocnerogyia amanda, Staud.), destructive to Figs in Ble aGU OfatA ay 58. onl) soe euetics op on LW Mes Ade, s, Sole NRE 181 Insect Pests of Dates and the Date Palm in Mesopotamia and else- er ere eee may fee 9 sl een het, «age ap 287 CorBETT, G. H. Observations on Cotton Thrips in the Gezira, Blue Nile Province, ant) mi See ero cad deo eed Me A pean! fees ah 95 D EMMEREZ DE CHARMOY, D. Notes on Insects accidentally introduced into the Island of Mauritius 171 D’ EMMEREZ DE Cuarmoy, D. & Tempany, Dr. H. A. The Campaign against Phytalus smithi in the Colony of Mauritius.... 159 DauziEL, Dr. J. M. Crab- holes, Trees and other Mosquito Sources in Lagos............. 247 Epwarps, F. W. Notes on the Mosquitos of Madagascar, Mauritius and Réunion...... 133 On the British Species of Simulium—II. The Early Stages ; with Corrections and Additions to PartI.................ceeeee eee: 211 Fraser, Mayor A. D. Notes on the Blood-sucking Flies in North Russia during the Summer Se AOLS,..1 hall eet dh. ORRIN cage ns x9 oot dubisinieses 195 IneraM, Dr. A. & Macrin, Dr. J. W. 8. The Karly Stages of West African Mosquitos—V. Culex decens, Thee., and Culex imvidioaua, Theo... .... 6.66 e eee ese eee: 105 (741) Wt. P2/154, 1,000. 7.21. B.&F.Ltd. Gli/i4. Vi. CONTENTS JACK, RupPErRT W. Some Notes and Remarks on the Bionomics of Glossina morsitans.... 113 LAMBoRN, Dr. W. A. Some further Notes on the Tsetse-Flies of Nyasaland.............. 101 The Habits of a Dipteron predaceous on Mosquitos in Nyasaland .. 279 Luoyp, Capt. LI. On the Reasons for the Variation in the Effects of Formaldehyde as Sl OlMGmst OPEL ONISP ted: i naicts «cede ia Peis jet a hd nis. > ee’ oT] 9? 99 25 36 10 24 » writer 99 D. read (viii, pp. 93-102) 9? 3° » Rev. App. Ent. i ». (PL , fig. 46) winter Dr. | Rev. App. Ent., B, i (Pl. XVII, fig. 1) - VOL. XI. Part 1.—pp. 1-94. ; AUGUST, 1920, BULLETIN OF — ENTOMOLOGICAL RESEARCH rss sU7) tee SMe. . ) L 2 « wf AW*ga 4 . APRi¢ 1923 \ . f ie hiya Qf Luo; bh \ NiAN OS LONDON. THE IMPERIAL BUREAU OF ENTOMOLOGY, 88, QUEEN’S GATE, S,W. 7. Price 5s. net. All Rights Reserved. IMPERIAL BUREAU OF ENTOMOLOGY. HON ORARY COMMITTEE OF MANAGEMENT ‘THE VISCOUNT HARCOURT, Cbafrman. Lrevt.-Cotonen A. ALCOCK, C.LE., ERS. Mayor E. E. AUSTEN, D.S.0. 7 a ff Dr. A. G. BAGSHAWE, C.M.G: | Major-GEenerat Siz JOHN R. BRADFORD, K.C.M.G., E.B.S. - Masor-GenERaL Siz DAVID BRUCE, K.C.B., F-.R.S. Mr; J: C. F. FRYER. | Siz SF. HARMER, K.B.E., FBS. | Pros. H. MAXWELL LEFROY. i Hon. E. LUCAS. “Dr. R: STEWART cachoseate Sir JOHN McFADYHAN. : Str PATRICK MANSON, G.C.M.G., F.R.S. | Siz DANIEL. MORRIS, K.CMG ee Pror. R. NEWSTEAD, F.R.S. : Pror. G. H. F.. NUTTALL, F-.R.S. Pror. E. B. POULTON, FBS. Lizut,-Cotonet Str DAVID PRAIN, C.M. G., Ci LE., E.R.S. Sir H. J. READ, K.CMG., CB. 4 Tun Hon. N. C. ROTHSCHILD. : | De. HUGH SCOTT. — | : * Sin A. E, SHIPLEY, G.B.E., PRS. | | ) Mr. R. A. C. SPERLING, C.MLG. om Sm STEWART er Mr. F. V. THEOBALD. Mr. C, WARBURTON, | | Director. | Dr. GUY A. K. MARSHALL, C.M.G. Fssistant Director. ie Dr. S& A. NEAVE, ~ | Secretary. | Mr. A. C. C. PARKINSON. IMPERIAL BUREAU OF ENTOMOLOGY. BULLETIN OF mNTOMOLOGICAL RESEARCH. Vou. XI. 1920. THE COCCIDAE OF SOUTH AFRICA—V.* By Cuas. K. Brain, D.Sc. Division of Entomology, Pretoria, South Africa. Subfamily LecanimnaE Key to Genera Genus Lecanium, Burm. Saissetia, Deplan. Hemilecannum, Newst. Aclerda, Sign. Allopulvinaria, gen. nov. .. Protopulvinaria, Ckll. Pulvinaria, Targ. Ceronema, Mask. Tnchtensva, Sign. Filippa, Targ. Conofilippra, gen. nov. Ceroplastes, Gray Inglisia, Mask. Cryptinglisia, Ckll. .. Parafairmairea, Ckll. Ceroplastodes, Ckil. .. Idiosatssetia, gen. nov. Membranaria, gen. nov. (Piatses I-LV.) CONTENTS. PAGE 25 26 36 38 39 40 40 4] Subfamily LECANIINAE. This subfamily comprises a large number of genera, which are, as a rule, easily distinguished by the presence or absence of secretionary matter in one form or another. All agree, however, in having the anal ring with a number of hairs and, further, in having the hind margin cleft. The anal ring is covered above by a pair of more or less triangular plates which are usually densely chitinised. The eighteen genera known to be represented in South Africa are tabulated below :— *For Part II, sce Bull. Ent. Res. ix, p. 107; Part III, op. cit. ix, p. 197; Part IV, op. cil. X, p. 95. (681) Wt.P4/140. 1,000 7.20 B.&F.,Ltd. Gp. 11. A 2 CHAS. K. BRAIN. Key to South African Genera. A. Female naked, or only covered by a very thin, inconspicuous film of secretion. B. Adult 9° with well-developed legs and antennae. C. Adult 2 with posterior extremity cleft. (1) Female flat or slightly convex, derm not very hard when mature, without polygonal areas containing pits .. A Lecanum, Burm., p. 3. (2) Female becoming hard and convex at maturity, derm with + oval or polygonal areas containing _ pits ae Sarssetra, Deplan., p. 2. CC. Adult © with posterior Baas not cleft; female with four large groups of spinnerets .. ; a .. Hemilecanmum, Newst., p. 14. BB. Adult 92 without legs or ae - a .. Aclerda, Sign., p. 15. AA. Female secreting a greater or less amount of cottony or powdery matter. B. Female naked above, secretion beneath or behind the insect. C. Female with only a narrow fringe of secretion. (1) Female very convex, dark red, abdominal segmentation very distinct Allopuluinaria, gen. nov., p. 16. (2) Female flat or slightly convex, segmentation not pronounced, Protopulvinaria, CkIL., p. 16. CC. Female secreting a definite protruding ovisac Pulvinaria, Targ., p. 17. BB. Female partly covered with secretion. (1) Adult 2 covered above, except for the median area, with a thick coat of wavy threads which project beyond the margin in all directions Ceronema, Mask., p. 22. (2) Adult 92 enclosed in a felted sac, ne the cephalic extremity which is + exposed Bs ; . .Inchtensia, Sign., p. 22. BBB. Adult 9 entirely enclosed. C. Sac fitting closely to body of adult 9 at maturity and later serving as ovisac ; insect leaf-infesting .. 2 . Filipyra, Targ., p. 23. CC. Sac felted, stout, conical, insect ah ee Ganctienal gen. nov., p. 25. AAA. Female secretion waxy, glassy or horny. B. Covering of 2 consisting of wax, generally soft and thick ; no marginal fringe nor marginal processes ; female with a + conspicuous caudal process visible on removing the wax .. ie Ceroplastes, Gray, p. 26. BB. Covering of 9 glassy, or at ie pide, thin. (1) Scale divided into Se and striated with rows of air-cells ; not causing galls vas : me Inglisia, Mask., Dp. 36. (2) As in Inglisra iit eateaae aaille on ene of host-plant Cryptinghsta, Ckll., p. 38. (3) Scale divided into two halves not striated with air-cells, but with grooves radiating from apex bi ne

with an apical pore, those of the pale broad margin of the ordinary polygona! type. There are also four groups of spinnerets; two towards the anterior and two towards the posterior extremity, each group composed of several hundre:i spinnerets forming well-defined dark chitinised areas. Antenna of nine segments, of which the third is the longest ; there are a few long hairs on the four terminal segments ; formula 3, 4 (5, 2, 9) 6, 1 (7,8). Legs highly chitinised ; slender but small compared with the size of the insect ; coxa almost equalling the length of the femur; digitules simple. Marginal spines with their broad bases suddenly contracted, each fitting into a well-defined socket, the latter being attached to a short subcutaneous tube. Stigmatic channels and spines absent. “ Length 13-15 mm. ; width 12-13 mm. er THE COCCIDAE OF SOUTH AFRICA. 15 “Second stage female broadly ovate, slightly narrowed posteriorly, marginal spines continuous, resembling those in the adult. Antennae of seven segments, the third being the longest, the rest of the segments subequal in length. Legs scarcely longer than the antennae; coxa rather broad. Mentum uniarticulate. Groups of spinnerets occupying relatively the same position as in the adult, but there are only about 60-70 individual spinnerets in each group; they are also larger and more distinctly separated than in the adult. Derm in the region of the anal cleft finely squamose, with a large subcutaneous tube ; there is also a similar tube just within the margin opposite the anterior stigmata. “ Larva elongate; position of the compound spinnerets as in the adult and nymph. Mentum monomerous. Antennae of six joints, the third equalling the length of the fourth, fifth and sixth together. Marginal spines forming a continuous series.” Habitat : On tree euphorbia, East London, and stems of oleander, Capetown (Fuller, 1898). Collection No. : 107. Genus Aclerda, Sign. Adult 2 naked, a mere sac containing eggs at maturity, with normal mouth parts, but legs and antennae absent. Larva greatly elongated, with parallel sides. 210. Aclerda digitata (CklI.), Pseudolecanium digitatum, Ckll., Ann. Mag. Nat. Hist. (7) ix, p. 24, 1902. Aclerda digitata (Ck) Fernald, Catalogue, p. 210, 1903. Professor Cockerell’s description is as follows :— “ , pyriform, with the hind end pointed ; about 2} mm. long; dark ferruginous, shiny, producing some cottony material. End of abdomen strongly chitinized, with the form usual in the genus. The diagnostic characters, as in all species of the genus, are derived mainly from the larva, the female being a mere bag of eggs. “ Larva extremely long and narrow, length 600, breadth 160. Anterior extremity truncate, crenulate, with six short blunt, finger-tip-like spines; a row of fifteen blunt spines down each side of body; no dorsal spines; legs well developed, anterior tibia 60, its tarsus 30u long; the tibia has a constriction about 27 from base, making it look almost 2-jointed, this being more or less apparent on all the legs ; tarsal digitules long, with small knobs; claw-digitules shorter, filiform, knobbed. Antennae 24 apart at base and 30u from anterior end of head; 6- jointed, joints measuring: (1) 18, (2) 12, (3) 27-33, (4) 20, (5) 21, (6) 30. Anal ring small, circular, without bristles on its margin, but posterior to it are set four bristles, of which the inner two are shortest ; anterior to anal ring is a row of four finger-like blunt spines ; on each side is a caudal bristle, about 190 long, and beyond each of these a pair of finger-like spines; anterior to and a little laterad of each caudal bristle is a small round gland. The last two segments have each a pair of bristles on the ventral surface, those on the last being twice as long as those on the penultimate segment.” Habitat : On grass, Natal; collected by C. Fuller, 1901. This species has not been found again. Collection No. : 134. 16 CHAS. K. BRAIN. Genus Allopulvinaria, nov. Adult 2 naked above, very convex and distinctly segmented. Lower surface concave, filled with a compact powdery mass of wax which projects around the margins. Antennae somewhat rudimentary, of 4, 5 or 6 segments. Legs well developed. Anal ring with numerous (12?) hairs. Larva elongate and narrow ; anal tubercles produced, each with one long seta. Type, A. subterranea, sp. n. 211. Allopulvinaria subterranea, sp. n. (Plate vi, fig. 246). Larva active, pale pink, eyes pigmented, body elongate and narrow. Anal tubercles long, each with one long seta. One-fourth grown 2 with 4-jointed antennae, and marginal hairs as in Lecaniwm. Adult @ viviparous, 7 mm. long, 45 mm. broad and 4 mm. high. Dorsum roundly arched ; segmentation distinct, the segments roundly ridged. The body is red, faintly flecked with white wax. Underside flatly concave, abdominal portion plainly segmented. Legs small, brown. Anal plates small, brown, from between which protrudes a double fine pencil of white wax. Lower concavity of body filled with cushion of white wax which extends slightly beyond the margin of the body; this is not cottony but compact. No eggs were seen, but larvae were present in the white waxy cushion and in the body of the Q. Antennae 4, 5, or 6-jointed, usually 5. Anal ring with numerous (12?) hairs. Anal plates raised, appearing excavate so as more or less to enclose the anal tube from the sides as well as above. Anal groove closed. Integument with numerous simple glands and short acute spines scattered over the surface; the former especially noticeable in transverse series across the segments. The following measurements may be useful for comparison. Antennae: (qa) 4-jointed type :— I W I IV. (a) 40 27 61 34u (b) 34 40 65 b4u (b) 5-jointed series, range in yu :—(1) 34-45, (2) 3440, (3) 51-68, (4) 15-34, (5) 23-40; (c) 6-jointed form, e.g. (1) 41, (2) 34, (3) 61, (4) 17, (5) 10, (6) 20. Leg II: coxa 81 by 108, femur + trochanter 150, tibia 74, tarsus 74. Habitat: On stems of “ quick” grass, near bank of stream ; collected by T. L. Watermeyer, Jonkershoek, Stellenbosch, C.P., July 1917. Collection No.: 138. Genus Protopulvinaria, Ckll. The few species which have been described in this genus are similar to Lecanwum in their early stages, but may be readily distinguished at maturity by the small white ovisac which is secreted beneath the adult 9. In the normal condition this is but slightly larger than the body of the female and protrudes as a white ring around the margins of the body (fig. 248). This genus is included in Pulvinaria by Mrs. Fernald in her Catalogue (p. 128), but it would appear that its correct position is between Lecanium and Pulvinaria. THE COCCIDAE OF SOUTH AFRICA. 17 212. Protopulvinaria piriformis (Ckll.), Lefroy. (Plate il, fig. 244; iv, fig. 248.) Pulvinaria (Protopulvinaria) pyriformis, CkIl., Jl. Trin. Nat. Club, i, p. 309, 1894. Pulvinaria (Protopulvinaria) pyrvformis, Ckll., Jl. Trin. Nat. Club, ii, p. 307, 1896. Pulvinaria newsteadi, Leon., Riv. Pat. Veg. vi, p. 279, 1898. Pulvinaria pyriformis, Ckll., Psyche, vii, p. 311, 1899. Protopulvinaria pyriformas, Lefroy, Scale Ins. Lesser Antilles, p. 43, 1901. Pulvinaria pyriformis, Fernald, Cat., p. 138, 1902. Pulvinaria plana, Ldgr., Jahrb. Hamb. Wiss. Anst. xxxii, p. 34, 1911. Protopulvinaria piriformis, Ldgr., Die Schildliuse, p. 199, 1912. Ovisac slightly larger than the body of the adult 9, from which it extends as a uniform ring around the margins (fig. 348). Ova pale greenish white, regularly oval. Adult 2 about 3 mm. long and about as broad slightly behind the middle ; broadly rounded behind, suddenly narrowing in front so as to be short pear-shaped or drop-shaped. The margins of the body are flat ; the central part is flatly rounded. In living specimens the general body colour is yellowish or pinkish with the margins reddish brown. The subdorsal area is irregularly suffused with mauve or violet. The anal cleft is deep, extending almost to the centre of the body, but the sides are in close proximity, so that the posterior margin in most cases appears uniformly broadly rounded. The anal plates are very long and very narrow and are brownish in colour. Dead, dry specimens are pale to deep brown according to age. When stained, cleared and mounted the following characters may be noted :— Antennae 7-jointed ; range in w: (1) 27-31, (2) 37-44, (3) 44-51, (4) 51-54, (5) 20-24, (6) 20-24, (7) 51-54. Leg I: coxa 75, femur + trochanter 170, tibia 109, tarsus 54, claw 20wu. Leg II: coxa 99, femur + trochanter 180, tibia 112, tarsus 68, claw 24. Anal plates very long (470) and narrow (fig. 244). Marginal spines short, thin, with ends deeply branched, or occasionally bifid from base, in a close-set series. Stigmatic spines 3, laterals short and acute, median two or three times as long. Young form, about 1:2 mm. long, oval, translucent, appearing faintly greenish yellow, with six deep purple lines radiating from median area to margin, the four posterior being nearer together than the anterior two. Habitat : On undersides of leaves of avocado pear (Persea gratissima) ; collected by A. Kelly, Pietermaritzburg, April 1916. Collection No.: 77. Genus Pulvinaria, Tare. ‘* Adult female resembling Lecanium in the early stages and until the commence- ment of oviposition, when a loose cottony ovisac is secreted from below the posterior extremity of the insect, for the protection of the eggs. The body of the insect is tilted up during the formation of the ovisac, and often becomes much shrivelled and distorted, finally remaining as a small shapeless scale at the anterior extremity of the mass of ovisac. It is sometimes rendered still more inconspicuous by the partial overlapping of the secretionary matter, but is never completely enveloped. In all purely structural characters there is nothing to distinguish species of this genus (681) | B 18 CHAS. K. BRAIN. from those of Lecaniwm, so much so that, until the period of oviposition, it would be impossible to determine whether an individual should be placed in the one or the other genus.” (Green). 213. Pulvinaria aristolochiae, Newstead. (Plate ui, fig. 242). Pulvinaria aristolochiae, Newst., Bull. Ent. Res. vin, p. 19, 1917. Adult 9, immediately prior to forming the ovisac, about 8 mm. long and 4:5 mm. broad, with thin margins and the dorsum rising to a distinct central prominence : colour brown, with darker markings. The margin is supplied with a short fringe of thin pale hairs ; from this to the raised centre extend a number of dark lines. Male puparium about 3 mm. long, moderately convex, dull white, not glassy, with a distinct transverse ridge about the middle. Two white filaments protrude from the posterior extremity some days prior to the emergence of the adult ¢. Adults were emerging in July 1916. Antennae 8-jointed, range in « :—(1) 57-75, (2) 68-75, (3) 187-190, (4) 95-102, (5) 78-85, (6) 34-47, (7) 24-34, (8) 41-44. Leg I: coxa 185, femur + trochanter 400, tibia 262, tarsus 123, claw 34u. Anal plate about 250u long. Marginal spines short (27-35), many appearing truncate. As Professor Newstead did not have the opportunity of examining living material the above particulars are given to supplement his description which is as follows :— “Female, adult. More or less cordate in outline and rather flattened ; dorsum almost completely covered with a well-defined layer of flake-like wax which varies in colour from dirty grey to greyish brown. Antennae of eight segments ; 3rd, 4th and 5th unusually long, the two first-named swollen distally ; 3rd about twice the length of the 4th ; three long hairs on 2nd, one of which lies (in three examples) close up to the succeeding segment ; there is also a long distal hair on the 3rd ; two on the 5th; a single spine on the 6th, 7th and 8th, the last-named also with a few very short hairs. Legs stout, long; tarsus relatively very short, less than one-third the length of the tibia ; lower digitules long and very broadly spathuliform. Marginal spines stout, pointed, and placed rather close together ; stigmatic spines broken away in all the specimens, their points of attachment being continuous with the marginal series. Submarginal pores very large, continuous, but rather widely separated. Anal lobes with four stout spines near the apex, on the inner edge. Anal cleft short, usually a little less than one-sixth the entire length of the body. Anal ring with eight hairs. No derm cells present; but there are numerous circular spinnerets (? ventral), each having an inner concentric ring. Length 57-76 mm. “ Ovisac pure white and closely felted, long and generally tortuous. Length 10-20 mm.” This insect shows great similarity to P. jacksoni, Newstead. Both have remark- ably long joint 3 to the antennae, which in jacksoni is almost 120y, in aristolochiae about 180u. Further both produce extremely long ovisacs. Habitat: On stems of Erythrina caffra Thunb., (Kaffir-boom), Natal Coast. Collection Nos.: 79 and 84. THE COCCIDAE OF SOUTH AFRICA. 19 214. Pulvinaria floccifera (Westw.). Coccus floccifera, Westw., Gard. Chron., p. 308, 1870. Pulvinarva brassicae, Ckll., Can. Ent. xxvu, p. 135, 1895. Pulvinaria floccifera, Green, Ent. Mon. Mag. xxxiu, p. 72, 1897. Pulvinaria pharae, Lull, Ent. News, x, p. 237, 1899. Pulvinaria floccosa, Newst., Jl. Roy. Hor. Soc. xxii, p. 26, 1900. Completed ovisac about 6 mm. long and 3 broad, with parallel sides and rounded extremity ; pink in colour, from the numerous ova which are of that colour; the small amount of white cottony matter which is incorporated in the mass is incon- spicuous owing to the large number of eggs. ©, young stage: about 2 mm. long, translucent greenish yellow, with the median dorsal area suffused with a pinkish or brownish tint. Adult 2 about 3mm. long and almost as broad, moderately convex. The margins are pale translucent greenish yellow, the median dorsal area is reddish brown and the whole dorsum is more or less wrinkled. The crests of the ridges, especially those running to the margin, blackish brown. Anal plates small, of the same colour as the dorsal area or slightly paler. When the ovisac is almost complete, the body of the Q has shrunken to such an extent that the brownish part of the dorsum is not conspicuous, while the black crests of the ridges are close together and thus give the insect a much darker appearance. Antennae 8-jointed, range in w: (1) 34-88, (2) 44-51, (3) 65-72, (4) 34-44, (5) 34-41, (6) 24-27, (7) 24, (8) 44. Leg I: coxa 78, femur -++ trochanter 190, tibia 163, tarsus 102, claw 24w. Upper digitules very long, slender, with spherical knobs. Lower digitules stout, normal. Anal plate about 145u long. Just in front of the anal plates there are a few scattered, “rosette” glands and two pairs of long (108) hairs. Marginal spines about 50u long and 100u apart, slender, often with finely divided tips, alter- nating with a submarginal row of similar hairs of about half the size. Stigmatic spines comparatively short and stout, with the outer ends tapering to blunt points. From these to the spiracles extends a double row of small rosette glands. Puparium of 3 of the usual type with the segments of the plates very distinct. The whole puparium is opaque white, appearing finely striate in a longitudinal direction. Habitat: On a native plant (Solanwm sp. ) Durban ; collected by C. P. v. d. Merwe, July 1916. Collection No. : 82. A very large percentage of the young forms (about 1:5 mm. long) are very convex and shiny black, obvious!y parasitized. 215. Pulvinaria jacksoni, Newst. Pulvinaria jacksoni, Newstead, Jl. Econ. Biol. i, p. 155, 1908. Professor Newstead’s description is as follows :— “ Ovisac from two and a half to seven times the length of the female ; breadth equal to the width of the insect ; low convex, sides parallel ; closely felted, tough and web-like in texture. Length 16°75-42 mm. ‘681) Rn? 20 CHAS. K. BRAIN. ‘* Adult female—Dried examples sienna brown or dull ochreous; the pale examples with yellowish mottlings indicating the position of the derm-cells. Form normal. Antennae of eight segments, the third being much the longest; last four subequal ; there is a very long hair on the third and fifth, the one on the former the longest. Legs stout ; trochanter about one-third the length of the femur, with a very long apical hair; digitules to claws and tarsi normal. Derm with very small incon- spicuous glands. Marginal spines short, closely set, truncate, the ends often notched. Stigmatic clefts with pointed spines and two large tubercles. “Male puparium, much stained with brown, normal in shape, but thick and un- usually wax-like in appearance; dorsal plate with one large central, white, wax- like felted spot and several brownish tubercles immediately beyond it. Length 2°25 mm.” The following measurements are added for comparison with the other South African species :— Antennae 8-jointed ; range in w: (1) 48-54, (2) 54-68, (3) 102-116, (4) 68-75, (5) 61, (6) 31-34, (7) 31-34, (8) 37-41. Leg I: coxa 139, femur + trochanter 354, tibia 230, tarsus 108, claw 27. Anal plate about 240u long. Habitat: On Trichitla sp., Durban ; collected by A. Kelly, May 1918. Collection No.: 79a. 216. Pulvinaria lepida, sp. n. Adult °° with ovisacs clustered on stems and leaves of grass. Adult 9 averaging 2 mm. long by 1°3 mm. broad, moderately arched, ight brown in colour, dorsum with transverse wrinkles; marginal area flecked with white wax. Ovisac white, elevated, extending behind the insect for about the length of 2 body, faintly fluted on upper surface. Numerous 9° parasitised. Mounted specimens contain numbers of well developed embryos. Antennae variable, of 7 or 8 segments; terminal joimt appearing truncate ; range in wu :—T-jointed : (1) 34-37, (2) 37-48, (3) 54-65, (4) 70-82, (5) 24-27, (6) 24-27, (7) 37-41; 8-jointed : (1) 34-87, (2) 44-48, (3) 51-58, (4) 34-41, (5) 44-48, (6) 24-27, (7) 24-27, (8) 37-41. Leg I: coxa 78, femur + trochanter 187, tibia 153, tarsus 102, claw 24y. Anal plate about 153 long, with 5 comparatively short hairs at the apex. Mar- ginal spines in a single row, simple, pointed, about 41u long and 20-30 apart, straight or only slightly curved. Stigmatic spines: laterals short, acute ; median stouter, longer, blunt. Halitat: On stems and leaves of common veld grass, Standerton and Pretoria (K. Munro). Collection Nos. : 139 and 328. 217. Pulvinaria mesembryanthemi (Vallot), Sign. (Plate i, fig. 235). Pulvinaria mesembryanthemi, Vall., Bull. de Ferussac, xxu, p. 469, 1830. Calypticus mesembryanthemi, Costa, Ann. Acad. Asp. Naples, p. 273, 1844. Pulvinaria biplicata, Targ., Catalogue, p. 34, 1869. Pulvinaria mesembryanthema, Sign., Ann. Soc. Ent. Fr. (5) ii, p. 39, 1873. THE COCCIDAE OF SOUTH AFRICA. 21 Adult 9.to time of forming ovisac, and younger stages, green, of the same tint as the fleshy leaf of Mesembryanthemum edule, its most common food-plant ; attaining a length of 5 mm. and an almost equal breadth, moderately convex, dorsum smooth. As the ovisac is produced the body of the 2 becomes yellowish and later yellowish brown, much shrunken, with four transverse ridges and, ultimately, contorted or bent backward (fig. 235). Antennae 8-jointed; range in w: (1) 48-55, (2) 48-54, (3) 75-85, (4) 44-58, (5) 27-37, (6) 24-34, (7) 27-31, (8) 46-51. Leg I: coxa 122, femur + trochanter 290, tibia 211, tarsus 126, claw 34u. Anal lobes approximately 165 long. This insect is common throughout South Africa on Mesembryanthemum spp., especially M. edule, Linn. It becomes so numerous in some seasons that it kills patches of this plant when grown in parks, etc. This was the case at the Eastern Sports Grounds, Pretoria, in November 1914. The 3g emerged from the next generation about Christmas 1914 and the following particulars are given from fresh material obtained at that time :— Male test, transparent, white, glassy, about 1‘5 mm. long, margins depressed, central plate raised, slightly keeled. Body, legs and antennae dark brown. Wings broadly rounded, iridescent, dorsal sclerites and eyes shiny black. Two caudal waxy filaments, white, as in Pseudo- coccus, length equal to head + body without antennae. Length of head + body 1:17 mm. (without genital spike) ; length of wing 1:0 mm.; width of wing 0°5 mm.; length of antenna 0°84 mm.; length of genital spike 0°27 mm. ; length of caudal setae 1:0 mm. Antennae 10-jointed, the segments measuring: (1) 34, (2) 48, (3) 51, (4) 180, (5) 112, (6) 105, (7) 85, (8) 68, (9) 61, (10) 74. Habitat: On Mesembryanthemum spp., chiefly M. edule, Linn., throughout the Union. Collection No: 78. 218. Pulvinaria psidii, Mask. Pulvinaria psidu, Mask., N.Z. Trans. xxv, p. 223, 1892. Adult 9 pale transparent yellow with irregular black markings; 2 mm. long and 1:7 mm. broad. The ovisac is white, as wide as the body, always flat, attaining twice the length of the body. Ova pale yellow. Larvae pale yellow, hatching when received, 16th November 1914. Adult 9 flatly convex with slightly raised median keel. Anal plates pale brown. Antennae 8-jointed; range in ww: (1) 24-34, (2) 37-44, (3) 51-58, (4) 27-34, (5) 27-34, (6) 20-24, (7) 14-20, (8) 41-44. Leg I: coxa 68, femur + trochanter 190, tibia 145, tarsus 85, claw 271. Anal plate about 130u long. Margin with a row of thin spines about 34, long, tips divided but not dilated. These spines are set much closer together on the anterior part of the body. Integument clear, with a few simple glands with short 29, CHAS. K. BRAIN. spines and a few long hairs near the insertion of the antennae. Stigmatic spines ; laterals short, conical ; median longer (51y). Habitat : On guava, Botanic Gardens, Durban ; collected by A. Kelly, November 1914. Collection No.: 81 Genus Ceronema, Mask. Maskell’s original diagnosis of this genus, as given in the Transactions of the New Zealand Institute, p. 55, 1894, is as follows :— “ Female insects in the adult stage covered wholly or partially by tests of threads more or less closely woven, neither glassy nor felted, never forming homogeneous plates. No fringe. Form of insect Lecanid, with normal cleft and lobes. “ Larva Lecanid, showing cleft and lobes. ““ Male pupa covered by a glassy test of normal Lecanid form, composed of plates more or less homogeneous.” 219. Geronema mobilis, sp. n. Adult 9° about 3 mm. long, broad oval in outline, flat, with the dorsum almost entirely covered with long, white, waxy filaments. Around the margin the filaments are long and coarse, but those on the dorsum are short and fine and + curly and densely matted and appear almost felted. The colour of the insect, beneath the secretion, is caramel brown. When cleared and mounted the body is broad oval ; integument thin and trans- parent with numerous scattered, small, simple gland openings. Margin very closely set with a very short fringe of marginal spines, which are only about 20u long, thin, linear, two or three, usually two, arising from the same pore; they are unusually close together, averaging about 14u apart. Stigmatic clefts shallow, but distinct because of their chitinised sides; each cleft with two or three curved, blunt processes which are about as long as the marginal spines. Antennae 7-jointed; range in w: (1) 27-37, (2) 41-51, (3) 34-54, (4) 68-78, (5) 27-31, (6) 17-27, (7) 41-44. Joints 2 and 4 each with two long setae, 5 with one, 6 with two and 7 with eight. Leg I: coxa 75, femur + trochanter 187, tibia 105, tarsus 68, claw 24u. Anal plate about 140u long with | apical and 3 subapical spines. Habitat: On leaves of common native bush (Celastrus cordata, E.M.) (Celastri- neae) ; collected by C. Fuller, Illovo River, Natal, August 1916. Collection No.: 97. The adult 2 moves about quite readily. A number of mature specimens were put on one side when received with a view to photographing them, but next day, when the leaves had slightly dried, the insects had all crawled off and so disarranged the waxy filaments. Genus Lichtensia, Sign. Adult 9 oval or suboval, like a Pulvinaria, except that the secretionary sac covers the whole body with the exception of the cephalic end, which is usually more or less exposed. Antennae 8-jointed. THE COCCIDAE OF SOUTH AFRICA. 23 220. Lichtensia asparagi, sp. n. Female ovisac about 4:5 mm. long, elongate, narrow and very convex, sordid white in colour, thin but dense, with the anterior extremity open, exposing the front portion of the dry female. Male puparium moderately large, elongate, very convex, with perpendicular sides, thin, semi-opaque, dorsum flatly rounded, entirely without median plate. There is a faint indication of sublateral lines, otherwise the surfaceis uniformly stippled. Adult 9, mounted, broad oval, about 4 mm. long, with the two extremities about equally rounded ; anterior extremity not produced. The integument appears remarkably free from conspicuous glands or hairs. Margin with a single scattered row of short sharp curved spines. Stigmatic spines stout, somewhat curved, with rounded ends. Antennae 8-jointed, in one antenna examined 9-jointed, with the following range in w: (1) 35-48, (2) 57-60, (3) 68-78, (4) 51-71, (5) 40-61, (6) 30-37, (7) 25-34, (8) 30-38. Leg I: coxa 88, femur + trochanter 228, tibia 190, tarsus 102, claw 24. Anal plate about 185 long, with longitudinal folds. Habitat: On Asparagus capensis, L. (Liliaceke), Eastern Cape Province. Collection No. : 131. Genus Filippia, Targ. Female flat, oval, legs and antennae normal. Side of body and dorsum with numerous tubular glands. Female, at maturity, entirely enclosed in a flat felted sae which, after oviposition, serves as an ovisac. 221. Filippia chilianthi, sp. n. Adult 9 entirely covered with a dense, closely felted layer of white secretion, about 6 mm. long and 3 mm. broad, + regularly oval, with the dorsum moderately convex. ‘This secretionary layer is secreted slowly from all the upper parts of the body, the glands concerned being scattered. This is not the case in the Pszupo- COCCIN AE, in which the ovisac is formed by secretionary glands in the posterior and marginal portion of the body only. Little or no secretion appears on the ventral surface, so that the leaf or twig forms the floor of the sac; the marginal secretion becomes attached and entirely encloses the 2. As oviposition proceeds the body shrinks away from the posterior end of the sac which becomes almost filled with ova. At death the dry shrivelled body of the 9 occupies a small portion at the anterior end. Prior to secreting the ovisac the 2 is from 4 to 5mm. long, and rather more than half as wide. The sides are almost parallel for more than half the length of the body, the posterior end appearing broadly rounded, and the anterior end more pointed. In colour it appears greyish brown with a lighter median, flatly rounded, keel. When seen under a hand-lens, however, it is noticed that the ground-colour of the body is yellowish, and semi-transparent, speckled with black. The anal cleft is short, with the sides closely adjacent. The anal plates are small and brown. The lower surface is slightly concave, greasy-yellowish in colour, the black speckling of the dorsum only showing faintly through at the extreme margins. The stigmatic bands are represented by two short faint white lines on each side. 24 CHAS. K. BRAIN. At a younger stage—when about 2°5 mm. long—the body is red-brown, rugose at the edges, with the median keel not prominent but smoother than the remainder of the body. The antennae are 8-jointed; range in uw: (1) 31-45, (2) 50-62 (3) 95-108, (4) 54-65, (5) 32-44, (6) 27-34, (7) 27-34, (8) 31-37. Leg I: coxa 102, femur + trochanter 245, tibia 165, tarsus 85, claw 20u. Tarsal digitules moderately long (70x), slender hairs with their distal ends slightly clubbed. Lower digitules short, broad. Marginal spines short (27), truncate. Anal plate about 155u long. Stigmatic cleft with three comparatively slender spines, laterals about 34, median similar but longer (51,1). Male puparium about 2 mm. long and 1 mm. wide, dark red when containing the insect ; white and glassy when empty. The divisions of the test are very prominent, opaque white ; the remainder being + wrinkled and almost hyaline. Adult 3g were emerging in large numbers when the material was received, 5th April 1916. Copulation was observed in a number of cases, the 9 in each case measuring about 4mm. in length. The body of the male was curved downward admitting the sheath into the cleft of the Q. During this process the two long waxy filaments of the J stood erect. The adult ¢ is about 1:0 mm. long, rich red-brown in colour, with two long waxy caudal filaments about half as long again as the head and body combined, without the antennae. The body, legs, and antennae are all red-brown. The head is slightly darker. The scutellum is polished at the edges, with a sunken, matt, central patch. The wings are hyaline, matt, extremely iridescent, with a strong sub-costal, cochineal red band. When mounted, the following measurements may be taken as an average :— Head and body, without antennae .. se mn os ul taba pe Antennae... nf ai ae fe vf .. 0847 mm. Genital spike .. sys i oh Ss - Je fe. em. Wings : length 1-2 mm., width Hs she be 3} OG, ain Caudal filaments 20 2 ft £20087, om: Habitat: On a native shrub ; collected by C. Fuller, [lovo River, Natal, April 1916. On leaves of Chilianthus oleaceus, Burch (Loganiaceae), Bloemfontein, 0. F.S.; collected by J. C. Faure, April 1916. Collection Nos. : 127 and 128. This insect, with the ovisac completed, is similar in appearance to F. oleae (Costa) but, according to Signoret’s description of that species, the adult has 6-jointed antennae. 222. Filippia carissae, sp.n. (Plate i, fig. 236). This species differs from F’. africana in the following respects :— The ovisac is a little larger and generally less compact ; often split around the edges (fig. 236). The antennae are more variable, 7 or 8-jointed, range in wu :— THE COCCIDAE OF SOUTH AFRICA. Zo (a) 8-jointed: (1) 41-54, (2) 51-54, (3) 102-119, (4) 68, (5) 37-44, (6) 34, (7) 24-27, (8) 34-37; (b) a specimen with 7-jointed antennae gave the following measure- ments :— 1. 54 68 105 122 34 27 41 il. rt aR 109 68 51 37 68 Leg I: coxa 95, femur + trochanter 240, tibia 204, tarsus 88, claw 20u. Anal plate 176u, rather elongate, with one or two long spines at apex. Margin with a single row of + plate-like spines, narrow at the base and gradually broadening at the distal ends. Stigmatic clefts somewhat sunken, with three spurs of moderate length. Habitat: On leaves of Carissa grandiflora, A. DC. (Apocynaceae), Natal Coast ; common. Collection No. : 129. Genus Conofilippia, nov. Female Lecaniid, flat, with well developed legs and antennae; dorsum with numerous sharp spines; margin with slender spines and stigmatic cleft similar to Lecanium. At maturity the insect is entirely enclosed in a dense felted sac, which is elevated in the form of a high cowl. Antennae 7-jointed. Type, C. subterranea, sp. n. 223. Conofilippia subterranea, sp.n. (Plate iui, fig. 245). Adult °° enclosed in conical felted tests on the roots of a native shrub. Test about 6 mm. long, 5 mm. broad and 5 mm. high (figs. 245a, 6). Adult © flat, about 5°5 mm. long, 4mm. wide, smooth, glossy, pink to red in colour, brown and wrinkled when dry, on a thick mat of dense powdery material (the floor of the test) which is buff to yellow-brown in colour and dusted with white wax beneath the insect. In one or two cases the top of the test shows a circular orifice, but this does not appear to be normal and looks as though it had been made for the exit of some Hymenopterous parasite. Largest 2, when mounted, about 6 mm. long, and 4:2 mm. wide. Integument clear, with a number of simple glands which are particularly abundant near the margin, and with numerous short (17) sharp spines. The margin is thickly set with a compound series of slender spines about 24 long, which are somewhat curved near their tips. Stigmatic clefts with three blunt spurs, laterals about 27, median about 371. : Antennae 7-jointed; range in mw: (1) 35-40, (2) 25-34, (3) 40-47, (4) 37-42, (5) 34-44, (6) 17-30, (7) 37-51. In one case joints 6 and 7 were united, forming a segment 64 long. Leg I: coxa 102, femur + trochanter 220, tibia 170, tarsus 100, claw 40u. The junction of the tibia and tarsus is swollen and suddenly narrows towards the distal half of the tarsus. The anal lobes are rounded, about 230m long. 26 CHAS. K. BRAIN. Habitat: On roots of native shrub with red stems; collected at De Wildt, Pretoria District, Transvaal, by Claude Fuller, May 1915. The drawings (figs. 245-b) were kindly made from living material by the collector. Collection No.: 137. Genus Geroplastes, Gray. Adult 2 completely enveloped in a more or less dense covering of wax ; no marginal fringe or radiating processes. Sometimes the waxy covering exhibits a variety of arrangement in the form of definite plates or plaques. In other species the covering is uniform in texture and is then most often soft, and contains a large per- centage of watery fluid. On removing the wax a caudal prominence is generally visible. Legs and antennae present, well developed. 224. Geroplastes bipartitus, Newst. Ceroplastes bipartitus, Newst., Bull. Ent. Res. viu, 1, p. 25, 1917. “Female test. Colour of dried specimens very like pale dirty beeswax. In the young adults the test is broadly oval, somewhat hemispherical and divided into nine plates: three bilateral, one cephalic, one anal and one dorsal, the last- named with a conspicuous dark brown or blackish oval spot, with a central elongated patch of pure white wax; the nuclear spots to the lateral plates are smaller and generally much less conspicuous than the dorsal one. Margin over the stigmatic areas with a pair of laterally compressed and somewhat disc-shaped extensions,, each extension carrying on its edge a narrow strip of opaque white wax, the tip of which sometimes reaches the dark nuclear spot of the lateral thoracic plate. In very old examples the test has increased in thickness considerably, but this has been so much damaged in transit as to render it useless for descriptive purposes ; however, one can trace the curious marginal extensions, which are somewhat like a narrow-waisted and distorted bobbin, or the toy used in the once popular game ‘diabolo.’ Average length of young adults, 3 mm.; height 1-6-2 mm.; average length of old adults, 6 mm. ; height doubtful. “Female adult. Denuded of wax, hemispherical; caudal process very long, varying in length from one-half to a little less than one-half the length of the remaining portion of the insect. Submarginal tubercles small, but generally clearly defined ; one cephalic and three bilateral, the two over the stigmata slightly more pronounced than the rest. When examined under a high power lens, by transmitted light, these tubercles are seen to be traversed by clear cell-like tracts. forming an irregular reticulated pattern. It is reasonable to assume, therefore, that these may be the special set of glands which secrete the nuclear spots in the centre of the plates in the test. Derm relatively thin, but strongly chitinised. Pores minute, separated over a large portion of the dorsum by slightly varying distances equalling the length of one of the short segments of the antennae or two of them together. Stigmatic clefts relatively shallow, but very clearly defined ; spines. short, obconical, those at the extreme margin very minute and stud-like. Marginal spines or hairs not traceable. Antennae of six segments, the 3rd longer than the last three together. Legs normal. Length of denuded female, inclusive of the: caudal process, 4°5-4°6 mm. ; length of caudal process, 13-15 mm. TNE COCCIDAE OF SOUTH AFRICA. 27 “Male puparium. Consisting of two distinct parts ; the lower half boat-shaped, and of a glassy vesicular texture, as in those typical of the genus Lecaniwmn ; the upper portion opaque, low, convex, and of a dirty beeswax colour, with nine narrowly rectangular, submarginal patches of snow-white secretion. Anal cleft apparently obsolete. On the emergence of the male the whole of the upper portion falls away, leaving the ventral half attached to the food-plant. The line of cleavage between the upper and lower portion is clearly defined in those puparia from which the imprisoned male has not escaped. Length 1:6 mm. ‘South Africa, 1914 (de Charmoy).” (Newstead.) This species is apparently not represented in the collection of this Division. 225. Ceroplastes candela, Ckll. & King. Ceroplastes candela, Ckll. & King, The Entom. xxxv, p. 113, 1902. “©. Long 23, lat. 34, alt. 44; dark red-brown, elevated, with vertical s:des. Caudal horn a prominent stout spine, hardly $ mm. long, placed nearer the top of the scale than the base. Dorsum smooth and shining, with only a very small central raised line. Sides of insects with vertical stripes of dense secretion ; no wax, except that composing these stripes, between the insects, which are densely crowded together, their vertical sides contiguous. They rest on a thin substratum of wax, and are covered above with yellowish-white wax, about 1 mm. thick. The outlines of the insects are vaguely marked on the surface of the covering wax by a brownish stain. The wax, with the insects beneath, surrounds the twig as the wax does the wick of a candle; the whole mass is about 20 mm. diameter, that of the t vig being about 5 mm. “ Mr. King found the antennae to measure thus in 1 :— 4. halide Abeta 4 diag 029 ml ¢) Haan ©: i (3) a) Pou ok er. GS 8! 60 a8 ag 49 Rees ee ag 4 gh BB! SOR gd “Found by Mr. Fuller at Richmond, Natal. The nearest ally is an undescribed species from Paraguay, collected by Professor Bruner.” (Ckll. & King). This species is not represented in the collection. 226. Ceroplastes combreti, sp. n. Test of 2 about 3 mm. long, broad and high, conical, with the anterior side a little more precipitous than the posterior, which is somewhat excavate above. There are no plaques, but the whole body of wax is arranged in + distinct columns, three on each side. The apex is blunt, bearing an opaque white ridge surrounded by six opaque white spots—the tops of the columns. Between the two most prominent lateral ridges are the distinct white stigmatic bands which extend from the base to the crown. The colour, when fresh, is bright rose-red with darker transverse marks. The aninnae are 8-jointed ; range in w: (1) 37-44, (2) 56-61, (3) 34-48, (4) 27, (5) 37-4 ', (6) 20-24, (7) 17-24, (8) 37-41. Leg I: coxa 75, femur + trochanter 180, tibia 129, tarsus 85, claw 20. 28 CHAS. K. BRAIN. Anal plates + semicircular, about 160u long and 85u broad. The derm is thin and transparent, with numerous small, scattered, simple glands, from some of which short, tubular projections arise. Stigmatic clefts with large numbers of very small, roundly conical, thimble-shaped spines. The distinct, long, opaque white lines, the reddish colour, soft wax, etc., suggest quadrilineatus of Newstead, but this latter species has 6-joited antennae and is obviously quite distinct by other characters of the test. Habitat: On stems of Combretum sp., De Wildt, Pretoria District ; collected by Claude Fuller, July 1918. Collection No.: 317. 227. Ceroplastes destructor, Newst. Ceroplastes ceriferus (Anderson) Newstead, Bull. Ent. Res. 1, pp. 66, 195, 1910. Ceroplastes destructor, Bull. Ent. Res. vii, p. 26, 1917. “ Female test. White, creamy white or dirty white; exceedingly soft and con- taining an excess of moisture. Form irregular, with large but ill-defined gibbose protuberances ; sides usually with two narrow opaque lines of secretion from the stigmatic clefts. No trace of lateral plates. Length, 4-8 mm. “Female, adult. More or less hemispherical, with the sides often slightly com- pressed ; caudal process long; integument castaneous and highly chitinised, smooth and shining, and without fovea or lateral tubercles. Antennae of six segments, the 3rd being as long as the 4th, 5th, and 6th together; the last three segments with stiff and bluntly pointed, spimose hairs. Legs small; hind femora very short and often distinctly incrassate ; hind tarsi equal in length to the tibiae, or sometimes slightly longer. Claw very short ; lower digitules very long and stout ; upper digitules normal. Stigmatic clefts well defined, but relatively small; stig- matic spines very small and pointed, bases not constricted; basal attachment (disc) very large. Caudal process (after maceration) transparent and somewhat flexible ; sides with an irregular double row of short spinose hairs, and in addition to these there are two pairs of longer hairs (one pair of which is twice the length of the others) slightly ventral to the row of short ones and towards the distal extremity. Anal lobes short and highly chitinised. Dorsal pores very small, rather widely separated. Ventral integument opposite the caudal process, with rather extensive groups of circular pores, many of which, in well cleared specimens, are linked together with lines of dark chitin. Length 4-7 mm.” (Newstead.) Habitat: On custard apple and avocado, Nelspruit, Transvaal; collected by D. Gunn, September 1915 (Coll. No. 93). On syringa (Melia azedarach), Bechuana- land, October 1918 (Coll. No. 336). Collection Nos. : 93 and 336. 228. Ceroplastes egbarum, CkIl. Ceroplastes egbarum, Ckll. The Entom. xxxii, p. 127, 1899. Ceroplastes cristatus, Green, Ann. Mag. N.H. (7) iv, p. 190, 1899. Professor Cockerell’s description is as follows :— “Waxy female scales often crowded on the twigs, two or more coalescing ; about 11 mm. long, 10 broad, and 6 high, the wax extremely thick, not at all divided into plates, snow-white, here and there with a suffused pinkish stain. THE COCCIDAE OF SOUTH AFRICA. 29 ‘* ©. Denuded of wax 53-7 mm. long, 4 broad, 2}—3 high, very dark, with a dorsal hump but no lateral humps; anal horn a mere mammiform prominence. Boiled in caustic soda the denuded females give a purple colour, which on dilution with water appears pink, and soon forms a flocculent pink precipitate. On adding nitric acid a flocculent white precipitate appears, but the pink precipitate is not altered. Skin after boiling remains yellowish brown, chitinous, with scattered minute gland-dots. Stigmatic areas with numerous crowded gland-spots, and many short and rather thick simple spines, but no capitate spines. Legs dark brown, the parts measuring thus in w: Coxa, 120; femur with trochanter, 180: tibia, 128; tarsus with claw, 96 to 114. Tarsal digitules 60, slender, with a small knob. Claw digitules with very large round knobs, extending about 15u beyond tip of claw. Antennae apparently only 6-segmented, but the segmentation towards the end very obscure. The segments measure in w: (1) 45; (2) 60-69; (3) 66-78 ; (4) 51; (5) 69; (6) 72. Segment 5 has a deep notch which makes it look as if divided into two. “Young larvae under female about 4301 long and 230 broad, tinged with a warm reddish colour. Male scales small, elongate, and glassy.” In the description of C. africanum var. cristatus Green states that specimens from Natal differ from africanum (1.e., mimosae) only in the presence of a small dorsal crest corresponding to the position of the central scar. They were, however, larger, being 12 mm. indiameter. This would seem to indicate, moreover, that the antennae may be 6, 7, or 8—jointed. Habitat: On Acacia, Natal. Collection No.: 87. 229. Geroplastes egbarum fulleri, T. & W. Ckll. Ceroplastes egbarum subsp. fullert, T. & W. Ckil., The Entom. xxxv, p. 113, 1902. Adult 9, with waxy covering about 14 mm. long, 12°5 mm. wide and 7 mm. high. Waxy covering regularly domed with a slight depression in centre. Margins widely crenulate, with usually two conical deflected waxy masses which clasp the stem. Stigmatic cleft waxy appendages conspicuous, snow-white, long and _ slender. There are two of these on each side, usually 3 to 4°5 mm. long, about 1 mm. thick at attachment to waxy covering, but gradually tapering to their extremities. They usually he closely pressed to the stem of the host plant. The colour of the waxy covering is whitish to coral, pink, with distinct brown patches on the intermediate area and lighter zones around the dome. Female, denuded of wax, 7 mm. long, 5°5 mm. broad and 4 mm. high. Colour coral pink, more yellowish than the darker coloured wax. The extreme frontal margin, the stigmatic clefts and the caudal projection are dark castaneous and shiny. In older specimens, after oviposition, the integument becomes brown. The venter is flat or concave with the median zone sunken and segmented. The 4 white stigmatic bands extend inwards to this sunken area. The lateral margins are slightly excavate, the upper edge of the excavation being formed by seven depressed conical projections from the intermediate area. The anterior of these extends forward until nearly level with the front edge of the rounded anterior lobe. 30 CHAS. K. BRAIN. The three lateral projections of each side are shorter and more bluntly pointed. The depressed area which separates the central cone from the intermediate area is coarsely punctate. The central cone is regular, elongate, oval at the base and roundly pointed at the apex. There is no sunken area in the centre. The caudal projection is exceptionally short, bluntly conical, projecting at an angle of about 45° with the ventral surface. Antennae 8-jointed; range in mw: (1) 44-51, (2) 48. (3) 58-68, (4) 34-37. (5) 27-34, (6) 20-24, (7) 20-24, (8) 37-48. Integument clear, hyaline, except for the extreme margin, marginal expansions and caudal prominence, which are densely chitinised. The anterior margin is broadly rounded with a distinct parallel-sided chitinous band. Lateral margins with two almost circular expansions on each side. These are much larger than those found in mumosae. Habitat: On “ monkey rope,” Natal coast, Umbilo and Equeefa Rivers, Collection No.: 88a. 230. Geroplastes eucleae, sp. n. (Plate u, fig. 239). Adult 9 tests sometimes single on stem, often aggregated in dense masses, Test of adult 2 about 6 mm. long, 5 mm. wide and 5°5 mm. high, without plaques but with the lower portion forming a wrinkled fold at the base of a highly conical dome. The colour is a delicate green, when alive, with the stigmatic bands conspicuous (fig. 239); when dry, it is semi-transparent, greenish yellow, with two white thin streaks on each side just above the stigmatic clefts. The central dome is pointed, without any central pit or depression, and is distinctly separated from the lower portion of the test by a groove. There is no indication of a caudal prominence on the test. Female, denuded of wax, smooth, regularly domed, without caudal prominence, chitin pale brown and moderately thin ; portion surrounding the anal plates deep castaneous. Caudal prominence rudimentary, indicated by deeper coloured chitin. When cleared, the derm is moderately chitmous. That of the dorsum is very finely rugose with very scattered small, transparent spots. The denser, marginal folds have in addition a few larger holes. Antennae 6-jointed; range in w: (1) 24-30, (2) 37-40, (3) 91, (4) 17-24, (5) 24-27, (6) 40. The legs are moderately developed, normal. Stigmatic clefts with a series of short, conical, thimble-shaped spines, which extend in a single row for some distance along the margin on either side of the cleft, where they compose a double row. About the middle of the group is a large spine, about twice as large as the others. Within the double row is a collection of small simple glands like the circumgenital glands of the D1asPINaz. Habitat: On stems of several native shrubs, including Huclea sp., Ochna sp. ?, Pavetta sp. ?, etc., Pretoria ; collected by Miss E. Impey, January 1915. Collection Nos. : 90 and 342. THE COCCIDAE OF SOUTH AFRICA, 3] 231. Geroplastes longicauda, sp. n. Adult 2 covered with a very thick layer of soft, white wax forming a test like a large ceriferus specimen, i.e., a little more elevated than egbarum. Largest specimen seen measured 18 mm. long, 11 mm. wide and 12 mm. high; marginal area prominent, forming a wide fold at the base of the central dome. The waxy appendages from the stigmatic clefts only project slightly from the main mass of the fold. Adult 9, denuded of wax, bright brown in colour, about 5 mm. long without caudal process, which alone measures 3°5mm. The body is + star-shaped with three short, lateral pointed spurs and one anterior. These,are sharper and more prominent than those of ceriferus. The dorsum is very convex, rising with straight sides to an acute point. The caudal process is exceedingly long, two thirds the lengt' of body, piceous, shghtly tapering, and extends in a horizontal direction, 1.e., flat along the twig. Cleared and mounted, the insect is remarkable for the broad oval, thin, trans- parent body with a very long dense black tail. The integument is uniformly hyaline. without chitinous marginal discs such as those found in fulleri, etc. In stained material the integument of the dorsum illustrates a strange segmentation -+ in plates, 5 elongate transverse median ones and shorter laterals. The antennae are 7 or 8-jointed, e.g., (1) 40, (2) 37, (3) 40, (4) 47, (5) 68, (6) 30, (7) 27, (8) 374; or (1) 23, (2) 34, (3) 40, (4) 88 (with pseudarticulation), (5) 27, (6) 27, (7) 37. Stigmatic cleft thin, hyaline, with a patch of scattered, short, thimble-shaped spies and, within this, a large group of simple glands similar to the circumgenit.l glands of the DIASPINAE. Legs comparatively short, otherwise normal. Habitat: On stems of native shrub; collected by C. Fuller, Natal Coast, July 1915. Collection No. : 334. This species is very similar in many respects to C. ceriferus but may be readily separated by the larger size of the adult 9 test, the comparatively longer caudal process and the 7 or 8-jointed antennae. C. ceriferus has antennae 5 or 6- jointed, usually 6. 232. Ceroplastes mimosae, Sign. (Plate ii, fig. 241). Ceroplastes mimosae, Sign., Ann. Soc. Ent. Fr. (5) ii, p. 46, 1872. Ceroplastes africanus, Green, Ann. Mag. N. H. (7) iv, p. 188, 1899. “Tnsects crowded on the stems of the plant, so much so that the waxy covering of adjacent individuals becomes more or less confluent and the normal form of the test is difficult to determine. The tests appear as rounded masses of cream-coloured wax, each with a more or less distinct nipple-like prominence at the apex bearing a small spot of whiter substance. ‘ The usual opaque white bands from the spiracular regions are present, but very inconspicuous, scarcely extending beyond the margin. In some specimens a series of CHAS. K. BRAIN. of impressed arches on the sides of the test marks the position of the marginal plates. The waxy coating being thinner on the impressed parts, the arches appear darker, the colour of the body of the insect showing through the covering-matter. An isolated test averages 7°75 mm. long, 6°50 mm. broad, 5°75 mm. high. “Female, denuded of wax, reddish brown to dark brown, the whole surface strongly chitinised ; irregularly globose ; apex often with an oblong scar correspond- ing with the position of the early larval pellicle, but which becomes almost obliterated in the oldest examples. In the early adult the median is separated from the marginal area by a more or less distinct furrow, which is particularly marked where it meets the anal tubercle. In the older examples only this hinder part of the furrow remains. Cephalic area constricted off from the globose body, forming a trowel-shaped projection in front. Spiracular clefts deeply indented, thickly set with small conical spines, not constricted at the base. Marginal hairs very small, few and inconspicuous. Anal scales minute, inner edge straight, base and outer edge together forming a semicircle. Anal tubercle blackish, directed upwards. Derm with numerous glandular pores, which are more distinct on the darker marginal area. Antennae with either 7 or 8-joints. It is difficult to say which is the normal number, as the two varieties are about equally represented in the series under examination. With the 8-jointed form the formula runs :—3, (1, 2), 8, 4, 5, (6, 7). When there are seven joints only the formula is 3, (1, 2, 4), 7, (5, 6). In this latter case there is a tendency for the fourth joint to separate into two, and there is always a more or less distinct false joint in the terminal segment. Legs well developed; tarsus more than half length of tibia. Foot with 4 digitules, the unguals broadly spatulate, the tarsals fine knobbed hairs. “Length of fully developed female 5°50 mm., breadth 5:0 mm., height 4°25 mm. “The male insect is unknown in any stage.” (Green.) The material I have examined has a characteristic odour and has uniformly 8-jointed antennae, with the following range in 4: (1) 27-34, (2) 41-48, (3) 58-68, (4) 27-41, (5) 17-87, (6) 17-24, (7) 20-24, (8) 34-37. Habitat : On Acacia karroo, Cape Colony ; common but local. Collection No.: 85. 233. Ceroplastes myricae (Linn.). Coccus myricae, Linn., Syst. Nat. Ed. xu, i, p. 741, 1766. Columnea myricae, Targ., Catalogue, p. 35, 1869. Ceroplastes myricae, Sign., Ann. Soc. Ent. Fr. (5) 1, p. 39, 1872. “ Habitat ad Cap. B. Spei, in Myrica quercifolia. ““Magnitudo pisi minoris, semi-ovatus secundum perpendiculum, pallide incarn- atus, vertice obtuse acuminatus cum poro tenuissimo, postice supra marginem etiam porus est, margo cartilagineus, crassior albus, utrinque circiter septem torutis protuberans. “Dans Olivier, Encyclopédie, VI, 96, 8, nous trouvons une description presque identique : la femelle est presque de la grandeur d’un petit pois, le corps est d’une couleur rouge pale et de forme demi-ovale, le vertex est elevé et percé d’un petit point, tout le bord est cartilagineux, épais, blanchAtre, marqué de chaque cété de petits cordons élevés. THE COCCIDAE OF SOUTH AFRICA, 33 “Ce sont ces sept cordons élevés qui, spécifiant bien ’espéce, nous empéchent de Yattribuer aux nombreux individus que nous possédons et decrivons sous le nom de C. Vinsonu.” (Sign.) Myrica quercifolia, Linn. (Myricaceae) is a near relative of the waxberry plant which is common on the Cape Flats. Mr. C. W. Mally, the Entomologist for the Cape Province, has recently kindly examined a large number of plants in an endeavour to re-discover this species of Ceroplastes, but as yet without success. 234. Ceroplastes pallidus, sp. n. Test of adult 2 to 85 mm. long, 6 mm. broad and 4:5 mm. high; very much like alarge C. rusci in form, with 8 lateral plaques and median dome. The lateral plaques, however, are without “ nuclei ” and are a little more perpendicular. In old specimens the waxy covering is pale, semi-transparent, yellowish, or having an indistinct greenish tint. The central dome is moderately elevated, somewhat tapering, slightly glossy, with longitudinal and concentric striae. The central “nucleus” is elongate, glossy and a little darker in colour. Stigmatic clefts indicated by small white dots. When cleared and mounted the integument is all thin and hyaline, without anterior or marginal thickenings and with only a small area surrounding the anal lobes chiti- nised. Stigmatic clefts shallow, not chitinised, with short, conical, pointed spines which are in a single row for a short distance on each side, but broaden out to form a small triangular patch opposite the spiracle ; at no point, however, are there more than 5 rows. Within this series is a compact group of simple glands like the circum- genital glands of the Drasprnaz and also a few smaller glands and short simple hairs. The remainder of the margin has a scattered, single row of short, slender, sharply pointed spines. The antennae are 8-jointed; range in w: (1) 37-44, (2) 37-47, (3) 44-52, (4) 34-44, (5) 51-56, (6) 27-32, (7) 27-34, (8) 35-47. Legs well developed, normal; tibia long, with a constriction at about the middle. The caudal tubercle is comparatively small, a mere plate with the chitin appearing -- streaky and somewhat perforate. Habitat : On fig, Church Square, Pretoria. Collection No.: 102. 235. Geroplastes quadrilineatus, Newst., var. simplex, nov. Adult °° clustered in large masses on the twigs of the host plant. @ test about 5 mm. long, 4°5 mm. broad and 3 mm. high, very dark in colour, almost black, with the thinner parts of the wax appearing reddish. Marginal fold corrugated, lighter and more transparent in colour than the central dome, with two very distinct lines of opaque white on each side. The central portion is comparatively flat and trun- cate, separated from the marginal fold by a deep furrow in dry material, with the apex hollowed into a shallow cup containing an opaque white ridge. The ridge of this cup-like depression is not regular but consists of an anterior transverse ridge and two posterior lateral spurs, and the floor usually slopes gently forward. The (681) , C ot CHAS. K. BRAIN. anal opening to the test is surrounded by a distinct circular ring which is itself sunken in a deep depression with broadly rounded sides. In dry specimens the whole waxy material is roughened, hard and very brittle. The pair of larger divergent pytiform bodies referred to in the description of quadrilineatus are not present. When cleared of the waxy covering the insect is of a characteristic form, with the dorsum uniformly rounded and smooth and the marginal area produced in a series of broadly rounded bead-like lobes. The caudal prominence is extremely rudimen- tary, appearing as a flatly rounded hump, black in contrast with the deep red brown of the remainder of the insect. The integument, when cleared, is dense and is pitted like that of many species of Sazssetia. The stigmatic spines are in an elongate row around the shallow cleft, the series becoming double in the centre. The antennae are 6-jomted, three being very long and variable, e.g. :—(1) 27, (2) 37-40, (3) 88-119, (4) 20, (5) 27, (6) 44-51 u. Legs long, normal. Chitin of caudal protuberance appears as though bossed w.th a honey-comb pattern. Habitat: On stems of Rhus sp. (probably R. vimanalis), Victoria West, C.P.; collected by Mr. van Heerden, October 1915. Collection No.: 346. 236. Ceroplastes rusci (Linn.) Coccus ruscir, Linn., Syst. Nat. Ed. x, 1, p. 456, 1758. Coccus caricae, Bern., Mem. Acad. Marseille, p. 89, 1773. Coccus artemisiae, Rossi, Mant. Ins. 11, pp. 56, 514, 1794. Calypticus radvatus, Costa, Faun. Reg. Nap. Coce., p. 12, 1835. Calypticus testudineus, Costa, Faun. Reg. Nap. Cocc., p. 12, 1835. Calypticus hydatis, Costa, Faun. Reg. Nap. Cocc., p. 14, 1835. Columnea testudinata, Targ., Atti dei Georgofili, n.s. xi, p. 31, 1866. Coccus hydatis, Targ., Studi sul. Coce., p. 12, 1857. Columnea testudinforms, Targ., Studii sul. Cocc., pp. 8, 11, 12, 1867. Chermes caricae, Bdv., Ent. Hort., p. 320, 1867. Ceroplastes rusci, Sign., Ann. Soc. Ent. Fr. (5) u, p. 35, 1872. Lecamum artemisiae, Sign., Ann. Soc. Ent. Fr. (5) i, p. 37, 1872. Adult 9. Test about 7-5 mm. long, 5:2 mm. broad and 5 mm. high, regularly domed, rounded in front, rather excavate behind. Colour greasy white, suffused with purplish red. Dorsal dome and margins of plaques lined with greenish grey lines ; anal pore and middles of plaques of same colour. Stigmatic bands, 2 on each side, pure white, broad at stem and tapering inwards. Middle of dorsum with a depression in which there is an elongate, glossy white prominence similar to those in centres of plaques, but longer. Antennae 6-jomted; range in uw: (1) 34-40, (2) 37-40, (3) 114-120, (4) 17-24, (5) 20-27, (6) 44-57. Leg II: coxa 102; femur + trochanter 170; tibia 120; tarsus 85; claw 20wu. Margin of sides of body with a single row of small thimble-like spines ; amongst these are mixed a few simple spines. Stigmatic clefts very shallow, not chitinised, THE COCCIDAE CF SOUTH AFRICA. 35 with a few additional spines like those of the marginal series and a small group of simple glands like the circumgenital glands of the Diaspinag. Anal plate heavily chitinised, with scattered perforations. Halitat: On stems of quince, Stellenbosch, C.P.; collected by F. W. Pettey, July 1916. Collection No.: 91. 237. Ceroplastes tachardiaformis, sp. n. Adult tests aggregated in huge masses on stems of the host-plant, often com- pletely covering the stems for a distance of several inches. © test Tachardia-like, globular, slightly flattened above, hard, thin, brittle, almost transparent resinous brown, due to the colour of the insect within ;_ without protuberances but with a slight apical depression containing the opaque white larval exuvia. Adult 2, denuded of wax, moderately dense, globular, smooth and shiny. Caudal protuberance short, very dense, surrounded by a coarsely perforated plate. Antennae 6-jointed, e.g. (1) 34, (2) 27, (3) 74, (4) 20, (5) 17, (6) 30u. Legs short, e.g., coxa 50, femur ++ trochanter 110, tibia 78, tarsus 50. Integument moderately dense, appearing, under the high power, finely rugose, with numerous, widely scattered, small, transparent pores. Stigmatic clefts almost obsolete, indicated by a small group (8 or 10) of short obconical spines and a few simple glands. Habitat: On rhenosterbosch (Elytropappus rhinocerotis, Less.) ; collected by Messrs. Watermeyer Bros., Aberdeen, C.P., November 1915. Collection No.: 94. 238. Ceroplastes zonatus, Newst. Ceroplastes zonatus, Newstead, Bull. Ent. Res. vii, p. 32, 1917. “Female test. Broadly ovate in outline, highly convex; marginal plates very faintly indicated, but apparently without nuclear spots; dorsal plate very large, with a central nuclear spot of white wax ; cephalic margin slightly clypeate ; lateral margins in very old examples with a pronounced foot-like extension from each of the stigmata, from which there extends a thick white waxen appendage. In the younger forms the foot-like extension is wanting, but the white waxen appen- dages are present and always porrected. Colour creamy white, suffused with very pale brown; dorsal plate surrounded by a shaded wavy zone of dark brown and brownish black, with here and there a suffused patch of dull flesh-colour. In very old examples the zone of colour extends to the margins and is of a shining madder- brown to piceous colour. On the removal of the outer surface of the test with chloroform, it is seen to be divided into seven areas by pale orange-coloured lines ; a central polygonal area, corresponding to the area occupied by the dorsal plate, from the angles of which radiate to the margin single lines marking off the areas of the lateral and cephalic plates. “Female adult (denuded of the test). Ovate; cephalic margin clypeate ; dorsum low and wrinkled; two large, bilateral, submarginal extensions, both longitudinally striated : the space between these extensions of the body-wall and (681) eo 36 CHAS. K. BRAIN, the margin markedly constricted. Dorsum with a large keel-like process. Caudal process very short and conical. Stigmatic clefts deep. Antennae of eight seg- ments ; the articulations relatively very broad ; 3rd about equal in length to the 7th and 8th together ; a very long hair on the 2nd and 5th, and a slightly shorter one on the 8th ; there are two spines on the 8th and one on the 7th. Legs normal. Stigmatic spines covering a large and somewhat pyriform area, the length of which is nearly equal to twice the length of the antennae ; the spines, with the exception of a small group near the stigmata, are obconical and the space between them with bands of dark granular bodies, which collectively form a polygonal reticulation ; the small proximal group of spmes are longer than the others and pointed. No trace of marginal spines. Derm thin and transparent after maceration ; rather thickly set with minute pores and minute scattered spines. Caudal process sur- rounded by a porose zone of brown chitin. Length 3°9-4°6 mm.” (Newstead). The above description does not refer to the mature female forms which attain, with the waxy test, 14 mm. long, 12 mm. broad and 8 mm. high. The colours remain the same except that the lighter parts become yellowish. When mature the denuded female is densely chitinous, brownish black, with the dorsum smooth, shining. In boiling KOH it stains the liquid deep purplish brown. The antennae are sometimes 7-jomted with joint 4 very long, obviously 4 + 5 of the 8-jointed form. Halitat: On Acacia sp., Pretoria, November 1914. Collection No.: 344. Genus Inglisia, Mask. Adult female more or less conical, covered above by a glassy shield which is divided into plates and striated with rows of air cells. Legs and antennae well developed. 239. Inglisia elytropappi, sp. n. Test of adult 9 small, 1°6-2 mm. long, 1°2 mm. wide and high, like a small bivalve shell with its hinge uppermost standing on the stem. It is very like a small zizypha test but paler in colour, with the upper angles of the two halves smooth instead of tuberculate. The colour is white to pale buff, shiny, often pearly, with the vertical striae conspicuous. 2, with the test removed, shiny dark brown, of the same general shape as the test, with the apex depressed between two rounded lateral humps. Cleared and mounted the body is hyaline. The margin has a close-set row of short conical spines with broad bases and moderately sharp points. The stigmatic clefts are obsolete, but their presence is indicated in the marginal row of spines by the addition of a single, slightly longer spine with a narrower base. The antennae are rudimentary, appearing in varied forms with very indistinct segmentation ranging from 4 to 7-jointed. When 7-jointed the segments are generally very short—mere rings. The measurement of such an antenna gave the following in wu :—(1) 17, (2) 6, (3) 17, (4) 6, (5) 10, (6) 6, (7) 6. THE COCCIDAE OF SOUTH AFRICA. 37 Legs rudimentary, appearing as though composed of three, almost equal, cylindri- cal segments with a minute claw. The total length varies between 70 and 90. Habitat: On the thinnest twigs of rhenosterbosch (Hlytropappus rhinocerotis, Less.), Groot Drakenstein, Somerset West and Cape Flats, C.P. (Cape Coll. No. 1244). The stems of the host-plant are thickly covered with “ sooty”’ fungus, apparently grown on secretion from the numerous specimens present. Collection No.: 100. 240. Inglisia geranii, sp. n. (Plate iv, fig. 249). Insects congregated on the main stems at or near the nodes. Adult insect, with test, about 2°5 mm. long and 1°3 mm. broad at the base, brown, with the air spaces of the test almost colourless and appearing as transverse lines radiating from the centre of each half of the test. The test is composed of two similar halves, each of which simulates a shell or tortoise-shell, with their apices widely separated. The median line at the point of union of the two is almost flat, very little depressed. The test easily flakes away from old specimens and is then white, almost hyaline. The antennae are 7 or 8-jointed ; range in « :—7-jointed: (1) 24-31, (2) 14-20, (3) 41-48, (4) 17-20, (5) 14-20, (6) 14-17, (7) 20-24; 8-jointed: (1) 24, (2) 17, (3) 31-34, (4) 14, (5) 14-17, (6) 14-17, (7) 17, (8) 24. Leg I: coxa 68; femur + trochanter 136; tibia 109; tarsus 85u. The tro- chanter has a long (90m) spine. The upper digitules are long and slender, hardly perceptibly clubbed ; lower digitules comparatively short and slender. Anal plate about 115 long, with 2 or 3 stout spines. Margin with a single row of sharp pointed spines of varying lengths ; these are all broad at the base and taper rapidly to the point. Stigmatic cleft with a single spine a little longer than the longest of the marginal series and a little less tapering. Habitat: On geranium, King Williamstown, C.P.; collected by A. Kelly, March 1916. Fresh material sent by Mr. J. Hobson, King Williamstown, May 1916. Collection No.: 99. This species is remarkably close to Inglisia theobromae, Newst., which _ was described on cacao from Uganda, but it is smaller, and joint 3 of the antennae is apparently always longer than 4. It is also somewhat like J. bivalvata, Green, but the dorsal shield, formed by the inner sides of the two halves of the test, is wider and more oval. 241. Inglisia zizyphi, sp. n. Test of adult 92 shaped like a small bivalve shell, standing erect, with the two halves separated by a conspicuous furrow, and the hinge uppermost, represented by two small, rugose “crowns” of the two halves. The lower margins, which rest on the stem, are surrounded by a slight fringe of fine glassy filaments. The test is pale horn-colour, with the distinct vertical striae appearing slightly iridescent. Length 2 to 3 mm.; width 15 to 2 mm.; height about 2 mm. Adult 9, with the waxy test flaked off, red-brown, of the same shape as the test but with the dorsum shiny and wrinkled. 38 CHAS. K. BRAIN. Cleared and mounted the integument is thin and hyaline. Margin with a dense series of stout conical spines, so closely set that the row appears, in places, double. The derm is clear except for a double row of large simple glands extending along the median line from the anal plates to the middle of the dorsum. Antennae 5 or 6-jointed, the individual segments badly defined even in stained preparations ; range in «:—0d-jointed : (1) 24, (2) 14, (3) 41-51, (4) 24-34, (5) 24-27; 6-jointed : (1) 20-24, (2) 10-14, (3) 48-51, (4) 10-17, (5) 14, (6) 20-24. Legs short but normal, e.g., coxa 41; femur + trochanter 102; tibia 75; tarsus 58; claw 17u. Anal plates about 90u long, each with two fairly stout spines about 40u long. Stigmatic clefts obsolete. Larva about 390u long; antennae 6-jointed ; caudal setae very long (170,). Habitat : On Zizyphus sp., Pretoria ; collected by the writer, December 1914. Collection No.: 101. This species is very close to I. conchiformis, Newstead, but is smaller and has an antennae of 5 or 6 segments instead of 7. I thought at first that I was dealing with young individuals of Newstead’s species but mounted specimens contain well developed embryonic larvae. . Genus Cryptinglisia, Ckll. “* A Lecaniine Coccid having a glassy covering containing air-spaces, and retaining the legs and antennae (7 or 8 joints) in the adult. Living in galls on the roots of Vitis. Differs from Inglisia in its mode of life; in the glassy scale not being divided, tortoise-like, into plates ; and in the air-cells running together, forming long air- spaces. Larva ordinary, with six large bristles on the cephalic margin. Male unknown.” (CkIll.) 242. Cryptinglisia lounsburyi, Ckll. Cryptinglisia lounsburyt, Ckll., The Entom. xxxiil, p. 173, 1900. , Cryptinglisia lounsburyi, Lounsb., Rep. Ent. Cape Good Hope, p. 54, 1900. “ Adult female about 24 mm. long, soft, shiny, very dark brown, covered with a semi-transparent, brittle, glassy scale. Skin transparent and colourless on boiling in KOH; mouth-parts moderate, rostral loop not very long ; margin with a row of simple spines, brownish, about 24, long, placed close together ; anal lobes ordinary, about 160u long, yellowish brown, surrounded basally by a large, thick, dark brown, chitinous plate, more or less semilunar in form, with the ends produced ; a row of small round glands in the middle line from one end of the body to the other, but best developed posteriorly ; antennae and legs pale ; legs ordinary, femur plus trochanter about 120, tibia about 96, tarsus about 78, claw about 20u ; claw-digitules about as long as claw, with large knobs ; tarsal digitules long, with distinct knobs. Antennae 7 or 8-jointed, having three types, thus: (1) 7-jointed with a short 3, all the joints subequal, 21 to 30u; (2) 7-jointed with a long 3, which is about 41 long; (3) 8- jointed, with 2 quite short, and 3 and 4 each about 30u long. The terminal joint is always short, 21 to 26. THE COCCIDAE OF SOUTH AFRICA. 39 “These insects occur underground on the roots of grape vines, living in galls which are more or less globular, 4 to 5 mm. diameter, dark, rough and often nodulose on the outside, often aggregated together in numbers, or even coalescing, so that the root presents a nodulose thickening 6 or 7 mm. in diameter and over 20 mm. long. On breaking open the galls, which are quite hard, one finds a cavity containing the Coccid. Small stones are frequently embedded in the sides of the galls. “ Hab. Constantia, Cape Colony, at the roots of Stein and Reisling grapes (Vitis vinifera). Mr. Chas. P. Lounsbury, sending the specimens, says; ‘ None were observed more than eight or nine inches from the surface, and all were on fibrous roots. As you will observe from the specimens, they are somewhat gregarious ; ofttimes. one or two rootlets will be quite covered, while all the others are free. Most of the infested vines were backward in growth—some almost dead ;_ but their condition, I think, is due to other causes than the attack of the insect. Some apparently healthy vines were noticed to be affected.” (CkIl.) This species has not been re-discovered. Collection No.: 105. Genus Parafairmairea, Ckll. Female scale divided by a longitudinal, median suture into two halves, each with minute grooves radiating from its apex but not striated with air cells. Legs and antennae well developed, the latter 7 or 8-jointed. 243. Parafairmairea patellaeformis, sp.n. (Plate iv, fig. 250). Adult ° dull brown, 9 mm. long and 5 mm. broad, covered above by a stout shield which is divided longitudinally into two halves. This shield has the wavy lnes and exact appearance of one of the common shells of the more elevated “ Patella ” type. With the covering removed the insect is glossy brown with wrinkles and ridges radiating from a two-fold dorsal peak to the margin. Antennae 7 or 8-jointed ; range in u :— i, II. lg bal ts va Ve. Oe eee 20-31 20-24 41-68 34-41 20-27 14-24 20-34 - 20-31 20-24 44 #17-34 27-34 17-20 1420 24-34 Leg 1: coxa 88-102 ; femur + trochanter 187-238 ; tibia 160; tarsus 90; claw approximately 20. Anal plate about 1604 long. The integument is thin and hyaline, without con- spicuous hairs or glands. The margin has a single row of short, conical spines of the Inglisia ty pe, amongst which are intermingled, at intervals, smaller spines of a thinner type. The stigmatic clefts are obsolete, merely indicated by a scant series of small circular, simple glands extending inwards from the margin towards the spiracle. Remarks : This is the third Parafairmairea to be described. The other two were described on grass, one from France and the other from Surrey, England. Habitat: On stems of Acacia karroo; collected by A. E. Kelly at Port Alired, C. P., March 1915. Collection No.: 98. 40) CHAS. K. BRAIN. Genus Geroplastodes, Ckll. Female scale convex, but not cone-shaped ; not divided into two halves, nor into distinct plates, but rough or beset with protuberances. © with antennae 7- or 8-jointed. 244. Ceroplastodes bituberculatus, sp.n. (Plate iv, fig. 251). Test of adult 9 about 4°5 mm. long and 3 mm. broad and high, white, with two prominent humps, one at each end of the median ridge; margin with a distinct white fringe. The test indicates the anal cleft and has a small prominence over the anal lobes. . Male scale dull white, with the median area denser, about 1°5 mm. long, divided into plates in the normal manner. Adult 2 with the waxy covering removed, dark brown, of similar shape to the test. Integument, when mounted, clear, hyaline, without glands or hairs. Margin with a single, close-set row of tubular spines somewhat of the Inglisia type but more linear and truncate. Stigmatic clefts obsolete but indicated in the marginal series of spines by the addition of three stigmatic spines ; laterals slender, a lictle longer than the marginal spines ; median stout, two and a quarter times as long as laterals (1161). Legs and antennae long, well developed. Antennae 8-jointed; range in yw: (1) 27-34, (2) 20-34, (3) 63-68, (4) 31-34, (5) 37-48, (6) 17-20, (7) 17-20, (8) 34-41 (from 5 antennae). Ley I: coxa 78; femur + trochanter 176; tibia 120; tarsus 90; claw 24y. Tarsal digitules very long, slender, clubbed ; claw-digitules comparatively slender. Anal plates about 145 long, apex with 4 very stout blunt spurs. Habitat: On stems of native shrub, Somerset West, C.P.; collected by T. F. Dreyer, November 1906. Also at Stellenbosch (Fuller). Collection No. : 327. : Genus Idiosaissetia, nov. Adult 9 secreting a thin brittle covering of waxy material not divided into plates or two halves, and without air cells. Legs and antennae present but rudimentary. Anal cleft not median but to one side. Type, L. peringueyt, sp. n. 245. Idiosaissetia peringueyi, sp. n. Adult 9 about 2°6 mm. long and 1°3 mm. broad, elongate, very convex, with almost perpendicular sides and sloping abruptly in front and behind. Female raised on a hollow waxy sheath which extends over the sides of the body, leaving the dorsum naked. The material is now ten years old and appears as though the waxy covering had once extended over the dorsum, but had become detached from it. If this were covered the insect would look like an Inglisia, except that the waxy covering is thin, pale buff-coloured and solid, 1.e., without air-tubes. Female, denuded of wax, densely chitinised, brown, shiny. When cleared and mounted the integument of the dorsum is moderately dense, with numerous + circular clear spaces as in Savssetia bitt of greatly varying size. The venter is thin and clear. The antennae are rudimentary, usually with about THE COCCIDAE OF SOUTH AFRICA. 41 three indistinct joints, always exhibiting pseudarticulations. In the longest antenna seen, after staining, one could observe 6 ill-defined joints measuring (1) 17, (2) 14, (3) 14, (4) 14, (5) 14,(6) 20n. Leg I: coxa 34; femur + trochanter 85; tibia 85; tarsus 27; claw 17w. The margin has a scant series of short, curved, moderately stout spines, except near the anal cleft, where the spines are closer set, long (41), straight and very acute. The anal cleft is invariably thrown to one side and is not in the median line as in Lecanium. The anal ring has numerous (12?) hairs; anal plates about 120ulong. Embryo-larva large, about 420, long, with 6-jointed antennae. Caudal extremity produced, with prominent caudal lobes, each with one long, stout seta (136) and several shorter spines. Habitat: On grass or thin reed. Label in tube reads: “ From Dr. Peringuey, S.A. Museum, Nov. 1908.” Collection No.: 140. Genus Membranaria, nov. Lecaniid, somewhat like Pulvinaria but with the cottony ovisac replaced by a membranous receptacle. Antennae and legs well developed, the former 7 or 8-jointed. Type, M. pretoriae, sp. n. 246. Membranaria pretoriae, sp. n. (Plate iv, fig. 247). Adult 2 with ovi-receptacle about 5°5 mm. long, 2 mm. broad and 3 mm. high. Receptacle membranous, secreted from the margins, particularly the posterior margin, of the insect, so that, when completed, the insect, except the head end, is raised from the stem. The membrane consists of two coats, both thin, but which may be readily separated. The outer coat is pale, honey-comb yellow, not striated, and is produced from the upper margin of the insect. The inner coat is paler in colour and has longitudinal striae. The completed receptacle is hoof- shaped (fig. 247) with a distinct, dense, longitudinal, median keel. Its greatest length, along this keel, is about 3 mm. The adult ° is deep caramel-brown, sometimes speckled with black, glossy, with a rounded, median ridge and marginal corrugations and depressions. Cleared and mounted the adult 2 is moderately dense. Margin with a single row of simple, hair-like spines set wide apart. Submarginal area, except of anterior end, with a broad band of very numerous, small, gland pores, each with a curved, linear gland tube. Stigmatic cleft obsolete, but indicated by a pair of curious, short, broad, cup-like; protuberances. Anal ring with 8 hairs; anal plates surrounded in front by a dense, rugose, chitinous plate. Antennae 7 or 8-jointed ; range in mu :— : I III IV V VI VII Vill 20-31 31-34 17-41 20-48 17-24 17-24 =. 24-27 — 24-27 «27-37 = «24-3400 sd14-24 020-27 17-20 =16-20 24-27 Leg I: coxa 102, femur + trochanter 238, tibia 153, tarsus 51, claw 20. Halntat: On crowns of grass, in front of Union Buildings, Pretoria ; collected by the writer, October 1914. Collection No.: 83. a9 % . “ue Yj nig 4 m : : 4 Bis 3. 2 | : et ee iatag #9 a ihe Sg [ee a ‘ 1 iat Ot, ike . ! Twa eo ant . i me shad 7 o+% at a = i » eese ; 7 mn . - i % » ‘ roe t a ~~ @¢ é ie) . " 7 7 ‘ ’ . J . . e a Ls » - > 4 a See ies hae: 6), - —i - - ?. - = ‘ j M ‘ wv a is ¥ ao i ¥ 4 - , 4 fe ieeewy? Ss . _ ‘ A > * x? 2 a > 4 A = pire ” a ne * yi o¢ 1" hong =. ’ ; ae? oe el 1 7 - - . De: oS 7 a 7 : EXPLANATION OF PLATE I. Fig. 233. Lecanium elongatum, Sign., adult 929. 234. Saissetia persimilis, Newst., adult 99. 235. Pulvinaria mesembryanthemi (Vall.) Sign., adult 99 with ovisacs. 236. Filippia carissae, sp. n., adult 2 with ovisacs com- pleted and one ¢ puparium. 237. Sarssetia subpatelliforme, Newst., adult 29. 238. Hemilecanium theobromae, Newst., adult 99. PLATE I, PART 1. Yoru. AL BuLL. ENT. RESEARCH. SOUTH AFRICAN COCCID&~. EXPLANATION OF PLATE II. Fig. 239. Ceroplastes eucleae, sp. n., adult 9 9. 240. Ceroplastes sp. 241. Ceroplastes mimosae, Sign., adult 2 9. 242. Pulvinarta aristolochiae, Newst., g tests and © ovisacs with larvae. 243. Sarssetia sp., to illustrate dermal cells. 244. Protopuluinaria piriformis (CkIL.) Lefroy, anal lobes of adult 9. BULL. ENT.- RESEARCH. Vo... XI. Part .1. Pcvate II, 243. SOUTH AFRICAN COCCID&. a 44% : re ee —_ , ; ’ a ’ ~ ~~ 7 @ { 7 . s J he 2 a. t 7 Fy mie . rom o- 8 2 ar _ oo an hy aa ial =) - ‘oe - ; - 7 . > Ppa — ae s A aes ‘ b we a} Uy r 7 for j ' Not F 4 Vs ! \ ee , air iy yae - Ld aa 2 Sh EXPLANATION OF PLATE III. Fig. 245. Conofilippra subterranea, sp.n., Q ovisacs on roots ; a, 2 with top of ovisac tilted to one side ; b, section of ovisac with 9 ; c, swollen tibio-tarsa joint; d, antenna; e, stigmatic spines and glands ; f, anal lobes of 9. ET}. PLATE PART 1. VOL, (20n. Butt. ENT. RESEARCH. SOUTH AFRICAN COCCID/4 EXPLANATION OF PLATE IV. . Allopulvinarva subterranea, sp.n., adult 29. . Membranaria pretoriae, sp. n., adult 2 9. . Protopulvinaria pirifornis (CkIl.) Lefroy, adult 22. . Inglsia gerann, sp.n., young and adult 99. . Parafairmairea patellaeforms, sp.n., adult 99. . Ceroplastodes bituberculatus, sp. n., adult 99 and dS puparia. BuLt, ENT. RESEARCH. VOL. XI. PART 1. PLATE IY. SOUTH AFRICAN COCCID&. ae At Me ‘ ‘ ; ‘ ay, ; . suit, Om 6 : ¥ é * i \ ‘ whe f; ** J ot 4 ee SOS RL ea ee tee, P A ‘ é Py 1 P ] ao ; = a" b ‘ ae ‘ + .» * “¥ 43 THE PERCY SLADEN TRUST EXPEDITION TO THE INDIAN OCEAN IN 1905, AND 1907-1909, UNDER Mr. J. STANLEY GARDINER, M.A. DIPTERA: TABANIDAE. By Major E. E. Austen, D.S.O. The material belonging to this family brought back by the expedition includes representatives of only four species, one of which, however, is apparently new. That three out of the four species obtained should belong to the Aegophagamyra- Bouverella group is not surprising, in view of the predominance of this group in the Malagasy Subregion. Through the courtesy of the authorities of the Percy Sladen Trust, the type of the new species described below has been placed in the British Museum (Natural History). PANGONIINAE. Genus Aegophagamyia, Austen. Ann. Mag. Nat. Hist. (8) 1x, p. 12 (1912). Aegophagamyia terticeps, Austen. P. ct., p. 15. Three 9 9, from Astove I., 1907 ; four 99 from Aldabra, 1908-9 (J. C. F. Fryer). The type and para-type of this species, in the British Museum (Natural History), are from Astove I. (P. R. Dupont). In the female sex, at any rate, this species bears a superficial resemblance to Tabanus albipectus, Big., for which on casual inspection it might easily be mistaken. The four specimens noted above as having been taken in Aldabra bear the same number as that attached to the examples of Tabanus albipectus collected in the same island, by Mr. J. C. F. Fryer, showing that the field note printed below under the heading 7. albipectus must be taken as applying to the present species also. Genus Bouvierella, Surcouf. Bull. Mus. Hist. Nat., Paris, 1909, p. 176. Bouvierella alluaudi, Giglio-Tos. Scione allawudi, Gigho-Tos, Ann. Soc. Ent. France, Ixiv, p. 357 (1895). Of this species, originally described from specimens from the Seychelles Is., the present collection includes no fewer than twenty-two examples, as follows :— One 3, one &, from Mahé, Seychelles, v-xu1. 1905 (J. S. Gardiner); eleven $6, five 99, from Mahé, Seychelles (5 gg, 2 29, from the Forét Noire district, x-xi. 1908, and 6 93,3 929, from Cascade Estate, at 800 to 1,000 feet or more above sea- level, 1908-9); one 9, from Félicité, Seychelles, xi. 1908; 2 gg, from Praslin, Seychelles, x1. 1908; and one 9 from Cosmoledo Is., 1907 (H. P. T.). (681) D2 44 FE. E. AUSTEN, Bouvierella inornata, sp. n. 2.—Length (2 specimens) 10 to 11°2 mm. ; width of head 3°25 to 3°6 mm. ; width of front at vertex 0°4 mm., at lower extremity 0°25 mm. ; length of wing 9 to 10 mm. Sombrely coloured species, entirely devoid of markings ; ground-colour of dorsum of thorax drab* or dusky drab, that of dorsum of abdomen greyish fuscous ; front in © narrow, slightly narrower at lower extremity than at vertex, about six tumes as long as its width at upper end ; wings, especially portion beyond basal and anal cells, strongly suffused with brownish, stigma well marked ; coxae drab, legs otherwise blackish brown, extreme tips of femora, at least of those of front parr, ochraceous buff. Head: front, subcallus, face, sides of face and jowls drab, occiput smoke-grey ; vertex clothed with minute, erect, blackish-brown hair, sides of face with ochraceous- brownish hair, basi-occipital region with longer fine brownish hair ; frontal callus represented by a fine, dark, median line, commencing at some distance above level of lower inner angle of eyes, and not reaching anterior ocellus ; proboscis, including labella, narrow and elongate, total length of proboscis about two-thirds of that of head ; proximal joint of palpr: dark brown or brownish, terminal joint similarly coloured on outer surface, narrow, curved, elongate and very slender; antennae dark brown, first and second joints clothed above and below with short, blackish hair, expanded portion of third joint ovate or elongate ovate when viewed in profile, without any angle on upper margin, annulate portion of third joint narrow, curved and elongate. Thorax: dorsum clothed with short, recumbent brownish hair, which in front of prescutellar groove may be mixed with paler hair, scutellum bearing jonger brownish hair, or a mixture of brownish and ochraceous hair; pleurae and pectus agreeing in coloration with dorsum, and clothed with fine, brownish hair. Abdomen : dorsum clothed with short, appressed dark brown hair, venter agreeing with dorsum in coloration and hairy covering. Wings: lengths (relative as well as actual) of stalks of first and fourth posterior cells variable in different specimens ; stigma elongate, dark brown. Squamae light drab or drab-grey, borders drab. Halteres: knobs clove-brown or blackish brown, stalks paler (cinnamon-buff). Legs sparsely clothed with short or minute dark brown or blackish hair. Seychelles Is. Type from Silhouette (plateau of Mare aux Cochons, about 1,000 feet, x. 1908) ; a paratype from Mahé (Cascade Estate, 800-1,000 feet), 1908-9. Bouvierella wmornata is distinguishable from all previously described species belonging to this genus, inter alia, by its uniformly sombre coloration, the ground- colour of the thorax and abdomen not being relieved by either lighter or darker markings. In particular, it is distinguished from B. brunnea, Surcouf (Bull. Mus. Hist. Nat., 1909, no. 4, p. 179), a species found in Madagascar, by its smaller size, and by the more uniformly dusky coloration of the body and legs, the dorsum of the thorax not being longitudinally striped with grey, and the legs not being reddish yellow, clothed with golden pile. In the length and slenderness of the proboscis, including the labella, the species just described agrees rather with Aegophagamyia, Austen, with which it constitutes in this respect an annectent link; as regards antennal characters and the shape of the first posterior cell, however, the species seems more correctly placed under Surcouf’s genus. * For names and illustrations of colours, see Ridgway, ‘‘ Color Standards and Color Nomenclature’ (Washington, D. C.: published by the Author, 1912). DIPTERA: TABANIDAE. 45 TABANINAE. Genus Tabanus, Linn. Fauna Suecica, Ed. 1, p. 462 (1761). Tabanus albipectus, Bigot. T. albipectus, Bigot, Ann. Soc. Ent. France, (3) vii, p. 125, pl. ix, fig. 2 (1859). ‘Twenty-one specimens, from various localities as follows :—One 9, from Poivre Is., Amirante Is., v-xii. 1905 (J. S. Gardiner) ; one 9, from I. des Roches, Amirante Is., 15. x. 1905; one 9, from Aldabra, 1907 (H. P. T.); 15 929, from Aldabra, 1908-9 (J. C. F. Fryer); one g, two 99, from Long Island and Mahé, Seychelles (Cascade, about 1,000 feet), 1908-9. A comparison of the foregoing material with the type of the species, taken in Madagascar by Dr. Coquerel and now in the British Museum, places the correctness of the identification beyond doubt. In addition to the type of T. albipectus, the National Collection includes a 3 and @ from Astove I. and Aldabra respectively (P. R. Dupont), and a 2 from Pigaduli, Zanzibar, “on goat,” 14.x.1912 (Dr. W. M. Aders). To the specimens from Aldabra in the present series, taken by Mr. J. C. F. Fryer, the following interesting field-note was attached :—“ Most. numerous in December [1908]: attack sea-turtles, biting them between the plates on the back: apparently attack them in water as well as on land. Were seen flying all over the lagoon, and also some way out at sea. They fly strongly, and bite man also.” As already explained, this note must be taken as applying equally to Aegophagamyia terticeps, Austen (vide supra). With reference to the biting of reptiles by blood-sucking Diptera, it may be remembered that the typical series of Tabanus crocodilinus, Austen, was taken on a crocodile in Nyasaland (cf. Austen, Bull. Ent. Res., 1, p. 285 (1912)), while in Uganda, as is well known, Glossina palpalis feeds readily upon crocodiles and monitors. 47 ON THE REASONS FOR THE VARIATION IN THE EFFECTS OF FORMALDEHYDE AS A POISON FOR HOUSE-FLIES, By Capt. Lu. Luoyp, D.Se., R.A.M.C. (T. F.). The following is an account of an investigation carried out on behalf of the Hygiene Department of the Royal Army Medical College at the suggestion of Prof. Maxwell Lefroy, who, after his visit to the Mesopotamia War Area in 1916, was insistent on the importance of the discovery of some poison for house-flies (Musca domestica and allied species) that could be used indoors with safety. The work was carried out in the Entomological Department of the Imperial College of Science in the early part of 1917. Large stocks of flies which were kept breeding in two rooms were available for the purpose. Formaldehyde has been frequently recommended as a poison for flies, but those | who tested its effects gave varying accounts of it. Miss Lodge (1), who studied it in some detail, was unable to account for these variations and stated that no definite conclusions could be drawn from her work. An analysis of the various factors which might cause such variations was made, and it was found that there were three which play an important part. These are (1) impurities in the formal- dehyde, (2) the humidity of the atmosphere, and (3) the strength of the solution used. The following subjects will also be discussed, (4) the advantage of adding a bait, (5) the most effective mode of administration, and (6) comparison with other poisons. “Formaldehyde”? must be understood throughout to mean the 40 per cent. solution in which it is usually sold, and all strengths are given as percentages of this. “ Formalin,” the trade name of one brand, is often loosely used for this solution. Impurities in the Formaldehyde. | Formaldehyde is prepared by passing a mixture of heated air and the vapour of methyl alcohol over some heated contact agent, such as copper gauze or silver. The methyl! alcohol is crude or pure, but should be 90 per cent., and should not contain more than | per cent. of acetone. The purity of the product depends largely on that of the methyl alcohol, and on the freedom of the air from ammonia. The following substances are liable to be present: trioxymethylene, methylal, methyl and ethyl alcohols, acetone, hexamethylene tetramine, methylamine, and formic acid. Prof. Baker, of the Imperial College of Science, provided samples of formaldehyde to which certain of these substances had been added in proportions in which they are liable to occur. Solutions of these, generally in water and soaked up in clean sand, were exposed side by side with solutions of the control formal- dehyde in the fly room, and left for twenty-four hours. The numbers of dead flies in the immediate neighbourhood of each were then counted. Many flies left the table to die and no account could be taken of them. There was always a definite group dead in a circle of six inches radius around each jar, and the comparisons were based on these. In order to obviate the influence of the varying numbers of flies on different days, all numbers are given reduced to percentages of the controls. All decisions are based on a number of tests, with the positions of the various solu- tions and the controls interchanged. 48 CAPT, LL. LLOYD. Of the impurities tested, hexamethylene tetramine, methylamine and formic acid were found to be of importance, and these only will be discussed. The effects of the presence of the other substances mentioned above were investigated and found to be neutral, or of very little importance. Hexamethylene tetramine, (CH,),N,. This substance is produced by the action of ammonia on formaldehyde. It is liable to occur in the solution owing to the presence of ammonia in the air during manufacture. It breaks down with the formation of methylamine, CH, NH,. Formaldehyde containing these bodies has a yellowish colour and a fishy odour, both these characteristics being intensified by the addition of a strong alkali. These substances are distasteful to flies. A sample containing 5 grms. of hexamethylene tetramine added per litre was tested against a control. The residue from the manu- facture of this substance, containing ammonia and methylamine, and some methyla- mine hydrochloride, with and without an alkali to set free the methylamine, were also tested with the following results. TaBLeE I. Showing the Influence of Hexamethylene tetramine in Formaldehyde. Soluti | Percentages of flies | Average | Total flies olution | killed | of | killed Formaldehyde control, 10% 100 in each case 100 324 Formaldehyde with 5 grms. hexamethylene tetramine added per litre, 10% T45/430; 932071. obi 85 267 Tas_e II. Showing the Influence of Methylamine in Formaldehyde. Boise Percentages of flies killed (actual numbers in brackets) | Formaldehyde control, 109; 100 (50) 100 (20) es with 2°5% hexamethylene tetramine residue 10% 36 (18) #} with 1% hexamethylene | tetramine residue 10% 65 (13) os with 1% methylamine | hydrochloride 10% | 40 (8) re with 5% methylamine hydro- chloride and 0:2% KOH 10% 30 (15) 3 with 1% methylamine hydro- chloride and 019% KOH i ae ae ae a pou 30 (6) A sample, naturally containing these substances in unknown proportion, with a pale yellow colour and fishy odour (previously unsealed bottle from the store of the Imp. Coll. Sci.) was tested against a colourless control with a clean odour. FORMALDEHYDE AS A POISON FOR HOUSE-FLIES., 49 The results of the tests are summarised in the next table. In nearly every case the purer sample proved the more effective. No suggestion can be given for the purification of formaldehyde of this nature, and it should be avoided if any other is obtainable. TABLE IIT. Comparing a Yellow Formaldehyde with Fishy Odour (Methylamine) with a Clean Sample. Soluti Percentages of flies Average Total flies kcal | killed % killed Formaldehydecontrol, 10% | 100in each case 100 | 1509 a impure, ae | ted. Gi, 96, -89, 230, 36; | 17, 68, 58. 89 1118 | Formic acid, H’CO’'OH. Formaldehyde is prepared by the dehydrogenation of methyl alcohol: CH,OH + O = CH,O + H,0. Theoretically in the process there is no formic acid formed. The samples of formaldehyde examined, however, were all found to be acid. Three which were titrated were found to be +24, +39, and +60 acid (Eyre) respectively. (A fluid ““ +a acid ” requires x ccs. of normal alkali to make one litre neutral.) The acid present was proved to be formic by a test suggested by Dr. A. M. Whiteley. A few ccs. of the formaldehyde are shaken up with lead carbonate and quickly filtered. If formic acid is present crystals of the sparingly soluble lead formate separate out in the filtrate after a short interval. The crystals are minute, very characteristic needles. On the assumption that formic was the only acid present, these samples contained 1°1, 1°8, and 2°76 grms. of formic acid per litre respec- tively. When the dilute poison 5-10 per cent. formaldehyde is made up the acid would be present in amounts varying from ‘0055 to ‘0276 per cent. Hven these small quantities are detected and disliked by flies. TaBLe IV. Showing the Deterrent Effect of Formic Acid in Water. euriian No. of Av. flies | Av. flies Av. percentage tests | tasting | drinking which drank Tap water control 2 35 24 69 Formic acid ya'8 1 33 | 3 10 i 05% 3 32 | 15 47 ag 02 :049 4 36 20 54 in 01 005% 3 36 | 18 50 Temperature 77°9° F. ; relative humidity 57%. Testing saucers, two inches in diameter, were filled with solutions of formic acid in water and covered with discs of blotting paper to soak up the fluid. These were then exposed one at a time in the fly room and watched for five minutes, a count 50 CAPT. LL. LLOYD. being taken of the numbers of flies which visited and tasted the solution and then left, and of those which remained to drink. Controls were made with tap water. A fly only “ tastes ’’ when it presses its proboscis one or more times against the wet surface, and quickly withdraws it. It “ drinks ’’ when the proboscis remains pressed against the surface for a few seconds or longer. Apart from the count, an impression may also be gathered from the behaviour of the flies. On a fluid which they find attractive they cluster, but on an unattractive one there are never more than two or three flies at one time. The total number of visiting flies was about constant in each test. The disadvantages of the presence of formic acid in formaldehyde are shown by the following tests, where a solution containing 12 grms. per litre was tested against the one containing 1°8 grms. per litre. In ten tests the less acid solution killed twice as many flies as the more acid one. TABLE V. Showing the Disadvantages of the presence of Formic Acid in Formaldehyde. Soluti Percentages of Av. per cent. Total ete toon flies killed killed killed Formaldehyde 1:8 grms. | | ) ECF OE | per lit. 10% | 100in each case 100 808 39 12-0 grms. | | | He OOF E. | | per lit.10% | 54, 78, 54, 27, 34, 38, 34, | | 51, 43, 50 | 50°6 | 366 j Flies are less readily repelled by dilute alkaline than by acid solutions, and will drink solutions containing 0°1 to 0°2 per cent. caustic potash, or 0°07 per cent. lime, as readily as water. There is possibly even a little attraction in it. The fol- lowing Table VI shows this, the data being obtained in the same way as those in Table LV. Tasie VI. Showing that certain weak Alkaline Solutions are not deterrent. | | No. of Av. flies {| Av. flies Av. percentage Solution tests tasting drinking which drank | Tap-water control 5 38 ol 83 KOH2% 1 35 0 | 0 » 0°3-0°4.% 2 31 14 | 45 » 0°1-0:2% 2 | 34 30 | 91 Ca(OH):, 0°14% (lime | water: 2 37 20 54 » _ 0°078% 2 | 33 20 61 » 007% 3 | 37 30 81 Temperature 77° F.; relative humidity 57%. FORMALDEHYDE AS A POISON FOR HOUSE-FLIES. 51 It is therefore advisable always to add as much alkali to the formaldehyde as possible, without deterring the flies, to neutralise any formic acid which may be present, or any which may be formed by oxidation during the exposure of the solu- tion. Table VII shows the benefit of the procedure. TaB_eE VII. Showing the Advantages of adding certain Alkalis to Acid Formaldehyde. ye | Percentages killed (actual Solution, 10% in each case numbers in brackets) | Control formaldehyde (12°5 grms. H-CO-OH per lit.) 100 in each case +50% lime water in dilution 283 (155), 241 (82) » +K,CO,, 15 grms. per. lit. 300 (36), 268 (147), 270 (92) (equiv.). +(NH,),CO;,10 grms. per lit. | (equiv.). | 68 (37), 82 (28) Control formaldehyde (1:8 grms. H-CO°OH per lit.) 100 in each case . +100% lime water indilution | 76(47), 90 (923) * +P. . ss * - " | 114 (269), 104 (312), 142 (57) ” apa POG CO. | 125 (50) > + 1% NH,0OH | 50 (20) Control formaldehyde (1:1 grms. H:CO-:OH_ per lit.) 100 in each case = +502, lime water in dilution | 131 (297), 124 (31) 9 +25% ” 2? oh 2” 110 (28) Ces cepa ye cena It will be seen that the addition of the alkali improved the effect in every case except when the whole fluid of dilution was lime water, which made it too strongly alkaline, and when ammonia was used, methylamine being formed in these cases. Time water is recommended in practice in preference to other alkalis because it is a weak solution of definite strength. It is prepared by allowing cold water to stand over lime, slaked or unslaked, for a few hours and then filtering the solution before use. Such a solution contains 0°14 per cent. Ca(OH),. If made with hot water the solution is weaker. The maximum amount, half the total fluid of dilution, should always be used to counteract the formic acid which would otherwise be formed after dilution, during exposure to the air, or by the bodies of dead flies falling in. In a trap which will be presently described, but an imperfect model, which was exposed in the fly room, this was very clearly shown. The fluid in the trap was 7°5 per cent. formaldehyde, 35 per cent. lime water (equivalent to the acid present), and 5 per cent. sugar. On the first three days this trap killed 1,800 flies, and then became ineffective. Four flies had crept into the poison, and when this was titrated it was found to have become +2 acid. An equivalent of KOH was added and it became again effective. Humidity of the Atmosphere. A test was made at a temperature of 96° F., and the poison was found to be effective. There is no reason apparent why it should not be effective in a tropical country. The humidity of the air, however, plays a considerable part, as indeed 5? CAPT. LL. LLOYD. it must with any poison which is administered as a drink. In wet weather flies have less inclination to drink and there is more moisture available, so that they are less likely to discover the trap. Experiments were made to test this. (i). A jar with an area of 2°5 sq. in. was filled with sand wetted with a 10 per cent. solution of formaldehyde without any addition. It was placed in the centre of a dish of dry sand, area 95 sq. in., and this was placed in a greenhouse containing a large number of flies. No water was available and the atmosphere was dry. After twenty-four hours 1,560 dead flies were counted on the surface of the sand, or within a few inches of the dish. The experiment was repeated with approximately the same number of flies available and the same atmospheric conditions, but in this case the sand in the surrounding dish was wetted with water, equally with the poisoned surface. After twenty-six hours only 20 flies were found dead, the majority having obtained their water from the unpoisoned surface. (1). This series was carried out in a glass cylinder 24 in. high and 10 in. in diameter. Wet and dry bulb thermometers were included. The formaldehyde (10 per cent. in water) was supplied soaked up in sand, area 2°5 sq. in. The duration of each experiment was twenty-two hours. The air was in no case quite dry, as the amount of moisture given off by the poisoned surface is considerable in this confined space. It was found that the formaldehyde was effective with a relative humidity of 76 per cent. and 72 per cent., that its effect was about halved with a relative humidity of 81 per cent., and that above 90 per cent. it was much reduced, available water being present throughout in this case. The results are given in Table VIII. From these it may be concluded that liquid poisons will not be very effective if much other available moisture, such as dew, is present. Formaldehyde works well in a room where there is a wet sink, as was the case in the fly room where these experiments were performed. It would be useless, however, to place the poison in the sink or immediately over it, though it will be found very effective a few feet away. Flies are usually going about looking for food, and seem to taste all moisture they come across in their wanderings, either in the hope of its containing food or simply for the sake of the drink. Their capacity for utilising solid food depends. on whether the crop contains fluid with which to dissolve it. Tase VIII. Showing the decreasing effect of Formaldehyde with increasing Humidity. | Conditions in Cylinders Temp. | Rel. Humid | Total flies | Perooniage | | | CaCl, added to dry the air . | Tor" | 7 pay 229 94 Air not treated | TaOAT we 76% | 125 89 Wet blotting paper as base | da°0° 81% 66 47 Wet blotting paper as base and top | 76°3° | 93%, 90 17 (ui). Experiments with baited formaldehyde under saturated conditions were made. It was found that the baits about doubled its efficacy, but the results were still very unsatisfactory. The details are given below in Table IX. Sugar in water was the bait and the formaldehyde was employed in various strengths, being supplied FORMALDEHYDE AS A POISON FOR HOUSE-FLIES. 53 in two-inch saucers with blotting-paper covers. Each series was continued for two days and the cyliriders were kept as wet as possible, though in several cases they dried somewhat towards the end, with a corresponding increase in the death- rate. The controls were on a dry table with muslin tops to the cylinders, the relative humidity being about 76 per cent. The fourth experiment in Table VII is taken as a control for the saturated conditions with unbaited formaldehyde. To obtain a proper estimate the percentages dead at the end of twenty-two hours should be studied. Another series carried out in precisely the same way, but with bread-water as a bait instead of sugar, gave very closely similar results and will therefore not be detailed. TABLE IX. The effect of Barted Formaldehyde compared under Wet and Dry Conditions. Bed Se ae 5 at SER Se Oe es ac | Percentages dead after : Solution _ Conditions | No. of flies _K-—_, —_____—_—____- 5 hrs. 22 hrs. | 48 hrs. No bait, formaldehyde 10%, | wet 90 | 17 Sugar 5% As ne dry 38 | 32 97 100 * ” » » wet 34 | 3 35 79 a d TH, | dry Ce ee 9] 100 e. Be a ~ wet 28 3 Zo 69 ar ¢,: . 50% | dry 21 43 100 100 Nes A .. wet Ce as 4 22 78 * 4 . es ey a diy 19 | 2) 79 | 100 a “A te i oe wet 17 | 0 6 | 82 Humidity and the use of formaldehyde on ships. Lt.-Col. W. Byam, R.A.M.C., reported that while he found formaldehyde very effective on board ship, it failed to be of use in Egypt. This is in accordance with the variations in the accounts of its utility. That used in Egypt was possibly too acid. Experiments were devised to discover what its effect would be on board ship where the atmosphere is moist, but not necessarily saturated, and where most of the available water is salt, owing to all the woodwork being saturated with this substance. (1). A cylinder was washed inside with 10 per cent. salt solution, flies were intro- duced and blotting-paper soaked in the salt solution placed above and below; a shelter was erected over the poison dish to prevent the salt water dripping in. A control was set up, fresh water being used instead of salt solution. It was found that the flies were poisoned about twice as rapidly in the one with salt water as in the one with fresh water. (2). This was carried out in cylinders, a bait being used in the poison which was soaked up in jars of sand (area 2°5 sq. in.) that were placed in dishes of sand (area 13°5 Sq. in.) wetted in one case with 10 per cent. salt solution and in the other with water. The atmosphere would be very moist from the vapour given off by the wet sand, 54 CAPT: LL. LLOYD. TaBLE X. Showing the Effect of Formaldehyde when available Water is strongly saline. Percentages dead. after : Solution Conditions Total flies - : a | j | 3 hrs. 28 hrs. | 48 hrs. | Formaldehyde Saturated with 10% in water salt solution | 29 14 69 93 Formaldehyde | Saturated with | | 10% in water | fresh water | 27 4 | 39 78 | and the conditions would approach those on board ship when it is not actually raining. The results, which are given in Table XI, are very striking, as the flies died rapidly when the available water was saline, and slowly when it was fresh. By further experiment it was found that the flies drink as readily as water a solution of 1 per cent. salt, that stronger solutions deter them, while 5 per cent. and upwards repels them entirely. Formaldehyde properly neutralised and baited will be found very effective on board ship. TABLE XI. Showing the Probable Effect of Formaldehyde on Board Ship. | | Percentages dead after : } | | Solution Conditions Total flies = oe te Ries fl Sekai, 24 hrs. | | 7 | eae Ae | ro) 1) oO Percentage of tasting flies which drink, 1a) °o io F secoscccceveceee Lormaldehyde and water, and control of water. aS ¥ 5 59 % lime water £ ,, 00 % lime water. oe ee ee ” ” 5 5% sugar 3 5 of sugar. 50 % lime water and 5 % sugar. Diagram 1. Showing the deterrence of acid formaldehyde and its remedy. former gives the factor to the solution. Controls were made with water alone, with 50 per cent. lime water, and 5 per cent. sugar solution, and each series was checked by one or more repetitions. The time allowed to each exposure was five minutes. The following solutions were thus studied: formaldehyde in water ; formaldehyde in 50 per cent. lime water ; formaldehyde in sugar solution ; formal- dehyde in 50 per cent. lime water with sugar. The detailed results are given in 56 CAPT MLL: LLOYD: Table XII, and in Diagram 1. In the diagram the points represent the percentages of the tasting flies which remained to drink. The bottom dotted line represents formaldehyde and water, and shows that any mixture of these is deterrent as com- pared to the water control, of which an average of 82 per cent. of the tasting flies drank. The 10 per cent. formaldehyde deterred from drinking nearly 82 per cent. (67 out of 82) of the tasting flies which would have drunk had there been no formalde- Taste XII. Showing the Deterrence of Formaldehyde, and us Remedy. | | | | | No.oi/Total| No. of flies | No. of flies | Percent. of tests flies tasting drinking drinking to ; | tasting flies Solution | | | | 1). | | Av Max Mir Av Max) Min Av |Max/ Min Water controls | 8 |. 263 | 33 ae P26 eer 1185 | 20"! Sasa ree ,, formaldehyde 10%! 6) | 227 |S SO COR ras Peles ita! Sel ee eal ees * i‘ 75%, 6 | 190 | 32 | 35) 27 | 7) 13) 0} 16) 38} v i f 50%!) | 6 195 | 32° 1°38 126.) | es oan Saag ; 7 25% 6 -|} 297°| S50 49ekoG, | 92/97 Nia BOr aSrlise . ds 20% | 4. | 148 ]-39-|-634) -24en ol hoot sel Ou AT 9) Be ¥ : 15%, | | 4} 151 | 38 | 47 | 29) |° 26.) 34 1G: | ee ee ; ‘ 10% | 2 715.1 37 | 38. 37) | 230318 26) | ised ee 33 - me} ule 74. | 37 | 28 | 36.) 30 | Tals 7 30) 2 san em Lime water controls 50S, 3 112 | 37 | 39 | 35 | 30 | 31 | 29 | 80 | 86 | 72 ,. formaldehyde 10% 4 }| 186 | 34. | 42°°255" Sed 1 Gu 29) eu cae a * 75% a0} 121-30" |. 407723) tes os) ae ie ¥ 5-09, 4 |} 122 | 30| 33 | 28)) 16 | 2E Uh oad ee ee - . 2:59!) | 14 [| 114) 28 | 320s) 265) 20 23 a aaa eel ae ‘ - B0% bo D264} B24 -BB- 1B BB125 4) abel ee 6) Be ‘ P 15% 2.4 HO /°35 | 89-1 BLS) 28:1 27 7 20) G6 ae Sao Water, sugar controls wy ae wl | 45 | 45 | | 40 | 89 | , formaldehyde 10% Bh 6 | BS |. 45°30 Tae) Ol ls axl at 6 cae ss 5 T5O,, 2 63 | 31 | 36 | 27 | 17 | 23 | 11 | 52 | 63 | 41 : ? Oo, 2 7. | 35 | 38 |-33" 22 | 284 2h | 62°|.69:) ae > ? Deb, 2 66 9/933"| 35°| 31-28. 31; | 265/86 | 88 | BA Lime water 50% | | | sugar 2 a | | | , formaldehyde 10% | 2 60) 30 | 804 30°] Ia tre 47 eae 99 : 1 ae a 72.| 36 | 41 | 31 | S429 20.) 68) Fes : “ BUcG, | 2 60 | 30 | 35'| 25 | 23 | 27 | 20 | 78 | 80 | 77 , es oy Se 52.126 | 27 | 25 | 22) 24) 21 eG eae | | | | Temp. av. 77° F. (78:9-76:0°). Rel. humid. av. 50% (58-39%). hyde present, and so on. The line of dashes represents the percentages which drank the formaldehyde in 50 per cent. lime water. It is higher than the previous one because, while the deterrent influence of the formaldehyde remains as before, that of the formic acid is removed. It also is always below the water and lime water controls. The next line of dots and dashes represents formaldehyde and 5 per cent. sugar. without lime water. The presence of the bait has partly counteracted the FORMALDEHYDE AS A POISON FOR HOUSE-FLIES., 57 deterrence, so that at 3 per cent. it is equal in its attraction to the water control, and at 2°5 per cent. approaches very near to the control of sugar and water. The continuous line represents the mixture which is finally recommended, 50 per cent. lime water and 2°5 per cent sugar. It is equal to the water control at 5°5 per cent., and then becomes more attractive than water because the flies recognise in it a food, while the formaldehyde is not repellent enough in weak solution to deter them in large numbers. At 5 per cent. only 12 per cent. fewer flies drank than would have done so if the solution were sugar and water only. At 2°5 per cent. formaldehyde, only 3°5 per cent. declined to drink. The neutralisation of the formic acid and the addition of the bait have at this point nearly counteracted the deterrent action of the formaldehyde. It is of considerable interest to notice that the lines which represent the formaldehyde with and without lime water are parallel in pairs, while those with and without sugar converge in pairs at 2°5-1°5 per cent. This shows the influence of the presence of formic acid in very dilute solutions. The formalde- hyde used proved on titration to be + 60 acid (Eyre) which is equivalent to 2°76 grams of formic acid per litre of formaldehyde. At 5°0 per cent. dilution this is 0°014 per cent. formic acid, and at 2°5 per cent. it is 0°007 per cent. formic acid. The presence of sugar is sufficient to counteract the latter but not the former. A factor which has not been taken into consideration is that the formaldehyde vapour may be deterrent and that fewer flies may taste stronger than weaker solutions. This was not apparent in the open, but showed itself when the experiment was repeated in enclosed spaces. There was then progressive deterrence as the strength of the solution increased, as was shown by the smaller percentage of the available flies which settled on it. The curve of this deterrence is represented by the dotted line in Diagram II. Consideration of it will show that its fall is so gradual that it would not affect the argument even if its influence were as powerful in the open as in the enclosed space, since the influence of the other factors is so much more potent. The Toxicity of Formaldehyde. In determining this factor, the solution used was 50 per cent. lime water and 2°5 per cent. sugar, only the percentage of formaldehyde varying. About 50 flies were enclosed in each of several glass basins of capacity of 600 cc. The basins were then placed over saucers containing the poison and a count was taken of the flies that tasted and those that drank. In the first series the time allowed was fifteen minutes, but it was found that all the flies that intended to drink did so in the first five minutes, and the exposure was subsequently limited to this time. The poison was then re- moved and the edge of the basin slightly raised to allow air to circulate. At the end of one hour and of two hours the numbers lying inert were counted, the difference between the two counts representing the recoveries, as formaldehyde seems to act as an anaesthetic, and its effect in weak solution, or when a very little is taken, is sometimes only temporary. No recoveries were noted after two hours. All the flies were then anaesthetised and counted. Four series of each percentage were thus made. The complete results are summarised in Table XIII. This shows the total and average numbers of flies exposed to each percentage of formaldehyde used (from 2°5 to 20 per cent.) ; the average, maximum and minimum numbers of flies which tasted ; the average, maximum and minimum percentages of those which drank (681) r 58 CAPT, LL, LLOYD. compared to those which tasted ; and the percentages of those temporarily and permanently incapacitated compared with those which drank. The calculation of the numbers of those which died against those which drank is more satisfactory than calculating them against those which only tasted. The numbers thus obtained are frequently above 100, as at a strength of 5 per cent. and upwards the amount taken by a fly in its taste is often enough to kill it. Reference to the table will show that not only the averages, but also the maxima and minima, run in series corresponding to the strength of the poison, allowance being made for occasional variation due to: experimental error. The percentage of those which died compared to those which drank gives the factor of toxicity to each strength of formaldehyde. These factors are plotted in Diagram 2 and are represented by the line of dashes. The toxicity increases rapidly from 15 per cent. upwards, and falls rapidly from 5 per cent. down- wards, the rise from 5 per cent. to 15 per cent. being gradual. The dot and dash curve is taken from Diagram 1, with the addition of the factors for 15 per cent. and 20 per cent. strengths which are taken from Table XIII. This curve represents the per- centage of tasting flies which remain to drink a solution of formaldehyde containing 50 per cent. lime water and 2°5 per cent. sugar. It is the curve of deterrence. By combining these two curves a hypothetical curve is constructed which is represented in the diagram by the continuous line. For example with a strength of 10 per cent. formaldehyde exposed in the open, 47 per cent. of the flies that taste it will remain to: drink. Some however will absorb enough poison in the taste to kill them, and the number which finally die is not 47 per cent. of the tasting flies but 133 per cent. of 47, that is, 62°5 per cent. of the flies that taste die. The relative effects of different. strengths of formaldehyde, as represented by the curve, were thus found. Itis a natural curve with its maximum at 5°0-6°0 per cent. formaldehyde, and this is therefore the optimum strength to use in practise. The straight dotted line in the same diagram represents, as explained above, the deterrence of the vapour at different strengths in enclosed spaces. Its tendency is to recommend the use of as weak a solution as possible, but the decline in the effect below 5 per cent. is so rapid that this factor could play no part in the decision as to which strength to use. TaBLe XIII. | | | | | % to drinking flies At ; | | | % of flies | % of tasting Solution =| ln ,.,| Av. | tasting | flies which |— a. 50% lime \Tests |Total) jo. | s eee de ee ae | no. | flies | Aies | | Downin| Dead after | ie fea gyi ee ee ee rn Tee 2 hous _ |Av. Max. Min.|Av. Max. Min. Av. Max. Min. Formald. 20% | 4 | 154] 39] 483 67 30/20 25 15 | 229 | 219 300 125 o 165% | 3 (188) 44} 41 53. 85 \| 80.3625.) 146 9) 1437 150 "140 2 1%, 4 | 220; 55 |} 53 55 45 | 54 67 48 136 133156. 110 a 75%! | 4 | 1821 45159 69 49|68 96 63) 111 | 112 136°. 87 2 50% 4 | 219! 55) 61 65 53 | 86 93 79 106) 10G°"°12T ' 86 . 275% 4 | 174, 43 | 66 80 54/9196 87): 61.4 39 962) 23 FORMALDEHYDE AS A POISON FOR HOUSE-FLIES, 59 cS _~ _—— =, = Percentages of Fiies 12:5 10 7°5 Percentage strength in AGilution. —..—..—.. Flies tasting fluid which remain to drink. Percentages killed of those drinking. Relative effect of various strengths. rae eae eee Deterrent effect of formaldehyde vapour in enclosed spaces. Diagram 2. Showing the most effective strength of formaldehyde. The Optimum Concentration. The conclusion drawn from the experiments just described was checked by the following series of further experiments. In each series concentrations of 10 per cent., 75 per cent., 5°0 per cent., and 2°5 per cent. formaldehyde, with 50 per cent. lime water and 2°5 per cent. sugar, were placed in saucers with blotting paper covers and these were exposed to flies in cylinders (10 in. diam., 8 in. high). Each experiment was continued for twenty-four hours, and counts were made at intervals of the flies lying inert. Seven series were completed, and about 100 flies were used in each cylinder. The summary of the whole is given in Table XIV. In this are shown the average number of flies in each cylinder, and the average percentages inert at the end of various intervals. These results are plotted in Diagram 3; the dot-and-dash line represents 2°5 per cent. formaldehyde; the continuous one 5 per cent.; the line of dashes 7°5 per cent.; the dotted line 10 per cent. Although recoveries are only (681) 2 60 CAPT, LL. LLOYD. indicated in the case of 2°5 per cent., they occurred occasionally also in the case of 5 per cent. and 75 per cent. strengths. This method confirms the conclusion that the most effective strength to use is 5 per cent., then 7°5 per cent., then 10 per cent., while 2°5 per cent. 1s the least effective. The factors of effect of these concentrations had been found to be 80°3, 7671, 62°5, and 33°5 respectively. TABLE XIV. Showing the Death-rate of Flies exposed to various strengths of Formaldehyde. Formaldehyde No. of Av. no. . ; strength Vouie flies Av. percentages inert after : EO: Ba : 1 hour | 3 hours | 4 hours | 5 hours | 24 hrs. a 81 arr ae wee: 92 1096 7 92 4a | 53), |. 63 |. 6p 93 5-0% 7 98 St ee oer oe 96 25%, di 96 42D hd ASE Ca 53 89 | g pests — oe ee Lee | a =, = a a cS e oe ae e - oe .° st x + e— i - a | er ae _ ry =-_ el? e ee = e? . c a° ae ° } 4 °° e : 2° hs °°? . f . or 60 ~ © ag Fae p $s 60 y v S 5p n S c 40 v : he i Y Q 30-4 Le) p op 7G "gh em ame Mete ia Tht) ie ! IS) 14S 1G) 7) 16 MISE 2021 Wee (26. es /nterva/ls of time in hours. One 2S Formaldehyde 10 % ie eee 3 13% 9 5°0 wf 0 "Fh O — +? au 6 6 «<= © 6 9 2°5 Yo 20 sO Diagram 3. Showing the relative strengths of formaldehyde. FORMALDEHYDE AS A POISON FOR HOUSE-FLUIES., 61 Advantages of adding a Bait. Reference to Table XII and to Diagram 1 will show that to get the best effect from the formaldehyde it is necessary to add some bait to counteract the deterrent effect. By this means a larger number of flies are induced to drink. Sugar is the most attractive and most permanent bait, and 2°5 per cent., or half an ounce to the pint, in the dilution is sufficient ; to make it stronger does not make it more effective, while at 1 per cent. it begins to lose its influence. An advantage of this substance is that it has no interaction with the formaldehyde. Glucose, maltose and lactose have also been tried, but are less effective than cane sugar. Molasses will be found effective. An infusion of hay, boiled and strained, and a cold infusion of horse-dung have been tested and found to be effective, but less so than sugar, and they are not recommended for this reason. The formula to be employed is :— 40 °, Formaldehyde, 50-60 ce. ag i 1 fluid ounce. Filtered lime water, 500 cc. .. ae i! 4 pint. Sugar, 25 grams Ns my a ounce (dessertspoonful). Water, add to make 1,000 ce. . * add to make 1 pint. Method of Administration. It is necessary to protect the formaldehyde from the action of air, and to prevent flies from falling into it, both of which factors make it acid, and soon cause the lime water to be neutralised. A simple form of trap has been devised for this. The poison is placed in a bottle, and the mouth is closed by means of a platform of absorp- tive material from the centre of which a stem of the same material passes down into the fluid. The top is wetted with the fluid at the commencement and is kept wet by rol— bo Fig. 1. Form of trap for use with formaldehyde. capillary attraction. Such an arrangement may easily be made from blotting paper. A circle of the paper two inches in diameter is cut out, and a small slit is made in the centre. ‘Two strips of the paper of the same width as the slit are then passed through it and the ends are doubled outwards and downwards. The strips are then drawn down till the doubled-back portions rest on the platform, and the trap is ready for use (see fig. 1). 62 CAPT. LL. LLOYD. This form of trap may be used under cover or outside when it is not raining. The paper should be renewed every third or fourth day in any case. A more permanent form may be made from plaster of Paris or some other durable material, porosity being an essential. The stem in this case is circular and should be in one piece with the top, or fit into it very tightly, otherwise the fluid will not rise. It may be necessary to soak this trap im water occasionally and dry it. These traps have been submitted to the very severe tests of being placed near dishes of the food which has been found most attractive to house-flies, casein, banana and bread, mixed and wetted. One trap with the top made of plaster of Paris was placed about a foot away from a large dish of this material, and in three days poisoned 1800 flies, which were found on, or within a few inches of it. After further neutra- lisation it became effective again and remained so for ten days, about half the original fluid (200 cc.) having been used. Another one, with the top made of filter paper two inches in diameter, was placed in a corner of the fly room, and was active for ten days until it was removed, the paper top having been renewed once; 1,560 flies were found dead around it. A third one with a paper trap was still active at the end of a month. Whether a trap is still effective may be determined by sweeping away the dead flies from its vicinity. If no more flies accumulate, and there are still flies about, the fluid has probably become acid and should be renewed, as it will not be found worth while in practice to neutralise it again. The traps should be placed wherever flies are numerous, especially near rubbish bins, in stables, near manure heaps, in kitchens and latrines. Observation will show which are the best positions. Comparison with other Poisons. Formaldehyde as a fly poison was not tested against sodium arsenite, but it has two great advantages over it : it is safe and it is more economical. With 2 per cent. sodium arsenite solution 10 per cent. sugar is recommended, while formaldehyde only requires a quarter of that amount. The use of sodium arsenite would always be dangerous to man and to domestic animals, especially if issued to the general public. That the poison used shall be safe in any hands is a great desideratum. The action of formaldehyde was tested against that of sodium fluoride, recently recommended as an effective fly poison (Jackson and Lefroy). Formaldehyde (5 per cent.) and sodium fluoride (1 per cent.) in saucers were placed in two separated halves of a deep glass dish, equally lighted and heated. No baits were used in either. After twenty-four hours in the formaldehyde section 650 flies were dead as against 234 dead in the sodium fluoride section. As more might have escaped from the section containing the fluoride than from the other section, a control comparison was made with cylinders of flies, four containing as poison | per cent. sodium fluoride with 3 per cent. sugar, and four containing the formaldehyde mixture recommended above. Notes were made of the death-rate and it was found that there was no appreciable difference in the actions of the two poisons. The results are summarised in Table XV. However from the first experiment it is seen that flies poisoned travel less from the formaldehyde than from the fluoride, which is an advantage. Authorities differ as to whether sodium fluoride is a poison to man, but as it is used in food preserving it is certainly not a very dangerous substance. FORMALDEHYDE AS A POISON FOR HOUSE-FLIES, 63 TABLE XV. | Aver. percentages dead after : No. of Aver. ad Solution. tests. | flies. | | 1 hour 4 hours 24 hours } _ a A : hes 2 1% sodium? fluoride, 3% sugar ieee ae 57 76 94 5% formaldehyde, 50% | : lime water, 3% sugar 4 | 74 6G: {| 74 96 Summary. The effective action of formaldehyde as a fly poison depends on the freedom ol the-exposed fluid from formic acid, and, to a less extent, from methylamine. That used for fly poisoning should therefore be colourless and free from a fishy odour, and a weak alkali, in slight excess, should be added in dilution to neutralise any acid present and that which will be formed during exposure. The formula recom- mended is :— 40% Formaldehyde .. + aa s be em oily Clear lime water oy oe 1 rm al - 50% Sugar .. ne Zs 2% ae “ee y Pe OF Water .. ; re: add to make 100 This should be exposed in a trap which will protect it from the air and will prevent flies from falling in. Formaldehyde, like any other stomach poison for flies, will be most effective under dry conditions. References. 1. O. C. Lodge, Proc. Zool. Soc. Lond., Sept. 1916. 2. Thorpe, Dictionary of Chemistry, on formaldehyde. 3. Mally, C.W. “On the use of poisoned bait for controlling the house-fly ; ” S. African Journal of Science, June 1915. 4. Jackson and Lefroy, “Some Fly Poisons for Outdoor and Hospital Use ;”’ Bull. Ent. Res. vii, p. 327. Lees, 7 aa ; ‘if Lee » 4? AFRICAN APHIDIDAE.—Parr IV. By Frep. VY. THEOBALD. The following new species of Aphids have been sent from Africa ; four are from a collection forwarded by Mr. Claude Fuller of the Department of Agriculture of the Union of South Africa (Macrosiphum cassi, Phorodon violae, Capitophorus chrysanthemi and Forda natalensis) ; one has been sent by Mr. C. C. Gowdey from the British East African Protectorate which has to be placed in a new genus (Cer- craphis). Macrosiphum cissi, sp. nov. (fig. 1). Alate viviparous female. Antennae longer than body, rather dark, pale at base of third segment ; basal segment wider and a little longer than second ; third longer than fourth, not so long as sixth, with a line of 15-18 sensoria extending nearly to the end of the segment the distal one usually markedly separated from the others ; fourth segment a little longer than fifth, the latter with the usual sub-apical sensorium ; sixth with long flagellum and usual sensoria at apex of basal area; fourth to sixth markedly imbricated ; with short, simple, scattered hairs. Eyes large and dark. Frontal lobes small, but prominent, median ocellus prominent; head with slightly spatulate Fig 1. sagen eae cissi, sp. n.: A, heed af alate 2; B, cornicles and cauda of apterous 2; ©, cornicles and. cauda of alate 9°; D, cauda of alate 2; E, head and antenna of apterous eH hairs. Proboscis rather broad, dark at apex, reaching to or just beyond the second coxae, apical segment a little longer and narrower than second. Legs long and thin, dark at apices of femora and tibiae, tarsi dark, most of hind femora dark, short hairs over apex of femora and tibiae. Cornicles black, long and cylindrical, about as long as third antennal segment, apex reticulate, remainder imbricated. Cauda about one-half the length of cornicles, pale, with three pairs of long lateral hairs, a smaller apical pair and a median sub-apical one. Anal plate darker and with many hairs. Cauda not quite level with apices of cornicles. Wings normal. Length, 2-2-3 mm. 66 FRED. V. THEOBALD. Apterous viviparous female. Antennae longer than body ; first to third segments pale, except apex of latter, remainder dark ; first segment much larger than second ; third longer than fourth, but shorter than sixth, with a group of 2-6 sensoria near the base ; fourth longer than fifth, the latter with usual sub-apical sensorium, both combined a little shorter than sixth to about the same length ; flagellum of sixth long; hairs normal. Head with well formed frontal lobes and with a small median prominence showing only in some specimens ; other mounted specimens show the vertex almost flat. Eyes large and dark. Proboscis reaching to or just beyond the second coxae ; apical segment longer and considerably narrower than the penultimate. Cornicles dark, cylindrical, somewhat expanding basally, reticulate at apex, remainder imbricated ; shorter than third antennal segment. Cauda pale, about half as long as cornicles and slightly thicker, with three hairs on each side and some apical ones ; not quite reaching the apices of the cornicles. Legs relatively shorter and thicker than in alate female; apices of femora and tibiae and tarsi darkened ; hairs moderate. Anal plate dusky, with rather long hairs. Hairs of body slightly spatulate. Length 2°2-2°5 mm. Nymph. Antennae longer than body ; darkened on fifth and sixth segments ; basal segment larger than second ; third longer than fourth ; fourth longer than fifth ; the sixth much longer than 4-++5 ; flagellum long, but relatively shorter than in mature forms ; a few scattered hairs. Proboscis with dark apical segment, reaching to the third coxae ; last two segments nearly equal. Wing pads dusky; cornicles dark ; about as long and as thick as third antennal segment, imbricated ; reaching past the cauda. Cauda triangulate, pale, with two pairs of lateral hairs. Legs with dusky apices to tibiae, but not to femora; latter with hairs on one side; tibiae hairy. Length 1°8 to 2 mm. Food-plant. Cissus sp. TRANSVAAL: Pretoria 18.x.1918. Described from three alate females and a number of apterous females and nymphae. No colour notes were sent, but the species is so well marked that I have described it. It comes nearest to Macrosiphum, but also approaches Myzus, especially in the alate stage. Mounted apterae show the vertex in different forms, some have it almost flat, others of typical Macrosiphum form, whilst others show a median prominence. In the various larval instars the differences are great; the first has very short thick cornicles, in later stages they increase in length. The length of the cornicles varies in the adults. The plant upon which these Aphids were found—C?ssus—is one of the Order Ampelideae. This genus is often merged into Vitis. None of the species recorded from Vtis agrees with Mr. Fuller’s specimens, and so far no Aphid has been recorded from Cissus. The Aphids recorded from Vitis are Aphis allinorsensis, Shimer ; Aphis rvparvae, Oestlund ; Aphis vitis, Scopoli; Macrosiphum viticola, Thomas ; Hyalopterus arundinis, Fabricius ; Peritymbia vitisana, Westwood ; Rhizoctonus ampelinus, Mokrzeki ; Schizoneura ampelorhiza, Del Guercio ; and three species of Phylloxera. AFRICAN APHIDIDAE.—PART IV. 67 Rhopalosiphum carduellinum, Theobald (fig. 2). Bull. Ent. Res. vi, p. 112. figs. 9 and 10, (1915). Apterous viviparous female. Antennae about as long as the body, arising from prominent frontal tubercles ; basal segment much larger than second ; third not quite as long as the sixth, with 17-20 small sensoria along one side; fourth slightly longer than fifth, with 3-6 sensoria ; fifth with normal sub-apical sensorium ; sixth about as long as 4-++5, its basal area about one-fourth the length of the fifth ; the apices of segments 3-5 are darkened and there is a dark area at junction of basal area and flagellum of sixth ; all the segments are imbricated ; the sensoria on 3 are of varied size and shape ; Fig. 2. Rhopalosiphum cardueliunum, Theob., apterous female; A, head and antenna; B, cauda and anal plate; C, cornicles ; D, sensoria on 3rd antennal segment. those on 5 are more regular in outline. Proboscis reaching to just past the second pair of legs ; two bands of a few hairs on penultimate segment ; two marked sub- apical ones on the apical arising from clear circular areas. Legs with apices of femora and tibiae darkened and dark tarsi, a few hairs on apex of femora and numerous short ones on the tibiae. Cornicles about as long as the fourth antennal segment, about twice as wide and somewhat swollen on the apical half; apex with a few striae, remainder faintly imbricated. Cauda pale, broader than cornicles, more than half their length, with three long hairs on each side and a curved sub-apical one ; reaching just beyond the apices of cornicles. Length, 1°8-2 mm. Food-plant. Thistles (Carduus sp.). TRANSVAAL: Pretoria, 20.v.1914 (C. Fuller); Onderstepoort, 28. vii. 1914 (G. B. Bedford). Type in the British Museum. The apterous female sent with the alatae of this insect appears to be a distinct species. The sensoria in the two apterous females sent by Mr. Claude Fuller appear so much more closely allied that I have little doubt that they are true carduellinum. Numbers of nymphae sent were mostly all of a uniform pale colour, evidently vellow or green, but a few show the tips of the wing pads dusky. One of the sensoria on the third antennal segment is markedly bean-shaped, and the varied size of the others is somewhat characteristic. ; Phorodon violae, sp. nov. (fig. 3). Apterous viviparous female. Yellowish green to dull pale yellow. Antennae longer than body, of same colour as body except extreme tip of fifth and whole of sixth segments, which are dark ; 68 FRED. V. THEOBALD. basal segments (1 and 2) somewhat darker than the rest ; the basal one much larger than second and somewhat projecting on its inner side ; third segment longer than fourth, but not so long as sixth ; fourth and fifth about equal in length ; flagellum moderately long, about as long as third ; a few scattered hairs over the segment and the third with short stiff hairs arising from small projections along one side, some extending on to the fourth ; basal segment imbricated. Frontal lobes large, with Fig. 3. Phorodon violae, sp. n., apterous viviparous female: A, head of apterous2; B, posterior end of body and end of antenna; B. 1, cornicle; C, C.1, cauda. a marked, slightly inwardly directed, cone-like process on each side, imbricated and with two minute hairs. Two stiff hairs also project from the vertex. Eyes large, red to dark red. Proboscis pale, apex slightly darkened ; reaching to or just beyond the third coxae, narrow ; apical segment a little narrower and longer than penultimate. Legs of same colour as body, except the tarsi which are darkened ; very short stiff hairs on the tibiae. Cornicles of same colour as body, moderately long, but shorter than the fourth antennal segment, slightly swollen from the middle onwards ; faintly imbricated, sometimes slightly expanded at the base. Cauda pale, short, bluntly triangulate, spinose, with two short hairs on each side ; projecting well beyond the cornicles. Anal plate slightly darkened, narrow, spinose, with four long apical hairs and others somewhat shorter. On each side of pronotum a small lateral papilla and traces of four others on each side of the abdomen. The body is contracted at the apex. Length, 1°3-1'8 mm. Nymph. Colour similar to adult. Antennae about as long as body and of same colonir ; sixth segment and apex of fifth dusky ; basal segment much larger than second, third a little longer than fourth ; fourth and fifth about equal; sixth as long as or a little longer than 4++5, its basal area about half the length of the fifth. Frontal processes as in alate female. Eyes bright red. Wing-cases of same colour as body, also the legs, except apices of tibiae and the tarsi, which are smoky, with numerous fine short hairs. Cornicles and cauda pale, the latter bluntly triangulate and AFRICAN APHIDIDAE.—PART IV. 69 extending beyond the cornicles, which are shorter than the fifth antennal segment. Cauda finely spinose, but showing no lateral hairs. The pale proboscis reaches nearly to third coxae. Length, 1°2-1°5 mm. Food-plant. Pansy. Natat: Durban, 1.x.1912. Described from a number of apterous females and some nymphae. Nearly all the females have the posterior portion of the gut, etc., evaginated, giving them a curious ragged appearance. The shape of the body is also marked. I have placed this species provisionally in the genus Phorodon on account of the two marked frontal processes and the swelling on the basal segment of the antennae, but the cornicles and cauda do not exactly fit in with that genus, taking P. hwmuli, Schrank, as the type. The flattened anal plate occurs, however, in some stages of the hop aphid. Capitophorus chrysanthemi, sp. nov. (fig. 4). Apterous viviparous female. Apparently a pale green or yellow species, with black eyes ; dusky apices to the cornicles and antennae; tarsi all dark and to some extent the apices of the tibiae ; hairs capitate. Antennae longer than body ; basal segment larger than second and somewhat angulated on the inner side ; third segment much longer than fourth Fig. 4. Capitophorus chrysanthemi, sp. n.: A, head of apterous 9; A. 1, enlarged cephalic hairs; B,analregion ; a,cauda; b,anal plate; b. 1, hairs of anal plate; c, dorsal process ; d, body hairs. but not so long as the sixth ; fourth a little longer than fifth, the latter with the usual sub-apical sensorium ; sixth with long flagellum, basal area about one-fourth of the fifth segment. Head with marked frontal lobes and a median raised process ; with two capitate hairs on each lobe, two in front on the median lobe, and two just above them, Eyes prominent, varying from black to deep red or reddish black. Proboscis pale, reaching to just past the second coxae; apical segment 70 FRED. V. THEOBALD. narrow and acuminate, penultimate nearly as long, expanded. Legs pallid, except tarsi, which are dusky except at extreme base ; tibiae with moderately long fine hairs and a few on apex of femora. Cornicles long and very thin, about as long as fourth antennal segment ; pale, apex dark. Cauda pale, rather long and thick, bluntly elongate-triangular, with two hairs on each side, close together towards the apex and a median sub-apical one ; finely spinose. Anal plate slightly darker, spinose and with a few long hairs, broadly rounded. The thin cornicles project to about the level of the cauda, which is very much wider and about half to less: than half their length. Body hairs strongly capitate, some with fan-shaped extremities, whilst those on the anal plate and cauda are simple. The lateral view shows a small tubercle above the cauda (fig. 4, B.c.). Length, 1'8-2 mm. Food-plant. Chrysanthemum. ORANGE FREE StaTE: Bloemfontein, 18.v.1914. Described from several mature apterous females. It resembles a typical Myzus, but the median frontal projection places it in Van der Goot’s genus Capitophorus ; which appears to be undoubtedly well marked, but I find so many species between the true Myzus and typical Capitophorus that it is doubtful if this genus should be accepted, unless many others are to be initiated. Genus Cerciaphis, nov. Described from apterous viviparous females. Head flat to very slightly concave in front. Antennae shorter than body, of five segments, the third about as long as 4+5; the fifth longer than fourth, its flagellum only half the length of basal area. Kyes small. Cornicles small, cone- shaped. Two thick, acuminate anal cerci, about as long as fourth segment of antennae. Cauda small, flattened. Anal plate rounded. Proboscis rather short. Cerciaphis bougainvilleae, sp. nov. (fig. 5). Apterous viviparous female. Pale-coloured ; apices of antennae dusky. Eyessmallanddark. Legs, cornicles, anal plate, cauda and base of antennae of same colour as body. Head narrower Fig .5 Cerciaphis bouganvilleae, sp. n.: A, head of apterous 9 ; a. 1, enlarged eye; a. 2, and a. 3, sensoria; B, posterior end ; b. 1, cerei; ¢, cornicle; d, cauda and anal plate; C, cornicle. than body, which is more or less oval ; head flattened to slightly concave in front, integument strongly striate. Antennae shorter than body, of five segments, the AFRICAN APHIDIDAE,.—-PART IV. 71 basal one larger than second, the third the longest, as long as or longer than™4-+5 ; the fourth shorter than fifth, with a simple round sub-apical sensorium ; the fifth with the flagellum short, about half the length of the basal area, which is equal in length to the fourth segment ; third to fifth imbricated. Cornicles cone-shaped, of Lachnid appearance, more or less circularly striate. Two long, pale, acuminate anal cerci, striate and about the length of the fourth antennal segment to a little longer. Anal plate pale, small, with a few longish hairs. Cauda inconspicuous. Legs rather short and thick. The posterior integument is striate, but not so mark- edly as the cephalic. The proboscis in all the mounted specimens is bent forwards! but appears to reach to about the second coxae. Length, 1-4—1'8 mm. Food-plant. Bougainvillea. Ucanpa : Kampala, 14.v.1919 (C. Gowdey). Described from a number of apterous viviparous females. It is a very marked insect on account of the two posterior cerci. The head and body seem almost J devoid of hairs, except posteriorly. The nymphae have long narrow wing-buds. Forda natalensis, sp.nov. (Fig. 6). Apterous viviparous female. Globular, much domed dorsally, flattened ventrally. Grey, greyish white to brown; antennae, proboscis and legs brown. Segmentation marked posteriorly. Antennae from one-fourth to one-fifth the length of body ; of five segments, basal one about half the length of second, third the longest, about as long as 4++5 ;}fourth shorter Fig. 6. Forda natalensis, sp. n.: A, head; B, cauda, C, integument of apterous viviparous 2; D, nymph. than fifth ; fifth with a short blunt nail, a large rather projecting sensorium at its base and traces of a secondary one, also a round sensorium at the apex of the fourth ; slightly hairy ; last two segments darker than remainder. Vertex flat, with a few short blunt hairs. Eyes very small, black. Proboscis reaching to near the third pair of legs or just past them, moderately broad ; apical segment a little longer and narrower than the penultimate, with a few short hairs. Legs rather short and thick, prothoracic shorter than mesothoracic, the latter shorter than the Te FRED. V. THEOBALD. metathoracic ; tibiae and tarsi with short fine hairs, a few on the femora ; femora of first and second pairs nearly as long as tibiae, the third a little shorter. Body with short fine scattered hairs, except at the apex, where they are longer. Length, 2-2°5 mm. Immature viviparous female. Paler, legs pale brown ; antennae dark on last two segments only. Antennae about one-fifth the length of the body ; basal segment shorter than second and very little broader ; third a little longer than second ; fourth about as long as the first ; fifth a little shorter than third, with a small blunt nail, with normal sensoria at its base ; a few scattered hairs. Proboscis reaching to or beyond the third pair of legs. Legs short, just projecting beyond the body, the first slightly shorter than second, third the longest ; tibiae and tarsi with hairs as in adult. Eyes very small. Body with a few rather short scattered hairs. Length, 1:5-2:0 mm. Food-plant. Roots of a weed and im ants nest. NataL: Umzinto, 23.i1v.1911. Described from several apterous viviparous females and some immature forms. It somewhat resembles Forda rotunda, Theobald, but can at once be distinguished by the long second antennal segment and the short fourth segment, also by the flat vertex and different posterior structure. The eyes also do not project in the same way, and the body is clothed with scattered fine short hairs. Some specimens show a partial constriction on the third antennal segment, but I could find none in which it was complete, so that they are only of five segments. Species not previously recorded or from new Localities. Macrosiphum dahliafolii, Theobald, Bull. Ent. Res. vu, p. 273, 1918. Ucanpa: Kampala, 2.x11.1917 (Gowdey) ; on thistles (Carduus). Macrosiphum sonchi, Linnaeus. Syst. Nat. u, p. 735. UGanDA: Kampala, 2.xii.1917 (Gowdey) ; on thistles (Carduus). One apterous 2, which agrees with British specimens. The darkening of the antennae around the sensoria of the third antennal segment is very marked. ~I Ww ANOPHELES BREEDING AMONG WATER LETTUCE—A NEW HABITAT. By James ZETEK, B.A., Entomologist, Ancon, Panama Canal Zone. With the filling up of Gatun and Mira Flores Lakes, water lettuce (Pistia stratiotes, Linné) and water hyacinth (Hichornia crassipes, Solms.) rapidly increased in numbers until large areas became completely covered by them. Masses of these plants would become detached and float about in these lakes. It was found necessary to destroy these “floating islands,” not because they were serious mosquito habitats, but because they interfered with navigation and the operation of the spillways. The water lettuce is the habitat of the very specialised larvae of Mansonia, of which titillans is the commonest member. Knab (1913) and Busck (comments appended to Mr. Knab’s paper) refer to the changes in the mosquito fauna brought about by the rapid increase in Pistia. Both Mansonia titillans. Walker, and Aédomyia squamipennis (Lynch-Arrib.) Theo., have been collected by me in very large numbers since 1912, and to-day Mansonia is the dominant species caught in the army barracks on the west side of the canal. Knab expressed the opinion that perhaps measures would have to be taken to destroy the Pistia habitats ; now that we have found Anopheles larvae in them, this prophesy comes closer home. The only question involved is how much of a menace are these floating islands to our towns. It is my belief they are a menace to the towns on the west side of the canal, and if allowed to float toward Paraiso and Pedro Miguel, or close to Gatun, would be a source of danger there. It has been shown conclusively that our common malarial Anopheles do fly more than a mile (Zetek, 1915) ; in fact, it appears that they will fly as far as they must in order to get food. Should Las Cascades become a negro settlement, the dangers from malaria would be increased. I doubt if we can depend upon screening and mosquito-catching indoors to control the Anopheles and malaria at such a settlement. At any rate, the idea seems to be getting firmly established that the sanitation of the Panama Canal Zone is so efficient that there is virtually little danger from the Anopheles, and as a result our people are becoming somewhat negligent. It will depend mainly upon the cumulative evidence gathered by the District Sanitary Inspectors, whether or not these floating islands will need to be controlled. That this evidence may be available and accurate, it is very necessary that the sanitary Inspectors make it their practise to take ample field notes. These should contain as complete a statement of the particular habitat, date and locality added thereto, and specimens of the larvae should be sent to the laboratory for the accurate determination of the species. At the close of each month a general summary of these notes should be made, a copy of which should be sent to the entomologist. It is a regrettable fact that much valuable information is lost or buried in the heads of observers. In 1918 I found Anopheles larvae among water lettuce at Juan Mina, a citrus plantation about five miles up the old Chagres River. On 21st November 1919, Messrs. Picket and Tolar, sanitary inspectors at Pedro Miguel, D>. D. P. Curry, (681) F 74. JAMES ZETEK. Assistant Chief Health Officer, and I, made an inspection of floating islands at Gamboa, C. Z., both in the canal prism and across in the bayous ; we went as far as Empire. Mr. Pickett previously reported the presence of Anopheles larvae in these islands at this place, and that they were going southward toward his station. Our joint inspection revealed larvae of Anopheles albimanus, Wied., and A. tarsimaculatus, Goeldi, to be plentiful among the leaves of the water lettuce. An unidentified Culex was also present. This particular habitat is very favourable to the Anopheles. The larvae are afforded much protection from the schools of young fish that usually feed upon insect larvae. They are also protected to some extent from predaceous insect larvae, such as those of dragon-flies. Shade is another important factor in this habitat, and this means a more even temperature, and an almost total absence of the direct hot rays of the sun. Wave action does not disturb these larvae very much ; but the most important advantage to them is the exceptionally favourable presence of oxygen, and this certainly must be effective in prolonging the life of the larvae and pupae. This factor is very easily explained. The Pistia, being a green plant, gives off oxygen. Its leaves are at or just above or below the surface of the water. Therefore this oxygen enriches the local atmosphere available to the larvae, and some of it is undoubtedly taken up by the water and made use of by the larvae through their cutaneous respiration. During the summer of 1911, while engaged by Dr. 8. A. Forbes to study the Anopheles at Havana, Illinois, a similar condition was found to exist, excepting that in place of Prstia the plant in question was the tiny Lemna. This plant covered the surface of the water in the bayous, ponds and quiet recesses of the rivers. The Chatagua grounds at Havana were invaded each year by Anopheles, but their source was not known until we found them in and among the floating islands of this Lemna. These islands broke away from the almost continuous mass across the river from Havana and by current and wind were swept across this deep body of water and came to rest along the shores of the Chatagua grounds ; here the larvae developed, pupae formed and adults emerged in rapid succession. There is one important aspect of this sort of habitat which must not be over- looked. It is that wind and current detach portions of this floating vegetation and sweep it across deep bodies of water to new localities, where were it not for this exceptionally good vehicle these larvae would be unable to cross the barrier of deep water. Searching through the literature available at our laboratory, I find that Ingram and Macfie (1917) found larvae of Anopheles costalis, Lw., and A. marshalli, Theo., in exactly the same sort of habitat, composed of the same species of Pistia, at Christiansborg, near Accra, West Africa. Associated with them were: Aédomyra africana, Nev.-Lem., Mimomyia splendens, Theo., Culex quasigelidus, Theo., and Mansonwoides africanus, Theo. In a previous paper, in which he discusses the limitations of kerosene as a larvicide, Macfie (1917) states (p. 278) that Mansoniordes africanus flies over one mile and that its breeding places can be located with fair accuracy, inasmuch as they are composed of Pistia. On page 278 he states it is useless to hope for a total abolition of such mosquitos by ordinary oiling, and that a layer of kerosene (p. 294), no matter ANOPHELES BREEDING AMONG WATER LETTUCE. 75 how thick, applied to the surface of the water in which larvae of M. africanus live is absolutely innocuous. He is right, and this is corroborated by our observations upon Mansonia titillans, which is our counterpart to his M. africanus. But nowhere in this article, nor in the preceding one, does he touch upon the control of the Ano- pheles which live in this same habitat, nor does he refer to the use of emulsions which mix with the water and kill by contact, or the destruction of the Pistia by means of sodium arsenite or their removal by boats. The logical control is to destroy the habitat. This may be costly and prolonged, and perhaps prohibitive where large numbers of bayous, ponds, cut-ofis, etc., exist. If Anopheles alone are to be controlled, spraying with a phenol-resin soap emulsion every six days will probably be all that is necessary. It may even be effective against the Mansonia. Macfie (1917) goes into great detail in his discussion of the favourable oxygen factor in this Pestia habitat and relates experiments made with certain mosquito larvae which corroborate his contentions. That this exchange of gases is a favour- able one is quite certain. Our common Anopheles have been known to stay on the bottom for five minutes ; if through the presence of green plants they have a more favourable habitat, then it follows that not only may cutaneous respiration be much more prolonged, but the general death-rate among the larvae may be greatly reduced. This death-rate is reduced also because the leaves of the plants protect the larvae from the numerous swimming enemies in the water, and from the hot, direct rays of the tropical sun. References. Incram A. & Macriz J. W.S. (1917). The Early Stages of Certain West African Mosquitos.—Bull. Ent. Research, vii, pt. 2. Kwap, Frederic (1913). Changes in the Mosquito Fauna of Panama.—Proc. Ent. Soc. Washington, xv, pp. 41-43. Macrie J. W.S. (1917). The Limitations of Kerosene as a Larvicide, with some Observations on the Cutaneous Respiration of Mosquito Larvae—Bull. Ent. Research, vii, pt. 3. ZETEK, J. (1915). Behavior of Anopheles albimanus, Wied., and tarsimaculata, Goeldi.—Ann. Ent. Soc. America, viii, pp. 221-271. (681) F2 at 7 Pre a rome wd we in iE we ¢ 17 LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO WITH SPECIAL REFERENCE TO ADULTS AND LARVAE OF NEW OR INCOMPLETELY DESCRIBED SPECIES OR VARIETIES. By N. H. SwELiteNnGREBEL, Ph.D. and J. M. H. SwWELLENGREBEL DE GRAAF. The following species and varieties have been found in the Malay Archipelago* :— Myzomyia: ludlowi, rossii, vaga (indefinita), flava, minima, minima var. aconita. Neomyzomyia : leucosphyra, punctulata var. tesselata. Cellaa : kochi. Myzorhynchus : sinensis, sinensis var. vanus and var. separatus, barbirostris, barbirostris var. pallidus, albotaeniatus, mauritianus, umbrosus, gigas. Nyssorhynchus: fuliginosus, maculatus, schiiffneri, karwari, jamesi, annulipes var. moluccensis. Stethomyia : aitkenii, aitkenii var. insulae-florum and var. papuae. Comparison with the neighbourmg Malay Peninsula shows the Anopheline fauna to be almost identical in the two countries. The exceptions are :— M. aurirostris, P. watson, M. sinensis var. peditaeniatus, M. albotaeniatus var. montanus, M. pseudumbrosus, M. hunterr and L. asiatica, which are not found in the Archipelago; whereas the Peninsula lacks M. mauritianus, M. gigas, M. barbirostris var. pallidus, S. aitkenw var. insulae-florum and var. papuae, and N. annulipes var. moluccensis. I.—DeEscRIPTION OF NEW SPECIES AND VARIETIES. 1. Myzomyia immaculata, Theo. Myzomyia flava, Swellengrebel (1917).T A yellow mosquito, with light orange-coloured unbanded legs, creamy unbanded palpi, brownish yellow proboscis and creamy wing scales. Wings unspotted. 2. Head scaled like M. rossw, on the occiput broadly expanding upright forked scales, all scales creamy. Palpi unbanded, creamy. Proboscis brownish yellow, labella yellow. Thorax: prothoracic lobes with brownish yellow hairs. Mesonotum with hair-like scales, narrow curved ones in front and a few at the sides, all scales creamy. Scutellum with dark yellow hairs and creamy hair-like scales. Halteres * Contrary to the present custom we adhere (with a few exceptions) to Theobald’s nomenclature, not because we attach to his ‘genera’? any intrinsic value as indicating a reaily existing relationship, but because the division of the Anophelines into genera is a practical measure, expressing the idea present in the mind of every one who studies these insects in nature, that, e.g., M. sinensis is quite distinct from M. aconita, but closely allied to M. barbirostris, and that ‘8. aitkentt i issomething apart from all other Anophelines. By calling all these species ** Anopheles’ one simply causes confusion and not simpliliea- tion. { [On a recent visit to England Dr. Swellengrebel presented to the British Museum a 6 and 2 of his Myzomyia flava. On comparing the latter with the type of Theobald’s A. immaculatus both he and Mr. F. W. Edwards concluded that they were conspecific. The older name immaculatus must therefore be used for this interesting and little known form.—EbD. | 78 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. creamy, light brown at the curve. Abdomen with golden hairs, narrow curved golden scales on genital lobes. Legs unbanded, light orange. Ungues equal and simple. Wangs unspotted; scales narrow, creamy or light yellow. Base of Ist submarginal cell nearer the wing base than that of 2nd posterior ; the cell is longer than its stem, 2nd posterior cell shorter than its stem. Supernumerary and anterior cross-veins meeting, the posterior one more than twice its length from the latter ; 3rd longitudinal vein not ending at the meeting of the supernumerary and anterior cross-veins. Length, 4-5 mm. §. Like 9. Club-shaped end of palpi yellow, owing to the presence of long yellow hairs mixed with the creamy scales ; a dark narrow band caused by partial desquamation. Narrow-curved creamy scales on the two apical abdominal seg- ments. On the wing, anterior cross-vein situated basally from the supernumerary. Larva unknown. JAVA: Soerabaia. SumaTrRA: Mandailing. Rare. Described from 3 99 and 3 J taken at Soerabaia, June 1917. Easily distinguished from other unspotted Anophelines by the long palps (M. brevipalpis) and the broadly expanding upright forked scales on the head (S. aitkeni). The species may perhaps be an albinoid variety of M. vaga (indefimita), as this species sometimes shows a marked decrease of the black portions on the wings and palpi. 2. Nyssorhynchus annulipes var. moluccensis, Swell. (1920). A rather large brown mosquito, with spotted legs, 4-5 light bands on palpi, the apex of the 2nd joint being black; proboscis dark, except labella; narrow curved scales on the whole dorsal surface of the mesonotum ; wings much spotted, resembling those of NV. punctulata var. tesselata. 2. Head scaled as in other Nyssorhyncht. Antennae light. brown, basal segment with small white scales, verticils and tomentum golden. Proboscis dark brown, Fig. 1. Palpi of Nyssorhynchus annulipes var. moluccensis 9, three variations. labella lighter. Palpi (fig. 1): 1st joint black with creamy apex and sometimes a dorsal yellow spot in the middle; 2nd joint black, with a broad white band on apical half and a narrow yellow one at the apex, sometimes a yellow dorsal spot in the middle of the black basal portion, the sub-apical white band sometimes narrow or even absent; penultimate and terminal joints with a narrow black basal band, then a narrow yellow band, the remaining portion creamy. Thorax: prothoracic lobes blackish brown, with black chaetae and a tuft of black scales in front. Meso- notum with a dark spot to the right and left at one-third from the anterior border and one in front of scutellum, covered all over with narrow curved white scales, . LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO, T9 which also are present on scutellum. Pleurae dark, with a few narrow curved white scales. Metanotum light brown, with a dark median line. Halteres white, with the knob dark at the apex. Abdomen dark brown; on apex of 6th and 7th segments narrow curved golden scales, on 8th black ones likewise. Legs dark brown with white spots, many of them forming incomplete rings; 1st tarsal joint banded apically, 2nd—-4th at both ends (except in hind legs, which have an apical band only) ; 5th banded at both ends in front legs, unbanded in the others, in front and mid-legs this joint is white on one side. A sub-apical elongated light spot on the tibiae of front legs. Coxae yellowish brown, with narrow white scales. Wings (fig. 2): ee — ee ~ ae a ce, ot Co aa sae =" leg tea a == i? Fig. 2. Wing of N. annulipes var. moluccensis , 2. on the costa from base to apex 3 short and 4 long black spots and a short one between the 1st and 2nd long spots, these last three extending on to the sub-costa. The 1st-4th long costal spots extend on to the Ist longitudinal in the form of 2, 3-4, 34, small spots and one long one; moreover a small spot below the interval of the Ist and. 2nd and the 2nd and 3rd costal long spots, and 1-3 at the base. Other small spots are present as follows :—3-5 on upper branch, 5-7 on lower branch, 24 on stem of 2nd vein; 5-6 on 3rd vein; 2 apically, one longer basally on upper branch, 2-3 on lower branch of 4th vein, on its stem below the cross-veins a long one, above them 2 long or 1 long and 2 short ones; 4 on the upper branch and lower one, 3 on the stem of the 5th vein; 5-6 on the 6th vein. Fringe dark with bright incisions at the tips of the veins and upward from the 3rd longitudinal vein, except a narrow black spot between the branches of the 2nd vein. Length of 1st submarginal cell more than twice that of stem, its base situated basally from that of 2nd posterior cell. Supernumerary and anterior cross-veins almost meeting, posterior one twice its length from anterior. Length, 4—5°5 mm. 3. Palpi: 1st joint black, mixed with white on one side; articulation with 2nd joint light, owing to absence of scales ; 2nd joint with one side of basal half and top white, the latter white portion fringed with yellow. Club-shaped end white, black on one side, with basal and median black ring, separated from the white by a yellow margin. Length 5-6 mm. Mouuccas. New GUINEA. Described from 592° and 34. Larva (fig. 3). Length 5 mm. Colour greenish brown or dark greyish-green ; head brown, with dark bands and spots; white spots on thorax in front and on abdominal segments iii and vii or ili and vi-viii. Antennae without branched hair. Median clypeal hairs far apart, with a few short hairs ; external clypeal hairs often nearly as long as medians, with long hairs ; posterior clypeal hairs as far apart as medians, reaching to the base of the latter, unbranched or with 2-3 branches. 80 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. External occipital hairs with 4 branches, internal ones with 3 branches or simple. A pair of small fans* on thorax, with 10-14 leaflets, sometimes pigmented, no serrate margin. Small fans on abdominal segment i, with 8-16 leaflets, usually serrate, not pigmented. Complete fans on segments li-vii, with about 18 leaflets, pigmented, serrate near the tip, the latter long and pointed, sometimes ending in a true filament. Shoulder hairs f with thick stem and numerous branches ; median and internal one with stout roots which often unite into one. Fig. 3. Larva of NV. annulipes var. moluccensis: a, b, two types of clypeal hairs; c, occipital hairs; d, shoulder hairs (right); e, f, leaflets. Breeds in all kinds of water, brackish or salt and fresh, running and stagnant, clean and dirty ; also in artificial collections of water (coconut-shells, water in native boats). * Fan = stellate tuft. t+ Shoulder hairs = anterior thoracic hairs. LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO. 81 ' This variety much resembles N. annulipes, Walker, as described by Theobald, but the female palpi are different in the latter, the apical half of the 2nd joint being all white. Also, there is a marked likeness to N. punctulata, Dén., but the proboscis of this species has a light apical half and the black apical ring of the 2nd palpal joint is much narrower. 3. Stethomyia aitkenii var. insulae-florum, Swell. (1920). The imago is not to be distinguished from S. aatkeni. Larva (fig.4). Length5mm. Colour dark green. Antenna with small branched hair. Median clypeal hairs close together, simple, nude; external clypeal hairs simple, nude, length 4 of median ones; posterior clypeal hairs far apart, short, with 3-6 branches ; occipital hairs with 3-6 branches. Shoulder hairs with only the median one inserted on a strong root, carrying about 11 branches, the internal one with about 8. Small fans on thorax and abdominal segment 1, carrying 13-15 Ve Fig. 4. Larva of Stethomyia aitkenit var. insulae-florum : a, clypeus ; b, antenna; ¢c, shoulder hairs (left); d, leaflets. = non-serrate leaflets ; segments ii-vii with complete fans, the leaflets very indistinctly serrate, ending in a long point, no filaments. Breeds in running water in the weeds along the edges, in hilly regions. With the variety papuae it is the only representative of this species in the Australian region of the Archipelago. Only once has it been found in Java (Island of Noesa Kembangan on the South Coast). The larva is somewhat like that of S. culiciformis, James, but the club-shaped antennal hair is wanting and the balancing hairs on abdominal segments 1-11 are of the usual type and number. 4. Stethomyia aitkenii var. papuae, Swell. (1920). Although no adults could be bred from this larval variety, we feel sure that it likewise belongs to this species. It much resembles the larvae of the preceding variety, but is separated from it by the fact of the internal shoulder hairs being very small and carrying two branches only. The leaflets of all the fans show no signs of serration. 82 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. The larvae described here, were full-grown, the pupal hairs already showing through the integument of the 1st abdominal segment. Breeds in the same places as the foregoing variety, but was only found in New Guinea (Kokas, Kaimana, May 1919). 4. Myzorhynchus barbirostris var. pallidus, nov. Adult like M. barbirostris. The relation of this variety to its type is the same as that of M. peditaeniatus, Leic., to M. sinensis, viz., the larvae differ from the typical ones by the external clypeal hairs bearing a smaller number of branches (11-22 against 60 or more in the type) (fig. 5). | Fig. 5. Larva of Myzorhynchus barbirostris var. pallidus, right half.of clypeus. Breeds in slowly running water, in springs in the jungle and in rice-fields, but always in hilly country. It is rare in Java and Sumatra; in the Australian portion of the Archipelago it is as common as the typical form. II. LARVAE OF KNOWN SPECIES NOT YET OR INCOMPLETELY DESCRIBED. 1. Myzorhynchus albotaeniatus (fig. 6). Length 5-5°5 mm. Colour dark rusty brown, lighter on dorsum of 3rd abdominal segment, less conspicuously so on the 6th, the 4th almost black. Antennae more stumpy than in M. sinensis, length : breadth: :1:4 (1:5 in snensis) ; a branched hair on antennae, its stem half the length of the hair (in sen msis 4-1), the branches emitted nearly at right angles (in sinensis at very sharp angles). Median clypeal pairs simple, nude, the distance between them equal to that separating them from the external clypeal hairs (close together in sinensis) ; external clypeal hairs with 18-24 terminal branches; posterior clypeal hairs short, with 3 branches ; occipital hairs with 6-7 branches. Internal shoulder hair very small, with 2-3 short apical branches. Small fans with 12-16 non-serrate leaflets on thorax and abdominal segments i-ii; complete fans on segments iii-vil ; leaflets as in M. sinensis. LIST OF THE ANOPHELINES OF TilF MALAY ARCHIPELAGO. 83 Breeds in slowly running jungle streams with scanty vegetation but much vege- table detritus. Java and Sumatra. 6b “ Fig. 6. Larva of Myzorhynchus albotaeniatus: a, clypeus; 5b, antenna ; c, occipital hairs’; d, shoulder hairs (right); e, variation of inner shoulder hair ; f, leaflet, thoracic fan; g, leaflet, fan on abdominal segment i; h, leaflet, segmentiv ; j, leaflet, segment v. 2. Neomyzomyia leucosphyra (fig. 7). As our own observations differ from those of Stanton (1915) the larva is redescribed here. | Length 4:°5-5°5 mm. Colour light yellowish brown or brownish grey. Antennae without branched hair. Median clypeal hairs long and slender, simple, sometimes carrying a few minute hairs; external clypeal hairs nude and simple; posterior ones short, nude, simple; occipital hairs with 2-3 branches. Shoulder hairs usually pigmented, median and internal ones with strong stems, numerous branches and stout roots, often uniting into one. Small fans on thorax, with 10-11 non- serrate leaflets. On abdominal segment i very small fans with 5-7 hair-like leaflets ; 84 N. H. SWELLENGRELBEL AND J. M. H. SWELLENGREBEL DE GRAAF. on segment ii small fans with 8-11 very narrow leaflets ; on segments iii-vii complete fans with about 18 leaflets each, showing 2-3 indentations near the apex, the latter pointed or rounded, no filament. The leaflets are pigmented throughout. Breeds in stagnant shady freshwater pools. e Fig. 7. Larva of Neomyzomyia leucosphyra: a, clypeus; b, shoulder hairs (left) ; ¢, occipital hairs ; d, leaflet of thoracic fan ; e, leaflet of fan on abd. segment i; f, leaflet on segment ii; g, leaflet on segment iii ; h, leaflet on segment v. 3. Myzorhynchus umbrosus. Addition to Stanton’s description. Another larval type was usually met with in salt or brackish water near the coast, with median clypeal hairs bearing numerous short hairs on the apical half. The adult emerging from this was of the usual umbrosus type. LIST OF THE ANOPHELINES OF TRE MALAY ARCHIPELAGO, 85 4. Myzomyia minima var. aconita (fig. 8). Addition to Stanton’s description : In running water a larval type was common with completely nude median and simple long posterior clypeal hairs. The adult emerging from this larval variety showed the usual characteristics of aconita. The breeding places of the two larval types differ somewhat: out of 644 typical larvae 42 per cent. were found in rice fields, 11 per cent. in running water, 31 per cent. in fish-ponds, 15 per cent. in marshes and | per cent. in dirty stagnant water ; the percentages of 562 specimens of the larval variety found in the same breeding- places were 1°7, 88, 10, 0°3 percent. and mil. Consequently the latter is more adapted to life in running water. Rents eae ie Fig. 8. Myzomyia minima var. aconita, larval variety : a, clypeus ; D, leaflet of fan on abdominal segment Vv. It is to be noted that the larva of the typical MZ. minima (fig. 9) is almost identical with the larval variety of aconita, the only difference we could detect existing in the length of the filaments of the leaflets which is }—3 of that of the whole leaflet (in the larval variety of aconita this relation is usually 35-3). 5. Differences between the Larvae of Nyssorhynchus punctuilata var. tesselata and Cellia kochi. Stanton enumerates them as follows: In N. punctulata there are no fans on abdominal segments 1 and 1, whereas in C. kochi they are present on the latter. We noted in both species on segments 1 and ii small fans with very narrow or hair-like leaflets ; in C. kochi on segment i sometimes a cockade only. Consequently in the Malay Archipelago the fans cannot be used to distinguish these larvae. We use the following characters: In N. punctulata: (a) Antennae usually much pig- mented ; (b) occipital hairs short with 3 or more branches; (c) internal shoulder hairs short, usually with no more than 3 apical branches. In C. kochi: (a) Antennae light ; (6) occipital hairs longer, unbranched or bifur- cated ; (c) internal shoulder hairs long, with numerous long branches inserted at intervals along the stem. 86 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. In the Australian region of the Malay Archipelago, a larval variety of N. punctulata — is commonly met with, bearing fans with non-serrate leaflets, no fans but cockades on the 2nd abdominal segment, and a small fan with hair-like leaflets on the Ist. 2 d te Fig. 9. Larva of Myzomyia minima: a, clypeus ; b, occipital hairs (left); ¢, leaflet from fan on abdominal segmenti; d, leaflet on segment ii ; e, leaflet on segment vi. 6. Difference between the Larvae of Myzorhynchus sinensis and M. barbirostris. In addition to Stanton’s differential characters we noted the following, which however cannot be used if the larvae do not show white spots on the dorsum of the abdomen. But if these are present, they show in M. sinensis on abdominal seg- ments ii, v, vill, in M. barbirostris on the segments iii and vi-vui but never on v. N. annulvpes var. moluccensis has abdominal spots like M. barbirostrvs. 7. Abnormal Larvae of Myzorhynchus sinensis and M. barbirostris (fig. 10). In both species one of the median clypeal hairs may be placed in close proximity to the neighbouring external clypeal hair, at the same time exhibiting on a shortened LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO. 87 stem a number of long apical branches (10 or more), somewhat resembling the ex- ternal clypeal hairs of M. uwmbrosus. Fig. 10. Abnormal clypeus of larva of Myzorhynchus sinensis. 8. Variations in the larvae of Myzomyia ludiowi and M. rossii. As yet we have not succeeded in differentiating these species ; in both we have noted the following variations of the mature forms: (a) The posterior clypeal hairs are of normal length and bifurcated, or much shorter and trifurcated ; (6) the median clypeal hairs are bifurcated ; (c) all clypeal hairs bear 2-5 long apical branches. These variations are infinitely more common in M. rossi than in M. ludlowt. Sometimes we found M. rossit emerging from larvae resembling those of M. vaga (endefinita). 9. Variations of the Larvae of Myzomyia vaga (indefinita). We never found M. vaga with larvae like M. ludlowi, they always adhere to the type first described by Strickland (1915), viz., with the short external and posterior clypeal hairs, the latter placed at a short distance behind the median clypeal hairs and closer together than these. As variations we have noted the bifurcation of the median and posterior clypeal hairs, the latter being placed farther backwards ; also they may be placed close together, the one a little in front of the other. 10. Nyssorhynchus schiffneri (fig. 11). The larva was described by Mangkoe Winoto (1918). They are difficult to distin- guish from those of N. fuliginosus. The following are the characters which separate them : (a) the median clypeal hairs bear more and longer branches ; (}) the posterior clypeal hairs are unbranched or bifid ; (c) the stems and roots are less thick than in N. fuliginosus ; (d) the filaments of the leaflets on the fans are somewhat shorter than in the same organs of N. fuliginosus. 88 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREEEL DE GRAAF, Fig. 11. Larva of Nyssorhynchus schuffneri: a, clypeus ; b, leaflet of fan on abdominal segment v. Ill. QUESTIONS CONCERNING THE VALIDITY, DEFINITION AND NOMENCLATURE OF SOME SPECIES. 1. Nyssorhynchus fuliginosus var. nivipes. This variety is characterised by the followmg points: ‘~) the white area on 2nd joint of the hind tarsus is very extensive, 3-3 of the jot. —_‘b) wing much lighter, with smaller spots, notably on 3rd vein and upper branch or 5th. The same wing markings are present in the male of the typical N. fuliginosus, consequently they cannot be used to separate the males of the type and the variety. 2, Myzomyia minima, and var. aconita. Christophers (1916) separates these two forms on the grounds that the type (qa) has the proboscis in the 9 dark, or with only a light ventral spot near the apex ; (b) lacks a light frmge spot at the apex of the 6th vein ; (c) has the apical half of the 6th vein completely dark; (d) has a long black spot on the base of the 3rd vein. To this we may add that (e) the type shows a broad black band on the middle of the light apex of the palps (in aconita this band is narrow). The two forms seem well separated by these characters, but we have met with otherwise typical aconita exhibiting the pomts (b) and (c), or (a), or (b) (c) (e). Moreover, we have found forms with the points (b) and (c) complete and (qa), (d), (e) intermediate ; 7.e., with a large light patch on the proboscis but ventrally only, a small black patch on the base of the 3rd vein and a band on the apex of the palps in breadth intermediate between that of manima and aconita. On the other hand, we have observed speci- mens bearing a spot on the base of the 3rd vein at least equalling in length 4 of the stem of the lst submarginal cell, but with all the other minima characters present likewise. LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO. 89 As we do not think it advisable to throw these two forms together, we propose a purely artificial but easily observable distinction, viz. :— (a2) No spot on base of 3rd vein... M. aconita, Don. (6) A small spot present (length about } sf iifts Ist sihwiscomnal cell), M. aconita var. cohaesa, Don. (c) A long spot present (from half as long to as long as this cell)... minima, Theo. 3. Myzomyia vaga (indefinita). The species known under the name of M. indefinita, Ludl. (1904) is separated from M. rossit by (a) a narrow black band at the base of the light apex of the female palpi ; (b) a light spot at the apex of the proboscis in the female, behind the labella ; (c) a lighter hee (d) its larger size ; (e) the larva. The points (a), (c) and (d) are subject to great variation, (b) and (e) being constant and reliable ; as (5) is not present in the males and (a) only refers to the females, the males cannot be dis- tinguished with certainty. In 1902 Dénitz described under the name A. vagus a mosquito distinct from although resembling M. rossu. The proboscis of the female is described as being black, with light apex and whitish labella. The palpi of the female are black, the terminal joints white, the penultimate with a broad black band around the base. A distinct variety from Celebes has the penultimate joint of the female palpus white only at the apex. As this is the condition actually existing in M. rossii we infer that the specimen on which the species vagus was founded had a narrower black palpal band. The species was described from a female caught at Fort de Koch (Sumatra) and we know that no M. rossw has ever been found there. From the description of the proboscis and palps of the female (in connection with the description of the variety from Celebes) and the origin of the type we believe that there can hardly be any doubt that Dénitz’s vagus and the species now called A. indefinitus are identical. Christophers (1916) suggests that Giles’ M. rossi might be the one with the broad white apex of the palps. As Giles (1900) states that the apical half of the end joint (2.e., of the two terminal joints taken together) is white in the female, we believe this supposition to be not well founded. On the other hand we believe him to be quite right in suggesting that Ludlow’s indefinita is indeed M. rossii, Giles, as she distinctly states (1904) that the palpal markings and general colour of her new species are like those of M. ludlowz. 4. Difference between Myzomyia ludiowi and M. rossii. The following wing marking is very constant and reliable: on the upper branch of the 5th vein, below the cross-veins, both species show two spots, which however in M. ross (and M. vaga) are short, whereas in M. ludlowi they are much longer. An extensive biometrical research has convinced us that exceptions to this rule are rare (Swellengrebel 1916, Mangkoe Winoto 1918). 5. Neomyzomyia punctulata and N. tesselata (A. deceptor, Don.). One of us has shown already that the characters differentiating these two species, so far as they are based on the wing markings, cannot be accepted as valid. Con- sequently the species are separated only by the palpal markings: in N. tesselata (681) G 90) N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. the second palpal joint of the female is white on the apical half, in N. punctulata it bears a black sub-apical ring, separated however by a narrow light ring from the basal black ring of the 3rd joint. We do not consider this difference of sufficient value to justify a new species and consequently we rank J. tesselata as a variety of N. punctulata. Donitz believed the var. tesselata to be peculiar to the western regions of the Malay Archipelago only, but we have found it as far eastward as Ceram (Moluccas). Dénitz’s typical punctulata we have never met with. The eastern punctulata var. tesselata differs however from the common western type, (a) by the larvae (p. 7.), (b) by the light portion of the proboscis being distinctly diminished in size. 6. Nyssorhynchus schuffneri. Christophers (1916) seems to suggest that this species resembles N. fuliginosus var. adie, James. It is said to differ from it by the very short ultimate joint of the female palpus. We found the length of this jomt to be 11 per cent. of the palpal length, the palpal index being 0°6, consequently it is an orthodactylous mosquito. But there are other, more striking differences. The palps have 3 white bands only (4 in adier), with a long white tip resembling M. vaga. The wing, which in the var. . adiei is like that in N. fuliginosus (judging by James’ description), is much lighter in schiiffnert ; the black spots on the costa are rather narrow, on 2nd longitudinal _ vein one small spot under the 3rd costal spot (reckoned from the apex). Stem of 4th longitudinal vein nearly all yellow, so is the 3rd vein (in adie it is black, except for a few minute light spots). IV. GEOGRAPHICAL DISTRIBUTION. A considerable portion of the Malay Archipelago has been searched for Anophelines, and although much remains to be done, we wish to draw attention to the following striking features in the geographical distribution of certain species and varieties. 1. Nyssorhynchus annulipes var. moluccensis. Wherever it occurs this is a very common species, breeding everywhere, and so it is difficult to overlook it. Still it has never been found in Sumatra, Java, Borneo, Celebes and the smaller islands of the western portion of the Archipelago, but it is the commonest Anopheline in the Moluccas and New Guinea. It is closely allied to the Australian N. annulipes and its distribution is in accordance with other zoological findings, showing the fauna of the eastern Archipelago to be mainly Australian. It is well known that Wallace divided the Archipelago into an Asiatic and an Australian region by a line following the Straits of Macassar and of Lombok. The Anophelines (and with them other animals ; cf. Weber 1902) do not conform to this scheme, as the Anopheline fauna of Celebes, so far as we know at present, is purely Asiatic. 2. Nyssorhynchus schuffneri. So far as we yet know, this species is confined to the western provinces of Java (Batavia, Bantam) and the neighbouring Lampong districts of South Sumatra, which are separated from the former by the Soenda Strait. LIST OF THE ANOPHELINES OF THE MALAY ARCHIPELAGO. 91 3. Myzomyia rossii. Christophers (1916) states that M. rossw, Giles, is the common western form, whilst M. vaga, Don. (indefinita, Ludl.) occurs throughout Malaya. To this should be added that M. rossii does so likewise, as far as the Moluccas ; only in Sumatra it has not been found, up till now. 4. Myzomyia !udlowi and Myzorhynchus sinensis. We failed to find both species east of Celebes. But we know that in districts where M. ludlowi usually is of common occurrence, it may be absent for months and even for a whole year (Citroen 1917). As suitable breeding places were present, likewise its common satellite M. rossi, we are not sure that M. ludlowi is really absent from the Australian regions of the Archipelago. 5. Stethomyia aitkenii The form with the typical larvae—including all variations described by Stanton (1915) and Stickland (1915)—was not found in the Australian region of the Archi- pelago. The species was represented there by the varieties insulae-florum and papuae, which in their turn were absent from the western regions, except var. insulae-florum, which occurs on the south coast of Java (Island of Noesa Kembangan). The following table gives a summary of our present knowledge of the distribution of the Anophelines in the Malay Archipelago. — Ea ae] = a 3 < © . = ‘a = = 5 = ra ee 2 ee 5 > Pe ies zisielnisias| sei Si/./ els esi sls slofgisisisis zlels leas VISES Lek | se PS iS i ee | ess is] Si sie s (els) slsell sparsely Sir letetsiecls. |e, siealsice sles Stelstele less etsise4ls STE S| 5] °S/ 8218/7] 42 Si S818] 8] 8 SST S/S Sle S| Se] ee] § < ear ys aiS] siSieit] Jerseys 2 =) Sumatra = Sears cere a See te eile aay —— (hy } — Java .. vs Sa ee ee se eee tle — 1474] + Borneo, incl. findings | ' ! ily | ' of Roper (1914) ../—|—/+}/+/+] + | + /+]/+] + }+/— |4+/-—/+ ne 5-3} 3: | Celebes a »-(—|—-T-]+]4+] + J + J4+]—] + [4+}/— a8 Moluccas and New Guinea as --}—||—} +4. J] + | + KH — /+} + i ce ee ee ee ee Riouw and Linga Archipelagos ff —|—|— | +1 — |— sian Himae es banal ep Es ee ee eo References. CiTROEN (1917). Geneesk. Tijdschr. Ned. Indié, lvii, p. 763. CHRISTOPHERS (1916). Ind. Jl. Med. Research, i, p. 454. _ Donirz (1902). Zeitschr. f. Hyg. u. Infectionskrankh. xli, p. 15. Gries (1900). A Handbook of Gnats or Mosquitoes. London, 374 pp. Lup.tow (1904). The Canadian Entomologist, xxxvi, p. 297. Mane@kor WinorTo (1918). Geneesk. Tijdschr. Ned. Indié, lvi, p. 462. STANTON (1915). Bull. Entomol. Research, vi, p. 159. STRICKLAND (1915). Ibid. v, p. 321. z (1915). Short key to the Identification of the Larvae of the common Anopheline Mosquitoes of the Malay Peninsula. Kuala Lumpur ; 18 pp. 92 N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL DE GRAAF. SWELLENGREBEL (1916). De Anophelinen van Nederlandsch Oost-Indié. Amsterdam, 182 pp. Weser (1902). Die Indo-Australische Archipel und die Geschichte seiner Tierwelt. Jena, 46 pp. 93 COLLECTIONS RECEIVED. The under-mentioned collections were received by the Imperial Bureau of Entomology between Ist January and 3lst March, 1920, and the thanks of the Managing Committee are tendered to the contributors for their kind assistance :— Dr. J. H. Ashworth, F.R.S. :—2 Chironomidae, 16 Culicidae, 6 Tabanus, and 17 Glossina ; from the Sudan. Surg.-Comm. E. L. Atkinson :—4 Tabanus and 3 larvae of Musca domestica. Mr. EK. Ballard, Government Entomologist :—46 Culicidae, 478 other Diptera, 28 Coleoptera, and 47 Rhynchota ; from Madras. Capt. P. J. Barraud :—49 Culicidae, 29 other Diptera, 8 Hymenoptera, 377 Coleoptera, 36 Lepidoptera, and 45 Rhynchota ; from Palestine. Mr. G. E. Bodkin, Government Economic Biologist :—6 Tabanidae, 37 other Diptera, 350 Hymenoptera, 34 Coleoptera, 100 Isoptera, 1 species of Coccidae, 51 other Rhynchota, 52 Spiders, 3 Centipedes, and 2 Millipedes; from British Guiana. : Mr. P. A. Buxton :—14 Trypetidae, 5 Weevils from Cotton, and 1 Bug; from Mesopotamia. Mr. J. B. Corporaal, Entomologist to the Algemeen Proefstation :—136 Rhynchota ; from Sumatra. Division of Entomology, Pretoria :—14 Diptera, 189 Hymenoptera, and 7 Coleoptera ; from South Africa. Government Entomologist, Sudan :—3 Diptera, 11 Hymenoptera, 10 Lepidoptera, and 5 Rhynchota ; from the Anglo-Kgyptian Sudan. Mr. G. F. Hill, Entomologist, Australian Institute of Tropical Medicine :— 125 Culicidae, 13 Tabanidae, 24 other Diptera, 10 Termites, and 5 species of Coccidae ; from Queensland. Imperial Department of Agriculture, West Indies :—2 Nests of the Wasp, Polvstes annularis and 6 Moths bred from them ; from Barbados. : Dr. A. Ingram :—87 Culicidae, 1 Tabanus, 2 Glossina, 15 other Diptera, 1 Moth, 1 Cockroach, and 2 May-flies ; from the Gold Coast. Mr. Nigel K. Jardine :—73 Parasitic Hymenoptera ; from Ceylon. Mr. Ll. Lloyd :—Collections of the Tomato Moth, Polia oleracea, L., and the Aleurodid, Asterochiton vaporariorum, Westw ; from Hertfordshire. Mr. A. Loveridge, Curator, East Africa and Uganda Natural History Society :— 27 Culicidae, 30 Tabanidae, 20 Glossina, 344 other Diptera, 37 Dipterous pupa cases, 834 Ants and 37 pupa cases, 6 other Hymenoptera, 1 Caddis-fly, 1 species of Coccidae, 1,497 other Rhynchota, 2 Orthoptera, and 228 Odonata ; from British East Africa and Tanganyika Territory. Mrs. W. Smith :—66 Culicidae, 6 Tabanus, 1 Glossina, 50 other Diptera, 21 Hymenoptera, 23 Coleoptera, 2 Lepidoptera, 2 Caddis-flies, 41 Rhynchota, 6 Orthoptera, and 1 Tick ; from the Gold Coast. Mr. F. V. Theobald :—2 Chalcids (Pteromalus egreqius, Forst.), bred from the Brown-tail Moth ; from Romney, Kent, 94 COLLECTIONS RECEIVED. Mr. Robert Veitch :—4 Diptera, 2 Hymenoptera, 3 Coleoptera, and 21 Lepidoptera ; from Fiji. Mr. Morris N. Watt :—7 Agromyzid Diptera; from New Zealand. Wellcome Bureau of Scientific Research :—9 Culicidae, 39 Tabanidae, 1 Hippo- boscid, and 23 other Diptera ; from Macedonia and Africa. Mr. C. B. Williams :—17 Scolytidae ; from Trinidad. Mr. Rodney ©. Wood :—4 Tabanidae, 1 Hippoboscid, 77 other Diptera, 165 Coleoptera, and 2 Rhynchota ; from Nyasaland. a IMPERIAL BUREAU OF ENTOMOLOGY. — HONORARY COMMITTEE OF MANAGEMENT . THE VISCOUNT HARCOURT, Chairman. Isgvt.-Coronen A. ALCOOK, OLE., F.RS. | Masor KE. E. AUSTEN, D.S.0. Sol ee Dr. A. G. BAGSHAWE, O.M.G. | Be Be a sd Masor-GENERAL Sir JOHN R. BRADFORD, K.C. M. G., F.R.S. Masor-GENERAL Sr DAVID BRUCE, K. O:B., F.R.S. Mr. J. CO. F.. FRYER. Sir §. F. HARMER, K.B.E., ERS. ‘Pror. H. MAXWELL LEFROY. Hon. E. LUCAS. Dr. R. STEWART MACDOUGALL. Sm JOHN McFADYEAN, Fe , Sir PATRICK MANSON, G.O.M.G., E B.S. Siz DANIEL MORRIS, K.0.M.G. Pror. R. NEWSTEAD, F.R.S.- . Pror. G. H. F- NUTTALL; ¥.RB:8. ~ Pror. E. B. POU: E.R.S. aS Lrzvt,-CoLonet Sir *DAVID PRAIN, OMG. CLE. ERS Sir H. J. READ, K.O.M.G., O.B. Ta Hon. N. OC. ROTHSOHILD. Dr. HUGH SCOTT. 3 | | Sim A. E. SHIPLEY, G.B.E., ERS. 5 — Mr. R. A. 0. SPERLING, O.M.G. se ae Sm STEWART STOCKMAN. | Mr. FE. V. THEOBALD. ee ee Me, C. WARBURTON. &. | - . Director. Dr. GUY A. K. MARSHALL, OMG. eee aS “Bssistant Director. SS Rig ote ‘Dr. S. As NEAV Bie oo eee es Secretary, ee S ‘e Mr. Ba: 0:9, PARKINSON. coe - ~~ \ " . a“ ie 95 OBSERVATIONS ON COTTON THRIPS IN THE GEZIRA, BLUE NILE PROVINCE, SUDAN, IN 1918-19. By G. H. Corsett, B.Sc. From conversations that the writer has had, it would appear that this thrips {identified by Mr. R. 8. Bagnall as Heliothrips indicus, Bagn.) has been known on the Gezira cotton for about three years, and is popularly called “‘ dry asal”’ in con- tradistinction to the work of the “asal fly’ (Aphis sorghi, Theo.) The natives have been acquainted with this insect for some time. The reason suggested for the enormous damage done by it is that the cotton is too scattered, and it is stated that if cotton were cultivated in larger blocks, the thrips might be viewed with equanimity. The cotton land at Barakat for this year (1919-20) has been, as far as possible, prepared in larger areas, and it will be interesting to compare the result of this arrangement with cultivation at Tayiba, where no alteration in grouping has taken place. Towards the end of October 1918 the thrips attracted serious attention ; by the middle of November the cotton looked “ all withered up,” but by the middle of December it began to recover. It was not until the beginning of February that work on this insect was commenced—most of it being done in the field whilst preparing a laboratory to conduct experiments under more immediate control. At that time the cotton had practically recovered, in spite of the fact that the thrips was present in large numbers. In most cases the infested cotton was first noticed on the north side of a “‘ hosa,”’ and this implies that the thrips either came from weeds growing on fallow land on that side, or was brought froma distance by the prevailing wind, and finding cotton and conditions favourable, propagated to such an extent that the immediate study of the bionomics and control of this insect was considered necessary. The occurrence of this minute insect in enormous numbers caused, during the season 1918-19, a considerable reduction in the yield of cotton, and it must at present be viewed as a great menace to cotton cultivation in the Sudan, though it may prove to be quite local in its attack and occur only spasmodically. In any case, at a very conservative estimate, an average loss of one and a half kantars of cotton per feddan has occurred during the past season on well supervised and cultivated land. This is the most important pest with which the cultivators of cotton in the Gezira have to contend. On account of its smallness, the rapidity of its spread, and the seeming suddenness of its attack in large numbers, the control of the insect is one which might present very great difficulties. If the thrips is found on cotton when young, it is probably practicable as well as economical to spray ; but on the other hand, if 1t comes suddenly when the cotton plants are-about five months old, the writer considers the policy of spraying doubtful. The injury caused to the cotton plant is due to the larvae and adults feeding on both the upper and lower surfaces of the leaves. By piercing the epidermis of the leaf and removing the sap, they lower the vitality of the plant. At the points of attack spots are formed which, often running together, produce distinct white streaks (687) Wt.P2/154. 1,000. 8.20 B.&F.,Ltd. Gp. 11. A 96 G. H. CORBETT. where the chlorophyll has been removed ; and later, if feeding is continued, the lower surfaces of the leaves acquire a silvery appearance. As the feeding of this insect is prolonged, the leaves become brownish and tough, owing to the efforts of the plant to repair the injury. In severe infestations the leaves wither up and fall to the ground, preventing the young bolls from developing. Leaves frequently present a distorted appearance, curling inwards, and in the folds the larvae are often abundant. Larvae have been observed neither on the bolls nor in the flowers, but are occasionally seen on the bracts. On plants grown in the laboratory, larvae have been noticed feeding on the stem and root just beneath the surface of the soil. This undoubtedly was due to the leaves drying up and the thrips. migrating for food. In no case was the larval stage completed in such situations. Though this insect was first observed as attacking fully-grown cotton trees, it may be mentioned that the youngest cotton plants are not immune. For egg-laying, preference is shown by the adults for the lower leaves, and then the middle leaves. On young succulent leaves at the upper portion of the plant, the thrips are never so abundant as they are on the much eaten and tough lower leaves. Habits of the Adult. The usual mode of progression of the adult is walking, though when disturbed it runs, and frequently takes long quick jumps. Flight has been observed in a number of cases, but this is not general. Emerging from the pupae in the ground, the adult. crawls up the stem to the leaves. In breeding cages it has been noticed to climb up the side of the chimney, and when on a level with the leaves to take a leap to the plant ; this motion is aided by the wings. By far the larger number of the adult thrips feed on and lay their eggs in the tissue of the lower surface of the leaf, and confine their attention for the most part to the lower and then the middle leaves of the plant. Flowers seem to be immune from attack, but larvae have been found on the bracts. Very often the adults are found resting and feeding along the veins of the leaves, but, as a rule, no particular part of the leaf is preferred. Adults enclosed in vials with plugs of cottonwool die rapidly, but under more natural conditions they have lived for a maximum of fifteen days, though the average works out at five days. The males and females emerge from the pupae about the same time, but there are indications that the pupal stage is slightly longer in the males. than in the females. The female is capable of laying eggs on the second day after emergence. On an average, six eggs each day are laid in the tissue of any part of the leaf, but though eggs have not been observed, the hatching larvae are more usually found along the veins. Habits of the Larva. The young larva, on hatching, looks like a starchy-coloured segmented worm, with red eyes and with no indications of legs or antennae. Gradually it works itself out of the leaf by a slight swaying motion, and standing as it were on its anal segment, unfolds antennae and legs. Later, bringing its legs to the leaf surface, it pulls itself free. OBSERVATIONS ON COTTON THRIPS. 97 Sometimes it moves only two or three centimetres, but at other times greater distances, before commencing to feed. As it imbibes the sap of the plant, the body takes on a yellowish hue, and usually by the second day the reddish bands, so very characteristic of all stages except the adult, begin to develop. Only one moult has been observed—the change from the larva to the prepupa in the soil. Like the adult, the larva feeds by piercing and sucking the juices of the plant. It prefers the lower surface of the leaf, though it does not by any means confine itself there. Movement from one surface of the leaf to the other is common, but from one leaf to another is exceptional. Larvae are often found segregated along the veins of the leaf, and occasionally in colonies on other parts of the leaf. Like many species of thrips, the larva carries a globule of brownish black excreta, (on the first day the excreta are colourless), which, after attaining a considerable size and dropping on the leaf, dries to a blackish spot. These spots discolour the leaf, and their presence is a sure indication that the plant has been, or is being, attacked by the thrips. This globule of excreta is held off the surface of the leaf owing to the anal segments of the larva being slightly turned upwards. The larvae when full grown drop from the leaf, and, crawling into a convenient crack in the soil, there change to prepupae. They prefer damp earth, and the prepupae are usually found at the junction of the damp and dry soil, which varies from about three to six inches down. Immature larvae falling to the ground will die if they cannot find food in the nature of weeds or crawl back to the plant, a very doubtful possibility. It may be noted that practically all weeds growing amongst the cotton support larvae. If the land is kept clean [whilst the cotton crop is growing, this will not only prevent the larvae from maturing, but will make the adults seek other plants for oviposition. According to experiments, the adults do not live more than a day without food. The larvae usually like to secrete themselves away from the sun, and are often found in large numbers in the curled up parts of the leaf. Larvae in the soil do not feed, and, if disturbed, move more quickly than on the leaf. Habits of the Prepupa and Pupa. The prepupae and pupae are found in the soil, though under unnatural conditions they have been observed on four occasions on the leaves. This peculiarity was not seen in the field. The larva, before prepupating, does not appear to make any cell, but crawls into any convenient place in the soil. The cast skins of the larvae and prepupae are found together, indicating that no movement of prepupae and pupae takes place. Both stages are capable of movement when disturbed, and are often seen in groups, and at other times singly or in pairs. The want of moisture in the soil does not have any deleterious effect, but from experiments ‘and observations it would appear that if, after the larvae have crawled into the soil, the soil is kept moist and not allowed to crack, the adult is unable to emerge. The prepupae and pupae possess red bands like those of the larvae ; but whilst in the prepupae the antennae are carried forward in front of the head and the wing-lobes are partially formed, the antennae of the pupae are bent backwards over the head and the wing-lobes are well developed. The darkening of the body of the pupae previous to the emergence of the adults is very evident, (687) a2 98 G. H. CORBETT. Food-Plants. This thrips seems to feed on most plants, cultivated or wild, and below is a list of the plants on which the larvae have been okserved. In course of time, no doubt, other plants will be found to sustain this insect, and attention may be called to the wide range of botanical orders represented in the list. Mr. R. E. Massey, Government Botanist, kindly identified most of the wild plants. Cultwwated Plants. Potatoes, field beans, French beans, butter beans, ground-nuts, wheat, barley, peas, turnips, cabbages, cowpeas, bamia, beetroot, sweet potato, lettuce, carrot, clover and cotton, both American and Egyptian. It has been reported as attacking Lubia (Lubia afin) and Dura, but this the writer has not been able to confirm. American cotton suffers very much more than the Egyptian, and whilst Eygptian cotton recovers remarkably after a severe infestation, in the American variety the percentage of recovery is much smaller. Wild Plants. Malata, the most common plant in the Gezira, generally found with thrips. Leucas nubica, Benth. (ungalot) ; very common, and always supporting thrips. Phyllanthus mruri, Linn. Tephrosia emeroides, Linn. (surep). Cucumis melo, Linn. Corchorus olitorius Linn. (molokhia). Digera arvensis, Forsk. (heyrayrah). Solanum incanum, L nn. (gibbein). Ipomoea cordofana, Choisy (hantut or tubba). Heliotropium europaeum, Linn. Heliotropvum supinum, Linn. Ocomum basilicum, Linn. (rihan). Polygala triflora, Linn. Crotalaria sp. Abutilon sp. (hambook). Abutilon sp. (saphirah). Rhyncosia memnonia, Linn. (myoda). The local native names for most of the weeds are given in brackets. Life-Cycle. Observations were made on this insect only from the month of‘November to the beginning of August. It is doubtful if there is any aestivating period, and the writer is inclined to think that there is no resting stage though, after the rains commenced considerable search was needed to find specimens of the insect. The differences in the lengths of each stage are remarkable, but under absolute field conditions the life-cycle is a short one, and the thrips quickly reproduce in large numbers. OBSERVATIONS ON COTTON THRIPS. 99 As the result of a number of observations and experiments it was found that the length of the egg stage varied from 4 to 15 days, with an average of 8°2. In February the minimum and maximum were, respectively, 9 and 15 days; in March, 6 and 11; and in April, 4 and 9. The period during which the larva remained on the plant varied from 3 to 6 days, with an average of 4:2. From 4 to 14 days elapsed between the entrance of larvae into the soil and the emergence of the adults, the greatest number of emergences taking place on the sixth day. From this it will be seen that the developmental cycle may be as short as 11 days, or may take as long as 35 days; but the writer is of opinion that 18 days may be regarded as the usual time, and the adults do not begin egg-laying until two days after their emergence from the soil. Control Measures. In discussing the control of this thrips, it should be borne in mind that its be- haviour after the old cotton crop is removed till the new one is well advanced is not yet precisely known. Larvae were difficult to find on wild plants at the beginning of August, but until the rains commenced they were numerous. Control may be considered under three heads, namely, suggestions, cultural methods, and insecticide applications. 1. Suggestions. (a). If this insect spends the “ dead season” on vegetation, all plants of every description should be removed ; but if, on the other hand, it rests in the soil, the ploughing of the old cotton land should take place as soon as convenient so as to expose the aestivating stage. (6). Larvae and adults are very numerous on the last standing cotton, and the writer suggests that, since there is so little to be gained and so much to lose, the watering of the cotton should cease slightly earlier and the cotton should be cut out and burnt sooner; by this means the number of thrips that could attack the suc- ceeding cotton crop would be greatly diminished. (c). Observations tend to show that heavy waterings have a detrimental effect on the numbers of thrips emerging from the soil. The probable reason for this is to be found in the fact that the soil does not crack so quickly, and the adults, unable to _ escape from the ground, succumb. The effect of irrigation on this thrips is a line on which further investigations are required. It is suggested that one or two heavy waterings about the beginning of November might have a very beneficial effect. (d). Owing to the wide range of food-plants, experiments with trap crops should be tried. 2. Cultural Methods. The cultivators of cotton in the Gezira appreciate the extent of the damage this insect can do, and the destruction of all weeds should be the first preventive measure to be adopted. Practically every plant in the vicinity of cotton has been found to harbour the larvae, and all vegetation growing amongst the cotton or in its vicinity should therefore be eradicated. This not only applies to the growing season, but is equally applicable to land which will be under cotton the following year, and to the banks of all canals and gadwels. > 100 G. H. CORBETT. It should always be borne in mind that clean cultivation not only prevents, in» many cases, the completion of the life-history of this insect, but also yields other advantages to the crops. 3. Insecticide Applications. Four insecticides were experimented with, namely, pyridine sulphate, quinoline sulphate, nicotine sulphate, and soap. Pyridine sulphate was found unsuccessful. Quinoline sulphate possesses insecticidal properties, but the fact that it had to be used in strong solution, and therefore in large quantities, rendered it unsuitable. Nicotine sulphate was tried with soap (“‘ Sunlight ’’) in varying strengths, and the best solution found was one part of 32 per cent. nicotine sulphate to 1,200 parts of the soap solution, which was obtained by dissolving 1 lb. of the soap in 21 gallons of water. Soap (“Sunlight’’) spray, consisting of one pound of soap to ten gallons of water, will kill a large percentage of thrips, both larvae and adults. Soap alone is only recommended if nicotine sulphate is unobtainable. It should be stated that spraying is considered impracticable, as well as unecono- mical, when the cotton is four or five months old; but if the thrips should appear when the plant is small, the nicotine sulphate spray with soap will be found very useful. Spraying should be done in the late afternoon and evening, and at least twice, so as to kill the larvae which have hatched from eggs in the leaves since the first spraying. 101 SOME FURTHER NOTES ON THE TSETSE-FLIES OF NYASALAND. By W. A. Lamporn, M.R.C.S., L.R.C.P. | Glossina morsitans. A fortunate stay of a few days in July 1919 at Fort Johnston, Lake Nyasa, while waiting instructions, enabled the writer to proceed to the village of Mzeze near the Livingstonia Peninsula, where in 1914-15 some work was carried out on the various insects parasitic on the puparia of this fly. The success at that time of the various parasites, especially Mutilla glossinae, Turner, as judged by the numbers of their cocoons found in old puparia of the fly and bred from recent ones, had been phenomenal, and so the chance, though one lasting only a few hours, of ascertaining the progress made by these insects in the course of the three subsequent years was a very welcome one. The early rainy season was quite similar in character to that of 1915, when long series of the flies had readily been captured in the district, on passing through which one had been constantly harassed by a buzzing swarm. Far different was one’s present experience, for though the journey was made on a cycle, which attracts more flies than a pedestrian, none at all appeared, and it was only on the following day that a single specimen was taken in the course of a hunt for puparia. Local natives agreed as to the diminution in numbers of the flies thereabouts without there being any falling off of game, and added that the pests had now become more numerous to the north, a statement substantiated by personal experience a few weeks later on. A gang of six boys, who had on a previous occasion worked as puparia collectors and were therefore experts, were got together to assist again in collecting puparia, and the results were confirmatory ; for though two days were devoted to the search, only 107 old puparia and 15 new ones were forthcoming, mostly from the artificial breeding places designed in 1915. Examination of the old puparia showed that nine definitely had been destroyed by Mutillid parasites, but it was not possible to determine with certainty the full extent of the parasitism either by this species or others, owing to the age and broken condition of many of the puparia. Of the fifteen new puparia three only afforded flies, a male on 23rd August, a female on lst September, and another female on the 4th. Five yielded Mutilla glossinae, two males on 17th September, one male on 18th, and one of each sex as long after as 29th. The remaining puparia all died, possibly as a result of the vicissitudes of temperature to which they were exposed, for they were taken on a month’s foot journey in the uplands of Angoniland, at an average elevation of from 3,000 to 4,000 feet, hot days being generally succeeded by chilly nights. Whereas in 1915 at Mzeze parasites would be readily obtained, at Domira Bay none at all were either captured or bred out, this being about fifty miles to the north and separated from the Mzeze area by the south-west arm of Lake Nyasa. No opportunity of visiting the area at Domira Bay occurred, but in September 1919 a visit was made to another section of it, at Kotakota, about forty miles north of Domira Bay, where between the 6th and 10th of the month hundreds of old G. morsitans puparia, and sixty-five new ones were obtained. None of these old pupa-cases showed the Mutillid cocoons, nor did any of these parasites emerge from 102 WwW. A. LAMBORN. the new puparia, of which twenty-six yielded flies (twelve males and fourteen females) during late September and the first week of October, the remainder dying. The morsitans area in the neighbourhood of the Livingstonia Peninsula and running inland for miles, so as probably to be continuous with the great area running down the west bank of the bed of the former Shire River, had undoubtedly been infested with these flies for many years, various natives, well on in years, stating that they remember them even when they were children ; whereas the area running from Domira Bay up north to Kotakota, and beyond it, and south towards Dedza, in which the fly is far more abundant, has undoubtedly been more recently occupied by them; for many of the younger Government officials remember when large herds of cattle grazed on its wide dambos, now entirely destitute of all domestic animals, though there are large settlements of natives. The evidence would seem to suggest that the abundance of the fly thereabouts may be due to the failure of the natural parasites to keep equal pace with the fly in the migration which has occurred. Glossina brevipalpis. This species was found by the writer in the same area with morsitans in the course of a journey through a stretch of twenty miles of country between the mouth of the Bua River, twelve miles north of Kotakota, and that of the Duangwa, in every ravine and hollow where large shade trees were associated with heavy undergrowth. As showing how readily the presence of this fly may be overlooked it may be remarked that on the outward journey not one was seen, and that it was only on the return that the capture of two males in the early morning suggested a search. Many were then taken at dusk, and the numbers of the puparia obtained will show how numerous the flies really must have been. On the 12th of the month a systematic search for their breeding places was com- menced. A little scratching with a bit of stick in the loose sandy soil, much lightened by humus, among the herbaceous vegetation beneath a huge shade tree soon brought to light twenty pupa-cases ; more were found during the pitching of the tent, anda _ gang of raw natives who had been attentive to the mode of search were then put to the work. Within four hours the most industrious man had got together 390 puparia (old and new), two others bringing in 267 and 257 respectively, and between the 12th and 15th a total of no less than 13,838 old puparia, with 248 new ones, were found by them. At the same time in these breeding places only 23 puparia (17 old and 6 new) of G. morsitans were found, showing how different are the breeding places selected, for morsitans was present in swarms. Twelve of the old puparia of G. brevipalpis were whole except for a small puncture _which had probably given exit to parasites, though neither on this occasion, nor when the puparia were previously studied, were any of these obtained. The perforations were of two kinds, one about the size of the head of a small pin and situated usually midway between the two poles, and the other a much larger breach less regularly rounded and more commonly towards one or other pole. Twenty- eight only of the new puparia yielded flies (13 males and 15 females), and all the rest perished. All these flies came out an hour or two before dusk, except five, all males which emerged, four at about 10 a.m., one at 2 p.m. NOTES ON TSETSE-FLIES OF NYASALAND. 103 On two successive nights camp was pitched in localities where these pupae were being obtained with a view to ascertaining whether brevipalpis is nocturnal as well as crepuscular in its habits. Though 32, all males, were taken at dusk, between 6 and 6.15 p.m., all within 50 yards of the tent, two only were taken inside it. None were taken after dusk, though the nights were bright and starlit, nor were they attracted to the camp lamps. At about 8 p.m. a search was made in the bush by the light of a bullseye lantern, but again the flies were neither seen nor heard ; and though the party of 23 natives slept out in the open, few having blankets, none complained of having been troubled by the fly. The breeding grounds of this fly were further studied in October, some two hundred miles north in the vicinity of the Lufira River, North Nyasa, close to the Lake, a locality long notorious for the fly, the places selected by it having the same character as all the breeding places previously examined. In this neighbourhood 13 new puparia and 9,094 old ones were collected between the 7th and 18th of the month, only 14 of which showed perforations suggesting the exit of parasites. In the course of the journey through the fly-infested villages north of Kotakota the method was witnessed by which the natives of that part, while sitting down, catch, with the least effort to themselves, flies settled within reach on their bodies. The blade of an old sheath-knife or an old spear blade is placed almost flat on the skin and is then advanced, with the edge slightly raised, by no means slowly or even very carefully, towards the fly, which, though very alert in regard to menace from above, evinces no apprehension of that below it. When the edge is well over its feet the native presses them on to his skin with it, and the fly, so trapped, is then taken in the fingers and subjected to treatment such as is doubtless considered justifiable by reason of the annoyance it has caused. © Suggestions in regard to Tsetse Control. One of the most pressing entomological problems at the present day in the British Empire is that of the control of the various species of tsetse-flies. As shown by a number of workers, the parasites destructive to their puparia are not few, at all events in the case of G. morsitans, and in some areas these parasites are known to exert a considerable influence in reducing the numbers of the fly. But, generally speaking, no very material reduction, from the point of view of man, would seem to be effected by these agencies; though at this early stage of research in regard to the fly it may perhaps be premature to make any unqualified statement as to their control value. Instances have been recorded where the tsetse-fly in a district has unaccountably diminished without there being any marked diminution in the numbers of the game animals therein. Major E. E. Austen in his ‘‘ Handbook of the Tsetse- flies ’’ (1911, p. 65) mentions the disappearance of G'. morsitans from the Victoria Falls, where at one time it abounded. The late Captain F. C. Selous, in his book “ African Nature Notes,” speaks of the disappearance of the flies at the same time with buffalos from the valley of the central Limpopo and its tributaries, where other game—kudu, zebra, wildebeeste, hartebeeste, impala and bushbuck— continued to exist in considerable numbers, and he suggests that the flies died out because they were unable to maintain themselves on game other than buffaloes. More recent study of the flies has shown that this is not the case, and the writer 104 W. A. LAMBORN. suggests that a more reasonable explanation is to be found in the local extermination of the flies by their parasites, a process, as he believes, now steadily proceeding in the area near Mzeze, Nyasaland. The question therefore naturally arises—how far is it possible to increase their influence either by the introduction of new parasites or by breeding on a larger scale those already known to science ? The genus Glossina being now limited to the Ethiopian region, it is doubtful how far the parasites of other Diptera, brought in from other lands, would seek out and destroy its puparia. It is to be apprehended that those obtained from the puparia of other Muscids, breeding under more or less similar conditions, might doso. A more hopeful line of action could possibly be found in the search for other species naturally parasitic on Glossina but having a different geographical distribution. For instance a Bombylid fly, Villa lloydi, Austen, parasitic on G. morsitans, has so far only been discovered in Rhodesia, and a second species of parasitic Mutillid, MW. benefactriz, Turner, in Nyasaland only. ‘The parasites of the fly on the West Coast are probably different from those in the East and South, the insect fauna, generally speaking, being largely distinct. The various species of tsetse-flies may each have different parasites yet to be discovered, which might be interchangeable. It is probable, and some of the evidence on this point is very definite, that in the case of rapid extension of a fly area, parasites less endowed than their hosts with the power of ranging far and wide—strength on the wing having doubtless played a considerable part in the present-day success of the tsetse as an insect—have lagged more or less behind in the foci first inhabited by the flies. And this is parti- cularly likely to be the case with Mutillids, the females of which are wingless. One of the present suggestions in connection with fly control is to endeavour to enhance the value of the parasites already known to science, the Chalcids in parti- cular. They could unquestionably be bred readily enough in the laboratory on the puparia of their natural host, though owing to the difficulty im obtaining these in any numbers, the laboratory output would necessarily be very limited. It has occurred to the writer that it might be possible to employ as alternative hosts the puparia of some of the common Muscids, then liberating in great numbers the little insects, each with its dominant aim in life to seek out and destroy the greatest of the insect scourges of man and beast in the African Tropics. The general upset of work as a consequence of war conditions made any attempt to put the idea to a practical test impossible, but an opportunity was made to carry out, during 1918 in Kast Africa and during the last six months of 1919 in Nyasaland, research work in regard to the question of alternative hosts for Chalcids normally parasitic on certain Muscids. The results (which will be written up later) showed some promise, though this work has again, unfortunately, been brought to a prema- ture close. 105 THE EARLY STAGES OF WEST AFRICAN MOSQUITOS. V.—CULEX DECENS, THEO. AND CULEX INVIDIOSUS, THEO. By J. W. S. Maorie and A. InerRam, West African Medical Service. Both Culex decens and Culex wnvidiosus are widely distributed in British West Africa. In the Gold Coast both have been taken in all the three divisions into which the country is divided, namely, the Colony proper, Ashanti and the Northern Territories ; the records at Accra showing the following distribution :— Culex decens. Colony—Accra, Cape Coast Castle. Ashanti—Kintampo, Kumasi, Sunyani. Northern Territories—Batiasan, Bawku, Binduri, Bole, Daweni, Dogan Kade, Jefisi, Kalleo, Kpalgu, Lorha, Mayoro, Nandaw, Nandawli, Nasia, Navarro, Pinna, Salaga, Sambolugu, Savelugu, Tamale, Tanina, Tiana, Tishi, Ulu, Wa, Yaga, Zuaragu. Culex invidiosus. Colony—Accra, Akuse, Bibianaha, Koforidua, Nsawam, Sekondi. Ashanti—Akrokerri, Bjere, Kumasi, Obuasi, Odumase, Sunyani. Northern Territories—Bawku, Bole, Gambaga, Kugri, Lorha, Maliki’s Zongo, Navarro, Tumu, Turu, Wa, Zuaragu. So far as they go these records do not point to any cleavage in the areas of distri- bution of the two mosquitos, both species having been collected in a number of the same places. The majority of the records for the Northern Territories were made by one of us (A.I.) during a tour undertaken in 1918, the observations made at this time suggested that C. decens might be a domestic variety of C. invidiosus. Edwards (Bull. Ent. Res. ui, p. 381), however, considers the two to be distinct. In the adult, he says, ““ C. decens can be distinguished by the reddish thorax (that of C. invidiosus being brownish), and (in the male) the banded abdomen. In the female the abdominal banding is not constant ; the bands in C. decens are always narrow and may some- times be interrupted.” But he failed to detect any differences in the genitalia of the males (an observation which he repeated later, Bull. Ent. Res., v, p. 70), and was unable to separate the larvae, considerations which led him to admit that it was ** quite possible that the two are really only forms of one species.” We have compared a number of larvae and pupae of these mosquitos in an endeavour to find some means of distinguishing them. We may say at once that we 106 J. W. S. MACFIE AND A. INGRAM. were unable to detect in the larvae any constant difference of specific importance, and as this observation is in accord with that of Edwards further details will not be given. In a previous paper (Bull. Ent. Res. vin, p. 85) we have described the pupa of C. nvidiosus. The pupa of C. decens (= C. mgrocostalis, Theo., and C. lividocostalis, Graham) has been briefly described by Wesché (Bull. Ent. Res. 1, pp. 40 and 45), but the details given are meagre. As it was necessary on this account to re-describe the pupa of C. decens, and as it seemed advisable to give at the same time further details regarding C. invidiosus because of the very close resemblance of the two pupae, we decided to compare seta by seta five specimens of each. The result of this examination will be recorded here in detail. The pupae were found to be similar both in their general features and as regards the arrangement of their cephalo-thoracic and abdominal setae. In describing the setae the same method has been adopted as that employed by one of us in des- cribing the pupa of Stegomyza fasciata (Bull. Ent. Res. x, pp. 161-172). General Features. In Table I the general features of the pupae are tabulated. The figures given are the averages of the measurements made in each case. The specimens of C. invidiosus examined were on the whole slightly smaller than those of C. decens. It will be observed that the general features of the two pupae were similar. TABLE I. General Features of the Pupae of C. decens and C. invidiosus. OC. decens C. invidiosus. Length of pupa, extended .. | 4-44 mm. 4-44 mm. Respiratory trumpet, total length 640u 580u length of the open ‘portion (pinna) 170u 160u ratio of the ‘length of the closed portion (meatus) to the total length ee a alee 1:14 Length of the tuft (A) at the posterior angle of the seventh abdominal seg- ment ee ae at aOR 1904 Length of the tuft (A) at the pos- terior angle of the eighth abdominal segment .. | 210u 190u Paddle, length .. mn .--| 750n 740u breadth .. 530u 530u ratio of length to breadth | 1:4: 1 L431 Cephalo-thoracic Setae. The setae on the cephalo-thorax of these pupae are alike, and are similarly situated to those of Stegomyia fasciata (loc. cit.) ; they do not require special description. Details regarding the subdivisions of the cephalo-thoracic setae are furnished in the following comparative table (Table II). THE EARLY STAGES OF WEST AFRICAN MOSQUITOS. 107 Fig. 1. Diagram showing the arrangement of the setae on the cephalothorax of the pupa of Culex decens or CO. invidiosus. TaBLeE II. A Comparison of the Cephalo-thoracic Setae of the Pupae of C. decens and C. invidiosus. | Seta. OC. decens. C. invidiosus. Post-ocular, | Superior ; rather long, not especially strong ; 3 | 3-4, usually 3 Median; rather Tong, not especially strong 4-5, usually 4 4—5, usually 4 Inferior ; rather long, not | especially strong .. | 3-4 3 Antero-thoracic, Lower anterior; moderate length, constituent hairs rather delicate .. 4-5 4-7 Upper anterior; rather small and delicate on tee 3 Lower posterior ; long, but not very strong .. Ry te 2-3, usually 2 Upper posterior; rather small and delicate, situated above the lower posterior seta .. | 1-4 2-5 Dorsal; moderate length and Strength . 2-4, usually 4 3-5 Supra-alar ; moderate length and strength . .s spine 2-3, usually 2 Postero-thoracic, Internal; well-developed tuft of moderate length and strength .. 5-8 5-8 Median; rather long and strong . 2-3, usually 2 y External ; tuft of moderate length ‘and strength, or rather long tg 3- Cr he Or | | 108 J. W. 8S. MACFIE AND A. INGRAM. Dorsal Abdominal Setae. The setae on the dorsal and lateral aspects of the abdomen may be described together. Details as to the subdivisions of these setae will be found in the tabular statement which follows; the facts given here are either of a general nature, or such as may be required to make the table comprehensible. The following setae, which form series, are recognisable on each side of a typical segment of either pupa :— 1.—The lateral seta (A), situated a little above and internal to the posterior angle. 2.—The seta belonging to the sub-median row (C), on the dorsum, situated near the posterior margin of the segment about half way between the posterior angle and the middle line of the abdomen. This is the row of setae which on a previous occasion (Bull. Ent. Res., x, p. 59) was referred to as the wnner lateral row. 3.—A very small seta (C’), situated near the posterior margin of the segment and a little internal to C. . Fig. 2. Diagram showing the arrangement of the setae on the dorsal and ventral aspects of the pupa of Culex decens or C. invidiosus; dorsal setae shown on the upper, ventral setae on the lower half. 4.—The seta belonging to the sub-lateral row (B) on the dorsum, situated near the posterior margin of the segment about half way between the posterior angle and C. This is the row of setae which on a previous occasion (Bull. Ent. Res., x, p. 59) was referred to as the outer lateral row. 5.—Two setae (B’ and B”) situated a little anterior to the posterior margin of the segment ; B’ outermost and always more or less external to B, B” internal to B’ and on some segments external, on others internal to B. 3B” is sometimes associated more closely with C than with B, for example on the sixth segment. . 6.—A very minute seta, the anterior dorso-central seta (D), near the anterior margin of the segment in line with the setae of the sub-median row. These seven setae are easily recognised on the third to the sixth segments; the setae on the other segments do not conform completely to this plan and require a few words of explanation. Segment I. The setae on this segment are arranged in a similar manner to the corresponding setae of Stegomyia fasciata, and the same names are used in describing them, namely, Dendritic tuft for the large branched seta in the middle of the segment near the median line, and, for the setae along the proximal border, from within outwards, Antero-internal, Antero-external, Medio-internal, Medio-external, Postero- internal, Postero-external and Lateral. THE EARLY STAGES OF WEST AFRICAN MOSQUITOS. 109 Segment II. Nine setae may be seen on this segment in a dorsal view, of which C and D are as defined above. Internal to C, near the posterior margin of the seg- ment, is a small dendritic tuft (D. T.). The setae external to C are as follows: a small single seta (B) a little external and anterior to C, a tuft of moderate size (B”) a little external and anterior to B, and a small tuft (B’) a little external and posterior to B”. Laterally, or ventro-laterally, there are three setae: the most anterior (A’) a very long, stout seta ; rather posterior to A’ a seta (A) similar to the Lateral setae on the more posterior segments ; and posterior to A a very small seta (S) similar in appearance to the Postero-lateral setae on the ventral aspect of the abdomen. Segment VII. A, B, B’, B’, C, and D as defined above. There is no small seta internal to C, but a little external to it there is asmall seta similar to C’ on the other segments ; this seta is referred to in the Table as C’. In addition there is a small tuft (A’) close to and a little posterior to A. Segment VIII. A and D as defined above. (P) situated close to the external margin of the root of the paddle. In addition there is a delicate seta At the distal end of the midrib of the paddle there are two very small setae (P’ and P”) ; usually single, but sometimes forked at their ends or even divided more deeply. TaB.eE IIT. A Comparison of the Dorsal Abdominal Setae of the Pupae of C. decens and C. invidiosus. a Seta. O. decens. O. invidiosus. I. |Dendritic tuft ; large and well- developed, strong primary branches | 8-12 8-11 Antero-internal ; short, rather stout | 1, end occasionally 1 forked. Antero-external; rather long and strong be ls na .. | 2-3, usually 2 2 Medio-internal ; small ne .. | tora amalltatt* 1, or a small tuft Medio-external; small, but longer | Tuft of about 2-5 Sunilar tuft than the Medio-internal oii hairs. Postero-internal; very long, strong | 1-2, usually 1 1 Postero-external; rather long and slender - ee dy? 2 1-3 Lateral; small, delicate 1, end sometimes | 1, end sometimes divided. divided. II. |A_ ; rather long and slender. . Z 1-3 A’ ; very long, strong 1 1-2 S ; very small and feebly chitinised | 1 1 B_; small, rather stout > e | 1 B’ ; small tuft 7 Bs .. | Smalltuftof2—5hairs) Similar small tuft B’ ; small tuft, but considerably larger than B’, situated between B and B’ and anterior to them | Tuft of 4-6 hairs Tuft of 4-6 hairs C ; moderate length, reaching nearly half way across the third ae ment. oa 2-3, usually 2 2 D.T. ; small dendritic ‘tuft Small dendritic tuft | Small dendritic tuft D ; minute ; 1 1 III. |A_ ; rather long and slender 2-3 2-4 B_ ; rather long and strong .. ra 2 B’ ; small tuft between A and B” .. | Smalltuft, 1-7 hairs) Small tuft, 1—4 hairs B’” ; tuft, considerably longer than B’, situated between B and B’ .. Tuft, 3-7 hairs Similar tuft *In the case of small tufts, the number of constituent hairs, when given, is only approxi- mate as it was often impossible to determine with accuracy the degree of sub-division. 110 J. W. S. MACFIE AND A. INGRAM. sch Seta. C. decens. C. invidiosus. C ; well-developed tuft, reaching | about half-way across the fourth segment a: Be 6-9 6-12 C’ ; small AG ee 1 D ; minute i « Wal 1 Iv. |A_; rather long and stoner 2-4 3-4 B_ ; long, strong, reaching almost across the fifth segment 3-4 4—6 B’ ; small tuft, external to B Small tuft, 1-3 hairs | Small tuft, 1-3 hairs B’” ; small tuft, anterior to B : Small tuft, 4-6 hairs | Small tuft, 4~7 hairs C ; well- developed tuft reaching fur- ther than half-way across the fifth segment 4-8 6-10 C’ ; small 1 1 D ; minute I 1 1 Vv. |A_ ; rather long and slender . 2 3-4 3-4 B_ ; very long and strong, reaching beyond the posterior margin of the sixth segment 5 2 2 B’ ; rather small tuft, external to By Small tuft, 2-5 hairs | Small tuft, 2-6 hairs B’ ; moderate length, external and anterior toC .. 1—2, usually 2 2-3, usually 2 C 3; well-developed tuft, reaching fur- ther than half-way across the sixth segment 4-8 6-7 iC”. svsmall 1 1 D ; minute 1 1 VI. |A_ ; rather long and alontes 3-4 3-5 B_ ; very long and strong, reaching beyond the posterior margin of the seventh segment sia Ae 2 B’ ; small tuft, external to B 2-4 2-4 B” ; moderate length, situated exter- nal and anterior to C 2 2 C ; well-developed tuft, reaching about half-way across the seventh segment : 4-7 4-8 C’ ; small i 1 D ; minute : = : i 1 VII. A ; constituent hairs strong, usually sub-plumose and often branched, sometimes simple : 3-5 , 2-4 A’ ; small tuft ie Small tuft Small tuft B_; moderate length and strength . 2 1-3 B’ ; moderate length, external to B. 2 2 B’” ; moderate length, internal to Bo. ees 2-5 C ; moderate length, reaching about half-way across the eo Seg: ment... 2-5 4-7 C’ ; small, external to C 1 1 D ; minute : 1 1 VIII. |A ; well-developed tutt, ame anes hairs strong, u usually branched and sub- aaa t 6-7 5-9 D «-minute’— ; 1 1 P ; moderate length, ‘slender 2 2 P’ and P”; very small Two small setae Two small setae THE EARLY STAGES OF WEST AFRICAN MOSQUITOS. Lt Ventral Abdominal Setae. The setae present on each segment, as well as the characters of the setae, are shown in the following tabular statement (Table IV). TABLE IV. A Comparison of the Ventral Abdominal Setae of the Pupae of C. decens and C. invidiosus. Seg- Seta. C. decens. CO. invidiosus. ment. III. | D ; very small, feebly chitinised ../ 1 1 E_ ; small tuft, longer than E’ Small tuft, 4-5 hairs | Small tuft, 2—5 hairs E’ ; small tuft Small tuft, 2-5 hairs | Small tuft, 4-5 hairs C ; moderate length 2 ; 2 B’ ; small 1 1 A; minute 1 1 IV. | D ; very small, feebly chitinised 1 1 fos sroall tuit. Ms 1, or small tuft, 2-5 | Small tuft, 2-4 hairs hairs. E’ ; small tuft, longer than E 1, or small tuft, 2-6 | 1, orsmall tuft, 2-4 hairs. hairs. C ; moderate length B’ ; small t 1 | A ; minute 1 1 V. | D ; very small, feebly chitinised 1 1 E ; small tuft ah it 1, or small tuft, 2-4 | Small tuft, 2-4 hairs hairs. E’ ; small tuft, longer than E Small tuft, 2-5 hairs | Small tuft, 4~7 hairs C ; rather long, reaching fully half way across the sixth segment ..| 1 1 B’ ; small a 1 1, end sometimes forked. A ; minute 1 1 VI. | D ; very small, feebly chitinised 1 1 E ; small .¢ is 1, or small tuft, 2-4 | 1, or small tuft, 2-3 hairs. hairs. C ; rather long, reaching fully half way across the seventh segment} 1 1 B_; rather long, shorter than C 1 1 B’ ; small t 1-2, usually 1 1-2, usually 1 A : minute .. 1 1 VII. | E ; small tuft Small tuft, 2-4 hairs | Small tuft, 2-4 hairs C ; rather long, reaching fully half way across the eighth oo ] 1 B_; moderate length 1 1 B’ ; small, end sometimes forked or sub- divided, a little external and anterior to B 1-2 1—4, usually 1 A ; minute 1 1 Vath, | A’ ;. minute 1 1 The setae on the ventral aspect of the abdomen are few, and mostly inconspicuous. The following setae, which form series, may be recognised on each side of the abdomen of either pupa on two or more of the segments :— 1.—A small seta, the medio-lateral seta (E), a little internal to the lateral border and rather posterior (junction of the posterior and middle thirds) to the middle of the segment. (687) 112 J. W. 8S. MACFIE AND A. INGRAM. 2.—A very small seta, the postero-lateral seta (D), in a lateral position just above the posterior angle. 3.—A small seta (E’) a little posterior and external to E. 4.—The seta belonging to the inner ventral row (B). 5.—The seta belonging to the outer ventral row (C). 6.—A small seta (B’) near the posterior margin of the segment a little internal to C. 7.—A very minute seta, the anterior ventro-central seta (A), at the anterior margin of the segment and very close to the middle line of the abdomen. Two setae require special mention. On the eighth segment there is a minute seta (A’) near the anterior margin of the segment a little internal to the level of B on the seventh segment. This seta resembles A, but is placed more externally and pos- teriorly than A on the more anterior segments. On the seventh segment the small seta internal to C is situated close by and a little anterior to B; we have therefore marked it B’, and have retained this symbol for the similar setae on the more anterior segments which, however, are situated nearer the posterior margin and more exter- nally. It is clear from the descriptions which have just been given that the setae of the pupae of C. decens and C. invidiosus are similar in character and situation, and that although they are somewhat variable as regards their sub-divisions the degrees of variation of the two overlap. In fact, they provide no feature of differential im- portance. The question then arises, are C. decens and C. wnvidiosus separate species or are they varieties of a single mosquito. There are indeed certain differences in the adults, but the genitalia of the males are identical according to Edwards, the larvae cannot be separated, and the same remark applies to the pupae. Under these circumstances we think there can be little doubt that they should be regarded as varieties and not as distinct species ; we propose to retain for the species the name C. decens. 113 SOME NOTES AND REMARKS ON THE BIONOMICS OF GZOSSINA MORSITANS. _By Rupert W. Jack, F.ES., Agricultural Entomologist, S. Rhodesia. It is with considerable diffidence that the writer has prepared the following paper, not only on account of several previous contributions to the subject, but also, and chiefly, because no opportunity has been available for more than comparatively brief visits to the fly areas in Southern Rhodesia, and the observations to be recorded are therefore of a somewhat meagre nature compared with work that has been carried out of recent years in other parts of Africa. The tsetse-fly problem is, however, of such paramount importance to Africa as a whole, and the continued extension of the fly belts such a grave hindrance to the full development of Southern Rhodesia, that it is felt that too much time and thought cannot be devoted to the subject, and that every contribution whether of actual observations or of theoretical deductions, or of both, must have a certain value, if only in provoking discussion and criticism. In the following pages the important question of whether or not Glossina morsitans is vitally dependent upon the larger mammals is not directly dealt with. The strong trend of scientific opinion in the direction of an affirmative conclusion must be admitted at the present day by the most vigorous opponent of the theory. There is hardly an investigator of any standing who has not contributed some observations of weight, indicating that the fly is mainly dependent upon the larger mammals for its food supply. For the purposes of the present discussion this dependence has been to a large extent assumed with a view to an attempt to explain certain phenomena upon this basis. The discussion is, however, of a very disconnected nature and aims merely at touching upon a few points in connection with the bionomics of the fly that need clearing up, and if the writer’s views provoke criticism, or even direct con- tradiction, supported by actual observations, the paper will have served its purpose, as our knowledge of the subject cannot fail to benefit thereby. It will be found that there is a very substantial agreement between observations in this territory and those recorded further north, particularly in Northern Rhodesia, but there are several important points bearing on the fly problem concerning which a considerable difference of opinion appears to exist, and others that have not apparently been touched upon as yet. The points raised include :—(1) the conditions in the dry season in respect to the welfare of the fly ; (2) the effect of grass burning on the numbers of the fly ; (3) the value of mopani country to the fly ; (4) the question of migration under the influence of hunger or other stimulus ; (5) the question of fly moving about infested areas with game; (6) the distance at which the fly can detect its hosts; (7) the following distance of female flies. In the following pages it is desired to emphasise the importance of these points, stating the writer’s own attitude of mind towards them in the light of available evidence, but without preferring any claim to having reached a final conclusion. Direct investigation of several of the points is planned for early attention if cir- cumstances permit, and it is hoped that other investigators will carry out similar researches so that definite solutions may be reached. (687) B2 114 RUPERT W. JACK. Conditions in the dry Season. Dr. H. Lyndhurst Duke, of Uganda, in a recent paper* refers to the dry season as a time of stress and reduced breeding rate for the fly in that territory. Lloyd’s records concerning Northern Rhodesia are quoted as parallel observations, although the latter’s statement to that effect applied only to the earlier part of his work in the Luangwa Valley (Bull. Ent. Res. 11, p. 234) and his later work on the plateau led to exactly the opposite conclusion, namely that breeding was practically confined to the warmer part of the dry season (Bull. Ent. Res. v, p. 58). Lamborn in Nyasaland has made several statements on this pomt. For instance, in his Third-Report on Glossina Investigations in Nyasaland (Bull. Ent. Res. vii, pp. 29-50) on page 30 we have the statement: ‘‘ The small number of living pupae is to be accounted for by the breeding being at its nunimum during the late dry season” (July). On page 36 he states: ‘‘ Now as the dry season draws to a close pupae are being produced more freely;’’ and on page 46, “ The breeding season for G. morsitans is in full swing, just as the rains are about due.” It would seem that the period when the dry season “‘ draws to a close”’ and “ the rains are about due” might without inaccuracy be described as “ the late dry season ”’ in preference to July. Nevertheless Lamborn’s contention is quite clear, namely that the breeding season is at its minimum in July, and is in full swing in October, when the first rains may be anticipated both in Northern and Southern Rhodesia. In Southern Rhodesia, however, the October rains are normally very local and may not occur at all, and dry season conditions frequently prevail well into November. October and November (before heavy rains set in) are well known as the hottest months of the year and the term “ latter part of the dry season ”’ would appear naturally to include the period from the beginning of August to the opening of the heavy “ planting ” rains, which usually commence about mid-November. In some seasons they are, however, postponed until well into December, and in such circumstances, as the veld still maintains a dry season character, observations on breeding rates, etc., are still influenced by late dry season conditions, and this should be made clear. | Observations on the seasonal breeding rate in Southern Rhodesia are insignificant compared with those in Northern Rhodesia and Nyasaland, owing to the fact that no officer has been available to devote even the bulk of his time to the tsetse-fly problem. There is not the least doubt, however, that the fly may, under favourable conditions, breed very freely during the latter and warmer part of the dry season (August to November). Observations at Sipane Vlei, already published, leave no doubt upon that point. It would seem therefore that conditions in Nyasaland, and Northern and Southern Rhodesia are similar in this respect, at least as regards the plateaux. In Uganda on the other hand the published records of rainfall in the Masindi area and at Mpumu show that the climate is very different, the dry season being much less clearly marked and of much shorter duration. ~ _In 1911 the writer called attention to a marked difference in the distribution of fly in the dry and wet seasons.f The dry season haunts of the fly in Southern Rhodesia have been described more than once. They consist of centres where shady evergreen *Bull. Ent. Res. x, p. 20, lines 11-13, and elsewhere. Bull. Ent. Res.‘ii, pp. 357-361. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 115 trees exist and apparently correspond more or less to the areas termed “‘ foci,” “primary centres,” etc., described later in Central and East Africa. These dry season haunts undoubtedly vary to a considerable extent in their efficacy as favour- able habitats for the fly, as is indicated by the fact that some of them are intensely and others little infested. The type of dry season haunt where the fly is generally met with in the greatest numbers consists of a vlei, with the necessary evergreen trees along at least some portion of its margin, in a part of the fly area where vleis are scarce. Surface water may be present, or not, but the underground supply is always much nearer the surface than in the surrounding country, with the result that green grass starts growing with the warm weather long before the rains commence. The green grass in the vlei, especially after the dry grass has been burnt off and the young grass has begun to grow, is intensely attractive to grass-feeding animals, which often concentrate in considerable numbers around an isolated spot of this nature and move in regularly during the afternoon to feed, passing perforce through the haunts of the fly in doing so. In this way the flies obtain more or less regular meals and the whole conditions appear to be ideal for rapid increase during the latter part of the dry season, namely, regular food, warm weather and shelter for the larvae and pupae. Such centres might conveniently be termed “first grade foci.” In other parts of the fly areas rivers and vleis are much more prevalent, and intense concentration of both game and fly does not occur. The dry season haunts of the fly correspond, however, with the best grazing grounds at this time of the year, on account of the fact that the superior underground water supply supports both ever- green trees that shelter the fly and green grass that attracts the game. The term “second grade foci’ might be permitted in this connection. In other places again, the fly affects shady forest bordering rivers, streams or dry water-courses, and although not the main grounds, game has a tendency either to lie up in such situations during the day, or to pass through for the purpose of drinking. In the writer’s experience, wart-hog, water-buck, kudu, impala and small buck are very commonly found in this type of forest during the daytime, whereas certain types constitute the permanent home of bushbuck. Such spots are rarely heavily infested and may be termed “ third grade foci.” It must be borne in mind that animals of all sorts seek the shade during the heat of the day in hot weather, and as the fly is also dependent upon shade, fly and game must in general tend to meet at the time of year when shade is restricted more readily than at other times when shade, water and grass are everywhere to be found. Provided, therefore, that the temperature is suitable for breeding to proceed at the maximum rate, the dry season should in a general way be more favourable to the fly, assuming its dependence on the larger mammals for its food supply, than the wet. It does not follow, however, that this is necessarily the case in particular instances, Certain tracts of country dry up so thoroughly towards the end of the dry season that they are deserted altogether by most species of game. As has been remarked by several observers, however, wart-hog and duyker are usually to be found even after the grass has been burnt off over a wide area. These species, especially wart- hog, may serve to keep the flies from starving, but conditions cannot be said to be favourable to increase. Fly under such conditions is almost invariably very hungry, the proportion of the sexes caught tends to approach equality and pupae are 116 RUPERT W. JACK. extremely scarce.* Fly appears to be extremely numerous and is certainly extremely pertinacious. Half-a-dozen flies that are eager for blood give a far greater impression of numbers than half-a-dozen “following” males. Fifty or sixty really hungry flies would give the impression of hundreds. Several investigators have indeed drawn attention to the fact that the numbers of tsetse under such conditions may be more apparent than real. . The writer can only speak with experience of one locality of this nature (Gorai River, Lomagundi), but a noteworthy point was the extreme localisation of the fly, few being met with except at the point where the path crossed the water-course, and here they seemed to swarm. When the writer camped at this spot for several days, the flies became very much less troublesome. Many pre- sumably fed on the two dogs accompanying the party, and a few on the natives. Some 80 odd were caught, and the party was then left in comparative peace. No- where else in the dried-up triangle of country formed by the Hunyani and Ambi Rivers with the escarpment were flies met with in any numbers. The following April no fly was encountered at the river crossing, but they were present in the neighbouring mopani forest in very moderate numbers, giving the impression of a decrease since the previous November. In regions such as this it is judged that the wet season must constitute the main breeding period, owing to the return of game with the rains, but that under these conditions the fly does not attain the same numbers as in areas where more favourable winter haunts are found. The visitor in the dry season meets, however, a large proportion of the fly population at once, whereas if game is plentiful he probably meets only a fraction. This argument may not apply universally to localities where fly seems numerous and game very scarce. In the area referred to it was actually shown that game was moderately plentiful when the country was not dried up. Data for times of year other than the dry season appear to be lacking in respect to most of the classic instances where fly apparently abounded in the absence or extreme scarcity of game. During the earlier part of the dry season, in May, June and July, the coldest weather occurs, and there is little doubt that the fly breeds less freely at this time of year and that the pupal period is more prolonged, both of which facts tend to check increase. On the whole, the fly tends to be less numerous at this time of year, particularly in July, but durmg August and onwards to the advent of the rains con- siderable increase occurs. With the advent of heavy and persistent rains the fly scatters and therefore appears to decrease suddenly, although this is probably only apparent. The answer to the question as to whether the dry season is in general a period of stress for the fly or not, except in so far as the breeding is influenced by temperature, would appear therefore to be in the negative as far as conditions similar to those of this territory are concerned, and the writer is of opinion that, excepting certain localities, far from being a time of stress, the latter part of the dry season is probably the most favourable and important period of the year under normal conditions. Furthermore, it would appear that the later the rains are in commencing, the greater the insects’ capacity for increase. That such conditions are not inimical to the fly *The writer is largely repeating results obtained by others in Central and East Africa, although observations in 8. Rhodesia are entirely in accord. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 117 is shown by the fact that the latest onset of the rains recorded in recent years in the territory, namely in 1912, when no appreciable rain fell till towards the middle of December, was followed next year by the largest extension of fly recorded in connec- tion with the Sebungwe fly area. The Effect of Grass Fires. Another point is the stress laid by certain investigators on the efficacy of grass fires in reducing the fly. The present writer has been unable to obtain the slightest evidence that grass fires in this territory have any material effect on the number of fly, indeed the evidence to the contrary is almost conclusive. One of the most striking instances was recorded on the Gaori River below the escarpment in the Lomagundi district. Fly in November 1910 was still confined to a dry season haunt, which consisted in this case of shady trees on the bank of the dry water-course mentioned above. For some reason the grass had not been burnt until three days (according to the natives) before the writer’s visit, and the presence of still smoul- dering trees and the general appearance of the ashes of the grass showed it to have been burnt extremely recently. The fire had swept through the country over a very wide area. Fly was present in considerable numbers, as already mentioned, attacking the party very persistently. When a second visit was made to the spot in April of the next year, it was found that the grass throughout the “ focus ”’ stood as high as a man’s chin and was so thick that it was far too laborious to attempt to make headway through it. The fire the previous November must therefore have thrown up flames twelve to fifteen feet into the air. These fires are an annual occurrence, here as ‘ elsewhere, and yet this little water-course is well known to all using the path which crosses it, as a fly centre, where the pest is always very much in evidence in the dry season. This was, as a matter of fact, the reason why that particular foot-path was followed. Many other less striking instances have come to the writer’s notice during the past ten years, and the accumulated evidence has induced an entire loss of faith in the efficacy of grass fires in reducing the fly. As a matter of fact if tsetse-flies could not avoid grass fires they would surely long ago have been exterminated in parts of the country inhabited by natives. Of course even in Southern Rhodesia much patchy burning of the grass occurs, but the long, intensely dry winter pre- sumably favours burning over wide areas more than conditions in Uganda, for instance, It is difficult to see how the flies could avoid a roaring furnace like that along the Gorai River by mounting in the air unless they went to a very great height (for a fly), but the contiguous mopani would afford a refuge, as in this type of forest the grass is always very short, and is sometimes almost altogether wanting. If large areas of country infested with fly were covered with long thick grass better results might be anticipated. Possibly this is the reason that greater effect on the fly has been noticed elsewhere. The Value of Mopani Country to the Fly. Both Duke and Fiske, referring to Uganda, speak of the fly showing hunger in the long grass season, which is of course the latter part of the wet season and earlier part ofthedry. This is attributed to difficulty in locating game. The point is extremely interesting, because from Duke’s description of the Masindi area comparatively long 118 RUPERT W. JACK. grass apparently occurs over wide areas. Conditions differ in various parts of the fly-infested areas of Southern Rhodesia, but in the great majority mopani belts are a feature of the country, and the fly certainly shows a strong preference for this type of forest in the wet season, deserting its dry season haunts where the grass is usually long and thick, sometimes very long. The sweet short grass of the mopani is very attractive to many species of game in the wet season, though the reverse in the dry, and the fly certainly has the best chance of meeting and perceiving game in this type of country during the rains. The mopani grass, however, dries up very quickly after the cessation of rains, being shallow-rooted and of little substance, so that it loses. its special attractiveness very shortly after the close of the wet season. Nevertheless, this type of forest is much haunted by impala and wart-hog even at that time, and many other species are commonly seen in it, if game is at all plentiful in the neigh- bourhood. Whilst, therefore, it is probable that m Southern Rhodesia the long- grass season is not in general a time of stress for the fly, the fact is m no way opposed to Fiske’s observations. In this connection it may be mentioned that the only sex count made by the writer in the long-grass season was in April 1911, with fly taken in mopani, and resulted in 53 males to 3 females, by far the most extreme disparity yet recorded in the territory. The fly’s habit of avoiding thick undergrowth would indeed suggest a difficulty in finding its hosts in such an environment. It is quite possible that in certain parts of the fly areas in Rhodesia a time of stress for the fly may occur in the long-grass season, as the mopani is not ubiquitous, being absent, so far as the writer is aware, from most of the Jetjenini fly area and from certain parts. of the Sebungwe area. Prospectors and hunters in this territory have constantly associated mopani and tsetse, and although obviously not essential to the tsetse, it is quite possibly of value to the fly under certain conditions. According to Fiske’s observations it ought theoretically to be of value in the long-grass season, and possibly it affords a refuge from grass fires. The Question of Migration. The next point is the question of the migration of fly, corresponding to the move- ments of game, or under stress of hunger. Now it is far from the writer’s intention to adopt a dogmatic attitude on this subject, but it would seem difficult to reconcile any habit of this nature with a number of known facts. First, in regard to the ques- tion of fly migrating with game, we have the phenomenon of restricted fly-areas and their mode of extension. It has already been pointed out (Bull. Ent. Res. x, p. 88) that the limits of a fly area are not necessarily permanent. The permanent limits are marked by the impingement of favourable on unfavourable country; that is to say, the country beyond the limit is for some reason or other unsuited to the tsetse. Transitory limits are, however, formed by the high-water-mark, so to speak, of the advancing flood when the fly is spreading, as has been the case in this territory since the rinderpest. Transitory limits would also be apparent if the pest were receding. The advance of the pest is, however, comparatively slow, and in no way com- parable with the powers of movement of the fly itself or of game. Much potential fly area, known to have been infested in pre-rinderpest days, is still free from the pest, although the latter is gradually occupying more and more of its old country. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 119 Is it possible for these conditions to obtain if the fly had a habit of migrating with herds of game, apart, of course, from the males’ habit of temporarily accompanying moving animals and human beings? Herds of certain species of game are more or less constantly on the move in and out of the fly area, and if the fly accompanied them in their wanderings the whole of the potential fly area would certainly be very quickly occupied. > It may be contended, however, that the fly only follows “ trekking ” game under stress of hunger. This appears to imply a rather marvellous instinct on the part of the fly. Any fly suffering from hunger that encounters a herd of game is likely to be a full-fed fly in a space of time measured by minutes, and is then not in a condition to follow anything for some time. Must we, therefore, suppose that the fly restrains its appetite in order to accompany the herd into another part of the country ? Again, game travels as a rule by night, and though there are numerous records of fly “‘ biting ” on warm nights, it is essentially of diurnal habit. Finally, we have apparently no records to show that the female flies follow animals or human beings for any appreciable distance at all, whether they are hungry ornot. The writer’s observations are all to the contrary, and published accounts apparently give no record of other than males exhibiting an obvious following habit. Much argument has in fact been based upon the supposition that the female flies only seek animals and human beings for the purpose of feeding, whilst the males follow for other reasons. These considerations have led the writer to be extremely sceptical concerning the fly’s alleged habit of migrating with game. Regular movements of game in a given direction might, it is admitted, help the spread of the fly in that direction. The allusion is to the daily movement of game from its feeding grounds to water, as, for instance, game which feeds in the neighbourhood of a dry river and moves across to a flowing one to drink (e.g., the Mzola and Kana Rivers in this territory). In the same way, regular traffic along a road would tend to spread the fly, if condi- tions admit of spread, but this is quite apart from the question of fly migrating with migrating game. The second aspect of the question is that of fly migrating without help from moving game, when pressed by hunger. This would be a serious state of affairs ; as they might be attracted by the settler’s herds of cattle, take up their residence in the vicinity of his “ kraal” and deposit their larvae under his orange trees. Such a development might well have happened in the writer’s experience if hungry fly had a habit of migrating even three or four miles to a food supply. The important point is that it did not happen. The idea of this form of migration has, as a matter of fact, always puzzled the writer. How is it supposed to take place? Do the tsetse send out scouts like bees to locate a better piece of country and then rise in a swarm to follow these guides to the new locality 2? Does some instinct inform them that better conditions prevail elsewhere, so that they rise with a common impulse and migrate thither? Do they, under stress of hunger, migrate en masse at random guided by chance, or perhaps the direction of the wind? Do they migrate by a series of short journeys, gradually tending day after day in one direction ? Unless the movement takes place at random, this presupposes a knowledge of where 120 RUPERT W. JACK. better feeding grounds are to be found. Or, finally. do they migrate individually ? In the latter case they might take many directions and a corresponding increase in another part would not be likely to accompany the decrease in the original centre. Fortunately for the neighbouring settlers, where we have had a clear case of hunger conditions as regards game, in the Hartley district, the diminishing fly became gradually restricted in range as the game decreased, and still adhered to their old haunts, whilst numerous cattle worked with more or less impunity amongst the Shagari mines a few miles away on the one side, and settlement proceeded along the railway line with rapidly decreasing losses on the other. The destruction of the usual haunts by the felling of the forest did, however, appear to cause a scat- tering of the tsetse, which died out, probably in consequence. The fly along the Gorai River, mentioned above, were obviously very hungry at the time of the writer’s visit in November 1910 (out of some 82 flies caught actually more were females than males), and from all accounts this is an annual occurrence at this spot; but although game was more abundant on the Hunyani a few miles away, the flies apparently preferred to endure hunger in this area rather than migrate. They were not, it must be confessed, absolutely starving, as wart-hog and duiker were actually seen in the fly haunts; but their meals were without doubt sufficiently irregular,to keep the appetite of the majority very keen. Fly, on the other hand, was comparatively scarce on the Hunyani where more game was to be found. The writer has, in fact, been able to find no evidence during the past ten years in this territory that the fly has any tendency to migrate, apart from the seasonal scattering during the wet season, as a result of which the pest is much more uni- formly distributed over the infested area than during the dry season. Although new areas have been invaded by its gradual spread, no incursion of fly into localities other than those immediately adjacent to the former infested areas have occurred. It is readily admitted that an apparent migration might take place in the course of time, the fly increasing in a newly invaded area and decreasing in the old, but this does not imply migration. In any case in Southern Rhodesia the invasion of new areas, in the writer’s experience, has not been accompanied by any corresponding reduction elsewhere, nor has reduction in an old area been accompanied by any noticeable increase elsewhere. The fly appears to increase or decrease in any area solely in relation to its ability to breed successfully. On the whole there seems little reason at present to believe in any form of migration in respect to Glossina morsitans, other than the seasonal scattering with the advent of the wet season, and forced movements induced by destruction of the forest. In the latter case the fly naturally follows the receding shade. The Question of Fly moving about Infested Areas with Game. The idea that fly locates game and follows its movements more or less continuously, at least within the limits of infested country, is not absent from the writings of serious investigators. In this connection it is essential to distinguish between following for a limited distance, as is the well-known habit of the males, and possibly of the females to a much less degree, and the alleged habit of the fly attaching itself to the herds and accompanying them in their wanderings day after day. The writer has been able to find no direct evidence adduced in support of this idea, NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 12] and there are certainly many reasons why such a habit is, to say the least of it, unlikely. The question is intimately connected with that of fly migrating with game ; for if the fly followed herds of game within the infested area, there appears to be no obstacle to its following the herds beyond it, provided that the country entered is potential fly country. This, as an attempt has been made to show, is at least not its general habit. The further difficulties in the way of accepting this idea, mostly already mentioned, may be briefly summed up as follows :— (1) female flies are apparently not known to follow to any great distance at any time ; (2) gravid females seek seclusion, and are not in the least likely to attempt to follow a moving herd ; (3) hungry flies tend to feed fully, and full-fed flies abandon their hosts, which are unlikely to remain in the vicinity until the flies recover; (4) the fly is diurnal in habit, whilst game moves largely at night. It would seem therefore that the only individuals capable of following game even for a few hours are those which do not desire to feed. The majority of observations indicate that only the males exhibit a tendency to follow, although the possibility of non-gravid females doing so is not altogether excluded. In any case it is well known that even the males do not follow human beings for more than an hour or so, but gradually fall away, and there seems little reason to think that they would exhibit a different habit in respect to game. Even supposing they followed till nightfall, they must surely lose touch with the herd after dark. It appears probable that the tsetse-fly neither ranges the forest in search of its prey, nor follows it when encountered for any great length of time, but that it waits for the animal to come within the range of its perceptive powers. The fact that fly is constantly met with at the same spots* strongly suggests ambush rather than pursuit, and the writer has certainly no experience of having encountered fly in circumstances that suggested anything else. The Maximum Distance at which the Fly can detect its Hosts. Direct experiments are needed and planned to determine this point as far as possible. It appears necessary to distinguish between the maximum distance at which the fly readily perceives its prey, so that a man or animal would not pass by without being observed, and the distance at which it might locate say a herd of game grazing for several hours in one spot. Some of the early explorers and hunters who encountered tsetse-fly were most emphatic that a very short distance might separate complete safety from certain death for their cattle, a narrow stream being sometimes mentioned as the dividing interval. Like other statements emanating from untrained observers this assertion, though undoubtedly greatly exaggerated, appears to have some foundation in fact. It would be possible to mention numerous personal observations of having been very *The writer is aware that observations have been published of fly having been encountered in quantity at certain spots on one occasion and only in very small numbers at the next visit. During the wet season this is quite likely, as the fly is not confined by lack of shade, and when carried even for a short distance is hardly likely to return to . exactly the same spot. Even in the dry season the passage of a herd of game might make a very material difference for a few hours, the hungry flies feeding and seeking seclusion, and the non-hungry males following the herd. Nevertheless, allowing for the difference in distribution in the wet and dry seasons, the statement is substantially correct. 123 RUPERT W. JACK. little molested by fly on one side of a vlei, whereas on crossing to the other, some two hundred yards or so away, the party was immediately assailed by great numbers. A rather striking instance of the fly’s limitations in this respect occurred in October 1919. The writer passed along a foot-path about 9 a.m. to examine a salt vlei near the Shangani for fly. None was seen till this vlei was reached and an hour was spent walking all round the margin, during which time only four tsetse were caught. Returning, the route taken aimed to cut the path at an acute angle, and when within 100 yards of the path on reaching a shady tree a number of tsetse suddenly attacked the party and “ bit ” with extreme voracity. Yet about an hour earlier the writer with his natives had passed within a hundred yards without being perceived. It. cannot be proved definitely that the fly were there an hour earlier, but there was no spoor to indicate that any game had passed in the interim, and as the morning was. hot the fly would hardly have crossed the leafless mopani to this particular tree. The writer was extremely impressed by this particular instance, and as a matter of fact the limitations in the fly’s capacity for readily detecting a food supply has been a constant source of surprise throughout investigations now extending over ten years. A move- ment of fifty yards or so in a fly haunt frequently brings an accession in numbers of hungry insects and in numerous cases when all the flies seen were being collected, this phenomenon has been particularly marked. If it be accepted as an axiom that: an animal or human being attracts all the hungry tsetse within a radius not exceeding the fly’s powers of perception, then the limitations of the fly in this respect have been proved again and again; for fresh hungry flies have undoubtedly been constantly encountered after a very short movement, whilst some two hundred yards or so has. been found sufficient to keep a party comparatively free from attack whilst the fly swarmed that distance away. To mention another instance, in September 1913, the writer halted for breakfast on the south-east side of Sipane Vlei, between the Sengwa and Sassame Rivers in the Sebungwe district. A certain number of tsetse had been picked up en route, and the boys fetching water brought in more, so that a few bites were received. Some time after breakfast the writer took three natives to the north-west side of the vlei; some 300 yards away, and found the fly in numbers, some 87 being caught with one net in an hour. The wind was from the south-east and had been blowing over the party towards the fly haunts for at least two hours ; yet the bulk of the flies had not been attracted over this comparatively short distance, though they attacked the party in considerable numbers when their haunts were entered. Further visits to this spot have been made since, and it has been found that by camping on the south-east side of the vlei the attentions of fly could always in a great measure be avoided. From the foregoing considerations it is believed that the tsetse-fly is only capable of readily detecting its hosts at a comparatively short distance, probably less than a hundred yards. It might, however, be longer in the wet season, when scent pre- sumably carries further than in the dry. If this limitation be proved, the fact increases the difficulty of accounting for the presence of fly in numbers in areas where large mammals are markedly scarce. It apparently remains to be proved, however, that fly occurs in numbers in any tract of country where large mammals are markedly scarce at all seasons of the year. If such a phenomenon does exist, the writer admits frankly that it appears irreconcilable with the theory of the vital association of the two forms of life. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 123 Is it not possible that this limitation might account to some extent for the gre- garious habit of the tsetse, the flies benefiting by the combined perceptive powers of a number instead of relying on those of single individuals ? The value of numbers in regard to perception will hardly need urging upon anyone who has hunted antelope or other gregarious game. The Following Distance of Female Flies. It appears very important that attempts should be made to determine as accurately as possible the maximum distance to which female flies may be carried by game, man, etc. The point has a great bearing upon the question as to what would con- stitute an effective barrier to the fly’s advance, and also upon the question of the fly as a whole migrating with game or moving about the fly area in company with its hosts. The writer.has as yet been able to prove following on the part of the females up to only about 400 yards, although one instance has occurred where a specimen had apparently followed very much further than this, over a mile in fact. The ob- servation in the latter case was, however, liable to considerable error. Some experiments in this connection were planned for the writer’s visit to Sipane Vlei in August 1919, but could not be carried out owing to the unexpected scarcity of fly in that centre. Sipane Vlei and some other Localities in 1919. In a recent paper * the writer pointed out that the somewhat delicately balanced state of affairs which exists at the most favourable dry season haunts of the tsetse, namely the margin of vleis containing green grass in the dry season, is apparently hable to be somewhat easily disturbed by any agency tending to prevent game from visiting the vlei with its normal regularity. This disturbance, it was pointed out, would enhance the effect of game reduction by hunters, and might also have operated in the rinderpest epizootic, which undoubtedly induced abnormal movements of the panic-stricken remnants. These views have received some indirect support during the dry season of 1919, when for the first time since the rinderpest a reduction of fly in areas of the territory unaffected by civilisation has been recorded. This reduction is local, and extension since the previous year was recorded in one other part of the country, and may have occurred elsewhere. A reduction has been reported in the northern part of the Umniati fly area by Dr. Alec Mackenzie of Gatooma. Dr. Mackenzie’s statement is to the effect that he found no fly at all in December 1919 in many parts where they are usually numerous near the Sakugwe and Umniati Rivers on the road from Gatooma to the Emerald Mine, and only a few finally at the headwaters of the Mvumvudzi and Urungwe Rivers, visited for the express purpose of locating the fly. There is, however, one spot where reduction can be proved on something like a mathematical basis, namely Sipane Vlei, lying between the Sengwa and Sassame Rivers in the Sebungwe district. The rapidly-expanding Sebungwe fly area reached the Sengwa River from the west by 1910, and fly was first recorded at Sipane in 1913, though it may have been present at least a year earlier. In late September of that year it was present in great numbers, and 87 were caught with a net in one hour, in order to establish some sort of basis for *Bull. Ent. Res. x, p. 83. 124 RUPERT W. JACK. later observations. In November 1914, the flies in the vlei were “ like a swarm of bees ” around the individual who penetrated their haunts, and the ease with which pupae were collected during that visit has already been recorded. In August 1916, fly was extremely abundant, and living pupae could be found (some 30 were collected in a few hours), though the latter were not as abundant as in November 1914.* In late August 1919, the writer spent ten days at the vlei, for the purpose of chopping down the evergreen trees as an experiment. Hardly half-a-dozen fly were seen in the course of the day, where formerly they swarmed, and search for living pupae was entirely unsuccessful, notwithstanding that much more time was available than on previous visits. Only a few empty cases were found in the usual breeding places, and from these the flies had apparently emerged normally, there being no indications of parasitism. In seeking a probable explanation of the decrease it was impossible to ignore the marked influence that the heavy rainfall of the two previous wet seasons had had on the country. The wet season of 1917-18 was the heaviest recorded since 1890-1, and was of a peculiar nature, the skies remaining overcast for weeks together, so that the maize crop suffered severely from excessive wet and lack of sunshine ; conditions were thus ideal for maximum penetration. The following season’s rainfall was also above the average. The first difference noted in the appearance of the country was the fact that the gusu (Brachystegia) forest bordering one side of the vlei had not yet lost its foliage, although quite leafless in August 1916 (see photograph, Bull. Ent. Res. x, p. 90, pl. u, fig. 2). The same remark applied in a lesser degree to the thorn thicket (isi-nanga). The mopani, on the other hand, was practically leafless, except half-a-dozen trees in a wet situation on the edge of the vlei itself. These were in full leaf, having apparently grown out again since the fall. The gusu, it may be remarked, showed no young foliage and the leaves were falling sufficiently rapidly to show an appreciable diminution of shade during the ten days of the writer’s stay. Careful inspection of the environment revealed a considerable extent of green grass to the south-west and west of the vlei, which was not in evidence during previous visits. Further, game was remarkably scarce in comparison with previous experience, and such as occurred was not feeding at the vlei itself, but on the green grass mentioned away from the vlei. With the exception of two small herds of impala, one kudu cow, one wart-hog, a rhinoceros, a wild dog and several duiker, no game was actually seen, in spite of daily excursions in all directions. Fresh spoor of a herd of zebra and another of sable was found in the vicinity. A lion drank one night at the vlei, as also did a solitary buffalo bull. This sounds a rather formidable list, but it represents the results of ten days explora- tion of the country within a radius of about eight miles of the vlei. A pack of wild dogs was apparently hunting the neighbourhood, and this may account to some extent for the scarcity of antelope, but the main point ascertained was that the few antelope present in the vicinity were not visiting the vlei for the purpose of feeding as is usually the case. They apparently found more attractive fare in the stretches of green grass away from the vlei (see sketch-map). *In connection with the comparative smallness of the numbers quoted here and elsewhere it should be noted that no trained natives were available and that the pupae and flies were practically all collected by the writer personally. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 125 Further evidence in support of the theory lies in the fact that fly was actually found more readily, though not at all abundantly, in the green grass areas away from the vlei, where such game as occurred was grazing, than at the vlei itself. The general prevalence of shade apparently rendered this possible under the prevailing conditions, although this is not usually the case at this time of year. A few fly were even met with in the gusu itself, which was unattractive to the majority of game, but only three were seen in a walk of several hours. - enya Manyoni R. Scale 9miles to linch. The obvious question is whether the fly had decreased from failure of the birth rate to equal or exceed the death rate, or from the majority having migrated. There was certainly no corresponding increase of fly within a radius of seven or eight miles of the vlei, as the country within this radius was thoroughly explored at the time. Nor had they shifted to the nearest known dry season haunt, which is Matjagenya on the Gadzi River, some seven miles away. There is no evidence of the appearance of fly at any point in this region outside the limits of the known area, and if the fly had migrated en masse in this direction it must have been noticed, as the country is inhabited. As a matter of fact the natives questioned stated that they knew of no fly outside the previously marked limit in this locality. It would appear, there- fore that if the fly had migrated it must have moved back into the infested country 126 RUPERT W. JACK. and for a distance exceeding the radius explored around Sipane Vlei. It is obviously possible that this may have occurred, but it appears very unlikely. Whatever be the correct explanation, it is certain that Sipane Vlei had ceased for the time being to serve the purpose it had served in the past, as a favourable breeding haunt for the fly during the latter part of the dry season. During the season of dispersion a considerable area of country would be affected by this fact. It is interesting to note that some years ago an experiment was planned in connec- tion with Sipane Vlei with a view to determining whether the fly there was dependent upon the visits of big game or otherwise. The plan was to surround the whole vlei with a game-proof fence and thus keep the game away during the season of concen- tration. It was not carried into effect on account of the doubtful practicability of effective control at such a distance from communications. It would almost seem, however, as if meteorological conditions, in reducing the peculiar attraction of the vlei to game, had effected a demonstration instead. Unfortunately further observations bearing on this point will not be possible at this centre as the ever- green trees on the margin of the vle1 have now been cut down. Time was not available for a detailed inspection of other dry season haunts in the same region, but fly was scarce at the crossing of the Gadzi River, known to the natives as Matjagenya, on the way to Sipane, though formerly numerous. The natives resident at this point also bore out this observation. Again, at Depota Vlei some seven miles east of Matjagenya, no fly at all were seen, and some natives who had taken up their residence there the previous year had a number of apparently healthy dogs. They stated that they never saw fly near the vlei, although a very few were seen there by the writer in 1916. Winter shade is scarce at Depota, other- wise it would be a typical first-grade focus, as it is attractive to game as a rule. The writer has in the past seen eland, kudu, sable, impala, zebra, and sessaby in the vlei, and the fresh tracks of buffalo and rhinoceros in addition. In August 1919 only one duiker was seen in the neighbouring forest, and the tracks of three kudu. For the moment, therefore, the fly seems to be retrogressive rather than otherwise in this region, and it is essentially a region of isolated dry season haunts. On the other hand, at the point where Advance has been recorded definitely, namely up the Shangani River, the region is one of diffused dry season haunts (second and third grade foci), and game was very plentiful in October 1919, though their numbers had doubtless been augmented owing to the organised shooting further down the river which commenced in June. Here then we have an illustration of “‘ the fly following the game,” only as the area of increase is some ninety miles, as the crow flies, from Sipane Vlei the fly can hardly have migrated thither. The presumption is that conditions the previous year favoured peal in the one area, and were unfavourable in the other. Whilst conditions in the northern part of the Umniati fly area have no doubt been affected to some extent by the efforts of hunters in shooting game for the “Emerald ” and “‘ Copper Queen ” Mines, and the portion to the east of the river considerably affected on account of being in the open shooting area, the writer has been, so far as is known, the only European visitor to Sipane Vlei for several years past. NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 127. It is not claimed that the only way in which one or two abnormally heavy wet seasons could affect the tsetse-fly lies in inducing a wider distribution of water and attractive grazing during the dry season. The involved interrelationship between the tsetse fly, its diseases and enemies, the enemies of its enemies, and so on, is almost a closed book at present, and this interrelationship is no doubt affected in some degree by the season. It is only intended to point out a possible explanation of the observed phenomenon as a basis for future research. The abnormally heavy tains of the past wet season have certainly affected the appearance of some parts of the country in the dry season to a marked degree, and the fly has undoubtedly decreased in certain areas, one of which at least, as the writer has endeavoured to show, has also been affected by the increased rainfall. | = @ 1°) ~ me 6 Sketch Plan of Sipane Vlei and Neighbourhood. Aug. 1919. E; vergreen trees indicated by black dots. Scale about 1-2 m. to theinch. Curiosity naturally suggested reference to the records of rainfall in the territory with a view to ascertaining whether any record of one or two seasons of exceptional severity preceded any definite information concerning reduction of fly. The available rainfall statistics unfortunately date no further back than 1888 and in the (687) o 128 RUPERT W. JACK. intervening period there is only one record of two consecutive wet seasons comparable with those of 1917-18 and 1918-19. These are the seasons 1889-90 and 1890-1. The total fall in these two seasons at Bulawayo considerably surpassed that of the recent two seasons, the number of inches recorded being 87°81 in the earlier couplet against 76°13 for the later. Was this heavy fall followed shortly by any marked reduction of tsetse? There is some suggestion to the effect that it was, at least in certain parts of the territory, although it is of such an indirect and unsatisfactory nature that it is practically valueless. It is curious, however, that during this very time, a marked recovery from a period of decrease seems to have taken place in the region of the Zambesi. In a foot-note to a paper by the writer (Bull. Ent. Res. v, p. 100), Captain Selous’ opinion is recorded that by 1888 “ the belt shown to the west of the Victoria Falls had disappeared,” and “ that much of the fly to the east of the Falls had also gone by that date.’ This statement was doubtless substantially true in 1888, in fact there is other evidence that fly was decreasing in this area during the preceding years (v. Dr. Holub, Austen’s Monograph, p. 203). Nevertheless Selous’ statement is misleading and should be corrected in case other investigators may use it as a basis for argument. The writer has first-hand information to the effect that fly ‘was numerous between the Falls and the Zambesi-Chobe confluence in 1893, and it is apparent that an increase of fly took place subsequent to 1888. The first-hand evidence referred to is contained in a report by Mr. A. Giese, Cattle Inspector, to the Chief Veterinary Surgeon, Salisbury, dated 9th July 1918. The report is worth quoting in full, although some of the matter does not bear on the present argument. It appears that the fly did actually disappear from this part of the country at the time of the rinderpest, whatever may have been the position of affairs near the Limpopo. Mr. Giese’s report is as follows :— “ 9,vii.1918. The following notes are reminiscently penned because they may ‘serve aS a guide as to where to expect the tsetse-fly, now re-appearing west of the ‘Gwaai,* to spread to should this re-appearance be only a return to localities driven out from or killed by the rinderpest, and not as well to localities which the fly had already voluntarily disappeared from before the rinderpest arrived. “Tn 1861-2 Baldwin(?) brought the first waggons to Deka commencing the track which eventually grew into what is now known as the Pandamatenga or western border road. He had to leave his wagons at Deka and proceed to the Zambesi on foot as the intervening country was infected by fly. ‘“ Somewhere about the middle seventies Geo. Westbeach established his later far- known trading station at Pandamatenga, when the fly had receded further north somewhere north of Gazume Vlei. At the same time, or a little later, transport could be taken to the Victoria Falls and to Bingua Spruit—about railway cottage 277—from where Westbeach’s waggons used to fetch grain traded in the fly-infested country to the east ; also that broken stony country north of the present Matetsi siding was believed to be free from fly right up to Victoria Falls. * No fly has as yet been found west of the Gwaai, but outbreaks of trypanosomiasis have occurred during the past three wet seasons. NOTES ON THE BIONOMICGS OF GLOSSINA MORSITANS. 129 “In July 1893, the writer’s first appearance in these parts, ox transport was left at Gazuma, but fly was found between Leshuma and Kazungula, Chobe and Zambesi junction, and from Kazungularight up the Chobe and right down to within a few miles of the Victoria Falls, fly being particularly aggressive on the latter stretch. “Tn 1894-5 during various trips to what is now the Wankie Colliery all animals had to be left at Deka; in fact when the writer pegged the present Deka farm in 1895 animals were not allowed near its 8. E. beacon. “ During these latter years the southern boundary of the fly area ran roughly from Galobe—Gwaai junction more or less west but north of that open park-like country round and to the south of Chumagu Malishi and Makolo to places near the present Kennedy’s Halt, and hunters used to take their ox waggons thus far north. “* At the end of 1897 the writer took a horse up the Chobe and in 1898 six spans of donkeys to the present Wankie Coal Mine, all fly having disappeared then.” In answer to direct queries by the Chief Veterinary Surgeon Mr. Giese replied as follows under date 30.vii.1918 :— 1. “ What species.died of rinderpest ? “Eland, buffalo and kudu principally, and in the order mentioned as regards numbers found dead and the occurrence of the species at the time. The greater mortality among the eland was not due to their being more susceptible to disease but to their occurring in much greater numbers than the other game. This they still do. Of other game reedbuck probably suffered most. 2. “ What was the rate of mortality ; were there any survivors % “In 1898 the only surviving buffalo, between Kazungula and the Gwaai were one bull, 3 cows and 3 calves, running 7 miles west of the present Colliery, and a somewhat similar number on the lower Inyantue ; which means that these were the only buffalo left in Matabeleland west of the Gwaai, because their habitat did coincide with the tsetse-fly. “What the number of buffalo was in pre-rinderpest days one cannot say, as they were scattered owing to the natives carrying guns ; but the increase of these 7 animals (survivors) is a fairly compact herd of some seventy animals now confining themselves to a belt of country 20 miles by 30 miles to the west of the railway. The Inyantue herd is a little larger. “There were stray survivors of kudu and they also have more than recovered owing to the absence of firearms; there are more kudu than any other antelope, Sharpe’s steinbuck excepted, in the broken and hilly part of the Wankie District, 3. ““ Did the fly disappear at once ? “ The writer left Deka on 15th January, 1896, when fly were as stated in his notes, arrived in Bulawayo mid-February and was sent for from the C. C’s office to give an account of what he knew of rinderpest which was then approaching. When he returned after the Rebellion in 1897 the fly had disappeared.” The record is extremely interesting in showing that fluctuations in the distribution of fly on a moderately extensive scale have occurred in the past apart apparently from overwhelming game destruction, although the hunters appear to have been of opinion that the decrease of fly during the period preceding 1888 was due to their (687) o2 130 RUPERT W. JACK. inroads into the number of buffalo. As a matter of fact, events in this region seem to have contributed largely to the solidifying of Selous’ opinion that the fly was dependent upon buffalo. The rainfall in the Zambesi valley frequently differs very considerably from that in the remainder of the territory, and it is unfortunate that no records in this area _ for the period concerned appear to be available. The records, are however, instructive in regard to the influence of a series of seasons of low rainfall. At Kariyangwe in the Sebungwe district the rainfall recorded for the three seasons 1909-10, 1910-11 and 1911-12 was extremely low. This station was then closed down, but referring to the Wankie records, somewhat further from the scene, we find that the next two years were also very low indeed. During this period the extension of the Sebungwe fly area proceeded at its maximum speed, as may be seen by a glance at the author’s map (Bull. Ent. Res. x, map 1). The records of advance in this area are moderately accurate as regards the three year intervals marked on the map, with the exception of the north, where the discovery of tsetse-fly on the lower Sengwe River in 1913 indicated that the previous limits in this area, visited only at long intervals by officials who traversed the country rapidly, were probably faulty. It is not possible, however, to obtain accurate yearly records. The dry season of 1913 was remarkable for advance in almost every direction, ex- cluding a portion of the western limit where the fly appears to have reached the division between favourable and unfavourable country. The rains were very late in 1912, no serious showers falling until near the mjddle of December. Dry season conditions were thus unusually prolonged, and it is believed that such conditions favour increase of fly. To recapitulate, whilst the writer is far from attaching any exaggerated importance to these limited records, they are certainly suggestive of abnormally heavy wet seasons having a deleterious effect on the fly in certain areas, and of a series of seasons. of low rainfall, especially if late, favouring increase. Intense and prolonged dry season conditions, such as occur during a cycle of years of low rainfall, are judged to favour the fly at centres which resist the action of drought best, owing to the con- sequent concentration of game. A rise in the water-table after an exceptionally heavy season or two causes a disturbance in the usual conditions by inducing a more general distribution of good grazing and water throughout the country, and a pro- portionately reduced attraction of the “ focus ” to game. It is much to be regretted that the writer’s itinerary in 1918 did not include Sipane Vlei. The tremendous rains of the 1917-18 season had a very marked effect on the forest lying between the railway line and the Gwaai River on the western side of the territory, and it is reasonable to suppose that conditions at Sipane Vlei may have been similar to those in 1919, seeing that at Gokwe, the nearest meteorological station, 43°03 inches fell in the 1917-18 season and 33°10 inches the 1918-19 season, although even the latter is well above the normal for that district. If the writer’s theory is correct, fly should have been numerous and hungry in the dry season of 1918, pupae being, however, scarce, resulting in a great diminution of fly by 1919. This theory is very much in accordance with observations published by Lloyd (Bull. Ent. Res. vii, pp. 67-79) with reference to Northern Rhodesia. Lloyd draws NOTES ON THE BIONOMICS OF GLOSSINA MORSITANS. 131 a most necessary distinction between apparent and real abundance of fly, and shows how the percentage of females caught in the ordinary way varies in inverse ratio, and the number of pupae secured in direct ratio, with the abundance of game. Here we have direct observations indicating that scarcity of game induces hunger and hunger inhibits breeding. Both these observations have received support from other investigators, including the comparatively scanty observations in this territory, and appear to be in a fair way to become established. If this proves to be the case the effect of hunting in the early days in the Transvaal and the apparent effect of the rinderpest would appear to be quite comprehensible, in spite of the fact that the game was not completely eliminated in either case. The result of a heavy reduction of game would seem to follow logically, quite apart from complete elimination, namely irregular meals at long intervals—perhaps some individual starvation—in any case a greatly reduced birth rate. If the birth rate were checked at all suddenly, especially during the season of concentration when the larvae are deposited within a limited area, the fly’s parasites might be expected to destroy a much higher percentage than before. The writer is very much in agree- ment with Duke’s contention as to the probable effect of a sudden or “‘ cataclysmic” change in the distribution and numbers of game animals. The local distribution of G. morsitans in relation to the habits of game animals, even during the time of year when the fly is dispersed, is frequently very marked. In this way it appears that the effect of shooting and of the rinderpest might be explained, but it is only com- prehensible on the supposition, which the writer believes to be justified, that the fly’s ability to perceive its hosts is strictly limited, that it 1s incapable of following up a trail for any great distance by scent and of attaching itself to and keeping in touch day after day with such game as may be present in the infested area. In the latter part of the dry season its limitations in locating game, except such as may visit its haunts, seem to be obvious, and if, as is probable, this is one of the most important breeding periods, the effect of both shooting and rinderpest seems clear enough. It may seem improbable that a species of insect should be so ill-provided for taking care of itself under varying conditions as the views outlined in this paper suggest, but if there is one thing more striking than another about the tsetse it is just this disability to adapt itself to a change in its environment, otherwise it would not have retreated before the advance of civilisation, in the absence of any direct effort to drive it back. Summary. In the foregoing pages the following tentative views have been put forward in the hope that they may receive either confirmation or criticism from other investigators : (1). The latter part of the dry season is probably, in general, the most favourable portion of the year to the fly. (2). The later the onset of the rains the greater the capacity for increase. (3). Seasons of unusually heavy rainfall are inimical, at least in certain areas. (4). A series of years of low rainfall is favourable. (5). Regions where the fly commonly attains the greatest concentration of numbers are the regions where it is most liable to be reduced or to die out, namely, parts of the infested areas where first-grade foci predominate. 132 RUPERT W. JACK. (6). In Southern Rhodesia the passage of grass fires has, as a rule, no appreciable efiect on the numbers of the fly. (7). Mopani belts are probably of considerable value to the fly in providing limited areas attractive to game in the wet season, where the grass remains short; they possibly also afford a convenient refuge from grass fires in the dry season. (8). The fly does not migrate under the stimulus of hunger or in company with game. (9). The fly as a whole does not follow game about infested areas. (10). The maximum distance at which a hungry fly readily detects its hosts is a short one, possibly less than 100 yards. (11). The maximum following distance of the females remains to be determined ; the writer is not yet convinced that the females seek animals and human beings only for the purpose of feeding, though this appears probable. 133 NOTES ON THE MOSQUITOS OF MADAGASCAR, MAURITIUS AND REUNION. By F. W. Epwarps, (Published by permission of the Trustees of the British Museum.) Through the kindness of Mons. E. Séguy, of the Muséum d’Histoire Naturelle, Paris, I have recently been enabled to examine a considerable number of mosquitos from Madagascar and the neighbouring French islands, including the types or co- types of all the species described from Madagascar by Ventrillon. As the examination of this material has disclosed the existence of several species hitherto unrecorded from these islands, as well as some new synonymy, it has been thought desirable to record the results at once. In the following pages all the species of mosquitos hitherto known from this region are dealt with. The total number is only 28, as compared with about 180 in the African fauna, so that it is highly probable that further collecting would produce many others. Anopheles mauritianus, Daruchy de Grandpré & d’Emmerez de Charmoy. Ambatofanghera and Ambohipomana (Bouet, 1905); Tananarive (Nevret and Ven- trillon, 1904 ; Bouet 1905). ; The commonest Anopheles in Mauritius, according to d’Emmerez. Anopheles maculipalpis, Giles. Occurs rarely in Mauritius (d’Emmerez de Charmoy). Anopheles squamosus, Theobald. Cellia tananariviensis, Ventrillon. According to Ventrillon this is by far the commonest Anopheles in Madagascar, and his statement is certainly borne out by its abundance in the collections which Ihave examined. No constant difference was observable between Madagascan and African specimens. Anopheles pharoensis, Theobald. A single specimen, much damaged but undoubtedly this species, from Tananarive (Ventrillon, 1905). Anopheles funestus, Giles. Four females from Moratsiazo, Lac Itasy (Bowet, 1904). Anopheles marshalli, Theo. Females from Tananarive (Nevret, 1904, and Ventrillon, 1905) and Ambohipomana (Bouet, 1905). Anopheles transvaalensis, Carter. Two females from Tananarive (Bouet, 1905). 134 F. W. EDWARDS. ( Anopheles costalis (Loew) Theobald. Females from Mauritius, where, according to d’Emmerez, the species is common, are in the Paris Museum. Aedes (Stegomyia) albopicta, Skuse. Stegomyra lambert, Ventrillon, Bull. Mus. Paris, x, p. 552 (1904) and Arch. Parasit. ix, p. 441 (1905). Stegomyva scutellaris, Theobald et auct. (nec Walker). Ventrillon’s types included both sexes, so that the identification of S. lamberti is beyond doubt, notwithstanding the fact that Ventrillon states that the female claws are toothed, whereas in all specimens I have examined they are simple. Hither Ventrillon made a mistake, or the species is variable in this respect, which is not unlikely. This species also occurs at Réunion as shown by specimens in the Paris Museum, collected at St. Denis, v-vi, 1913 (Surcouf). It has been recorded from Mauritius, where it is‘said to be very common (d’Emmerez de Charmoy). It is surprising that it does not seem to occur on the African mainland, since it has such a wide range throughout the Oriental region. Aedes (Stegomyia) argenteus, Poiret (fasciata, Fb.). Culex insatiabilis, Bigot, Ann. Soc. Ent. France, (3) vu, p. 118 (1859). Bigot’s description and figurecan apparently only apply to this species, which is common on the islands, especially near the coast (d’Hmmerez de Charmoy). Aedes (Skusea) cartroni, Ventrillon. Stegomyia cartroni, Ventrillon, Bull. Mus. Paris, xii, p. 143 (1906). This species, which I had not previously seen, evidently belongs to the sub-genus Skusea, although the male is unfortunately unknown. It is very close to S. pembaensts, the type of the subgenus, so close indeed that it may eventually prove to be a variety only. However in all the examples of S. pembaensis received at the British Museum the abdomen is unbanded dorsally, and the scutellar scales are black, while in the few specimens of S. cartroni sent the abdominal segments have narrow pale basal bands, and the few remaining scutellar scales are white. Aedes (Ochlerotatus) nigeriensis, Theobald. Culex fowlert, d’Emmerez de Charmoy, Ann. Trop. Med. u, p. 258 (1908). Mr. H. F. Carter informs me that he has examined a male of C. fowleri from Mauritius, and established its identity with O. nigeriensts, which was already sug- gested by the author’s allusion to the two pale spots on the sixth abdominal tergite of the female. A female in the Paris Museum from Madagascar may be either this species or O. fryer, Theo. NOTES ON THE MOSQUITOS OF MADAGASCAR. 135 Orthopodomyia arboricollis, d’ Emmerez de Charmoy. Qulex arboricollis, d’Emmerez de Charmoy, Ann. Trop. Med. i1, p. 257 (1908). Newsieadina arboricollis, Theobald, Ann. Trop. Med. u, p. 297 (1908). This species, which is so far only known from Mauritius, is a typical Orthopodomyra. I have examined the types in the Liverpool School of Tropical Medicine and find that the female has the very short fourth tarsal joint on the front and middle legs, characteristic of this genus, while the male hypopygium is very similar to that of the other species. Taeniorhynchus (Mansonioides) uniformis, Theobald. Females of this species in quite recognisable condition from Ambatofanghera (Bouet, 1905). Other specimens, probably of T. (M.) uniformis, were from Morous- lava (Dr. Petit, 1901). Taeniorhynchus (Coquillettidia) flavus, Ventrillon. Culex flavus, Ventrillon (nec Motschulsky), Bull. Mus. Paris, x, p. 550 (1904). Culex grandidiert, Blanchard, Les Moustiques, p. 627 (1905). Several females, including cotypes of Ventrillon, from Tananarive (Dr. Ventrillon, 1905). This is very close to the African 7’. fuscopennatus, of which it may be a mere local form ; the hypopygia are very similar, as is shown by a drawing taken from Ventrillon’s male cotype which M. Sécuy kindly sent me; this indicates the clasper as having a rather different shape. J’. flavus, however, in all the examples I have seen, is readily distinguishable by the coloration of the scales on the wing, particularly the apical half. The costa is yellow, with a line of black scales on the outer margin on its basal two-thirds ; subcostal and first longitudinal veins predominantly yellow, apically entirely so ; stems of fork-cells and tips of veins round wing-tip yellow ; remaining veins and fringe mainly black. This description is considerably at variance with that of Ventrillon, but an examination of the cotypes showed that his was inaccurate. T. fuscopennatus differs in having the dark and light wing-scales fairly evenly mixed, and also in having the integument of the thorax blackish or dark brown, that of 7. flavus being light brown. T. aureus, Edw., has an almost greater resemblance even than 7’. fuscopennatus, but it again has a rather differently shaped clasper, and has the costa entirely yellow. I have adopted the name flavus rather than grandidiert, because the Culex flavus of Motschulsky is probably an Aéles, Ventrillon’s later flavus being therefore valid in the genus Taeniorhynchus. Culex ventrilloni, sp. n. 2. Head scales narrow, blackish ; a rim round the eyes and a longitudinal band down the middle pale golden. Proboscis with a distinct yellow median ring, broader below than above. Palpi one-fourth as long as the proboscis, black, with some white scales about the middle. Thorax blackish brown, bristles dark. Prothoracic lobes with a few narrow golden scales ; mesonotum with golden and dark brown scales, the former occurring in an indefinite line in the middle and in some large 136 F. W. EDWARDS. rather irregular patches on each side on the front half. Scutellum pale, with narrow pale golden scales. Pleurae with four or five small patches of pale ochreous flat scales. Abdomen dark brown, tergites with white basal lateral patches and ochreous basal bands, which are broadest in the middle; sternites with dark apical bands. Legs black ; all the femora black to the base above, ochreous beneath on the basal half. Femora and tibiae all with conspicuous whitish tips, no scattered pale scales. Tarsal joints conspicuously ringed with ochreous at the base only ; the rings on the last two joits very narrow. Claws simple; empodia large, nearly as long as the claws. Wings with brown scales, those in the lateral series linear. First fork-cell more than twice as long as its stem, its base nearer the base of the wing than that of the second. Cross-veins separated by rather more than the length of the posterior. Length, 6 mm. Mapacascar : Tananarive, 1904, 19 (Dr. Neiret) and 1905, 292 (Dr. Ventrillon). Type in the Paris Museum ; paratype in the British Museum. This species is noteworthy for the ornamentation of the head, the basally ringed tarsal joints, and the large empodia. It has no near ally among the described African and Oriental species, and in the absence of the male it is impossible to say in what. subgenus it should be placed. Possibly it may be a Lutzva. Culex annulitarsis, Macquart. The brief description of Macquart will not apply to any known species. It was. referred by Theobald to the synonymy of Stegomyia fasciata, but Macquart’s descrip- tion of the legs (hind tibiae with a broad white ring before the tip ; hind metatarsi white, with a narrow black ring) makes the reason for this far from obvious. Culex (Lutzia) tigripes, Grandpré & Charmoy. Tamatave (Jaillet). Described from Mauritius. Culex (Culex) quasigelidus, Theobald. Pseudoheptaphlebomyia madagascariensis, Ventrillon. This synonymy, previously given by me, can be confirmed from an examination of the types, although Ventrillon’s description does not agree. Additional specimens. are from Ambohipomana (Bouet, 1905) and Tananarive (Waéerlot, 1916). Culex (Culex) giganteus, Ventrillon. Culex giganteus, Ventrillon, Bull. Mus. Paris, xn, p. 100 (1906). Culex nevreti, Ventrillon, Bull. Mus. Paris, xi, p. 103 (1906). The original description of C. giganteus and C. neireti are very similar, except as regards the size given and the markings of the thorax. Study of the types shows that the insects are actually the same, the differences in the description being due to inaccurate observation. In particular, Ventrillon’s measurement of 10 mm. body length for C. giganteus is a considerable exaggeration, the actual length being only about 6 mm., excluding the proboscis. Additional specimens are from Ambatofanghera and Ambohipomana (Bouet, 1905). The species is a close ally of the African C. annulioris, Theo., differing chiefly in the greater extension of the yellow markings of the abdomen; and in some small details of the structure of the male hypopygium, notably the shorter and stouter NOTES ON THE MOSQUITOS OF MADAGASCAR. 137 basal arms of the tenth sternites. Perhaps the single specimen of C. annulioris recorded by d’Emmerez de Charmoy (Ann. Trop. Med. u, p. 260) may have been C. giganteus. Culex (Culex) sitiens, Wied. Culex ronaldi, d’Emmerez de Charmoy, Ann. Trop. Med. ii, p. 259 (1908). Recorded from Mauritius, but not from Madagascar, where however it is very likely to occur, since it has a wide distribution from East Africa to Australia. Culex (Culex) univittatus, Theobald. Heptaphlebomyia montforti, Ventrillon, Arch. Parasit. ix, p. 448 (1905). My previous statement as to the identity of C. uwnivittatus and H. montforti is confirmed by an examination of the male hypopygium of one of Ventrillon’s specimens. Culex (Culex) pipiens, L. Tananarive (Nevwret, 1904 ; Ventrillon, 1905). This record forms a further notable extension of the known range of this species, which until recently was thought to be confined to the temperate regions of the northern hemisphere. It is now known also from South and East Africa and from the Argentine. Culex (Culex) ? laurenti, Newstead. Some female specimens from Tananarive (Neiret, 1904) are perhaps this species, though it is possible that they may belong to the variety of C. pipiens with an unbanded abdomen, which is known from East Africa and from the Mediterranean region. Culex (Culex) fatigans, Wied. Culex cartroni, Ventrillon, Bull. Mus. Paris, xi, p. 429 (1905). Culex anxifer, Bigot, Ann. Soc. Ent. France, (3) vil, p. 117 (1859). The hypopygium of Ventrillon’s type male was mounted and proved it to be this species. Other specimens from Réunion (Surcouf). The commonest mosquito in Mauritius (d’Emmerez de Charmoy). Bigot’s description of C. anxifer is unrecognisable, and he had no type, but Blanchard has referred his name to the synonymy of C. fatigans, and there seems no sufficient reason to dispute this. Culex (Culex) argenteopunctatus, Ventrillon. Heptaphlebomyva argenteopunctata, Ventrillon, Arch. Parasit. ix, p. 446 (1905). Heptaphlebomyva kingi, Theobald, Ann. Trop. Med. vii, p. 601 (1913). A number of females, all in poor condition, some barely recognisable, from Tananarive (Ventrillon, Bouet, Guérin Méneville, Waterlot). A male in the British Museum named by Ventrillon lacks the abdomen, but fortunately there is another, in perfect condition, from the Gold Coast (Dr. A. Ingram). I have mounted the hypopygium of this and find that it is a true Culex bearing a rather considerable resemblance to C. decens, Theo. The two silvery spots on the thorax are remarkable and absolutely diagnostic, being comparable only with certain species of Aédes, e.g., punctothoracis, Theo. 138 F. W. EDWARDS. Uranotaenia neireti, sp. n. ¢S Head scales and integument blackish, a spot of pale blue in the middle in front, and another on each side at the eye-margin. Proboscis slightly longer than the abdomen, slightly swollen at the tip. Thorax brown, lower half of pleurae somewhat lighter. Prothoracic lobes with light blue scales; proepimera bare, lower half blackish. Mesonotum with narrow dark brown scales, those on the scutellum flat ; no lateral line of flat scales. A spot of pale blue scales in middle of pleurae, and a few more scattered on sternopleura ; integument darker on each side of the blue spot. Abdomen dark brown above, lighter brown beneath. Legs dark brown ; coxae and bases of femora ochreous ; tip of third and whole of fourth tarsal joint of hind legs whitish; fifth darker. Tarsi and claws normal. Wings with brown scales only, those towards the apex rather narrowly spatulate. Upper fork-cell more than half as long as its stem. Length, 3 mm. MapacascaR: 1 $1 9 (Dr. Ventrellon), without exact locality, labelled “ Ficalbia neirett. Type: Ventr.” Type g in the Paris Museum. The female and a second male from Tananarive (Ventrillon, 1905) have the blue scales replaced by greyish white, probably owing to fading. Eretmopodites quinquevittatus, Theobald. Eretmopodites condei, Ventrillon, Arch. Parasit. ix, p. 144 (1905). No males were present, and the few females in Ventrillon’s collection were rather rubbed, but the above synonymy, previously suggested by me, is highly probable. 139 NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE (SUB-FAMILY PANGONIINAB), WITH DESCRIPTIONS OF A NEW GENUS AND NEW SPECIES. By Major E. E. Austen, D.S.O. Some six years ago, it was correctly pointed out by Bréthes (Bull. Soc. Ent. de France, 1914, p. 59) that the generic designation Diatomineura, Rond. (Archiv. Zool. Anat. Fisiol. ii, p. 84, 1864), is a synonym of Osca, Walk. (Ins. Saund., “Diptera, Part i, p. 10, 1850—nec Walk., 1864, nec Stal, 1871). The first species mentioned by both Walker and Rondani is Pangonia depressa, Macq. (=Erephopsis lata (Tabanus latus), Guér.), and this species is designated by Bréthes as the genotype. Erephopsis lata, Guér., of Kertész’s Catalogus Dipterorum (iii, p. 165, 1908), must therefore be known henceforth as Osca lata, Guér., and the South African Tabanus barbatus, L. (Pangonia barbata, Auct.), and Pangonia fulvifascia, Walk., must be transferred to the genus Osca. It is not, however, simply a question of substituting Osca for Diatomineura, since Osca lata, Guér., is certainly not congeneric with any, or at any rate with the majority of the species at present grouped under Diatomineura, sub-genus Corizoneura. So far as regards Ethiopian and Oriental species, the difficulty can be met by, as is hereby proposed, raising Corizoneura, Rond. (loc. cit., p. 85) to generic rank, desig- nating as its genotype the first species mentioned by Rondani under Corizoneura, viz., Tabanus aethiopicus, Thunb. (syn. Pangonia appendiculata, Macq.), and erecting a new genus, which may be termed Buplex,* for certain Ethiopian species included by Kertész under the Corizoneura division of Diatomineura, but not congeneric with C. aethiopica, Thunb. The arrangement indicated may be expressed in tabular form as follows. Eyes bare ; first posterior cell open. Ocelli absent ; face, especially in 9, markedly produced (forming a snout-like prolongation), with a shining callus, or at least a more or less shining area, on each side ; proboscis long or very long, usually much longer than thorax including scutellum, and generally horizontal or nearly so; distal extremities of first and second joints of front tarsus in g generally each produced above into a tongue- or lappet-like process, which, in case at least of second joint, is often of considerable relative length be ‘ .. Corizoneura, Rond. (Genotype, Tabanus ens Thunb. —syn. a ests aethvopica, Thunb. ; Pangoma appendiculata, Macq.). Ocelli usually present ; face not produced, merely convex, and without shining callus or area on each side; proboscis of only moderate length, shorter than or about as long as thorax including scutellum, and generally slanting downwards at an angle of 45°; distal extremities of tarsal joints never produced .. as Buplezx, gen. nov. (Genotype Pangoma suavis, Lw.). * Bou-rdné 6, aN Ox-goad. 140 E. E. AUSTEN, Among other species referable to the new genus thus briefly characterised are Pangonia brunnipennis, Lw., P. subfascia, Walk., Corizoneura albifacies, Ric., CO. dissimilis, Ric., and the new species described below. The geaus Corizoneura, Rond., as defined and restricted above, includes, in addition to the genotype and the new species described in this paper :—Diatomineura virgata, Austen ; D. inornata, Austen ; D. neaver, Austen ; Pangonia sagittaria, Surc.; Corizoneura distincta, Ric. ; Pangonia lateralis, Wied. (Fabr. ?); Diatomineura hastata, Austen; D. lineatithorax, Austen; D. penetrabilis, Austen ; Corizoneura pallidipennis, Ric. ; and C. umbratipennis, Ric. The foregoing all belong to the Ethiopian Region, and congeneric with them are the Oriental Pangonia taprobanes, Walk. (syn. P. rufa, Macq.), and P. longirostris, Hardw. Another change which may as well be made now, since in any case it is almost certain to be effected sooner or later, is the replacement of Macquart’s generic designation Cadicera by Phara, Walk. (Ins. Saund., Diptera, Part i, p. 9, 1850). The latter name, accompanied by a brief diagnosis, was applied by its author to one of “* several groups or subgenera” of Pangoma, Latr. ; eighteen species were mentioned under Phara, the first three of these being, in the order given, Pangonia melanopyga, Wied., P. chrysostigma, Wied., and P. crassvpalpis, Macq. The present writer hereby proposes to regard Pangonia melanopyga, Wied., as the genotype of Phara, Walk. ; to raise the latter to generic rank ; and to restrict it so as to include besides the genotype, among the species mentioned under Phara by Walker, only Pangonia chrysostigma, Wied., and P. crassipalpis, Macq., since these three species are obviously congeneric. Phara, Walk., therefore, as thus restricted, replaces and must be substituted for Cadicera, Macq. (Mém. Soc. Imp. des Sc. Lille, 1854, (2) p. 42, 1855), founded for C. rubramarginata, Macq. (loc. cat, p. 23). Except where otherwise stated, the types of the new species described in the following pages are in the British Museum (Natural History). Genus Buplex, nov. Buplex fuscinervis, sp. n. 0,.—Length (2 specimens) 12 to 13 mm. ; width of head 4 to 4:2 mm. ; width of front at vertex 0°8 mm. ; length of proboscis 3°4 to 3°5 mm. ; length of wing 12°4 to 12°5 mm. Body, except certain areas at sides of abdomen, black above, with scutellum, lateral borders and three narrow longitudinal stripes on scutum, and deep posterior borders to abdominal segments smoke-grey* or pale smoke-grey pollinose ; lateral extremities of tergites of first (visible) and second abdominal segments smoke-grey, more or less tinged with ochraceous tawny owing to the subjacent ground-colour ; wings with base, costal and sub-costal cells and stigmatic area ochreous, veins after first longitudinal in distal half (from base of discal cell onwards) dark brown, and for most part more or less strongly suffused with mummy brown. *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). NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE. 141 Head : ocelli present ; face, jowls and lower half of front light buff pollinose, clothed with pale yellowish hair, occiput smoke-grey pollinose, and likewise clothed with pale yellowish hair; upper part of front with a pair of narrow, blackish stripes, commencing at ocellar tubercle, descending somewhat lower than middle and diverging below, each stripe clothed with short, curving black hairs, interspersed with some glistening yellowish hairs ; palpi clothed with yellowish hai, longer on proximal than on terminal joint, latter russet or ochraceous tawny, proximal joint mouse-grey, russet towards distal extremity ; first and second joints of antennae clothed partly with yellowish hair, partly with black hair, first jomt mouse-grey or deep neutral grey, second joint russet (third joint missing in case of type). Thorax: smoke-grey longitudinal stripes on dorsum extending from front to hind margin of scutum, confluent posteriorly ; transverse suture smoke-grey, forming a connection between lateral border and outer smoke-grey stripe on each side ; pleurae and pectus light greyish olive pollinose ; thorax clothed with yellowish hair, black stripes on dorsum clothed posteriorly with fine black hair. Abdomen: black area on first (visible) and second tergites restricted to a transverse band occupying middle third of anterior half, with rounded posterior angles in each case, and, especially on second segment, indented in middle line behind ; lateral margins of fourth and following tergites, and lateral extremities of hind border of third tergite ochraceous tawny, though normally concealed by hair and smoke-grey pollen ; seventh tergite entirely grey ; lateral extremities and hind borders of all tergites clothed with glistening hair, longer and ochreous on lateral extremities of second and two or three following segments, shorter and paler elsewhere ; black area on second tergite clothed with yellowish hair, corresponding areas on two following tergites clothed with short, erect black hair; venter smoke-grey, clothed with appressed, glistening, cream- coloured hair ; second sternite with a transversely elongate, somewhat reniform, fuscous black blotch in middle line, resting on or close to anterior border and confined to anterior half of segment, third and three following sternites each with a sooty black, transverse band on front border, widely separated in each case from lateral margins, and from twice to four times deeper on third than on either of the other sternites, these bands clothed with short, erect, black hair. Wangs with auxiliary and first longitudinal veins, second longitudinal from base to a little beyond origin of third longitudinal, praefurcal portion of fourth longitudinal, and extreme bases of fifth and sixth longitudinal veins ochreous ; veins otherwise more or less dark brown. Squamae cream-bufi. Halteres maize-yellow, knobs sometimes darker (mummy-brown) at base. Legs: coxae and femora deep mouse-grey, clothed with yellowish hair ; tibiae and tarsi clove-brown, clothed with minute black hairs, tarsal joints sometimes paler towards base. South Africa (Cape Province) ; type and para-type from Bizana, East Pondoland, 1912(—Goodall : : presented by Mr. T. B. Goodall). The species described above shows a decided resemblance to Buplex (Pangonia) suavis, Lw., but, apart from its smaller size, is distinguishable, inter alia, by the presence of the two blackish stripes on the front ; by the outer two of the three longitudinal smoke-grey stripes on the disk of the scutum being narrower ; by the absence of a brown blotch near the tip of the wing, above the fork of the third longi- tudinal vein; and by the veins in the region of the discal cell being suffused with mummy-brown. 142 E. E. AUSTEN. Genus Corizoneura, Rond. Corizoneura formosa, sp. n. o.—Length (4 specimens) 20 to 21:4, mm. ; width of head 6 to 6°4 mm. ; distance from upper margin of occiput to anterior extremity of face 5 to 5°25 mm. ; length of proboscis 13 to 15 mm. ; length of wing 17-4 to 18 mm. ; wing-expanse 21°5 to 23 mm. Dorsum of thorax dark olwwaceous black, with broad smoke-grey lateral borders clothed with cream-coloured hair, and with two narrow, smoke-grey, admedian, longitudinal stripes ; dorsum of abdomen ochraceous tawny, with distal extremity (fifth to seventh segments) infuscated (iron-grey to olwaceous black), a large black median blotch on each of the first four visible segments, and fourth segment bordered posteriorly with appressed, silvery white and ochreous hair ; venter cinnamon-buff or connamon-coloured, with last three segments sometimes mouse-grey and lateral extremities of first visible scute neutral grey ; wings with a sepraceous tinge ; processess at tups of firstand second joints of front tarsus in 3 very long. Head : face bluntly conical, moderately produced ; front and face (except sides of facial prominence below, which are clove-brown and shining) pinkish buff pollinose, front and borders of face adjacent to eyes clothed with longish, cream-buff hair ; occiput smoke-grey, clothed above with cream-buff, below with whitish hair ; jowls and basi-occipital region pale smoke-grey, clothed with long whitish hair ; palpi russet, dark brown on outer side, proximal joint with a bunch of whitish hair at base below; first and second joints of antennae isabella-coloured pollinose, clothed above and below with longish hair, generally black or blackish but sometimes mainly cream- coloured on first joint, third joint ferruginous or vinaceous rufous, clove-brown at tip. Thorax: dorsum, including scutellum, clothed for most part with somewhat appressed, cream or cream-buff-coloured hair, often but little visible when regarded from above, hair above lateral borders between bases of wings often largely or mainly dark brown, pleurae and pectus clothed with fairly long, whitish or yellowish white hair. Abdomen: except on first (visible) tergite, on which median blotch extends’ to or is but narrowly separated from hind margin, black median blotches, which rest on front margin, do not reach hind border ; posterior angles of first and second tergites clothed with yellowish or whitish hair, hind border of second tergite with a patch of glistening, appressed, ivory-yellow or cream-bufi-coloured hair in middle line, light-haired hind border of fourth tergite expanded in middle line and at each extremity, lateral extremities of sixth and seventh tergites generally clothed with glistening silvery white hair, lateral extremities of second to fourth tergites inclusive, except as already stated, clothed with black hair, ochraceous tawny area of first and second tergites clothed mainly with minute, glistening, appressed, ochreous hairs, corresponding area on third and fourth tergites clothed with minute black or blackish hairs, fifth and sixth tergites (except lateral extremities in case of latter) clothed with black hair; venter clothed with minute, appressed, glistening, cream-bufi- coloured hairs, fifth sternite sometimes largely clothed with minute, black or blackish hairs. Wings: veins sepia-coloured, adjacent membrane suffused with same colour at base, and to a less extent at level of proximal extremity of discal cell. Sqwamae waxen ochraceous buff or pale orange-yellow, fringed with minute yellowish hairs. Halteres light chestnut-brown, tips of knobs light buff or light ochraceous buff. Legs: coxae neutral grey, clothed with whitish hair, which is longer on front pair ; NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE. 143 femora russet (extreme tips ochraceous buff), clothed mainly with whitish or yellowish white hair, antero-inferior area of front pair with minute, erect, blackish hairs ; front and middle tibiae ochraceous buff, clothed with minute, glistening, yellowish or ochreous hairs, hind tibiae and hind tarsi russet-brown, clothed with black hair, distal extremities of hind tarsal joimts blackish ; front tarsi ochraceous buff or ochraceous tawny, distal extremities of last three joints dark brown, body of second joint very short, tongue-like process at its distal extremity very long (longer than following joint), process at end of first joint overlapping and closely applied to first two-thirds of following joint ; first joint of middle tarsi cream-buff, remaining joints dark brown or mummy-brown, blackish brown at tips. South Africa (Cape Province): type and three para-types from Deelfontein, 7.xi1.1902 (presented by Colonel Sloggett, A.M.S.—now Lneut.-General Sir A. T. Sloggett, K.C.B., K.C.M.G.). This handsome species presents a decided resemblance to Corizoneura aethiopica, Thunb., another inhabitant of the South African portion of the Ethiopian Region, but can at once be distinguished, inter alia, by the presence of the conspicuous, smoke-grey, admedian, longitudinal stripes on the dorsum of the thorax. Corizoneura schwetzi, sp. n. 3 9.—Length, ¢ (6 specimens) 16°4 to 18 mm., 2 (6 specimens) 16 to 18°6 mm. ; width of head, g just over 5 to 5°4 mm., 9 52 to 5°75 mm.; distance from upper margin of occiput to anterior extremity of face, f 4:2 to 45 mm., 9 45 to 48 mm. ; width of front of 2 at vertex 0°6 to 0°75 mm. ; length of proboscis, 3 15 to 15°25 mm., 9 11°75 to 13°75 mm. ; length of wing, ¢ 15 to 15°2 mm., 2 15°5 to 17°25 mm. Dusky species, with base of abdomen paler; dorsum of thorax dark brownish olive, with a broad, indistinct, blackish brown longitudinal stripe along each side above lateral margin, a narrow, elongate, pale spot (composed of Naples yellow or cream-buff hair) above base of each wing wm front of postalar callus, and a short but conspicuous stripe of black hawr between base of costa and humeral callus on each side ; dorsum of abdomen with first two (visible) segments connamon-buff or tawny olive in 3, ochraceous tawny m 2, 3 with a blackish brown median triangular spot at base of second segment, tergites of third and following abdominal segments blackish brown or black, fourth segment posteriorly with a conspicuous transverse band of appressed, glistening, silvery white hair ; venter abruptly bicoloured, proximal portion as far as base of third (visible) stermite, wory-yellow, cream or cream-buff-coloured, remainder blackish brown ; wings strongly and uniformly tinged with sepia. Head drab pollinose, posterior orbits paler (light greyish olive above, smoke- grey or pale smoke-grey below), shining callus on each side of face below antenna black, sharply defined, sides of face below calli more or less dark brown or blackish brown owing to ground-colour showing through pollinose covering, lower border of sides of face in ¢ shining blackish brown, deeper posteriorly, distal extremity of face in @ shining blackish brown, sparsely clothed with drab pollen, in both sexes a more or less distinct, somewhat triangular, olive-buff or pale olive-buff, pollinose spot on each side of lower part of front, between base of antenna, margin of eye and shining callus ; front in Q with following series of dark brown marks— (687) D 144 E. E. AUSTEN. a transversely oblong spot occupying ocellar region of vertex but not extending to eyes, a pair of somewhat guttate spots in centre of front, indistinctly connected with foregoing spot but likewise not in contact with eyes, and an oblique spot on each side below, extending from base vf antenna to eye, above the light (olive-buff or pale olive-buff) spot already mentioned ; front in ¢ clothed with black hair, in 2 with shorter blackish hair, mingled with paler hair in upper portion, posterior orbits in both sexes fringed above with yellowish hair and below with longer whitish hair, basi-occipital region and jowls densely clothed with fairly long, whitish hair, lower borders of sides of face clothed posteriorly with black or blackish hair, more conspicuous in g than in 2; palpi russet-brown in J, russet in 9, terminal segment (elongate and very narrow in 4) clothed with minute black or blackish hairs, proximal segment, especially m Q, clothed below with longer hair, brownish or blackish brown anteriorly, yellowish or whitish posteriorly ; antennae russet, first and second joints more or less smoke-grey pollinose, clothed with black hair. Thorax: dorsum clothed anteriorly with silky, tawny olive or ochreous hair, posteriorly, including base of scutellum, with fine, erect, black hair, postalar calli, and scutellum, except as stated, clothed with Naples yellow or cream-buff hair, hairy covering of central portion of dorsum visible only when viewed from side ; pleurae and pectus clothed with longish hair, for most part cream-buff (whitish on propleurae) in colour, a tuft of black hair (less conspicuous in Q, in which sex it is often much reduced) below base of wing on each side ; sternopleurae in 3 sometimes clothed mainly with black or blackish hair. Abdomen: hind border (distal third or rather less) of second (visible) tergite clothed with minute, appressed hairs, which are silvery white or whitish at and towards lateral extremities of segment, and sparser and more yellow- ish (often glistening cream-buff) in vicinity of median line (owing to paler ground colour, whitish transverse band thus formed is, except at lateral extremities, less ‘conspicuous than corresponding band on fourth segment); base of first (visible) tergite with a blackish brown, median area, extending bevond scutellum, but not or barely reaching hind margin except in middle line in 4, in which sex it is more or less distinctly connected with triangular spot on second segment; blackish brown median triangle on second tergite in ¢ with its base resting on or close to anterior margin, and its apex reaching or extending somewhat beyond middle of segment ; third tergite sometimes irregularly paler (cinnamon-brown) at base and on hind border ; hairs in silvery white transverse band on fourth tergite somewhat yellowish in vicinity of middle line; seventh tergite in Q, as well as frequently hind borders of fifth and sixth tergites, often dull fuscous ; lateral extremities, or at least posterior angles, of fifth and sixth tergites each clothed with a prominent tuft of silvery white hair; posterior angles, as well as in 9 hind margin of first (visible) tergite clothed with shining ochreous or pale ochraceous orange hair ; dorsum of abdomen except as already stated clothed with minute, appressed, black hairs ; first (visible) sternite bare, second sternite clothed with minute, appressed, glistening cream-coloured hairs (anterior border of second sternite fringed with fine, erect, black or blackish hair, shorter and less developed in @ than in 3, in which sex fine, recumbent, black hairs are also present among the cream-coloured hair, especially towards lateral extremities of the scute), blackish brown portion of venter clothed with black hair, among which a few pale hairs are occasionally present NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE. 145 on one or more of fourth and following segments. Wangs: appendix to anterior branch of third longitudinal vein, though of variable length, as a rule relatively somewhat long. Squamae isabella-coloured. Halteres mummy-brown, knobs in 2 often paler (cream-buff) at tip. Legs: front coxae neutral grey, clothed with cream-coloured or cream-buff hair, middle and hind coxae deep neutral grey or dark neutral grey, clothed with black hair, hind coxae in 9 also with some yellowish hairs ; hind femora in ¢ and bases of front and middle femora in same sex blackish brown, front and middle femora in ¢ except at base russet-brown or cinnamon- brown, all femora in ¢ clothed with black hair, femora in Q paler, and clothed largely with ochraceous buff or ochreous hair as well as with black hair ; coloration of tibiae and tarsi and of their hairy covering alike in both sexes, front and middle tibiae ochraceous buff or ochraceous tawny, clothed with minute, appressed glisten- ing ochraceous buff hair, hind tibiae and hind tarsi blackish brown, clothed with black hair, tips of second and two following tarsal joints clothed below with ferruginous hair, similarly coloured hair also largely present, at least in 9, on under side of first tarsal joint, front and middle tarsi ochraceous tawny, clothed above with black hair, last joint in each case, as well as distal extremities of preceding joints mummy-brown, processes at tips of first and second joints of front tarsi in S$ of moderate size, in neither case reaching distal extremity of following joint. Belgian Congo (North Katanga) and Tanganyika Territory. Typical series from Kakanu (between 15 and 16 miles south of Kisengwa, R. Lomami), N. Katanga, vi. 1918 (Dr. J. Schwetz). Type of 3, type of 9, 36 gd and 2 9 para-types, in Musée Royal d’Histoire Naturelle de Belgique (Brussels) ; 6 J and 6 2 para-types, in British Museum (Natural History)—presented by M. G. Severin ; 1 9, from Tangan- yika Territory (formerly German East Africa), 30° 55’ E. Long., 2° 5’ S. Lat., 16. vi. 1916 (Dr. G. D. H. Carpenter), in British Museum (Natural History), presented by Imperial Bureau of Entomology. This fine species, with which the author has much pleasure in connecting the name of its discoverer (the well-known student of tsetse-fly bionomics in North Katanga), was met with by the investigator in question in large numbers in the vicinity of Kakanu.* On 6th June 1918, in a belt of forest several hundred metres in breadth and about 33 miles from Kakanu, Dr. Schwetz’s native carriers succeeded in catching some 2,000 specimens of C. schwetzi, from 80 to 90 per cent: of which were males.f Unlike Corizoneura inornata, Austen (see below), which is found in the open, C. schwetzi does not occur outside the forest. According to Dr. Schwetz,t in the case of the present and the following species (C. inornata, Austen) at least, the labium itself is the piercing organ, and in the act of biting is thrust by the insect deeply into the skin of its victim. It is therefore interesting and possibly suggestive to note that, in the Q taken by Dr. Carpenter in Tanganyika Territory, the labium * Of. Schwetz, ‘‘ Dix Jours d’Observations sur les Moeurs de la ‘ Pangonia zonata’ et de la ‘ Pangonia oldii’ (Deuxiéme Note)”: Revue Zoologique Africaine, vii, pp. 92-106 (1919).—Cf. also the earlier paper by the same author, ‘* Quelques Observations Pré- liminaires sur les Moeurs de la ‘ Pangonia zonata’’’: ibid., pp. 46-54. In both of the memoirs cited Corizoneura schwetzi is referred to as Pangonia oldii, while the species termed Pangonia zonata is really Corizoneura inornata, Austen. + Cf. Schwetz, loc. cit., p. 103. t See below, p. 147. (687) D2 146 EK. E. AUSTEN. only projects 7°5 mm. beyond the extremity of the clypeus, exceeding the length of the labrum-epipharynx and the other mouth-parts by little more than the labella, while the proximal portion of the labium is bent backwards underneath the head, and beneath the cleft between head and thorax forms an angle of 45° with the distal portion. Although superficially presenting a decided resemblance to Pangonia oldi, Austen, Corizoneura schwetz can be distinguished from that species inter ala by the processes (entirely wanting in P. oldw) at the tips of the first and second joints of the front tarsi in the 9; by the much greater development of the dark spot (often scarcely more than vestigial in P. oldi) at the base of the second abdominal tergite in the - same sex ; and in both sexes by the short stripe of black hair on each side of the dorsum of the thorax, in front of the base of the wing. From Corizoneura inornata, Austen, apart from obvious differences in coloration, especially the sharply bicoloured venter of the species just described, C. schwetzi is distinguished by the inferior development of the processes at the tips of the first and second joints of the front tarsi in the g. Whereas in C. inornata 3 each of these processes is so long as to project beyond the tip (excluding the process in the case of the second segment) of the following joint, in C. schwetzi 3 neither process reaches the tip of the succeeding joint. Corizoneura inornata, Austen. Diatomineura mornata, Austen, Bull. Ent. Res. 1, p. 282 (1911). This species was described from a single 9, obtained in September 1907, in S. Katanga, Belgian Congo, between Bunkeya and Kambove, at an altitude of 3,500 ft., by Dr. Sheffield Neave. The kindness of Dr. J. Schwetz, in presenting to the British Museum (Natural History) a large number of specimens of both sexes, now renders it possible to indicate the distinctive characters of the ¢. C. inornata, Austen, g.—Apart from ordinary secondary sexual characters, agreeing generally with the 2 except as follows. Head: hair clothing jowls and basi-occipital region often hoary or nearly white; first two joints of antennae clothed mainly with black hair. Thorax: pleurae on each side with a tuft of black hair below base of wing. Abdomen: first (visible) tergite with a median brownish black area at base, projecting somewhat beyond scutellum, but not reaching hind margin ; second tergite with a conspicuous, median, brownish black, triangular spot resting on front margin, and varying in size in different individuals, but not extending beyond middle of segment, if so far; fifth and sixth tergites mainly brownish black, blackish brown or clove-brown, each of the two preceding tergites often with an ill-defined median blotch of same colour occupying anterior two-thirds. Legs : front and middle femora clothed largely with black hair, at least towards base, hind legs clothed mainly with black hair, processes at tips of first and second joints. of front tarsi very long, in each case projecting beyond end of following joint (excluding process in case of second segment). In the papers already referred to,* under the name “ Pangonia zonata,” Dr. Schwetz has furnished a series of interesting field notes on the behaviour of this * P, 146, note.* NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE. 147 species, which, at the end of May 1914 and at the close of the same month and beginning of June 1918, was found by him in great abundance in the vicinity of Kakanu, N. Katanga, at and about the flowers of Acanthus montanus (termed by the natives ‘‘ Nafimbia’’), which grows in sheets in open spaces outside the forest. According to Dr. Schwetz, the existence of C. inornata is “ intimately connected ” with that of the flowers in question, the nectar of which is imbibed by both sexes, though the females also suck blood on occasion. It is interesting to note that, having in a number of cases observed the process of sucking blood by females of this and the foregoing species (Corizoneura schwetzi, Austen), Dr. Schwetz states that, in these PANGONIINAE at any rate, the actual piercing organ is the proboscis (labium) itself, which is “‘ evidently sufficiently rigid ”’ to penetrate the human integument, and is driven by the insect “ fairly and squarely into the skin to the extent of one-third or one-half of its length. . . .” * Genus Pangonia, Latr. Pangonia discors, sp. n. ©.—Length (1 specimen) 19°6 mm. ; width of head 5°6 mm. ; width of front at vertex 0°75 mm.; distance from upper margin of occiput to anterior extremity of face 45 mm. ; length of proboscis 5°5 mm. ; length of wing 17°6 mm. Deep black, somewhat shining ; first (visible) and second abdominal tergites densely covered with pale gull-grey pollen, and thickly clothed with closely appressed, silvery white hair ; area beneath scutellum free from pollen and clothed with black hairs, some black hairs also present in middle line on anterior border of second tergite, latter likewise exhibiting a broad, median, triangular area (its base resting on front margin, its apex directed backwards and reaching beyond middle of segment) which, like front border of same segment, has a blackish look owing to pollen on it being thinner than elsewhere ; lateral extremities of second tergite clothed with black hair ; wings mummy-brown, proximal half (as far as base of discal cell) and stigma pale orange-yellow or light orange- yellow. Head : ocelli wanting ; face moderately prominent, front above relatively some- what narrow ; area from middle of front to anterior margin of clypeus olive-buff, upper half of front fuscous black, sides of face clove-brown, occiput and basi-occipital region smoke-grey or pale smoke-grey, jowls clothed with yellowish cream-coloured hair ; palpi clove-brown, terminal joint elongate, tapering to a point and somewhat curved ; first and second joints of antennae dark mummy-brown, clothed with minute black hairs mixed with some minute yellowish hairs, first joint short, expanded portion of terminal joint cinnamon-brown (terminal portion missing in case of type). Thorax: postalar calli chestnut-brown ; dorsum including scutellum clothed with short black hair, which on front border of scutellum is interspersed with minute, glistening, appressed, golden hairs, a few golden hairs also on upper portion of swelling occupying depression at each end of transverse suture, hair on pleurae and pectus entirely black or fuscous black. Abdomen: third (visible) and following tergites clothed with appressed, black hair ; venter, except second sternite, clothed with appressed, black or brownish black hair, ventral surface of second segment | * Of. Schwetz, Rev. Zool. Africaine, vii, pp. 101-102 (1919). 148 E. E. AUSTEN. clothed with minute, appressed, glistening, yellowish white hairs, fore border and lateral extremities clothed with black hairs. Wangs: transition from orange- yellow proximal to mummy-brown distal portion sharply marked, at least in case of type. Squamae ivory-yellow. Halteres ochraceous buff, stalks and knobs towards base brownish. Legs: coxae and femora dark brown or blackish brown, clothed with black hair, which on under side of hind femora is mixed with a certain number of minute, glistening tawny hairs, and on anterior surface of front coxae with a few golden hairs ; tibiae and tarsi russet, clothed for most part with minute, appressed, glistening, ochraceous tawny hairs, last three joints of tarsi dark brown above. Angola (J. J. Montevro). Pangona discors resembles and is closely allied to the East African P. beckeri, Bezzi, but is readily distinguishable oyving to, inter alia, its more prominent face and narrower upper part of the front, the entire absence of white hair on pleurae, postalar calli and front coxae, and the distal portion of the abdomen being entirely black and covered with black hair, instead of having the tip ochraceous tawny and clothed with glistening ochraceous orange or ochreous hair. Pangonia lautissima, sp. n. ¢g 2.—Length, g (9 specimens) 17 to 19 mm., Q (3 specimens) 16 to 17°4 mm. ; width of head, g just under 5 to 5'4 mm., 9 5 to 5°6 mm.; width of front of 9 at vertex just under 1 to 1-2 mm. ; length of proboscis, g just under 4 to 4°25 mm., 2 3°4 to 4-2 mm. ; length of wing, 9 14 to 15:2 mm., 9 14 to 15°6 mm. Shining black ; basi-occimital region clothed and lower halves of posterior orbits Jringed with orange-buff hair, and patches of similarly coloured hair on pleurae ; wings with base and a deep anterior border extending to end of third costal cell ochraceous orange, and remainder of surface uniformly brownish black, with a strong purplish metallic sheen. Head black, frontal triangle in § and region of subcallus (area immediately above bases of antennae) in Q shimmering silvery white pollinose, a similar pollinose — patch (clothed with a few whitish hairs, and usually more distinct and sharply defined in 9 than in Q) on each side of upper part of face in both sexes ; face in both sexes tumid below antennal prominence, then indented or somewhat receding, not produced into a nose-like prolongation, front in Q deeply furrowed ; occiput pallid neutral grey pollinose, clothed with whitish hair, posterior orbits silvery white, their upper halves fringed behind with minute blackish hairs ; palp: and antennae black, proximal joints of both sparsely clothed with short, black or blackish hairs, third joint of antennae from certain aspects appearing dark olive-grey, mouse-grey or brownish grey pollinose. Thorax: dorsum including scutellum clothed with short black hair, humeral calli inconspicuously neutral grey pollinose and clothed on sides in front with pale orange-buff hair, postalar calli fringed posteriorly below with orange-buff hair; pleurae on each side with a thick tuft of orange-buff hair below humeral callus, and more posteriorly with two further tufts of similar hair arranged somewhat in the shape of a wide V, of which the anterior branch fringes the hind margin of the mesopleura, while the posterior runs back to the squamae ; pleurae except as stated, and pectus except in front of front coxae clothed with black hair, pectus in front of front coxae neutral grey pollinose, clothed with pale orange-buff _ NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE, 149 hair. Abdomen: first six (visible) tergites in both sexes each with its lateral fourth on each side (rather more in case of 2) clothed with close-set, appressed, minute, glistening smoke grey hairs (extreme lateral extremities of seventh tergite in 9 clothed with hairs of same kind), so that dorsum of abdomen exhibits on each side a broad grey stripe, which in certain aspects contrasts fairly sharply with remainder of surface, which is clothed with minute appressed black hairs and thus forms a broad, longi- tudinal, median, black stripe ; first (visible) tergite clothed with greyish pollen on each side at base, second tergite in both sexes on each side with a roughly semi- circular whitish pollinose spot on hind margin, from most points of view concealed by the smoke-grey hair, but clearly visible when abdomen is looked at obliquely from behind, in @ a pair of similar but smaller whitish pollinose spots on hind margin of each of the two following tergites also; venter clothed with minute, appressed, glistening, smoke-grey hair, last two sternites, and sometimes median area of hind border of preceding sternite also, clothed mainly or entirely with black hair. Wings: extreme base of costa and first longitudinal vein brownish black ; ochraceous orange base extending into bases of basal and anal cells, similarly coloured anterior border including upper margin of first basal and proximal fourth of first submarginal cell, thence tapering obliquely to end of third costal cell ; veins within ochraceous orange area similarly coloured, elsewhere dark brown. Squamae light orange yellow. Halteres: stalks and knobs mummy-brown above and below, tips of knobs cream-buff. Legs black and clothed with black hair, front coxae anteriorly dark neutral grey pollinose, thinly clothed towards base with longish pale orange-buff hair. Tanganyika Territory (formerly German East Africa) : Itigi, iv, 1917 (Dr. G. D. H. Carpenter). Type of 3, type of 9, and 2 9 para-types, taken 18.iv.1917; 7 g para-types, taken 15.iv.1917, “on composite flower”; 1 3 para-type, taken 6.iv.1917, “ among low herbage.” All foregoing presented by Imperial Bureau of Entomology, which retains possession of six additional para-types, taken by Dr. Carpenter at same time and place as specimens already enumerated. The extremely striking and unusual-looking Tabanid just described belongs to the group of the genus Pangonia that includes P. elongata, Ric., P. becker, Bezzi. and P. discors, Austen, but is readily distinguishable from all three species by, apart from its sharply defined wing-markings and other characters, its unbanded abdomen and the patches of orange-buff hair on the pleurae. The coloration of the wings, in conjunction with the shining black body, would seem to suggest that P. lautissima possibly mimics some species of wasp (perhaps Rhynchium cyanopterum, Sauss.), and it is worthy of note that the same colours, though differing widely in extent and arrangement in the case of the wings, are exhibited by “ Pangonia” mesembrinordes, Surc., of which the type was also obtained in Tanganyika Territory (Amani). The latter species, however, was incorrectly assigned by its describer to the genus Pangonia, and really belongs to a new genus allied to Thriambeutes, Griinb. Pangonia carpenteri, sp. n. Q.—Length (8 specimens) 15 to 16°6 mm.; width of head 4°25 to just under 5 mm. ; width of front at vertex 0°6 mm. ; distance from upper margin of occiput to anterior extremity of face 3°5 to 4 mm. ; length of proboscis 11°75 to 13°75 mm. ; length of wing 13°5 to 15 mm. 150 E. E. AUSTEN. In Q sex, at any rate, a somewhat sombre-coloured species looking more like a small or medium-sized Corizoneura than a Pangonia, and not unlike a smaller and more dusky form of Corizoneura hastata, Austen, of Portuguese E. Africa.—Face moderately produced ; dorsum of thorax olivaceous black, with a pair of broad, paler (greyish olive), narrowly separated or sometimes confluent, admedian, longitudinal stripes ; dorsum of abdomen shining blackish brown, on each side of base rather more than lateral third of first (visible) tergite ochraceous tawny, a similar area on each side of second tergite, or of second and third tergites, usually russet or chestnut-brown, lateral borders, posterior angles, and (at least in part) hind borders of second and fourth tergites clothed with glistening silvery white hair; wings strongly tinged with sepia; legs for most part ochraceous tawny. Head: ocelli wanting ; face and front clothed with yellowish grey or isabella- coloured pollen, and front thinly covered with yellowish hair, lower half of front, above antennal prominence, with an ill-defined, shining black, median longitudinal mark, upper half of front usually streaked with mummy-brown or dark brown along each side, while a narrow mummy-brown streak, which starts from base of antenna on each side, runs obliquely upwards and outwards, and joins corresponding eye just above base of antennal prominence ; below antennae, a shining black transverse band unites and includes the shining black facial calli; jowls and basi-occipital region clothed with long, whitish hair ; occiput greyish olive or smoke-grey pollinose, sparsely clothed with cream-coloured hair, which also forms a short fringe behind posterior orbits above, lower portion of posterior orbits fringed with whitish hair ; palpi elongate, proximal joint russet-brown or deep mouse-grey, sparsely clothed with whitish hair on outer side and below, terminal joint russet, somewhat expanded in middle, clothed with minute black hairs ; first and second joints of antennae pinkish cinnamon or cinnamon-buff pollinose, both sometimes clothed with black hair though in some specimens hair on first joint is for most part yellowish, second joint usually with outstanding black hair above and below, third joint orange- cinnamon. Thorax: admedian stripes on dorsum entire, extending from front margin to prescutellar groove, outer border of each stripe paler in front of transverse suture ; dorsum including scutellum thickly clothed with fine yellowish hair, with which in some specimens on central part of posterior portion of scutum a few fine black or blackish hairs are intermixed ; postalar calli and lateral margins of dorsum behind transverse suture fringed with whitish hair, pleurae and pectus thickly clothed with similar hair. Abdomen: central portion of base of first (visible) tergite olivaceous black, basal angles of same tergite more or less distinctly neutral grey or dark neutral grey ; rather less than median third of second tergite and a broad median triangle on each of following tergites, in each case with its base resting on hind margin and its apex in contact with front margin, and on fourth and following tergites with its base expanded laterally so as to include entire hind border of segment, dull dark olive-grey ; olive-grey median area on second tergite (sometimes that on third tergite also) indented on each side ; second tergite at base with a median, semicircular, dead black spot, or in some cases with a pair of smaller spots, narrowly separated in middle line by a dark olive-grey longitudinal stripe ; first (visible) tergite with a median patch of glistening yellowish or pale yellowish hair, and clothed elsewhere with ochraceous hair; second tergite, except on black spo NOTES ON THE NOMENCLATURE OF CERTAIN AFRICAN TABANIDAE. ld1 (or spots) on which hair is usually black or ochreous, clothed with minute, appressed, glistening silvery white hairs, which however, at least on each side of median olive- grey area, are often largely replaced by similar ochreous hairs ; lateral borders of fourth and following tergites, as well as (at least in part) hind border of fourth tergite, clothed with glistening, silvery white hair, similar hair also usually visible on lateral margins of third tergite, towards posterior angles ; dorsum except as stated clothed with appressed black hair ; venter isabella-coloured or light brownish olive, clothed with minute, appressed, whitish or yellowish white hairs. Wangs: veins mummy-brown ; first posterior cell variable as usual as regards distance from hind margin at which it 1s closed, sometimes closed on margin itself, or even in one or other wing narrowly open. Squamae cream-buff. Halteres: knobs fuscous, stalks cinnamon-buff. Legs: coxae neutral grey, clothed with whitish hair, anterior and inferior surfaces of femora clothed, at least in part, with black hair, femora elsewhere clothed with yellowish hair, tibiae and tarsi clothed with minute, appressed, ochreous hairs, extensor surfaces of hind tibiae and hind tarsi clothed, at least in part, with black hair ; front femora blackish brown at base and sometimes also on greater part of under side, joints of front tarsi often mummy- brown or dark brown at tips above, those of hind tarsi similarly marked, or some- times entire upper surface of hind tarsi, except base of first joint, dark brown. Tanganyika Territory: Itigi, iv, 1917 (Dr. G. D. H. Carpenter). Type and 3 para-types, taken 18.1v.1917; 1 para-type, taken 15, iv, 1917, “ on low herbage ”’; 1 para-type, taken 6.iv.1917, “ hovering while feeding from composite flower ; looking much lke a Bombylius.” All foregoing presented by Imperial Bureau of Entomology, in whose possession are two other para-types, taken by Dr. Carpenter at same time and place as specimens already mentioned. So far as itis possible to judge from the 9 alone,this species, which is named in honour of its discoverer and does not resemble any African Pangonia hitherto described, presents, apart from the venation, all the characteristics of a Corizoneura, to which genus it would have been assigned were it not that its posterior cell seems normally to be closed before reaching the wing-margin. Should the 3 prove to have processes at the tips of the first and second joints of the front tarsi, Pangonia carpentert, despite the transitional character of its venation, would more fittingly be placed under Corizoneura, so long as the independence of the latter be maintained. Genus Thriambeutes, Griinb. Thriambeutes fuscus, sp. n. §.—Length (1 specimen) 11:14 mm.; width of head 4°25 mm.; length of wing 10°5 mm. Dorsum of thorax sepia-coloured, with traces of a faintly marked, paler, longitudinal median stripe im front of transverse suture, pleurae and pectus mummy-brown ; abdomen uniformly blackish brown ; wings mummy-brown, with a clear oblique transverse streak, commencing on anterior transverse vein (its base extending from commencement of lower border of distal fourth of first basal cell to proximal extremity of first posterior cell), wmeluding rather more than proximal third of discal cell, proximal extremity of fourth postervor cell, distal extrematy of second basal cell, and upper border of proximal two- thirds of fifth posterior cell, but not reaching hind margin ; legs blackish brown or black, middle tarsi cream-buff, last joint and tips of preceding joints cinnamon-brown. 152 E. E. AUSTEN. Head black or blackish brown, occiput dark neutral grey pollinose, basi-occipital region thinly clothed with fine blackish brown hair ; ocelli present, enlarged facets of eyes very coarse, area occupied thereby same as in ¢ of genotype (Thriambeutes singularis, Griinb., of Togoland and 8S. Nigeria), 7.e., small facets confined to a deep lower border and a narrower hind border of uniform width running up to ocelli ; palpr blackish brown and clothed with fine hair of same colour, terminal joint elongate and curved but not conspicuously swollen, considerably smaller and narrower than in g of genotype; antennal protuberance large and prominent, considerably larger and more prominent than in 3 of genotype; first joint of antennae blackish brown, short, swollen, cylindrical, and clothed like second joint with blackish brown hair, second and third joints sepia-coloured, expanded portion of third joint rather broad. Thorax and abdomen thinly clothed with fine blackish brown hair. Wdngs: anal angle and lower region of distal extremity short of actual tip paler than elsewhere with exception of clear, transverse streak, a close scrutiny, when wing is viewed against a light back-ground, revealing beyond clear streak an ill-defined transverse band, which appears somewhat darker than remainder of surface ; stagma well developed, elongate, cinnamon-brown when seen against a. light background. Sguamae blackish brown. Halteres: knobs ivory yellow, stalks sepla-coloured. Legs: coxae, femora and tibiae clothed with blackish brown or blackish hair, middle as well as front tibiae swollen (front and hind tarsi, and hind tibiae missing in case of type). South Africa, Bechuanaland Protectorate : N’Gami Country, 1897 (Sir Frederick Lugard, G.C.M.G., C.B., D.S.O.). The species characterised above is readily distinguishable by its wing-markings. alone, apart from all other characters, from Thriambeutes singularis, Griinb. (the only other member of its genus as yet described), in which moreover the body as well as. the head and its appendages are in the § uniformly tawny. So far as it is possible to judge from a photograph, which is all that 1s at present available for comparison, what appears to be another ¢ of Thriambeutes fuscus 1s in the possession of Mr. R. W. Jack, Government Entomologist, Southern Rhodesia, and was taken by him in November 1914, in Sebungwe District, Southern Rhodesia, on the jacket of a companion. In Mr. Jack’s specimen, however, the clear streak in the wing reaches the hind margin, while the margin of the anal angle, and a further portion of the hind border embracing part of the distal extremity of the second submarginal cell and the distal extremities of the first three posterior cells are also. hyaline. 153 APHIDIDAE OF PERSIA. By F. V. THEOBALD. Aphis buxtoni, sp. nov. (fig. 1). Alate viwiparous female. Head and thorax dark ; abdomen pale, apparently pale green, with dark lateral spots. Antennae, cauda and cornicles dark. Legs with mid and hind femora dark, front pair paler ; apices of tibiae and the tarsi dark. Antennae shorter than body, rather thick ; basal segment larger and paler than second ; third much longer than fourth and about as long as the sixth, with 18-20 small round pale sensoria over its whole length ; fourth segment very slightly longer than fifth, with 5-8 sensoria ; fifth with 4-7 sensoria of unequal size, exclusive of the usual subapical one ; sixth with the rather thick basal area less than one-third the length of flagellum, in some however almost half its length ; all the segments from the third imbricated. yes Fig. 1. Aphis buxtoni,sp.n.: A, B, alate 9 ; a,, lateral bladder-like papilla; b,, lateral papilla; C,D, apterous ¢; d,, lateral papilla. large. Proboscis reaching to the third coxae. A large pale bladder-like papilla on each side of pronotum and five pairs of pale rounded papillae on sides of abdomen. Cornicles about as long and as thick as third antennal segment, cylindrical, markedly imbricated, not reaching to the level of the cauda. Cauda bluntly pointed, about half the length of the cornicles, spinose ; with numerous fine hairs, curved apically. Anal plate spinose, with fine pale hairs. Tibiae with numerous fine, pale, short hairs. Wings with normal venation; veins and stigma brown, the veins very faintly and narrowly clouded on each side. Length, 2-2°2 mm. Apterous viviparous female. Pallid, with dark head, cornicles, cauda and anal plate, also dark antennae and mid and hind legs, the fore pair paler. Antennae shorter than body, rather thick ; basal segment larger than second, both very dark ; third paler at base, longer than fourth and about same length as sixth ; fourth a little longer than fifth, the latter with the usual sub-apical sensorium ; sixth with basal area nearly half the length of 154 F. V. THEOBALD. flagellum ; segments 3-6 imbricated. Eyes dark, of moderate size. Proboscis reaching to the second coxae. Cornicles cylindrical, longer than third antennal segment, imbricated. Cauda blunt and rather short, not half the length of the cornicles, spinose, with numerous pale hairs ; broader than cornicles ; the hairs curved apically. Anal plate spinose, with long pale hairs. The rather thick set legs have the coxae and trochanters darker than the rest of the legs; a few short hairs on the femora, many on the tibiae. Length, 2 to 2°5 mm. Food-plant: Umbelliferae. Persia : Enzeli, Caspian Coast, 6.v1.1919 (P. A. Buzton). Types in the British Museum. Described from a single perfect alate female and several apterae. Oneof the latter shows the cornicles not reaching the cauda, others passing it. It 1s a very marked species, the black cornicles, cauda, etc., showing up prominently against the pale body. The antennae are short, as in the genus Cavariella and the cauda is markedly spimose. I can see no trace in the mounted apterae of any lateral pronotal or abdominal papillae, as seen in the alate female. The larval stages are more uniform in colour, but in succeeding instars the blackness of the cornicles, etc., gradually becomes more pronounced. This species was found in association with an ant that has been identified by Mr. W. C. Crawley as Lasius emarginatus, Latr., var. nigro-emarginatus, Forel. Myzus mespiliella, sp. nov. (fig. 2). Apterous viviparous female. Green or yellowish green ; antennae of same colour as body, dusky at apex of fifth segment, the sixth dusky, with a more or less darkened area at the junction of basal portion and flagellum. Eyes deep reddish-black. Proboscis of same colour as body, dark at apex. Legs same colour as body, apex of tibiae and tarsi dusky. Cornicles and cauda of same colour as body ; in some the former seem a little darker. Anal plate darker than cauda. Fig. 2. Myzus mespiliella, sp. n.: A, B, apterous 2; C, D, alate 9. Antennae longer than body ; basal segment larger than second ; third segment longer than fourth, but not quite so long as the sixth ; fourth a little longer than fifth ; the sixth about as long as 4+ 5, with moderately long flagellum ; a few short hairs on third and fourth. Proboscis rather long, reaching to or just beyond the third coxae. Cornicles slightly longer than fourth antennal segment, cylindrical, slightly broadening at the base, markedly imbricated. Cauda reaching not quite to the level of the cornicles, less than half their length, narrowly triangulate or pointed, spinose, with two pairs of lateral hairs. Anal plate spinose, with several long hairs. Tibiae with numerous small hairs, becoming longer near apex. The APHIDIDAE OF PERSIA. 155 vertex shows two short and two longer slightly capitate hairs. A few short blunt lateral abdominal processes. Length, 1°5-1°8 mm. Alate viviparous female. Only imperfect specimens sent. Apparently dark and of reddish hue with dark head, thorax, cornicles and cauda ; legs paler, with dark apices to tibiae and dark tarsi; two basal segments of antennae dark, third pale. The abdomen has dark lateral and dorsal spots and the proboscis is pale towards the tip, but extreme apex dark. The basal segment of antennae larger than second ; the third with 6—7 large round sensoria in a line. Eyes large and deep reddish-black. The black cornicles are cylindrical, shorter, but a little thicker than third antennal segment, slightly expanding basally and not quite reaching the level of the cauda, markedly imbricated. Cauda black, more than half the length of the cornicles, more or less constricted near the middle, spinose, with long, fine hairs; blunt apically. Anal plate black, with long pale hairs. Wings large, venation normal. The abdomen shows four small dark blunt lateral processes before the cornicles and one between the cornicles and cauda. Length, 1°8-2 mm. Food-plant : Medlar (Mespilus). N. W. Persia: Enzeli, 29.iv.1919 (P. A. Buzton). Types in the British Museum. Described from several perfect apterous females, but both alatae damaged, neither showing complete antennae. ‘The sensoria on the third segment are marked, The only other Mespilus species is Van der Goot’s Myzus mespili, which is very . distinct. The vertex and lobes of the apterae and the basal segment of the antennae approach to some extent those of the genus Phorodon, but in the alate stage it is a distinct Myzus. It was found on the young twigs of the medlars, not doing much harm. Ants were found in attendance. Aphis punicae, Passerini (fig. 3). A. punicae, Pass., Aphid. Italicae, pp. 32 et 42, n. 26 (1863). Alate viviparous female. Antennae shorter than body ; two basal segments dark and a dark area at apex of fifth and around the sensoria on sixth, or all dark except base of third segment : Fig. 3._ Aphis punicae, Pass.: A, B, C, alate 9; D, E, F, apterous 2; dj, dark areas on antennae. basal segment wider but no longer than second ; third nearly as long as the sixth, with 6-8 sensoria spread over its whole length, two frequently smaller than the others ; fourth segment longer than fifth, the latter with the usual subapical sensorium ; the sixth about as long as 4 +5; fourth to sixth imbricated. Cornicles a 156 F. V. THEOBALD. dark, about as long as fourth antennal segment and thicker, cylindrical, imbricated, not reaching to the tip of cauda. Cauda pale, shorter than the cornicles, with three pairs of lateral hairs, curved apically and arising from prominent tubercles. Anal plate darker than cauda. Jemora of mid and hind legs dark, except just at base ; front pair paler ; apices of tibiae and the tarsi dark. Wings normal. Length, 1-15 mm. Apterous viviparous female. Antennae shorter than body, pallid, except for a small dark area at apex of fifth segment and one around sensoria on sixth; basal segment wider but scarcely longer than second ; third a little shorter than sixth, but longer than fourth ; fourth and fifth about equal; sixth with flagellum about twice as long as basal area. Cornicles about as long as third antennal segment, broadening basally, pale, dark at apex, not reaching as far as cauda. Cauda pale, with three hairs on each side, curved at their tips and arising from marked tubercles ; not quite as long as cornicles. Anal plate dark. Marked pronotal lateral tubercles and one each side of body before the hind legs. Apices of tibiae and the tarsi dark. Proboscis not reaching to third coxae. Length, 1-15 mm. Food-plant : Wild Pomegranate. N.W. Persta: Tula Road, Talish District, 8.W. Coast of Caspian, 9. vi. and 5. vil. 1919 (P. A. Buxton). Ivraty (Passerint and Theobald). Apparently the apterae only were described by Passerini. The alatae seem to vary in the colour of the antennae in the mounted specimens, some are pale with two dark areas, others all dark except the base of the third segment. The specimens taken by Mr. Buxton were on the twigs, attended by ants (Cremastogaster scutellaris schmidti, Mayr, and Tapinoma erraticum mgerromum, Nyl.). Passerimi records it from Punica granatum and P. sylvestris. The Aphis punicellae I described from Egypt on P. granatum (Bull. Ent. Res. vi, p. 125) is quite a distinct species. Lachnus pyri, Buckton (fig. 4). L. pyri, Buckt., Indian Museum Notes, iv, p. 271, pl. xvi (1899). Taken on pears at Karmanshah, W. Persia (4. xu. 18) by Mr. P. A. Buxton. This species was described by Buckton (with notes by E. E. Green) from pears in Ceylon. The Persian specimens agree with the Pear Lachnid I have from Ceylon and answer generally to Buckton’s short description. I append some characters of the antennae, etc. ‘ey Fig. 4. Lachnus pyri, Buckt.: A, antenna of alate 2; a, basal segments ; b, eye; B, cornicle; C, lst to 3rd antennal segments of apterous 9. APHIDIDAE OF PERSIA. 157 Alate viviparous female. Antennae of six segments; first two large and dark, third the longest, with 13-15 round sensoria of unequal size and a few hairs ; fourth about as long as fifth, with 3 sensoria ; fifth longer than sixth, with one large subapical sensorium ; sixth with ‘nail’ shorter than basal area and a large sensorium at its base, a few moderately long hairs. yes large. Cornicles dark, of normal form, hairy. Apterous viviparous female. Basal segments of antennae large and dark; third segment with 3-4 uniform round sensoria situated near apex. Cornicles as in alate 9. A marked median abdominal tubercle. The specimens sent by Mr. Buxton were all very much damaged. Another Lachnid (Nippolachnus piri, Matsumura) occurs on Pirus sinensis in Japan, but is very distinct (vide, “ A list of the Aphididae of Japan, with descriptions of New Species,” Journal of the College of Agriculture, Tohoku Imperial University, vi, pt. 6, p. 382, July 1917). e. ‘ a ee ros « "A a = Lae ies Tie Ate *, CTP A ey 159 THE CAMPAIGN AGAINST PHYTALUS SMITHI IN THE COLONY OF MAURITIUS. By H. A. Tempany, D.Sc., Director of Agriculture, and D. d’EMMEREz de CHARMOY, Entomologist. (Pate, V.) The following paper contains an account of the campaign which has been in pro- gress against Phytalus smithi during the past 8} years. The history of the pest in Mauritius and the measures which have been adopted for its control are of interest both from an entomological and from an administrative point of view. Published accounts of the pest are contained (1) in a monograph, prepared by one of us at the time of its first appearance in the Colony, and printed in Mauritius ; (2) ina paper on the introduction of Tiphia parallela into Mauritius, from Barbados, for the purpose of the control of the insect, published in the Bulletin of Entomologi- cal Research (iii, pp. 93-102), also published subsequently as a Bulletin of the Depart- ment of Agriculture, in Mauritius ; (3) in the annual administrative reports on the Department of Agriculture in Mauritius for the years 1913-1918 ; and (4) in special reports on the progress of the work of control which have been prepared from time to time for presentation to the Government and to the Board of Agriculture in Mauritius. As, however, the information is scattered, and many of the reports and papers are difficult of access to the general reader, the following summarised account has been prepared for publication. . Historical. The first record of the appearance of Phytalus smithi as a pest in Mauritius occurred in the month of July 1911, on lands of Mon Rocher Estate adjoiming the Royal Botanical Gardens, Pamplemousses. Once the pest had been detected, it became apparent that damage on a very considerable scale was being occasioned thereby, many fields of both plant and ratoon canes being found to be almost totally des- troyed by the grubs. The insect was apparently imported into Mauritius, either from Barbados or South America, probably in earth containing rooted sugar-canes. The exact date of its introduction is uncertain, but it is probable that it must antedate its discovery as a pest in the Pamplemousses district by five or six years at least. Its observation in Mauritius was followed by prompt and energetic measures for its control, which have been continued up to the present time. The effect of these (687) E 160 H. A. TEMPANY AND D’EMMEREZ DE CHARMOY. measures has been to circumscribe to a large extent the infested area, and to mitigate the damage caused thereby. Had these measures not been applied, it seems reason- ably certain that the cane-growing industry, at any rate in the lower regions of the island, would have sustained much more serious damage than has actually been experienced. At the time of the first outbreak, no properly constituted authority existed for dealing with administrative measures directed to the control of plant pests and diseases, and consequently two committees were appointed, charged with the super- vision of the work of control, one being a consultative and the other an executive body. At the same time the present Entomologist of the Department of Agriculture, who then held the office of Curator of the Mauritius Institute and Museum, was appointed temporarily for executive work in connection with the operations, the work being attached, for administrative purposes, to the Department of Forests. On the formation of the Department of Agriculture in 1913, the control of these and other operations against plant pests and diseases was transferred definitely to the control of this Department and has remained thereunder ever since. The Habits of the Beetle and the Damage done by it. The character of this Melolonthid beetle and the nature of the damage occasioned by it are now too well known to need any detailed description, and for a complete - account of the various stages in its life-history reference may be made to the report. on Phytalus smith by one of us (D. d’Emmerez de Charmoy), previously referred to. Only one generation is produced per annum. Though the period of the greatest emergence does not last more than 3 to 4 months, from November to January or February, these insects can be found in the adult stage all the year round in varying numbers, according to the season, the development being to some extent dependent upon climatic conditions. Soil temperature and soil moisture are factors influencing the duration of the pupal and larval stages, as well as the incubation of the eggs. The following table brings out the difference existing between insects reared at: Phoenix, at an altitude of 1,311 ft., and those reared at Pamplemousses, where the altitude is 208 ft., the temperature being in the latter district much higher than in the former. Egg stage Larva: active Lif inactive—prepupal Pupa prior to emergence from soil .. first copulation after emergence Imago 4; copulation af 2 laying perod .. Ae after the laying period No. of days Minimum | Maximum 5 23 210 240 30 45 18 25 10 15 10 15 60 75 30 120 30 90 Remarks Maximum at Phoenix in September. Minimum = at Pamplemous- ses in Decem- ber THE CAMPAIGN AGAINST PHYTALUS SMITHI. 161 It would appear that climate is a most important factor in the control of the pest ; to exemplify this it may be stated that the insect was first detected in the year 1907 at Reduit at an elevation of 1,100 feet, where the climate is distinctly cooler than at Pamplemousses. The insect has been kept under close observation at this point for a number of years past and has not up to the present shown any signs of becoming a serious pest. In this connection mention may be made of the very recent discovery of the beetle on a small area at Joli Bois in the Savanne district of the island at an elevation of about 900feet. Atthis particular point it seems probable that the insect was introduced from Pamplemousses in planting materials trans- ported from forest nurseries formerly existing in Pamplemousses Gardens, for the purpose of restocking Government forests adjacent to this point. No planting material has been sent out from these nurseries since the discovery of the pest at Pamplemousses in 1911, but just prior to this, important consignments of such materials were delivered at Riviere du Poste Railway Station, which is in the centre of the infested zone and undoubtedly is the focus from which the infestation originated. The fact that during an interval of eight years the pest has extended only to an insignificant extent is further evidence of the effect of climate on its incidence. The extent of the infested area in the Pamplemousses district now measures approximately twelve square miles. When the pest was first detected in 1911, the infested area measured only three square miles, the increase in the intervening period being due to outward spread. Hnergetic measures, described in a later section of this paper, are now being taken with a view to circumscribing its further extension. The Royal Botanical Gardens are approximately in the centre of the infested zone and are apparently the nucleus from which infestation originally started. Remedial Measures. The remedial measures that have been applied to combat the pest may be grouped under the following heads :— (a) the collection and destruction of adult insects on emergence ; (b) the digging out and destruction of larvae in badly infected fields ; (c) census and survey work designed to ascertain the location and distribution of foci of infestation throughout the infested area ; (d) the introduction and dissemination of the natural parasitic enemies of the pest ; (e) legislative measures directed towards preventing the spread of the pest to uninfested regions, and affording the necessary powers to the authorities to carry out protective and destructive work in infested areas ; (f) experimental work on other possible means of control. The control work in the Pamplemousses area has assumed very considerable proportions, and the local headquarters for dealing with the work have for some years been established in the Royal Botanical Gardens, In the following paragraphs an account is given of the administrative and practical methods employed under each of the above sub-heads, after which information is supplied concerning the results of the work. (673) E2 162 F. A. TEMPANY AND D. D EMMEREZ DE CHARMOY. Collection of the Insects. The collection of insects by hand has proved one of the most effective methods of controlling the pest, and in this connection organised work under Government supervision on a considerable scale has been carried out systematically each year. The methods adopted have comprised the collection of insects by the general public, combined with collection by gangs in localities where the infestation is light. In regard to public collections the principle observed is the purchase of the collected beetles at rates dictated by the abundance of the insects at any particular period, as gauged by the number of insects captured ; a sliding scale of payments has been worked out in this connection, and the rates paid are fixed in accordance therewith. The work of collecting usually commences in the month of October and continues uninterruptedly till March. The months of greatest emergence are December and January, during which period the daily catches not infrequently run into several millions. The capture of the beetles commences at about 6.30 p.m. and continues up to 10.30 p-m. The habit of the insect lends itself to this particular form of control measure, inasmuch as at nightiall it emerges from the soil, where it lies buried during the day, takes a short flight to the nearest bush or tree, and remains there for hours nibbling slowly at the leaf on which it has settled. The imsects are undisturbed by lights and are readily collected by searchers who look for them with lanterns. Beetles caught each night are received, measured and purchased at Pamplemoussses Gardens on the following morning, and immediately after measurement destroyed by fire. During busy seasons the work of measuring beetles occupies many hours. The method employed in computing the number of captured beetles is by means of measurement in containers of standard size and known capacity. To facilitate the work of collecting beetles, purchasers are licensed by the Agricultural Department, who buy beetles from the general public at certain specified points on the night they are col- lected and resell to the cfficer in charge the next morning ; while many estates also follow the same practice, in some cases paying slightly enhanced premiums in order to attract collectors, and also supplying oil for lanterns, collecting receptacles, &c.- With regard to collection of sects by gangs as opposed to public collection, this form of control has of necessity to be practised in regions where the infestation is light and beetles are not numerous, since under the system of purchase, beetle hunters are obviously attracted only to localities in which insects are plentiful, while the scale of operations renders impossible any disciimination in the matter of prices in relation to the localities in which captures are effected. Two quite distinct cases occur in which collection by gangs is indicated; the first is that of places well within the infested area at which the beetles are not numer- ous ; the second is in the case of the peripheral zone, where infestation is light and where slow outward spread to previously uninfested regions is taking place. This outward spreading constitutes one of the most serious aspects of the situation. By 1917 the increase in the infested area had become so marked that special measures to restrict the spread became necessary. With this object in view the periphery of the infested zone was resurveyed in 1917; following this the periphery was divided between a series of patrols, and to each patrol a special collecting gang was assigned. Hach gang consists of from 8-10 labourers under the charge of a head- man ; it is provided with powerful acetylene lights to assist in the detection of THE CAMPAIGN AGAINST PHYTALUS SMITHI. 163. insects, and is required to patrol the section allotted to it on a depth of } a mile at least once in each night, and to collect insects found occurring on this area. The gangs are regularly inspected by a travelling inspector, when attendances are verified, captures of insects checked, and the general working of the gang examined. As the result of work on these lines during two seasons there appears reason to believe that the rate of outward spread is becoming considerably reduced. The collection of beetles during busy seasons affords part time employment to large numbers of people, and at such times the actual number engaged may total several thousands. The largest number of beetles captured on any one night amounts to 4,600,000, which was recorded on 12th January 1919. Emergence of beetles usually takes place in flights, large numbers appearing simultaneously for a short period and then rapidly diminishing. It is usual for two or more flights to be recorded in one season, and on occasions when beetles are numerous it is remarkable, after nightfall, to observe the country-side dotted with the lanterns of collectors. The transport of the beetles from the points of collection gave rise some time ago to fears concerning possibilities of reinfestation owing to their escape from unsuitable containers. A special form of container, intended to preclude the escape of captured insects deposited there, has in consequence been devised and efforts,made to induce all persons engaged in the work of beetle collection to provide themselves therewith. Daily returns of captures of beetles and of larvae are supplied to the Head Office ; by means of these, detailed control of the work is maintained, and it is found possible in this way to detect readily localities which show signs of being especially heavily infested and if requisite special attention can then be devoted to them. From the outset the collection and destruction of larvae, by digging them out of the soil in badly infested areas, has formed an essential part of the control measures adopted. As at present conducted, it is principally combined with survey work and is carried out during those months of the year when beetle emergence is at a minimum and the larvae are attaining full growth, wz., May to August. With the oradual spread of Tiphia parallela, the introduced parasite, it has been found advisable to restrict the destruction of larvae in localities in which this parasite has established itself, as the indiscriminate destruction of larvae in these circumstances may do more harm than good, owing to the risk of destroying those that are para- sitised. Census work in the infested area has been systematically carried out for some years past. Under this method in every field in the zone 10 to 15 trial holes of a depth of 8 inches are dug in every acre, the larvae found in each hole are counted, and an infestation rate per acre is worked out. The results are subsequently tabulated in the form of a plan, a system of tinting being adopted to indicate varying degrees of infestation. Following this, orders are issued to estate proprietors to dig out the larvae on fields which show more than 20,000 larvae per acre. The prescribed fields are subsequently investigated again after the digging operations have been completed, and if the work has not been satisfactorily performed, the Department of Agriculture is empowered to repeat the process and to charge the estate owner with the cost. ; 164 H. A. TEMPANY AND D. D EMMEREZ DE CHARMOY. The Introduction of a Natural Parasitic Enemy of the Pest. The identification by Nowell, in 1912, of Tephia parallela as the principal insect enemy of Phytalus smithi in Barbados led to a series of attempts being made between the years 1913-1915 to introduce the insect into Mauritius. A full account of the efforts made in this direction are contained in the paper on this subject previously referred to, published in the Bulletin of Entomological Research. Since the publication of the paper m question Tiphia parallela has established itself firmly at several points in the infested zone and at these points has already exercised a very marked influence on the incidence of the pest. The effect was first seen in 1917 at Esperance estate, where, as the result of the action of the parasite, the total number of adult insects captured in the season 1917-1918-showed a very marked reduction over that of the previous year, the reduction being still more marked in the year 1918-1919. In the year 1918-1919 similar reductions became visible in the case of Maison Blanche estate and also on the group of small properties classed under the head “Small Planters’ which occur in this vicinity. During the present season, though completed figures are not available, there is evidence to show that a similar reduction has become evident at Beau Plan Estate; while at Esperance where the insects were formerly present in great numbers they are now only met with occasionally. The following figures for the number of insects captured at these pots during the years 1916-17, 1917-18 and 1918-19 illustrate these remarks :— Year Esperance . MaisonBlanche Small Planters 1916-17 21,419,586 — a 1917-18 7,650,461 1,895,752 8,322,130 1918-19 518,320 937,950 4,303,625 The spread of Tiphia throughout the infested area has been systematically assisted by the liberation of insects in places where they did not previously exist. An important point, which has materially assisted in the dissemination of the insects, is the existence of the wild plant Cordia wnterrupta (Herbe Conde), the sweetish exudation from the leaves of which provides food for the Tvphia adults. It is also noteworthy that the plant attracts Phytalus ; i consequence planters are now growing the Cordia around the borders of the cane-fields, and this has materially facilitated both the capture of Phytalus and the dissemination of Tiphia. Legislative Measures. Legislative measures for the control of plant pests and diseases are provided under Ordinance No. 8 of 1911. This ordinance provides for the proclamation of any prescribed area as being infected with plant disease, including both insect and fungous pests, and for the making of general or special orders for the following purposes :— (a) for prescribing or regulating the destruction, removal, uprooting, disposal or treatment of plants and products of a vegetable nature within an infected district, area or place ; (b) for prescribing or regulating the cleansing and disinfecting of infected places or areas, or parts thereof ; (c) for prescribing the period within which it will not..be lawful to plant or replant with any plant whatsoever, or with any particular kind of plant, the whole or any portion of an infected district, area or place ; THE CAMPAIGN AGAINST PHYTALUS SMITHI. 165 (d) for prohibiting or regulating the movement of persons, animals, earth, soil, manure, vegetable product or other things into or out of a district, area or place declared to be a place infected or suspected of being infested with a plant disease ; (e) for the appointment, pay and regulation of the duties of persons to be charged with the execution of any of the provisions of or orders made by virtue of this Ordinance ; (f) generally for the purpose of in any manner preventing the spreading within the island of any plant disease. It further provides for the removal of such restrictions in the event of infection being stamped out, for the entry on occupied lands of duly authorised persons for the purpose of examining plants and for carrying out remedial, disinfective or prophylactic measures, for the notification of plant diseases and for penalties for contravention of the provisions of the Ordinance. Under this Ordinance, Proclamation No. 10 of 1913 prescribed the area infested with Phytalus smithi, and made certain orders in relation thereto. In 1918 the spread of the pest and experience with its control rendered desirable the rectification of the prescribed area and slight modifications in the order ; accord- ingly the above proclamation was superseded by Proclamation No. 42 of 1918 the terms of which are as follows :— (a) the removal of living specimens of Phytalus snuthi either larvae, pupae or adults from the area prescribed is prohibited. This, however, shall not apply to the transport of living insects in captivity by properly authorised officers of the Department of Agriculture for purposes of scientific study. (b) the removal of all earth, manure, leaf mould, decayed vegetable matter and all living plants or parts of plants in soil is prohibited from the area pre- scribed, but the above shall not apply to fruits or vegetables growing above the ground and detachable from the tree or stem on which they grow. (c) cane-tops for propagation purposes in such limited quantities as may be decided upon by the Director of Agriculture may, after examination and disinfection if necessary at Pamplemousses Gardens or at such convenient point as the Director of Agriculture may select, be removed in sealed packages from the infested area. (d). the Director of Agriculture or other person deputed in writing by him may at any time inspect sugar-canes or other crops within the specified area and issue, if considered necessary, an order in writing containing instructions as to the cutting, reaping, disinfection or other treatment of such sugar-canes or crops and as to the disinfection or treatment of the soil in which they are growing. Owners, managers or occupiers shall within 48 hours of such service commence to carry out the treatments ordered, and failing to carry them out, the Director of Agriculture or person aforesaid may cause such treatments to be carried out at the expense of the owner, manager or occupier notified. (e) the Director of Agriculture may plant, raise or cultivate such trees, shrubs or plants as may be deemed expedient for the purpose of attracting or trapping, and may spray or otherwise treat the same for the purpose of destroying the pest. These trees, shrubs or plants shall be protected by the owner, manager or occupier of the land on which they are planted or cultivated for such time as the Director of Agriculture may consider necessary. 166 H. A. TEMPANY AND D. D’EMMEREZ DE CHARMOY. The measures above indicated have on the whole provided adequate machinery for the control of the pest. One point, however, requires mention, namely the risk of artificial dissemination of the pest by interested parties, as a result of the policy of buying the beetles. To guard against this a supplementary Ordinance was passed in 1918, whereby any person wilfully disseminating a plant disease is rendered liable to imprisonment without the option of a fine. Other Methods of Control. In the early days of the outbreak, a considerable number of experiments were tried on methods of control, other than those enumerated above. These included the trial of various stomach and contact poisons and also the injection of carbon bisulphide and other poisons into the soil. The pest, however, exists in such numbers and on such very extensive areas that the cost of methods of this description renders their employment prohibitive for the most part. The principal methods which have been tried are described below. Numerous experiments were undertaken at the outset with a view to finding an efficient insecticide, moderate in cost and easily applicable, that would permit of the control of the pest, more particularly on young plantations, as these suffered most from its attacks. Powder insecticides, such as Paris green and vaporite, proved unsuitable on account of the labour expenses involved in their application. Neither potassium cyanide nor carbon bisulphide could be utilised, not only on account of the expense, but owing also to their harmful effects on plants. Liquid insecticides, involving the use of kerosene, creoline and carbolic acid, though rather expensive, gave more satisfactory results when dealing with young plantations and first ratoons, but their application could not be generalised, in most cases for want of the water needed in their preparation. Experiments were also tried with the object of finding some substance that would attract the larvae. Good results were obtained with refuse cane-cuttings and tops. These were split in halves and buried in small furrows 2 or 3 inches deep. Every fortnight the furrows were opened, the cuttings turned over and the larvae removed, the same cuttings being replaced and revisited. This method not only allows of the destruction of the pest at a comparatively cheap rate, but also prevents the larvae from entering the planting holes when in search of food. Other substances such as molasses, ashes, scums, &c., have been experimented with, but without success ; attempts were in addition made to infect adults and larvae with cryptogamic diseases. Those affecting larvae proved contagious, but could not be transmitted experimentally ; while the one that attacks the beetle and was introduced from Porto Rico (Metarrhizium anisopliae) gave poor results. It was found subsequently that this fungus existed already in the Colony, attacking various species of SCARABAEIDAE, but did not exert any very great influence in checking their spread. The Number of Insects captured and the Results of the Campaign. The statistics for the number of insects captured afford the best indications of the total prevalence of the pest. These are given below in tabular form, the figures showing both insects and larvae destroyed on all the properties concerned. 167 THE CAMPAIGN AGAINST.PHYTALUS SMITHI. < OCSOY OPO »» SUIPNPOUT 196 °606'F | | Igs‘eT | 006°ST a 00F | 000°6 009°TLO"E | 000°F +006°L60°T 8I-LI6L LOVISL'S | 000‘8 L6S‘€0I eke 0O8‘Sl | 006°SF9°S OOL‘9ZE «00618 LI-9I6T O¢8'F69°9 | *0GL°6E os Z10°S 000‘019°% 00€‘*EZ0'T | 00E*6 9I-SI6I 0f0°S9S*L OOL*ESI 0S0°OLS OOT*F9T‘9 | OOS*LEE | SI-FIGT C6F'9E0'S OS TSS OO€*FZ8'T | SFO°L8T | FI-EI6I SIS*1IZ | | SIS‘LIZ | €I-ZI6I 9L8‘S16 | *8&6°1 GSI‘ LES LFL‘901 999°ES¢ | SI-II6I | | | | | OVAIV] | suepreyg | oipuy HUSA | OTCSOY | $1OFUVT | goursodsy urvjg neeg, PVA gunoyy oy, UU s1v0T Jo “ON [890], ‘au "48 “nog =| epuviy | yeurg | te uOSsteyy uoW ‘SIG OUNL OF [I61T AME Wo yno Sup ovAIv'T jo Joquny | | FIL 69S" OL SZSFIL | OSTZLE‘E | OST9ET | 0095099 | SZ9°EOEF | OZE*STS OFL'LIG 6G Of6°LE6 LZF‘98L‘8% FC9'169'T 61-8161 F6E‘8ES‘ OL | FES IG | €6E°S | OGOOE | LS8‘ET IF9°SFL OLT‘SES | 0G0'6G9‘9 | TIT69F6 | FIL‘6SP°SS | 9160S 1% EE0'E89'FS [6L°196 1 —LI6I 6FOOETEL| OSL‘TIL| €S8 | 6LEL |LES‘SOT | SES'hs | 166‘T G63‘SS8‘L | 9LOFI9‘TS | OST°SS8‘9S | O86'FZI‘E | SSL°SLE'SI | TLL°060' F| LI-9I6I GEL'SPZ'EF *9PS°LIT L66‘09S‘8T OZL‘*Z19‘SI €L66SS‘OL |€96°ESFT| YI-SIGI SOF‘ LETS | *666 ZIS*LZS*LI 986°618‘TI S8S‘TPEe‘ST | 1ZE*L9E‘E | SI-FIGI O€ 1 F89‘9E | O88 ‘ESPs 888 ‘FEF's ESP‘ LLL‘OG | 6E6°L86°F| FI-SIGI 9€9°S40'9T | LIZ‘9EV'P L98°609°G GSE*LLL‘'G |L90°@CS°E| EI-ZIGI L8T‘O9F'SZ | 6¢¢‘60E‘T 296 ‘698'F | PIPESLTOL | Z91°8Z6°6| ZI-II6I | | | | soTJoog JO suepaey | opny imof{ | g9apuy IIUeA | O1[Vsoy | s10jUeT | oyourl eg LOTOOY ‘ON IMIOL “aa | mos -eg-a | 39 nog | opursy | yeu Seis ee umes ee Rog | nee SI6Il 9uNL 0F TI6LT ATE Wor porngdvo sopjoog jo rtequinN 168 H. A, TEMPANY AND D. D EMMEREZ DE CHARMOY. The figures for the 1919-20 campaign are as yet incomplete, but the data so far accumulated appear to give reason for the expectation that the total captures re- corded will be markedly less than those of the three preceding years. It will be observed that during the years 1916-17 to 1918-19 the total number of insects captured per annum has remained practically stationary ; before that date they showed a fairly steady annual increase, mainly due to the increase of the infested area ; while it is clear that the observed check to this increase is the the result of the control measures which have been applied, the actual result experienced is the outcome of a number of superimposed factors. As already poimted out, marked reductions have taken place in the intensity of infestation at Esperance, Maison Blanche and in the case of small planters, largely as the result of the introduction of Tiphia parallela. In localities where Tiphea has not yet established itself, the tendency is on the whole still slightly towards an increase in numbers, while in addition there is the factor of outward spread, and it is estimated that if Tuphoa had not been introduced, the insects captured in 1918-19 would probably have totalled 100,000,000. It would appear that collection methods by themselves have proved efiective in reducing the incidence of the pest and have greatly mitigated the damage done. Collection and the work of T7phia parallela combined appear to afford reasonable hope that in infested areas they will ultimately render the incidence of the pest comparatively unimportant. The problem of preventing the spread of the pest to uninfested areas is, however, a difficult one and if this is to be effected, a continua- tion of some measure of quarantine and peripheral patrol collection seems essential. Even if Tiphia becomes established throughout the whole of the infested area, it can never by its nature establish a complete control of Phytalus, since for its existence it is dependent on the production of a supply of beetles each year. It is moreover powerless to check completely the spread of the beetle, since Tiphia will naturally congregate where Phytalus is plentiful, that is to say, well within the infested area, while the pest spreads at the circumference where the attraction for Tiphia is least. Whether in the event of the successful establishment of Tiphia throughout the infested area such a slow spread can be regarded in the end with equanimity, it is impossible at the present stage to say. It is, however, quite clear that although the results obtained with T7phia are up to the present in the highest degree encouraging, it is not possible to predict what the future may hold in store for it. No natural enemies of Tvphia of any importance are known, but in a foreign environment it is conceivable that hitherto unknown inimical influences may develop ; in consequence it would for the present appear unwise to relax the control at present practised and in point of fact more stringent control measures are in contemplation. The Cost of the Campaign. The total expenditure on the campaign against Phytalus smithi up to 31st December 1919 amounted to Rs.232,100 from funds administered by the Department of Agriculture. Up to the year 1915, the money expended was provided entirely by Government; since then a system of contributions by estate owners has been inaugurated. During the year 1918-19 Government contributions amounted to Rs.24,000 and the estates’ contributions to Rs.8,500. Apart from the direct THE CAMPAIGN AGAINST PHYTALUS SMITHI, 169 expenditure, a considerable sum of money has also been spent by estates in operations for beetle destruction, both in labour in the digging out of larvae from infested fields and in the provision of extra premiums, supervision and materials for the work of beetle collection. Throughout, the closest co-operation has been maintained with the planters concerned, meetings are periodically held for the purpose of discussing plans and details of the campaign with interested proprietors and managers. In this way harmonious and efficient work has been secured, and many obvious administrative difficulties have been successfully avoided or overcome. Very recently the planting body of the Colony have decided unanimously that, in view of the fact that the pest may be held to constitute a serious menace to the greater part of the sugar industry of the Colony, the entire cost of control should be borne by the planters ; and they have recommended to Government the imposition of an export tax of two cents of a rupee per hundred kilos of sugar exported for the purpose of defraying the cost of the campaign. The sum of money likely to be provided as the result of this measure will suffice for a further extension of activities, and, given the favourable results so far achieved, there is reason for hoping that within a comparatively short period the incidence of the pest may become reduced to comparatively small proportions. It is impossible to gauge exactly what would have been the effect if these measures of control had not been undertaken. It is, however, certain that the pest would have extended to a very much greater extent than at present, and in badly infested fields would have probably rendered cane-growing impracticable. The lower sugar-growing lands of the island, which provide a suitable environment for Phytalus, probably constitute two-thirds of the total cane-growing area; in the absence of rigid control measures it is probable that by now the beetle would have become endemic in the greater part of them. While therefore the exact loss which has been avoided cannot be estimated, one is safe in saying that it would certainly amount to several millions of rupees. On this account the measures taken and the expenditure incurred must be regarded as abundantly justified by the results. in! dis fi ‘ : * fa nal | | ' ” _-« ‘ 4 eeig Wy ft ; . : ir j ha Pay} 7 a oie, qwenmasit Oeil Yr DOnlaey bine sd hts Hebe ong Bilal % ‘on i ts ate th A iy , et %) + 4 lh pil et | _ ey prone data aie ' llama hp sae tigate ive rt Ns take sa ya] art hy ee ‘ feu iy t Raret ein + Ve mbes tiph ides ab bot SEN Soayeiarrys pobre vii) ish! baal ii iNetond "5 ay bie: | | 1 A CCE! aie aR nit | ay x pulrile | meee ‘ et vi LR a | Santo niu Pa bine a sek svetr: NB SANTEE TAN EME ESS AD arp es fester J ; Din VY Jf 28 Ante. clan al to wha er et - 4 bes EP Gl mS f sph ; ae 7 en Fe Piece. i> ory a cord Lib ine ee | 1S 5 aes Lk cael nabesthi dh wh * f* goa ay , if ’ ee j ig Ue q niet ay" ; i d 7 “\ isis Dae % » tai og youd . ribs 4 ee ahaa ye OG, Ka ks, ae i ' 1 ; 3 f ' a ee oe ie 7 “ ' a4 Mea | unde Ne ES A ie Ta.) i . As bie, 0h cts wee , A , Ty fal 5 mr pe pa rf ie i Hn ’ ’ By Pat rit Prem Nee - Le, nara Cr t A y * } A i) M j , c he ae a a Ds PERO) ay, i oe b , LO oe Ma GC WS I / ia wy i 4 rh ive nayt ee a BBN AAR i | hh genie ne eyelet ROME VOLES StS 1650 SAY RRR le ihe — a a ner Fie sake PACT MES i 2 UMA aa rs 5 ¥T wir mend o pe TH is ay BuLL. ENT. RESEARCH. Fig. 1. Fig. 2. Votes al. PART. 2; A night’s capture of Phytalus smithi. Destroying captured beetles by burning. PLATE a ¥% = 7 Zz =o Coe oe 7 q oe 171 NOTES ON INSECTS ACCIDENTALLY INTRODUCED INTO THE ISLAND OF MAURITIUS. By D d’Emmerez de CHarmoy, Government Entomologist. In respect of insect and, incidentally, fungous diseases of plants, Mauritius is remarkable in two respects ; the first being the very small number that are indigenous to the island, and the second the very large numbers that have been introduced. The desire to introduce exotic plants of a useful or ornamental nature is very widely spread in this Colony, and very extended plant introductions have taken place. Until comparatively recently no restrictions existed on such importations, and as a result insect and fungous pests in large numbers have become established to the sreat and permanent detriment of the major and minor agricultural industries of the Colony. For example, the cane-sugar industry was threatened with ruin in 1842 as the result of the introduction of the borer, Proceras saccchariphaga, Bojer, from Ceylon ; at a somewhat later date great damage was done by the insect known locally as the “ Pou a Roche Blanche ” (Pulvinaria gasteralpha, Icery) ; and again quite recently the introduction of Phytalus smithi has proved a serious menace to the industry and has resulted in the loss of large sums of money. Minor industries have suffered in a corresponding degree. For example, the cultivation of coffee and vanilla has been in large part abandoned, partly as the result of the incidence of introduced cryptogamic diseases ; while in general it has been shown in another paper that out of 80 species of insects recorded as pests of minor crops in the Colony, 50 (or 62°5 per cent.) are of exotic origin. In relation to official measures of control, the first action taken was in the year 1882, when an ordinance was passed dealing with the introduction of pests of the grape vine, the legislation being the outcome of a convention entered into by the local authorities with the Government of the Cape of Good Hope. Until 1910, no other protection was afforded ; in this year a second ordinance was passed empowering the Collector of Customs to destroy, or cause to be treated, plants and other articles, the importation of which was liable to lead to the introduction of pests and plant diseases. The ordinance remained inoperative until 1911, when, by proclamation, the powers thereunder were transferred to the Director of Forests. In 1913, the regulations under these ordinances were again revised, under Pro- clamation 81 of 1913, whereby the restrictions were made more comprehensive, and executive powers transferred to the Director of the then newly organised Depart- ment of Agriculture. Since that date a regular and efficient control has been exerted over plant importa- tions. The text of the Ordinance and the regulations made thereunder are given as an appendix to this paper. It is impossible, at the present time, to trace the manner in which many of the older insect pests of crops were introduced into the island, or even to give the exact date of their introduction. Wey D. D EMMEREZ DE CHARMOY. On the other hand, information regarding occurrences during the past twenty years is much more precise and a review of it is of interest, inasmuch as it affords indications of the channels of introduction and the harm that has been caused thereby. Phytalus smithi. . Though this beetle was detected in Pamplemousses only in the year 1912, there is no doubt that its introduction dates back to at least ten years before its existence was noted. The infested area, the large number of insects destroyed in the first year 1912, the subsequent infestation of other localities, and its rate of spread in these localities, leave no doubt as to the time that the insect must have taken to spread over so large an extent, which in 1912 was about 1 mile in radius. Though it is impossible to make any definite statement, the probabilities are that the insect was originally introduced through the Royal Botanical Gardens, Pamplemousses, by means of cane plants imported from Barbados, packed in soil in Wardian cases. A Cecidomyiid Fly. This Cecidomyiid, which it has not been possible to identify up to now, and the origin of which cannot therefore be traced with certainty, was probably imported in the year 1909 with a lot of mango plants from India by a sugar estate proprietor of Grand Port. A few years later, the presence of this insect was revealed simultaneously at Port Louis and at Mahebourg, on mango trees, the leaves of which bore small conical galls grouped together irregularly and containing Cecidomyiid larvae. These localities were at first the only two affected, but the infestation was then so widespread that there was no hope of destroying the pest. After having remained localised in these two places until 1918, this fly was found at Grande Riviére, then at Riviére Noire, and can now be seen in almost all localities in the island, in varying numbers. Dacus d’emmerezi. | This Trypetid fly was described by Professor Bezzi in 1917 in the Bulletin of Entomological Research. It is not an indigenous species, as its presence was only detected in 1915, in the course of investigations concerning Dacus sygmoides. It would be difficult to say whence it came, though one might reasonably suppose that it has been introduced from Madagascar by means of cucurbitaceous fruits, as it would have been difficult in infested fruits coming from a more distant place to pre- serve the healthy appearance which alone could have warranted their safe transit. This species is nowadays far more common than Dacus sygymoides, and exhibits a partiality for marrows and pumpkins, constituting a pest in the cultivation of these vegetables. Echidnophaga gallinacea. This flea found its way into Mauritius on fowls imported from Vohemar in 1913, and others which were introduced from South Africa in 1914. The insect proved troublesome at Mapou during the early days of its introduction ; it has not spread to the other districts up to now. NOTES ON INSECTS INTRODUCED INTO MAURITIUS. 173 Sarcopsylla penetrans. The jigger has been repeatedly introduced from Madagascar and Farquhar Island since 1910. Rare cases of infection were noticed at Port Louis several years ago. The flea has since never been found elsewhere. Solenopsis sp. (Red Ant). It has not been possible to identify this ant up to now, and it is not therefore possible to say what is its country of origin. It is stated to have existed in Pample- mousses for the last twenty years. The writer first encountered it at Rose Hill in the district of Plaines Wilhems some fifteen years ago, when it was only to be found in a few places. It is now widespread throughout the island, being quite common on the coast and in hot places, where it has become a regular pest ; it is still relatively rare in the higher districts, where it has not hitherto attracted special notice. It rarely establishes itself in houses ; on the other hand it is a serious pest of food crops wherever it has established itself, building its nests deeply in the soil and extending its subterranean galleries in every direction for long distances. It steals small seeds from seed beds and eats on the spot larger seeds which it cannot carry away ; it also attacks young seedlings. Coccidae. As one would expect, importations have been more numerous in this family than in any other, not only from the year 1900 but also previous to that date. In a monograph published by the author in 1900, 54 species were mentioned, among which three only could be considered as indigenous. This number has since been increased by eight more ; these are :—Asterolecanium spectabile, Newst., Aspidiotus destructor, Sign., Aspidrotus mauritianus, Newst., Chionaspis simplex, Green, Lecanium mangiferae, Green, Lecanwum hesperidum, L., Pulvinaria antigoni, Green, and Pulvinaria sp. Asterolecanium spectabile and Aspidiotus mauritianus. These scales were observed for the first time in 1915 in Pamplemousses Gardens and at the Botanic Gardens, Curepipe, on various species of palms and on Cycas revoluta and Cycas sp. They cause much harm to these plants in spite of the fact that the insects are highly parasitised by various Chalcids during a certain part of the year. Aspidiotus destructor. This species also appears to be of relatively recent introduction, as it is only quite lately that guavas, the plants on which they are most abundant, show visible signs of infestation. It is probable that the insect was imported from South Africa, whence a considerable number of plants have been introduced during the last ten years. Chionaspis simplex. This species has hitherto been found only on bamboos and is relatively rare, causing no apparent damage to this plant; it presents no importance whatever from an economic standpoint. 174 D. D EMMEREZ DE CHARMOY. Lecanium mangiferae. The importation of this scale probably does not date back further than 1907, as it was only in 1908 that the first infested trees were recorded in Moka. It would be difficult to state its origin, as the author has observed it in Madagascar and in Réunion in 1917, where many trees, especially mangoes, were severely attacked. In this Colony, it is in Grand Port and especially in the town of Mahebourg that this insect causes most damage. Mangoes, bread-fruits, Connamomum zeylanicum (cannellier) and litchis are often infested to such a degree that they do not bear fruit for several years. In the north of the island, where it exists, its presence is not apparent, either on account of unfavourable climatic conditions, or owing to a small Chalcid which parasitises 1t intensely. Lecanium hesperidum. It seems curious that this species, the distribution of which is world-wide, should not have been found in Mauritius before 1900. The effects of the insect are so striking that it is inadmissible that 1t could have escaped the author’s notice at the time when this group of insects was being worked upon. It was introduced from the Seychelles or from Madagascar with oranges. On the higher plateaux its presence is hardly perceptible, but it 1s much in evidence along the coast, where limes, particularly, suffer severely from its attacks. Pulvinaria antigoni. This species was introduced some 10 years ago from the Seychelles ; it was first discovered in Pamplemousses Gardens and a little later at Phoenix on Solanum wildenit and on Aristolochia. It is far from beimg widespread in the island and is relatively common on these plants for only a few months of the year, being highly parasitised by two minute Chalcids. It is up to now of no economic importance. Gagrella feae. This Phalangid, which was imported from Burma some 10 years ago with cargos of rice from Rangoon, is a striking instance of the rapidity with which an exotic insect can spread itself, when conditions prove favourable to its development and when nothing intervenes to check its spread and multiplication. The first specimens were seen by the author in 1910 and had been captured in the town of Port Louis by Mr. Lauricourt Olivia, taxidermist to the Museum. Three years later it had reached Grand River, 7.e., 26 miles from the original point of occurrence, and in the course of the five following years, the other districts of the Island had all become more or less invaded ; nowadays it is to be found everywhere. It has established itself in all towns and has crept over to the forests, where it exists in considerable numbers. It is sometimes so common in houses as to constitute a regular domestic pest. These Phalangids shun light and have the habit of collecting in considerable numbers in dark places. They spread out rapidly whenever they are disturbed, but will come back soon afterwards. NOTES ON INSECTS INTRODUCED INTO MAURITIUS. 175 Anoplodesmus saussurel. This millipede was introduced into the island about 1904-1905, in Wardian cases containing rubber plants imported from Ceylon by an estate situated in Vieux Grand Port district. One year later the writer was in a position to say with certainty that the creatures occupied already the whole mountain range that separates Grand Port district from Flacq, and that the western side was no less infested than the eastern side. Conclusion. It follows from the above that within the last 20 years, 14 exotic species of insects are known to have been accidentally introduced into this Colony. Of these 14 species, two have no significance from an economic standpoint; the 12 others have already caused considerable damage. These facts demonstrate emphatically that the restrictions established through the Department of Agriculture in 1913, though they may hamper to some extent freedom of importation, fulfil an urgent requirement in protecting the agricultural products of this Colony, which in the past have suffered considerable losses that could have been avoided altogether by the adoption of the measures now in force. APPENDIX. ORDINANCE 4 oF 1910. 99 1. In this Ordinance “ article’? means any seeds, plants, cuttings, or any package, covering or thing that may have come directly or indirectly from any Country or place named in any Proclamation issued under this Ordinance. 2.—(A). It shall be lawful for the Governor in Executive Council by Proclamation to prohibit the importation of any articles from any country or place named in such Proclamation which, in his opinion, are likely to be a means of introducing any plant disease from such country or place. (B). In like manner the Governor in Executive Council may prescribe the con- ditions under which alone the importation of any articles shall be permitted that may have come directly or indirectly from any country named in such Proclamation. 3. Any article coming from any country or place the importation from which is prohibited, and any article arriving from a country or place the importation from which is permitted upon certain conditions only, until and unless such conditions shall have been complied with to the satisfaction of the Collector of Customs, shall be deemed to be prohibited goods within the meaning of the Customs Ordinance No. 28 of 1892; and any such conditions as aforesaid shall be deemed to be restrictions within the meaning of the said Ordinance. 4. So long as any Proclamation as aforesaid is in force any articles mentioned therein coming from parts beyond the sea may be deemed to have come from a place the importation from which is prohibited, and may be treated accordingly, unless the importer satisfies the Collector of Customs to the contrary. (687) . F 176 D. D EMMEREZ DE CHARMOY. Proclamation 81 of 1913 issued under the above Ordinance provides :— 1. The importation into Mauritius from any country or place whatever, including the Dependencies of Mauritius, is absolutely prohibited of :— (a) grape-vine cuttings and plants, except when covered by a certificate from the Board of Agriculture (or other competent authority) of the country of origin that the vines have not been exposed to the infection of phylloxera. for the six weeks prior to the date of shipment ; (6) earth, and leaf and garden mould ; (c) sugar-canes or cuttings thereof, live plants of all sorts, including roots, tubers, cuttings and grafts in any description of earth or soil ; (d) dung or animal droppings (except guano) ; (e) forage ; (f) timber with the bark on. 2.—(A). The following articles may not be introduced unless written permission has previously been obtained from the Director of the Department of Agriculture :— (a) sugar-canes or cuttings thereof ; (b) living plants or bulbils of Agave or Fourcroya ; (c) tea plants. (B). The issue of a permit shall be in the discretion of the Director of Agriculture, who may attach conditions to the permit and may limit the number of plants, cuttings and others aforesaid to be introduced. 3. The following will be inspected at the port of entry :— (a) sugar-canes or cuttings thereof ; (b) live plants of all sorts, including roots, tubers, cuttings, grafts, and buds ; (c) fresh citrus fruits from all countries except the Dependencies of Mauritius. The consignee or his agent shall, upon request, open the coverings and afford the inspecting officer every facility for conducting the examination. 4. If, on such inspection, the articles are found to be not free from pest or diseases, they may be ordered to be destroyed by the consignee or his agent under the supervision of the inspecting officer, or to be subjected to such process of disinfection or treatment as the inspecting officer may prescribe, and the consignee or his agent. shall pay in respect of such treatment such fees and charges as are prescribed. 5. If the removal is authorised, all sugar-canes or cuttings thereof and live plants. of all sorts, including roots, cuttings, grafts, and buds, shall be planted in a nursery apart from growing plants of the same kind and shall be subject to inspection by the Department of Agriculture from time to time during twelve months from the date. of importation. _ 6. If, on such inspection, the articles are found to be not free from pests or diseases not known to occur in the Colony they shall be rooted out and destroyed by the owner under the immediate supervision of the inspecting officer, and if found to be not free from pests or diseases known to occur in the Colony, they shall be so treated by the owner as may be directed by the Director of the Department of Agriculture. 7. Non-compliance with the conditions imposed in articles 4, 5 and 6 with respect. to the disinfection, treatment and planting of the articles imported will render them liable to destruction at the expense of the owner. NOTES ON INSECTS INTRODUCED INTO MAURITIUS. Rit 8. Provided that nothing in the above shall prohibit or prevent the Government from making such importations and introductions of such live plants, cuttings and others aforesaid, as it may require for scientific investigations. The provision for prohibition of certain imports is to protect our staple industries from pests and diseases which could be introduced in shipments, and which would at the same time be very difficult to detect at the port of entry. The sugar-cane, fibre and tea industries are also protected against the introduction of pests and diseases by imports only being allowed from countries and localities where serious diseases are non-existent. The Department of Agriculture through its exchanges of scientific publications is well informed of the occurrence of pests and diseases in other countries, and therefore can afford assistance to the principal industries of the Colony by having control of the imports of certain living plants. Fumigation and disinfection of certain imports after inspection is provided for, in order to safeguard against the introduction of diseases and pests of a cosmopolitan character, and the provision for plantings to be made in nurseries is made so that insects or diseases which may by chance have passed through the port of entry inspections and treatment may be detected before they have become spread in the country. By Proclamation 52 of 1913 it is provided that the importation into Mauritius from any country or place whatever, including the Dependencies of Mauritius, of all living insects except such as may be required by the Government for scientific purposes is absolutely prohibited. (687) F2 ar ] _ 1 Nit, ORR, cael Lew Rony mink is a ih mh ut 4 dinoan’ ae ‘y a a nine 8 we , i VAP cesta 048 au is eas “ i j ’ wii: , eee a ioe j 4 abe yet veoneih We ete akg Male sid ti Al teby «ik done Wh debt pean tain ; edtaiye af ry citer: wii So sage { uh fs ry _ tied ) pega ae ebm nba als hora i thi it Cn . piyashee 0 atirenisnt od Pages , “ya uigveind eto ad iy ) ven be tye aE Beri nee eet wes “as . Mei hel i Me | lee Pilla weit Lite aie ‘ a Heh cage 7 +) PINE (0, OTs danevry: MOLT : Ky i aceite ‘tind 1 ay devin Laren HE eels FE ARV E) yy hie cai ait “yh ACEP ty el Veg Fads 1H ib vec Bt ite Pee cand wit aa ah = L | i eroit agent wel waky tactis 140 bray go tid inka , ah eer Yet vole ry oid J LS Tae wiih ir ; ay Ha ie Pmt Aes fe me sarily: ta) v6 ak) eS OU ange OP A a Hiab te | iio is d . ; as ait a i 94 ttn t See trie at a on were Qlaliy we eae ra a ion Jqueze ete as yaives yh a) Vial itt at | ; pad & ‘ vj i $. t an Powit ~) as ; " : ee a jie her 2 rv 6 Om A iS a Y ts Us Gait Sin 7.5 ae my . ce is Wee ex hve deere ". 7 PAO™, fe N aputeyn d Fl OG, h (tee Fig. 2. Apex of mentum of Simuliwm larvae ; all x 150. a, S. ornatum; b, S. subornatum; ec, S.reptans ; d, 8. equinum ; 'e, 8. latipes ; f, S. subex- cisum ; g, S. hirtipes (after Malloch) ; hs S. tredeci- matum. The hairs on the sides of the mentum are not shown, but are of diagnostic importance, as described in the text. \ ame ia, || | l | b } d e f. Fig. 3. Antennae of Simuliuwm larvae; all x50. a, S. ornatum; b, S. subornatum ; c, S. equinum ; d, S. latipes; e, S. subexcisum; f, S. hirtipes (this last after Johannsen). Smaller species; gills with 4-6 (rarely 8) branches ; abdominal banding usually conspicuous ; dark “ eyebrow ” usually absent ; clypeal markings various but never of the ornatum type .. Ho std ne ey. i, 10. Thoracic proleg with strong chitinisations behind the circle of hooks... variegatum, Mg. These chitinisations much less evident ; skin round anus with more numerous scales as oh fs ra by ies .- monticola, Fried. ki. 12. 13. 14. 15. 16. ON THE BRITISH SPECIES OF SIMULIUM—II. 221 Clypeus with a single dark spot posteriorly, or none iP x i= .. Clypeus otherwise marked... wi Yr, we ie ar Pity OA A dark area in middle of posterior margin of clypeus, and another on each side of the head before the eyes ; body colour greenish. . reptans, L. Head all pale ; body colour greyish .. ds ig .. tuberosum, Ldst. A large black mark in the form of an H near the middle of the clypeus ; body colour yellowish-green .. SP on Pe .. morsitans, Kdw. A pair of dark spots on oduaned more or less connected anteriorly ; body colour bright green sit .. venustum, Say. Gills branched ; clypeus with six a: ee not Aino well defined; ‘“‘ eye- brow ” without dark dot es ¥ em . latipes, Mg. Gills simple (at least in S. aureum): a dark dot in the anterior part of the ** eyebrow ” re i. Ss a bee my a ad BS: Clypeus with a median dark streak, more or less interrupted, and four other spots aig sig Ee e™ be. os i aureum, Fries. Clypeus with eight distinct dark spots, including two transverse bars at each posterior corner ake 5a fe - angustitarsis, Ldst. Antennae very long, blackish, 9—joited ; central tooth of mentum simple subexcisum, Edw. Antennae short, pale at least ame 3- tes central tooth of mentum strongly trifid oe + a hirtipes, Fries. Pupae. td Prothoracic respiratory organs in the form of thick tubes, the main branches of which encircle the thorax at the nearly circular mouth of the cocdon 2 We - ve ne .. equinum (L.) Edw. Respiratory organs consisting of a number of slender filaments... sn. es Respiratory filaments with numerous ultimate branches (50-60 or more) : cocoon a loose shapeless network i af hirtipes, Fries. Respiratory filaments with not more than thirteen ultimate branches ; cocoons closely woven and shaped like the toe of a slipper, usually with semi- circular opening one ap = - os im Fl adits 9 Respiratory filaments eight .. Wi i. - 2s Se i ae 2 a SIX se at Te a i “he ts 9, _ os Tic 2. ~ es uh se - ees | s thirteen a as a tredecematum, sp. Nn. Cocoon with a long anterior dorsal projection : the upper two pairs of respiratory filaments with long vertical stalks .. - 2) yerburyt; spr Nn: Cocoon without anterior dorsal projection ; respiratory filaments otherwise 5. Filaments arranged in four shortly-stalked pairs .. via 8 Filaments not in pairs ; all rather widely divergent slniee to the Sati towards which 3 of them are noticeably thickened . subornatum, sp. n. (713) C 222 F, W. EDWARDS. 6, Filaments much shorter than the pupa, all of equal thickness, the upper pair shorter than the others ; all four pairs with extremely short stalks : cocoon with ‘‘ windows” on each side in front reptans, L., and var. galeratum, n. Filaments about as long as the pupa; upper pair slightly thicker but no shorter than the others ; cocoon without ‘“* windows ” % ae Fig. 4. show branching; allx35: a, S. ornatum ; b, S. ornatum, var. ; ¢, S. ornatum var. nitidifrons ; d, S. subornatum ; e, 8. reptans; f, S. morsitans ; g, S. variegatum ; h, S. monticola ; i, S. venustum ; 9, S. tuberosum; k, 8S. argyreatum ; 1, 8. latipes; m, S.aureum; n, 8S. angustitarsis ; 0, 8S. subexcisum ; p, 8. yerburyt ; q, S. tredeci- matum. Bases of prothoracic respiratory organs of Simuliwm pupae, side view, to 7. The four lower filaments arising from a short common stalk at the base ; then with longer separate stalks, of which that of the lower pair is the longer morsitans, Edw. The three upper pairs of filaments separate from the lowest pair at the base 8. 8. The stalks of the two lower pairs of filaments usually longer than those of the two upper pairs All four pairs of filaments with short stalks of about equal length ornatum var. mitidifrons, n. ornatum, Mg. ON THE BRITISH SPECIES OF SIMULIUM—II, 223 9. Cocoon with a long anterior projection ; upper pair of filaments with a long vertical stalk ; lower filaments twice dichotomously branched from base subexcisum, Edw. Cocoon without anterior dorsal projection ; filaments otherwise .. 10. 10. Filaments all widely divergent at the base, the dapat one mere vertical for a short distance _ argyreatum, Mg. Filaments not widely divergent at bie, all more or less in the same vertical plane, arranged in three shortly stalked pairs yy A Ova: 11. Filaments somewhat divergent, all about the same thickness ; cocoon all closely woven, as long as the pupa ye .. venustum, Say. Filaments scarcely divergent, the lower pair cae thinner than the upper 12. 12. Cocoon short, usually not extending beyond the middle of the thorax ; filaments longer than the pupa, directed straight forwards .. tuberosum, Ldst. Cocoon longer than the pupa; filaments shorter... . 4 ites 13. Cocoon all closely woven ; upper pair of filaments considerably thicker than the lower fe ; ne *. os monticola, Fried. Cocoon loosely woven at hd sides in dant, upper pair of filaments not quite so stout ¥: oo af iE 4 ¥ .. varregatum, Mg. 14. Filaments in two distinctly stalked pairs ; cocoon with anterior projection . latipes, Mg. Filaments all arising from a common base, ornearlyso .. ey gee 15 15. Filaments very stout at the base ; cocoon with anterior projection (short or long) o Ye +. angustitarsis, Ldst. Filaments slender at the ee cocoon without anterior projection aureum, Fries. 1. §. ornatum, Mg. (figs. la, 2a, 3a, 4a & Db). Larva.—General colour variable, usually greenish grey or dingy greenish-yellow, without special markings on the body. Head light yellowish-brown ; a median wedge-shaped dark mark near posterior margin of clypeus ; a spot on each side of this and sometimes a trace of a fourth anterior to the wedge-shaped mark ; usually also pairs of marks on each side at the posterior margin ; a dark stripe above and in front ofthe eyes. Antennae 5-jointed ; first joint usually a little shorter than second, which is often very imperfectly separated from the first and sometimes obscurely divided into two in the middle ; fourth longer than third ; last joint very short and conical. Mentum* with three teeth much larger than the others ; between the central tooth and each of the other large teeth are three small ones ; sides of mentum with four to six long hairs. Last abdominal segment without ventral papillae. Anal sucker with about 70 rows of hooks each containing 12-15 hooks about 10ulong. Anal gills 3, simple. Skin round anus beset with numerous minute scales dark in colour, and plainly visible under a magnification of 100, split apically into a number of spines (3-5). * Often incorrectly spoken of as the labium. (713) a2 224 F. W. EDWARDS. Pupa.—Respiratory filaments about as long as the pupa, on each side 8 in number, arranged in 4 shortly stalked pairs, the stalks of the two ventral pairs usually the longest ; ventral pair slightly thinner than the others ; all the filaments more or less in one plane (the vertical). Cocoon tough, closely woven, about twice as long as its breadth in front ; height and breadth about equal ; anterior margin thickened, but without any forwardly-projecting piece in the middle ; no floor to the front half of the cocoon, which has its mouth widely open, and entirely covers the pupa. Habitat.—Principally a river species, but occurs also in small streams, some- times even those of a temporary nature. Where water-plants are present, the larvae and pupae attach themselves to the plants, and show a preference for those with long narrow leaves, such as Sparganium and water-grasses ; Ranunculus is less favoured, Potamogeton less still; usually the larvae are found chiefly on the lower sides, the pupae chiefly on the upper sides of the leaves. In the smaller streams, where water-plants may be few or absent, the larvae and pupae are to be found on or under stones or on any suitable objects. The species is rarely met with in streams less than 4 feet in width. Breeding Season.—This species appears to breed in suitable localities all the year round, almost regardless of the season. Pupae have been collected in every month of the year except January, from the middle of February until Christmas. On one occasion (19.11.1917) fully developed pupae were collected immediately after a period of three weeks continuous frost ; these produced adults of both sexes a few days later. I took a newly emerged male in the open on Ist March 1919. There appear to be three main broods, in April, July and September. Material collected.—Larvae and pupae have been collected as follows :—Bucks: | R. Misbourne, near Denham, 10.ix.15.; Burnham Beeches, iv. 1916, a few in stony stream with S. latvpes. Middlesex: Harefield, 21.1x.15, in stony and weedy stream ; Pinner, 22—24.vii.15. Surrey: Beddington, near Mitcham, 31.vu.19 ; larvae in millions in river Wandle, pupae also numerous, but no adults seen; a few pieces of leaf (totalling one yard in length) of Sparganium simplex carried approxi- mately 500 pupae and 4,000 larvae, apparently all of this species. Devon: Larvae, and pupae numerous in many small streams in the valleys of the Dart and Teign, up to a height of 1,000 ft. ; also in the rivers Otter and Sid, and in small runnels on the sea-cliffs near Sidmouth, vi. 1920. Hants: New Forest, 1.v.1920, a few in small streams and in Beaulieu River, in company with large numbers of var. nitidifrons ; River Test (M. E. Mosely). Herts.: River Chess, near Chorley Wood, 30.vi1.15; River Lea, near Hatfield, v.1916 and 12.11.1917, on Ranunculus ; Pegsdon, 21.iv.16, on grass in small stream in “ The Meg,” larvae very large and very variable in colour; Hitchin, 5 i.15, under stones in stream 2-3 inches deep ; Radwell, 24.v.20, larvae of all sizes, very variable in colour, also some pupae, on stones in small rapid streams ; R. Ivel, near Baldock, vi.1917, on Ranunculus, Spar- ganium, etc. Beds.: Barton, 21.x.16, under stones. Cambs.: Bartlow, Hinxton, and Ashdon, 2.x.15, on grass and willow rootlets in small streams, none on stones in stream bed; Wittlesford, 29.xii.15 and 3.vii.16, larvae on stones in small stream ; Stapleford, 28.iv.16, on grass at edge of small stream ; Shepreth, 19.1.17 and 14.1v.17, on Ranunculus and Oenanthe in small stream; R. Granta, near Hauxton, 7.iv.17, on sedge at edge of river; Cambridge Botanical Gardens, 1.x.15, one pupa with ON THE BRITISH SPECIES OF SIMULIUM—II. 225 numerous S. awreum on tree-roots in streamlet. Suffolk: River Lark, near Milden- hall, 25.iv.16, in large numbers on various water-plants, chiefly Ranunculus. Kent: Bushyruff, Dover, 21.ii1.1898 (B. L. Hesse). Wates: Anglesey, 19.ix.15, on stones in small stream across field (C. B. Williams). Channels Is.: Guernsey, Christmas 1910, all stages together (K. G. Blair). Scotland: near Edinburgh, 27.1x.15, on grass at edges of small stream flowing through fields; current flowing at rate of 6 yards in 10 seconds (W. Evans). Variation.—In some localities the larvae of S. ornatum are fairly constant, in others they vary greatly in colour. Thus among 4,000 larvae from Mitcham, roughly examined, no variation was noticeable, while specimens from Pegsdon, Herts., and Harefield, Mx., varied in general colour from almost white to dark greenish, blackish or reddish; occasional specimens have dark brown markings as in S. aureum. The head markings are usually distinct, but may be obsolete ; very dark specimens sometimes have the head almost all black. The larvae found in small streams seem to exhibit much more variation than those found in rivers, the greatest variation being shown by parasitised larvae. This may possibly indicate that S. ornatum is normally a river species, the variation being induced by the change of conditions to those found in small streams. The pupae exhibit considerable variation in the length of the stalks of the pairs of respiratory filaments, but only in one specimen, out of hundreds examined, was any divergence seen in the number of filaments. This specimen had seven filaments on one side, eight on the other. The adults vary in size, and as is the case in some other species, the spring broods seem to be larger than the summer ones, thgugh this is not always very marked. The male varies to a remarkable extent in the silvery markings on the thorax, but only individually ; a close study of the variation might however possibly indicate the existence of definite local or seasonal forms. One such form has been recognised in the New Forest, and is described below. Most males have the thoracic markings similar to those of the female, and no more conspicuous ; some have conspicuous rounded silvery spots much as in the males of S. reptans, etc., while I have seen one or two beautifully marked examples with the whole of the margin of the mesonotum silvery, as well as the usual spots. All these are structurally typical; the only other character which I have noted to vary to any extent is the amount of black on the hind tibiae. The chief variation in the female is in the amount of white on the hind margins of the abdominal segments, which is sometimes very conspicuous, outlining three blackish areas on each segment. To this form Curtis’ name trefasciatum will apply. It is evident that S. ornatum is at present a plastic species some derivatives of which have attained stability, while others are probably now in process of development and fixation. It is therefore not surprising that the American representative (S. pzscicidiwm) should appear to be specifically distinct. Blood-sucking.—Since my previous notes on this subject were published, I have had abundant evidence that this species is a blood-sucker, but it would appear not to bite except on warm and rather still days, chiefly in the afternoon sun. Mr. C. L. Walton sent me some specimens from Crosswood, Cardiganshire, which he took in August 1916 attacking horses about the nose and eyes. I have also taken S. ornatum 226 F. W. EDWARDS. (Letchworth, 5.iv 18) on horses’ belles, and (Letchworth, 19.x.17) in numbers on calves’ bellies, some containing blood. Besides this I have frequently taken females, sometimes apparently gorged, in cowhouses. Mr. W. Evans has noted the species biting his hand. I have also myself received bites from them on my nose, eyebrow and hand (25.iv.16, 19.x.17, 15.x1.17, 5-18 iv.19). The bites were not particularly painful and left no after-effects ; most of them were inflicted while I was in a stooping position (e.g., when gardening), and I frequently noticed that the flies were more in evidence then than while I was standing. Perhaps they mistake a man stooping for a horse or cow, which would seem to be their normal hosts. heferences.—The early stages of this species have been described in more or less detail by a number of writers. The best figure of the pupa is that of Castellani and Chalmers (Manual of Tropical Medicine, Ed. i. p. 812) ; the figure is not named specifically as that of S. ornatum, but is quite unmistakeable. A larva and pupa of S. ornatum have also been figured by Kollar (Sitzb. k. Akad. Wiss. Math.-Nat. KI. i, pp. 92-107, pls. iu, 1848, under the name S. serrceum, Mg ) ; Griinberg (Die Siisswasserfauna Deutschlands, 2a, p. 107, figs. 128-130, 1910); Verdat (Naturw. Anz. alle. Schweiz. Ges. v, p. 65, 1822—as S. serzcea) ; Liebe (Zeits. Naturw. 82, p. 345, 1910—larva only); Meinert (K. Danske Vid. Selsk. Skr. (6) ii, p. 373, 1886). The most recent writer is Friederichs (Zeits. f. angew. Entom. vi, pp. 61-83, 1919), who figures the male hypopygium and the pural filaments of this species as S. reptans; with this use of the name veptans I cannotagree. It isnot clear what species he would indicate as S. ornatum ; possibly the one described below as S. subornatum. la. S. ornatum var. nitidifrons, nov. (fig. 4c). Closely resembles S. ornatum, Mg., except that the frons of the female is shining and not dull ; the mesonotum of the female appears distinctly more shining than in the typical form ; and there is little or no white on the hind margins of the abdominal segments. The male does not differ perceptibly from that of S ornatum. Larva with the head markings rather more distinct. Pupa with eight branches to the respiratory organs, all about equal in diameter, arranged in four pairs, all almost sessile or with very short stalks of about equal length. This form was abundant in the New Forest at the beginning of May 1920; larvae and pupae were collected on different water-plants in the Brockenhurst River, the Beaulieu River, and some small streams running into the latter. In the small streams it was the most abundant form Specimens of the normal form of S. ornatum were also reared from the same streams, but these were comparatively scarce. A few females were taken on the wing at Gidleigh Park, 8. Devon, 2.v1.1920, but no larvae or pupae were found. Both sexes were reared from pupae found in company with S. ornatum near Shelve, Salop, vu.1920, on grass in a small stream. Dr. G. Enderlein has since sent me specimens of both sexes (identified as S. reptans) from Berlin; he also sent specimens, probably from the same batch, to Friederichs, who has described them as S. reptans. | This form bears a great resemblance to the Scotch and Welsh S. monticola; the latter is regarded as a distinct species, since it differs markedly from S. ornatum in the larvae, pupae and male hypopygium, whereas the New Forest form apparently does not. ON THE BRITISH SPECIES OF SIMULIUM—II. 207 2. §. subornatum, sp. n. (figs. lc, 2b, 3b, 4d, 7b). Adult.—Male: Thorax black, with fine golden pubescence and double silvery markings in front as in typical S. ornatwm. Membranous area of pleura bare. Silvery markings on the sides of the abdomen not very conspicuous, but extending on to the 8th segment as well as the 6th and 7th (which is not the case in S. ornatum). Genitalia resembling those of S ornatum, but the stylus of the adminiculum (viewed. from beneath) is much longer and narrower, more like S. reptans, but distinctly pubescent. Front tibiae silvery; front tarsi thickened, the metatarsus about 51 times as long as broad. Middle tibiae yellow at the base Hind tibiae and metatarsi practically all black or dark brown, the latter moderately swollen. Female: Face and frons dull grey. Thorax marked as in S. ornatum, but rather less distinctly ; pubescence of mesonotum finer. Abdomen with the first five segments dull, blackish, with white hind margins, the last three moderately shining, but rather less so than usual. Legs coloured practically as in S. ornatum, but rather less brightly ; middle and hind tibiae sometimes with traces of dark rings near the base, especially on the under-side ; claws simple. Other characters as in the male. Larva.—Head markings similar to those of S. morsitans, but the H is not so dark and much less distinct. Abdomen without distinct banding. Antennae apparently 6-jointed, but the first four joints not very distinctly separated ; the first about as long as the second and third together, the second, third and fourth about equal in length. Mentum with 9 poimted teeth in the terminal row, the middle one and the one at each end larger. Sides of mentum with 5 or 6 long hairs. About 80 rows of hooks in the anal sucker, 12-15 hooks in each row. Skin round anus with a number of minute spines or scales, which are smaller, less numerous and less conspicuous than those of S. ornatum. No ventral tubercles on last segment. Anal gills retracted in all specimens examined. Pupa.—Respiratory filaments 8 in number, the upper four rather widely divergent at the base, not arranged in pairs ; the uppermost one nearly vertical for some distance ; the upper three arise from a common base or a very short common stalk ; the next one is somewhat isolated ; the remaining four are in two pairs with very short stalks ; the three filaments on the inner side of the bunch are rather distinctly thickened towards the base. Cocoon differs from that of S. ornatum in being more loosely woven, with “ windows ”’ on each side in front. Material collected.—I first recognised this species from two females sent me from Bayswater, Oxford, 15.vi.16 (A. H. Hamm); the same collector subsequently obtained a third at the same place, 8.vi.18. About 10 ¢ and 59 were received for determination in 1917 from Bulwell Hall Park, Notts., 21.vi.16. and 18.vii.16 (J.W. Carr). Some of these were presented to the British Museum, and I designate one of the males as type. I myself found a small swarm of the males hovering under a lime tree at Norton Bury Farm, Letchworth, 16.vi.17, and again at exactly the same spot on about the same date in 1918 ; in each case all the members of this swarm were this species, although S. ornatum was swarming under another tree close by ; I also obtained a single male by the river Severn at Shrewsbury, 24.vii.1920. A scrutiny of the series of S. ornatum in the British Museum revealed a single male of S. sub- ornatum from Llangammarch Wells, Brecknock (Lt.-Col. Yerbury). Finally 228 F. W. EDWARDS. Mr. M. E. Mosely has presented to the British Museum a number of pupae and a few larvae ; masses of the former were in a jar in Mr. Mosely’s possession collected by the late F. M. Halford and labelled “ Crickmere, 16.ix.1894;” Mr. Mosely also collected a few pupae himself in the River Test, Hants. The species must be very similar to the North American S. piscicadium, but that is described as having yellow legs. 3. §. variegatum, Mg. (figs. 1b, 49). Larva.—General colour brownish or dark grey, without special markings on the body. Head usually dark at the sides above and below the eyes ; clypeal markings often faint, when present usually ill-defined, but arranged much as in S. ornatum. Antennae 5-jointed, the first as long as the next two or three together ; second some- times indistinctly divided in the middle, twice as long as the third ; third shorter than the fourth. Mentum with 9 teeth in the terminal row, all shghtly trilobed, the small lateral lobes pale ; middle tooth very large ; about 8 long hairs on each side almost in a straight line. Thoracic proleg with a pair of rather strongly chitinised plates above the ring of teeth. Last abdominal segment without ventral papillae. Anal sucker with about 85 rows, each with 12-15 hooks. Anal gills each with 8-10 rather long branches. Skin round anus with numerous minute dark scales, composed of short rows of spines (5-8 in each) ; the scales are much smaller and less numerous than those of S. ornatum. Pupa.—Respiratory filaments about as long as the pupa, six in number, in three very shortly stalked pairs ; all the filaments more or less in the same plane ; upper pair rather thicker than the others. Cocoon very large, fully a third longer than the pupa, in texture close-woven behind and above, but with an open network on the sides towards the front ; this network often, but not invariably, continued round the front of the cocoon and enclosing the ends of the respiratory tubes, the cocoon then approaching the “ boot ’’-shape of S. equinum. Habitat—Swift stony rivers and streams, both larvae and pupae on the upper surfaces of stones where the current 1s strong but smooth and the water not too deep. Material collected—Larvae and pupae were found in great abundance in early June 1920 in South Devon: in the River Dart through Holne Chase, the River Teign between Fingle Bridge and Christow, and the river Yeo near Ashburton, associated with S. tuberosum and S. reptans but in much larger numbers. A female specimen hatched from a pupa collected in 1912 at Ewbank Scar, Kirkby Stephen, Westmoreland (C. B. Williams), in company with larvae and pupae of S. monticola, Fried. Blood-sucking.—In August 1916 Mr. C. L. Walton sent me females from Crosswood, Cardiganshire, which were taken in company with S. ornatum “ attacking horses about nose and eyes and on belly.” Variation.—There seems to be great variation in the colour of the larval head, some being much darker than others, Reference.—Friederichs has described the larva, pupa and adult of this species as S. venefica, sp. n. ON THE BRITISH SPECIES OF SIMULIUM—II, 229 4. §. monticoia, Fried. (figs. 1b, 4h, 5a). Adult.—Resembles S. ornatum in all respects except the following -—Membranous area of pleura bare ; front metatarsi about 6} times as long as broad ; male genital claspers broader, the external margin convex in the middle ; frons of female blackish, moderately shining ; abdomen of female entirely black. From 8S. variegatum the main if not the only differences are the shape of the male claspers, and the shining frons and dark femora of the female. This description agrees with that of S. arcticum, Mall., described from the female only from British Columbia, but without definite proof the two can hardly be considered the same. Larva.—General colour dirty greenish-grey, with obscure darker abdominal banding (on first 5 segments). Head very dark at the sides, except just round the eye-spots ; clypeus dark on its posterior margin, in the middle with a short dark longitudinal mark in contact with the dark margin, and slightly anterior to this a pair of inconspicuous dark spots. Antennae as in S. ornatum. Mentum with 9 teeth in the terminal row, the central tooth and the one at each side rather larger than the others ; sides of mentum with 8-10,long hairs. No ventral papillae on last abdominal segment. Anal sucker with about 100 rows, each containing 15-18 hooks. Skin round anus as in S. ornatum. Gills each with about 8 branches. Fig. 5. Male hypopygium, seen from beneath : a, S. monticola ; b, S. subornatum. Pupa.—Respiratory filaments barely as long as the pupa, six in number, arranged in three pairs with very short stalks ; all the filaments in a vertical plane ; upper pair slightly swollen at the base and thicker than the other two pairs. Cocoon as in S. ornatum. Habitat.—Moderately small swift streams, usually on rocks or stones ; both larvae and pupae usually on the upper or down-stream side. Sometimes present on stones covered with moss. Breeding Season.—The first brood hatches in May, and there is certainly another brood, since a gravid female was taken near water in the middle of June, and full- grown larvae have been found in August. Materval collected.—The species was fairly common in most of the small burns in the north half of the Isle of Arran, 22.v.—7vi.19 ; there were few larvae left by the end of this time. A specimen was also taken in a wooded valley at Ffrith, Flintshire, 10. vi.19, and in the stream near by were numerous small larvae, possibly representing a second brood of this species. One or two specimens were reared from pupae 230 F. W. EDWARDS. collected on stones in a rapid mountain stream at Kirkby Stephen, Westmoreland, 22.vii.1912 (C. B. Williams). Inverleithen, near Peebles, 7.viii.1910, adult males on the wing, larvae of various sizes and some pupae on stones and some on grass (Dr. J. Rettie, per Dr. J. H. Ashworth). Various streams in the neighbourhood of Pontesbury and Church Stretton, Salop, vu.1920. Blood-sucking.—No definite evidence, but a few specimens (of both sexes) entered tent. ; References.—Friederichs, in the paper cited in the introduction (Oct. 1920), describes this species from the Harz Mountains. I have distributed it to various correspondents under the MS. name obreptans. 5. §. reptans, L. (figs. 1d, 2c, 4e). Larva.—General colour greyish. Head with a single dark spot of varying size and shape in the middle of the posterior margin of the clypeus ; no dark mark above eye-spots. Abdomen with conspicuous dark greenish bands on all the segments dorsally, those on 6-8 nearly confluent. No ventral papillae on last segment. Gills each with 4 to 5 short branches. Skin round anus with numerous minute almost colourless simple scales. Teeth of mentum rather blunt, the nine in the terminal row all slightly trifid, the small lateral lobes pale. Antenna 5-jointed, gradually tapering to the end of the 3rd; first scarcely as long as the next two together, second nearly twice as long as the third. About 75 rows of hooks in the anal sucker, 12-15 hooks in each row. Pupa.—Respiratory filaments much shorter than the pupa, 8 in number on each side, spreading but little from the vertical plane ; arranged in 4 very shortly stalked pairs ; uppermost pair rather shorter than the others ; all the stalks equal in length and all the filaments of equal diameter, slightly and evenly tapering towards the tips. Cocoon as in S. ornatum, except that on each side towards the front there is a rather large open space, usually crossed horizontally by one or two bands of silk. Halitat—Swift shady rivers in mountainous regions, but not at high altitudes. Where water-plants (such as Ceratophyllum) are present the larvae and pupae occur on them in some numbers, but are most typically found on the flat upper surfaces of large stones where the current is strongest, and particularly in those places where the flow is smooth and even and where there is no great depth of water. No larvae were found in the bare valleys above the tree-line, and very few in places where the rocks were much covered with moss. Breeding season.—The first brood of the year does not appear until the end of May or the beginning of June. I have had no opportunity of ascertaining whether there is a second brood, nor what is the method of over-wintering. Material collected.—The species was abundant in Machrie Water and in the burn flowing through Brodick Woods, Arran, v—vi.1919, in company with S. tuberosum. It did not occur in the smaller burns on the island. Also found in Devonshire, in small numbers in the Upper Teign and in the Yeo near Ashburton, and more numerous in the Otter at Tipton St. John and in the Sid at Sidmouth, in the former case in company with large numbers of S. variegatum, and in the latter with S. ornatum and other species, 1-18.vi.1920. A single male was taken on the wing near Church Stretton, Salop, 21.vii.1920. ON THE BRITISH SPECIES OF SIMULIUM—II. 231 Variation —None noticed, except, as already mentioned, the size of the single dark spot on the head of the larva. Blood-sucking.—I have little to add to my previous notes except that up to the time I left Arran (5.vi.19) I was not attacked, and no females were even seen on the wing in Devonshire (up till 17th June). Probably the insect requires to be on the wing for some time before proceeding to its blood meal. Mr. P. H. Grimshaw writes “The specimens from Rannoch repeatedly flew at my face, in fact chiefly at my eyelids, and proved very irritating. They did not actually bite, but possibly would have done if I had given them the chance.” 5a. §. reptans var. galeratum, nov. (fig. le). Adults.—Not positively identified, but probably identical with S. reptans. Larva.—General colour light greenish, with darker but not very conspicuous trans- verse bands on theabdomen. Head yellowish, clypeus witha black spot inthe middle of the posterior margin, connected anteriorly with a large roundish black mark which occupies almost the entire breadth of the clypeus ; its edges are not very sharply defined and it has a small anterior emargination. A dark rim above and behind the eye-spots, sometimes entirely surrounding them. Antennae apparently only 4-jomted, the first joint longer than the remaining three together, second about equal to the third (the division between what are usually the first two joints seems to have disappeared). Mentum as in S. reptans. The pigment-spots within the sides of the sixth and seventh abdominal segments are darker and more conspicuous than usual. No ventral papillae on last abdominal segment. Gills either quite simple or with one or two (rarely three) short branches. Skin round anus with a few minute colourless scales. About 65-70 rows of hooks in the anal sucker, 10—15 hooks in each row. Pupa.—Apparently indistinguishable from that of S. reptans. The respiratory filaments as dissected from the fully developed larva are as in S. reptans, and the single pupa collected that can with certainty be referred to this species (owing to its still retaining the larval head-capsule) showed no obvious difference either in the pupa itself or in the structure of the cocoon from S. reptans. This pupa was unfortunately not isolated, and the adult issuing was lost among a number of S. reptans. Material collected—Larvae were abundant on stones and water-plants (Spar- gamum, etc.) in the river Otter at Tipton St. John, 8. Devon, 14.vi.1920, in company with many S. ornatum and S. reptans, and a few S. equinum and S. aureum. Although this species was the most abundant in the larval state, even more so than S. ornatum, no adults were reared from pupae which could be correlated with these larvae, all the specimens obtained being either S. reptans (in the largest numbers), S. ornatum or S. equinum. This would seem to suggest that the form under consideration may be only a local form of S. reptans which has developed distinct characteristics in the larval state only ; but against this it must be stated that a small number of larvae were obtained at the same time and place which agreed in every respect with the Scotch form of S. reptans as described above. Moreover, in the river Teign, and even in the Sid, two or three miles distant, only normal 252 F, W. EDWARDS. larvae of S. reptans could be found. Since, however, the only structural difference between this form and S. reptans is in the simple or less branched gills, it seems on the whole most probable that it is a variety of reptans, and I have so regarded it pending further investigation. The name has reference to the cap-like black patch on the head of the larva, a most striking feature. 6. S. morsitans, Edw. (figs. Lf, 4/). Adult.—Both sexes have now been obtained by breeding. The female is very difficult to distinguish from that of venustum ; there seems to be no constant difference in the colour of the base of the abdomen, as I previously thought, but this species may perhaps be separated (in the female sex) by the greater amount of yellow on — the middle tibiae. The only clear distinction in the adult, however, appears to be in the shape of the male claspers. Larva.—General colour rather pale yellowish-green without distinct darker markings. Head yellowish, the clypeus with a conspicuous and rather sharply defined black mark in the form of an H. Antennae 4-jointed, the second joint a little longer than the first, third scarcely as long as second, fourth minute. Mentum with 9 teeth in the terminal row, the middle one long, the one at each end very broad, intermediate ones indistinctly trilobed ; four or five long hairs at the sides. Last abdominal segment without ventral papillae. Anal sucker with 65-70 rows of hooks, about 15-20in eachrow. Skinround anus bare. Gulls with 5 or 6 branches all approximately equal in size. Pupa.—Respiratory filaments 8, about as long as the pupa, in four pairs, the two lower pairs slightly thinner than the upper, and arising from a short common stalk ; each of the four pairs also with its own stalk, that of the lowest pair distinctly, often much, longer than the others ; all the filaments in the vertical plane, not at all swollen towards the base. Cocoon as in S. ornatum. Halitat—Weedy rivers of moderate current in company with 8S. ornatum, S. venustum and other species. Not found as yet in small streams. Breeding season.—There may be two broods in the year, as adults have been obtained in May and July, but the May brood is certainly the more numerous. Material collected—Two males were reared from pupae found in the river Granta at Hauxton, Camb., 5.vi.15, on Scirpus in company with large numbers of S. equnum and S. argyreatum. A fair number of larvae and pupae were found on grass and Ranunculus at the edge of the Lymington River, near Brockenhurst Bridge, New Forest, 2.v. 20. 7. S. venustum, Say (austeni, Edw.) (figs. lg, 42). Adult female-—A number of females have been obtained by rearing ; they appear to differ from those of S. morsitans as shown in the table. I have failed to discover a better distinction. Larva.—General colour bright green, the abdomen with rather indistinct bands of darker green. Head light in colour, with a pair of somewhat oval dark spots at some distance from the neck, lying close together, connected at or near their anterior ends by a transverse bar ; usually also a dark area along the posterior margin of the clypeus. (The areas which in most species are dark are here pale, with dark areas ON THE BRITISH SPECIES OF SIMULIUM—II. 233 more or less surrounding them.) Antennae and mentum as in S. morsitans, the mentum however with a broader black terminal portion. Anal sucker and gills as in S. morsitans. Skin round anus with some very minute scales, inconspicuous _ even under a magnification of 300. Pupa.—Respiratory filaments about as long as the pupa, six in number, arranged in three distinctly stalked pairs, the stalk of the uppermost pair nearly vertical, slightly longer and thinner than that of the middle pair. All the filaments in the vertical plane at the base, slightly and evenly tapering towards the tips, bases not at all swollen. Cocoon as in S. ornatum. Habitat.—Weedy rivers of moderate current. Breeding Season.—The first brood appears at the end of April; we have yet no clear evidence as to whether there isasecond. grahami i as a 3 yA | Ochlerotatus irritans .. - ee! | 3 2:7 Stegomyia fasciata .. ed ees 3 td Culex tigripes .. ; | 2 1:8 » thalassius 2 1:8 Total | 111 This list might be regarded as corroborating the remarks above on the influence of the agbo pot, and of the sanitary inspector, on the preponderance of Culicoomyia over Stegomyia, were it not that Ingram and Macfie warn us against accepting such a collection (probably made chiefly in the evening) as evidence of the relative absence of S. fasciata, a scarcity which is often demonstrable even in blocks where its larvae are predominant, and for which a possible explanation is that this species “ whilst entering houses to obtain a feed of blood, does not roost indoors ” (Bull. Ent. Res. vii, p. 173). More abundant collections under proper conditions of observation should settle this point, but material of this nature from Lagos is at present insufficient. There seems no doubt that under native social conditions in Lagos C. nebulosa has become a highly domesticated species, and this habit along with its actual abundance suggests the need for a study of its bionomics and potentialities as a possible carrier of disease. Many mosquitos have been captured for identification at various times in the Kuropean quarters to the east of the Race-course and around the Golf-course. The following list (p. 268) shows one such classified collection, in which note was taken of the numbers obtained during one month only; in a total of 1,682 specimens a rich mosquito fauna is represented including 26 species. June is the month of greatest rainfall and A. costalis preponderates. Much has been done since then in the area in question in the way of filling hollows and draining. ‘The abundance of crab-holes in the locality no doubt influences the order of precedence amongst the Culicines. O. domesticus takes a comparatively high place, though its larvae have only once been discovered in a neighbouring swamp ; they have previously been found in borrow-pits by the railway on the mainland. A smaller collection of 150 mosquitos from the same houses, covering the period 15th May to 21st July, 1917, is shown overleaf; 16 species are represented. J. M. DALZIEL. TABLE IX, Adult Mosquitos captured in European Quarters, June 1914. Species. No. me eto Percentage of Culicines. Ochlerotatus irritans 112 26:23 s nigricephalus 70 16°39 Culiciomyia nebulosa 53 12-41 Culex thalassius .. # 45 10°54 Ochlerotatus domesticus .. 25 5°85 Culex decens 23 5°38 Stegomyia fasciata 13 3°04 Mansonioides africanus TO * 2°34 Culex consimilis .. 9 2°11 Ochlerotatus cumminsi 8 1°88 Culex fatigans 7 1:64 Uranotaenia annulata 7 1°64 Culex grahami 6 1-40 » ligrupes .. a 5 1°16 Banksinella punctocostalis 5 1:16 Culex duttoni ae 4 "93 » quasigelidus 2 °47 Ochlerotatus caliginosus es 2 °47 > argenteopunctatus .. 1 23 a punctothoracis 1 23 Culex salisburiensis ‘ 1 "23 Taemorhynchus annettt .. 1 23 % aurites .. a 1 23 Others undetermined .. ae 16 3°75 Total Culicines .. 427 Anopheles costalis se oe 1251 Culicines. 25°4% re mi .. dis a 3 Anophelines 74°6 % 9 mauritianus .. ats 1 Total Anophelines 1,255 Total Mosquitos 1,682 TABLE X. Species. Ochlerotatus irritans .. st. % migricephalus Culiciomyia nebulosa . . Oulex decens Anopheles costalis Ochlerotatus domesticus Culex grahami : » insignis 3 TING .. ay Taeniorhynchus aurite Culex thalassius » aduttont .. » pruina .. Stegomyia fasciata : luteocephala Anopheles nili S Total No. of Specimens obtained. ee DO OO HP IR CO 150 Percentage. mem DD 09 Pi igh: Agsindh. Aye woWIASAIASS each 0°7 CRAB-HOLES, TREES, AND OTHER MOSQUITO SOURCES IN LAGOS. 269 The following species obtained in this locality on other occasions may be added to the above lists, viz.:—Hodgesia sanguinis, Theo., O. ochraceus, Theo., C. unwittatus, Theo., Mansonioides uniformis, Theo., Taeniorhynchus metallicus, Theo., Urano- taenva balfouri, Theo., U. bilineata, Theo., var. frasert, Theo. The last-named was at one period fairly abundant in the neighbourhood of Government House, but the others are of rare occurrence. On 31st August 1915, a strong tornado from N.E. occurred about 6 p.m., in the course of which very numerous mosquitos were blown into the quarters occupied by the Medical Officer of Health. An examination of 36 specimens captured for identification gave the following curious result :—Culex decens, 339, 13; Culex rma, 12; M. africanus, 19. There is here a lack of correspondence with the record of mosquitos taken in the same rooms during the immediately preceding 3 months, which is as follows :— 403 mosquitos comprising A. costalis, 375 or 93 per cent.; O. nigricephalus, 8 ; C. nebulosa, 6; C. thalassius, 3; C. decens, 2; O. writans, 2; O. domesticus, 2; S. fasciata, 2; C. quasigelidus, 1; M. africanus, 1; T. metallicus, 1. It is probable that on this occasion some particular nidus close at hand was stirred up. A search further off in the Ikoyi Plains to windward of the populated area, where there are swamps and small surface pools, crab-holes, etc., as well as a few out-lying hamlets, revealed chiefly A. costalis, O. wrritans, U. annulata, and M. africanus. The numerous breeding places on the island beyond the limits of the inhabited town, and again the mangrove and brackish swamps beyond the island itself across the creeks to south and east, are doubtless capable of furnishing Lagos, at least intermittently, with wind-blown mosquitos, and possibly some species rarely found as latvae in the local sources may have their origin in these regions beyond. In the period dealt with in these notes which may be taken as from May 1914 to August 1918, the number of different species which have come to notice in Lagos Municipal Area has been 46,7 viz.:— ANOPHELINAE. Anopheles costalis, Theo., A. funestus, Giles, A. mauritianus, Grandpré, A. obscurus, Griinb.,* A. nili, Theo.* CULICINAE. Banksinella punctocostalis, Theo.* Culex ataenratus, Theo., C. consimilis, Newst., C. decens, Theo., C. duttoni, Theo., C. fatigans, Wied., C. grahami, Theo., C. guiarti, Blanch., C. imsignis, Carter, C. pruina, Theo.,* C. quasigelidus, Theo.,* C. rima, Theo., C. salisburiensis, Theo., C. sitiens, Wied., C. thalassius, Theo., C. tgripes, Grandpré, C. univittatus, Theo.* t To these 46 may be added 10 species absent from this list but included amongst the 41 species recorded from Lagos and published in the pamphlet of the Yellow Fever (West Africa) Commission, ‘ Distribution of Mosquitoes in West Africa,” Aug. 1913, viz. Anopheles pharoénsis, Theo., Aédomyia africana, N.L., Banksinella lineatopennis, Ludl., Hretmopodites inornatus, Newst., Mimomyia mimomyiaformis, Newst., Ochlero- tatus longipalpis, Grinb., O. marshalli, Theo., Stegomyia apicoargentea, Theo., Toxorhyn- chites brevipalpis, Theo., Uranotaenia mashonaensis, Theo. (713) F 270 J. M. DALZIEL. Culiciomyia nebulosa, Theo. Eretmopodites quinquevittatus, Theo. Hodgesia sanguinis, Theo.* Micraédes inconspicuosus, Theo. Mansonioides africanus, Theo., M. uniformis, Theo.* Ochlerotatus apicoannulatus, Edw., O. argenteopunctatus, Theo.,* O. caligi- nosus, Grah.,* O. cumminsi, Theo.,* O. domesticus, Theo., O. irritans, Theo., O. nigricephalus, Theo., O. ochraceus, Theo.,* O. punctothoracis, Theo.* Stegomyia africana, Theo., S. fasciata, F., S. luteocephala, Newst. Taenorhynchus annettr, Theo.,* T. aurites, Theo.,* T. metallicus, Theo.* Uranotaenra annulata, Theo., U. balfourt, Theo.,* U. bilineata, Theo. var. frasert. (N.B.—Species marked with an asterisk have not been found in the larval state within the period dealt with in this paper, but the larvae of C. quasigelidus, O. caliginosus, O. punctothoracis, and U. balfoury have been recorded by Graham from borrow-pits at Yaba (Bull. Ent. Res. 1, pp. 32, 39, 50). Breeding-places of the other species are referred to in the text.) Of the 5 species of Anopheles, A. costalis is universally distributed over the island and is very abundant in the rains ; it also makes an appreciable figure in the returns from domestic receptacles. Its eradication is one of the chief objectives of sanitary effort. A, mauritianus is a rarity, but it has been bred from larvae in a roof-gutter and captured once or twice in European quarters ; the larvae have also been taken in a pond and in road puddles at Yaba on the mainland. A. funestus was bred from slightly brackish water in a canoe. A. obscurus was captured once in a native house, but has not been taken in the larval state. A. nili is found now and then in European houses in the Golf-course and Race- course residential area, and in offices and native houses in the town. Its larvae have not been obtained, but no doubt have their origin in the stagnant waters which abound on the island. The imagines when found have always been in such very scanty numbers that its influence as a malaria carrier seems almost negligible at present. The circumstances which favour the prevalence of one species over another are, however, incompletely understood, and as these may conceivably alter in unexpected ways or in conditions beyond immediate control, it is well to bear its presence in mind, and to discover and deal with its sources. For similar reasons the above notes and record of species found in Lagos within recent years, even if little more than a bare recital of names, may at any time prove of interest and possibly of practical use to future workers. SOME NEW INJURIOUS WEEVILS. By Guy A. K. Marswatt, C.M.G., D.Sc. (Puate VIL.) Subfamily BRACHYDERINAE. Tanymecus destructor, sp. nov. (Pl. vii, fig. 5). Colour piceous black ; the head with dark brown scaling above, except for a line of pale scales bordering the eye, and with pale buff scales beneath ; scaling of the prothorax brown or grey brown on the disk, an ill-defined dark brown stripe at the side, and grey or buff beneath ; scaling on the elytra grey or brown, with a very ill-defined darker stripe between striae 2 and 5, usually vanishing at a short distance from the base, but often traceable to behind the middle in the form of spots. 32. Head with shallow, confluent and longitudinally striolate punctation ; the forehead at its narrowest not narrower than the base of the rostrum, and without any central fovea or carina ; the eyes very short oval, moderately convex. Rostrum a little shorter than the head, gradually narrowed from base to apex, and the dorsal area more markedly so, so that both scrobes are visible at the same time directly from above for almost the whole length; the dorsal area flat, sculptured like the head, and with a mere trace of a carina on the anterior half, the apical margin only shallowly simuate. Antennae with the scape reaching the hind margin of the eye ; the funicle with joint 1 longer than 2, and 2 as long as 3+4; 3to7 subequal. Pro- thorax a little longer than broad (9) or a little broader than long (9), broadest at the middle, the sides rather strongly and regularly rounded ; the apex and base of equal width and both slightly arcuate, the apical margin rather oblique in lateral view ; the dorsum evenly convex and with fine confluent punctation throughout. Scutellum inconspicuous. LElytra subtruncate at the base, separately rounded at the apex, with very sloping shoulders and without any humeral angle, the sides gently rounded and markedly narrower in the than in the Q, and without any posterior callus; the striae fine, shallow and shallowly punctate; the intervals even, almost flat and finely shagreened ; the numerous short recumbent scale-like setae hardly distinguishable from the narrow true scales. Legs densely clothed with variegated scales and setae ; the front femora thicker than the others, the two anterior pairs bearing long erect hairs beneath on the basal half inthe ¢; the hind tarsi with joint 1 as long as 2+3, and 2 longer than 3. Length, 6-8 mm.; breadth, 2-3 mm. S. Ruopesia: Salisbury, 5,000 feet, xi-xii. 1894 (G.A.4.2.); Mazoe, 4,000 feet, xii. 1919 (R. W. Jack). Described from 17 3 g and 19 9 9Q. Of the described South African species, this insect is most nearly allied to T. inafectatus, Fhs., which has similarly sloping shoulders, the scape reaching to the hind margin of the eye, and similar hind tarsi. On the other hand the latter is a larger insect (10-12 mm.), with the rostrum as long as the head; joint 2 of the (713) F2 pi" a GUY A. K. MARSHALL. funicle is as long as 1, and longer than 2+3; the scales on the elytra are much smaller, etc. This wingless terrestrial species is abundant in the Salisbury district, and accor- ding to Mr. Jack, Government Entomologist, Southern Rhodesia, the adults some- times cause considerable damage to the maize crops by eating off the young plants as soon as they appear above ground, They also feed on the leaves of other crops, such as sweet potatoes. — Tanymecus agricola. sp. nov. (Pl. vu, fig. 6). Colour black, variegated with greyish buff scaling and light and dark brown setae; the pronotum with three indistinct dark stripes; the elytra with several indefinite interrupted dark lines. $2. Head with rather coarse longitudinally confluent punctation ; the forehead not narrower than the base of the rostrum and without any central fovea or carina ; the eyes almost circular and rather strongly convex. Rostrum as long as the head and parallel-sided, the apex asymmetrically emarginate in the middle, the nght side of the angle being rounded and the left straight; the dorsal area parallel- sided, shallowly impressed in the middle, with a well-marked narrow median carina, the sculpturing like that of the head. Antennae with the scape elongate, reaching beyond the front margin of the prothorax; the funicle with the two basal joints equal, 2 equal to 3+4, and 3 to 7 subequal. Prothorax longer than broad in both sexes, gently rounded at the sides, broadest a little before the middle, and with a faint apical constriction; the apex slightly narrower than the base, and both margins very gently arcuate, the apical margin oblique at the sides; the dorsum evenly convex, with close confluent punctation throughout ; the short broad recumbent setae pale on the areas of pale scaling and dark brown elsewhere. Scutellwm elongate, with dense whitish scales. Hlytra jointly sinuate at the base, with rounded humeral angles, parallel-sided from the shoulders to beyond the middle, and impressed on each side before the apex, which is jointly rounded ; the striae rather deep, especially towards the apex, but partly hidden by thescaling towards the sides, the punctures large at the base and becoming smaller behind ; the intervals not broader than the punctures on the disk, slightly convex and shagreened ; the pale areas formed of short oval or almost circular scales, mixed with numerous pale short flat curved setae, the dark areas only with similar dark brown setae. Legs grey, the femora with an indistinct pale spot beyond the middle; the front femora slightly thicker than the others, the two anterior pairs in the g clothed with long hairs beneath on the basal half; the hind tarsi with jomt 1 as long as 2 + 3, and 2 equal to 3. Length, 7-8 mm. ; breadth, 2:°2-2°8 mm. S. Rnopesta: Gwelo, i. 1920 (R. Lowe Thompson). Described from 2 3g and 2 99. Although very similar in general facies to 7’. destructor, this species may readily be distinguished by its longer and parallel-sided rostrum, more convex eyes, much longer scape, broad scales, etc. It has also been found feeding upon maize. SOME NEW INJURIOUS WEEVILS. 273 Subfamily OTIORRHYNCHINAE, isaniris ater, sp. nov. (Pl. vii, fig. 1). Colour uniform dull black, the elytra with a small spot formed of small elongate white scales near the base of the 5th row of punctures, and sometimes smaller ones on the 7th and 9th, also a lateral row of similar spots near the margin, and occasionally traces of another a little above it. $9. Head with coarse shallow punctation and a deep median furrow, and sparsely set with short spatulate white setae; the eyes moderately convex, deepest in the middle. Rostrum broad, gradually dilated at the apex, which is deeply sinuate in the middle, the epistome not carinate on its posterior edge; the dorsal area broad, regularly widening to the apex, broadly impressed in front, almost flat in the basal half, and there coarsely punctate and with a broad smooth median carina, which is highest at the base; the basal incision gently curved; a longitudinal lateral ridge running from the scrobe to the eye. Antennae black, with recumbent pale setae ; the scape not much curved, moderately slender and rather abruptly clavate ; the funicle with joints 1 to 6 progressively diminishing, 7 as long as 4. Prothorax broader than long, strongly rounded at the sides, broadest at the middle, truncate at base and apex, the base being narrowly carimate and a little broader than the apex ; the dorsum covered with large flattened contiguous granules mostly forming irregular pentagons, each with a setiferous puncture ; laterally the granules become elongate and less defined. Hlytra narrowly ovate in the 3, much broader in the Q, broadest not far from the base, each with about 18 rows of punctures running in pairs, but becoming rather irregular towards the sides and apex; the intervals between the pairs more regular than the intervening ones, no broader than the septa in the rows and becoming granulate on the apical third, each bearing a spaced row of short subrecumbent setae, which are longer and more erect on the apical area. Legs rather stout, clothed with coarse pale setae and sparse setiform white scales, the setae being much longer, denser and softer on the lower edge of all the tibiae in the g; the hind tibiae of the ¢ deeply excised on the inner edge for more than half their length from the apex, the excision flattened and smooth. Length, 8-8°75 mm. ; breadth, 3:2-4°2 mm. S. Ruopesia: Salisbury, ix. 1893, ete. (G.A.K.M.—type). NYASALAND: Blantyre, iv—v. 1910 (Dr. J. E. S. Old). The sculpturing of the pronotum will distinguish this insect from all the described species of [saniris, except I. acuticollis, Fst., which differs in having the prothorax sharply angulate behind on each side and in lacking the granules on the elytra. In general form and structure it much more resembles such species of Systates as S. brevicollis, Fst., and S. crenatipennis, Frm., but these differ in having only 10 rows of punctures on the elytra, which Faust considered to be the only character separating Systates from Isanirvs. This weevil occurs very commonly on trees of the genus Brachystegia, which form the principal component of the woodland on the Mashonaland plateau ; but Mr. R. W. Jack informs me that it is a very general feeder, attacking various cultivated plants, including citrus trees. Systates exaptus, sp. nov. (PI. vii, fig. 2). Colour black, with brownish grey “scaling, which is sufficiently thin on the pro- notum and disk of the elytra for the shiny integument to show through, but is denser at the sides and especially on the declivity of the elytra ; above the declivity an ill-defined common transverse dark chevron-shaped patch, the posterior edge of which is deeply sinuated. 274 GUY A. K. MARSHALL. 32. Head obsoletely punctate, the forehead with a short deep median furrow ; the furrow separating it from the rostrum almost straight ; the eyes strongly convex, deepest in the middle. Rostrum as long as broad ( Q) or slightly longer than broad (3), subparallel-sided in its basal half, then gradually dilated to the apex; the dorsal area alutaceous, shallowly impressed on each side of the sharp median carina, the lateral margins with shallow confluent punctures, very slightly costate, parallel in the basal half, then gradually diverging; the central carina broadly forked a little beyond the middle, the fork enclosing a large smooth subtriangular impressed area just behind the epistome; a low longitudinal ridge running from the scrobe. to the eye. Antennae with the scape slender, cylindrical, abruptly clavate and clothed with recumbent white setae and narrow setiform scales; formula of funicle: 1, 2, (3, 7), (4, 5, 6). Prothorax broader than long, gently rounded at the sides, broadest at the middle; the apex truncate or shallowly sinuate, a little narrower than the base, which is feebly marginate ; the dorsum closely set throughout with low contiguous shiny granules, each bearing a puncture containing an erect scale- like seta; laterally the granules are much reduced and interspersed towards the base with coarse shallow punctures. Elytra without any truncate margin at the base, very broadly ovate in the 9, much narrower in the 4, widest not far from the base, the shoulders broadly rounded, the posterior declivity steep in the § and almost perpendicular in the 2; the striae rather shallow, closely punctate, the punctures. as broad as (9) or narrower than (Q) the intervals, each of the latter bearing a single row of short erect flattened setae; the scales small, narrowly ovate, the palest ones broader. Legs with short pale setae and narrow scales; all the tibiae denticulate internally in both sexes, the hind pair in the ¢ broader, with the inner face flattened and smooth throughout and incurved at the apex, the lower edge. deeply bisinuate and with an obtuse tooth a little before the middle; in the 92 the hind tibia has a short smooth shallow hollow on the inner face near the apex. Venter of 3 with two widely separated small tubercles near the posterior edge of the Ist segment. Length, 6°5-7°5 mm. ; breadth, 3-3°6 mm. S. Ruopesia: Salisbury. i. 1895—i. 1905 (G.A.K.M.—type); Unmtali, xi. 1901 (G.4.K.M.); Mpudzi R., Umtali dist., xi. 1901 (G.4.K.M.); Mt. Chirinda, Melsetter dist., i. 1907 (C. F. M. Swynnerton). This common species appears to feed on a number of different herbaceous plants, and the adults have been observed by Mr. R. W. Jack to do appreciable injury to young maize plants,* * The following closely allied species occurs on the coast in the same latitude :— Systates beiranus, sp. n.—Closely resembling S. ewaptus in form and colouring, but smaller and differing in the following particulars : rostrum with the dorsal area slightly convex longitudinally in the basal half, the anterior impressed portion not triangular, but separated off by a low transverse ridge ; no longitudinal ridge behind the scrobe ; the epistome slightly asymmetrical, the apex of the curve formed by the delimiting carina being a little to the left of the median line, and the carina on the right side rather longer and higher than that on the left; the prothorax with a very low median costa in the anterior half ; the hind tibiae of 3 smooth and flattened internaily, but less incurved at the apex and without any tooth on the lower edge ; the elytra vertically truncate at the base; Ist ventrite without tubercles inj}. Length, 5-6 mm. ; breadth, aa mm. PortTuGUESE East AFRica: Beira, x.1900—x.1904 (P. A. Sheppard) ; S327 QO bo ~I Ol SOME NEW INJURIOUS WEEVILS. Systates chirindensis, sp. nov. 32. Dull black, uniformly and thinly clothed with very small grey or light brown scales, sometimes with a greenish reflection ; in the ° only there is a large common triangular denuded dark patch at the top of the declivity with its apex directed backwards. Head separated from the rostrum by a straight furrow, impunctate, and with a shallow median furrow that is longer than half the transverse one; the eyes moderately convex, deepest in the middle. Rostrum a little broader than long, very slightly narrowed from the base to the middle, then gradually widening to the apex ; the dorsal area almost flat, finely alutaceous, parallel-sided to the middle and then dilated, the margins slightly costate and very shallowly punctate, the median carina low and broadly forked beyond the middle, the fork enclosing a shallowly impressed subtriangular smooth bare area; no costa behind the scrobe. Antennae long and very slender, clothed with sparse fine recumbent setae; the scape cylindrical, abruptly clavate ; the relative lengths of the joints of the funicle thus: 1, 2, 3 (4, 5), (6, 7). Prothorax somewhat broader than long, moderately rounded at the sides, broadest a little before the middle, truncate at base and apex, the former not marginate and a little broader than the apex; the dorsum evenly set with low shiny granules, each with a setigerous puncture ; laterally the granules are much reduced. Slytra very narrowly ovate in the g, much broader in the 9, without any truncate margin at the base, the apices jointly rounded and produced downwards like a beak, especially in the 9; the striae shallow and containing punctures that are deep towards the base but diminish much behind ; the intervals broader than the punctures, slightly convex, each with a row of small low setigerous granules, the setae being longer and thinner in the Q, and also duphi- cated on the declivity of interval 1 in that sex. Legs thinly clothed with narrow pale scales, the hind femora with a broad dark band in the middle; in the 4, the front tibiae slightly bent inwards at two-thirds from the base and denticulate internally, the mid pair also denticulate and the lower edge very strongly curved in the apical half, the hind pair only slightly curved, the inner face broadly flattened and smooth almost to the base, with a single row of granules along its lower edge in the apical half and the upper edge granulate ; in the Q, the tibiae only slightly curved, the hind pair with a smooth flat area on the inner face bounded above and below with a ‘row of granules. Venter with ventrite 1 in the ¢ very broadly impressed at the base and with two widely separated small conical tubercles on the hind margin. Length, 64-8 mm. ; breadth, 2°6-3°6 mm. S. Raopes1a: Mount Chirinda, 4,000 ft., xu. 1901 and x. 1905 (G.A.K.d1L). Described from 14 gg and 15 99. From the very few species of Systates having granulated elytra, such as S. granosus, Fst., this insect may be distinguished by the absence of the vertically truncate edge at the base of the elytra. In 1915, Mr. C. F. M. Swynnerton found the adults of this species feeding in numbers on the leaves of coffee bushes on Mt. Chirinda. 276 GUY A. K. MARSHALL. Subfamily CRyPTORRHYNCHINAE. Genus Elytroteinus, nov. nom. Pteroporus, Fairmaire (nec Schonh. 1843), Ann. Soc. Ent. France, 1881, p. 307. To the generic characters given by Fairmaire the following may be added :— The mesosternum parallel-sided, much broader than one of the mid-coxae, the posterior edge of the cavity markedly overhanging ; the intercoxal process of the metasternum very broadly truncate, no metepisternal suture; ventrite 2 in the middle much longer than 3 + 4; the elytra with stria 10 complete, and the lateral margin rather strongly smuate at the base; the front femora with a stout tooth, and all the femora very shallowly sulcate beneath; the tibiae with a sharp carina on the dorsal edge ; the tarsal claws simple and divaricate. Elytroteinus subtruncatus, Frm. (Pl. vii, fig. 8). This is the only known species of the genus and was originally described from Fiji (op. cit. p. 308), but it has recently been found by Mr. O. H. Swezey at Honolulu attacking the roots of ginger (Hedychium coronarvum). Subfamily CALANDRINAE. Calandra shoreae, sp. nov. (Pl. vii, fig. 4). 3 2. Colour dull black or piceous, without scaling ; the rostrum, legs and antennae red-brown. Head with three transverse rows of punctures behind the eyes, the forehead with an elongate fovea. Rostrum elongate, slightly curved, gradually widened at the apex, with four dorsal rows of punctures and three lateral ones on each side, of which the uppermost disappears in the basal half; in the 9 the rostrum is very little longer than that of the g and the punctures are finer. Antennae with joints 3 to 5 of the funicle about as long as broad, the others longer. Prothoraz a little longer than broad, widest near the base, the sides slightly rounded and gradually narrowed to the apical constriction; the base slightly rounded in front of the scutellum and truncate on each side, the apex broadly sinuate; the dorsum coarsely and closely punctate, the transverse septa between the punctures being narrower and lower than the longitudinal ones, producing the effect of numerous undulating longitudinal carinae; the median carina well-marked, fairly straight, and running from the apica! constriction almost to the base; each puncture with a very short stout erect pale seta. Hlytra with the usual sculpturing characteristic of the genus, intervals 1, 3, 5, 7 and 9 being rather higher and much broader than the others, and each with a row of punctures almost as large as the quadrate ones in the striae ; the posterior callus feeble, and intervals 3 and 9 not elevated at the apex. Abdomen with the exposed part of the pygidium forming almost an equilateral triangle, flat and coarsely scrobiculate. Length, 3°2-4'4 mm. ; breadth, 1:2-1°8 mm. Inp1a: Dehra Dun, United Provinces (C. F. C. Beeson); Kheri, U.P. (Beeson) ; Gorakpur, U.P. (A. E. Osmaston); Shwegu, Burma (Beeson); Pusa, Bengal ; Adoni, Bellary dist., Madras; Pattikonda, Kurnul, Madras; Waltair, Madras (S. W. Kemp); Nilgiri Hills (WH. ZL. Andrewes). Mauritius (J. £. M. Brown). SOME NEW INJURIOUS WEEVILS. 277 Very closely allied to Calandra sculpturata, Gyl., from South and East Africa,* which has the pronotum somewhat similarly sculptured, but the longitudinal cari- nation is not so well marked in the basal half and the median carina is less distinct in the African species ; this also has a small spot composed of a few pale scales in the middle of the base of the pronotum, and another on the shoulder of each elytron ; and the exposed part of the pygidium is more elongate and transversely convex. C. shoreae attacks the seeds of the sal tree (Shorea robusta), as well as those of Dipterocarpus turbinatus. Prof. A. Hustache had already drawn up an M.S. description of this species upon two specimens from Mauritius, but he has kindly suggested that the present description should be published instead. Calandra glandium, sp. nov. (Pl. vu, fig. 3). Closely allied to the foregoing species, and differing principally in the following characters :—A rather larger insect and longer in proportion to its breadth ; rostrum of the 9 finely aciculate, rather dull, and as strongly punctate as that of the ¢ of C. shoreae; the rostrum of the g much shorter and straighter, rugosely punctate and with four dorsal rows of minute granules; the pronotum with large reticulate punctures, all the interspaces being equally raised and the median carina usually indefinite and abbreviated ; the elytra longer in proportion, the posterior callus more developed and intervals 3 and 9 elevated at the apex, so that the apical junction of striae 3 and 8 lies in a deep impression ; the rugose area of the pygidium more elongate and with a sharp median longitudinal carina. Length, 4°6-5'2 mm. ; breadth, 1°6—-1°8 mm. Inp1a: W. Almora, Kumaon, numerous specimens bred from acorns of Quercus ancana (H. G. Champion—type) ; Naini Tal, Kumaon, bred from acorns of Quercus dilatata (H. Marsden). Stenommatus musae, sp. nov. (Pl. vii, fig. 7). Colour dull piceous black, usually with a greyish sheen ; the antennae, legs and apical portion of the rostrum red-brown. Head short globular, smooth, rather shiny, with minute sparse punctures ; the eyes actually contiguous beneath. Rostrum () elongate, as long as the head and prothorax together, slightly curved, moderately stout, almost cylindrical but some- what flattened beneath, not dilated at the apex, opaque and with small close shallow punctures for four-fifths of its length, the apical fifth impunctate and very shiny; a deep lanceolate median furrow from the base to the antennae ; the lower surface impunctate, with a low shiny median carina; colour red-brown, with only the basal third or fourth part blackish ; in the 3, a little shorter and broader pro- portionately, and blackish from the base to the middle or beyond. Antennae with * Schénherr records this species from Bengal, but his description agrees well with a South African species. Prof. Aurivillius has kindly confirmed the identification on comparing one of my African specimens with the type of C. sculptwrata, which, he informs me, bears the label ‘‘ Cap. B. spei ? Drége.”” The Indian record seems almost certainly erroneous. Drége resided in South Africa, and all the other species described from his collection by Schonherr are from that country. 278 GUY A. K. MARSHALL. the scape short, just reaching the eye. Prothorax a little longer than broad, the. sides subparallel from the base to beyond the middle, then roundly narrowed and very deeply constricted near the apex, the constriction continued across the dorsum as a deep curved furrow ; the apical margin truncate or very slightly sinuate, the base truncate ; the dorsum gently convex transversely and evenly set throughout with small separated punctures, the interspaces being as broad as the punctures. themselves ; on the sides the punctures are rather larger and much closer together. Elytra elongate-ovate, slightly sinuate jomtly at the base, and without any hu- meral prominence ; the sides gently curved, broadest at a short distance from the base, then rather rapidly narrowing to the apex, which is broadly rounded ;. the striae containing rows of shallow distant punctures, which encroach slightly on the sides of the convex impunctate intervals. Length, 1°6-1'8 mm.; breadth, 0°6-0'7 mm. Hawalan Is.: 2 ¢ 3, 8 2 Q, Kaimuki, Oahu, 19.11.1920 (0. H. Swezey). The genotype, S. fryz, Woll., from Mexico, differs in its slightly larger size ; the. scape extends behind the eye ; the eyes are not quite confluent beneath; the head is opaque and coarsely punctate ; the rostrum is markedly shorter ; the prothorax. is broader and much more coarsely punctate, the interspaces being much narrower than the punctures, etc. This species was found by Mr. Swezey breeding in numbers in the root (corm): of a banana plant. ame me *. « wy ico > . as" ped Fee ’ : at : - 7 : > 7 4 ‘2 1) e * ! ‘ “i r. \ eG. | 8 ; ' i Fig. EXPLANATION OF PLATE VII. . Lsamris ater, sp.'n., c- . Systates exaptus, sp. D., . . Calandra glandium, sp. n., &. ‘ shoreae, sp. n., &. Tanymecus destructor, sp. n., . ms agricola, sp. D., 3. . Stenommatus musae, sp. Nn. . Elytroteinus subtruncatus, Frm. PraTe VIf. PART 8. XI. VOL, Butt. ENT. RESEARCH. INJURIOUS CURCULIONIDA. 279 THE HABITS OF A DIPTERON PREDACEOUS ON MOSQUITOS IN NYASALAND. By W. A. Lamporn, M.R.CS., L.R.C.P. On 12th September 1919, while watching at a pool near Kotakota, Lake Nyasa, the emergence of mosquitos from their pupae, which in the case of this particular species takes place mostly between 4.30 p.m. and dusk (about 6.30 at that time of year), the writer saw at rest on some scum on the surface of the water, an An- thomyiid fly of the genus Lispa, which after a few minutes made a sudden rush at a mosquito just emerging from its pupa-case. The mosquito with its wings still unexpanded and part of its legs and the hinder segments of its abdomen still within the pupa-case was absolutely at its mercy. The fly grappling it from behind with its front legs proceeded to thrust its proboscis into the thorax. The fly was by no means timid, and it was possible to lean over and approach one’s head to watch the proceedings so closely that 1t was even possible to determine by its antennae the sex of the unlucky mosquito; it was a male. The fly was captured before it had finished its meal, lest it should suddenly fly up and escape. Next morning the pool was again visited and quite a number of the same flies were observed on the scum. One was seen to make a sudden thrust with its pro- boscis into the water, at the edge of the scum, followed by a series of similar but smaller efforts, as if it had seized some body, though insecurely at first. Presently it raised itself and partly drew out of the water a mosquito larva, about half-grown, impaled on its proboscis, which it proceeded to suck. The larva was not seen to move, possibly being paralysed by some buccal secretion from the fly. Advantage was taken of the fly’s lack of timidity to push beneath the scum a broad stick on which it and its victim were gradually raised, so that the feeding process could be studied as near at hand as one wished. A good series of the flies, subsequently taken, were all removed in this manner, the stick being thrust into the mouth of a net which was then closed. One fly was seen on the same occasion to grasp and thrust its proboscis into a mosquito, a female, which had failed to emerge completely from its pupa and may or may not have been dead at the time; it was not seen to move at all. The rest of the series were all seen at one time or another attacking mosquito larvae, not always with success, but when successful dragging them out of the water and sucking their juices. On 13th September, the emergence of a Culex was witnessed almost from the first ; the thorax bursting its covering, the imago then extracting its head, the rest of the body following until the wings and some of the legs were free. At this point a Lispa, which had been resting about four inches off on some scum, approached deliberately on the surface film, grasped the still soft Culex with its forelegs, and thrust its proboscis into the thorax, dragging the mosquito beneath its body to a little piece of leaf, about two inches off, floating on the water. As the captor walked, one could see quite clearly the legs of the victim trailing beneath. The fly with its prey was then taken. The mosquito, a female, was still alive and moved its legs feebly. 280 W. A. LAMBORN. The attack of another Lispa in a similar way on a mosquito pupa was witnessed on the same occasion. The pupa impaled on the fly’s proboscis, was drawn from the water, fully extended, to a little bit of leaf on which the fly had been sitting. It was not seen to move, so that one is unable to assert that it was living when seized by the fly. At a later date many instances were witnessed of attacks by the flies on mos- quitos which had been unable to fly away on the previous night owing to the failure of their wings to expand properly. One fly had impaled in the usual way a helpless mosquito, and had dragged it on to a small stick in the water. On an attempt being made to secure it, the insect flew across the water bearing its prey to another support, a foot further off in the water. Threatened by another attempt it flew up bearing its prey and settled fully another foot away on a blade of grass where both were captured. Its prey, a female mosquito, moved freely after capture. The manoeuvres of another fly were precisely the same when it was menaced with the net. In view of the unsuspicious attitude of most of the flies previously cap- tured, this wariness was a little surprising, more especially as they were taken, with the water still in the shade, at an early hour when most insects of diurnal habits are prone to be sluggish in their perceptive faculties. The pressure of other work made it impossible to study the behaviour of at least four other Diptera, all having business on the water. One species appeared to act as a scavenger, endeavouring to extract nourishment out of empty pupa- cases and larval skins of mosquitos. Another small species, which was not captured, was almost certainly predaceous on mosquito larvae. It gave from time to time most careful and deliberate attention to the cleaning of its long, pointed proboscis, aiter which it gradually stalked the smaller mosquito larvae breathing at the surface film, endeavouring, seemingly, to thrust its proboscis into their air-tubes, and repeating this action again and again, though in no instance was a successful issue witnessed. In the same pool in which these Culex larvae occurred in swarms were a few predaceous larvae of a much larger species of mosquito, which every now and again attempted to secure one of the Culex larvae, causing them instantly to scatter. On the far side of the pool, which was about two feet across, was noticed a “ rat- tailed ” larva struggling on the surface, and on near approach it was seen that one of the larger species of mosquito larvae, almost fully grown, had seized it firmly at the point of junction of the “ tail” with the body. The rat-tailed larva wriggled and contorted itself in every possible way; its tail became twisted round the head of its adversary at least twice; and it endeavoured to escape by diving into the depths of the pool, but unsuccessfully. The mosquito larva never slackened hold and, when its victim approached the surface, immediately put up its respiratory siphon and probably nibbled away constantly, for by and by a portion of the cuticle of the larva was seen to be partly detached near the mouth of its assailant and to be trailing the water. Towards the end of a prolonged and desperate struggle a second similar mosquito larva joined in the attack, seizing the Syrphid larva at a point midway between the two extremities of its body. The Syrphid struggled still more violently. Its new assailant, better situated for attack at a vital spot HABITS OF A DIPTERON PREDACEOUS ON MOSQUITOS. 281 than the original one, very soon bit a hole through which more and more of the Syrphid’s viscera gradually protruded. Eventually this mosquito larva dropped off and wriggled away, taking a small portion with it, much of which was consumed at leisure. The Syrphid’s struggles thereafter got more and more feeble and in two or three minutes it straightened out and ceased to move, floating on the suriace, the first mosquito larva still holding to it. An attempt was then made to secure this larva, which however alarmed by the movements of the numerous Culex larvae through the water being disturbed by the net, suddenly let go and vanished. The precise species of larva is therefore impossible of determination ; all that can be recorded in referrence to it is that 1t was considerably larger than any of the Culex larvae; it had posterior respiratory spiracles only; it also suspended itself in the water rather more obliquely than the Culex larvae did.* It was not possible to capture any with the means at one’s disposal. Such a fate for these Syrphid larvae would seem to be by no means uncommon, for on the following morning two more, both dead and with viscera protruding, were seen floating on the surface. One rather wondered how it was that any of the Syrphid larvae, which were securely grouped on a piece of rotting cassava one and a half inches below with their tubes thrust up to the surface, ever came up to the top, thereby incurring such risks. The explanation is probably to be found in the occasional failure of the larvae to withdraw their life-lines from the surface in sufficient time to escape the sudden attack of the mosquito larvae and in their subsequent endeavours to set themselves free. When the sun was off the pool full-fed Syrphid larvae, with their tails already partly shrivelled, were seen to crawl out and to burrow in dry sand well above waterline, where their puparia were readily obtained. Some of these larvae placed on 12th September in a box containing dry sand pupated in about forty-eight hours, and the flies emerged on about 23rd September, the pupal stage having therefore occupied about twelve days. The species has been identified by the Imperial Bureau of Entomology as Phytomia curta, Lw. During the heat of the day these flies were frequently seen ovipositing in the wet sand at the margin of the pool. [The fact that Lispa preys on mosquito larvae has also been recorded by Mr. J. Mitford Atkinson (Jl. Trop. Med. xii, 1909, pp. 255-6 ; Howard, Dyar & Knab, Mosquitoes of N. America, i, 1912, p. 170), who in Hong Kong observed Lispa sinensis, Schiner, devouring larvae almost as long as their own bodies.—Ed.] * [The predaceous larvae were probably either Toxorhynchites brevipalpis, Theo., or Culex tigripes, Grp., both of which are known to prey on other mosquito larvae.—ED. } th ot em " fat oo yi ep iy le ne RICE) ther’ nye ence Re A indir: la ott ne igatansti we ort f rep ae tie ee ai sf hander fibdetocgy A abaling iat Bite het nuh ie hk aR A allel if iitbon Neth te dtalstnicth mls is | Da Fea Ce PORE warnice ey yyy ai viiaut rolaitt par Mid ‘ WH leo yet per hyilht ovale Oat Exdabaiy Dima eiies han ary arr ae de ih Baila) since foi hiatal, UeAplat ragel) tapi Cel ae i i ar ten eb bt pa TEs ees lh iS eter Ate “aight, ri F noah sot fa R an tt ao ; ‘ . ‘ i ‘% ‘ . is , . we | i } : ‘ ‘ r if 7 eee pi Piel % : wt eee | ys is Adapt aes 4 24 a rR we day a) i Ma ‘yp! 1 iia) } ; ; \ 2 : j ; c" 2 Mar idx Ah as BO mi Riki ‘ re +e >) Lai ANSE, | tnt ve ty; ra rth Ay, 1 ; a easing 8 : SP ah ale ifn ead 9 basa ch isignrbnsts ia: fthsad rif if Sih or ee hi m4 bi ARP? pe A, diy ral beta TY or ‘oh x 1 ae | i , i = ‘a i ie ia i oe Pane 4 hides (ia? PRS SEL RE ANA ao BD EO NBN ee ‘i bats wi ch ‘ORR n lh aa wich vase f, s jt ay 4 “ty an) tr Hager f pith 1; sbi St fa "a oi i, "Rist vaet nei yt it ute btn dll la geal sk ag pe a yi . full) ; Mis aol deg il iy Gude: gergiby yt Pigs * e J wind. t a oe So es aeveniianieerd.e Ah ae pinuns-ote f Wye ‘vy wii. ana ay, LU AePSHAR lh tt fai aa vit One ack . s 7 4 j y f fi ‘ ‘ | AM et , / ) ee eed corn Pye | ry ie ie * & y A‘ eer of a =~ y ‘ ' iy { < > y se hw i, H d q a, atic ade tts poms sels a2 erp i cagal Kiwilis aie i sik, at: neesling weiter? Vasil pi fs 1. IRE: Be ht: VLoLeebOaKe et etontstoubey sma: ie his i ane OTT sath spidignaet a ae * nat ¥ " wih, ee eta alt Seek SARL are RY wee MTT OF Nets {hy des HNN tai ies ; J Fi a3 iy ‘fay Pas UM oo , foe’ is! ‘# thi ie Ky Hien city ‘hye by hg We ie: 7 i " 4 a Pe ele J ae ee ae , oer | ii Pac tilt a io! SEiy ahi v a , ee? it sd pai’ Mai Bid ver | } Heth any wil ee ey hae ete ae ts i - Re ED LACE: Whee aT “CLE Lt Cana Pa hey "SVR tne pees: ives e ‘a f 4 es “ee? mous) Pring, Patty ae i . ad oe | : , y f & q ais ~ nF iv} f ian Ae 7 Oars ee ai v4 a nat Ciatigy ty Wika) ta ye “eathe tudy TRR MEAN + oT ved: ie bid. thar rier Irs ely vq ny el wh ein ‘ahi ri ‘isihelih ; De ie Ue OMCs rT Moe Ue a itt | pe ne Fuh dee Ae? We ELD he Lee ay 4 APA 'r, is r) oY Pina grin. 9 ' { t vite: Fi *) ” pi ] ay dat b4 ‘ Whe " ie t rye ’ (y Whe i Me TA Kae eel i hay Vd ate) ci ids ky it a Vi aie Ht. Aine ) f beet a fy Wy i) , sith oF Cai ahi ) f q } 7 ve i bmiaby ny big nit s vy i a AM Csaba: low ' ps et Ae a a 2 ae i is — ‘de “ ray ait a te errata} hh Be i Ay Grate a a u si sti fir a! ie He a nie ae Ay) ) tg Lal r i Be F Hie as i ht ORE DIEM as Tei F Soarby and ner NT RATER Rt RID hihi ink fone ove me yee at ae Oe ae) rk] henarnekd) ¢ Tak | Cwnqit eal ‘s ee ath vee, 3 yh gs “At a rs i, ve ow 4 f° py” Re OOK eS i! } spel at ihe vet nse, Lai i ah wee ye none f i u 4 . ar ! hee A as “elton ach (7 ait " has bel ne han Wat seers { SAY CARON iA, ugh: Evi en rah a | ve Paki ae erin i, ane shines i "\ ae he Ie / Wy bc al Pus i y F : AR f i ft Pig / ; ry 1 hy 1 evs OY P P a Phe ®t 1 ihe i ‘| ey TNs i ey 4 ‘ f ry a hy) bs ‘ ; : f bi y an” bod ny oe tS ya ibd, 2 » vgs a An ; ‘ ; ‘ hes oF , i) ae »* v iy ey , we 1 ‘ iT? Dh a _ eh ate vei i A + y ‘ ad N \ [ is \ ; My y } (after PC PM a OE areas wee, ey Abeyt 4 hi ass as ~ oe a rae , ¢ ‘ i i ant, en i 7] 1s : F shi F Ni . Yu ) ; ; ry or Peet ‘ined Pie ay bia) ' ei) | ie a vol veer cate plat ip MY ugar ii sik iy ‘aA N \ : At a ‘ Fi i * Ady eqn ne woltnge ry ri aa A Hee ML NL, WL Ni ea Otick ee sa a hing . / { , ath: p Ph Bei Tih ave ve ry Bes iy eae 909 slaty eas AS TN ony ee eee Bi us ae rane M Py | , ead ‘eo 6 Porat art ae , : age hs re el rs. ’ ‘ou ahd ar hai ; a a i an a hcg ot | Mlk hot ae f Y . Ae es met . pM Hi a aoa: ) Nes, Bil) y sa mone : od indy : mht lane, MNS P: Ay, H, 283 A NEW TYPE OF ENTOMOLOGICAL KILLING-BOTTLE. By Matcotm E. MacGrecor, Wellcome Field Laboratory, Surrey, Wellcome Bureau of Scientific Research. The “cyanide bottle” is never a very attractive part of the entomologist’s equipment. When made up in the old way, 7.e., plaster of Paris poured over lumps of solid potassium cyanide and the plaster then allowed to set, the contrivance has an attractive appearance, but is an abomination to use, especially if small insects are being dealt with. The large amount of water utilised in making up the plaster of Paris can never be dried out properly, and the potassium cyanide being deliquescent to some extent the presence of moisture is unavoidable. Moreover the cyanide soon decomposes and turns brown, imparting its colour to the plaster and forming an ugly-looking mass on which small insects are not contrasted and may be easily overlooked and lost. The newer method of making up the killing-bottle by powdering the cyanide and mixing this with about an equal weight of powdered borax, compressing the mass at the bottom of the bottle, and afterwards pouring over it a small amount of plaster just sufficient to secure the powders, is a distinct improvement in so far as the approach to anhydrous conditions is concerned and the more rapid liberation of HCN gas. However, the mass decomposes more rapidly with the addition of the borax, and the unfortunate brown coloration is accentuated. There are too the methods that depend on fixing the potassium cyanide in the cork instead of in the bottle, which are certainly improvements in that the insects are left lying on the glass, against which they can be easily seen; but with this method the cyanide is not in a very safe situation, and furthermore cannot so rapidly evolve as much HCN gas as cyanide that is distributed over the bottom of the bottle. In addition to the faults peculiar to whatever method is employed in making up the cyanide killing-bottle there are the following disadvantages in the use of cyanide :—(1) the bottle is a potential danger to the ignorant if not properly looked after, owing to the highly poisonous nature of potassium cyanide ; (2) if the bottle is accidentally broken, cases of poisoning may result unless every scrap of the cyanide and plaster is collected and destroyed; (3) the killing power of the preparation speedily weakens if the bottle is constantly used ; (4) insects recover more often from HCN gas in weak concentration than they do from concentrated chloroform vapour; (5) when the cyanide bottle is exhausted it is preferable to throw the whole thing away (unless the bottle is of the type where the cyanide is secured in the cork) rather than to attempt to chip out the hard plaster to renew the apparatus, and a new killing bottle has therefore to be obtained. On the strength of these objections I have for several years preferred to use chloroform as an insect poison, utilising chloroform-saturated rubber in place of cyanide. The method is ideal as far as its action is concerned, but the one great drawback has been the rapid evaporation of the chloroform and the constant addition of the fluid that has been necessary. 284 MALCOLM E. MACGREGOR. In order to have a killing-bottle which will have the great advantage of chloroform as the poison, the bottle described below is now used. Jt can be made from ordinary laboratory material, unless a very large-mouthed bottle is required. Select an ordinary “ boiling-tube ” (large form of test-tube), and a large rubber cork which will just slide down the inside of the tube, making rather a tight fit. Bore a hole in the centre of the cork so that a small piece of glass-tubing (internal diameter, 2 to 3 mm.) can be forced through. The glass-tubing should be of such a size that it projects about # inch on one side of the cork, and about + inch on the other. Now slide the cork into the boiling-tube so that the ? inch length of glass tubing projects towards the bottom of the boiling-tube, and force it down until the end of the tubing touches the bottom of the tube. If the rubber cork is a tight fit roinary cork. -- Glass bolling-tube. _ Za’ projection of glass tubing. on - Red rubbercork. s %’, ojection Chleroform.~ of glass tubing. Fig. 1. A chloroform killing-bottle. (as it should be) moisten the cork slightly with water and press it down the boiling- tube by the aid of the open end of a small test-tube. Warm the apparatus so as to drive off any moisture that may remain, and when dry it may be charged with chloroform. This is accomplished as follows :—Pour into the boiling-tube suff- cient chloroform to stand about 2 inches above the upper surface of the rubber cork, and then carefully warm the air-space below the cork over a spirit-lamp. The air will be partly driven out by expansion. Remove the tube from the flame and as the space cools the chloroform will run in to fill the partial vacuum. Then again heat the bottom of the tube until the chloroform within boils gently, and all the air is driven out by the chloroform vapour. On cooling once more the space will this time be completely filled with chloroform. Now invert the tube and A NEW TYPE OF ENTOMOLOGICAL KILLING-BOTTLE, 285 pour out the surplus chloroform from the upper surface of the cork. None of the chloroform in the space below will escape, owing to the air pressure at the upper open end of the glass tubing. Gently warm the sides of the tube above the cork to drive off any remnants of liquid chloroform, and when an ordinary cork has been fitted, we have an apparatus similar to that shown in fig. 1. Although pure rubber is soluble in chloroform, it will be found that the compound from which the rubber corks (red variety) are made, swells when in contact with chloroform but will not dissolve. The following are the advantages of the apparatus :—(1) A high concentration of chloroform vapour is always available; (2) if the tube is accidentally broken there is no danger of poisoning ; (3) 5 cc. of chloroform is sufficient for a week’s continuous use, and the tube is therefore exceedingly economical. Care should, however, be taken that the lower end of the tube containing the chloroform be not unduly heated by the hand, otherwise gaseous pressure will force the liquid up the central glass tubing. Should this occur at any time it is only necessary to shake out the few drops of liquid at the upper end of the cork, and blow into the tube to dry the cork surface. If the tube is inverted, no amount of overheating will do more harm than to dissipate the chloroform to some extent, and for this reason it should be carried in this position. The open lower end of the glass tubing then stands above the surface of the liquid chloroform, and the apparatus is “ sealed.” via a ‘* wate tel “» Pn! ‘ mht Aenareat we ei yl 7 y ai Rad je ah Ja hel " bier on ny, rabid dina | elit wii i‘ arte rehir i ! , oy, | ae is ® "1 rad er! it baw shies fe: a idly vin bart: uv rig di . add & ae ‘ “p i ce of rh he rhs ae a ae 1) ta . niinelk 1} hh | rasta vy) cehaitint ) w } t rt ‘ti AY bn fas ¢ vs y ae ee a tte ; rigs te At yeh hid pI Li Ny Ae rit a 7? yattned Be ae beak’ tre h) Mc A 287 INSECT PESTS OF DATES AND THE DATE PALM IN. MESOPOTAMIA AND ELSEWHERE. By P. A. Buxton, M.A., F.E.S., Fellow of Trinity College, Cambridge ; formerly Entomologist, Mesopotamian Expeditionary Force. (Pirate VIII). In the summer of 1918, there was every prospect of an unusually good crop of dates in Mesopotamia, until the month of June, when a large number of the half- grown dates suddenly became “ hashaf’’—an Arabic word signifying that they - dried up and fell from the trees. This was a very serious matter, not only because the date is the principal resource of the country, but also because any and every disaster was at that time attributed by a section of the population to the British administration. At the end of July I was lent to the Directorate of Agriculture to carry out an investigation upon the failure of the date crop. I found that the principal pest, the larva of a Gelechid moth, had finished its work, and had wan- dered away from the palms to pupate ; and though I was unable to solve the prob- lem of its life-history, I had the good fortune to discover a good deal that was new about the insect pests of the date palm. At the time I published a report (1918), to which I only refer now in order to say that it was produced on the spot and without access to any books ; it contains certain inaccuracies, and anything of value which I may have discovered will be found in the present paper. During the course of my investigations I visited all the important date-growing areas of Mesopotamia, that is to say, Baghdad and its neighbourhood, the Diyala River from Ba’qubah to the Persian Frontier at Baba Pillawi, Mendali and Balad Ruz, Basra and the country round it, Mohammerah and Fao, the Lower Euphrates from Nasiriyeh to Sugq-ash-Shuyukh, Qurnah and Amara. At the time of my visit the fruit was nearly ripe, so that it was not difficult to learn the distinctions between the different varieties of date palms, a point of great importance in view of the fact that they are liable to the various pests in varying degrees. I strongly recommend future investigators to familiarize themselves thoroughly with these varieties, and also with the very considerable number of Arabic words which are used specially for the date palm and its cultivation. This is no small task, as will be realized when I mention that in Mesopotamia about 70 words are employed by the peasants for various parts of the date palm ; the vocabulary relating to cultivation and methods of employment is just as extensive. The literature relating to the pests of the date is so scattered, some of it in the most obscure periodicals, that I believe this paper will meet a want. My aim in writing it has been to deal with the pests observed in Meopotamia, but to give at the same time a summary of all that has been written up till the present time, and a full bibliography. From the bibliography I have excluded all purely popular references to well-known pests, all reports, etc., which merely record the presence of one of these pests in some place in which it is well known to occur, and papers relating solely to the systematics of the pests. (713) G2 288 P. A. BUXTON. I am indebted to Prof. R. Newstead, Mr. E. E. Green, Mr. G. J. Arrow and Mr. J. H. Durrant for much help most kindly given, and to the staff of the Imperial Bureau of Entomology, who have always assisted me in every possible way. LEPIDOPTERA. Family HESPERIIDAE. Padraona palmarum, Moore. This butterfly is recorded by Stebbing from leaves of date palms at Calcutta. Family PYRALIDAE. Subfamily PHycirinaE (= ANERASTIINAE). Myelois phoenicis, Durrant. This species was described by Durrant (1915) from a specimen bred by Lord Walsingham in 1904 at Hammam-es-Salahin, Constantin, Algeria. The larva was found in dates. The species has smce been found in England and France in im- ported dates, and as Durrant remarks, “may be widely spread by commerce.” Plodia interpunctella, Hb. According to Popenoe the Indian meal moth is a serious pest of stored dates in Arizona. This is the only record I have found of this omnivorous and very widely distributed species attacking dates. Spermatophora hornigii, Léderer. Moths which I bred at Basra in September 1917, and Baghdad in September 1918, from larvae found in August feeding in wind-fall dates, have been identified by Lord Rothschild as this species. He informs me that Léderer’s original speci- mens also came from dates, but I have been unable to consult the original reference. The species appeared to be quite common in Mesopotamia, but was in no sense & pest. . Ephestia calidella, Guenée. Gough (1917) mentions that in Egypt the larvae of this species sometimes attack living dates on the tree, and bore into them through the rind. These larvae spin sulk at all times in their life-history, and lower themselves on a silken thread when they are alarmed. In these respects the habits of this insect differ from those of the Ephestia in the Khargeh oasis. This species occurs at Baghdad and Amara, but I have no evidence that it is a pest. Ephestia cautella, Walk. (Hphestia passulella, Barrett.) This species is a common warehouse pest in Basra, and probably throughout Mesopotamia. The larvae also enter windfall dates in the gardens in Baghdad, Amara, and Basra, passing into the fruit at the poimt at which it has become detached from the perianth. These windfall dates are frequently harvested by the less well-to-do, who in this way infect and reinfect their stores and godowns with E. cautella. It is a common thing to see masses of dates which have been riddled by this and other Pyralid moths exposed for sale in the bazaars ; these spoilt dates are bought for feeding domestic animals. INSECT PESTS OF DATES IN MESOPOTAMIA. 289 The larva of Z. cautella not only eats ripe dates which have fallen to the ground, but also the small “ hashaf’’ dates which have become hard and withered owing to the attacks of the Gelechiid moth (see below) while they were half grown. According to Smyth the life-history from egg to adult occupies five weeks in summer at Smyrna; in April, May and June about 48 days. Chittenden says that this species has a wide distribution in Europe, Asia, North Africa and North and South America; it was originally described from Ceylon. It has been recorded as attacking pomegranates, pears, cotton cake, cacao beans, gall-nuts, flax-seed meal, tonka beans, walnuts, pecan nuts, pea-nuts, figs, pearl hominy, dried moths (specimens), stored corn, asparagus berries, white rice, chick pea, zizyphus berries, etc. Chittenden and Smyth’s Bulletin should be consulted for details of the life-history of this pest, and for an account of the damage it does to stored figs. Buckler described its life-history (under the name Hphestia passulella) in 1882. It is parasitised in Egypt by Habrobracon kitchenert, Dudgeon & Gough. Ephestia sp. ? Gough (1917) describes the life-history of an Hphestia which is not identified but differs from £. cautella and E. calidella. This larva is a pest of fallen and harvested dates in the Dakhlah and Khargeh oases, which lie in the desert west of the Nile. The . life-history is given as follows. The female lays as many as 190 eggs in 24 hours, and this is not the full number; the eggs are deposited im ones and twos on the surface of the dates after they have fallen or been gathered. The larvae wander about and are not able to enter living dates, a point in which they differ from the larva of H. calidella. They generally enter a stored date by the calyx end, and take up their abode in the space between the flesh and the stone. After feeding for a minimum of 20 days, during which no silk is spun, the larvae emerge from the date and wander about, leaving a silk thread behind them; they wander for two days, leaving their thread on all the dates over which they pass, and then spin a cocoon in some convenient corner. The minimum life period from egg to adult is 25 days in August, 27 in September, and 38 in October, in Egypt. The species passes the winter in the larval stage. The adult moths are crepuscular and nocturnal, and never fly far. Gough suggests that arrangements should be made for marketing the dates in two grades, and that the first grade should be treated with SO, after they are packed ; the second grade should be pressed, a process in which many of the larvae are crushed. He also speaks of the employment of moth-proof receptacles. I imagine that it will be some years before the inhabitants of these desert oases use either moth-proof receptacles or sulphur fumigation, but they can probably be induced to gather up their wind-falls and give them to their cattle to eat. Subfamily GALLERIINAE. Arenipses sabella, Hampson. The larva of this species is known to eat stored dates, to which so far as we know it is confined ; in this respect it differs from E'phestia spp., which are pests of the 290 P. A. BUXTON. most catholic tastes, devouring almost all forms of stored produce. The larva is described by Durrant as dull pinkish-brown, head and plates darker, brownish spots darker, length 1 im. This species has been recorded from Fao, at the head of the Persian Gulf, in Arabia, also from Algeria and Persia (Hampson). As an imported species it has reached England, and has been reported from Canterbury by Hampson; in the British Museum there is a specimen bred from the larva described by Durrant; this larva was found by Miss B. Reed in dates in London. The species was common at Amara on the R. Tigris, males and females coming to ight from Aprilto June. It seems probable that Mesopotamia is one of the native countries of this sect. There is no record of it as a serious pest, but this may only be due to our ignorance of the moths whose larvae attack stored fruits. Family GELECHIIDAE. Genus and species unknown (text fig. 1). This was by far the most serious pest of the date palms in Mesopotamia in 1918, and it was on account of its ravages that I was ordered to report on the failure of the date crop during that season. Unfortunately I received this order after the larvae had left the dates to pupate, and it is only owing to a fortunate chance that I was able to obtain a few larvae which had been preserved in whisky by one of the officers of the Department of Agriculture. These specimens enable me to say that the pest is a Tineid of the family GELEcHIIDAE, if the tables given by Fracker for determining Lepidopterous larvae are correct. I have found his paper most useful, and I believe it to be accurate, but I think we must accept it with a certain amount of reservation, at any rate where the Microlepidoptera are concerned, because Fracker has been forced to base his conclusions on the examination of a comparatively small number of identified species of larvae. Text fig. 1 shows the chaetotaxy of this larva, and with the aid of it future investigators should be able to determine whether any pest with which they are dealing is identical with the one of which I am speaking. The explanation of the nomenclature of the setae will also be found in Fracker’s paper. Apart from chaetotaxy the larva may be described as follows. The adfrontals reach the vertical triangle, and the frontals about half way up to the vertical triangle. The superior angle of the frontals is not prolonged into a fine point. The arrangement of the ocelli is that characteristic of the GELECHIIDAE and not the OECOPHORIDAE, a point which cannot be described without reference to a figure (cf. Fracker, Pl. vit, figs. 80, 81). The body is slightly flattened, and is widest at the third and fourth abdominal segments ; the setae in general are short and inconspicuous, as is usual in internal feeders. The prothoracic shield is weakly chitinised and is not darker than the rest of the body ; the legs are short. The abdominal prolegs are also short, and the crochets on them are uniordinal, and arranged in a complete series: the crochets of the anal prolegs (text fig. 1,D) are arranged in a continuous transverserow. The prothoracic spiracle is oval, the abdominal ones circular, that on the 8th segment the largest. [ was informed that the larvae when alive were pink, without any markings. This pest has apparently not been recorded before, and is so far as we know confined to Mesopotamia. Its biology is very little known, but ‘its ravages were first detected by other observers in June. The half-ripe dates turned brown and INSECT PESTS OF DATES IN MESOPOTAMIA. 291 became spherical, and inside them the larva could always be found eating the flesh of the date along the stone; early in July the larva left the dates, which were by now quite dry and about half the size of a healthy date, making its egress by a hole bored close to the attachment of the perianth. From the hole one could always see frass and silk projecting. It is extremely rare to see two emergence holes from one date, and I believe that the eggs must be laid singly, probably on the fruit soon after it is set, and that the larva burrows into it and remains in a single date throughout its period of growth. All my efforts to find the place of pupation Fig. 1. Structural details of larva (Fam. Gelechiidae) causing damage to half-ripe dates in Mesopotamia. A. Chaetotaxy of pro and meso-thorax (I, II) and 3rd, 7th and 8th abdominal segments (3, 7, 8). The shield of the prothorax is drawn rather too wide and too deep. The head of the larva is supposed to be towards the left. C. Left proleg of 7th abdominal segment seen from below; the seta ¢ is in the direction of the mid-ventral line. D. Anal proleg seen from above. Small Greek letters (a, 8, 7, w, ete.) indicate individual setae, according to Fracker’s terminology. Greek capitals (B, K, etc.) indicate groups of setae, of regular occurrence in certain families or many families of the Lepidoptera, e.g., B=a-+8, P (capital rho)=e+tp, T=r+w+¢ (2), K=n-+x. Or., crochets of proleg; leg, proleg, or leg; sp., spiracle ; sh, dorsal shield of prothorax. were unsuccessful. I searched cracks on the palm stem, the fibrous leaf-sheaths and the rubbish which accumulates in the crown of the tree; I also passed earth irom beneath the trees through a sieve, without finding any pupae except those of large Nocrurpaz. I believe that pupation takes place in the ground, and lasts throughout the winter, because we always observed that the incidence of the pest was slight in gardens in which a crop of winter and spring vegetables was cultivated beneath the trees. I opened a very large number of the dates which had fallen prematurely as a result of the attack of these larvae, and found no dead larvae, so it appears that the pest is not controlled by parasitic Hymenoptera or entomophthorous fungi. 292 P. A. BUXTON. This pest attacks all varieties of date palm, but it was particularly harmful to the Khustawi in Baghdad and Ba’qubah, and the Sair (Ista’mran) and Halawi in the Basra neighbourhood. The age of the tree appeared to make no difference to the severity of the attack, but it was an invariable rule that a heavy loss was sustained by owners of gardens far from the river, and by those who failed to keep the ground beneath the trees clear of weeds. Palms growing in rather salt ground, and those irrigated by the rise and fall of the actual sea at Fao were not particularly attacked by this insect. It will be understood that only a few sorts of palm will live under these conditions, but those that lived did not appear to be any less healthy than individuals of the same sort irrigated by fresh water. I have already said that the pest was never very serious in gardens in which green vegetables were cultivated beneath the trees ; I believe that this was due to the breaking up of the soil, and consequent damage to pupae, but we cannot regard this as proved until we know more of the biology of the insect. It was suggested that the very heavy incidence of this pest in the year 1918 was due to the fact that many of the cul- tivators had taken work in the Labour Corps, with the result that the gardens were less dug and ploughed than usual. Until the life-history of this very serious pest is known, one cannot do better than impress on land-owners and cultivators the need to dig the ground beneath the trees in winter ; but if it is found that the female moth oviposits on the flowers of the female palm, or on the young fruit soon after it is formed, it might be worth while to protect the whole inflorescence with a muslin bag. Popenoe figures such bags in use to defend the ripening dates from wasps and birds. It is impossible to estimate the loss due to this larva in 1918; I found evidence of its ravages in every garden I visited from Baghdad to Fao, and from Nasiriyeh on the Lower Euphrates to Khaniqin on the Kurdish border. Its predilection for the Sair dates of Basra is serious, as this is the most widely grown of all sorts and its fruit is exported to India in immense quantities. The more costly sorts suffer least, because no man owns more than a few trees, and these are always well- watered and carefully cultivated ; I refer to such sorts as Barhe, Braim, Maktum, and Mirhaji. I made a rough estimate that the following percentages of the crop of all varieties were lost, between the time when the fruit set and the harvest ; at Ba’qubah 50-70 per cent., at Qurnah 40-50 per cent., at Basra 30-40 per cent. The Basra date forest is the largest in the world, and contains about 7,000,000 palms. One may assume that the average yield of a palm is 60 lb. of fruit, and it will be realized that the loss of about a third of the crop is an extremely serious matter. This is the pest which I wrongly referred to in my report (1918) as a Pyralid. The attacked dates are called “ hashaf” in Arabic. COLEOPTERA. Family DYNASTIDAE. Oryctes elegans, Prell. In Mesopotamia this species, for the identification of which I am indebted to Mr. G. J. Arrow, is common throughout the date-growing area: I have seen speci- mens from Baghdad, Ba’qubah, Balad Ruz, Mendali, Amara, Basra and Fao. INSECT PESTS OF DATES IN MESOPOTAMIA. 293 Adults fly to light from April to September and do great damage by boring through leaf-bases and stems of inflorescences in the crown of the tree. I have seen a garden littered with bunches of nearly ripe dates felled by this insect, which can be a most serious pest. I presume that the females lay their eggs in the burrows they make. Beetles of this species may also frequently be found in burrows in the hard trunks of living or dead palms, and I believe that these burrows are made either by the adult or the larva, but I have no proof that this is so. When these burrows are numerous they interfere seriously with the nutrition of the tree, and they sometimes cause it to break off in a high wind. The larva of a large Lamelli- corn, which I believe is this species, lives in the crown of the date- palm among the leaf-sheaths, etc., and bores its way through the bases of leaves, the bases of fruit-bearing stems and the young leaves before they unfold. Sometimes it bur- rows inwards towards the actual growing point. It appears to have a predilection for fairly young trees, between the 10th and 20th year. It is an interesting fact that the larva of O. rhinoceros occasionally adopts the same habit, and lives in the crown of coconut trees, especially if they are decayed (Burkill); it is of course well-known that this larva normally inhabits piles of vegetable refuse, and old thatch and palm leaves which are lying on the ground. Such a habit would I think be impossible for the larva of O. elegans ; for owing to the extreme desiccation of Mesopotamia it is generally impossible to find any decaying material of any sort in the gardens. Treatment.—The presence of the adult or larva in the crown of a palm is known from the immense quantity of chewed vegetable fibre which passes through them and falls from the mouth of the burrow. When he sees this the Arab cultivator climbs the tree and endeavours to extract the insect with the short heavy knife with which he prunes his palms, and in so doing he frequently commits much havoc himself. As Prell’s O. elegans is very little known and was described from the female I have asked Mr. Arrow to re-describe the species. His description which is as follows is founded on a series of ten collected by W. E. Evans, C. F. C. Beeson and others. “ Oryctes elegans, Prell. This species has hitherto been known from its female only, examples of that sex taken at Fao having been described in a German perio- dical in July 1914 (Entom. Mitt. in, p. 210.) It is one of the smallest repre- sentatives of the genus Oryctes, males ranging from 28 to 34 mm. in length, females from 34 to 36 mm. It is narrower in shape and rather less convex than O. deser- torum, Arrow (found in the same region) and the upper surface is more smooth and shining, the elytra being almost devoid of the punctation occurring on the latter species, but possessing a deeply incised line on each side of the sutural edge instead of the row of dots found in O. desertorum. . The two points of the clypeus diverge more widely and are separated by a rounded, instead of an angular excision. The armature of the head and prothorax scarcely differs from that of the other species, but the posterior edge of the thoracic cavity is rather more sharply toothed in the female, and rather less so in the male, which has only a very slight blunt projection there, not distinctly bicuspid as in O. desertorum. The cephalic horn 294 P. A. BUXTON. ‘ wi in the male is short, but in well-developed specimens longer than in the female. The thoracic cavity is similar in both sexes, except for the sharper posterior tooth in the female just mentioned. The pygidium is recurved in both sexes and has a fringe of red hairs on its free edge only; in the male it is convex and very smooth and shining; in the female it is more strongly inflexed and rather strongly and closely punctured. The last ventral segment in the male is also very smooth and its front and hind margins are straight and parallel; in the female it is strongly punctured, and forms a rounded lobe.”’ Oryctes desertorum, Arr. Oryctes desertorum, G. J. Arrow, Fauna Brit. India, Coleoptera Lamellicornia (Cetoniimae and Dynastinae) 1910, p. 276, The distribution of O. desertorum is given by Arrow as “ India, Karachi; Persia; Arabia, Masqat (Muskat), Lahej and Fao.” He tells me that the Persian specimen came from South Persia, so that all the records are from countries in which the date-palm is grown, It appears probable that O. desertorum will eventually prove to be a date-palm pest; at present however we know nothing of its habits or life- history. Oryctes rhinoceros, L. Burkill gives Phoemx dactylifera in a list of palms attacked by “ the Rhinoceros Beetle.” Stebbing records O. rhinoceros from date palms at Madras. On the other hand Copeland in his very careful account of the pests of the coconut does not mention the date palm in the list of trees which O. rhinoceros is known to attack, nor does Ghosh. O. rhinoceros is found from India eastwards, so that its range and that of the date palm only coincide in parts of India. It may well be that in that country it attacks this palm, as Stebbing has stated, but I do not feel that the matter is proved in view of the fact that the species of Oryctes have until recently been much confused. The life-history of O. rhinoceros is well-known, as it is a very serious pest of coconut. Ghosh’s and Burkill’s papers and Copeland’s book should be consulted for details about this member of the genus and the measures which have been adopted to destroy it. Family NITIDULIDAE. Carpophilus dimidiatus, F. Popenoe speaks of this species (as ‘‘ Carpophyllus dimidiatus”’) eating stored dates. The ordinary “ Dried Fruit Beetle” is Carpophilus hemipterus, L, The parts of Popenoe’s book which deal with pests of the date tree are not as valuable as the rest of the book. | Family ANOBIDAE. Coccotrypes ? dactyliperda, F’. Willcocks records small beetles, which he believes to be this species, from Amry and Aglawy dates from the Sharkia province of Egypt. As many as 11°8 per cent. of the latter variety have been found infected. The beetles live ina cavity hollowed out of the date stone, in which as many as forty eggs, larvae, pupae and imagines may be found. A small circular hole communicates between the cavity and the exterior. ‘ ia INSECT PESTS OF DATES IN MESOPOTAMIA. 295 Family CURCULIONIDAE. Rhynchophorus ferrugineus, Oliv. Popenoe states that the “‘ Coconut Palm Borer, Rhynchophorus ferrugineus, has killed date palms in India, and to less extent in Mesopotamia.’’ Opposite p. 168 he gives a plate of two dying palms nearly leafless, under the title “ Ravages of the Palm Borer.’’ However in a footnote he says that “ he did not see the borer itself in Baghdad,” and his record appears to rest on descriptions received from Arab sources. As far as I am aware this very large and conspicuous weevil has never been obtained in Mesopotamia. Captain W. Edgar Evans, R.A.M.C., never saw it though he lived in Amara, an important date-growing centre, for about a year and made a very extensive collection of beetles; it is absent also from one or two smaller collections of beetles received from that country by the British Museum. In India, on the other hand, this weevil is definitely known to attack the date palm, and MacKenna tells us that it is capable of inflicting considerable damage, It is found that the palm may be saved from attack if mud enclosures are built round the trunk and kept filled with water. Milne figures palms the tops of which have bent right down owing to this weevil. It is important to cover with earth or tar the raw surface from which an offset has been detached. This weevil is distributed from India to the Philippines (and possibly to New Guinea also). Burkill gives a list of half a dozen palms from which it has been recorded, the chief of them being the coconut, of which it is a major pest. It will possibly be found that other species of Rhynchophorus, e.g., the African f. phoemcis, occasionally attack date palms. THYSANOPTERA. Family THRIPIDAE. Heliothrips haemorrhoidalis, Bch. This insect is recorded by Williams as having been found on date palm in Trinidad. It is the species known as the ‘“‘ Greenhouse thrips,’ and has been recorded from every continent and has a long list of food-plants. It is not a pest of the date palm. RHYNCHOTA. Family Coccipae. Parlatoria blanchardi, Targ. Aonidia blanchardi, Targioni Tozzetti, Mem. Soc. Zool. France, v, pp. 69-82, 1892 (from Ourir, Shott Melrir, Algerian Sahara). Parlatoria victriz, Cockerell, Arizona Exp. Stn. Bull. 14, 1895 ; and Entomologist xxix, p. 52, 1896 (on imported date palms). ? Parlatoria proteus var. palmae, Maskell, Trans. and Proc. New Zeal. Inst. (1897) Xxx, p. 229, 1898 (on date palms imported from Algeria into Australia). Parlatoria (Websteriella) blanchardi, (Targ.), H. H. King, Rept. Wellcome Res. Labs. Khartoum, iii, p. 240, 1908. 296 P. A. BUXTON. Cockerell himself states that his description of Parlatoria victrix was made from dark specimens of the species previously named Aonidia blanchardi by Targioni Tozzetti. Distribution.—This pest has been recorded from the following localities : ALGERIA: Ourir, 8. of Shott Melrir. Eqaypr: Assouan, Cairo, Luxor. Supan: Khartoum and WN. Sudan generally. Iranian SoMALitaAnD: Merka. MesoporaMIA: Probably throughout the date area; certainly Basra and Amara. Inpia: Lyallpur, Punjab. Uwnirep States: Arizona, California. AUSTRALIA : Northern Territory. I can find no record of the occurrence of this scale-insect in Morocco, Tunisia, Tripoh, Arabia, or 8. Persia. P. blanchardi was introduced into the Southern States of the U.S.A. in 1890 from Cairo and Algiers (Cockerell), or in 1889 (Popenoe). It was introduced into Australia about 1894 (Maskell), but has apparently not yet reached South Africa. (Lounsbury). Biology.—In Mesopotamia this scale occurs on the upper side of the leaflets, more rarely on the midrib; it attacks particularly recently planted off-sets and the resultant young trees during the first half dozen years of their life. I saw it on various types of date palms at and near Basra, and on the variety known as “ Zehedi’’ at Amara. I did not notice it in the other parts of Mesopotamia to which I travelled in the course of the investigation. I never saw a tree severely attacked, even in the most neglected gardens far from the river. I believe that the comparative scarcity of this insect is due to the attacks of Aphelinus mytilaspidis (EULOPHIDAE). Males were easily obtained in October by enclosing leaflets attacked by the scale in a bottle. This scale on the date leaf is known to the Mesopotamian peasant as “ remaj ”’ ; I have not been able to find this word in any dictionary. The most complete account of this scale is the pamphlet by Cockerell and Forbes (1907), to which I am indebted for a great deal of information. It is clear from all the American authors that it is a very serious pest, not only because it weakens the palm by almost covering the leaves, which turn yellow when heavily attacked, but also because it invades the fruit and renders the dates unsightly and hardly fit for the market. We know that it is also a very serious pest in the old world ; King speaks of dates exposed for sale in the Sudan “smothered with” P. blanchard, and Targioni Tozzetti’s original description of the species mentions date leaves from the Sahara covered on both surfaces by overlapping scales of this species. Essig says it may kill the tree in California. Cockerell (1907) states that it is attended by the ant, Phezdole cockerelli, Wheeler, at Tempe, Arizona. He and Forbes (1907) make the interesting suggestion that the young insects may be carried from tree to tree by wind, birds, or rats, or by fruit-pickers and pruners. Young offsets are frequently infested by both this and Phoenacococcus at the time when they are pushing their way through the old matted leaf-bases left on the trunk of the parent tree. INSECT PESTS OF DATES IN MESOPOTAMIA. 297 Other Hosts.—Fide Cockerell (1907), Parlatoria blanchardi has been recorded by Newstead (1906) from jasmine foliage. King (1908) says Draper (1907) has recorded it from yellow jasmine and periwinkle. It does not appear to have been recorded from any other plant. Parasites and other Enemies —Cockerell (1907) records Coccinella abdominalis and Chilocorus cacti (COCCINELLIDAE) as feeding on this scale in America ; to this (Mthly. Bull. Sacramento, 1913, p. 139) we can add Scymnus sp. Popenoe states that there is a parasite which perforates this scale in Algeria, but it has not yet been investigated. No doubt this is some small Hymenopteron. Cockerell (1907) has once seen the emergence-hole of a Hymenopteron in a scale of P. blanchardi in America. In October 1918, at Amara, R. Tigris, in Mesopotamia, I bred a small Eulophid wasp from P. blanchard:. This parasite has been determined by Mr. J. Waterston as Aphelinus mytilaspidis, Baron. It was extremely common, and it appeared probable that it was an important agent in controlling the scale, which so far as I know is never a serious pest in Mesopotamia. It would probably not prove worth while to import this Eulophid into other countries in which the date is a main crop, because we know that the parasite is already widely distributed. According to Mercet, Aphelinus mytilaspidis has been found in U.S.A., Ceylon, Italy, France, and Spain, and is known to parasitize the following scales :—in America, Lepido- saphes ulma (Mytilaspis pomorum), Chionaspis pinifolir, and Diaspis carueli ; in Ceylon, Chionaspis permutans and C. graminis ; in Italy, Aspidiotus hederae and A. betulae; in Spain, A. hederae and Diaspis rosae. It has not apparently been previously recorded from Parlatoria blanchardi. Treatment.—Forbes (1907) states that treatment designed to exterminate Parla- torva blanchardi and Phoenicococcus marlatti has to be extremely thorough, because the scale-insects persist, even in small numbers, among the closely crowded bases of the great leaves. The reader must realize that the whole crown of the tree is enveloped in sheets of fibrous matting which arise between the petioles of the leaves where they are inserted on the trunk. Any scale-insects which happen to be in this tightly-bound tangle escape soap and kerosene emulsion, whale-oil soap, distillate spray, resin wash and even fumigation with hydrocyanic acid vapour. There re- main two treatments which are said to be efficacious. Wilsie (1913) tells us that after the great fire in San Francisco “the ornamental palms withstood the great heat, and put out new leaves at once after the fire.” This observation suggested the treatment described by Forbes (1913) ; the trees are pruned down to the crown and all the great leaf bases, and the fibre in which they are set, are removed as much as possible. When the tree has been reduced to a bare pole it is drenched with paraffin and fired, after which the surface is flamed with a gasolene blast, several times in the first few months, in order to destroy such few pests as may be finding refuge in cracks and similar places. This treatment does not kill the tree but des- troys the Parlatoria and Phoenicococcus. Popenoe (1913) says that the gasolene torch treatment puts the young palm back two years and is unpopular with the growers for that reason. He states that all offsets are now dipped bodily in a tank of cresol dip, and adds that “ this destroys the scale so that no danger need be feared from 298 P. A. BUXTON. it.’ A plate is given showing the process in operation. I have no experience of this method but do not understand how any dip can penetrate the spaces between the leaf bases. No one but those who have endeavoured to separate date leaves from the trunk or even from a small offset can realize how firmly the petiole is attached to the plant, and what an ideal refuge there is among all the fibrous material which surrounds the bases of the petioles. Parlatoria zizyphi, Lucas. Various early authors refer to this scale as a pest of the date palm, and an un- signed editorial article in Insect Life (1891) definitely says that “ specimens proved on comparison to be identical with P. zazyphi Lucas.”” There appear to be no more recent records of P. zzyphi from date palms, and the above record dates from before the publication of the description of P. blanchardi, to which it possibly refers. Phoenicococcus marlatti, Ckll. Phoenicococcus marlattr, T.D.A. Cockerell, Proc. Acad. Nat. Sci. Philadelphia (1899), 1900, p. 262 (from imported Algerian date palms). Sphaerococcus drapert, R. Newstead, Quart. Jl. Inst. Comm. Res. in Tropics, 1906 (from Egypt). Prof. Newstead tells me in a letter that his S. draperi is certainly identical with Cockerell’s P. marlattz. Distribution —Egypt and Algeria, both the Tel or Mediterranean coastal plain and also the Saharan Oases. As an imported species, California and Arizona. Biology.—I failed to find this species in Mesopotamia. Very little seems to be known of its habits. Cockerell (1899) says that the female is 1-1} mm. long, and wine-red in colour; she secretes wax but makes no true scale ; the embryos grow to a large size in the body of the female. Marlatt found it, according to the same author, occurring packed in great numbers “in little cavities about 10 or 12 mm. long by 4 or 5 broad on the midribs of the leaves, communicating with the air by a narrow longitudinal slit,” on date palms imported into America from Algeria. The slight swellings containing the insects are very inconspicuous, and very likely to be overlooked even on a close examination. Later Cockerell himself found it on the sides of the sheathing bases of the leaves ; in this case it was not beneath the epidermis, and Forbes (1907) speaks of it as remaining close to the trunk among the great overlapping leaf-bases. Wilsie speaks of it as completely covering the stalk of the inflorescence near its base, where it is completely concealed by the leaf- bases. Probably Marlatt’s original description of the Phoenzcococcus living in cavities on the leaf-base was taken from specimens which had started growing on very young leaves ; as the leaf grew the tissues might have swollen round the colony of the scale-insect in such a manner as very nearly to cover it. Trabut’s most interesting note describes a date palm disease known to the natives of the oases of Zibans in the Sahara as ‘“‘ khamedj”’: this is caused by a heavy infestation of the stem of the inflorescence by P. marlatti. The disease becomes evident in the late spring soon a ter fertilisation ; the whole inflorescence dries up and no fruit develops. This calamity happens to 5 per cent. of the trees, nearly the whole of INSECT PESTS OF DATES IN MESOPOTAMIA. 299 which are to some extent attacked. The Arabs treat the condition by free excision of leaves down to the bole, and occasionally by removal of the terminal bud itself. Forbes (1913) records that this pest occasionally attacks exposed roots of date palms. The male of Phoenicococcus marlatti is unknown. No parasites or natural enemies of this scale have been recorded. Icerya aegyptiaca, Douglas. This scale-insect is recorded by Newstead (1917) from a young date palm at Zanzibar. So far as I am aware this is an isolated record. This species has a long list of host plants, and in Ceylon at any rate attacks various kinds of palms, some of which are badly infested. Selenaspidus (Aspidiotus) articulatus, Morg. The West Indian red scale has been recorded by Morstatt from ‘‘ Phoenix palm,” from German East Africa. Iam not perfectly certain whether this refers to the date palm, or to some wild member of the genus. In any case it is an isolated record of a Coccid which has been found on a variety of plants, including olive, rubber, citrus, coffee, bananas and cacao. Aspidiotus ostreaformis. Curtis. Aspidiotus hederae, Vallot. Chrysomphalus aonidum, L. Chrysomphalus aurantii, Maskell. Coccus hesperidum, L. These five scales are all general feeders ; they are all recorded by Essig as occasional minor pests of the date palm in California. Mr. HE. E. Green has been good enough to furnish me with the following list of CoccipaE which have been recorded from the genus Phoenix :-— Icerya aegyptiaca, and purchase. Phoenicococcus marlatt. Pseudococcus nvpae. Lecanvum hesperidum, tessellatum, oleae, and expansum. Diaspis phoenicis, and borsduvali. Chionasprs dilatata. Aspidiolus ostreaefornus, dictyospermt, ficus, rapax, lataniae, hederae, cyanophylli, and articulatus. Parlatorva blanchardi, proteus var. palmae, zizyphi, and pergandei. Frorima pellucida. Ischnaspis filiforms. ACARINA. Family TETRANYCHIDAE. Oligonychus (Paratetranychus) ? simplex, Banks. Mr. Hirst has examined some of my specimens of a small “red spider” from date palms and believes that they are the Paratetranychus simplex described by 300 F, A. BUXTON. Banks (Journ. Ent. Zool. Claremont, vi, 1914); it is apparently impossible to feel certain on this point. Mr. Hirst (1920) has re-described the species and figured the penis and the palp of the female. This mite is a most serious pest of growing dates in all parts of Mesopotamia which I visited. It was first noticed in early July, when immense numbers were found spinning webs all over the clusters of growing dates. These webs are of considerable extent and stretch from date to date over the whole cluster; dust becomes entangled in the web and the dates beneath never ripen properly. At the time when healthy dates are ripe the affected ones are discoloured and scaly, and as they are scarcely fit to eat and quite unsaleable they are generally used as food for cattle and horses. These diseased clusters are for some reason attractive to the wasp Polistes hebraeus, which may constantly be seen hovering over them and crawling on the surface of the webs. I was quite unable to find why the wasp acted im this manner, but I should imagine that its doing so must be a factor in the spread of the mites from tree to tree ; the wasp does not frequent the clusters of ripening dates unless they are attacked by the mite. The disease is found on many varieties of dates, especially the Khadhrawi, Dairi and Zehedi ; in 1918 1t was a most serious pest of the first of these, which is a widely grown date of superior quality and was not attacked by the Gelechud moth. I have seen clusters of dates on trees of all ages attacked by this mite, but it was always to be observed that its ravages did not cause serious damage except to trees far from water or growing in ill-kept unweeded gardens. It is not confined solely to such trees; mdeed I have seen it on a Khadhrawi which was growing on the very margin of a tidal creek at Basra. The disease is known to the Arabs as Toz, Trab, and Ajaj, all of which words mean solely dust ; it is stated to come from Allah, but more immediately from the dust of the roads. OTHER INsEcT PEsTS. ,Mr. H. H. King informs me that in some years the migratory locust (Schistocerca peregrina, Oliv.) eats all the leaves and fruit on date palms in the Sudan, and Popenoe mentions that locusts are occasionally troublesome in California and Arizona, and in Algeria. He gives no further information, but we may presume that the adults of some species devour the dates on the tree. He makes some even more indefinite remarks about ants and termites, and also mentions bees and wasps. I have alluded above to the fact that Polistes hebraeus frequents clusters of dates infested by the mite Oligonychus simplex, Banks. The Oriental hornet (Vespa orientalis) never, I believe, attacks fruit on the trees except that which has been already pecked by birds, but it devours the fruit exposed for sale in the bazaars, and is so large and abundant that it must cause appreciable loss in all parts of Mesopotamia in which dates are grown. ANIMAL AND BirpD PEstTs. Popenoe tells us that a horde of rats at Tempe, Arizona, once did great damage to the date crop, and mentions that gophers and jack rabbits eat the offsets; he states that at Baghdad a “ squirrel-like animal” eats the fruit. The Mammalia of Mesopotamia are now well-known, but I am not able to guess which animal he was referring to; probably it is the mongoose (Mungos persicus), which is common INSECT PESTS OF DATES IN MESOPOTAMIA, 301 ee in the date gardens and is known in Arabic as “ palm-tree rat’’; the mongoose is no doubt largely a vegetable feeder, but I do not think that it is in any way a serious pest of the date. In Mesopotamia birds undoubtedly eat a very large number of dates just before they are picked: Hypocolius ampelinus has been observed in Baghdad gorging itself and damaging many more dates than it could devour, and the bulbul (Pycnonotus leuco is mesopotamia) and house-sparrow (Passer domesticus) are also destructive. Mr. H. H. King informs me that the Spanish Sparrow (Passer hispaniolensis) pecks young dates in the Sudan. Milne adds parrots and monkeys to the list. Bibliography. Buckuier, W. (1882). Natural History of Ephestia passulella.—Ent. Mo. Mag., xix, pp. 104-106. Burkitt, L. H. (1913). The Coconut Beetles, Oryctes rhinoceros and Rhyn- chophorus ferrugineus.—Gardens Bull. Straits Settlements, i, pp. 176-188. Buxton, P. A. (1918). Report on the Failure of the Date Crop of Mesopotamia in 1918.—Bull. no. 6, Agric. Directorate Mesop. Exped. Force. (Provisional and inaccurate). CHITTENDEN, F. H. (1911). The Fig Moth.—U.8. Dept. Agric., Ent. Bull. no. 104. CocKERELL, T. D. A. (1899). Some Notes on Coccidae.—Proc. Acad. Nat. Sci., Philadelphia, 1899 (pub. 1900), pp. 259-275. CocKERELL, T. D. A. (1907). The Scale Insects of the Date Palm.—Bull. no, 56, Univ. Arizona Agric. Exp. Stn., 22 pp., 5 plates, 5 figs. Coox, A. J. (1914). Monthly Bull. Cal. State Commis. Hortic. Sacramento, i, pp. 440-441, 2 figs. CopeLAND, EH. B. (1914). The Coconut.—xiv and 212 pp., 23 pl. Macmillan & Co. Draper, W. (1907).* Notes on the injurious Scale Insects and Mealy Bugs of Kgypt.—28 pp., 16 pl. Cairo. Durrant, J. H. (1915). Myelois phoenicis, sp. nov., bred ee Dates in Algeria and in England.—Ent. Mo. Mag., li, pp. 303-304. Essie, E. O. (1915). Injurious and Beneficial Insects of California.—Suppt. to Mthly. Bull. Calif. State Comm. Hortic., 2nd edition. | Forses, R. H. (1907). The Extermination of Date Palm Scales.—Bull. no. 56, Univ. Arizona Agric. Exp. Stn. Forsrs, R. H. (1913). The Gasoline Torch Treatment of Date Palm Scales.— Jl. Econ. Ent., Concord, vi, pp. 415-416. Fracker, 8. B. Classification of Lepidopterous Larvae.—Illinois Biol. Mono- graphs, u, July 1915. GuosH, C. C. (1912). Life-history of Indian Insects. I1].—The Rhinoceros Beetle and the Red or Palm Weevil.—Mem. Dept. Agric. India, Ent. Ser. ii, no. 10. (713) H 302 P. A. BUXTON. GouaH, L. H. (1913).* Entomological Notes.—Agric. JL, Egypt, ui, pp. 103-106. Gouenu, L. H. (1917). Notes on an Ephestia, an Insect mjurious to Stored Dates in Khargeh Oasis.—Bull. Soc. Entom. d’ Egypte, pp. 133-140, 1 pl. Haceporn, M. (1913). Borkenkafer (Ipidae) welche tropische Nutzpflanzen beschadigen.—Der Tropenpflanzer, xvu, p. 43, etc. Hampson, Sir G. F. (1917). A Classification of the Pyralidae, Subfamily Gal- lerianae.—Novitates Zoologicae, xxiv, p. 36. Hirst, 8. Revision of the English Species of Red Spider (Genera Tetranychus and Oligonychus).—Proc. Zool. Soc. Lond., 1920, p. 49-60. Kine, H. H. (1908). Report on Economic Entomology.—Rept. Wellcome Res. Labs., Khartoum, iii, pp. 201-248. Lounssury, C. P. (1918). Union Dept. Agric., Division of Entomology, Annual Report, 1917-1918, jp. 87-107. Mackenna, J. (1918). Report on the Progress of Agriculture in India for 1916-1917. Maske, W. M. (1898). Further Coccid Notes with Description of New Species, and Discussion of Points of Interest.—Trans. and Proc. New Zeal. Inst., xxx. Minne, D. (1918). The Date Palm (Phoenix dactylifera) and its Cultivation in the Punjab.—Punjab Govt., Simla, 3rd edition. | Morstatt, H. (1913). Liste schadlicher Insekten.—Der Pflanzer, ix, pp. 288-296. NewstTEAD, R. (1906).* Quart. Jl. Inst. Commercial Res. in Tropics, April 1906. NewstTeEaD, R. (1917). Observations on Scale Insects (Coccidae) V.—Bull. Ent. Res., vii, pp. 125-134. Popenog, P. B. (1913). Date-growing in the Old World and the New.—316 pp., 40 pL, Publ. Altadena, California, West India Gardes. Romanorfr, N. M. (1901). Mémoires sur les Lépidoptéres, vii. Smytu, HE. G. (1911). Report on the Fig Moth in Smyrna.—U.8. Dept. Agric., Ent. Bull. no. 104. : STEBBING, EH. P. (1903).* Ind. Mus. Notes, v, pp. 126-127. TarGioni Tozzerti (1892). Aonidia blanchardi, nouvelle espéce de cochenille du Datier du Sahara.—Mém. Soc. Zool., France, v, pp. 69-82. Toumey, J. W. (1895).* Notes on Scale Insects in Arizona.—Ariz. Exp. Stn. Bull. .no. 14. Traput, L. (1912). Sur une Maladie du Datier, le Khamedj ou Pourriture du Regime.—C.R. Acad. des Sci., cliv, p. 303. Wixuoooks, F. C. (1914). The Date-stone Beetle-—Bull. Soc. Ent. d’Egypte, vi, 1918, pp. 37-39. Wituiams, C. B. (1918). Notes on some Trinidad Thrips of Economic Importance.—Bull. Dept. Agric. Trinidad and Tobago, xvii, pp. 143-146. Wiusiz, W. EH. (1913). The Date Palm Scales and their Control—Monthly Bull. State Comm. Hort. Sacramento, ii, pp. 538-539. (Unsigned editorial on Scales &c. in Sacramento).—Ibid., ti, p. 139-140. * Articles marked with an asterisk have not been examined by myself personally. INSECT PESTS OF DATES IN MESOPOTAMIA, 303 Postscript. As this paper goes to press | have had an opportunity of examining the MS. of a “ Note on Dates and Date Palms in Iraq” by Mr. V. H. W. Dowson, which will shortly be published by the Directorate of Agriculture, Mesopotamia. I have also seen the “ Administration Report of Agricultural Directorate for the Year 1919” (Baghdad Govt. Press, 1920), which contains the report of C. R. Wimshurst, Entomologist. From these sources I gather that a “‘ minute larva has been found in the inflorescence of a date palm immediately after its exit from the spathe ” ; this is presumably the newly hatched larva of the unknown Gelechiid. Spraying with cold water at intervals of a few days is reported to be a valuable means of controlling the red spider (Oligonychus). Silvanus surinamensis, L., and Laemo- phloeus sp. (Coleoptera, Cucujidae) are added to our list of pests of stored dates. (713) H2 Hh Figs Ah ne f u Nat t 4 soe F * ir gee 4 sw ah! bi ln Cee ee 4 a AS f F ; A , ; . j aon tbe Pal vs bri A edie * ve) 7 wi é : a { cu ; i Wa Nes Of Pa a! see Ad W n atP ee rf ha OE) A , whe, ‘ sy ob + fy oy 1 ae NN , bas Bile : ! x eee “, hee i " ‘ “ . ‘ ag: ‘ j _ ‘él fe ‘5 ye a k ‘ , Ci ! ; ; ; By 1s te : val See yA). | > i , ‘ i? oy ' wd ie ® in hy “! v way Ne: 4 a ak mith gt yee ¥ : m1. i } , - ae ‘ - bia ag py POSE F wy. # hg "Ad i Te ds aa ea " " Nj] ‘ ’ La; : > oo ¢ Ai i é hs ey ) i, 7 * hil . aM a , 4 Vigne’ Fe i x ‘ pp.4 ie | i Ly ie id PS i v te 4 i Ths . \ : f 4 Ee 4 a t i fea ; i in id ™, *) o) att ais) a nar %; ans ~* - j 3 : bee Li - j ; f 4 j Y fa ot 7“ f y } ; Bie mae oot Mae + my ay) te perch bile) as 4 \ ‘ Ng ft “ ay ‘ i Yule, bisey ‘ OB us Ap it as 7 ' : yh an ‘ sl snes Mbt iade. . LEP i { ma 4 ‘ 5 bd V : ae Te pp Seqs ; P - era. ya 4 ht ; sf ee a ; aS Sela en ) Ah ONESy Sitys Rh 3 ; ‘ | bas Wize y A ew | Ln i in + meh " 4. 4 ie Tee y - P 5 ur 7 ; s . y wee \ 1 i he ad a j : Wie blll | ; a" i, ie \ : ' af aw AUN ai { { “ : ie es ail 7 I ‘ ' : 4 . Sf t i “vr © vi biaeet , ' des i 5 iu iG i an a hy vy Poa : y ‘Dy an i gt ‘ ab ty » * ’ . iV * wi ‘ vie F iy! . Je +! 4 5 ‘ a. ’ f : j j ‘ i + oh A | ari ' i am my " , F " ! at a 4 ia ’ ‘ 4 ! 7 " , 4 Ve ry y y x 1 ' ae | af ’ i 4 u . ; i | i fA j : i, v ‘ ¥ i a 4 : ; y a Bi yi ) ; b hie ae \ ; | ap " i t ; ; f Led f i 1 \ , i ‘ ( 4 pis ' A ' z ee mane hn , ' 5 é J y ol , y hibe re q i ‘ 4) Vy ‘1 At i) / 4) iT! : » “' ; .* ' eae ’ e% , i] j nN ey ( ' , r a ay a MY H } ., i L5H ¥ ‘ 3) ° a = i ee 1! bah i rs ‘ f \ 1 re s ' 4 ‘ ty Pt é - my fea Hel d AB) iy ued a * a4 i } , 1 ‘ f ' i % be.’ P \ on! ke i i y } ule J 1 i. ie we wk te are wy - ‘ ¥ g * 7) € ; a 7 ki ’ j - 2 rn q : = vs ab . } . 7) - 7 4 -“), i i : L : : - - i ' | * mo ri j ~ ’ * " % f al ’ \ rd - _ - 4 & a) : ® aay * ” ra 7 ; Pd ia ; ’ ad © } “hs > Mey. #3 WET ‘ j ° 1 ia i ‘A x *) wae Ay 4 oa . ‘ ‘ »¥ fa ‘ : A, | ’ 4 i ee i bee ' nt ot FS ees a wd : &, : oe | { ; ‘ , y or Thy 22 | itn a. + J nes 4 ‘ ; ‘ + = * a he es ee, a Sean Y ¥ ' ; Pe, be ott fh +t pair te “N's is bese 5 aC at he ; ( it ti . , + ae \ ‘ E . ad ae piss aha 2’ WA Ete ahi. ast? ew dn Gee ~ ri > ny Sy e Gee bad cam on eae ‘ ise oa ds ar OMrate % ‘, * yep f ‘ pe ae uA i , Pat ¢> Lae , st Day ry ¢ elon 2g “adie AS viata a eB ee = 1 ae a z3 . i 4 y ’ x ai woe ’ a Pi =,” ih 7 > : we ’ 4 . : he 2 b2ten Oye ai hor te! etal sagioee a ; te whl a? 4 \ ipa : ' os ie bale 7s Pe | * > 4 he< ie c : rey rey , aT Pia mS Pee ot $ 2, " z } a Fs 42) gf 7 4 1 a oe | £ : bad ‘ 4: at : p ; i , 4 - ~ | <4 , mg vy, P } + f é "7 4 ) s* Pits! As parte at 7 ee PS ee ee . fv 4 % me) f \ ij % , J P ry Bot ie ae % ‘ t hig 5 - ae tal i yy “al if ; “Ti _ = F L, 4 Pekin 2 . 2 ft Sts ea) hedin Foetal daca abe Be ae hem > aaa -< We , : s Vw yy a ae t ' \ . ‘s . . » ov Mt ah, }- f ‘ Aon fa ral ty A aie a ae £5 | wo Prat hare | [ 4 7 es ‘ { ‘ i P ‘ t 1 + wide a * Z “<< - + é wi : ee Mae eae Alte ‘ ’ i 7 i . 7 . rey < y 4 : e igre j ” .. a " fs _ \] ~ ' ‘ } : : os 4 1} > * - Vs w ¢ a ~~ | ~ er: ‘ . r ! « 4 EXPLANATION OF PLATE VIII. A bunch of dates photographed in early August when growth of the fruit had practically ceased and ripening was just commencing : a, pedicel from which all the dates have fallen owing to the attacks of the larva of a moth (Fam. Gelechiidae) ; as will be seen, the perianth falls as well as the unripe date. b, dates destroyed by the larva, but which have not yet fallen; notice their small size and spherical shape ; they are quite hard and dry ; such dates are known in Arabic as “ hashaf.”’ c, healthy dates. BuLi. ENT. RESEARCH. VoL. XI. PART 3. PLaTE VIII. 305 LON THE GENUS PHLEBOTOMUS.—PART IV. By R. Newsteap, F.R.S., The School of Tropical Medicine, Liverpool. = The following notes may be regarded chiefly as materials towards the study of the geographical distribution of this group of the bloodsucking midges, which are of so much interest from a medical standpoint. For the opportunity of examining the very large series of specimens (570) herein recorded, | have to thank Dr. Guy A. K. Marshall, Director of the Imperial Bureau of Entomology, who sent the collections made by Drs. A. Ingram and J. W. Scott Macfie, in the Gold Coast ; and also that formed by Capt. J. Waterston, R.A.M.C., in Macedonia. I tender my thanks also to Capt. P. A. Buxton, R.A.M.C., for the collections made by him at Amara, Mesopotam a; to Capt. H. W. Leatham, R.A.M.C., for examples from Basra; and to Major J. A. Sinton, V.C., I.M.S., for the specimens which he collected in N.E. Persia, under exceptionally difficult circumstances, and also for the long series taken by him in the N.W. Frontier Province of India. I may add to this short communication that the determinations of the specimens herein recorded are based upon material which was carefully prepared and stained for microscopical examination, as by this means the minute taxonomic characters can be seen and studied with greater accuracy and ease. Phiebotomus antennatus, Newstead. Male (hitherto unknown). Pale form. Drab or straw-coloured ; fringe of costa, anteriorly, rather strongly infuscated, that of the posterior margin not so, Hairs NS Fig. 1. Phlebotomus antennatus, Newst., g; a, proximal segments of the antenna; a1, terminalsegments. P. minutus africanus, Newst., $: b, proxi- mal segments of antenna. P. simillimus, Newst., J: ¢. proximal segments of antenna. The figures a, b, and ¢c are drawn to the same scale; ai, ata much higher magnification. on wings and scales on legs very faintly iridescent in a strong light. Dark form. Wing very heavily infuscated, especially the costal fringe on both margins ; the hairs beautifully iridescent, and in certain lights strongly resembling the rich deep blue of lapis-lazuli. The scales of the legs are also iridescent in certain lights, 306 R. NEWSTEAD. the predominant blue colour being almost as intense as that of the wings. Hairs on the proximal segment of the abdomen erect ; those on the remaining segments recumbent, Antennae with the segments (fig. 1, a) only very slightly longer than those of the female ; the geniculated spines unilateral* and unpaired. Wing rather narrowly lanceolate. External armature precisely similar to that of P. minutus, Rond. Length, 2-2°4 mm.; wing 1'°2 mm.; leg 11, 2 mm. Gotp Coast: Accra, in laboratory. vi. 19, 38 §g,59 292 (Dr. J. W. Scott Macfie) ; Kwaman, 10.vii.18, 12; Navarro, on walls of M.O’s. quarters, 22.vi.18, 1 9; Yamalaga, 18.v.18, 1 2; Salaga, on walls of M.O’s. quarters, 10.v.18, 1 9; Yeji, 6.v.18, 1 2 (Dr. A. Ingram). Phiebotomus ingrami, Newstead. Gotp Coast: Chechewere, 9.vii.18, 1 g (Dr. A. Ingram). Phlebotomus minutus, Rondani. Maceponta : near Salonica, 1918, 1g, 19 (Capt. J. Waterston). MESOPOTAMIA : Amara, River Tigris, 1.1x.18, 1 3, 20 9° (at light); 81x.18,2 gg, 5 2 (at light) 3" 19.1%-18, 1°95" 30.1%.18, 1 gy 9 YO lat heht): Tis eo: ea 1 ¢ (at light) (Capt. P. A. Buaton). Inpt1a: N.W. Frontier Province, Dera Ismail Khan, vii. 1919, 33 gg, 34 92 (Major J. A. Sinton, V.C.). Phlebotomus minutus africanus, Newst. Gotp Coast: Yeji, 6.v.18, 1 9; Malowe, 29.vi.18, 1 g, 299; Accra, 4.1.19, 2 ¢3, 1 9; Lilxia, 4.vi.18, 3 gg, 3 29; Diari, on Rest House wall, 26.v.18, 1 2; Gambaga, in latrine, 1.vi.18, 7 gg, 1 9; 5.vi.18, 1 2 inside mosquito net ; Salaga, on walls of M.O’s. quarters, 10.v.18, 6 gg, 3 Q2; Jugbe, 1.viu. 18, 1 2; Sekodumase, 8.vii.18, 2 gg, 11 292; Til, on walls of Rest House, 14.vi.18, 2 §g, 12; Nangudi, on walls of Rest House, 15.vi.18, 3 dg, 7 99; Kul- masa, 22.vii.18, 1 9; Nandawli, 15.vu.18, 2 gg, 429; Bole, 26.vi.18, 1 ¢g; Chechewere, 9.vii.18, 1 g, 3 29; Bawku, 12.vi.18, mside mosquito net, 3 99; Savelugu, Rest House, 25.v.18, 2 3g, 2 99; Wa, Rest House, 20.vu.18, 1 9; Navarro, on walls of M.O’s. quarters, 22.vi.18, 2 3g, 4 99; Bawku, on walls of quarters formerly belonging to M.O., 10.vi.18, 5 gg, 1 2; Banda Nkwanta, 30.vu.18, 2 gg, 6 29; Maibindiga, on walls of Rest House, 3l.v.18, 2 99; Palbe, 16.v.18, 1 9; Kwaman, 10.vili.18, 1 9; Zovaragu, on walls of D.C’s. quarters, 16.vu.18, 1 g, 2 99; Tishi, on walls of Rest House, 1.vi.18, 1 g, 1 9; Kpalgu, on walls of Rest House, 30.v.18, 1 9; Yamalaga, 18.v.18, 1 2 (Dr. A. Ingram); Accra, in laboratory, vi.19, 3 3g, 4 992 (Dr. J. W. Scott Macfie). Phiebotomus papatasii (Scopoli). MrsopoTaMIA: 30.vi.18, 1 3, 3 99; Amara, River Tigris, 14.v1.18, 1 9; vili,18, 1 6, 5 O93 2.1%:18,3 99 (biting man); 211x18, 1. og (Cap F. A. Buxton). Macrponta: near Salonica, 1918, 14 gg, 35 29 (Capt. J. Watersion). Inp1a : N.W. Frontier Province, Dera Ismail Khan, vii. 1919, 3 ¢¢, 3 22 (Major J. A. Sinton, V.C.). Asta Minor: Trebizond, 5 gg, 3 99 (Dr. B. Uvarov). * This rather remarkable character is common also to the males of the following species: P. minutus, Rond., P. minutus africanus, Newst., and P. simillimus, Newst. (see fig. 1 b, ¢). ON THE GENUS PHLEBOTOMUS.—PART IV. 307 Phlebotomus perniciosus, Newstead. MaceponliA: near Salonica, 1918, 26 3g, 9 2° (Capt. J. Waterston), CaucasiA: Borzham, Province of Tiflis, 16.vii.1919, 8 gg, 3 22 (Dr. Kandelaki, per Dr. B. Uvarov). Phiebotomus sergenti, Parrot. Phlebotomus sergenti, Parrot, Bull. Soc. Path. Exot. x, pp. 564-567, figs. 1-11 (1917). Phlebotomus caucasicus, Marzinovsky, Medical Review, Moscow, lxxxvii, nos, 13-14-15-16, pp. 612-614 (1917). A large species somewhat resembling P. papatasi (Scop.) in colour and in the general arrangement of the body hairs. But the male of P. sergenti is abundantly distinct and may be recognised from all other species by the unique character of the genitalia, in which the great length of the sub-proximal spine on the superior claspers should serve at once to distinguish it from all the species possessing paired brushes (peniculz) of non-deciduous hairs between the superior claspers proximally. On the other hand, it is extremely difficult to give taxonomic characters for the separation of the female from that of P. papatasw. Ina typical example of P. sergentt, generously presented to me by M. Parrot, I fail to find any structural difference by which it could be separated from the 9 of P. papatasw, though it is decidedly smaller ; and this applies also to the female of the variety herein described from Mesopotamia. In the females from Persia, however, the geniculated spines of the antennae are slightly longer than those on the corresponding segments in the African examples before me ; and in one example the spines are also accompanied by a somewhat lunular-shaped, subcutaneous structure (? gland), which no staining has hitherto revealed in any of my preparations. On the other hand, the males from the two widely separated regions agree in all their morphological details and are clearly conspecific. There is, I think, no doubt also that the African and Caucasian species are one and the same. Marzinovsky (l.c.) notes the close affinity of his P. caucasicus with P. sergentz, Parrot, but claims that it is distinguished from the latter in some details. These ‘* details ’ are, however, not traceable either in his diagnosis or in his figure of the male genital armature. Thanks to Col. C. M. Wenyon I have been able to study a direct photographic print of the male genital armature taken from the original negative which was used by Marzinovsky in illustration of his paper, and this con- firms my views as to the specific identity of the Caucasian with the African form. The following description is based upon material collected in N.E. Persia by Major J. A. Sinton, V.C. :— Male.—Palpi of five segments; 2nd and 4th about equal in length; the 3rd a little longer than the 2nd, laterally incrassate, the incrassation seen only in dorso- ventral view ; sensorium well marked by the relatively large clubbed hairs ; 5th a little more than twice the length of the 4th. Antennae with the distal end of the 3rd segment almost reaching to the tip of the proboscis ; the tips of the paired geniculated spines on segments 3-15, inclusive, reaching a point very slightly in advance of the centre of all the segments with the exception of the 3rd. Wing 308 R. NEWSTEAD. (fig. 2, a) somewhat lanceolate, the tip rather acutely pointed. Genital armature (fig. 2, 6) with a pair of relatively large brushes or peniculi between the superior claspers proximally. Proximal segment of superior claspers very robust; distal segment about half the length of the first, strongly bifurcated distally, the tip of each Jmm Vega iar eee Oa 5 ates ‘ The local ‘‘ formations” are as follows :— (1). Very highly deciduous types, such as Ptlerocarpus sericeus (mumbhungu), Pterocarpus angolensis (bloodwood or mubvangazi)—the earliest shedder and latest reclother of all (Pl. x, fig. 1)—Acacia natalitia (ihlofunga), A. caffra (iguwha), Bridelia micrantha (mushungunu) and others. Considerable stretches of country are covered almost exclusively with bush of these species—open or dense, pure or mixed. Fly spreads in during the rains, but I failed completely to find tsetse in wooding purely of this type during the leafless period even in such shade as there was, and as it is quite unlikely that they can exist there then, it follows that such areas are cleared each year for from one or two to many months at a time. The country carrying these more sensitive types of wooding is characterised by good soil and permanent streams, and much of it by high elevation. It is particularly suited to settlement. (2). Lowland bush savannah (Pl. x, fig. 2), comprising many tree species, but with the mupoza (Combretum near leonense) dominant on the drier parts and the musekesa (Bauhinia reticulata) on those that are wetter in the rains. ‘The bush is mostly sparse and, the fires being early and the trees and shrubs low, the shade is destroyed at an early date in a normal year. Rare scattered fly, G. morsitans and pallidipes, was to be found in this type of country up to the time of burning, that is in 1918 the end of July, but permanent fly was near by in a more suitable type of wooding—Brachystegia, with vleis. I found no trace of the existence of permanent breeding centres in this type of country—the basalt. (3). Brachystegia wooding, here known as “tondo-bush” or “itondo” and evidently identical with what Jack calls “ gusu.” This again contains a variety of trees and shrubs, but different species of Brachystegia dominate, such as B. globiflora, the mutondo (which gives its native name to the type of wooding), and the mutsatsas, B. randii, B. bragaet and B. spicaeformis, the trees of the red spring foliage. Uapaca also—U. kirkiana (munjanje) especially in the highlands, U. sansibarica (mutongoro) below—is common and in many places forms pure wooding. Diplorhynchus mossambicensis (mutowa), with its weeping THE TSETSE PROBLEM IN NORTH MOSSURISE. 321 foliage, creviced stem and branches and furrowed, protectively coloured bark, is a particularly useful tree to the tsetse and is abundant in Brachystegia bush at low elevations. It is the foreground tree in PI. xvi, fig. 1. Tondo-bush is the tsetse bush par excellence. Brachystegia-Uapaca wooding in our present connection is divisible into three varieties, two of which carry each its own tsetse. It is divisible into (a) high “‘itondo,” characterized by the greater size of the trees and generally denser wooding and forming fine “ savannah ”’ forest (Pl. xvi, fig. 1); (6) lower, more open “itondo,” still mostly savannah forest, but poorer and intermixed with more definite savannah (PI. xv, fig. 1). It is (with other conditions correct) the special haunt of G. morsitans. The former, finer variety is divisible further into (1) clean itondo, devoid of woody undergrowth and of the large fly (Pl. xi, fig. 1), and (2) itondo with woody undergrowth, inhabited by G. brevipalpis (Pl. xu, fig. 2). This undergrowth, usually patchy, consists for the most part of saplings of the trees themselves, often in close-packed clumps (Pl. xu, fig. 2; Pl. xvi, fig, 1), and of large scattered shrubs such as the umtunduluku (Ximenia americana) and the chigwenderes (Bauhinia galpina and B. petersiana); but in parts of the high itondo may be found numerous clumps of “ravine” type primary forest undergrowth, extensive or small, of bird and baboon-carried genera that have made a start and maintained themselves where the ill-regulated fires have enabled them to do so. PI. xiv, fig. 1, gives some idea of their nature. Two highly important points about itondo in relation to tsetses are these :— (a) The grass-growth in general below high tondo-bush is sparser than in some of our other secondary sub-types and does not burn really effectively till late in the season ; (6) itondo as a whole, under our local conditions of rainfall, tends to retain its leaves through the winter better than any other of our secondary formations. The mutsatsas (Brachystegia bragaei, etc.) have often regained their leaves before the associated mutondo trees (B. globiflora) have lost theirs, and the Uapacas only lose theirs (and that not completely) when nearly all the other species are busy regaining their leaves. This tendency to keep in leaf without some s!ight interregnum is less marked at high elevations, in particularly cold and dry winters, and in the drier parts of the lowland areas; but on the Buzi, and in the “ Oblong” (see Map) generally, it was sufficient in 1918 to carry G. brevipalms in numbers everywhere through the winter, and it is probable that in most if not all years it would do so in most parts of this piece of wooding. (4). Dense secondary forest includes especially (a) the very fine type of wooding dominated» by musara and musunganyemba (Milletia stuhlmanni and Pteleopsis myrtifolia)—definite “‘ monsoon forest’ ; and (6) a more mixed type. In places, owing often to the coppicing effect of unthorough native cultivation with the good seed-bed offered by old gardens to seeds from the surrounding bush, and to the relative exclusion of fire consequent on the absence of grass, the ordinary trees of the savannah form more or less dense thickets (Pl. xvii, fig. 1). Eventually trees and shrubs that are normally found on the outskirts of primary forest introduce them- selves and finally dominate. Such are the munjerenshe (Albizzia fastigiata chirin- densis), the large umkadhlo (Rauwwolfia inebrians) and the fine shrubs Vernonia 322 C. F. M. SWYNNERTON. podocoma (mudambasese) and Brachylaena rhodesiana (ipahla). The annually burning grasses tend to disappear, the live carpet may become sparse or it may become composed in part of the non-burning grasses that occur in the primary forest of the same locality. Clumps and stretches of the handsome ribbon-like izeya grass (Setaria mauritiana), beloved of bufialos, also of Amomum (good shade for brevipalpis) are exceedingly common, and thickets occur of such tall, more or less herbaceous plants as Isoglossa mossambicensis and the unpleasantly smelling Mellera lobulata. Woody undergrowth is usually more or less scarce excepting when the wooding is under invasion by primary elements. In the simple coppice stages (Pl. xvu, fig. 1) these woods are highly attractive to G. pallidipes ; in the older stages where the overwood is tall and heavy, as on the Sitatonga base, and some underwood is present, they shelter G. brevipalpis. In the Mtshanezi-Puizisi area especially a large proportion of the groves of this type are of relatively low growth and are characterized by the presence of the very beautiful climbing fern, Lygodiwm subalatum, which carpets the ground and curtains the trunks up to a considerable height, affording low shelter and additional shade where the shrub growth is poor. This variety of the type is perhaps more commonly frequented by pallidipes than by brevipalpis. “ Dense mixed secondary ”’ occurs here chiefly on the dolerite and, like Mulletva bush, tends readily to be replaced by primary elements (which are more suitable to brevipalpis) where it is in contact with them. Thus Albizzia chirindensis and Piptadenia buchanani (umfomoti) are at present sharing the dominance of the Mafusi rubber forest, and the shrub growth shaded by them is already largely primary. The musara (Muilletia) is rather specially deciduous, and I have found relatively little dry season fly in this variety of dense secondary excepting where it was under incipient primary invasion. (5). Bauhima and Erythroxylon-Landolphia thickets. Two species of chigwendere (Bauhinia galpiom and B. petersiana), large and densely foliaged shrubs that lose their butterfly-shaped leaves more or less Jate in the season, occur from the British border to the Sitatonga Hills. B. petersvana continues to be present east of the hills. They are found in every type of secondary wooding, especially on ant-heaps, but in places they form almost pure coppice of their own. When well-grown and in leaf, their coppice is more attractive to G. pallidipes. Where, however, the Bauhinias occur isolated under higher shade, G. brevipalpis is attracted and rests under their coils—for under these conditions they tend to assume the habit of lianas. | Low dense thickets of Hrythroxylon emarginatum, Pleuroridgea zanguebarica and other erect shrubs, with a few rubber vines, have invaded (for the most part) Pterocarpus wooding at high elevations on Mount Umtareni. I found no fly there, for reasons probably of elevation and also of situation ; for between this and the permanent fly-bush (Brachystegia) lower down the slopes lay a belt of highly deciduous trees (formation 1, above). Lowland rubber forest (Pl. xi) is not essentially different. It also contains much of the Erythroxylon, but the invasion of Landolphia and other lianas has progressed further than on Umtareni and here fills the thickets. The trees of the highly deciduous and other (Brachystegia) types that have been invaded still in many places stand out above the new shrub growth, just as on Umtareni. I refer particularly to the Madanda forests, which I did not THE TSETSE PROBLEM IN NORTH MOSSURISE. 323 visit on this trip, but which (from what I saw of them in 1906-7) I should quite suspect of being capable of harbouring G. pallidipes and, in many places, G. brevipalprs. TV.—F actors INFLUENCING LEAF-FALL. I have described above the relative deciduousness of our various types of bush. We have seen that primary forest (Pl. ix), and bush which it is seriously invading, is never leafless enough here to discourage its special fly—brevipalpis; and that of the secondary types, Brachystegia wooding (PI. xvi, fig. 1) and particular varieties of the “dense secondary’ formation tend to retain in this district sufficient leaf to carry fly through the year. However, within any particular type of wooding actual leaf-fall is hastened locally, and the gaining of the leaf correspondingly deferred, by several factors. It is true that some of the differences are apparently individual, though unseen root disabilities might account for them. It is at.any rate of frequent occurrence to see two trees of the same species and size (e.g. of Pterocarpus) standing together, one of them leafless the other not. Ficus subcalcarata (the mutowetowe fig, an isolated large specimen of which will sometimes harbour fly) is an extraordinarily erratic leaf-shedder. But general rules are apparent also. (1). Season affects leaf-fall. After the abnormally heavy rains of 1917-1918 (2,514-5 mm. at Spungabera) leaf-fall generally took place, I should say, quite three weeks later than usual, where frost was absent. A low rainfall (such as that of 1913), or early and severe cold, hastens the phenomenon. It must be remembered further that rainfall, hence leaf-fall, varies with locality, particularly in relation to the mountain ranges. The rainfall appears to be distinctly lower east of the Sitatongas than on them or even west of them, and is lower here again than on the hills of Spungabera, the “ Jihu”’ (Gwenzi’s) and the British border. (2). Proximity of moisture in the soil affects leaf-fall in two opposite ways. On damp low-lying ground liable to frost the leaf-fall is hastened, and even where there is no actual frost, trees bordering on ill-drained, cold, swampy ground tend annually to lose their leaves earlier then those of the same species that stand back on well-drained soil. But trees standing on warm, damp soil lose their leaves later than the latter. These differences are to be clearly seen in fly-harbouring Brachystegia woods, but the particular illustration I shall quote is concerned with some scattered wooding of Pterocarpus angolensis (mubvangazi—PI. x, fig. 1) that is under my continuous observation. Here not only is there, in May, an increased leaflessness as one approaches the cold, damp lower valleys, but of the trees within the latter, those with their roots raised on ant-heaps are practically in full leaf when those on the general ground-level are completely leafless. (3). Differences in the amount of humus present has appeared to account for a difference in leaf-fall between woods of the same species—e.g. of Pterocarpus sericeus (mumbhungu) on the dolerite—and it is to be noted that ant-heaps exert an influence in this direction also. It is probably again mainly a matter of relative moisture retention. 7 (4). Geological formation affects leaf-fall once more on the same lines. I have noticed that Pterocarpus sericeus loses its leaves a good deal later on the fertile 324 C. F. M. SWYNNERTON. (and moisture retaining) upland dolerite than on the sandstone-shale formation at the same elevation. The Brachysteqia-Burkea groves of the far drier, deeply cracking, compact basalt soils of the plains, effectively drained by the low-flowing Buzi and highly insolated, were losing their leaves earlier than those of the granite, especially on stony outcrops ; while similar wooding on a quartzite strip on the Buzi, at the same elevation and distance from the river and only half a mile away, was far later than the latter, as was the Brachysteqia wooding (doubtless with better rainfall, however) of the sedimentary area of the higher Buzi. (5). Altitude, other things being equal, makes a difference. Travelling between Maruma and the escarpment at Makwiana’s kraal on 11th August I saw no Albizzia (munjerenshe) or other secondary wooding that was not completely leafless. On approaching the foot of the descent I began to find some in leaf—and at once found fly, both brevipalpis and pallidipes. I have come across many other examples both in our individual tree-species and in our woodland types that were thus remi- niscent of the instance of the plane tree—simultaneously leafless at Innsbruck, yellow-leafed at Garda, dark green at Palermo and in parts of Greece evergreen ; and I have little doubt that latitude must also count for much in this connection. Altogether it is certain that no unqualified statement disregarding rainfall, soil- moisture, frosts, elevation, latitude and exceptional seasons can be made for parti- cular types of wooding as regards the period for which they will be without leaves. and inhospitable to tsetse. It is, I think, equally certain that an exhaustive study of these factors may be of great assistance not merely in the explanation of local outbreaks, but in the actual control of the fly. Certainly no investigator can at present lay down the law for another area than his own as to the behaviour of particular types of wooding and, therefore, for the seasonal behaviour there of the fly. A point that cannot be too greatly emphasised is that, if we are to fight the fly economically as well as effectively, we must base our measures not on the normal but on the exceptional season. I do not refer to the season that is so exceptional as to occur only once in 50 years—as the great droughts of 1860 (approximately) and 1913; nor to cases in which haste is vital ; but, let us say, to the driest or coldest season in ten. The effect of the factors that I have mentioned above may be seen even when leaf-fall has been brought about artificially, as by a fire. V.—EFFECT OF THE ANNUAL GRASS-FIRES. These very commonly anticipate leaf-fall. Except in the loftier bush, and to some extent there, they scorch and wither even the higher leaves and defoliate the saplings and shrubs, destroying the shade. The very factors, local and seasonal, that accelerate or retard the date of the normal leaf-fall hasten or postpone that of the fires also. In 1918—except on the uplands where the frosts were considerable and dried the grass—the earlier fires were a month or more later than usual ; in 1913 they were exceptionally early (after a deficient rainfall) and provided some excellent instances of what might be accomplished by severer annual fires towards the destruction of tsetse-haunted bush. The Buzi flats with their cracking basalt THE TSETSE PROBLEM IN NORTH MOSSURISE. 325 soil (Pl. x, fig. 2) are stated always to burn earlier than the granite gneiss near them (PI. xiv, fig. 2; Pl. xv, fig. 1) with its impervious underlying bed, and the sandstone-shale areas of the uplands west of Spungabera definitely do burn considerably earlier than the dolerite beside them. The type of bush also influences the date of burning. High canopy-forming Brachystegia, such as occurs in Mossurise on the upper Buzi, burns effectively much later than the same bush where it is lower and more broken, and some types (as high dense secondary, including Muilletia) burn, if at all, late yet tamely. It is stated that under the Zulus burning was the subject of regulation for hunting purposes, a late, thorough burn being aimed at and usually achieved. Under the white man everyone burns when he pleases. It thus comes about that the first fires take place when the grass is but half ready to burn and there is relatively little recent leaf-litter on the ground. They attract but little wind to swirl the flames into the tree-tops and the thickets, and they are very small in extent, so that the flies driven from them find ample refuge all round. When later fires take place— at various dates—the areas burned earlier are already becoming fit for the reception of the fly and the latter suffers no inconvenience. A still worse effect of unregulated burning, from the point of view of tsetse control, is the encouragement given by too early burning to the wooding and particularly to the formation of the thickets en which G. brevipalpis rather specially depends (Pl. xvu, fig. 2). Late burning on the other hand finds the grass and fallen leaves at their driest and most abundant, and is so extensive as to draw in great wind. It achieves much actual destruction of young growth, and a temporary destruction of much high shade as well, at a hot, dry time that must be relatively critical for the tsetse. A failure to burn, again, may be made the means of destroying thicket growth if the next year’s burning is late, owing to the additional amount of inflammable material present. The fire no longer stops at the outskirts. This result actually followed the temporary effort to keep fires from the rubber forests, and I was shown where pieces of these forests (in Umtobi’s country—of the type shown in Pl. xii, fig. 2) were completely destroyed in consequence. VI.—DistrisuTion oF Rocks, WoopLaANpD Types, Foop ANIMALS AND FLy. Geological Formations. Geologically, the country investigated is divisible into four clear-cut blocks, two of them east of the Sitatongas, two west. The two to their east are a granite- gneiss and a basalt formation. The two to their west are (a) a sedimentary formation of shale, sandstone and quartzite ; (b) dolerite, breaking through and capping the sedimentary formation, which in the dolerite area is visible only in places. The granite-qneiss country consists of lowland plains, flat to undulating.* It is bounded on the west by the Sitatongas, on the north (I believe) by the Mabiti hills across the Lusitu and by their north-easterly continuation, on the south by a fairly well demarcated line a few miles south of the Umvuazi and apparently *My rocks, identified by Mr. Mennell as mica granite, gneissose granite and Muscovite anite or gneiss, Show more or Jess foliation, and the area appears to be that, east of the Sitatongas, described by Theile and Wilson as metamorphic. The general character- istics are those of the granite. 326 C. F. M. SWYNNERTON. following the same general direction as that river. The surface soil is a coarse, loose, whitish sand, and numerous vleis, still wet in the dry season, indicate the presence below of the usual impermeable layer. The basalt plains (PI. x, fig. 2) adjoin the granite-gneiss on its south. They cross the Buzi and cover a great extent of country to the south of it. The ground, reddish where the rock is exposed, browner or blacker in the interleaved alluvial patches, dries rapidly after the rains and shrinks and gapes. In the country actually investigated by me it is possible that the Buzi, flowing between fairly high banks, determines the height of the water-table. The sedimentary area—or rather the area where this stratum is continuously ae for it underlies both the basalt and the dolerite—fills the great oblong tract enclosed between the Mtshanedzi and Buzi rivers (see Map). A broad strip of the same formation runs thence up the Buzi to the British border. It is hilly throughout, the soils are mostly compact, and the rock frequently outcrops. Except in the strip just referred to, it is not well supplied with permanent streams. The sub-ophitic dolerite, as dyke or cap, dominates largely between the Mtshanedzi and the Lusitu, and especially at and south of Spungabera in Gwenzi’s country. It is accompanied by a red, rather clayey, yet well drained and highly fertile, typical “trap” soil, and abounds in excellent permanent streams. The Sitatongas themselves are of quartzite, flanked by schists, trap, ete. ; apparently (Thiele and Wilson) the schists dominate to the north of them, in the area traversed by my native expedition. Distribution of Woodland Types. The granite-gneiss carries open (mixed) Brachystegia bush usually of a rather poor type (Pl. xiv, fig. 2). Its vleis are nearly bare of wooding, except for the fine trees on the ant-heaps that stud both vleis and dry land. The finer vleis in particular are lined by a short green sedge (Fuirena). The male tsetses are found in rather special connection with this. The basalt, as might be expected from the above description, tends to carry a very poor type of bushy (Pl. x, fig. 2), mixed, but commonly more or less stunted and very open woods—the “‘ lowland bush savannah”’ of my classification (p. 320). Thorn groves occur in places; elsewhere, on exposed or nearly exposed sedimentary rock, Brachystegia patches. The grass, short in some places, is long in others. Fires are earlier here than anywhere else. The sedimentary rock of the “Oblong” and its northward extension carries itondo (Brachystegia-Uapaca forest) of a particularly fine type, with a good deal of sapling undergrowth and, in places, patches of invading primary shrub formation (v. p. 320 and the figures there referred to). The mutongoro (Uapaca sansibarica) often forms close groves beside vleis. The grass growth is seldom dense or very high, except in open places (Pl. xiii, fig. 1), but great and annoying burr areas, of four semi-herbaceous species of Triumfetta (dzunzu), occur. They dry up and are leafless in the winter. The dolerite carries (a) wooding of the dense mixed secondary types, largely composed of highly deciduous species, and including Bauhinia thickets; (b) open THE TSETSE PROBLEM IN NORTH MOSSURISE. 327 Pterocarpus sericeus (mumbhungu) wooding, highly deciduous. It is also, however, the chief seat of the surviving primary forest (Pl. ix; Pl, xin, fig. 2). A high-forest area composed of primary forest (largely ravine type), tall, dense, mixed secondary, and invasions of the latter by primary forest, covers much of the broad strip of country that comprises the Mafusi and Maronga rubber forests, and stretches interruptedly from somewhat near the British border at beacons 76-77 eastwards to the Musapa, crossing the Lusitu on the way. It is roughly bounded on the south by the Puizisi, but sends two broken, narrow extensions southwards along ‘the eastern and western base respectively of the Sitatongas. Between the Puizisi and the Mtshanedzi, and again in the Gwenzi country, lighter dense secondary types dominate, copse-like groves (Pl. xvii, fig.1 ), and fern-entwined wooding alternating with open, highly deciduous wooding of Pterocarpus sericeus and its Acacia and other associates. This more open wooding dominates rather specially at the higher elevations, though scattered close Albizzia groves, losing their leaves in the cold weather, are also somewhat prominent there. Narrow primary-type thickets —fringing forests (Pl. xiv, fig. 1)—often occur along the small streams, and harbour fly permanently at the lower elevations. The grass-jungle growth, except in the dense wooding, is immense and tangled. The fires are normally more or less late and, when postponed further till the grass is dust-dry, they are of extraordinary severity. Distribution of the Game. This again is most simply stated in terms of rock-formation. The lower-lying Mtobi-Gogoyo dolerite-dyke country that is bounded by the Mtshanedzi on the south, the lower Chikambwe on the west, the Mtobi piece of the Lusitu on the north, and the Sitatongas in the east, forms an area that is poorly stocked with game, and is surrounded by the heavily stocked areas of the gneiss, basalt, sedimentary rock and (latterly) the higher elevation dolerite. Of these well-stocked areas the basalt only (Gunye’s, etc.) is somewhat unattractive to the buffalos, though they wander into it in small parties in the rains. Their distribution is chain-like, the new buffalo colonies of the dolerite connecting occasionally with the herds of the sedimentary “ Oblong,” these with the herds of the Umvuazi—Lusitu Brachy- stegua-morsitans area (the gneiss), these with further herds in the Moribane district (at Muchamba’s, etc.), and these again, it is said, with those of Chermgoma. With them are found elands and elephants in numbers, animals that are also rather specially attracted by Brachystegia bush, yet have latterly invaded the dolerite also. Bush-pigs are exceedingly and increasingly abundant on the sedimentary rock and the dolerite, far less so on the gneiss and basalt, where they are largely replaced by wart-hogs. Herds of sable antelope are distributed through the country generally. Lichtenstein hartebeests are present in numbers throughout the two formations east of the Sitatongas, also (but much less so) on the sedimentary rock, but not at all on the dolerite with its jungly grass-growth. Waterbucks occur in small herds on the large rivers. Bushbucks, reedbucks and duikers are present throughout, where the conditions are suitable, and little blue duikers (Cephalophus monticola) are in some numbers in the dense forests and thickets. 328 Cc. F. M. SWYNNERTON. Baboons are abundant everywhere, especially perhaps in the Sitatongas, in the mountains of Spungabera and parts of the British border, and in the dense Lusitu wooding. The forest monkey, Cercopithecus albogularis bevrensis, is to be found wherever high dense forest is present, and the bush monkey, C. pygerythrus, is ubiquitous, in smallish numbers. Distribution of other potential Hosts of Tsetses. I shall refer sufficiently to man and the smallest mammals below under “ Food of the Fly.” But facts to be referred to also, and particularly Lloyd’s results from his examinations of the gut-contents of. G. morsitans, show that in considering the fly’s dependence on the local game population we cannot leave out of account the local bird and reptile population. Guineafowls and francolins, the former particularly, both crested and horned (Guttera edouardt and Numida mitrata), are everywhere abundant, the crested guineafowl especially frequenting dense forest. Of doves, all ground-feeders, two of five common species occur in dense forest, three throughout the other woodland types. Each of these two main woodland divisions boasts also its own considerable low-searching insectivorous bird population and its common touraco. The smaller seed-eaters—waxbills, whydahs, weavers, canaries—are a very important item in the bird population everywhere except in dense forest, and even there (as I noticed in the Madanda) in the open spaces that are covered with grasses that burn. They are nearly always found on or near the ground, and commonly in great flocks. They are most abundant in the non-forested, heavily seeding dolerite areas—with less permanent fly. Quails were seen everywhere in grass-country and much on the borders of the morsitans vleis. Another bird that haunts every type of woodland and has been recorded as being attacked by tsetses (Lloyd) is the ground hornbill (Bucoraz cafer), purely a ground- searcher. It walks in small spread-out parties and (in common with the insect- tivorous birds) makes up somewhat for any lack of numbers by adopting a definite beat, most parts of which it revisits at intervals of a few days. In common with them also it probably does the fly more harm than good, for it destroys many snakes and the young of ground-nesting birds. Secretary birds and the various storks may be classed with it in this respect and are found in rather special association with the two formations about to be mentioned. | The water-haunting birds are found mainly on the basalt (on the Buzi) and granite-gneiss. Jacanas, egrets, bitterns, hammerheads, moorhens, rails, grebes, and occasional herons are found at the larger pools of the vleis of the latter formation, constantly but in small numbers and in contact with some of the dry season centres of G. morsitans. The two largest rivers and such fine series of pools as those east of Chibabava boast a far larger water and waterside population, but, so far as I have seen, it does not come into really intimate and essential contact anywhere with the main tsetse populations. The ground-running birds—the various bustards and coursers, dikkops, sandpipers, plovers, etc.—are particularly addicted for many spring and summer months to the shorter-grassed and sparingly wooded basalt, and certainly also to the granite-gneiss, though I have not seen them there when the grass has been recently burnt. THE TSETSE PROBLEM IN NORTH MOSSURISE. 329 The main reptile-population follows closely the two biggest rivers and is not sufficiently intimately connected with the big permanent fly-belts to be worth discussing. In the fly areas I noticed more snakes, including several puff-adders, on the granite-gneiss than elsewhere. Other reptiles were a large arboreal Varanus, which occurs in various types of secondary wooding, including Brachystegia, but of which I have seen few individuals in all, and some not very abundant Agamid and smaller lizards in the Brachystegia areas. Distribution of the Fly in relation to the Geological Formation. The granite-gneiss, with its open-stemmed Brachystegia covering and frequent vleis, is the special home of G. morsitans, which in the dry season is still to be found in numbers and breeding actively beside particular vleis. G. pallidipes is also plentiful there, and G. brewpalpis is found in small numbers where the shade is particularly heavy. The basalt, with its sparse stunted bush, was still carrying an extremely thin sprinkling of G. morsitans and pallidipes in late July, when burning was barely beginning. I am inclined to regard it as, in the main, an overflow area from the gneiss, and it is quite likely that any measures that banished fly from the latter formation would incidentally banish it from the adjoining basalt. The sedimentary rock area, with its fine Brachystegia forest and considerable if patchy undergrowth, is apparently more or less heavily infested almost throughout with G. brevipalpis. G. pallidipes also occurs, but in small numbers, said to be greater in the rains. Of the two woodland sections of the dolerite area the Mafusi-Musapa high-forest section, with its extensions, appears to be more or less heavily infested throughout, at any rate in the dry season, with G. brevipalpis, G. pallidipes also occurring. Pupae of G. austent were taken in numbers near the Buzi-Mtshanedzi confluence. In the Puizisi-Mtshanedzi area on the other hand G. pallidipes, though in small numbers seems rather specially in prominence, perhaps in part owing to the relative absence of the bigger fly. The latter does occur here and there in small numbers wherever the shade is suitable. Behind the Makwiana escarpment the country appeared to be clear of the fly by early August and had probably been so for some weeks, though it would still doubtless be possible occasionally to encounter there an odd straggler carried from below, particularly on the Chikambwe’s tributaries, where leaf was present. At the lower elevations, where also the bush of the dolerite retained more leaf, fly was still met with ; but judging from the observations that Dr. Lawrence has most kindly carried out for me since my departure, it would seem that it must become specially scarce there also towards the close of the dry season, though not extinct. It is in part a matter of leaf-fall, for at his station (Gogoyo) pallidipes showed itself once more in some numbers with other biting flies when the leaves were well in evidence again in the hot weather shortly preceding the rains. Glossina brevipalpis was taken in small numbers here and was seen by me also on the Buzi below Spungabera. An expedition sent by me to the Chinyika in late September failed to take any tsetses, though it brought back numbers of other biting flies. 330 C. F. M. SWYNNERTON. There are three great permanent fly-areas in the country investigated :—(1) The morsitans-pallidipes area of the granite-gneiss ; (2) the brevipalpis area of the Lusitu rubber forests ; (3) the brevipalpis area of the “‘ Oblong.” Between the last two is sandwiched an area of more deciduous foliage. Out of the higher parts of this such fly as has got there is ousted in the cold weather by the leaf-fall, and in its lower parts the fly becomes scarcer and more localised before the rains, reappearing with the leaf in October. These last remarks refer to G. brevipal- prs and pallidipes, the latter fly occurring throughout the area. Finally, there is a narrow strip along some of the frontier, widest at Spungabera and on the hill-mass enclosed by the Puizisi-Muchamba confluence, that may be regarded as quite fly-free, excepting in so far as travelling parties of male flies may on occasions be carried into it during the rains and early dry season by buftalos, ete. Nearly all the country on the British side of the frontier may be regarded as falling within this category. G. austent was taken only near the junction of the Buzi and Mtshanedzi. The type of wooding in which it occurred is not uncommon on lowland streams, and there can be little doubt that further investigation will show it to have a more extended local range. More generally it may be concluded that each fly is confined closely to particular woodland types, which, in turn, are confined to particular geological formations, the conditions that suit pallidipes being the most widely distributed. But the distribution of the permanent fly is by no means thus conterminous with that of either the buffalos, the big game generally, or any particular species of mammal, though abundant mammalian life of some kind is present in each of the areas, fly- carrying or otherwise. Man is relatively scarce (except on the Lusitu), and very small mammals are abundant everywhere in the fly area. The dominant bird elements of the brevipalpis areas are insectivorous and gallinaceous. The morsitans area has in addition a stronger bustard-plover element and, at some of the permanent fly centres, a very limited water-bird element, as well (apparently) as a somewhat larger reptilian population than the other areas. I may add that various moth larvae, numerous, gregarious and often immense, are much more in evidence in Brachysteqia areas than elsewhere ; but, for reasons to be stated, it is doubtful in spite of Roubaud’s results whether they are of great use to the fly. The fact that brevipalprs does not come readily to man makes its presence difficult to detect unless one is lucky enough to come across a male queue some evening. However, my conclusion that the “ Oblong ”’ was well infested was come to after I had seen only two flies. Their hum in flying struck me as being identical with a sound I had been hearing continually since I came there. I then made a point of disturbing all large flies of any species I saw and made my natives bring me for a day or two all they could catch for release close to me, but I failed completely to match the sound, which I therefore concluded to be that of G. brew- palpis. I continued to hear it, often on the part of large flies which I could not then identify visually that came out to us and turned back, evidently rejecting us, and of flies that flew away from the shady sides of tree-trunks when these were beaten. An observation on pigs of a day or two later and subsequent results from cattle abundantly justified the conclusion. THE TSETSE PROBLEM IN NORTH MOSSURISE. aol Vil.—Umnzita’s EXPERIMENTS IN CLOSE SETTLEMENT. Earlier History of the District. About ninety years ago, following immediately on the invasion of Bulindhlela or lower Gazaland (the Delagoa-Inyambane country) by the Zulu chief Manikusa, what are now the Mossurise and South Melsetter districts became the scene of three invasions. First came that of “the Bulindhlela chiefs, who, though defeated by Manikusa, refused to submit to him. . . . They retired in a north-westerly direction, over-ran this country, seized all cattle (then plentiful) and later proceeded north to Matoko’s.” Secondly, these fugitives were “ followed by Unéaba (alias Mosani) as far as Chiamiti—now the farm Schaapplaats in South Melsetter. Here Unéaba and his people settled.”” Unéaba was a brother of Manikusa’s father, Isigode, and he and his people were part of the Zulu invading force under Manikusa. ““Unéaba became unpopular here, one of the principal grievances being that he would not allow members of his crack regiment, the Amakanda, to marry until their heads were grey. Representations were made to Manikusa and he was induced to come up here. Fighting occurred and Uncaba, forsaken by his best fighting regiment (the Amakanda) retired with a following up the Sabi valley, and subsequently in a north-westerly direction” (letter from Mr. W. M. Longden). I do not know whether the view is correct that this “ impi” founded the “ Angoni” tribe near Lake Nyasa. Manikusa (who returned to lower Gazaland) died in 1861. His second son, Umzila, disputed the claim of his elder brother, Mawewe, to the throne and, at first defeated by him, went north and settled in succession at Umpombo’s on the Buzi, and Dongonda north of the Lusitu (see Map). Mawewe eventually followed him thither, but was defeated with the aid of the Portuguese, and Umzila took possession of lower Gazaland. He himself remained at his northern capital, which he shifted more than once. From Dongonda he moved to Gandwa, on what is now the writer’s farm “ Gungunyana ”’ in South Melsetter, and thence, later, to the neighbourhood of the Chinyika River in Gwenzi’s country, south of Spungabera. Here he made his longest stay, eventually moving to Mwayamuhle’s, just south of the Mossurise in Garawha’s country ; and here he died. His son Gungunyana shifted the headquarters across the Mossurise to the foot of Mount Singuno, but later, in about 1889, he returned to the old southern capital Bileni, near Lourengo Marques. The population of Mossurise and South Melsetter was forcibly collected and taken thither. Soon after—in 1895—the Portuguese Government had to bring Gungunyana to account, and with his capture in the following January (Theale), the Zulu native domination ended. Since then the Mossurise natives have gradually drifted back from Bileni to their old homes, but it is stated on all hands that only a small proportion actually has come back, and that the population of the district does not at all approach its former size. In 1893, the land on the British side of the border was occupied and settled by Europeans, that on the Portuguese side, in part depopulated of natives as I have stated, remaining wild to this day. In 1896-7 the rinderpest swept through the country and, missing small areas here and there, decimated not merely cattle, but in particular buffalos, kudus and elands. (737) B oon C. F. M. SWYNNERTON. Fly and Cattle before Umzila’s Time (say 1820-1861). There can be little doubt that ever since the arrival of the Bantus particular pieces of country have been subject to local concentrations of population, and that where other circumstances have been sufficiently favourable, these concentrations have resulted, while they lasted, in the opening-up to cattle-farming of the piece of country involved. The “ old people ” had told my various native informants that previous to Unéaba’s massacres of about 1830, farming went on successfully even close to the Sitatongas, where they had shown them the old cattle pits. The country was very densely settled. The chiefs Mafusi and Mtobe have both told me that at that time cattle were kept in those portions of their countries in which all subsequent attempts to keep them have failed, and Gunye has stated that it had been handed down that in the days before Uncaba cattle-farming, accompanying a concentrated population, took place even in his country, but that many used to die. But when Umzila came here (in 1861) the whole country had reverted to wooding and game as the result of the depopulation arising from Unéaba’s massacres. It was much as it is now ; fly had become plentiful, and the mountains of the present political border were the boundary, then as now, between the fly and such cattle as existed. Umzila’s Experiments (1861-1889). | Both when at Umpombo’s, and later at Dongonda, Umzila made a determined attempt to keep cattle, but in each case he failed. The concentration of population was not so great as in his later attempts ; it was short-lived, and close by was a great belt of G. morsitans—a fly which follows man. The Zulus, it may be mentioned here, knew the fly well, and the disease caused by it, and they regarded the proximity of game as dangerous for cattle. But why did they place their cattle amongst the fly? They said, “this whole country is full of it—where shall we put our cattle ?” However, cattle had always thriven in the open hilly country which is now South Melsetter; they were being captured at this time (after Umzila’s defeat of his brother) in great numbers in raids on the Mashonas, and Umzila at last sent them all to Gandwa. Later he said ‘“‘ I cannot live away from my cattle,” and it was thus he left Dongonda. From Gandwa, Umzila sent an order to sondela enkosini (draw near to the King). Thereupon an immense compulsory movement of the population took place. The country to the east of the Sitatonga Hills, particularly in and south of Gunye’s, was at that time more fully populated than that to their west, but almost the whole of this population was deported, territorial chieftains and all, to the lower parts of the tract between the Sitatongas and the present British border, to Spungabera and Gwenzi’s country, and (with Umzila’s further movements) to the Umswirizwi (Mossurise) valley, to Zinyumbo’s Hills and Chimbiya’s, and even eventually to the Sabi east of this area. The chief Gunye told me that he himself was compelled to settle at Chimbiya’s. The Effect on the Wooding. Every one of my informants has described most graphically the result of this concentration. The bush simply disappeared and the country became bare, except THE TSETSE PROBLEM IN NORTH MOSSURISE. 333 for the numberless native villages and a continuity of native gardens. This complete clearing was confined to the lower-lying areas. The Puizisi-Mtshanedzi hill-mass, as Makwiana told me himself, was relatively little affected by the new settlement, and the bush was never at all completely cleared. It was, however, surrounded on three sides by a broad cleared cordon, and on the other, backed by the highlands. In Mtobe’s and Mafusi’s the bush was never fully cleared ; the rubber forests already existed and they continued to do so, although, according to Mafusi and. others, they then consisted merely of umfomoti wooding (Piptadenia) that was chiefly confined to the ravines. In Gungunyana’s time they had already spread greatly (under encouragement), and many Landolphia vines existed throughout the new areas not yet wrist-thick. They were a valuable asset to the Zulus, who traded the rubber for cloth on the coast, and used the latter in turn in the barter of cattle. Again, away from the neighbourhood of the Mtshanedzi and the Buzi the ‘“‘ Oblong ”’ remained completely uninhabited and uncleared. Zinyumbo’s area, like Gogoyo’s, was very completely cleared—“right to the Mwangezi it was gardens only,” as was Gwenzi’s country, the Mossurise valley and portions of the Sabi. The Effect on the Game. Large mammals became very scarce—not merely big game, but pigs and baboons. “The Mangoni (Zulus) are killers of everything, men and animals.” Drives with nets were organized across the entire country, and game, pigs and baboons were thus killed wholesale. If a herd of buffalos was reported subsequently anywhere west of the Sitatongas, it was at once hunted ; if pigs appeared in a garden, they were at once tracked down to their retreat, and, the people round having been called out, were surrounded and killed. Except on its fringes the “ Oblong,” then as now, was a great uninhabited game reserve, The game in it was thinned, it is true, and was kept well driven within its borders, but there still remained enough to attract the Mangoni hunting parties. In the heavily settled areas a few bush- bucks, duikers and pigs were still to be found throughout the period. The Effect on the Tsetses. There were still plenty of int-hesi (tsetse) in the tondo-bush on the granite-gneiss, and fly never disappeared in the “ Oblong.”’ In neither place could cattle ever be kept. In Mtebe’s country too and the eastern part of Mafusi’s—that is, within a short distance of rubber forest—they could never keep cattle. Mafusi told me that he remembered the fly there, with the same distribution as now, from the time he was a child, except that it has spread with the spread of the forest. At Maronga’s and beyond—that is, in the continuation eastward across the Lusitu of the rubber forests—cattle could never be kept, though Usele and other Zulus settled there and made the attempt. In such places the cattle required for ceremonial purposes used to be brought, as needed, from the safe areas. Cattle also continued to fail in Gunye’s country and south of the Buzi from the Mwangezi eastwards, though persistent and prolonged attempts were made to keep them in this area of sparse fly, the losses in some cases being made up time after time from outside. (737) B2 354 C. F. M. SWYNNERTON. In Zinyumbo’s hills and even just on the Mwangezi cattle succeeded, as they did thence westward through the Mossurise valley and northward to Spungabera. This tract was full of cattle. But few deaths took place, and these always along its extreme eastern margin, on the borders of the “ Oblong.” In the Gogoyo-Makwiana tract cattle were kept right under the Sitatongas both at and opposite the Rupisi and from the great bend of the Mtshanedzi to its source, also in the hills behind the cleared guard-area between the Mtshanedzi and Puizisi and up to and beyond the present British border. Many of the herds were large. On the Sabi (present British territory) the results were particularly interesting, for this had previously constituted a separate fly-belt, which was eventually almost completely wiped out by native cultivation. The rinderpest may have given the coup de grace to the surviving remnant or two (recent events render this doubtful), but at any rate cattle were already being placed and kept successfully all over the old fly area in the seven years between Gungunyana’s departure (with all the cattle he could take) and the advent of the rinderpest (W. M. Longden and others). When the country was closely settled, cattle were kept successfully in places where they had always died before ; and when the settlement was well established, they succeeded where in its earlier days they failed, though fluctuations still took place with successive shifting of the population. It is true that herds actually abutting on the fly still suffered small and occasional losses, as they. are doing to-day to a greater extent on the present fly boundary. What drove the Fly out ? From the failure throughout the 25 years or so of this experiment to keep cattle near the Mafusi rubber forests, well settled and well cleared of large mammals though they were, it is likely that it was not the mere destruction of its food supply that cleared the Zinyumbo, Gwenzi and Puizisi-Mtshanedzi areas of fly. We may compare also Mr. Pollard’s statement with regard to the Munshi division of Northern Nigeria (Bull. Ent. Res. 111, 1912, p. 221). “It is interesting,” he writes, ‘“‘to note that the Munshis are great hunters and that they have practically destroyed all the wild game in their district, and yet, in spite of this, the trypanosomiasis of cattle and horses is rampant.” But the game does not merely feed the fly ; it also carries it, each year, from the permanent fly-areas into the more deciduous areas adjoining them as soon as these have sufficiently regained their leaves. The heavy settlement of the broad river valleys and the base of the Sitatongas, by clearmg the bush, not only rendered those areas themselves inhospitable to the fly, but also opposed a barrier to the passage of game, and so protected from the fly the large enclosed piece of chiefly deciduous bush behind the guard-area. Previous to the establishment of the guard- area it had been impossible to keep cattle in the area behind it, as it is now again impossible, and the broad guard-area itself, to judge from the descriptions of the bush cleared, had probably harboured even more permanent fly then than it does to-day. That the fly-population itself was diminished by the Zulu system of late burning is also by no means unlikely. The case is most instructive and points the way for future measures. It must be remembered that the flies concerned—brevipalyis and pallidipes, especially the THE TSETSE PROBLEM IN NORTH MOSSURISE. BaD former, which is more abundant—are not such great followers of man as is morsilans, and that the chance of success in such a case, without wholesale clearing of the bush, is greater than where morsitans is concerned. On the Sabi it was different. Here the permanent fly-bush itself was destroyed to a far greater extent and the fly in most parts exterminated. Was the Experiment intentional ? These concentrations of population round the King were not primarily fly- measures ; they took place whether the country was fly-infested or not. But their effect in fly-country was well known to the natives, and I was informed that Umzila entertained considerable hopes in relation to the keeping of cattle from his successive concentrations at Umpombo’s, at Dongonda, and west of the Sitatongas. Success was attributed to the effect on the game rather than to that on the shade, and Umazila’s measures against the game were stringent. His intention of colonising the “ Oblong” was frustrated by the depredations of lions on the settlements placed there (e.g. in the Budu country), and there were difficulties connected with water and an immemorial and bloody feud with regard to the Dengaza tract between the chieftains Makwiana and Gogoyo;* but he settled its borders very heavily and kept the game well driven within them by means of frequent hunts. The Effect of the subsequent Depopulation. Before Gungunyana carried off the population to Bileni (near Lourengo Marques) in about 1889, he had already commenced to protect the game. He had decreed az-zale (let them multiply), and game had become more abundant both outside and inside the cattle-keeping areas. The guard-areas still opposed its passage into and out of these areas and no harm resulted to the cattle. When the population left, the game (in the words of my native informants) just “ burst forth” (za- dabuka). At the same time the wooding was let loose and soon re-established itself throughout the previously settled country. In a very few years (by 1896 according to native information) the fly had more or less regained its old wet season limit—not many miles east of the present political boundary—though the bush capable of supporting it was not then so ubiquitous as it is now. I myself lost a beast from fly on “Scott's Hope” (a Portuguese farm on the British border) in 1900, and a tsetse was taken south of Spungabera, on the tondo-covered hills west of the Buzi (spot shown on map), in the same year by Drs. Wilder and Lawrence. The fly caused little trouble then, but two things have happened since—a pro- gressive invasion of the upland dolerite (a) by the deciduous wooding, and (b)—more important still—by game of wandering species. Buffalos and elands especially, decimated by the rinderpest, are now extremely abundant, and having spread from the sedimentary area into the hills, wander much more frequently than they used to into the highlands. It is only in the last three years that bufialos have reached the Puizisi country in any numbers. They have there tapped another brevipalpis area, and the consequence has already been visible in the destruction of the cattle of the Muchamba-Puizisi valleys. * Not that this was allowed to continue actively under Umzila. He decreed * There must be no more war. We are the only King (Si’nkosi sodwa) ! ”’ 336 C. F. M. SWYNNERTON. VIil.—Tue Foon or tae Fiy. Observations by numerous workers elsewhere have shown that tsetses living in contact with large reptilian and avian populations very freely utilise the former, and perhaps also the latter as food. Gut examinations even of G. morsitans have shown in some cases (Lloyd) a remarkable proportion of non-mammalian contents ; the same fly has been seen to feed on the bare neck-wattles of an unconfined ground hornbill (also Lloyd) ; a case is on record (Ensor) in which a fly, “ almost certainly morsitans,’ was observed through glasses trying to get under a perched hawk’s wing ; and in the Entebbe laboratory Glossina palpalis “fed readily on captive fowls, creeping under the wings to reach the poorly protected parts”’ (Bagshawe). It follows from this last and extraordinarily important observation (which does not surprise me when I remember that morsitans rather reminded me of Olfersia ‘in its low poise and clever sidelong movements in response to attempts to catch it) that birds are not necessarily protected against tsetses by their armature of feathers, though many of them certainly are by their agility and insect-eating habits. Yet young birds, whether nestlings or runners both in the nest and for a time after leaving it, are defenceless—and abundant just when the fly is abundant. Also the fly (as met with on this trip) retires to rest later than do most diurnal birds, the latter being thus relatively unprotected from it during a very brief portion of each day when the fiy is particularly active—but the bird (it must be admitted) less conspicuous. Such facts, with the way in which birds, including ground hornbills, keep to definite beats, the whole of which they work every few days, and the fact that nearly the whole bird-population searches for its food and breeds either usually or very commonly within a few feet of the ground, the search being conducted with much bustle and conspicuousness on the part of the large combined parties of small birds, must be taken into account when we attempt to explain such cases as Umzila’s failure to keep cattle near the then much smaller rubber forests (in spite of a great hunting population and a strenuous eflort to exterminate game) and when we consider the question of destroying game to starve the fly. At the same time, so far as small birds are concerned, it seems to me that the fiy’s habit of choosing harmoniously coloured surfaces to rest on suggests that its effort will be to evade rather than attack them during their active hours and stages of existence. I feel aiso that the very striking connection between the fly’s breeding habits and game raust be given its full weight. My own observations dealt only with mammals. Mammals from the smaller Unguiates downwards. The natives of the morsitans area were unanimous in their statement that tsetses feed freely on baboons, and that “ wherever you find baboons you will also find fly.” They gave me many instances in which baboons driven from their gardens had left numerous replete flies behind and others in which flies were attracted in numbers to baboons that were killed. Simpson’s similar personal observations and native statements on the Gambia and Gold Coast will be recalled, as will Lamborn’s indirect evidence. Monkeys (Cercopithecus pygerythrus, which lives in lower bush than C. albogularis bewensis and raids natives’ gardens) were stated more rarely THE TSETSE PROBLEM IN NORTH MOSSURISE. 387 to have done the same, and—an important statement if true—it was stated that replete tsetses were occasionally found where cane-rats (abundant in much of this country and very largely diurnal) had been many together. Pigs were as universally incriminated as baboons, and here I obtained a quite excellent instance myself. In the “Oblong” (east end) in cloudy weather we walked right on to four bush-pigs sleeping. In their hasty rush they left the flies behind, and these streamed after them in great numbers and with quite a hum. We captured nearly twenty that through repletion could scarcely fly ; all but one were brevipalpis, the exception being a pallidipes, and more than half were females. Austen records both morsitans and brevipalpis as feeding freely on wild pigs and quotes Dr. Hearsey’s statement that G. morsitans was seen to settle literally in hundreds on the carcase of a wart-hog, behind which animal | also took these flies. My indirect evidence of the value of such animals as pigs to the fly was also interesting. In a mile-wide patch of primary forest east of the Sitatongas, in which I saw much brevipalpis, both native information and a careful search for spoor showed that it could have been feeding on nothing but pigs, baboons and smaller fry. The same applies to a piece of high, dense, secondary forest on the western foot of the hills, into which according to the owner of a kraal on the spot (confirmed by the usual search for spoor) no big game had entered for some months. It also applied, I am certain, to much of the rubber forest area in which pigs and little blue duikers—and these only—are abundant. In one place a length of about 300 yards for a very great width was continuously turned up by the pigs and looked like a hoed plantation. It made me wonder whether one of the forest hogs (Hylochaerus) may not occur in these forests. Finally, working for over a fortnight round my camp on the Buzi east of Spunga- bera, in an area in which game is relatively abundant, I made a special point of studying the daily spoor in relation to the distribution of G. brevipalpis. To sum up the result, there was a considerable area that I am certain was not entered by big game during my stay, or for some days before it. There was a smaller, inner | area, immediately round certain kraals, that had probably not been visited by such animals for many weeks or months. In genera] the spoor showed much less move- ment of the big game now than in the wet season, yet the fly was equally present throughout, lurking in all the thickets to attack passing animals. The only “ passing animal” that showed a similar ubiquity was the bush-pig, and I was convinced from the evidence that the fly was living practically entirely on bush-pigs at the time of my visit. Man was not being attacked. My friend Mr. G. D. Otterson spent a few days with me here, and declared on leaving that the evidence of the fly’s independence of the bigger game in the matter of food was a ‘‘complete eye-opener ”’ to him. I conclude from these observations that any attempt in north-west Mossurise to destroy the fly by starving it in its permanent haunts is doomed to failure if the bush-pigs, and perhaps the baboons also, are not destroyed ; and the destruction of the pigs in this type of country is not easy. Two species of lemur occur in the fly areas. The larger of these, Galago crassi- caudatus, is abundant in most of the brevipalpis country and sleeps exposed (but perhaps protected by its fur) in the daytime. Bats, including the largish fruit-bat, 338 C. F. M. SWYNNERTON. Epomophorus crypturus, are abundant, and insectivorous species were found sleeping in great numbers in hollow trees of Sterculia trvphaca on the ant-heaps of vleis frequented by G. morsitans and G. pallidvpes. Bushbucks and duikers are more or less common—and hard to exterminate. The little blue duiker (Cephalophus monticola), more or less abundant in thickets and dense forest, is less hard to reduce by fence-trapping. It has a habit of resting under logs and between tree-buttresses that might bring it into additional contact with the fly. It was one of the animals least liked by my carnivorous mammals. Of yet smaller mammals two species of Tatera came in great numbers to my traps at a spot where male clusters of morsitans were present all the time on the grass and females in the wooding (PI. xiv, fig. 2). Lloyd has suggested that nocturnal mammals, including Murip4g, are perhaps liable to the attack of tsetses owing to the latter’s habit of resting in similar holes. T have myself taken morsitans puparia from holes made by animals in a bank, and noted that brevipalpis and allidipes rest in holes in trees. Nevertheless, an argument exists against this view in the fact that some of the bats at any rate are very easily disturbed, and the part played by the small nocturnal mammals, including bats, needs far fuller testing than it has received. Species of Arvicanthis, Otomys, Myosorez and Petrodromus are largely diurnal, but their small relative size suggests both their ready perception and intolerance of the fly and their less ready detection by it. ; Dassies or rock-rabbits (Procavia) may be of much use to the fly, but very locally and only in rocky places. I found them in the coarse quartzite rocks that form the crest of the Sitatongas and support Brachystegia; and I spent many hours with several workers trying to find wet season tsetse puparia in the humus under raised and overhanging rocks, but without success. I searched similarly in the Maruma forest, frequented by dassies, but Maruma, I judge, in spite of the outbreak of trypanosomiasis there, is seldom if ever reached by the fly. The larger Ungulates. I have thought it well to give prominence to the evidence favourmg the view that the pigs, aided doubtless by the smaller buck and perhaps yet more by the baboons, will continue to support a fly population in parts of north Mossurise even if the larger game is banished. At the same time it goes without saying that the larger the permanent game population the greater will be that of the fly; and I obtained abundant indirect evidence that, as things stand now, the bigger ungulates present in the fly-areas are taking an important part in the feeding of the tsetse. As I shall describe elsewhere, I found numerous batches of puparia—mostly of morsitans, a few of brevipalpis—associated with spots where larger mammals had been lymg down. Buffalos were incriminated the most frequently, but Lichtenstein hartebeests, wart-hogs and other species were concerned. My finds of pupae in the morsitans area seemed rather to suggest a preference for the buffalo on the part of the fly. I found this definitely asserted by some of the natives and denied by others. My guide in the morsitans area, a very observant native, was particularly convinced of it. Questioned as to a connection between the two animals, he replied that one may find fly where there is no buffalo, but that where there is a choice it follows the buffalo. ‘“‘ The buck,” he said, ‘‘are much THE TSETSE PROBLEM IN NORTH MOSSURISE. 339 more restless under its attentions than the buffalo, the hartebeests especially keeping up their dance when tsetse are about them; so that the fly can feed more easily on the buffalo.”” Neave has already made a suggestion to this effect. I shall discuss this subject further under the next heading. Elephants were stated by the natives to be attended by tsetses when the latter are numerous, and I took a male brevipalpis waiting on a much used elephant path. Roubaud and Bouet (referring to G. longipalpis) are both quoted as speaking of a special association between tsetses and the elephant and hippopotamus. Man—and the Fly’s Preferences. Using cattle as I did, I obtained abundant and excellent evidence that G. morsitans and G. pallidipes attack these animals more readily than they attack man, that G. morsitans attacks man more readily than does G. pallidipes, and that the latter, in turn, attacks man far more readily than-does G. brevipalpis. It constantly and everywhere happened that the carriers passed through a place without drawing fly and that the cattle coming just afterwards attracted many. A dozen to twenty carriers are more conspicuous than two or three head of cattle and their scent is overpoweringly stronger. Similarly I have seen a fly (pallidipes) on a leaf beside the path allow carriers to pass it unmolested and then at once fly out to my donkey. Another piece of evidence was afiorded by a fly (brevipalpis) that 1 found perched on a projecting slab of shale in the path, facing away from me. I stopped short (about two yards away) and had the cattle (immediately behind me) turned into the grass. After about two minutes the fly flew back to me and, without alighting, returned to its perch. I then had the two cattle immediately brought up, the leading ox standing nearly level with me in the path. The fly shortly flew back again, first to myself, then, swerving off, to the ox, under which it at once settled and began to feed. A point that struck me in the morsitans area was that whereas when we were digging for puparia there was often delay before a fly attacked us (one fly was actually seen watching on a tree-trunk for a time before it came), the cattle were attacked quickly. Man is attacked by all these flies much more frequently in the rainy season, but the increased attacks are probably not more than proportionate to the great increase in the number of the flies at that time of year. Dr. Lawrence has sent me records of occasional attacks on natives at his homestead at Gogoyo’s even by brevipalpis, and as for morsitans (no doubt in company with pallidipes), I was informed by the natives across the Sitatongas that they become at times unbearable, forcing every man to carry a leafy twig with which he continuously switches his back and shoulders. G. pallidipes (it is stated) also becomes a nuisance in places west of these hills, Concerning the practical question, “if other sources of food were eliminated, could the tsetse still keep going with the aid of man?” it is certain from the observations of Lloyd, Maugham, Stephenson-Hamilton, myself and others that the tsetses would then attack man much more. Tsetses (as I have seen) constantly obtain full feeds from man and escape unscathed and, where the bush comes up to a village, so far from avoiding it, morsitans and pallidipes—but not brevipalpis— appear to become rather a nuisance. 340 Cc. F. M. SWYNNERTON. It is doubtful whether brevipalpis, a relatively easily detected and easily caught fly, would readily adapt itself to man. With regard to morsttans it may be said that the conditions favouring its dry season centres would have to be present as well as the villages. I only saw one such apparent coincidence here, but there is no reason why it should not be common. The size of the human population and its effect on the bush when really large are further factors to be considered. Of our fly areas here the “ Oblong” is uninhabited and the higher Mafusi and morsitans areas are not thickly inhabited. I believe it is very different on the actual Lusitu, but there brevipalpis would be the fly permanently present. I was myself bitten by morsiians far less than my natives, but this was presumably a matter of skin-colour and clothes ; Mabuzana, living near the Mtshanedzi south of Gogoyo’s, volunteered the information that tsetses were specially attracted by a black coat. I said “ How do you know that?” “ Because I have one!” Perhaps more surprising than the preferences shown in relation to man are those shown with regard to his domestic animals. I had three female native goats with me throughout the trip and in about the fifth week added a male; and until I obtained cattle, about a week later still, I used these goats continuously for bait. They proved most unsatisfactory, and although they were occasionally bitten, I bave little hesitation in saying that even man was much preferred to them both by morsitans and pallidipes and was not liked worse than they were by brevpalpis. Put ahead of the cattle they failed to draw fly where the cattle would draw many, A dog that was with me throughout the trip and still remains healthy was relatively httle bitten. Healthy-looking native dogs and goats were present throughout the fly-areas and were stated to be much bitten in the rains. I regularly fed my flies on the goats and once they bit they fed fully, but a relative reluctance to bite was distinctly seen, and this reluctance extended to a sheep on which I fed the flies that I finally brought home with me. The relative reluctance shown by Dr. J. W. S. Macfie’s flies to feed on guinea-pigs, dogs and cats is worth recalling, also the preference of Lloyd’s flies for monkeys as against goats. Experiments that I carried out on AstLIpAE showed that those flies possess graded preferences along the same lines as carnivorous vertebrate animals experimented on, so that there would be nothing extraordinary in the tsetses doing the same. Experiments on carnivorous animals, including lions, showed the following order of preference :—({1) Pig, wild and domestic ; (2) beef (nearly equal to pig); bushbuck and sable antelope were liked very nearly as much ; (3) goat, sheep, dog, man, blue duiker. These last were liked far less than (2). It seems probable from the evidence I obtained on this trip that the preferences of the tsetses I had to do with would follow much the same general lines. ASILIDAE often attack the highly nauseous Acraeine butterflies and tsetses attack man, but their preferences are very real nevertheless, in view of the fact that after a feed they do not wait till hungry enough for Acraea or man before feeding again should anything better turn up. Lloyd and others have shown, and my own observations confirm them, that when there is nothing better to turn up, the fly will at last “occur in such quantities ’—that is to man—“‘as to be a source of the greatest annoyance’ (Maugham). At Kanyezi’s vleis, with some game about THE TSETSE PROBLEM IN NORTH MOSSURISE. 341 but buffalos absent, we saw a good deal of fly and were often attacked by “ feeders ” (as distinguished from “ followers”). In the Masando vleis in 1900 with game temporarily very scarce and bufialos absent I was very greatly pestered. This time, in the same Masando vleis, with buffalos present, I found few flies, and at a series of vleis a few miles further east which the bufialos had just reached in numbers I found none, though the natives stated that they had been there in great quantities up till then. At first at these places I had not the cattle, and it was legitimate to suppose, as the natives did, and it was doubtless true of the males, that the tsetses were “away following the game.” But the female fly is a feeder, and apparently follows little, and the cattle quickly showed that flies were present, but that they were being kept too well fed to show themselves to less acceptable man. It is easy, understanding the working of the fly’s preferences, to see that a failure to find non-mammalian blood in an examination of its blood-contents is quite compatible with the possibility that it might live on birds and reptiles were mam- mals removed. It follows also from my observations on this expedition that the old idea that tsetses possess a preference for the bufialo may be perfectly correct, though it will show itself strongly only where, and while, buffalos are so abundant as to make the fly comparatively independent of less favoured food. The special enthusiasm I saw shown for an actual congener of the buffalo seems to be highly significant in this connection. Unfortunately there would seem to be no very practical method of utilizing the flies’ preferences except by using cattle to draw them out. With its favourites destroyed it will live well on the next best and so on. Arsenic-tolerant cattle, sufficiently saturated, might be worth experimenting with as bait. Drinking of Water by Tsetses. My flies frequently applied their proboscides to grass-stems put into the tube wet (to simulate the effect of dew), never to those inserted dry. My native informants on the Umvuazi agreed in asserting that they frequently saw tsetses in numbers “ drinking” on the wet sand of the river in very hot weather. I have watched Tabanus on the wet mud of pig wallows at the Amanzimhlope head-waters with its proboscis in definite contact with the moisture, as were those of non-biting Muscid flies of several species that were also present, and this habit of drinking is @ common one in hot weather on the part of both Diptera and butterflies. King- horn and Lloyd have both noticed tsetses at the edges of puddles ; Moiser obtained records of it from his natives, and saw his captive flies insert their proboscides into the wet soil of their bottles ; and Lloyd obtained the same result from moistened sponge and moistened blotting paper that I obtained from wet grass. Carpenter has gone further, for he has traced the presence of the liquid inside the fly. The point is not merely of importance in relation to the hot-weather needs of the fly ; it may also be important, it seems to me, in relation to the possibility of poisoning the highly localised male swarms of (. morsitans by spraying the grass they rest on. Lloyd’s positive results from slices of water-melon go to confirm Maugham’s obser- vations (under natural conditions) as to the sucking of vegetable juices. It is probably moisture rather than real nourishment that would thus be sought and obtained. 342 C. F. M. SWYNNERTON. [X —DISTRIBUTORS OF THE FLY. Game not only helps (a) to feed the fly, and (6) to provide the trypanosome, but (c) it helps to distribute the fly, carrying it back each summer into the areas from which the fall of leaf had driven it. That the fly does definitely spread at each raimy season, over the latter type of area is Shown by native statements, by nagana outbreaks, and by my finding in the highly deciduous country a mudstained pupa-case of G. brevipalpis, one of our most shade-loving tsetses. That the fly wanders back into these areas partly without the aid of game is possible enough ; but without their aid I doubt if it would have re-occupied much country by the time the leaf fell again. The tsetse is a distinctly sedentary insect, and throughout reminded me much of a tick, excelling it mainly in flying a little distance to its prey instead of waiting to be brushed off by it. This, with the details of some of the outbreaks of trypanosomiasis, the fact that tsetses actually do follow animals for long distances, and the cogent indirect evidence I shall refer to on page 358, support the view that the game is. of the first importance to the redistribution of the fly. Accepting the hypothesis, the most important pomt to note is that the fly’s chief distributors will be by no means necessarily identical, in a given area, with its chief food animals. The individuals, pairs or herds of most of our antelopes have definite circumscribed haunts and grazing grounds that they keep to. In the case of bushbucks and duikers these are small, and parties of pigs also for long periods together rootle and depredate from particular lairs or lair-areas. Harte- beests also have relatively small beats, and herds of sable antelope and kudu have their permanent grazing grounds, larger than these others, though the kudu is sometimes a wanderer. Animals with this localised habit may be regarded as the fly’s most reliable food-supply in its permanent haunts, even though some other animal may temporarily become more important through sheer numbers. Elands, on the contrary, are great wanderers ; elephants cover much ground ; and the buffalo, if it is on the whole less of a wanderer individually, and in the matter of long purposeful “treks” than the eland, nevertheless covers, with its great herds, very large grazing grounds, is continually wandering back and forth between its various centres inside and out of the permanent fly-areas, and breaks up into parties and individuals that, to judge from the spoor, leave little ground in the general range unvisited in the rainy season. Also, as I shall show elsewhere, the buffalo more than any other animal has been increasing its numbers and range here at a great rate for many years past. I believe then that these animals, and the buffalo in particular, are mainly res- ponsible for the annual spread of the fly, and that so far as we are concerned, in and near Mossurise, our main grievance against the game is not so much that it feeds the fly (which would be fed and contaminated in any case by the pigs) but that it carries it far and wide in the rainy season and so brings it into contact with the cattle. The effect of the difference in the travelling habits of the male and female flies must of course be duly taken into account. THE TSETSE PROBLEM IN NORTH MOSSURISE. 343 X.—EFFECT OF THE RINDERPEST. The effect of the rinderpest is of practical importance in relation to the question of destroying the game to starve the fly. In reply to my question, addressed to many natives, as to whether the fly became less after the rinderpest, I received two answers, one from south of the Lusitu, the other from north. It may be stated at once that both buffalo and fly are said to be much more abundant, and to have been so before the rinderpest, in the Brachy- stegia area north of the Lusitu than in that to its south (both of them on the gneiss and carrying both morsitans and pallidipes), so that the difference in the replies may perhaps be genuine. The reply from the former area (from the countries of Muchamba, Usambai and Udombe) was that the rinderpest was followed for two seasons only by a definite reduction in the numbers of the fly. In the second of these seasons it was already more numerous, in the third season and subsequently it was as numerous as ever. The buffalos were still scarce, so that if this statement is correct and there were also any real connection between the rinderpest and the fly’s reduction, one explanation might be that the fly had previously become specially dependent on the buffalo, owing to the latter’s immense relative numbers, and took a season or two to adapt itself completely to the habits of the other larger mammals. Wart-hogs and baboons were stated to have been particularly abundant. Another explanation will be suggested below. The answer I received from south of the Lusitu, in the fly-areas both east and west of the Sitatonga Hills, was that the rinderpest was followed by no noticeable reduction of the fly at all. One man stated on the contrary that they then bothered the natives terribly. Another made the quite sound suggestion that perhaps with their other food destroyed they attacked men more, so that, though fewer, they seemed as many as ever. All insisted that, though certain abundant species of game were killed off, plenty of other food remained, in the form of bush-pigs, wart- hogs, baboons and part of the larger game. To the above evidence I am able to add a little of my own. In 1900 I passed with carriers through a portion of the morsitans fly-area. Hunting on 3lst May and 1st June in and near the large wood and series of adjoining vleis called the “ Mahloka ” or “ Masando ” near the Umvuazi we were beset by very great numbers of tsetses, which also kept with us for considerable distances—far heavier attacks and “followings ” than any I experienced in 1918. At that time buffalos were still exceedingly rare as the result of the rinderpest of four years before. I saw no trace whatever of them during that trip—a shooting trip—and local natives I have consulted have confirmed the fact of their absence. I have already mentioned the loss of a beast beyond the limits of the present fly in 1900, and the capture in the same year of a tsetse south of Spungabera, just inside the edge of the present permanent fly. On the other hand, my information from north of the Lusitu, if reliable (it was given to the members of my native expedition thither, not to myself), certainly seems to suggest that in some fly-areas an effect may have been produced. The real, dense, busily breeding centres of a morsitans belt may in the dry season at 344 Cc. F. M. SWYNNERTON. any rate and under certain local conditions, be exceedingly localised. It is a fact also that the two animals that I have suggested to be the fly’s most important carriers are amongst the three that suffered worst from the rinderpest. It is possible, therefore, that the actual effect of the rinderpest in certain areas may have been not directly to starve the fly, but to confine it the year through to a far greater extent to its dry season centres by the destruction of its chief carriers. The narrow- ing in this way of the area tapped for food im the wet season might react also on the numbers of the fly. The suggestion is tentative and perhaps may not be borne out in the belts actually reduced, especially if man, a carrier of morsitans, was present in numbers, living or passing, about the dry season centres. The failure of the rinderpest to destroy the fly here to any appreciable extent tells decisively against the view that the rinderpest blood was in some way poisonous to the fly—for all my mformants agreed as to its severity everywhere, except in a portion of Makwiana’s highlands. The animals sometimes died “ whole herds together,’ and I myself can witness indirectly to the effect on the buffalos. Its failure tells also against the hope that we may exterminate the fly in the Mossurise district by destroying the bigger game only. XI.—FACTORS DETERMINING THE PRESENCE OF TSETSES. Shade and Undergrowth. It would appear from the present and other observations that the fly must be protected from the drying effect of continuous sunlight either by a sufficient supply of shade, or by a readily accessible moisture supply, or by both; and that this, with food, is its primary desideratum. Glossina brevipalpis, so far as I have seen (and I was with it for some weeks in all), relies very greatly indeed on shade and is rarely found away from fairly heavy shade. The requisite degree of shading is provided by wooding with leafy under- growth. This may be either primary forest (in which the undergrowth is sometimes such as to give sufficient shelter alone (P!. xui, fig. 2) or secondary bush in leaf with thickets and sapling clumps below (PI. xii, fig. 2). Of these types the primary forest (Pl. ix; Pl.xin, fig. 2) and, im many places and most seasons, heavy Brachy- stegia bush are the best capable of carrying this tsetse through the dry season. It is not at all dependent on the presence of vleis, and I have found it waiting in all the thickets at some distance from water of any kind and in hot weather in September with the ground baked. It is true that it was in greater numbers near certain little streams than in the Brachystegia thickets, but (as results from the same type of forest away from streams appeared to show conclusively) this was only because those streams were lined with fringing forest of primary type. So dependent does the fly appear to be on good shade that, except in the early morning, after sunset and on dull days, it will leave animals it is on as soon as they emerge from the shady bush into the sunlight. On dull and rainy days it will follow freely into the most open country and at high noon, so that the term “ crepuscular,” which has been applied to this fly, is not altogether justified by my observations. Thermometer readings taken at the same time in primary forest, primary forest with its undergrowth cleared, and, thirdly, in a sapling thicket in Uapaca wooding along- THE TSETSE PROBLEM IN NORTH MOSSURISE. 345 side, were identical, yet brevipalpis had deserted the cleared piece (v. Expt. 1, p. 373). This, with a failure to take brevipalpis in some high, dark secondary wooding on the Sitatonga base, excepting at the occasional thickets, suggests that extra shade may not be the only advantage gained from thickets. Protection from drying winds, additional protection from the eyes of enemies (such as the thickets very definitely afford) and some advantage to the pupae are the three that occur to me, but I am convinced that shade is the chief consideration. Glossina morsitans was found in a very different kind of country. It was very obviously far less dependent on good shade than G. brevipalpis, and I failed to find even stray individuals in the densest forest types—but the stray individuals taken were in any case very few. It occurs both on the granite-gneiss and the basalt of the lowlands, the latter usually with very poor bush indeed (Combretum, etc., Pl. x, fig. 2) and the former with savannah forest of a poorer type than that of the sedimentary area, though in each of these two it 1s Brachystegia. On the basalt and on most of the gneiss this fly occurred during my visits only as very rare individuals, and on the basalt and away from vleis only males were taken. The granite-gneiss, however, carries (as I have related) very numerous vleis, many of them with permanent water, and in spite of much painstaking search elsewhere, it was only at these vleis (and then only at some of them) that morsitans was found in great numbers together and its breeding going on with some vigour in the dry season. Here the male flies occurred in the usual bands, containing in some cases at least a few hundred flies at a time on the short or shortish grass and sedge. They were sometimes out in the sunlight amongst or beyond the scattered chidsgwati shrubs and stunted Parinariums that separate the open vleis from the Brachystegia bush surrounding them, sometimes a little way in the Brachystegia bush itself, never very far from the vleis. Some of these vleis with well populated fly-pockets consisted of mere glades in the bush, in which there was now no definitely exposed water and in some of which, had I not dug, I would hardly have expected water; but bigger vleis were part of the general series to which they belonged, and even in the glades experimental digging always showed moisture near the surface and actual water not many feet below. Some places would be moist on the surface, especially in the early mornings, when dew also was present in such glades generally. The flies were found in rather special association with a low, heavily-headed sedge (Futrena), with which I found the local natives had also learned to associate them. This remained green after the grasses generally were dry and, with one or two low-growing associates, covered considerable areas at the edges of certain vleis. Almost the only low-growing plant that was transpiring at the time of my visit, and doubtless cooler than the dried-up species, it possibly offered a greater certainty of the deposition of dew. The stray flies, of which I took fewer of morsitans than of pallidipes, appeared (as I shall indicate for pallidipes) to be finding their way to the vleis and rivers by the end of July, with leaf-fall already producing a visible effect in dry areas and at the actual vlei-edges, which latter, however, were also coming back into leaf. The natives described to me the increase of the fly in the hot weather, its spread in the early rains and its abundance then at places where, even with game temporarily absent, one now found few or none. In general it would be difficult 546 C. F. M. SWYNNERTON. to find a more fitting description of this morsitans-pallidipes area than has been given by Dr. Shircore of the “ proclaimed area” of Lake Nyasa (Bull. Ent. Res. v, pp. 87 and 88) ; although, owing to the lateness of the season in 1918 and the faet that grass-fires had now barely begun, I am not in a position to judge of the complete- ness and duration of the stripping of the drier country of the stray fly of which I seemed to have seen the beginning. This type of wooding comes back into leaf again very readily after the weather has warmed up, and it possesses an early element in the mutsatsas. The return of the leaf does not await the rains and the country 1s sometimes already very shady when it is otherwise absolutely at its driest. The relative value of these two factors—shade and moisture—in relation to the fly’s fresh redistribution should be interesting to study.* I did not find that G. morsitans appeared to be at all dependent on the presence of undergrowth. The bush in which it occurred was for the most part devoid of such growth, though some of it was itself low. JI mayadd here that in consequence of the very late fires of 1918, and of the fact that I worked the normally late-burning areas last, the whole investigation was carried out in unburned country (except for a few odd patches). Nevertheless I obtaimed no evidence in favour of the view, but much against it, that any of the flies concerned will live in open grass-country devoid of bush. Even the attacks of morsitens on the basalt were always in or beside shade, were it only that of a large shrub or a semi-leafless tree, and when noted definitely resting, either in my experimental net or in the field, this fly was always in the shade—under a log, on the shady side of a trunk, etc. A replete resting morsttans female that I disturbed repeatedly always settled again on the shady side of trees. I have seen waiting male clusters furthest from bush in large vleis, but there was some reason to suppose that the individual flies did not stay with the cluster indefinitely. Glossina pallidipes is distinctly more catholic here in its general tastes than either of the other two. It occurs both in the country favoured by brevipalpis and avoided by morsitans and in that favoured by morsitans and avoided by brevipalpis, as well as in wooding of its own; nor do the stray individuals, at any rate, avoid the extremes. Males were taken on the sparsely shaded basalt (Pl. x, fig. 2) m dry windy weather and with the leaf falling freely, and occasional individuals of both sexes were taken in primary forest dominated by Khaya nyasica in full leaf (Pl. ix). I do not feel that I disentangled its habits sufficiently from those of G. morsitans on the granite-gneiss. Here, in the Brachystegia bush, it appeared to occur in every place in which we found morsitans (Pl. xiv, fig. 2; Pl. xv, fig1). It was in great numbers only where morsitans was also numerous—namely at particular vleis and glades ; it was very sparse elsewhere, but less so than morsitans, and both flies (and once brevipalpis too) were on us or the cattle together. When pallidipes puparia were taken at all, they were under the same log as those of morsitans. * Mrs. Lawrence, writing for her husband well on in November and sending me flies, speaks of the pleasant and cooling effects of the first light rains—‘‘ but” (she adds) ‘“ the tsetses ! they are abundant.’ Tsetses had reappeared with the return of the shade, but the advent of the rains would seem to have led either to their freer breeding out or (as I think, by making more general the conditions under which they could exist in comfort or safety) to their freer dispersal from their breeding centres. THE TSETSE PROBLEM IN NORTH MOSSURISE, 347 A fact that is possibly of interest is that in my first week spent at the Kanyezi morsitans centre (28th June—-6th July) I saw and secured numbers of morsitans but relatively few pallidipes, though a good sprinkling of this fly was present. Dur- ing my second stay there (27th July—lst August) pallidipes had apparently been heavily reinforced and was now taken (both males and females) in greater numbers than morsitans. It did not seem that emergences could account for this difference ; the puparia, living and empty, taken in large numbers on each visit, contained an extraordinarily small proportion of those of pallidipes—perhaps one per cent.— though the bush generally was well searched. On the other hand the adult flies of both sexes, which at my first arrival were scattered more through the country generally, had latterly, with only two or three exceptions (these being on the basalt where moisture would be harder to find), been taken only at streams. There was a difference now in the fact that the country was drier, and that leaf-fall had com- menced and was in places appreciable at the time of my second visit, so that the evidence suggested in a remarkable manner that pallidvpes, previously to a larger extent scattered sparsely over the face of the Brachystegia area, had now congregated at the morsitans centres. It is perhaps difficult to disentangle this conclusion with certaimty from the fact that the cattle always drew, and were now drawing, pallidipes in much larger numbers than natives working without them; but on this occasion the latter were themselves securing a larger proportion of pallidipes than previously. I should say that on the granite-gneiss the presence of pallidipes was in no way dependent on that of woody undergrowth. West of the Sitatongas, where morsilans does not appear to occur at all, stray individuals of pallidipes were taken in very open bush—and in every type of bush— but it was never taken even two or three together except in Brachystegia and the less tall of the “‘ dense secondary” types. It certainly occurs in small numbers in the dry season throughout the “ Oblong” and its northward extension to the Buzi. It is even possible that this great Brachystegia block may be its main per- manent habitat here, and it is notable that my cattle sustained a severe attack from pallidipes on 27th July in a limited piece of Brachystegia wooding just north of the Dysart Concession—a wood that had previously supplied a pallidipes or two daily to Dr. Lawrence’s grass-cutters just outside it. I also took a few pallidipes between this station and Gogoyo’s kraal at a series of small vleis in very dry sandy country carrying a type of Brachystegia bush equal only to the poorest on the granite-gneiss and already losing leaf somewhat heavily, and was informed by a native formerly resident that this fly was exceptionally bad here in the rains. Apart from these instances my only real and repeated successes with this species in the “ Oblong ” area were in and near the piece of coppice shown in PI. xvii, ‘fig. 1 and representing a type of wooding that, resultmmg from native cultivation, is more abundant north of the Mtshanedzi. Here (north of this river) on a few occasions up to half a dozen—even a dozen—together showed themselves in such types as are illustrated in Pl. xi (rather low thicket) and Pl. xvii, fig. 2 (coppice). Largely on account of this predilection, partly no doubt on account of the greater scarcity of brevipalpis and absence of morsitans, pallidipes west of the Sitatongas appeared to be specially associated with the Puizisi-Mtshanedzi area, which caiTies a very great deal of wooding of these types. The coppice or fern-entwined (737) ¢ 348 C. F. M. SWYNNERTON. bush might be moderately open (as 1t was in Makwiana’s Makubvu through the trampling of elephants and buffalos). Very frequent—almost constant—con- stituents of it were the small trees or large shrubs, Brachylaena rhodesiana (ipahla), Markhamia lanata (mubfeya) and Vernonia podocoma (mudambasese), the latter bemg replaced near the Sitatongas and across them by another woody Vernonia. Bersama niassae (muyahawa), Vitex eyles (mukubvu) and Conopharyngia elegans were not infrequent, but a great mixture of other species, both of the savannah forest and the primary forest outskirts, were commonly present, and where the tree element overtopped the lower layer the wooding became fit for brevipalpis. Generally speaking, brevipalpis needs coppice with overwood, while pallidipes prefers coppice without it, but in the right types—such as rather poor Brachystegia —it can apparently dispense with yet lower growth. Clearing experiment No. | (below, p. 373) is of possible mterest in this connection. Here pallidipes attacked the cattle somewhat freely in tall primary wooding in which the undergrowth had been cleared, though before the clearing only brevipalpis had been present ; but it may have come direct to the cattle from the wooding (Pl. xvii, fig. 2) outside. It is curious that im all my special trials on the Buzi for brevipalpis in wooding of the types shown in Pl. xiu, fig. 1, I never took pallidipes either, though I did take it here in places with low growth. It may possibly best be described as a ‘‘low-wooding fly,” the low wooding varying from mere coppice and bush savannah to poor Brachysteqia, etc., though better wooding is to some considerable extent utilised. All vleis that I came across west of the Sitatongas were searched for male clusters of this fly and the neighbouring—and other—wooding for breeding places, but unsuccessfully. Though stated to be abundant at times in the rains it is generally by far the scarcer of the two flies where it occurs with brevipalpis. The latter often appears to be scarcer through its reluctance to show itself to man. Buti. ENT. RESEARCH. VoL. XI. PART 4. PLATE IX. Primary ‘“‘rain” forest; the large trees are African mahogany (Khaya nyasica’. At suitable elevations this type of forest harbours Glossina brevipalpis in the dry season. BuLc. ENT. RESEARCH. VoL. XI. PART 4. PLATE X. Fig. 1. Highly deciduous secondary forest on Mount Umtareni in August, with Pterocarpus angolensis and Terminalia sericea. Capable of sheltering Glossina brevipalpis and G. pallidipes in the wet but not in the dry season. Fig. 2. Lowland bush-savannah on the basalt at the younger Gunye’s, Gombretum dominating; harbours the overflow of G. morsitans from the Brachystegia for much of the year. BULL. ENT. .RESEARCH. VoL. XI. PART 4. PLATE XI. 7 %: mite aah * . Pn ~ Sy _* — Rem ct Runde 7 ¥ wo. 3 Veet, te: Chak he) Fo $ , P ~ BMG A typical scene in the Madanda rubber forest. An Erythroxylon-Landolphia invasion of tree-savannah and savannah-forest, capable of harbouring G. brevipalpis and G. pallidipes. a met Sipe aed Met = =x > a ' whe i a. = Ne + : Ss : ae Ds * ee i x ae OO i rs rola Kan ea Na : ee 4 ah hy ‘i Eee . j ~ + ’ cee) 7 i j pe 4a ae See he cet cee aes ee ayy = Bors Mb ae fi 2 eg 9 hile ” . a > ian 2 i = a el =i \ _ - —e * ” J a" ar a 2 y . vie ¢ - » r ~ = a > 7 i * 1 r " ° = = c y e = Ss « f , = oy Px - _ Me Se iit ? = a . © a’ =e - a ‘ er Sra “9 oes ” ‘ t . 7 - oF Lf a et ee ee ee ee ee rr a ee 2 ‘ PLATE XII. PART 4. XI. VOL. ENT. RESEARCH. BULL. In Uapaca-Brachystegia bush; two shade- layers, and G. brevipalpis present. Fig. 2. In Uapaca-Brachystegia bush; lower shade- layer only, therefore no G. brevipalpis present. Fig. 1. Buti. ENT. RESEARCH. VOL. XI. PART 4. PLATE XIII. Fig 1 In Uapaca-Brachystegia bush; only an upper shade-layer, and therefore no G. brevipalpis. Fig. 2. Landolphia wooding of the Lusitu rubber forest type; with a lower shade-layer of this high, very dense type G. brevipalpis was present even when the higher trees were leafless. + T) Ss “3 pe a ae | | i ' na - a ofS ; rs . ¢ - = a a on Ry al, ea ne _ ca f. % 7 1 é et) 7 s a “3 ia > ne ’ A oe =e ws di { - f y ~*~ ~~) ra 4 w 7 4 =a 7 a - . a - +) ; i 4 . iz 2 7 we ai ~ 7 ss Tis a = ) , ; ! -_ 106 7 oe : - 7 4 ae - - ’ f Yoh (se 2 - “ q — i, maid , = _ 3 a, a a Ub Oey : . £ - * as t mr = s _ ¢ 7 L ‘ a a oat | , > ni ‘ 7 t ‘ ae # , ‘ a4 = = ‘ A oad es. b is — a - i ‘ t ¢ a + - ‘ 4 . : * as . = 2 : A é . eu on walle z a i —— ee A apo Pobttannnes aut _—s S Z + - ‘ . = Me F) = ‘ i ’ » ~*% 6>. - 4 PT ; e Butt. ENT. RESEARCH. VoL. XI. PART 4. PLATE XIV. =a aN t . J ss 2 Fig. 1. Nearly 100 puparia of G. brevipalpis were taken under this fallen Piptadenia buchanani at the spots indicated by the papers. Fig. 2. A very typical breeding-place of G. morsitans under a long-fallen trunk of Burkea africana. — oo —"” a= -, — BuLt. ENT. RESEARCH. VoL. XI. PART 4. PLATE XV. Fig. 1. Site of a male cluster of G. morsitans; the short grass was very green and luscious, and the flies occurred both on it and on the longer grass near, Fig. 2. Experimental net and its furniture, as used for G. brevipalpis and G. paillidipes. ~~ a a aes als | J A PLATE XVI. RESEARCH. VoL. XI. PART 4. ENT. BULL. uncleared ; Undergrowth G, brevipalpis present in abundance. Clearing experiment No. 1. Ile Fig. ; G. brevipalpis has entirely disappeared. The same spot with the undergrowth cleared Fig. 2. Butt. ENT. RESEARCH. Vot. XI. PART 4. PLaTE XVII. Fig. 1. Coppice type of secondary bush, strongly favoured by G. pallidipes west of the Sitatonga Hills. This bush has grown up in 12 years on deserted native gardens. Fig. 2. A thicket on a piece of land belonging to the writer resulting from early and irregular burning for a good many years; it was previously clean-stemmed wooding. But. Fees: R SO —— =~ ae Explanati ostly sparse and low, bush-savannah, ‘*Umpoza” dominating; fly Breen areas.- ther poor but + continuous Brachystegia savannah forest with ma ans and, in places, a few brevipalpis. posure— sandstone, shale quartzite and good Brachystegia—Uapa rs. for the most part highly deciduous; permanent fly absent. wooding chiefly of the various dense-secondary types, especially fer a brevipalpis) more localised then than in the rose and dark-gre nd Umtshanedzi. - ts. Wooding contains an unusually large element of the primary ( Pteleopsis. Much Landolphia kirkii in North. -Brevipalpis appare Tleleopsis Pangolpnia Kirkil Brevipalpis ~—~~~~~ —Route of two native expeditions. pproximately, but distances (cyclometer and timing) probably nearl nce. ps. s.f. = secondary forest. f.p. = primary forest-patch o Portuguese frontier and numbered beacons. _—_ New kraal below Chipungambira forest. F. = Ferry and kraals of : From ‘1" north-eastwards brevipalpis was present almost conti ojd *‘5"' flies (pallidipes) seen on June 13th, 1918. ‘6° a combi &/uzi. 8" brevipalpis, do, +. “9” Jacks high-elevation pallidip e, 1900. “12°' and ‘*13"’ most westerly and southerly morsitans, idipes than Jack's were taken on the Inyamadzi. le-keeping under Umzila and Gungunyana; also, roughly, the line 2-keeping now pushed back to the old boundary of the pre-Umzila Alleged qe es. @ e__ Recent nagana outbreaks. Fly. matic and gives merely a fairly correct general idea of the prese map and the portions (taken by compass from frontier beacons) o ply fairly correct also the windings of the Buzi from Itamba northw and the positions of some Kafir kraals and river-crossings. O06), showed low country to S.E. covered densely with trees, stated parse. : Burt. Ent. Researce. Vou. X1., Part +- Chimanimani Massif (Peaks over 8000 feet) * é i Y ab aE ———=—- SAMBA 2" here, Je J os reported pa Pticulary ven roe torent N eae Ps és} oe (a, MUCHAMBA y Wee” Mou 5 rath cs wh ¥ — no se \5 m. To "Buzi Col “ nt Singwekwe : F: INYAMSIZWA_7,ru R- a Go BH ence’ saihing boat te Heine or Pene, 6000 ft. 3 Mee zs pe ; lg et CHIBUYE ‘ \ _7. \Chirenda : Z Y, : e GUDZA NAY io Seatarel SNE se? fram eh S : 7 ‘rom Chironda for a doz 1 Mag patcres SPS > Honecat lrren progeaten Seopa uareR arta . fiers! busts) 3 Scattered thickets (largely Pavetta galigna). Further : es vast more open, but with bigger trees (still open) 7 reported here Be near river. Approaching Imyamita (196 ny. from Guzt Colonial) more open still, BECC Pe scattered palms ang Map to Illustrate REPORT ON A STUDY of the J : ye SS Marowery. H Hh = Makwakwa Y ES YEW: e s> NORTH MOSSURISE A Moyet (and elder Gunye wae 3Gurdoda : §Mangunde 7) Mapapa vieis, etc. 1 TSETSE PROBLEM. 2 Chibabava - Subse “%, A Ce ‘\ Baayaberive rarer fe “i Ms To Arucate and u R@asanande net Madanda forests. " MO and snene Let iew! ane ger! BY pas: “WIPE ance, “HOS ee \ pare Bont mwwrncaNe,, BE caSe Sec ie case U more prominent points of a somewhat broken ° 5 f - es] Mainly basalt; wooding mostly sparse and low, bush-savannah, “*Umpoza” dominating; fly pi pes and morsitans, apparently chiefly as overflow from the salmon and 7 7 Pole fkrasl” . 3 1 PS wo (pallidipes) dark-green areas. (ae ficest piece of country ; MSE a Pp: s pi é : heh ; , a F eu OS™ Ot. Dos =] ainly granite-gneiss; rather poor but + continuous Brachystegia savannah forest with many vieis. Spunfebiea® be 5 oe Soe cae ranhe ~ FAO » Fly pallidipes and morsitans and, in places, a few brevipalpis. u ‘ YN owana * eo Fi 5 Dengedza f.p- (6) BLONG Rr > We BY eo pene A great sedimentary exposure—sandstone, shale quartzite and good Srachystegia—Uapaca savannah forest, continuous brevipalpis everywhere in the thickets, pallidipes 4 / with ari aecos small hillS) d Playyena ach srallerinura Bere, cs f ee | Mainly dolerite; wooding for the most part highly deciduous; permanent fly absent. <4 ae BA Mainly dolerite; broken wooding chiefly of the various dense-secondary types, especially fern-entwined. In the main deciduous, but late in the season, and the permanent a27 ZA fly (pallidipes ana brevipalpis) more localised then than in the rose and dark-green areas, Brachystegia (and fly) on scattered Sedimentary exposures, Ramturartesy ©, between) Puizise and) Umtshanedzi. a a Mainly dolerite and schists. Wooding contains an unusually large element of the primary (chiefly ravine”) and the finer secondary “monsoon” types « Khaya, Piptadenia TET Bon AU IEEh any Acacia caftras tn Albizia, Milletia, Pteleopsis. Much Landolphia kirkii in North. -Brevipalpis apparently abundant throughout, pallidipes present. eee i & reticulata, Albizzia hypoleuca Kigeli Bik EN not dangerous all the year an een Route of main expedition. ~ -————— — Route of two native expeditions. wee’ © . . « ees 28 POG: a Buzi route: shown very approximately, but distances (cyclometer and timing) probably nearly correct. The notes appended to it were made in 1906-7 and the wooding may ey iter RO ogee have increased since. Four main working camps. s.f. = secondary forest, fp. = primary forest-patch 1. = G. austeni vleis, 7 ao . o—~o—o 1 ADEE rah oo e806 gh” Anglo Portuguese frontier and numbered beacons. “——_ _ Buffalo passage areas, routes and approaches. U, = Umgazaza’s kraal, ZZ; oe Soo Corse oy 2 Ma. = Mapondo's. N. — New kraal below Chipungambira forest. F. = Ferry and kraals of Siyaghiya anal Shilling. wis u 7 (Cn s Ts ee bi ASebosher™ord © O.- Flies taken furthest west. From ‘1" north-eastwards brevipalpis was present almost continuously in numbers, "2" — a fly taken by Mr, Jack. ‘3" by Dr. Lawrence, pushy.an If UmbIya Ne. ee ae August. “4" and “5” flies (pallidipes) seen on June 13th, 1918. “'6" a combined attack by pallidipes and brevipalpis, 7 = brevipalpis continuous in thickets . i} ater from here to Buzi * brevipalpis, do. +. “9” Jacks high-elevation pallidipes, August, 1917 " = brevipalpis dry season limit here. ‘11 Wilder an 25 } 2 Lawrence's tsetse. pT “13 "" most westerly and southerly morsitans, respectively. * = highest elevation brevipalpis, 3000 ft. or somewhat over, “10 i Fan-palms — Hyphane — More westerly pallidipes than Jack's were taken on the Inyamadzi. ‘ yj) become common from id ae | here eastward, throughout. Eastern boundary of cattle keeping under Umzila and Gungunyana ; also, roughly, the line of maximum clearing in the Puizise-Mtshanedzi area. oGungunyana’s, 2 capita Kay great numbers — LE ver ~~ Much Bauhinia 99S> ~ © AS HLIT = = =====- Recent boundary of cattle-keeping now pushed back to the old boundary of the pre-Umzila period vane cS ZINYUMBO er Hot Diplorhynchus (Miatawaypeonumon— in paris Note___. The contouring is diagrammatic and gives merely a fairly correct general idea of the presence of hill-masses and their relative importance. The frontier is taken from the Brinn gmecmian i fran sin unos Ss faontoiChimbiys official Rhodesian map and the portions (taken by compass from frontier beacons) of the Sitatongas, higher points of the oblong, Umtarene and the Spungabera details are probably fairly correct also the windings of the Buzi from ltamba northwards and, in a more general way, the course of the Umtshanedzi and the entries of the tributaries and the positions of some Kafir kraals and river-crossings- Alleged . Some definite Zulu failures. @ e-_ Recent nagana outbreaks. _ View from here (Nov, 1906), showed low country to S.E. covered densely with trees, stated by natives to constitute Maguasha," or dense forest, that along Buzi route was seen to be very Sparse. XK Bale & Danielsson 4 387 MOSQUITOS COLLECTED IN PALESTINE AND ADJACENT TERRITORIES. By Capt. P. J. Barraup, F.Z.S., F.E.S. These notes include observations made on the mosquito fauna of Palestine, for the period July 1919 to August 1920. In September and October 1919 a somewhat rapid tour was made through Syria and Cilicia, while from time to time short visits have been made to Egypt. The climate of this part of the Mediterranean region is divided into wet and dry seasons, the latter usually prevailing from May to October, inclusive. In the northern parts of Syria and Cilicia the dry season is of rather shorter duration. From November to April there is a heavy rainfall over the larger part of Palestine and the countries to the north, but in the lower Jordan Valley, and in the deserts of the south, the amount is much less. The winter is mild, especially along the coastal belt and in the lower Jordan region, the thermometer seldom falling to freezing point. In some seasons there may, however, be a considerable snowfall in the mountains. With the advent of summer, the country quickly becomes arid. Rivers and streams are few and insignificant, and the inhabitants of the towns and villages are obliged to rely for the most part upon wells or supplies of rain-water collected in catchment tanks. In certain parts of the country there are, however, perennial streams, and in some cases these form extensive marshy tracts, affording breeding grounds for mosquitos during the greater part of the year. Irrigation is practised throughout the dry season in districts where the supply of water is sufficient from springs or surface wells. The summer weather in the Jordan Valley can only be described as tropical. Along the coastal plain there is usually a sea breeze, but when this fails, the moist heat is often trying. The higher parts of the mountain ranges enjoy a succession of hot sunny days, tempered by cool breezes, with a greater fall in the temperature at night. The maritime plain, consisting for the most part of rich grain lands, orange groves and olive yards, varies in width from a narrow strip where the mountains approach the sea, to many miles where they recede. The mountains of Judea and the Galileean hills, rising here and there to 3,000 feet, form the backbone of Palestine proper. They consist chiefly of barren limestone. To the eastward they descend in tumbled masses to the Jordan Valley, a unique and stupendous crack in the earth’s crust, sinking to 1,200 feet below sea-level. On the far side the mountains of Moab divide the valley from the tablelands of Bashan and the Arabian desert. The larger part of Palestine, especially the central and northern regions, Syria, and Cilicia are malarious. A fairly high percentage of the native population in some of the towns and villages has been found to be infected, and well known malaria-carrying species of Anopheles abound. The duration of the malaria season roughly corresponds to that of the dry period, from May to November. 388 CAPT. P. J. BARRAUD. In early summer the Bedouins, with their flocks, descend to the plains in search of pasture and water. From their habits it has been assumed by some authorities that the majority of these people are malaria-carriers, and in their wanderings assist in the spread of the disease, but I am not aware that this opinion is based upon any extensive investigations. It is possible that these people derive a certain protection from the fact that they are usually accompanied by numbers of the larger domestic animals, horses, donkeys, camels and cattle. I have observed that frequently some of these are stabled under one end of the curiously shaped bivouac tents used by the Bedouins. These conditions appeared to be favourable for the feeding and shelter- ing of Anophelines. Although I was unable to prove that Anopheline mosquitos prefer feeding upon animals rather than man in this part of the world, I have noticed, when searching camps, that they are usually to be found in larger numbers in tents pitched near horse or mule lines. In the towns and villages the number of large domestic animals would not be sufficient to supply food for the abundant Anopheles, and in these places, no doubt, they mostly subsist upon the blood of human beings. One of the most abundant domestic mosquitos of the country is Anopheles bifurcatus. I have seen it breeding in countless thousands in the basement rain- water cisterns, which are to be found almost universally beneath and around private houses. As this stored water often represents the sole supply for the inhabitants during the summer, anti-malarial work is difficult, and opposition is sometimes met with. In Jerusalem there are upwards of 4,000 basement cisterns, besides an extensive system of ancient drains and sewers, many of which have become blocked by subsidences. Some of these cisterns are of considerable size, having a rain-collecting platform as large as a tennis court, with a correspondingly large surface of water beneath. They are from 15 to 20 feet deep, and usually have only one opening about 2 feet square, so that a satisfactory examination is hardly possible. Hven where a lid or cover has been made mosquito-proof, and a pump fitted to draw up the water, there is always an opening through which the rain enters. Mosquitos find their way in and out through this. Large surface wells protected by masonry are another frequent source of trouble. One of the most important medical problems with which the new Administration has to deal is the provision of an adequate piped water supply to all the towns, and the abolition of basement cisterns, surface wells and catchment tanks. In Egypt, where the incidence of malaria is much less, the three chief Anopheline carriers of Palestine, viz. A. maculipennis var., A. bifurcatus and A. swperpictus, are practically unknown. A. multicolor is there considered to be the principal carrier. It is common in the Canal zone, and especially abundant in the oases. It is in the last-named places that malaria is most prevalent. Here again the popula- tion is largely a nomadic one, tending to spread the disease from place to place. It is interesting to note that all the species of Anopheles so far found in Egypt occur also in Palestine, in spite of a wide strip of waterless desert separating the two countries. On the other hand, there are four or five species in Palestine which are absent from Egypt. MOSQUITOS GOLLECTED IN PALESTINE. 389 Anopheles maculipennis, Myg., var. eu This is generally distributed over Palestine, Syria and Cilicia, but although recorded from Egypt by earlier workers, has not been found there in recent years. The adult has the wings less distinctly spotted, especially in the male, than the European form,* some specimens being unspotted. The resting position on a vertical surface is rather flat, this being more noticeable in the male. Gorged females resting on a tent roof hang down in the usual Anopheline position. Major S. R. Christophers discovered this faintly spotted form in Mesopotamia and wrote to me in regard to it, mentioning that the egg differs from that of the typical A. maculipennis (figured by Nuttall and Shipley in Journal of Hygiene, i, pl. 1, figs. 1 and 2) in that it has a frill of air cells all round the edge, instead of one pair of lateral floats. This observation I have been able to confirm in the eggs of Palestine speci- mens. The full-grown larva and the adult do not, however, appear to differ structu- rally from the type, and it remains to be seen whether the variety here referred to should be regarded as a distinct species. The larvae are usually to be found in natural water both fresh and brackish. The adults have been found in very large numbers in tents and huts in camps, in various parts of the country, and often cause a heavy incidence of malaria. They will sometimes travel a considerable distance from the breeding-grounds, especially when emerging in large numbers. On 27th September 1919 I visited a camp at Toprak Kali, Cilicia, which was situated on a hill 500 feet high, and one and a half miles from the nearest water. In the earlier part of the year this site had been considered safe from a malaria point of view. The tents were, however, found to be heavily infested with both A. maculipennis var. and A. superpictus. Very similar observations were made on Ist October at a camp in the valley of the Ak river, near Marash, Northern Syria, from which large numbers of cases of malaria were being evacuated. In the first tent examined (square Indian pattern) there were over 300 Anopheles in one corner, while a patient lying beneath had 70 inside his mosquito net. These were nearly all A. maculipenms var., with a few A. supermctus ; some of the former were noticed to be biting in the daytime. Although the river passed within a few hundred yards, very few larvae, or likely breeding-places, could be found. In marshy ground and rice fields at about one and a quarter miles distance, however, the larvae were in abundance. Breeding commences in April and May, according to the season, in the marshy areas along the coastal belt of Palestine. In the Jordan Valley, at depressions below sea-level, it probably begins earlier. During the first part of May 1920, I collected large numbers of larvae and pupae in a shallow brackish marsh lying amongst the sandhills, east of Haifa, and near the mouth of the Kishon river. Larvae of several other species, viz. A. multicolor, A. hyrcanus, A. mauritianus, Culex univittatus and C. tepuliformis, were found at the same time. By the end of May larvae were much less numerous, the day temperature of the water in the marsh having then risen to 88-90°F. *The North European form has not been found in Palestine, but the two forms occur together in Macedonia. 390 CAPT. P. J. BARRAUD. Adults of A. maculipennis var. were found in camps on the sands near the Kishon, and a few from time to time in camps above Haifa town at 300 feet ; also in the lower western part of the town, at least one and a half miles from the nearest breeding place, as far as I could discover. Larvae were never found within the town limits, either in covered or uncovered collections of water. At the end of April larvae began to appear in an area flooded by late rains, north of Acre town, in the neighbourhood of a battalion camp. This area soon became dry, but in the meantime large numbers of A. maculipennis var. had hatched out, and an outbreak of malaria occurred, lasting several months. The first primary case of benign tertian was reported on 28th May and of malignant tertian on 27th June. The species occurs from 1,200 feet below sea-level to 2,300 feet above. My collec- tion includes specimens from the following places :— . PALESTINE : Sarona (near Jaffa) ; Zummarin ; Athlt ; Haifa; Acre; El Afule ; Jenin ; Tabgha (north of Tiberias), along the course of the Jordan from the Sea of Galilee to near the Dead Sea. Syria: Damascus; Homs; River Ak Valley (near Marash). Crticta: Missis ; Jihan ; Toprak Kali. Anopheles bifurcatus, L. As mentioned previously, this is the most abundant domestic Anopheline mosquito of Palestine and Syria, breeding almost exclusively in basement rain-water cisterns and covered surface wells. I did not discover the larvae in any natural open water in Palestine, but at Beirut (Syria) larvae and pupae were taken from a small gently flowing stream containing much vegetation and over-grown with bushes. It appears to be essentially a cool-water species. The temperature of the water in basement cisterns in a coastal town, such as Haifa, varies between about 58°F. in winter and 78°F. in summer. During the spring and early summer, when breeding is most active, the water was found to remain at about 62°F. Larvae may be found through- out the year in their favourite haunts, but feed up more slowly in winter than during the warmer months. The majority of larvae must pass their whole existence in semi-darkness, the sun’s rays seldom penetrating to these deep cisterns. Examinations were made at Haifa in N ovember, before the autumn rains, when many of the cisterns were dry or contained very little water. In most cases they were clean, with very little, if any, debris upon the floor. In others, which had probably not been cleaned out for some years, the walls had become green, and there was a thicker floor deposit. Further investigations in J anuary, after about six weeks of heavy rains, revealed the presence of very large numbers of adults. Collec- tions were made of individuals resting on the walls, within reach of the opening. At first I was under the impression that they were all females, but while carrying out a series of fumigation experiments, it was found that in cases where all the mosquitos had not been killed by the fumes, most of the survivors were males. These were resting or flying about near the openings when the cisterns were examined. On several occasions in January I was bitten by females disturbed from cisterns. Examinations were continued in April in parts of Haifa not previously visited. Larvae were found to be present in nearly every case, together with adults. Usually some larvae were obtained in the first bucket of water drawn. After that it was MOSQUITOS COLLECTED IN PALESTINE. 391 difficult to secure many more, even though in one case twenty buckets of water were examined in succession. When disturbed they evidently move away from the vicinity of the opening. As far as I could judge, from a number of observations made, the larvae of A. bifurcatus living in these situations are chiefly surface feeders, existing upon debris which may be washed in by the rains. I successfully reared adults from the egg by keeping the larvae in a bowl of clean water and feeding them upon a diet of chopped flies. A few young larvae found on 18th November fed up slowly in the same way, living for ebout two months before pupating. The adults prefer cool and dark places in which to rest during the day, and were not found in any abundance in houses or tents. Except for occasional individuals the species is absent from Egypt. Specimens. were obtained in the following localities from sea-level to about 3,000 feet :-— PALESTINE : Jerusalem; Bethlehem; Jaffa; Haifa; Acre; Nazareth. Syria: Baalbek ; Zahle ; Beirut. Anopheles algeriensis, Theo. I think this must be the species referred to by previous workers in Palestine as A, fragilis, Theo., or A. aitkent, James. It does not appear to be very general, and I have no records for Egypt or Cilicia. One female was captured on 3rd December 1919 in a tent pitched in the marshes east of Acre; this had recently sucked blood. On 11th and 12th August 1920 it was found in numbers at Wadi Selhab, a marshy tract about three miles south of Jenin on the Nablus road and over twenty miles from the coast. The females were biting freely at sunset, and some fifty specimens were caught in half an hour. The only other mosquitos seen at the same time were A. hyrcanus and Culex pipiens (one). It would appear to breed chiefly in the larger marshes, away from human habita- tions, like A. hyrcanus, and is therefore probably not a frequent carrier of malaria. It has been found in the marshy areas of the Auja valley, between Mulebbis and Jaffa. A few specimens were bred from larvae found in a small marsh on the outskirts of Beirut, in September 1919. Anopheles hyrcanus, Pall. (sinensis, Wied. ; pseudopictus, Grassi). As mentioned above, this species breeds chiefly in the larger marshes and is to be found along the coastal belt of Palestine, at least as far north as Acre. It also occurs in suitable localities inland. I think there is no doubt that it is to be found in Syria and northwards, but I did not meet with it. I believe that up to the present it has’ not been found in Egypt. The adults may be found nearly all the year round, but in larger numbers in winter and spring. PALEsTINE: Marshes in the Auja Valley, near Mulebbis ; Athlit ; the Kishon Plain, from Haifa to Acre ; Wadi Selhab. near Jenin. 392 CAPT. P. J. BARRAUD. Anopheles mauritianus, Grandpré. This is one of the Anopheles which is to be found regularly in Palestine and Feypt. In the former country it breeds in much the same environment as A. hyrcanus, viz. the swamps of the Auja and Kishon rivers. I have not met with it inland, or in Syria. In Egypt it is found chiefly in the lake margins and swamps of the Delta. The adults appear to be most common in the early part of the year. Anopheles superpictus, Grassi (palestinensis, Theo. ; nurse, Theo.). Generally distributed over Palestine and Syria, and met with in Cilicia. The larvae are usually found in natural water, and prefer clear pools, near springs of fresh: water. The adults are no doubt responsible for a great deal of the prevailing malaria in country districts, as distinct from towns. My collection includes specimens from the following localities, from 1,100 feet below sea-level to about 3,000 feet above :— PALESTINE: Sarona (near Jafla); Ludd, Latron, Ram Alla (near Jerusalem) ; along the course of the river Auja from Rasel Ain to Mulebbis ; Beisan ; Tabgha ; at many places along the Jordan from Beirut Yakub bridge, on the Upper Jordan, down to Ghoranyieh (near the Dead Sea). Syria: Ak river, near Marash ; Zahle, Lebanon Mountains. Criicta: Bozanti; Toprak Kali, Taurus Mountains. Anopheles culicifacies, Giles, var. sergenti, Theo. This mosquito was not discovered until the end of my stay, and I think it is probably often overlooked amongst numbers of the preceding. A few females were found in tents at Tabgha, on 15th August 1920. Previously a few specimens were bred from larvae found in pools along the Nazareth road, near Haifa. Anopheles multicolor, Camb. (chaudoyei, Theo.). This insect has been previously known to workers in Palestine as A. turkhudi, Liston. It is now found to be distinct from the Indian species. The larvae are often found in large numbers in brackish marshes along the coastal plain of Palestine, in late autumn, spring and early summer. It is not, however, confined to the coast, and occurs in salty pools and streams in the Jordan Valley. As mentioned on a previous page, it is considered to be the chief malaria-carrier in Egypt, where it has a wide distribution. Anopheles pharoensis, Theo. Although abundant in some parts of Egypt, it is regarded as a doubtful carrier of malaria. In October 1919 I found it commonly at Kantara, on both sides of the Canal, and also in the northern suburbs of Cairo. It is very rare in Palestine, and T obtained only one female. This was found at Tabgha, north of Tiberias, on 15th August 1920. I have no records for Syria. Stegomyia fasciata, F. This is exclusively a domestic species, found in most parts of Egypt, Palestine and Syria, at least as far north and east as Aleppo. — ve MOSQUITOS COLLECTED IN PALESTINE. 393 Larvae were found in water receptacles of all descriptions, including flower-vases in a sitting-room, fire-buckets containing clean water, and a disused boiler in a kitchen. . The adults are most troublesome in the hot season, and are suspected of being carriers of the organism causing “Aleppo sores” and “Jericho boils.” Some of the Palestine specimens are very beautiful, having the dorsal surface of the abdomen almost completely covered with silvery scales. PALESTINE: Jaffa; Jerusalem; Haifa; Jenin; Tiberias. Syria: Beirut ; Damascus ; Aleppo. Ochlerotatus caspius, Pall. (dorsalis, Theo., nec Mg.) A very abundant and widely distributed species in Egypt, but so far as my observations go, it is not so common in Palestine. The larvae prefer brackish water. A few adults were found at Haifa in June 1919, The night of 31st March 1920 was spent at Jisr 2l Damie, Jordan Valley. During the evening we were attacked by clouds of mosquitos, but owing to an unfortunate accident to my supply of tubes, I was unable to bring back a large number of speci- mens. A few of this species were secured, together with O. detritus, Theobaldva annulata var., and Anopheles superpictus. Ochlerotatus mariae, Serg. One of the few mosquitos to be found breeding in sea-water. Large numbers of ova, larvae and pupae were found in pools above high water mark on the rocks at Athlit, Palestine, in August 1920. In June and July 1919 larvae were abundant in similar situations along the coast to the south-west of Beirut, Syria. The pools become very salt owing to evaporation, and the temperature of the water, by exposure to the sun, is usually well over 90° F. for the greater part of the day. The adults bite freely in the daytime, and cause much annoyance to people bathing from the shore. Ochlerotatus detritus, Hal. (salinus, Fic.). A very abundant and troublesome species in parts of Palestine during the wet season and spring. The larvae are found chiefly in brackish water, near the coast, but I also found them in large numbers in the Jordan valley at the end of March 1920, at 1,000 feet below sea-level. PALESTINE: Zummarin; Haifa; Acre; Jisr el Damie, Jordan. Culex pipiens, L. Abundant and widely distributed in Egypt, Palestine and Syria, and found as a rule near human dwellings. The species occurs from sea-level to 4,500 feet. 394 CAPT. P. J. BARRAUD. The larvae differ in some respects from Western European and Mesopotamian specimens, and appear to represent a distinct race. No diflerences can, however, be discovered in the adults. Larvae collected in Jerusalem, Haifa and Baalbek (Syria), are very uniform in structure and size and show the following obvious differences from typical specimens :—Full-grown larva smaller, with distinctly shorter siphon ; slightly smaller average number of pecten teeth ; antennae light, instead of dark. In one case larvae were found living in a well in which the water- level was 85 feet below the ground. The adults exhibit a certain amount of variation in the abdominal banding, in some specimens the white bands being incomplete dorsally. [Culex fatigans, Wied. | Although this mosquito was recorded from Ezypt by several of the earlier workers, it does not seem to be found there. An examination of the genitalia of all males in the collection of the Egyptian Ministry of Agriculture has proved them to be C. pipiens, I have been unable to find the species in Palestine or Syria. Culex laticinctus, Edw. The larvae may sometimes be found in very large numbers in covered surface wells, tanks, and similar situations in Palestine. The adults have been observed in quantities in tents in camps. In Egypt it is found commonly at Alexandria. PALESTINE: Jerusalem ; Wadi Hamish; Haifa. Syria : Baalbek. Culex univittatus, Theo. Distributed from Egypt to Syria, but never m any great abundance. The larvae live in marshes and stream pools, generally away from dwellings. One is lable to overlook them, as, when disturbed, they immediately leave the surface, returning for only a second or two at long intervals. There still seems to be some confusion regarding the larval characteristics. My specimens disagree in almost every particular with the description and figure given in Bull. Ent. Res. x, p. 69, 1919; neither do they fit in with the synoptic table given in the same journal vol. i, p. 381, Dec. 1912. I have handed some of my material to Mr. F. W. Edwards, who has kindly offered to go into the matter, and. publish the results in a future paper. PALESTINE: Wadi Hamish; Sarona; Haifa. Syria: Beirut. Culex tipuliformis, Theo. Found in similar situations to the preceding. It has not, so far as | am aware, been recorded from Egypt under this name, but there is a specimen from the Fayum. in the British Museum collection. PALESTINE : El Afule; Acre; Haifa. Syria : Damascus. MOSQUITOS COLLECTED IN PALESTINE. 395 Culex hortensis, Fic. Although I was on the look out for this species in Palestine, very few were found. Some females were taken at Haifa in May 1920. No doubt it is common enough in some parts of the country. At Damascus, in July 1919, larvae of this and of C. tupuliformis were dipped from the same pool. It appears to be absent from Egypt. Culex mimeticus, Noé. Not uncommon in suitable localities in Palestine, Syria and Cilicia, but absent from Egypt. The larvae have been found associated with those of A. swperpictus. PALESTINE: Wadi Hamish. Syria: Beirut. CrLicta: Bozanti. (Culex modestus, ['ic. | This was not met with, but is likely to occur, since it has been found in Macedonia and Egypt. Theobaldia longiareolata, Macq. One of the commonest gnats in Egypt, Palestine and Syria. It is so universal that it is hardly necessary to give localities. This and the following species were found up to 4,500 feet in the Lebanon Mountains. The larvae abound in water-barrels and other receptacles in gardens, surface wells, and covered and uncovered tanks and cisterns. Theobaldia annulata, Schrank. Much less common than the last-named, and absent from Egypt. The strongly marked type form was taken at Ramleh, Palestine, and at Ain Sofar, Lebanon Mountains. In the Jordan Valley at 1,000 feet below sea-level, I found some specimens of a pale variety which agree with the Mesopotamian form, referred to in my paper on mosquitos of that region (Bull. Ent. Res. x, p. 325, 1920). Theobaldia morsitans, Theo. At the end of April 1920 some larvae were found in a pool at the bottom of a small quarry at Haifa. The species was not met with elsewhere. [Taeniorhynchus richiardii, Fic. | This species has been found in Palestine, at Jell Zahmfil, Huleh, by Dr. J. Cropper, but was not met with by the author. Uranotaenia unguiculata, Edw. Widely distributed in Egypt, and found in parts of Palestine. The larvae as a tule are only to be found in the larger swamps. My thanks are again due to Dr. Guy A. K. Marshall, Director of the Imperial Bureau of Entomology, and Mr. F. W. Edwards, of the British Museum, for assistance in the identification and synonymy of certain of the species and forms mentioned above. Also to Mr. G. Storey, of the Egyptian Ministry of Agriculture, for informa- tion as to the distribution of mosquitos in Egypt. (737) F 8K fh ake tae ye ia (as) a oe ; 7 Kore eit . re ee ee eee vt Hey vopibe seo Th ‘ed ‘aad a dni dts f <7 PAM “i "hi 1. OEE ft Wy ‘in io ‘ ¥ i 1 r ‘ f fi} i H y ni ats frei te dal AF Ge edutone ty ‘ + r NX f ' ait , q es en, rin Reagan | if as ute m4 art waits | Nes copay tap. i a 4] ve nly Cintect Ay; “yt, é) i Palle CULE: Pgegit pee ; ck ayy in hae uh, y goth vv eal hg nae ae herbert as ee) f " if ¥E), ayaa CE he ti bapaues= i one oH ahaa ig a & eet ‘Midis cud » pallidus, sp. n., on fig in Transvaal, 33. > quadrilineatus var. sim- plex, n., on Rhus in Cape Colony, 33. >> rusci, On quince in Cape Colony, 34. i tachardiaformis, sp. D., on Elytropappus rhinocero- tis in Cape Colony, 35. = zonatus, on Acacia in Transvaal, 35. Ceroplastodes, 2, 40. a bituberculatus, sp. n., in Cape Colony, 40. Ceylon, Lonchaea in, 209; Aphelinus mytilaspidis parasitisng Parlatoria blanchardi in, 297 ; Icerya aegyptiaca infesting dates in, 299. chaudoyei, Anopheles (see A. multi- color). GENERAL INDEX. Chermes anthurii (see Saissetia hemis- phaerica). caraciae (see Ceroplastes rusct).. + hesperidum (see Lecanium). - laurt (see Lecanium hesperi- dum). nd oleae (see Saissetia). chilianthi, Filippia. Chilocorus cacti, predaceous on Parla- toria blanchardi in America, 297. Chionaspis graminis, parasitised by Aphelinus mytilaspidis in Ceylon, 297. “ permutans, parasitised by Aphelinus mytilaspidis in Ceylon, 297. x pinifolii, parasitised by Aphelinus mytilaspidis in U.S.A., 297. “A simplex, on bamboo in Mauritius, 173. chirindensis, Systates. chrysanthemi, Capitophorus. Chrysomphalus aonidum, on date palm. in California, 299. ‘5 aurantvi, on date palm. in California, 299. Chrysops, in North Russia, 196, 197. 2 caecutiens, in North Russia,. 197, 198. 3 divaricata in North Russia, 198. 3 nigripes, in North Russia, 198. Pe relicta, in North Russia, 198. ds sepulcralis, in North Russia, 198. Chrysophyllum flavum, relation of Glossina to, in 8.E. Africa, 319. chrysostigma, Phara (Pangonia). Cilicia, mosquitos in, 390, 392, 395. Cinnamomum zeylanicum, Coccids on,. in Mauritius, 174. cisst, Macrosiphum. Cissus, new Aphids on, 66. cttricola, Lonchaea (Carpolonchaea). Citrus aurantium (Orange), mosquito: larvae breeding in, in Lagos, 257. clypeatum, Lecanium (see Sarssetia hemisphaerica). Coccidae, South African, 1-41. GENERAL INDEX. Coccinella abdominalis, predaceous on Parlatoria blanchardi in America, 297. Cocos nucifera (Coconut), mosquito larvae breeding in, in Lagos, 257. Coccotrypes dactyliperda, in dates in Egypt, 294. Coccus artemisiae (see Ceroplastes rusci). » caricae (see Ceroplastes rusct). » coffeae (see Saissetia hemis- phaerica). » cycadis (see Saissetia oleae). » floccifera (see Pulvinaria). » hesperidum (see Lecanium). » myricae (see Ceroplastes). » oleae (see Saissetia). » palmae (see Saissetia oleae). » patelliformis (see Lecanium hesperidum). » rusci (see Ceroplastes). cockerelli, Pheidole. coffeae, Lecanium (Coccus) (see Saissetia hemisphaerica). cohaesa, Anopheles minimus. columbaczense, Simulium. Columnea myricae (see Ceroplastes). : testudinata (see Ceroplastes TUSCt). - testudiniformis (see Cero- plastes rusci). combreti, Ceroplastes. Combretum, in relation to distribution of Glossina in S.E. Africa, 320, 345, 349; Coccid on, in Transvaal, 27. cometes, Lachnosterna (Phytalus). . condei, Hretmopodites (see EH. quinque- vittatus). consimilis, Culex. continentalis, Lonchaea. Conofilippia, gen. n., 2, 25. subterranea, sp. 0., in Transvaal, 25. in relation to im toch. Conopharyngia spp., distribution of Glossina Africa, 319, 348. Corchorus olitorius, food-plant of thrips in Sudan, 98. Cordia interrupta, food-plant of Tiphia parallela in Mauritius, 164. (Myzomyia) 417 Corizoneura, systematic position of, 1 s aethiopica, Synonymy of, 139. _ albifacies (see Buplezx). ie dissimilis (see Buplex). 7 distincta, 140. formosa, sp. n., in Cape Colony, 142. 33 (Diatomineura) hastata, 140. ” % amornata, 140,145, 146,147.. 9 (Pangonia) lateralis, 140. ns (Diatomineura) lineati- thorax, 140. 35 (Pangonia) longirostris, 140. a (Diatomineura) neavet, 140. Ks pallidipennis, 140. 99 (Diatomineura) pene- trabilis, 140. » (Pangonia) sagittaria, 140 9 (Pangonia) taprobanes, 140. s schwetzi, sp. n., in Central Africa, 143, 146, 147. a umbratipennis, 140. 99 (Diatomineura) virgata, 140. corniger, Culex. costalis, Anopheles. Cotton Thrips (see Heliothrips indicus). Crabs, mosquitos breeding in holes of, in Lagos, 247, 248. crassicornis, Haematopota. crassipalpis, Phara (Pangonia). crassiusculus, Dociostaurus. in S.E. Craterispermum laurinum, Africa, 320. Crateropus, in 8.E. Africa, 365, 368. s kirki, possibly destroying Glossina pupae in §.E. Africa, 369. Cremastogaster scutellaris, associated with Aphids in Persia, 156. Crinum giganteum, mosquito larvae breeding in, in Lagos, 257. crocodilinus, Tabanus. 418 Crotalaria, GENERAL INDEX, food-plant of thrips in Sudan, 98. erucigerus, Dociostaurus. Oryptinglisia, 2, 38. lounsburyi, on Vitis in Cape Colony, 38. Oucumis melo, food-plant of thrips in Sudan, 98. Culex, breeding among water lettuce, 3° 39 3° 3° 39 39 3° - he) 3) 39 9° 3°) 39 3° 3) 3) >> 74; larvae of, attacked by Inspa, 279-281. annulioris, 136. annulitarsis, 136. anxifer (see C. fatigans). arboricollis (see Orthopodomyia). argenteopunctatus, in Mada- gascar, 137. ataeniatus, breeding places of, in Lagos, 264, 265, 269. (Aédes) cantator, breeding in salt water in America, 261. cartront (see C. fatigans). corniger, breeding in crab-holes in Brazil, 254. consimilis, breeding places of, in Lagos, 253-269. decens, 137; early stages of, in West Africa, 105, 251-269; pupa of, 106, 107, 108. duttoni, breeding places of, in Lagos, 259-269. extricator, in Tropical America, 254. fatigans, in Madagascar and Mauritius, 137; breeding places of, in Lagos and Gold Coast, 253-269; probably not occurring in Egypt, 394. fluviatilis (see Aédes). flavus (see Taeniorhynchus). fowleri (see Ochlerotatus nigeri- ensis). giganteus, in Madagascar, 136. grahami, breeding places of, in Lagos, 259-269. grandidieri (see Taeniorhynchus flavus). guiarti, breeding places of, in Lagos, 264, 265, 269. hortensis, in Palestine, 395. insatiabilis (see Stegomyia fasciata). Culex insignis, breeding places of, in 99 Lagos, 251-253, 264-269. invidiosus (see C. decens). laticinctus, in Palestine and Egypt, 394. laurenti, in Madagascar, 137. lividocostalis (see C. decens). mimeticus, in Palestine and Syria, 395. modestus, probably occurring in Palestine, 395. neireta (see C. giganteus). nigrocostalis (see C. decens). pipiens, in Madagascar, 137 ; in Palestine, 391, 393, 394. pruina, in Lagos, 268, 269. quasigelidus, in West Africa, 74, 268, 269 ; in Madagascar, 136. reticulatus, in Brazil, 254. rima, breeding places of, in Lagos, 251-254, 268, 269. ronaldi (see C. sittens). salinarius, breeding in salt water in America, 261. salisburiensis, breeding places of, in Lagos, 251-253, 268, 269. sitiens, breeding places of, in Lagos, 260, 261, 269; in Mauritius, 137. sollicitans, breeding in salt water in America and India, 261. taeniorhynchus, breeding in salt water in America, 254, 261. thalassius, breeding places of, in Lagos, 251-253, 260-268. (Lutzia) tigripes, in Madagascar and Mauritius, 136; breed- ing places of, in Lagos, 259- 269; predaceous larva of, 281. wunivittatus, in Lagos, 269; in Madagascar, 137; larva of, in Palestine, 389, 394. ventrilloni, sp. n., in Madagas- ear, 135. culiciformis, Anopheles (Stethomyia). Culiciomyia nebulosa, breeding places of, in Lagos and Gold Coast, 251—270. GENERAL Culicoides, in North Russia, 197. arcuatus, in North Russia, 39 198, eS fascipennis, in North Russia, 198. 9 festivipennis, in North Russia, 198. pulicaris, in North Russia, 198. cumminsi, Ochlerotatus. curta, Phytomia. cyanopterun, Rhynchium. cyanophylli, Aspidiotus. _eycadis, Coccus (see Saissetia oleae). Cyathomyia fusca, larva of, at Accra, 258. Cycas revoluta, Coccids on, in Mauritius, ia. cylindrica, Lachnosterna. dactyliperda, Coccotrypes. Dacus demmerezi, on Marrows and pumpkins in Mauritius, 172. » sygmoides, in Mauritius, 172. dahlifolii, Macrosiphum. damnosum, Simulium. dantini, Dociostaurus. Dasyops, 199. Dates, pests of, in Mesopotamia, 287— 303. -decens, Culex. decisus, Aulacobothrus. Deinocerites, in Tropical America, 254, WMemmerezi, Dacus. denticulata, Lachnosterna. depressa, Osca (Pangonia). depressum, Lecanium (see Saissetia nigra). -desertorum, Oryctes. destructor, Aspidiotus; Ceroplastes ; Tanymecus. detritus, Ochlerotatus. Dialium guineense (Black Tamarind), mosquito larvae breeding in, in Lagos, 257. Diaspis boisduvali, on Phoenix, 299. carueli, parasitised by Aphe- linus mytilaspidis in U.S.A., 297. phoenicis, on Phoenix, 299. rosae, parasitised by A phelinus mytilaspidis in Spain, 297. 99 INDEX. 419 Diatomineura inornata (seeCorizoneura). Dicrurus afer, predaceous on Glossina in §.E. Africa, 368. dictyospermi, Aspidiotus. Digera arvensis, food-plant of thrips in Sudan, 98. digitata, Aclerda (Pseudolecanium). dilemma, Lachnosterna. dimidiatus, Carpophilus. Diplorhynchus mossambicensis, in §.E. Africa, 320, 366, 379. Dipterocarpus _ turbinatus, shoreae on, in India, 277. discors, Pangonia. Dispholidus, in 8.E. Africa, 379. dissimilis, Buplex (Corizoneura). distincta, Corizoneura. divaricata, Chrysops. Dociostaurus, revision of, 397—407.- se albicornis, characters of, 400, 404. x anatolicus, characters of, 398, 400, 404; dis- tribution of, 405. ¥ crassiusculus, 398; char- acters of, 400, 401; synonomy of, 405; distribution of, 406. rs crucigerus, characters of, 399; synonymy of, 402; distribution of, 403, 404. crucigerus hispanicus, characters of, 399; peculiar to Spain, 403. crucigerus tartarus, 399, 403, 404. dantini, in Morocco, 397, 400, 405. genei, characters of, 400; distribution of, 404. gracilis, probably syno- nymous with D. genei, 397, 398. kervillei, in Persia, 397. kraussi, characters of, 401; distribution of, 406, 407. kurdus, sp. n., char- acters of, 401; dis- tribution of, 406. maroccanus, distribution of, 398, 402; char- acters of, 397, 399; synonymy of, 401, 404, Calandra 99 99 99 420 Dociostaurus plotnikovi, sp. n., charac- ters of, 400; synon- ymy of, 405 ; distribu- tion of, 406. < tartarus, 397. domestica, Musca. domesticus, Ochlerotatus. Dominica, new species of Lachnosterna from, 191. dorsalis, Cerodonta. Dracaena fragrans, mosquito larvae breeding in, in Lagos, 257 ; in relation to distribution of Glossina, 319. MA gazensis, in S.E. Africa, 320. dracaenae, Hretmopodites. drapert, Sphaerococcus. Dryomyza flaveola, seasonal variation or, 214: durbanense, Lecanium. duttoni, Culex. Echidnophaga gallinacea, imported into Mauritius, 172. egbarum, Ceroplastes. Egypt, date pests in, 288, 289, 294; mosquitos in, 156, 392, 393, 394, 395; Dociostaurus gracilis in, 397. ehretiae, Lecanium. Eichornia crassipes (Water Hyacinth), Anopheles breeding amongst, 73. elegans, Oryctes. elongata, Pangonia. elongatum, Lecanium. elytropappt, Inglisia. Elytroteinus, nom. n., 276. sh subtruncatus, on ginger in Fiji and Honolulu, 276. Enterolobium dulcis (Rain Tree), mos- quito larvae breeding in, in Lagos, 257. Ephestia calidella, in dates in Egypt and Mesopotamia, 288. Bs cautella, in dates in Meso- potamia, 288, 289. 3 passulella (see E. cautella). Epomophorus crypturus, in Glossina areas in S.E. Africa, 338. equinum, Simulium. Erephopsis lata (see Osca). GENERAL INDEX. Eretmopodiltes conde (see EF, quin- quevittatus). 9 dracaenae breeding places of, in Sierra Leone, 258. ‘4 inornatus, in Lagos, 269. > quinquevittatus, breeding places of, in Lagos, 264, 265, 270; in Madagascar, 138. Eriodendron orientale (Silk Cotton Tree),. mosquitos breeding in, in Lagos, 256. Eritrea, Lonchaea in, 201, 203. Erythroxylon Africa, 322. eucleae, Ceroplastes. Eugenia, in relation to breeding places. of Glossina morsitans in 8.E. Africa, 364. exaptus, Systates. emarginatum, in §.E. excisa, Lonchaea (Carpolonchaea). expansum, Lecanium. extricator, Culex. fasciata, Stegomyia. fascipennis, Culicoides. fatigans, Culex. feae, Gagrella. | ferrugineus, Rhynchophorus. fervida, Lachnosterna. festivipennis, Culicoides. Ficalbia neireti (see Uranotaenia). ficiperda, Lonchaea (Carpolonchaea). | Ficus (Banyan), mosquito larvae’ breeding in, in Lagos, 257, 258. » subcalearata, in 8.E. Africa, 323. jficus, Aspidiotus. Figs, bionomics of Ocnerogyia amanda: on, in Mesopotamia, 181-186. Fiji, new weevils from, 276. filamentosum, Lecanium. Filaria bancrofti, carried by mosquitos. in Lagos, 255. jfilicum, Lecanium hemisphaericum (see Saissetia). filiformis, Ischnaspis. Filippia, 2, 23. as africana (see F. carissae). Ae carissae, sp. 0., on Carissa grandiflora in Natal, 24. * chilianthi, sp. n., in Natal, 23.. GENERAL INDEX. Fiorinia pellucida, on Phoenix, 299. flava, Myzomyia (see Anopheles wm- maculatus). flaveola, Dryomyza. jflaveolum, Lecanium (see L. hesperidum). flavus, Taeniorhynchus (Culex, Coquil- lettidia). floccifera, Pulvinaria (Coccus). jloccosa, Pulvinaria (see P. floccifera). fluviatilis, Aédes (Culex). Forda natalensis, sp. n., in Natal, 71. h rotunda, 72. Formaldehyde, as a poison for house- flies, 47-63 ; impurities in, 47. Formic Acid, disadvantages of, in formaldehyde, 50. formosa, Corizoneura. fowleri, Culex (see Ochlerotatus nigeri- ensis). fragilis, Anopheles (see A. algeriensis). France, Aphelinus mytilaspidis para- sitic on Parlatoria blanchardi in, 297. fraseri, Uranotaenia bilineata. fryeri, Aédes (Ochlerotatus). Fryi, Stenommatus. Fuirena, in relation to distribution of. Glossina, 326, 345. fuliginosus, Anopheles (Nyssorhynchus). fulvifascia, Osca (Pangonia). funestus, Anopheles. fusca, Cyathomyia. fuscinervis, Buplex. fuscopennatus, Taeniorhynchus. fuscus, Pectinopalpus (see Ouliciomyia nebulosa) ; Thriambeutes. Gagrella feae, an imported pest in Mauritius, 174. Galago crassicaudatus, occurrence of, in Glossina areas in §.E. Africa, 337. gallinacea, Echidnophaga. Game, in relation to Glossina in Africa, 115, 117, 327, 328, 333, 336-339, 342,. 353, 376. gasteralpha, Pulvinaria. Gelechiid, unidentified species of, on date palms in Mesopotamia 290, 303. genet, Dociostaurus (Stauronotus). geniculatus, Ochlerotatus. geranui, Inglisia. 421 giganteus, Culex. glandium, Calandra. Glossina austeni, bionomies of, in S.E. Africa, 329, 330, 348, 349, 350, 351, 366. a brevipalpis, bionomics of, in Nyasaland, 102; bionomics. of, in 8.E. Africa, 315-384. by longipalpis, 339. o morsitans, bionomics of, in S.E. Africa, 315-388; in Nyasaland, 101 ; bionomics of, in 8. Rhodesia, 113-132. pallidipes, bionomics of, in S.E. Africa, 316—339. » palpalis, in Uganda, 45. glossinae, Mutilla. Glugea, parasite of Simulium, 215. Gold Coast, Lonchaea impressifrons in,. 202; mosquito larvae from, 105- 1 gracilis, Dociostaurus. grahami, Culex. graminis, Chionaspis. grandidieri, Culex (see Taeniorhynchus flavus). guiarti, Culex. Guttera edouardi, possibly destroying pupae of Glossina in §.E. Africa, 328, 369. ) Habrobracon kitcheneri, parasite of Ephestia cautella, in Egypt, 289. Haematopota, in North Russia, 196, 197. es crassicornis, in North Russia, 198. ie pluvialis, in North Russia, 198. haemorrhoidalis, Heliothrips. hastata, Oorizoneura. hauensteini, Dociostaurus (Stauronotus) (see D. crassiusculus). Hawaii, new weevils from 276, 277. hebraeus, Polistes. hederae, Aspidiotus. Hedychium coronarium (Ginger), weevil on, in Hawaii, 276. Heliothrips haemorrhoidalis, on date palm in Trinidad, 295. 99 indicus (Cotton Thrips), bionomics and control of, in Sudan, 95-100. 422 Heliotropium europaeum, food-plant of thrips in Sudan, 98. si supinum, food-plant of thrips in Sudan, 98. Hemilecanium, 2, 14. a theobromae, on Huphor- bia and oleander in Cape Colony, 14. hemisphaerica, Saissetia. Heptaphlebomyia argenteopunctata (see Culex). 59 kingi (see Culex argenteopunctatus). 4 montforti (see Culex uniwittatus). hesperidum, Lecanium (Cocecus, Calym- natus). Hexamethylene Tetramine, influence of, in formaldehyde, 48. hibernaculorum, Lecanium (see Sais- setia hemisphaerica). hirtipes, Simulium. hispanicus, Dociostaurus crucigerus. histrio, Oedipoda. Hodgesia sanguinis, in Lagos, 269, 270. Holotrichia, considered a synonym of Lachnosterna, 189. Hopperdozer, new type of, 179. hornigu, Spermatophora. hortensis, Culex. House-flies, formaldehyde as a poison for, 47-63. Hyalopterus arundinis, on Vitis, 66. hydatis, Coccus (Calypticus) (see Cero- plastes rusci). Hylochaerus, possible relation of, to Glossina in 8.E. Africa, 337. Hyphaene ventricosa, puparia of Glossina morsitans found near, in S.E. Africa, 364. Hypocolius ampelinus, injuring dates in Mesopotamia, 301. hyreanus, Anopheles. Icerya aegyptiaca, on date palm in Zanzibar, 299. ay purchasi, on date palm, 299. Idiosaissetia, gen. n., 2, 40. ” peringueyt, sp. ., on grass in 8. Africa, 40. ammaculatus, Anopheles (Myzomyia). ampressifrons, Lonchaea. GENERAL INDEX. inafectatus, Tanymecus. inconspicuosus, Micraedes. indefinitus, Anopheles (Myzomyia) (see A. vagus). India, pests of date palms in, 293, 294, 295; Lonchaea spp. in, 206, 208, 209; Anopheles breeding in salt water in, 261; new weevils trom, 276, 277. indicus, Heliothrips. Inglisia, 2, 26. = elytropappi, sp. n., on Elytro- pappus rhinocerotis in Cape Colony, 36. cs geranii, sp. D., on geranium in Cape Colony, 37. 9 zizyphi, sp. n., on Zizyphus in Transvaal, 37. ingrami, Phlebotomus. insulae-florum, Anopheles (Stethomyia) aitkent. inornata, Bouvierella; Cort.on ura (Diatomineura). inornatus, Hretmopodites. insatiabilis, Culex. insignis, Culex. interpunctella, Plodia. invidiosus, Culex (see C. decens). Ipomoea cordofana, food-plant of thrips in Sudan, 98. irritans, Ochlerotatus. Isaniris ater, sp. n., on Brachystegia in Nyasaland and Rhodesia, 273. Ischnaspis filiformis, on Phoenix, 299. Isoglossa mossambicensis, in 8.E. Africa, o22. italicus, Calliptamus. Italy, Ochlerotatus ornatus in, 259 . Aphelinus mytilaspidis parasitising Parlatoria blanchardi in, 297. jacksom, Pulvinaria. Jamaica, Lachnosterna in, 190, 191. jamaicensis, Lachnosterna. Japan, Nippolachnus piri in, 157. Java, Anopheles in, 77; Lonchaea in, 207, 209. kellyi, Saissetia. kervillei, Dociostaurus (see D. crasstus- culus). GENERAL INDEX. Khaya nyasica (East African Mahog- any), relation of Glossina to, in S.E. Africa, 319, 320, 346, 348, 349, 376. Kigelia pinnata, relation of Glossina to, in S.E. Africa, 348. Killing-bottle, description of new type of, 283-285. kingi, Heptaphlebomyia argenteopunctatus). kitcheneri, Habrobracon. kochi, Anopheles (Cellia). kraussi, Dociostaurus. kurdus, Dociostaurus. (see Culex Lachnosterna, synonyms of, 189. a acinosa, Sp. 0., in Trinidad, 190, 192. a anodentata, in Mexico, 192. eS antiguae, sp. n., in Antigua and Domin- ica, 190, 191. - (Phytalus) cometes, 193. s cylindrica, 190. = denticulata, in Mar- tinique, 192. a dilemma, sp. n., in Trini- dad, 190, 192, 193. a fervida, in Jamaica, 190. is jamaicensis, Sp. N., in Jamaica, 190, 191. e longipilosa, in Mexico, 192. be misteca, in Mexico, 192. Wy montserratensis, Sp. D., in Montserrat, 190, 191. ov (Phytalus) omiltemus, in Mexico, 189. is (Phytalus) platyrrhinus, in Mexico, 189. . setifera, in Central America, 193. ey (Phytalus) smithi, bion- omics and control of, in Mauritius, 159-169, P7E, bie ae’ * Weat Indies, 189. a trinitatis, in ‘Trinidad, 190, 193. ~ Lachnus pyri, on pears in Persia, 156. Laemophloeus sp., infesting stored dates in Mesopotamia, 303. 423 laevis, Calypticus (see Lecanium hespert- dum); Lonchaea (Carpolonchaea). Lagerstroemia regina, mosquito larvae breeding in, in Lagos, 257. Lagos, breeding places of mosquitos in, 247-270. lamberti, Stegomyia (see S. albopicta). lambiana, Lonchaea. Landecrab (see Cardisoma). Landolphia, relation of Glossina to, in §.E. Africa, 317, 320, 322, 333. lata, Osca (Hrephopsis). lataniae, Aspidiotus. lateralis, Corizoneura. laticinctus, Culex. latipes, Simulium. laurenti, Culex. laurt, Chermes peridum). lautissima, Pangonia. Lecaniodiscus fraxinifolia, in relation to distribution of Glossina in S.E. Africa, 349. (see Lecanium hes- Lecanium africanum, on citrus in Natal, 4. 25 alienum (see L. hesperidum). # angustatum (see LL. hes- peridum). at artemisiae (see Ceroplastes TUSC1). "p assimile var. amaryllidis (see L. hesperidum). - beaumontiae (see Saissetia hemisphaerica). és begoniae (see Saissetia nigra). 5 cassinae (see Saissetia oleae). x clypeatum (see Saissetia hemisphaerica). + coffeae (see Saissetia hemi- sphaerica). és depressum (see Saissetia nigra). es durbanense, sp. n., in Natal. ye ehretiae, sp. n., on Ehretia hottentottica in Trans- vaal, 4. " elongatum, on Acacia melan- oxylon in Transvaal, 5. 59 expansum, on Phoenix, 299. = filamentosum, on fig in Cape Colony, 6. 424 Lecanium flaveolum (see DL. hespert- dum). a hemisphaericum (see Sais- setia). 55 (Coccus) hesperidum, on date palms in California, 299; on citrus in Cape Colony, 3; on limes in Mauritius, 173, 174. 9 maculatum (see L. hespert- dum). mangiferae, food-plants of, in Mauritius, 173, 174. minimum (see L. hesperi- dum). nanum (see IL. hesperidum). 2s nigrum (see Saissetia nigra). he oleae (see Saissetia). AS persimile (see Saissetia per- similis). » platycerii (see L. hesperi- dum). a proteae, sp. n., on Protea in Transvaal, 7. 4 pseudelongatum, on Acacia caffra in Transvaal, 6. x pumilum, sp. 1., in Cape Colony, 5 re signiferum (see L. hesperi- dum). By subpatelliforme (see Sais- setia). se sylvestru (see Saissetia per- similis). Bs terminaliae (see L. hes- peridum). ™ tessellatum, on Phoenix, 299. x ventrale (see L. hesperidum). b> wistariae, Sp. D., on Wis- taria in Cape Colony, 8 Lemna, Anopheles breeding amongst, 74. lepida, Pulvinaria. Lepidosaphes ulmi, parasitised by Aphelinus mytilaspidis in U.S.A., 297. Leucas nubica, food-plant of thrips in Sudan, 98. Jleucosphyrus, mya). Leva apicalis, synonymy of, 398. » mundus, synonymy of, 398. Lichtensia asparagi, sp. n., on Aspara- gus capensis in Cape Colony, 23. Anopheles (Neomyzo- GENERAL INDEX. lineatithoraz, Corizoneura. lineatopennis, Banksinella. Lispa, bionomics of, in Nyasaland, 279-281. lividocostalis, Culex. lloydi, Villa. Lonchaea, key to Ethiopian species of, 199, 200; key to Indo- Australian species of, 204; key to Oriental females of, 210. » (Lamprolonchaea) aurea, wide distribution of, 200, 202, 204, 206. 7” (Carpolonchaea) atratula, in Malaya and Philippines, 205,207. ‘3 biroi, probably a synonym of L. atratula, 205, 207. 9 (Carpolonchaza) bisulcata, sp. n., in India, 206, 208, 210. mA brouniana (see L. aurea). f (Carpolonchaea) citricola, on oranges in Philippines and Australia, 205, 207. a continentalis, sp. ., in Natal, 200, 202. > (Carpolonchaea) excisa, wide distribution of, 201, 206, 209, 210. > (Carpolonchaea) ficiperda, in Philippines, 205, 207, 210. as impressifrons, sp. in Gold Coast, 201, 209. ” Soe gee ig laevis, sp. .,in Zanzibar and Eri- my 201, 203. Pe lambiana, in Seychelles, 200, 202. > (Carpolonchaea) lucens, in Java, Sumatra and Philip- pines, 206, 209, 210. wis metatarsata (see D. aurea). ” minuta, in Java and Philip- pines, 205, 207. ¥ mochii, from Eritrea, 200, 201. “ obscuripennis, probably a synonym of L. atratula, 207. * (Carpolonchaea) ophyroides, sp. n., in Tropical Africa 201, 203. oo, ___——< = —— —— ——_ —EE— 3 a a Se “2 — 7 6 tie - _ = GENERAL INDEX. 425 Lonchaea (Carpolonchaea) pallicarpa, sp. n., in Philippines, 205, 207, 210. a (Dasyops) phaeolepis, sp. 0., in Natal, 200, 201. (Carpolonchaea) plumosis- sima, wide distribution of, 199, 201, 203, 204, 206, 209, 210. (Carpolonchaea) setifera, in New Guinea and Philip- pines, 206, 209, 210. = splendida (see L. awrea). Lonchaeidae, classification of new African and Oriental species of, 199- 210 longiareolata, Theobaldia. longicorda, Ceroplastes. longipalpis, Glossina; Ochlerotatus ; Phlebotomus. 99 99 longipilosa, Lachnosterna. longirostris, Corizoneura. Loranthus, bird-lime made from, used in catching tsetses, 361. dlounsburyi, Oryptinglisia. ducens, Lonchaea (Carpolonchaea). ludlowi, Anopheles (Myzomyia). luridus, Tabanus. dutescens, Ochlerotatus. luteocephala, Stegomyia. LIutzia, 136. Lygodium subalatum, in S.E. Africa, 322. -maculatum, Lecanium (see L. hes- peridum) ; Simulium. maculicornis, Tabanus. -maculipennis, Anopheles. Macrosiphum cissi, sp. n., on Cissus in Transvaal, 65. ys dahliafolii, on Carduus in Uganda, 72. ; sonchi, on Carduus in Uganda, 72. a viticola, on Vitis, 66. “Madagascar, mos uitos of, 113-138. ‘madagascariensis, Pseudoheptaphlebo- myia (see Culex quasigelidus). Malata, food-plant of thrips in Sudan, 98. “Malaya, Anopheles in, 77-92; Lonchaea spp. in 204, 207, 209. Malapterurus electricus (Electric Cat- fish) feeding on mosquito larvae in | Lagos, 266. Mangifera indica (Mango), mosquito larvae breeding in, in Lagos, 257 ; Coccids on, in Mauritius, 174. mangiferae, Lecanium. Mansonia titillans, breeding among water lettuce, 73, 75. Mansonioides africanus, breeding places of, in West Africa, 14, 258; 266,268, 269, 270. 5 uniformis, in Lagos, 269 ; in Madagascar, 135. mariae, Ochlerotatus. Markhamia, puparia of Glossina au tent found under logs of, 366. ef lanata, in relation to dis- tribution of Glossina in S.E. Africa, 348, 382. marlatti, Phoenicococcus. maroccanus, Dociostaurus. marshali, Anopheles ; Ochlerotatus. Martinique, Lachnosterna denticulata in’ 192. mashonaensis, Uranotaenia. mauritianus, Anopheles ; Aspidiotus. Mauritius, mosquitos of, 113-138; bionomics and control of Lachnos- terna (Phytalus) smithi in, 159-169 : plant pest legislation in, 175-177; insects accidentally introduced into, 171-177 ; new weevil from, 276. melanopyga, Phara (Pangonia). Mellera lobulata, in 8.E. Africa, 322. Membranaria, gen. n., 2, 41. “ pretoriae, Sp. D., on grass in Transvaal, 41. meridionalis, Pericoma. Mermis, parasite of Simuliwm, 215. mesambrinoides, Pangonia. mesembryanthemi, Pulvinaria. Mesopotamia, bionomics of Ocnerogyia amanda on figs in, 181-186; date pests in, 287-303; Phlebotomus sergenti var. in, 310, mespili, Myzus. mespiliella, Myzus. Mespilus, new Aphid on, 155. metallica, Stegomyia. metallicus, Taeniorhynchus. Metarrhizium anisopliae, introduction of, into Mauritius from Porto Rico, 166. 426 metatarsata, Lonchaea (see L. aurea). Methylamine, influence of, in formalde- hyde, 48. Mexico, Lachnosterna spp. in, 189, 192. Micraédes ‘ . , F » \ Pe jhe ’ ay ® é : ’ nl ( é ; é ; +. a) ou ’ \ a we * » 5 ¥ af J + Ae 1¢-* ‘ut, hy 4 i “, f . y . . . ‘ 4 a PS *% ‘ * ’ ~ . 4 4 » 4 9 ‘ i a i ‘ ¥ ' * ' , # ’ | : + ’ . 7 t ; - . ( ‘ . \ é “e , } A ’ ‘ { fl Te is i ‘ P a 7 ’ y vt Pat , }! ‘ ok Uhh.) | * ¢ ‘ , ve wee ‘ a 29 Ay | . * é AS Sark 4 , s pe Fy | inf’: a o\ A "LF " <> —_— ry J a & 4 is Se ok in |) Bk “pat . iw i / STS Peter na Peg aan